Perioperative medicine is an evolving field with a rapidly growing body of literature, particularly in cardiology.

In this update, we review 6 articles to answer questions related to preoperative cardiac risk assessment, perioperative medication management, and postoperative cardiac complications. We surveyed perioperative literature from February 2018 through January 2019 and chose the final articles by consensus, based on relevance to clinicians who provide preoperative evaluations and postoperative care to surgical patients.

These summaries are derived from “Updates in Perioperative Medicine” presented at the 14th Annual Perioperative Medicine Summit (Orlando, FL, February 13–16, 2019) and the 2019 Society of Hospital Medicine Annual Meeting (National Harbor, MD, March 24–27, 2019).

PREOPERATIVE CARDIAC EVALUATION

How well do measures of functional capacity predict perioperative complications and mortality in noncardiac surgical patients?

Functional capacity is commonly assessed in preoperative evaluations to estimate patients’ risks of perioperative complications and death. The American College of Cardiology/American Heart Association¹ and the European Society of Cardiology² guidelines both include estimation of cardiopulmonary fitness as a step in preoperative assessment before major noncardiac surgery.

“Subjective assessment” is one way to estimate functional capacity. Simply put, clinicians try to form a rough idea about the fitness of patients by asking questions about routine activities such as walking or climbing stairs. Although commonly used, subjective assessment of functional capacity lacks strong evidence that it predicts adverse perioperative events.

Cardiopulmonary exercise testing is a third option. It measures peak oxygen consumption and anaerobic threshold during exercise. It is probably the best objective measurement of functional capacity, but not necessarily for predicting postoperative cardiac complications, and it is performed relatively infrequently.

[Wijeysundera DN, Pearse RM, Sulman MA, et al. Assessment of functional capacity before major non-cardiac surgery: an international, prospective cohort study. Lancet 2018; 391(10140):2631–2640. doi:10.1016/S0140-6736(18)31131-0]

In a multicenter, prospective cohort study, Wijeysundera et al⁴ compared subjective functional capacity assessment, the Duke Activity Status Index, cardiopulmonary exercise testing, and the preoperative N-terminal pro-B-type natriuretic peptide (NT-proBNP) level in their ability to predict complications and death in 1,401 noncardiac surgery patients older than 40 with at least 1 cardiovascular risk factor. After surgery, patients had daily electrocardiograms and troponin measurements until postoperative day 3 or discharge.

The primary outcome was the 30-day incidence of death or myocardial infarction (MI). Additional outcomes included the 30-day incidence of death or myocardial injury after noncardiac surgery (MINS), the 1-year mortality rate, and moderate to severe in-hospital perioperative complications.

Findings. Two percent of patients died or had an MI within 30 days of surgery.⁴

Subjective assessment had only a 19.2% sensitivity (95% confidence interval [CI] 14.2–25) but a 94.7% specificity (95% CI 93.2–95.9) for predicting inability to attain 4 metabolic equivalents during exercise.⁴

A lower Duke Activity Status Index predicted the primary outcome of death or MI within 30 days (adjusted odds ratio [OR] 0.96, 95% CI 0.83–0.99, P = .03), and it was the only measure that did so. Additionally, the Duke index and NT-proBNP level predicted the risk of death or MINS within 30 days.⁴

Only elevated NT-proBNP was associated with death at 1 year.⁴

On exercise testing, low peak oxygen consumption was significantly associated with perioperative complications.

Limitations. The number of primary outcome events (death and MI) was low, potentially affecting the statistical power of the study.

Conclusions. Subjective assessment of functional capacity misclassifies too many patients as being at low risk of perioperative complications and should not be used for preoperative risk stratification. Other tools, such as the Duke Activity Status Index and NT-proBNP levels, are better predictors of adverse perioperative cardiovascular outcomes and should be considered for use in preoperative cardiac risk assessment.

Although the Duke Activity Status Index is a better predictor of adverse outcomes than subjective functional capacity assessment, a specific perioperative threshold for risk classification has not been established. Its correlate for metabolic equivalents should be considered for use in clinical practice at this point.

Is perioperative aspirin beneficial in patients undergoing vascular surgery?

The Perioperative Ischemic Evaluation 2 (POISE-2) trial,⁵ a 2-by-2 factorial randomized controlled trial in which patients received perioperative aspirin, clonidine, both, or neither, demonstrated that perioperative aspirin did not reduce cardiovascular events and increased major bleeding. Patients with recently placed coronary stents and those undergoing carotid endarterectomy were excluded because aspirin is known to have a beneficial effect in these patients.

A subsequent substudy⁶ found perioperative aspirin to be beneficial in patients with coronary stents placed more than a year before noncardiac surgery. Whether perioperative aspirin is beneficial in other subgroups was unknown.

[Biccard BM, Sigamani A, Chan MTV, et al. Effect of aspirin in vascular surgery in patients from a randomized clinical trial (POISE-2). Br J Surg 2018; 105(12):1591–1597. doi:10.1002/bjs.10925]

Biccard et al⁷ investigated the effect of perioperative aspirin in the subgroup of patients from the POISE-2 trial who underwent vascular surgery. The primary outcome was death or MI within 30 days. Secondary outcomes in this substudy included vascular occlusive complications (amputation and peripheral arterial thrombosis) and major or life-threatening bleeding.

Limitations. There were few adverse events, and this substudy was underpowered for the primary and secondary outcomes.

Conclusion. Starting or continuing aspirin did not improve outcomes, and withdrawing it did not increase cardiovascular or occlusive complications.

Do ACE inhibitors affect risk in noncardiac nonvascular surgery?

Angiotensin-converting enzyme (ACE) inhibitors and angiotensin II receptor blockers (ARBs) are some of the most commonly used medications for treating hypertension. But whether patients should continue receiving them on the day of surgery or whether they should be held remains unclear.

Although current recommendations are inconsistent, the most recent American College of Cardiology/American Heart Association¹ perioperative practice guidelines say that continuing ACE inhibitors or ARBs is reasonable perioperatively. This recommendation, however, acknowledges that published evidence is limited. There is general agreement that preoperative exposure to ACE inhibitors and ARBs is associated with intraoperative hypotension, but whether this increases the risk of adverse clinical outcomes remains unclear. Needed was a study to determine the effect on perioperative morbidity and mortality of continuing vs withholding ACE inhibitors and ARBs before surgery.

[Shiffermiller JF, Monson BJ, Vokoun CW, et al. Prospective randomized evaluation of preoperative angiotensin-converting enzyme inhibition (PREOP-ACEI). J Hosp Med 2018; 13(10):661–667. doi:10.12788/jhm.3036]

Shiffermiller et al⁸ performed a randomized controlled trial comparing the effect of 2 preoperative ACE inhibitor management protocols in patients undergoing noncardiac nonvascular surgery. Patients were randomized to either receive or not receive their final preoperative ACE inhibitor dose, whether scheduled on the morning of surgery or the night before.

Exclusion criteria included hypotension or hypertension at their preoperative clinic appointment (defined as systolic blood pressure < 90 or ≥ 160 mm Hg, and diastolic blood pressure < 60 or ≥ 95 mm Hg), moderate to severe heart failure, and end-stage renal disease requiring dialysis. Excluded surgery types were cardiac, vascular, organ transplant, oncologic, and all outpatient procedures. Patients taking ARBs were also excluded.

The primary outcome was intraoperative hypotension defined as any systolic blood pressure less than 80 mm Hg from the time of anesthesia induction until transfer to the postanesthesia care unit. Secondary outcomes were measured until hospital discharge and included postoperative acute kidney injury, postoperative hypotension (systolic pressure < 90 mm Hg) and hypertension (systolic pressure > 180 mm Hg), major cardiac events (composite of acute coronary syndrome, acute heart failure, or new-onset arrhythmia), and death.

Findings. A total of 453 patients were screened for eligibility, and of these, 291 were included for randomization. Their average age was 64, 48% were men, and 87% were white. About 50% underwent general anesthesia, 25% spinal, and 25% regional. Over half of the surgeries were orthopedic, and 20% were spine surgeries.

The primary outcome of intraoperative hypotension occurred significantly less often in patients randomized to ACE inhibitor omission than in the continuation group (55% vs 69%, relative risk [RR] 0.81, 95% CI 0.67–0.97, P = .03). This translates to 1 case of intraoperative hypotension for every 7.5 patients continuing an ACE inhibitor perioperatively (number needed to harm 7.5). Intraoperative hypotension associated with vasopressor administration also occurred significantly less frequently in the ACE inhibitor omission group.

Patients in the ACE inhibitor omission group were also less likely to experience postoperative hypotension, but on the other hand, they were more likely to experience severe postoperative hypertension (defined as any systolic blood pressure > 180 mm Hg). The two groups fared the same in terms of rates of acute kidney injury and major adverse cardiac events (MACE) and hospital length of stay, and no patients died in either group.

Limitations. Several factors limit the generalizability of this single-center study, including the many exclusion criteria, the predominance of orthopedic and spine surgeries, and the low-risk patient population (the average Revised Cardiac Risk Index score was 0, range 0–3). Other limitations include not controlling for the specific ACE inhibitor used and not including the precise timing of the final dose in relation to surgery. Lastly, this study lacked power to measure postoperative outcomes.

Conclusions. Continuing ACE inhibitor treatment before noncardiac nonvascular surgery is associated with a greater frequency and duration of intraoperative hypotension, but it did not increase the incidences of acute kidney injury, MACE, or death nor the hospital length of stay.

[Hollmann C, Fernandes NL, Biccard BM. A systematic review of outcomes associated with withholding or continuing angiotensin-converting enzyme inhibitors and angiotensin receptor blockers before noncardiac surgery. Anesth Analg 2018; 127(3):678–687. doi:10.1213/ANE.0000000000002837]

Hollmann et al⁹ performed a meta-analysis to determine whether it is better to continue or withhold ACE inhibitors and ARBs before surgery. The patients were adults undergoing noncardiac surgery and receiving an ACE inhibitor or ARB, which was either withheld or continued on the morning of surgery.

Primary outcomes were all-cause mortality and MACE, while secondary outcomes included the incidence of acute kidney injury, heart failure, stroke, intraoperative and postoperative hypotension, and length of hospital stay. Randomized controlled trials and observational studies were included, while case reports and case-control studies were excluded.

Findings. This meta-analysis included 5 randomized controlled trials and 4 cohort studies, with a total of 6,022 patients; 1,816 had their ACE inhibitor or ARB withheld before surgery, while 4,206 continued therapy. It found no difference between the 2 groups in the incidence of death or MACE, and there were not enough data to determine a difference in heart failure, stroke, acute kidney injury, or hospital length of stay.

Seven studies, with 5,414 patients, examined intraoperative hypotension. The overall incidence was 30%, but was significantly lower if the ACE inhibitor or ARB was withheld (OR 0.63, 95% CI 0.47–0.85, ^P = .002). Findings were similar in an analysis of only the randomized controlled trials. No difference was observed in postoperative hypotension.

Limitations. There was no standard definition of the morbidity outcomes, including hypotension and MACE. The assessment of MACE included data only for MI and not MINS. The specific duration of hypotension was not reported, and this meta-analysis did not take into account different anesthetic techniques. The duration of follow-up varied widely among studies, ranging from the day of hospital discharge to 30 days after surgery. And the randomized controlled trial performed by Shiffermiller et al⁸ was not included.

Conclusions. While continuing ACE inhibitors or ARBs before noncardiac surgery was associated with intraoperative hypotension, it did not seem to affect other outcomes, including death and MACE. The authors propose that a large randomized controlled trial is needed to determine whether continuing or withholding ACE inhibitor or ARB therapy before surgery is safer.

How should we treat MINS?

MINS is associated with an increased risk of cardiovascular events and death in both the short term and long term. MINS is defined as an elevated postoperative troponin level related to an ischemic etiology. However, whether to routinely measure troponin after surgery is unclear, as most patients do not present with ischemic symptoms, and there is no standard of care for treatment of this entity. Limited observational data suggest that starting or intensifying cardiac medications, particularly aspirin and statins, may be beneficial in terms of reducing 30-day mortality rates in patients with MI or cardiac events at 1 year in vascular surgery patients with MINS.

The Management of Myocardial Injury After Noncardiac Surgery (MANAGE) trial was designed to evaluate the potential of the anticoagulant dabigatran to prevent major vascular complications in patients with MINS.

[Devereaux PJ, Duceppe E, Guyatt G, et al. Dabigatran in patients with myocardial injury after non-cardiac surgery (MANAGE): an international, randomised, placebo-controlled trial. Lancet 2018; 391(10137):2325–2334. doi:10.1016/S0140-6736(18)30832-8]

Devereaux et al¹⁰ randomized patients who were at least 45 years old and had developed MINS within the previous 35 days to receive dabigatran 110 mg orally twice daily or placebo for up to 2 years. Patients not already taking a proton pump inhibitor were also randomized to take either omeprazole 20 mg once daily or placebo.

The primary efficacy outcome initially was major vascular complications, which included vascular mortality, nonfatal MI, nonhemorrhagic stroke, and peripheral arterial thrombosis. However, amputation and symptomatic venous thromboembolism were subsequently added during the study.

The primary safety outcome was a composite of life-threatening, major, and critical organ bleeding. Major bleeding required a decrease in hemoglobin of at least 4 g/dL, transfusion of at least 3 units of red blood cells within a 24-hour period, or a procedure to stop the bleeding.

Findings. The original goal was to recruit 3,200 patients, but due to slow enrollment and loss of funding, the sample was reduced to 1,754 patients (877 in each group). Approximately 45% of each group stopped taking the study drug prematurely.

The primary efficacy outcome occurred in significantly fewer patients receiving dabigatran (97, 11%) than placebo (133, 15%, HR 0.72, 95% CI 0.55–0.93, P = .0115). The incidence of the primary safety outcome was similar in both groups: 3% with dabigatran and 4% with placebo (HR 0.92, 95% CI 0.55–1.53, P = .76). The only individual efficacy outcome meeting statistical significance was a lower rate of nonhemorrhagic stroke in the dabigatran group. Subgroup analyses showed a trend benefiting patients randomized within 5 days of MINS or with a diagnosis of MI, although it was not statistically significant.

Limitations. The efficacy outcomes were expanded to include venous thromboembolism and others not directly related to MINS, raising questions about the conclusions. Further, as defined by the protocol, bleeding had to be fairly severe to be deemed major. The high number of patients who discontinued the study drug is another limitation of this study.

Conclusion. Dabigatran lowered the risk of major vascular complications with no significant increase in major bleeding in patients with MINS.

What is the risk of thromboembolism in postoperative atrial fibrillation, and what are the benefits of anticoagulation?

Although nonvalvular atrial fibrillation is associated with increased risks of ischemic stroke and systemic embolic events in nonsurgical patients, the association of new-onset postoperative atrial fibrillation with long-term thromboembolic events in the noncardiac surgical population is not well established.

[Butt JH, Olesen JB, Havers-Borgersen E, et al. Risk of thromboembolism associated with atrial fibrillation following noncardiac surgery. J Am Coll Cardiol 2018; 72(17):2027–2036. doi:10.1016/j.jacc.2018.07.088]

In this retrospective cohort study using a nationwide registry in Denmark, Butt et al¹¹ assessed the long-term risk of thromboembolic events in noncardiac surgical patients with new postoperative atrial fibrillation. Patients were identified who had no previous history of atrial fibrillation and developed it after noncardiac, nonobstetric surgeries, and were matched in a 1:4 ratio with patients who developed nonvalvular atrial fibrillation during nonsurgical hospitalizations. Matching was based on age, sex, heart failure, hypertension, diabetes, known history of thromboembolic events, ischemic heart disease, and the year patients presented with new atrial fibrillation.

Patients were excluded if they received antiarrhythmic drugs or oral anticoagulants before hospitalization or surgery, had cancer in the year prior, or died in the hospital.

The primary outcome of the study was thromboembolic events—a composite of ischemic stroke, transient cerebral ischemia, and peripheral arterial thrombosis or embolism. Secondary outcomes included rehospitalization for atrial fibrillation and all-cause mortality.

Findings. Overall, 0.4% of patients developed new postoperative atrial fibrillation, of whom 3,380 were matched with 15,320 patients with nonvalvular atrial fibrillation. Over a median follow-up of 3.2 years, the risk of thromboembolic events was similar in both groups (31.7 and 29.9 per 1,000 person-years, HR 0.95, 95% CI 0.85–1.07). The groups did not differ in their CHA₂DS₂-VASc risk scores, HAS-BLED risk scores, or year in which patients were diagnosed.

Anticoagulation lowered the risk of thromboembolic events to a similar extent in both groups compared with no anticoagulation:

• In postoperative atrial fibrillation—HR 0.57, 95% CI 0.40–0.67
• In nonvalvular atrial fibrillation—HR 0.56, 95% CI 0.51–0.62.

Despite the similar reduction in thromboembolic events, only 24.4% of the postoperative atrial fibrillation patients were started on anticoagulation therapy within 30 days of discharge, compared with 41.5% of those with nonvalvular atrial fibrillation.

Limitations. Although this was a large study with excellent follow-up data, it was observational. It may have underestimated the number of patients who developed postoperative atrial fibrillation because episodes that were judged not to be clinically significant may not have been charted. Many patients are not monitored with continuous telemetry postoperatively, which also may have led to underestimation of the number of atrial fibrillation events.

The study also did not examine the number of atrial fibrillation episodes per patient, the heart rhythm at discharge or long-term, or indication for and duration of anticoagulation. There were no data regarding international normalized ratio levels.

Conclusions. Postoperative atrial fibrillation is associated with outcomes similar to those of nonsurgical nonvalvular atrial fibrillation. Anticoagulation decreases the risks of stroke and death. However, substantially fewer patients with postoperative atrial fibrillation receive anticoagulation. Anticoagulation should be considered in these patients, while noting bleeding risk.

Perioperative medicine is an evolving field with a rapidly growing body of literature, particularly in cardiology.

In this update, we review 6 articles to answer questions related to preoperative cardiac risk assessment, perioperative medication management, and postoperative cardiac complications. We surveyed perioperative literature from February 2018 through January 2019 and chose the final articles by consensus, based on relevance to clinicians who provide preoperative evaluations and postoperative care to surgical patients.

These summaries are derived from “Updates in Perioperative Medicine” presented at the 14th Annual Perioperative Medicine Summit (Orlando, FL, February 13–16, 2019) and the 2019 Society of Hospital Medicine Annual Meeting (National Harbor, MD, March 24–27, 2019).

PREOPERATIVE CARDIAC EVALUATION

How well do measures of functional capacity predict perioperative complications and mortality in noncardiac surgical patients?

Functional capacity is commonly assessed in preoperative evaluations to estimate patients’ risks of perioperative complications and death. The American College of Cardiology/American Heart Association¹ and the European Society of Cardiology² guidelines both include estimation of cardiopulmonary fitness as a step in preoperative assessment before major noncardiac surgery.

“Subjective assessment” is one way to estimate functional capacity. Simply put, clinicians try to form a rough idea about the fitness of patients by asking questions about routine activities such as walking or climbing stairs. Although commonly used, subjective assessment of functional capacity lacks strong evidence that it predicts adverse perioperative events.

Cardiopulmonary exercise testing is a third option. It measures peak oxygen consumption and anaerobic threshold during exercise. It is probably the best objective measurement of functional capacity, but not necessarily for predicting postoperative cardiac complications, and it is performed relatively infrequently.

[Wijeysundera DN, Pearse RM, Sulman MA, et al. Assessment of functional capacity before major non-cardiac surgery: an international, prospective cohort study. Lancet 2018; 391(10140):2631–2640. doi:10.1016/S0140-6736(18)31131-0]

In a multicenter, prospective cohort study, Wijeysundera et al⁴ compared subjective functional capacity assessment, the Duke Activity Status Index, cardiopulmonary exercise testing, and the preoperative N-terminal pro-B-type natriuretic peptide (NT-proBNP) level in their ability to predict complications and death in 1,401 noncardiac surgery patients older than 40 with at least 1 cardiovascular risk factor. After surgery, patients had daily electrocardiograms and troponin measurements until postoperative day 3 or discharge.

The primary outcome was the 30-day incidence of death or myocardial infarction (MI). Additional outcomes included the 30-day incidence of death or myocardial injury after noncardiac surgery (MINS), the 1-year mortality rate, and moderate to severe in-hospital perioperative complications.

Findings. Two percent of patients died or had an MI within 30 days of surgery.⁴

Subjective assessment had only a 19.2% sensitivity (95% confidence interval [CI] 14.2–25) but a 94.7% specificity (95% CI 93.2–95.9) for predicting inability to attain 4 metabolic equivalents during exercise.⁴

A lower Duke Activity Status Index predicted the primary outcome of death or MI within 30 days (adjusted odds ratio [OR] 0.96, 95% CI 0.83–0.99, P = .03), and it was the only measure that did so. Additionally, the Duke index and NT-proBNP level predicted the risk of death or MINS within 30 days.⁴

Only elevated NT-proBNP was associated with death at 1 year.⁴

On exercise testing, low peak oxygen consumption was significantly associated with perioperative complications.

Limitations. The number of primary outcome events (death and MI) was low, potentially affecting the statistical power of the study.

Conclusions. Subjective assessment of functional capacity misclassifies too many patients as being at low risk of perioperative complications and should not be used for preoperative risk stratification. Other tools, such as the Duke Activity Status Index and NT-proBNP levels, are better predictors of adverse perioperative cardiovascular outcomes and should be considered for use in preoperative cardiac risk assessment.

Although the Duke Activity Status Index is a better predictor of adverse outcomes than subjective functional capacity assessment, a specific perioperative threshold for risk classification has not been established. Its correlate for metabolic equivalents should be considered for use in clinical practice at this point.

Is perioperative aspirin beneficial in patients undergoing vascular surgery?

The Perioperative Ischemic Evaluation 2 (POISE-2) trial,⁵ a 2-by-2 factorial randomized controlled trial in which patients received perioperative aspirin, clonidine, both, or neither, demonstrated that perioperative aspirin did not reduce cardiovascular events and increased major bleeding. Patients with recently placed coronary stents and those undergoing carotid endarterectomy were excluded because aspirin is known to have a beneficial effect in these patients.

A subsequent substudy⁶ found perioperative aspirin to be beneficial in patients with coronary stents placed more than a year before noncardiac surgery. Whether perioperative aspirin is beneficial in other subgroups was unknown.

[Biccard BM, Sigamani A, Chan MTV, et al. Effect of aspirin in vascular surgery in patients from a randomized clinical trial (POISE-2). Br J Surg 2018; 105(12):1591–1597. doi:10.1002/bjs.10925]

Biccard et al⁷ investigated the effect of perioperative aspirin in the subgroup of patients from the POISE-2 trial who underwent vascular surgery. The primary outcome was death or MI within 30 days. Secondary outcomes in this substudy included vascular occlusive complications (amputation and peripheral arterial thrombosis) and major or life-threatening bleeding.

Limitations. There were few adverse events, and this substudy was underpowered for the primary and secondary outcomes.

Conclusion. Starting or continuing aspirin did not improve outcomes, and withdrawing it did not increase cardiovascular or occlusive complications.

Do ACE inhibitors affect risk in noncardiac nonvascular surgery?

Angiotensin-converting enzyme (ACE) inhibitors and angiotensin II receptor blockers (ARBs) are some of the most commonly used medications for treating hypertension. But whether patients should continue receiving them on the day of surgery or whether they should be held remains unclear.

Although current recommendations are inconsistent, the most recent American College of Cardiology/American Heart Association¹ perioperative practice guidelines say that continuing ACE inhibitors or ARBs is reasonable perioperatively. This recommendation, however, acknowledges that published evidence is limited. There is general agreement that preoperative exposure to ACE inhibitors and ARBs is associated with intraoperative hypotension, but whether this increases the risk of adverse clinical outcomes remains unclear. Needed was a study to determine the effect on perioperative morbidity and mortality of continuing vs withholding ACE inhibitors and ARBs before surgery.

[Shiffermiller JF, Monson BJ, Vokoun CW, et al. Prospective randomized evaluation of preoperative angiotensin-converting enzyme inhibition (PREOP-ACEI). J Hosp Med 2018; 13(10):661–667. doi:10.12788/jhm.3036]

Shiffermiller et al⁸ performed a randomized controlled trial comparing the effect of 2 preoperative ACE inhibitor management protocols in patients undergoing noncardiac nonvascular surgery. Patients were randomized to either receive or not receive their final preoperative ACE inhibitor dose, whether scheduled on the morning of surgery or the night before.

Exclusion criteria included hypotension or hypertension at their preoperative clinic appointment (defined as systolic blood pressure < 90 or ≥ 160 mm Hg, and diastolic blood pressure < 60 or ≥ 95 mm Hg), moderate to severe heart failure, and end-stage renal disease requiring dialysis. Excluded surgery types were cardiac, vascular, organ transplant, oncologic, and all outpatient procedures. Patients taking ARBs were also excluded.

The primary outcome was intraoperative hypotension defined as any systolic blood pressure less than 80 mm Hg from the time of anesthesia induction until transfer to the postanesthesia care unit. Secondary outcomes were measured until hospital discharge and included postoperative acute kidney injury, postoperative hypotension (systolic pressure < 90 mm Hg) and hypertension (systolic pressure > 180 mm Hg), major cardiac events (composite of acute coronary syndrome, acute heart failure, or new-onset arrhythmia), and death.

Findings. A total of 453 patients were screened for eligibility, and of these, 291 were included for randomization. Their average age was 64, 48% were men, and 87% were white. About 50% underwent general anesthesia, 25% spinal, and 25% regional. Over half of the surgeries were orthopedic, and 20% were spine surgeries.

The primary outcome of intraoperative hypotension occurred significantly less often in patients randomized to ACE inhibitor omission than in the continuation group (55% vs 69%, relative risk [RR] 0.81, 95% CI 0.67–0.97, P = .03). This translates to 1 case of intraoperative hypotension for every 7.5 patients continuing an ACE inhibitor perioperatively (number needed to harm 7.5). Intraoperative hypotension associated with vasopressor administration also occurred significantly less frequently in the ACE inhibitor omission group.

Patients in the ACE inhibitor omission group were also less likely to experience postoperative hypotension, but on the other hand, they were more likely to experience severe postoperative hypertension (defined as any systolic blood pressure > 180 mm Hg). The two groups fared the same in terms of rates of acute kidney injury and major adverse cardiac events (MACE) and hospital length of stay, and no patients died in either group.

Limitations. Several factors limit the generalizability of this single-center study, including the many exclusion criteria, the predominance of orthopedic and spine surgeries, and the low-risk patient population (the average Revised Cardiac Risk Index score was 0, range 0–3). Other limitations include not controlling for the specific ACE inhibitor used and not including the precise timing of the final dose in relation to surgery. Lastly, this study lacked power to measure postoperative outcomes.

Conclusions. Continuing ACE inhibitor treatment before noncardiac nonvascular surgery is associated with a greater frequency and duration of intraoperative hypotension, but it did not increase the incidences of acute kidney injury, MACE, or death nor the hospital length of stay.

[Hollmann C, Fernandes NL, Biccard BM. A systematic review of outcomes associated with withholding or continuing angiotensin-converting enzyme inhibitors and angiotensin receptor blockers before noncardiac surgery. Anesth Analg 2018; 127(3):678–687. doi:10.1213/ANE.0000000000002837]

Hollmann et al⁹ performed a meta-analysis to determine whether it is better to continue or withhold ACE inhibitors and ARBs before surgery. The patients were adults undergoing noncardiac surgery and receiving an ACE inhibitor or ARB, which was either withheld or continued on the morning of surgery.

Primary outcomes were all-cause mortality and MACE, while secondary outcomes included the incidence of acute kidney injury, heart failure, stroke, intraoperative and postoperative hypotension, and length of hospital stay. Randomized controlled trials and observational studies were included, while case reports and case-control studies were excluded.

Findings. This meta-analysis included 5 randomized controlled trials and 4 cohort studies, with a total of 6,022 patients; 1,816 had their ACE inhibitor or ARB withheld before surgery, while 4,206 continued therapy. It found no difference between the 2 groups in the incidence of death or MACE, and there were not enough data to determine a difference in heart failure, stroke, acute kidney injury, or hospital length of stay.

Seven studies, with 5,414 patients, examined intraoperative hypotension. The overall incidence was 30%, but was significantly lower if the ACE inhibitor or ARB was withheld (OR 0.63, 95% CI 0.47–0.85, ^P = .002). Findings were similar in an analysis of only the randomized controlled trials. No difference was observed in postoperative hypotension.

Limitations. There was no standard definition of the morbidity outcomes, including hypotension and MACE. The assessment of MACE included data only for MI and not MINS. The specific duration of hypotension was not reported, and this meta-analysis did not take into account different anesthetic techniques. The duration of follow-up varied widely among studies, ranging from the day of hospital discharge to 30 days after surgery. And the randomized controlled trial performed by Shiffermiller et al⁸ was not included.

Conclusions. While continuing ACE inhibitors or ARBs before noncardiac surgery was associated with intraoperative hypotension, it did not seem to affect other outcomes, including death and MACE. The authors propose that a large randomized controlled trial is needed to determine whether continuing or withholding ACE inhibitor or ARB therapy before surgery is safer.

How should we treat MINS?

MINS is associated with an increased risk of cardiovascular events and death in both the short term and long term. MINS is defined as an elevated postoperative troponin level related to an ischemic etiology. However, whether to routinely measure troponin after surgery is unclear, as most patients do not present with ischemic symptoms, and there is no standard of care for treatment of this entity. Limited observational data suggest that starting or intensifying cardiac medications, particularly aspirin and statins, may be beneficial in terms of reducing 30-day mortality rates in patients with MI or cardiac events at 1 year in vascular surgery patients with MINS.

The Management of Myocardial Injury After Noncardiac Surgery (MANAGE) trial was designed to evaluate the potential of the anticoagulant dabigatran to prevent major vascular complications in patients with MINS.

[Devereaux PJ, Duceppe E, Guyatt G, et al. Dabigatran in patients with myocardial injury after non-cardiac surgery (MANAGE): an international, randomised, placebo-controlled trial. Lancet 2018; 391(10137):2325–2334. doi:10.1016/S0140-6736(18)30832-8]

Devereaux et al¹⁰ randomized patients who were at least 45 years old and had developed MINS within the previous 35 days to receive dabigatran 110 mg orally twice daily or placebo for up to 2 years. Patients not already taking a proton pump inhibitor were also randomized to take either omeprazole 20 mg once daily or placebo.

The primary efficacy outcome initially was major vascular complications, which included vascular mortality, nonfatal MI, nonhemorrhagic stroke, and peripheral arterial thrombosis. However, amputation and symptomatic venous thromboembolism were subsequently added during the study.

The primary safety outcome was a composite of life-threatening, major, and critical organ bleeding. Major bleeding required a decrease in hemoglobin of at least 4 g/dL, transfusion of at least 3 units of red blood cells within a 24-hour period, or a procedure to stop the bleeding.

Findings. The original goal was to recruit 3,200 patients, but due to slow enrollment and loss of funding, the sample was reduced to 1,754 patients (877 in each group). Approximately 45% of each group stopped taking the study drug prematurely.

The primary efficacy outcome occurred in significantly fewer patients receiving dabigatran (97, 11%) than placebo (133, 15%, HR 0.72, 95% CI 0.55–0.93, P = .0115). The incidence of the primary safety outcome was similar in both groups: 3% with dabigatran and 4% with placebo (HR 0.92, 95% CI 0.55–1.53, P = .76). The only individual efficacy outcome meeting statistical significance was a lower rate of nonhemorrhagic stroke in the dabigatran group. Subgroup analyses showed a trend benefiting patients randomized within 5 days of MINS or with a diagnosis of MI, although it was not statistically significant.

Limitations. The efficacy outcomes were expanded to include venous thromboembolism and others not directly related to MINS, raising questions about the conclusions. Further, as defined by the protocol, bleeding had to be fairly severe to be deemed major. The high number of patients who discontinued the study drug is another limitation of this study.

Conclusion. Dabigatran lowered the risk of major vascular complications with no significant increase in major bleeding in patients with MINS.

What is the risk of thromboembolism in postoperative atrial fibrillation, and what are the benefits of anticoagulation?

Although nonvalvular atrial fibrillation is associated with increased risks of ischemic stroke and systemic embolic events in nonsurgical patients, the association of new-onset postoperative atrial fibrillation with long-term thromboembolic events in the noncardiac surgical population is not well established.

[Butt JH, Olesen JB, Havers-Borgersen E, et al. Risk of thromboembolism associated with atrial fibrillation following noncardiac surgery. J Am Coll Cardiol 2018; 72(17):2027–2036. doi:10.1016/j.jacc.2018.07.088]

In this retrospective cohort study using a nationwide registry in Denmark, Butt et al¹¹ assessed the long-term risk of thromboembolic events in noncardiac surgical patients with new postoperative atrial fibrillation. Patients were identified who had no previous history of atrial fibrillation and developed it after noncardiac, nonobstetric surgeries, and were matched in a 1:4 ratio with patients who developed nonvalvular atrial fibrillation during nonsurgical hospitalizations. Matching was based on age, sex, heart failure, hypertension, diabetes, known history of thromboembolic events, ischemic heart disease, and the year patients presented with new atrial fibrillation.

Patients were excluded if they received antiarrhythmic drugs or oral anticoagulants before hospitalization or surgery, had cancer in the year prior, or died in the hospital.

The primary outcome of the study was thromboembolic events—a composite of ischemic stroke, transient cerebral ischemia, and peripheral arterial thrombosis or embolism. Secondary outcomes included rehospitalization for atrial fibrillation and all-cause mortality.

Findings. Overall, 0.4% of patients developed new postoperative atrial fibrillation, of whom 3,380 were matched with 15,320 patients with nonvalvular atrial fibrillation. Over a median follow-up of 3.2 years, the risk of thromboembolic events was similar in both groups (31.7 and 29.9 per 1,000 person-years, HR 0.95, 95% CI 0.85–1.07). The groups did not differ in their CHA₂DS₂-VASc risk scores, HAS-BLED risk scores, or year in which patients were diagnosed.

Anticoagulation lowered the risk of thromboembolic events to a similar extent in both groups compared with no anticoagulation:

• In postoperative atrial fibrillation—HR 0.57, 95% CI 0.40–0.67
• In nonvalvular atrial fibrillation—HR 0.56, 95% CI 0.51–0.62.

Despite the similar reduction in thromboembolic events, only 24.4% of the postoperative atrial fibrillation patients were started on anticoagulation therapy within 30 days of discharge, compared with 41.5% of those with nonvalvular atrial fibrillation.

Limitations. Although this was a large study with excellent follow-up data, it was observational. It may have underestimated the number of patients who developed postoperative atrial fibrillation because episodes that were judged not to be clinically significant may not have been charted. Many patients are not monitored with continuous telemetry postoperatively, which also may have led to underestimation of the number of atrial fibrillation events.

The study also did not examine the number of atrial fibrillation episodes per patient, the heart rhythm at discharge or long-term, or indication for and duration of anticoagulation. There were no data regarding international normalized ratio levels.

Conclusions. Postoperative atrial fibrillation is associated with outcomes similar to those of nonsurgical nonvalvular atrial fibrillation. Anticoagulation decreases the risks of stroke and death. However, substantially fewer patients with postoperative atrial fibrillation receive anticoagulation. Anticoagulation should be considered in these patients, while noting bleeding risk.

Perioperative medicine is an evolving field with a rapidly growing body of literature, particularly in cardiology.

In this update, we review 6 articles to answer questions related to preoperative cardiac risk assessment, perioperative medication management, and postoperative cardiac complications. We surveyed perioperative literature from February 2018 through January 2019 and chose the final articles by consensus, based on relevance to clinicians who provide preoperative evaluations and postoperative care to surgical patients.

These summaries are derived from “Updates in Perioperative Medicine” presented at the 14th Annual Perioperative Medicine Summit (Orlando, FL, February 13–16, 2019) and the 2019 Society of Hospital Medicine Annual Meeting (National Harbor, MD, March 24–27, 2019).

PREOPERATIVE CARDIAC EVALUATION

How well do measures of functional capacity predict perioperative complications and mortality in noncardiac surgical patients?

Functional capacity is commonly assessed in preoperative evaluations to estimate patients’ risks of perioperative complications and death. The American College of Cardiology/American Heart Association¹ and the European Society of Cardiology² guidelines both include estimation of cardiopulmonary fitness as a step in preoperative assessment before major noncardiac surgery.

“Subjective assessment” is one way to estimate functional capacity. Simply put, clinicians try to form a rough idea about the fitness of patients by asking questions about routine activities such as walking or climbing stairs. Although commonly used, subjective assessment of functional capacity lacks strong evidence that it predicts adverse perioperative events.

Cardiopulmonary exercise testing is a third option. It measures peak oxygen consumption and anaerobic threshold during exercise. It is probably the best objective measurement of functional capacity, but not necessarily for predicting postoperative cardiac complications, and it is performed relatively infrequently.

[Wijeysundera DN, Pearse RM, Sulman MA, et al. Assessment of functional capacity before major non-cardiac surgery: an international, prospective cohort study. Lancet 2018; 391(10140):2631–2640. doi:10.1016/S0140-6736(18)31131-0]

In a multicenter, prospective cohort study, Wijeysundera et al⁴ compared subjective functional capacity assessment, the Duke Activity Status Index, cardiopulmonary exercise testing, and the preoperative N-terminal pro-B-type natriuretic peptide (NT-proBNP) level in their ability to predict complications and death in 1,401 noncardiac surgery patients older than 40 with at least 1 cardiovascular risk factor. After surgery, patients had daily electrocardiograms and troponin measurements until postoperative day 3 or discharge.

The primary outcome was the 30-day incidence of death or myocardial infarction (MI). Additional outcomes included the 30-day incidence of death or myocardial injury after noncardiac surgery (MINS), the 1-year mortality rate, and moderate to severe in-hospital perioperative complications.

Findings. Two percent of patients died or had an MI within 30 days of surgery.⁴

Subjective assessment had only a 19.2% sensitivity (95% confidence interval [CI] 14.2–25) but a 94.7% specificity (95% CI 93.2–95.9) for predicting inability to attain 4 metabolic equivalents during exercise.⁴

A lower Duke Activity Status Index predicted the primary outcome of death or MI within 30 days (adjusted odds ratio [OR] 0.96, 95% CI 0.83–0.99, P = .03), and it was the only measure that did so. Additionally, the Duke index and NT-proBNP level predicted the risk of death or MINS within 30 days.⁴

Only elevated NT-proBNP was associated with death at 1 year.⁴

On exercise testing, low peak oxygen consumption was significantly associated with perioperative complications.

Limitations. The number of primary outcome events (death and MI) was low, potentially affecting the statistical power of the study.

Conclusions. Subjective assessment of functional capacity misclassifies too many patients as being at low risk of perioperative complications and should not be used for preoperative risk stratification. Other tools, such as the Duke Activity Status Index and NT-proBNP levels, are better predictors of adverse perioperative cardiovascular outcomes and should be considered for use in preoperative cardiac risk assessment.

Although the Duke Activity Status Index is a better predictor of adverse outcomes than subjective functional capacity assessment, a specific perioperative threshold for risk classification has not been established. Its correlate for metabolic equivalents should be considered for use in clinical practice at this point.

Is perioperative aspirin beneficial in patients undergoing vascular surgery?

The Perioperative Ischemic Evaluation 2 (POISE-2) trial,⁵ a 2-by-2 factorial randomized controlled trial in which patients received perioperative aspirin, clonidine, both, or neither, demonstrated that perioperative aspirin did not reduce cardiovascular events and increased major bleeding. Patients with recently placed coronary stents and those undergoing carotid endarterectomy were excluded because aspirin is known to have a beneficial effect in these patients.

A subsequent substudy⁶ found perioperative aspirin to be beneficial in patients with coronary stents placed more than a year before noncardiac surgery. Whether perioperative aspirin is beneficial in other subgroups was unknown.

[Biccard BM, Sigamani A, Chan MTV, et al. Effect of aspirin in vascular surgery in patients from a randomized clinical trial (POISE-2). Br J Surg 2018; 105(12):1591–1597. doi:10.1002/bjs.10925]

Biccard et al⁷ investigated the effect of perioperative aspirin in the subgroup of patients from the POISE-2 trial who underwent vascular surgery. The primary outcome was death or MI within 30 days. Secondary outcomes in this substudy included vascular occlusive complications (amputation and peripheral arterial thrombosis) and major or life-threatening bleeding.

Limitations. There were few adverse events, and this substudy was underpowered for the primary and secondary outcomes.

Conclusion. Starting or continuing aspirin did not improve outcomes, and withdrawing it did not increase cardiovascular or occlusive complications.

Do ACE inhibitors affect risk in noncardiac nonvascular surgery?

Angiotensin-converting enzyme (ACE) inhibitors and angiotensin II receptor blockers (ARBs) are some of the most commonly used medications for treating hypertension. But whether patients should continue receiving them on the day of surgery or whether they should be held remains unclear.

Although current recommendations are inconsistent, the most recent American College of Cardiology/American Heart Association¹ perioperative practice guidelines say that continuing ACE inhibitors or ARBs is reasonable perioperatively. This recommendation, however, acknowledges that published evidence is limited. There is general agreement that preoperative exposure to ACE inhibitors and ARBs is associated with intraoperative hypotension, but whether this increases the risk of adverse clinical outcomes remains unclear. Needed was a study to determine the effect on perioperative morbidity and mortality of continuing vs withholding ACE inhibitors and ARBs before surgery.

[Shiffermiller JF, Monson BJ, Vokoun CW, et al. Prospective randomized evaluation of preoperative angiotensin-converting enzyme inhibition (PREOP-ACEI). J Hosp Med 2018; 13(10):661–667. doi:10.12788/jhm.3036]

Shiffermiller et al⁸ performed a randomized controlled trial comparing the effect of 2 preoperative ACE inhibitor management protocols in patients undergoing noncardiac nonvascular surgery. Patients were randomized to either receive or not receive their final preoperative ACE inhibitor dose, whether scheduled on the morning of surgery or the night before.

Exclusion criteria included hypotension or hypertension at their preoperative clinic appointment (defined as systolic blood pressure < 90 or ≥ 160 mm Hg, and diastolic blood pressure < 60 or ≥ 95 mm Hg), moderate to severe heart failure, and end-stage renal disease requiring dialysis. Excluded surgery types were cardiac, vascular, organ transplant, oncologic, and all outpatient procedures. Patients taking ARBs were also excluded.

The primary outcome was intraoperative hypotension defined as any systolic blood pressure less than 80 mm Hg from the time of anesthesia induction until transfer to the postanesthesia care unit. Secondary outcomes were measured until hospital discharge and included postoperative acute kidney injury, postoperative hypotension (systolic pressure < 90 mm Hg) and hypertension (systolic pressure > 180 mm Hg), major cardiac events (composite of acute coronary syndrome, acute heart failure, or new-onset arrhythmia), and death.

Findings. A total of 453 patients were screened for eligibility, and of these, 291 were included for randomization. Their average age was 64, 48% were men, and 87% were white. About 50% underwent general anesthesia, 25% spinal, and 25% regional. Over half of the surgeries were orthopedic, and 20% were spine surgeries.

The primary outcome of intraoperative hypotension occurred significantly less often in patients randomized to ACE inhibitor omission than in the continuation group (55% vs 69%, relative risk [RR] 0.81, 95% CI 0.67–0.97, P = .03). This translates to 1 case of intraoperative hypotension for every 7.5 patients continuing an ACE inhibitor perioperatively (number needed to harm 7.5). Intraoperative hypotension associated with vasopressor administration also occurred significantly less frequently in the ACE inhibitor omission group.

Patients in the ACE inhibitor omission group were also less likely to experience postoperative hypotension, but on the other hand, they were more likely to experience severe postoperative hypertension (defined as any systolic blood pressure > 180 mm Hg). The two groups fared the same in terms of rates of acute kidney injury and major adverse cardiac events (MACE) and hospital length of stay, and no patients died in either group.

Limitations. Several factors limit the generalizability of this single-center study, including the many exclusion criteria, the predominance of orthopedic and spine surgeries, and the low-risk patient population (the average Revised Cardiac Risk Index score was 0, range 0–3). Other limitations include not controlling for the specific ACE inhibitor used and not including the precise timing of the final dose in relation to surgery. Lastly, this study lacked power to measure postoperative outcomes.

Conclusions. Continuing ACE inhibitor treatment before noncardiac nonvascular surgery is associated with a greater frequency and duration of intraoperative hypotension, but it did not increase the incidences of acute kidney injury, MACE, or death nor the hospital length of stay.

[Hollmann C, Fernandes NL, Biccard BM. A systematic review of outcomes associated with withholding or continuing angiotensin-converting enzyme inhibitors and angiotensin receptor blockers before noncardiac surgery. Anesth Analg 2018; 127(3):678–687. doi:10.1213/ANE.0000000000002837]

Hollmann et al⁹ performed a meta-analysis to determine whether it is better to continue or withhold ACE inhibitors and ARBs before surgery. The patients were adults undergoing noncardiac surgery and receiving an ACE inhibitor or ARB, which was either withheld or continued on the morning of surgery.

Primary outcomes were all-cause mortality and MACE, while secondary outcomes included the incidence of acute kidney injury, heart failure, stroke, intraoperative and postoperative hypotension, and length of hospital stay. Randomized controlled trials and observational studies were included, while case reports and case-control studies were excluded.

Findings. This meta-analysis included 5 randomized controlled trials and 4 cohort studies, with a total of 6,022 patients; 1,816 had their ACE inhibitor or ARB withheld before surgery, while 4,206 continued therapy. It found no difference between the 2 groups in the incidence of death or MACE, and there were not enough data to determine a difference in heart failure, stroke, acute kidney injury, or hospital length of stay.

Seven studies, with 5,414 patients, examined intraoperative hypotension. The overall incidence was 30%, but was significantly lower if the ACE inhibitor or ARB was withheld (OR 0.63, 95% CI 0.47–0.85, ^P = .002). Findings were similar in an analysis of only the randomized controlled trials. No difference was observed in postoperative hypotension.

Limitations. There was no standard definition of the morbidity outcomes, including hypotension and MACE. The assessment of MACE included data only for MI and not MINS. The specific duration of hypotension was not reported, and this meta-analysis did not take into account different anesthetic techniques. The duration of follow-up varied widely among studies, ranging from the day of hospital discharge to 30 days after surgery. And the randomized controlled trial performed by Shiffermiller et al⁸ was not included.

Conclusions. While continuing ACE inhibitors or ARBs before noncardiac surgery was associated with intraoperative hypotension, it did not seem to affect other outcomes, including death and MACE. The authors propose that a large randomized controlled trial is needed to determine whether continuing or withholding ACE inhibitor or ARB therapy before surgery is safer.

How should we treat MINS?

MINS is associated with an increased risk of cardiovascular events and death in both the short term and long term. MINS is defined as an elevated postoperative troponin level related to an ischemic etiology. However, whether to routinely measure troponin after surgery is unclear, as most patients do not present with ischemic symptoms, and there is no standard of care for treatment of this entity. Limited observational data suggest that starting or intensifying cardiac medications, particularly aspirin and statins, may be beneficial in terms of reducing 30-day mortality rates in patients with MI or cardiac events at 1 year in vascular surgery patients with MINS.

The Management of Myocardial Injury After Noncardiac Surgery (MANAGE) trial was designed to evaluate the potential of the anticoagulant dabigatran to prevent major vascular complications in patients with MINS.

[Devereaux PJ, Duceppe E, Guyatt G, et al. Dabigatran in patients with myocardial injury after non-cardiac surgery (MANAGE): an international, randomised, placebo-controlled trial. Lancet 2018; 391(10137):2325–2334. doi:10.1016/S0140-6736(18)30832-8]

Devereaux et al¹⁰ randomized patients who were at least 45 years old and had developed MINS within the previous 35 days to receive dabigatran 110 mg orally twice daily or placebo for up to 2 years. Patients not already taking a proton pump inhibitor were also randomized to take either omeprazole 20 mg once daily or placebo.

The primary efficacy outcome initially was major vascular complications, which included vascular mortality, nonfatal MI, nonhemorrhagic stroke, and peripheral arterial thrombosis. However, amputation and symptomatic venous thromboembolism were subsequently added during the study.

The primary safety outcome was a composite of life-threatening, major, and critical organ bleeding. Major bleeding required a decrease in hemoglobin of at least 4 g/dL, transfusion of at least 3 units of red blood cells within a 24-hour period, or a procedure to stop the bleeding.

Findings. The original goal was to recruit 3,200 patients, but due to slow enrollment and loss of funding, the sample was reduced to 1,754 patients (877 in each group). Approximately 45% of each group stopped taking the study drug prematurely.

The primary efficacy outcome occurred in significantly fewer patients receiving dabigatran (97, 11%) than placebo (133, 15%, HR 0.72, 95% CI 0.55–0.93, P = .0115). The incidence of the primary safety outcome was similar in both groups: 3% with dabigatran and 4% with placebo (HR 0.92, 95% CI 0.55–1.53, P = .76). The only individual efficacy outcome meeting statistical significance was a lower rate of nonhemorrhagic stroke in the dabigatran group. Subgroup analyses showed a trend benefiting patients randomized within 5 days of MINS or with a diagnosis of MI, although it was not statistically significant.

Limitations. The efficacy outcomes were expanded to include venous thromboembolism and others not directly related to MINS, raising questions about the conclusions. Further, as defined by the protocol, bleeding had to be fairly severe to be deemed major. The high number of patients who discontinued the study drug is another limitation of this study.

Conclusion. Dabigatran lowered the risk of major vascular complications with no significant increase in major bleeding in patients with MINS.

What is the risk of thromboembolism in postoperative atrial fibrillation, and what are the benefits of anticoagulation?

Although nonvalvular atrial fibrillation is associated with increased risks of ischemic stroke and systemic embolic events in nonsurgical patients, the association of new-onset postoperative atrial fibrillation with long-term thromboembolic events in the noncardiac surgical population is not well established.

[Butt JH, Olesen JB, Havers-Borgersen E, et al. Risk of thromboembolism associated with atrial fibrillation following noncardiac surgery. J Am Coll Cardiol 2018; 72(17):2027–2036. doi:10.1016/j.jacc.2018.07.088]

In this retrospective cohort study using a nationwide registry in Denmark, Butt et al¹¹ assessed the long-term risk of thromboembolic events in noncardiac surgical patients with new postoperative atrial fibrillation. Patients were identified who had no previous history of atrial fibrillation and developed it after noncardiac, nonobstetric surgeries, and were matched in a 1:4 ratio with patients who developed nonvalvular atrial fibrillation during nonsurgical hospitalizations. Matching was based on age, sex, heart failure, hypertension, diabetes, known history of thromboembolic events, ischemic heart disease, and the year patients presented with new atrial fibrillation.

Patients were excluded if they received antiarrhythmic drugs or oral anticoagulants before hospitalization or surgery, had cancer in the year prior, or died in the hospital.

The primary outcome of the study was thromboembolic events—a composite of ischemic stroke, transient cerebral ischemia, and peripheral arterial thrombosis or embolism. Secondary outcomes included rehospitalization for atrial fibrillation and all-cause mortality.

Findings. Overall, 0.4% of patients developed new postoperative atrial fibrillation, of whom 3,380 were matched with 15,320 patients with nonvalvular atrial fibrillation. Over a median follow-up of 3.2 years, the risk of thromboembolic events was similar in both groups (31.7 and 29.9 per 1,000 person-years, HR 0.95, 95% CI 0.85–1.07). The groups did not differ in their CHA₂DS₂-VASc risk scores, HAS-BLED risk scores, or year in which patients were diagnosed.

Anticoagulation lowered the risk of thromboembolic events to a similar extent in both groups compared with no anticoagulation:

• In postoperative atrial fibrillation—HR 0.57, 95% CI 0.40–0.67
• In nonvalvular atrial fibrillation—HR 0.56, 95% CI 0.51–0.62.

Despite the similar reduction in thromboembolic events, only 24.4% of the postoperative atrial fibrillation patients were started on anticoagulation therapy within 30 days of discharge, compared with 41.5% of those with nonvalvular atrial fibrillation.

Limitations. Although this was a large study with excellent follow-up data, it was observational. It may have underestimated the number of patients who developed postoperative atrial fibrillation because episodes that were judged not to be clinically significant may not have been charted. Many patients are not monitored with continuous telemetry postoperatively, which also may have led to underestimation of the number of atrial fibrillation events.

The study also did not examine the number of atrial fibrillation episodes per patient, the heart rhythm at discharge or long-term, or indication for and duration of anticoagulation. There were no data regarding international normalized ratio levels.

Conclusions. Postoperative atrial fibrillation is associated with outcomes similar to those of nonsurgical nonvalvular atrial fibrillation. Anticoagulation decreases the risks of stroke and death. However, substantially fewer patients with postoperative atrial fibrillation receive anticoagulation. Anticoagulation should be considered in these patients, while noting bleeding risk.

Over the last 3 decades, the endoscope has become a highly valued visualization tool in neurosurgery, applicable to a broad range of neurosurgical procedures. Following the pace of technological innovations, the quality of the instrumentation has greatly improved along with the status of endoscopy in the neurosurgical field. The use of the endoscope in interdisciplinary extended transnasal approaches revolutionized skull-base surgery.¹ Transcranial neurosurgery took advantage of the endoscope for inspection, endoscope-assisted, and endoscope-controlled procedures, although the main visualization tool during these interventions remains the operating microscope.

At present, endoscopy has applications in a variety of neurosurgical procedures including transnasal approaches for pituitary and other skull-base tumors, third ventriculostomy, and resection of intraventricular tumors. The range of application is expanding to include extracranial procedures such as peripheral nerve and spine surgery.

From Li KW, Nelson C, Suk I, Jallo GI. Neuroendoscopy: past, present, and future. Neurosurg Focus 2005; 19(6):E1. Figure used with permission.

CURRENT CONCEPTS

Hopf and Perneczky² defined the terminology regarding endoscopic procedures and divided them into 3 categories:

Pure endoscopic neurosurgery, ie, procedures performed through working channels under complete endoscopic visualization and with endoscopic instrumentation (Figure 1).³

Endoscope-controlled microsurgery, ie, operations performed with standard microsurgical instruments under endoscopic visualization—the microscope is not used (Figure 2).

Endoscope-assisted neurosurgery, ie, the use of both microscope and endoscope during the same intervention. In endoscopic inspection the endoscope is solely used as an adjunctive tool for visualization and not for surgical manipulations.

Enhanced area and surgical dissection

Technical innovations are probably the major reason for the growing role of endoscopy in neurosurgery over the last 3 decades.⁴ High-definition imaging, neuronavigation, new instruments, an interdisciplinary approach mostly with ear, nose, and throat (ENT) surgeons, and detailed anatomic studies led to the breakthrough of endoscopic endonasal extended approaches in skull-base surgery.⁵

These endoscopic techniques allow the neurosurgeon to optimize tumor resection, increasing the area of surgical dissection without increasing the size of the surgical approach, thereby limiting perioperative morbidity due to surgical manipulation of eloquent brain structures. Endoscopy offers direct illumination of the operative field, magnification, and the ability to look around corners with angled optics.

However, while angled endoscopic optics provide various visual perspectives, the surgical issue is not only to see but also to work on and around remote structures. Microsurgical endoscope-assisted manipulations require optimal working angles that are guaranteed only by a sufficiently large craniotomy. As an example, a dissection study by Chaynes et al⁶ highlights that a craniotomy that is too narrow often hinders a sufficient exploration of the entire cerebellopontine angle. Most neuro­surgeons are familiar with the operating microscope. The microscopic field of inspection is 3-dimensional (3D) and of high quality. However, the light stream is straight and thus limited in the narrow and angled corridor of the cerebellopontine angle or in the perimesencephalic cisterns. In these situations, the angled optic of the endoscope offers the advantage of being able to look around the corner with the appropriate amount of direct illumination.⁷

Minimally invasive endoscopic approaches are also being used in peripheral nerve surgery, especially carpal tunnel decompression. The first carpal tunnel release treated endoscopically was performed by Okutsu et al in the late 1980s.⁸ Since that time, endoscopic carpal tunnel decompression has become very common and is the preferred method for many surgeons, using either single-portal or dual-portal techniques. Although the superiority of endoscopic over conventional minimally invasive microsurgical peripheral nerve surgeries has not been proven, large series of endoscopic carpal tunnel decompressions have reported low complication rates and excellent success rates with high patient satisfaction scores.^8,9

Visualization of the spinal canal

Expanding the use of the endoscope to spine surgery, endoscopic explorations of the interlaminar spaces after having completed open surgical laminectomies have been reported since the early 1980s,¹⁰ while endoscope-assisted interlaminar procedures started in the late 1990s.^11–13 The development of fully endoscopic transforaminal or interlaminar approaches for lumbar stenosis or lumbar disk herniation has been ongoing in the last 2 decades. The rationale for direct endoscopic visualization of the spinal canal is to reduce scarring of the epidural space, which might affect the outcome of possible revision surgeries (recurrent disk herniation), and to reduce injury to the paraspinal muscles, which may reduce postoperative incisional pain and length of hospital stay. Major limiting factors for fully endoscopic spine surgeries such as the narrow working channels (which are limited by the osseous perimeter of the neuroforamina, as well as the pelvis and abdominal structures) and the learning curve for the surgeons are, however, still matters of debate and restrict the use of endoscopy to very carefully selected cases.^14,15

Pediatric craniosynostosis

Recently, the use of the endoscope has extended to treatment of craniosynostosis in pediatric patients, historically treated with large and occasionally staged craniotomic approaches. A meta-analysis of the literature showed statistically significant reductions in blood loss and rates of perioperative complications, reoperation, and transfusion compared with open approaches.¹⁶

Technical limitations

While neurosurgeons increasingly advocate the use of the endoscope in their practice, the development of instruments for endoscopic surgery does not always follow the same pace. There are technical problems with current rigid endoscopes and ergonomic limitations of the endoscope-assisted techniques in trans­cranial neurosurgery. The endoscope itself occupies space in an already limited surgical corridor like the posterior fossa, the parasellar space, or the intraventricular region. The ideal endoscope is thin and sturdy, does not generate heat, and provides high-resolution images. In addition, a self-irrigating feature could minimize the need to remove and reinsert the endoscope for cleaning. Finally, most intracranial surgery is extremely delicate and requires bimanual dissection. The ideal endoscope should also be easily integrated with a holder that allows the surgeon to easily transition between static and dynamic endoscope movements.

Newer flexible fiberscopes with even smaller diameters are likely to be launched on the market in the near future. When working in a surgical corridor less than 10 mm wide, this difference could be substantial.

In addition, surgical instruments specifically designed for endoscopic endonasal procedures are needed for microdissection in these regions, which were previously only visible but not reachable endoscopically. These include tools such as malleable suctions and curettes, rotatable back-biting microscissors, and malleable bipolar instruments (Figure 3).

IMPACT OF NEUROENDOSCOPY IN CURRENT CLINICAL PRACTICE

The introduction of endoscopy in neurosurgery changed many treatment paradigms and had an important impact on morbidity and outcomes. In this section, we discuss the specific indications, contraindications, and expected benefit of endoscopic vs open surgical approaches applied to neurosurgical pathology at the present time.

Skull-base tumors and CSF leaks

The use of the endoscope in skull-base surgery was originally applied to purely midline intrasellar tumors without suprasellar or lateral extension beyond the carotid cave. Ideal cases were intrasellar pituitary microadenomas not responding to medical treatment or Rathke cleft cysts.

These pathologies were traditionally addressed via microscopic craniotomic approaches and later through sublabial or transnasal transsphenoidal approaches. Traditional transsphenoidal approaches were highly invasive for the oral mucosa, causing delayed healing, oral dysesthesia, and, in some cases, loss of the superior dental arch (sublabial) or limited visualization and surgical maneuverability (microscopic endonasal).

The endoscope offered better visualization and surgical freedom, thus allowing higher resection rates to be achieved. Resection of purely intrasellar pathology with preservation of the diaphragma sellae as a barrier to the subarachnoid cysterns and third ventricle guaranteed a lower incidence of cerebrospinal fluid (CSF) leaks.

New endoscope optics with varied angles, together with dedicated long surgical instruments with low steric volume, offered a large variety of new endonasal surgical corridors, so-called expanded endonasal approaches on the sagittal and coronal planes, as discussed in detail by Kassam et al.^17–19 These allowed endoscopic treatment of invasive tumors extending on the coronary plane into the suprasellar region or invading the cavernous sinuses (pituitary macroadenomas, craniopharyngiomas).

Highly specialized centers with expertise in endoscopic skull-base surgery can now also offer pure endoscopic treatment for some selected cases of lesions located far laterally to the cavernous sinus, such as trigeminal schwannomas, or along the sagittal plane like olfactory groove or tuberculum sellae meningiomas and clival lesions (chordomas, chondrosarcomas).

As one might expect, the increase in surgical complexity corresponded to an increase in complication rates. For example, the incidence of CSF leaks varied from 5% for standard midline transsphenoidal approaches to 11% for expanded endonasal approaches.^20,21 The consolidation of the use of the endoscope and the cooperation with ENT surgeons led to the development of surgical strategies to prevent and reduce the incidence of CSF leaks, such as the use of “rescue flaps,” nasoseptal flaps, or temporoparietal fascia flaps.^21–23

The development of such techniques allowed endoscopic endonasal approaches to be used in treatment of other pathologies, such as spontaneous CSF leaks, treated in the past with large transcranial repairs that carried high morbidity rates due to the surgical frontal lobe retraction and injury to the olfactory mucosa.^24,25 Progress in the field of neuroendoscopy therefore led to the creation of specialized endoscopic skull-base surgery centers, including neurosurgery, ENT, ophthalmology, and endocrinology services.

In clinical practice, when evaluating a patient with intracranial skull-base pathology amenable to endoscopic resection, one should consider referring the patient not only to a neurosurgeon, but also to an ENT surgeon for preoperative assessment of the sinonasal cavities. The same concept applies to postsurgical follow-up, which is mostly performed by the ENT physician to assess nasal mucosa healing and nasal hygiene.

The introduction of endoscopic third ventriculostomy created the opportunity to offer a more physiologic treatment in selected patients with obstructive hydrocephalus by creating an internal CSF diversion through the basal cisterns. Two advantages of this procedure are that it does not create dependence on a CSF shunt, and it eliminates the related risks of shunt infection and malfunction. Its drawback is the recurrence rate of hydrocephalus (around 58% at 2 years of follow-up) due to formation of scarring in the perforated Lilie­quist membrane, which may require repeat surgery or conversion to CSF shunting.^26,27

Neuroendoscopic approaches are also used in cases of purely intraventricular pathology such as colloid cyst or choroid plexus papillomas. The concept behind neuroendoscopy is to achieve maximal resection in a minimally invasive way, using the natural cavity of the cerebral ventricles and reducing the need for brain retraction and, in particular, the risk of injury of the fornix (therefore causing memory deficits) of open transventricular approaches and of the corpus callosum necessary in inter­hemispheric approaches. Large tumor size and inability to tolerate a longer surgical procedure can be relative contraindications to a pure endoscopic approach to these lesions.

Degenerative spine disease

In recent years there has been a growing interest in the use of endoscopy for selected cases of degenerative lumbar spondylosis (generally, lateral disk herniation above the L5-S1 level or spinal canal stenosis). This approach has been shown to reduce postoperative incisional pain, scarring of the epidural space affecting the outcome of possible revision surgeries (recurrent disc herniation), and length of hospital stay.^14,15 Information on surgical nuances should be provided when consulting on selected patients with lumbar degenerative disease resistant to conservative treatment.

Carpal tunnel syndrome

Although endoscopic carpal tunnel release is controversial, its supporters report smaller incision size and lower recurrence rates due to better visualization of the entire carpal ligament compared with open surgery, with high patient satisfaction scores.^8,9,28

Craniosynostosis

Increasing data from specialized centers show that early endoscopic suturectomy is an effective treatment option alone or when combined with open surgeries for patients with syndromic and nonsyndromic craniosynostosis. The aesthetic advantage of small incisions (which can also be achieved with some open techniques) is accompanied by significant reductions in blood loss (median 162.4 mL), operative time (median 112.38 minutes), length of stay (median 2.56 days), and rates of perioperative complications (odds ratio 0.58), reoperation (odds ratio 0.37), and transfusion (odds ratio 0.09) compared with open approaches.¹⁶

SURGICAL TRAINING

Today’s patients expect high-quality healthcare, and they approach their surgeons with an enormous amount of information collected through unlimited Web-based access or peer-group blogs. In this respect, the pressure on young surgeons to achieve excellent results is high and growing from the very beginning of their careers.

Residency training programs differ in each country, and surgical standards usually focus on open microscopic procedures rather than newly developed endoscopic techniques. Endoscopic pituitary adenoma surgery, the most frequent neuroendoscopic procedure, is still performed mostly by experienced neurosurgeons, not trainees. Moreover, many training institutions might not offer pediatric neurosurgery care, limiting exposure to endoscopic third ventriculostomy procedures. The European Union of Medical Specialists, responsible for harmonizing and improving the quality of training of medical specialists in Europe, set low neuroendoscopic surgical requirements for trainees to complete their residency programs (minimum of 0 to optimum of 5 total transcranial or transsphenoidal pituitary adenoma resections as first operator, 10 procedures as assistant, and a minimum of 2 to an optimum of 4 endoscopic third ventriculostomies as first operator).²⁹

The need to develop training programs in neuroendoscopy is especially urgent because endoscopic surgery has a steeper learning curve than conventional microneurosurgery. In particular, endoscopy requires a good deal of dexterity and hand-eye coordination, which surgeons consider the main pitfall of neuroendoscopy. For such reasons, many accredited clinical fellowship programs have been developed inside and outside North America that offer intensive training in endoscopic skull-base surgery and pediatric neurosurgery after residency.

Some clinical studies have shown that the complication rate of neuroendoscopy is 15% to 18%.^27,30 In view of this statistic, it is ethically questionable to perform a randomized study to prospectively compare microscopic and endoscopic procedures. Surgeons specialize in one technique or the other, experience their own learning curve, and do not randomly decide which tool to use. Furthermore, every intracranial surgical exploration is unique and somewhat difficult to compare with each other without the risk of bias.

Multivariable rigid endoscopes like the EndoCAMeleon (Karl Storz, Tuttlingen, Germany) or the EndActive (Karl Storz, Tuttlingen, Germany) for cerebellopontine angle surgery represent a starting point to overcome some of the aforementioned limitations.^31,32 They are inserted in the surgical field with a direct 0° angulation view into the operative site beyond neurovascular structures that need to be preserved and that obstruct the microscopic view. Once the final position is reached, the field of view is directed toward the region of interest without moving the endoscope tip.

The EndoCAMeleon is a rigid rod-lens endoscope, steerable in one plane from –10° to +120° by a fine optomechanical mechanism. Anatomic laboratory testing found it to be superior in terms of usability and visualization compared with rigid fixed-angle endoscopes.^₃₁ The first clinical experiences have been promising; however, ergonomics and the limited perspective of a single plane of rotation leave room for improvement.

The EndActive endoscope might overcome such limitations.³³ This device is a rigid videoendoscope connected to a laptop (video data) and USB port (control and power supply); thus, it weighs less and can be held in one hand like a microsurgical instrument. The endoscopic imaging system allows the operator to simultaneously see a 160° wide-angle view of the site and an inset of a specific region of interest. The surgeon can hold the device like a microsurgical instrument in one hand and control movements precisely due to its reduced weight and ergonomic shape.

The multiplanar variable-view rigid endoscope has proven to be useful for working on diverse anatomic structures such as intracranial vessels and cranial nerves. The device is effective in narrow working spaces where even small movements can jeopardize the delicate surrounding structures. The multiplanar variable-viewing mechanism in a compact device offers advantages in terms of safety and ergonomics. Improving the usability will probably optimize the applicability of those endoscopic devices in neurosurgery. A major drawback of the current prototype is poor image resolution, which will probably soon be overcome with the ongoing progress in electronic microchip technology.

The addition of laser technology to endoscopic techniques offers a huge potential to neurosurgery but has achieved little acceptance to date. The reasons include concern regarding heat production, uncontrollable and distant penetration, and tissue interaction. Experiences with a 2-micron continuous­- wave laser (RevoLix Jr, LISA Laser Products, Katlenburg-Lindau, Germany) for neuroendoscopic intraventricular procedures proved this laser to be a valuable and useful tool with safe applicability for endoscopic intracranial procedures in patients of all ages.³⁴

Parallel to the launch of video screens for other uses with higher image definition, the image quality on the 2D endoscope cameras has been constantly improving over the last years. At the same time, the introduction of modern 3D endoscopic monitors is promising. However, 3D endoscopes have some disadvantages compared with the 2D endoscopes. First, the smallest 3D endoscopes are 4 mm in diameter, compared with 2.7 mm for 2D endoscopes. Moreover, the field of view with the 3D endoscope is less than half of that with conventional 2D endoscopes.³⁴ When working in and around a region with critical neurovascular structures in close proximity, this loss of field of view can result in an increase in iatrogenic injury from the endoscope. In addition, 3D endoscopes require special glasses, generating a potential obstacle to the seamless integration of visual information from the microscope and endoscope. Finally, some surgeons experience vertigo when looking at the 3D picture through the glasses, which limits its universal applicability.

CONCLUSIONS

Using the endoscope and microscope as complementary and not competing tools allows surgeons to benefit from both technologies at the same time.^35,36 The intraoperative combination of these 2 powerful visualization tools expands the effectiveness of microsurgical procedures and has the potential to further improve surgical results and reduce surgical risks. With endoscope-assisted microsurgery, visualization is often far superior to surgical maneuverability.

Endoscopic neurosurgery will likely be influenced by further innovations in optical physics, electronics, and robotics. Specific implementations in endoscopic systems are likely to pave the way for remarkable progress in minimally invasive surgery, such as robotic surgical technology, further miniaturization of devices, improvements in 3D endoscopy, multiport endoscopy, and new designs for surgical instruments. Future progress in flexible endoscopes and wireless capsule or camera technology may reduce our dependence on rigid rod lens systems. Rigid variable-view endoscopes will bring endoscopes closer to ideal attributes utilizing newer instrumentation that is tailored to specific indications and techniques.^37,38 Extension of the visual field by the feature of a movable optic lens may allow the neurosurgeon to use tailored keyhole approaches to treat pathologies in smaller surgical corridors with less trauma and greater efficacy.

Over the last 3 decades, the endoscope has become a highly valued visualization tool in neurosurgery, applicable to a broad range of neurosurgical procedures. Following the pace of technological innovations, the quality of the instrumentation has greatly improved along with the status of endoscopy in the neurosurgical field. The use of the endoscope in interdisciplinary extended transnasal approaches revolutionized skull-base surgery.¹ Transcranial neurosurgery took advantage of the endoscope for inspection, endoscope-assisted, and endoscope-controlled procedures, although the main visualization tool during these interventions remains the operating microscope.

At present, endoscopy has applications in a variety of neurosurgical procedures including transnasal approaches for pituitary and other skull-base tumors, third ventriculostomy, and resection of intraventricular tumors. The range of application is expanding to include extracranial procedures such as peripheral nerve and spine surgery.

From Li KW, Nelson C, Suk I, Jallo GI. Neuroendoscopy: past, present, and future. Neurosurg Focus 2005; 19(6):E1. Figure used with permission.

CURRENT CONCEPTS

Hopf and Perneczky² defined the terminology regarding endoscopic procedures and divided them into 3 categories:

Pure endoscopic neurosurgery, ie, procedures performed through working channels under complete endoscopic visualization and with endoscopic instrumentation (Figure 1).³

Endoscope-controlled microsurgery, ie, operations performed with standard microsurgical instruments under endoscopic visualization—the microscope is not used (Figure 2).

Endoscope-assisted neurosurgery, ie, the use of both microscope and endoscope during the same intervention. In endoscopic inspection the endoscope is solely used as an adjunctive tool for visualization and not for surgical manipulations.

Enhanced area and surgical dissection

Technical innovations are probably the major reason for the growing role of endoscopy in neurosurgery over the last 3 decades.⁴ High-definition imaging, neuronavigation, new instruments, an interdisciplinary approach mostly with ear, nose, and throat (ENT) surgeons, and detailed anatomic studies led to the breakthrough of endoscopic endonasal extended approaches in skull-base surgery.⁵

These endoscopic techniques allow the neurosurgeon to optimize tumor resection, increasing the area of surgical dissection without increasing the size of the surgical approach, thereby limiting perioperative morbidity due to surgical manipulation of eloquent brain structures. Endoscopy offers direct illumination of the operative field, magnification, and the ability to look around corners with angled optics.

However, while angled endoscopic optics provide various visual perspectives, the surgical issue is not only to see but also to work on and around remote structures. Microsurgical endoscope-assisted manipulations require optimal working angles that are guaranteed only by a sufficiently large craniotomy. As an example, a dissection study by Chaynes et al⁶ highlights that a craniotomy that is too narrow often hinders a sufficient exploration of the entire cerebellopontine angle. Most neuro­surgeons are familiar with the operating microscope. The microscopic field of inspection is 3-dimensional (3D) and of high quality. However, the light stream is straight and thus limited in the narrow and angled corridor of the cerebellopontine angle or in the perimesencephalic cisterns. In these situations, the angled optic of the endoscope offers the advantage of being able to look around the corner with the appropriate amount of direct illumination.⁷

Minimally invasive endoscopic approaches are also being used in peripheral nerve surgery, especially carpal tunnel decompression. The first carpal tunnel release treated endoscopically was performed by Okutsu et al in the late 1980s.⁸ Since that time, endoscopic carpal tunnel decompression has become very common and is the preferred method for many surgeons, using either single-portal or dual-portal techniques. Although the superiority of endoscopic over conventional minimally invasive microsurgical peripheral nerve surgeries has not been proven, large series of endoscopic carpal tunnel decompressions have reported low complication rates and excellent success rates with high patient satisfaction scores.^8,9

Visualization of the spinal canal

Expanding the use of the endoscope to spine surgery, endoscopic explorations of the interlaminar spaces after having completed open surgical laminectomies have been reported since the early 1980s,¹⁰ while endoscope-assisted interlaminar procedures started in the late 1990s.^11–13 The development of fully endoscopic transforaminal or interlaminar approaches for lumbar stenosis or lumbar disk herniation has been ongoing in the last 2 decades. The rationale for direct endoscopic visualization of the spinal canal is to reduce scarring of the epidural space, which might affect the outcome of possible revision surgeries (recurrent disk herniation), and to reduce injury to the paraspinal muscles, which may reduce postoperative incisional pain and length of hospital stay. Major limiting factors for fully endoscopic spine surgeries such as the narrow working channels (which are limited by the osseous perimeter of the neuroforamina, as well as the pelvis and abdominal structures) and the learning curve for the surgeons are, however, still matters of debate and restrict the use of endoscopy to very carefully selected cases.^14,15

Pediatric craniosynostosis

Recently, the use of the endoscope has extended to treatment of craniosynostosis in pediatric patients, historically treated with large and occasionally staged craniotomic approaches. A meta-analysis of the literature showed statistically significant reductions in blood loss and rates of perioperative complications, reoperation, and transfusion compared with open approaches.¹⁶

Technical limitations

While neurosurgeons increasingly advocate the use of the endoscope in their practice, the development of instruments for endoscopic surgery does not always follow the same pace. There are technical problems with current rigid endoscopes and ergonomic limitations of the endoscope-assisted techniques in trans­cranial neurosurgery. The endoscope itself occupies space in an already limited surgical corridor like the posterior fossa, the parasellar space, or the intraventricular region. The ideal endoscope is thin and sturdy, does not generate heat, and provides high-resolution images. In addition, a self-irrigating feature could minimize the need to remove and reinsert the endoscope for cleaning. Finally, most intracranial surgery is extremely delicate and requires bimanual dissection. The ideal endoscope should also be easily integrated with a holder that allows the surgeon to easily transition between static and dynamic endoscope movements.

Newer flexible fiberscopes with even smaller diameters are likely to be launched on the market in the near future. When working in a surgical corridor less than 10 mm wide, this difference could be substantial.

In addition, surgical instruments specifically designed for endoscopic endonasal procedures are needed for microdissection in these regions, which were previously only visible but not reachable endoscopically. These include tools such as malleable suctions and curettes, rotatable back-biting microscissors, and malleable bipolar instruments (Figure 3).

IMPACT OF NEUROENDOSCOPY IN CURRENT CLINICAL PRACTICE

The introduction of endoscopy in neurosurgery changed many treatment paradigms and had an important impact on morbidity and outcomes. In this section, we discuss the specific indications, contraindications, and expected benefit of endoscopic vs open surgical approaches applied to neurosurgical pathology at the present time.

Skull-base tumors and CSF leaks

The use of the endoscope in skull-base surgery was originally applied to purely midline intrasellar tumors without suprasellar or lateral extension beyond the carotid cave. Ideal cases were intrasellar pituitary microadenomas not responding to medical treatment or Rathke cleft cysts.

These pathologies were traditionally addressed via microscopic craniotomic approaches and later through sublabial or transnasal transsphenoidal approaches. Traditional transsphenoidal approaches were highly invasive for the oral mucosa, causing delayed healing, oral dysesthesia, and, in some cases, loss of the superior dental arch (sublabial) or limited visualization and surgical maneuverability (microscopic endonasal).

The endoscope offered better visualization and surgical freedom, thus allowing higher resection rates to be achieved. Resection of purely intrasellar pathology with preservation of the diaphragma sellae as a barrier to the subarachnoid cysterns and third ventricle guaranteed a lower incidence of cerebrospinal fluid (CSF) leaks.

New endoscope optics with varied angles, together with dedicated long surgical instruments with low steric volume, offered a large variety of new endonasal surgical corridors, so-called expanded endonasal approaches on the sagittal and coronal planes, as discussed in detail by Kassam et al.^17–19 These allowed endoscopic treatment of invasive tumors extending on the coronary plane into the suprasellar region or invading the cavernous sinuses (pituitary macroadenomas, craniopharyngiomas).

Highly specialized centers with expertise in endoscopic skull-base surgery can now also offer pure endoscopic treatment for some selected cases of lesions located far laterally to the cavernous sinus, such as trigeminal schwannomas, or along the sagittal plane like olfactory groove or tuberculum sellae meningiomas and clival lesions (chordomas, chondrosarcomas).

As one might expect, the increase in surgical complexity corresponded to an increase in complication rates. For example, the incidence of CSF leaks varied from 5% for standard midline transsphenoidal approaches to 11% for expanded endonasal approaches.^20,21 The consolidation of the use of the endoscope and the cooperation with ENT surgeons led to the development of surgical strategies to prevent and reduce the incidence of CSF leaks, such as the use of “rescue flaps,” nasoseptal flaps, or temporoparietal fascia flaps.^21–23

The development of such techniques allowed endoscopic endonasal approaches to be used in treatment of other pathologies, such as spontaneous CSF leaks, treated in the past with large transcranial repairs that carried high morbidity rates due to the surgical frontal lobe retraction and injury to the olfactory mucosa.^24,25 Progress in the field of neuroendoscopy therefore led to the creation of specialized endoscopic skull-base surgery centers, including neurosurgery, ENT, ophthalmology, and endocrinology services.

In clinical practice, when evaluating a patient with intracranial skull-base pathology amenable to endoscopic resection, one should consider referring the patient not only to a neurosurgeon, but also to an ENT surgeon for preoperative assessment of the sinonasal cavities. The same concept applies to postsurgical follow-up, which is mostly performed by the ENT physician to assess nasal mucosa healing and nasal hygiene.

The introduction of endoscopic third ventriculostomy created the opportunity to offer a more physiologic treatment in selected patients with obstructive hydrocephalus by creating an internal CSF diversion through the basal cisterns. Two advantages of this procedure are that it does not create dependence on a CSF shunt, and it eliminates the related risks of shunt infection and malfunction. Its drawback is the recurrence rate of hydrocephalus (around 58% at 2 years of follow-up) due to formation of scarring in the perforated Lilie­quist membrane, which may require repeat surgery or conversion to CSF shunting.^26,27

Neuroendoscopic approaches are also used in cases of purely intraventricular pathology such as colloid cyst or choroid plexus papillomas. The concept behind neuroendoscopy is to achieve maximal resection in a minimally invasive way, using the natural cavity of the cerebral ventricles and reducing the need for brain retraction and, in particular, the risk of injury of the fornix (therefore causing memory deficits) of open transventricular approaches and of the corpus callosum necessary in inter­hemispheric approaches. Large tumor size and inability to tolerate a longer surgical procedure can be relative contraindications to a pure endoscopic approach to these lesions.

Degenerative spine disease

In recent years there has been a growing interest in the use of endoscopy for selected cases of degenerative lumbar spondylosis (generally, lateral disk herniation above the L5-S1 level or spinal canal stenosis). This approach has been shown to reduce postoperative incisional pain, scarring of the epidural space affecting the outcome of possible revision surgeries (recurrent disc herniation), and length of hospital stay.^14,15 Information on surgical nuances should be provided when consulting on selected patients with lumbar degenerative disease resistant to conservative treatment.

Carpal tunnel syndrome

Although endoscopic carpal tunnel release is controversial, its supporters report smaller incision size and lower recurrence rates due to better visualization of the entire carpal ligament compared with open surgery, with high patient satisfaction scores.^8,9,28

Craniosynostosis

Increasing data from specialized centers show that early endoscopic suturectomy is an effective treatment option alone or when combined with open surgeries for patients with syndromic and nonsyndromic craniosynostosis. The aesthetic advantage of small incisions (which can also be achieved with some open techniques) is accompanied by significant reductions in blood loss (median 162.4 mL), operative time (median 112.38 minutes), length of stay (median 2.56 days), and rates of perioperative complications (odds ratio 0.58), reoperation (odds ratio 0.37), and transfusion (odds ratio 0.09) compared with open approaches.¹⁶

SURGICAL TRAINING

Today’s patients expect high-quality healthcare, and they approach their surgeons with an enormous amount of information collected through unlimited Web-based access or peer-group blogs. In this respect, the pressure on young surgeons to achieve excellent results is high and growing from the very beginning of their careers.

Residency training programs differ in each country, and surgical standards usually focus on open microscopic procedures rather than newly developed endoscopic techniques. Endoscopic pituitary adenoma surgery, the most frequent neuroendoscopic procedure, is still performed mostly by experienced neurosurgeons, not trainees. Moreover, many training institutions might not offer pediatric neurosurgery care, limiting exposure to endoscopic third ventriculostomy procedures. The European Union of Medical Specialists, responsible for harmonizing and improving the quality of training of medical specialists in Europe, set low neuroendoscopic surgical requirements for trainees to complete their residency programs (minimum of 0 to optimum of 5 total transcranial or transsphenoidal pituitary adenoma resections as first operator, 10 procedures as assistant, and a minimum of 2 to an optimum of 4 endoscopic third ventriculostomies as first operator).²⁹

The need to develop training programs in neuroendoscopy is especially urgent because endoscopic surgery has a steeper learning curve than conventional microneurosurgery. In particular, endoscopy requires a good deal of dexterity and hand-eye coordination, which surgeons consider the main pitfall of neuroendoscopy. For such reasons, many accredited clinical fellowship programs have been developed inside and outside North America that offer intensive training in endoscopic skull-base surgery and pediatric neurosurgery after residency.

Some clinical studies have shown that the complication rate of neuroendoscopy is 15% to 18%.^27,30 In view of this statistic, it is ethically questionable to perform a randomized study to prospectively compare microscopic and endoscopic procedures. Surgeons specialize in one technique or the other, experience their own learning curve, and do not randomly decide which tool to use. Furthermore, every intracranial surgical exploration is unique and somewhat difficult to compare with each other without the risk of bias.

Multivariable rigid endoscopes like the EndoCAMeleon (Karl Storz, Tuttlingen, Germany) or the EndActive (Karl Storz, Tuttlingen, Germany) for cerebellopontine angle surgery represent a starting point to overcome some of the aforementioned limitations.^31,32 They are inserted in the surgical field with a direct 0° angulation view into the operative site beyond neurovascular structures that need to be preserved and that obstruct the microscopic view. Once the final position is reached, the field of view is directed toward the region of interest without moving the endoscope tip.

The EndoCAMeleon is a rigid rod-lens endoscope, steerable in one plane from –10° to +120° by a fine optomechanical mechanism. Anatomic laboratory testing found it to be superior in terms of usability and visualization compared with rigid fixed-angle endoscopes.^₃₁ The first clinical experiences have been promising; however, ergonomics and the limited perspective of a single plane of rotation leave room for improvement.

The EndActive endoscope might overcome such limitations.³³ This device is a rigid videoendoscope connected to a laptop (video data) and USB port (control and power supply); thus, it weighs less and can be held in one hand like a microsurgical instrument. The endoscopic imaging system allows the operator to simultaneously see a 160° wide-angle view of the site and an inset of a specific region of interest. The surgeon can hold the device like a microsurgical instrument in one hand and control movements precisely due to its reduced weight and ergonomic shape.

The multiplanar variable-view rigid endoscope has proven to be useful for working on diverse anatomic structures such as intracranial vessels and cranial nerves. The device is effective in narrow working spaces where even small movements can jeopardize the delicate surrounding structures. The multiplanar variable-viewing mechanism in a compact device offers advantages in terms of safety and ergonomics. Improving the usability will probably optimize the applicability of those endoscopic devices in neurosurgery. A major drawback of the current prototype is poor image resolution, which will probably soon be overcome with the ongoing progress in electronic microchip technology.

The addition of laser technology to endoscopic techniques offers a huge potential to neurosurgery but has achieved little acceptance to date. The reasons include concern regarding heat production, uncontrollable and distant penetration, and tissue interaction. Experiences with a 2-micron continuous­- wave laser (RevoLix Jr, LISA Laser Products, Katlenburg-Lindau, Germany) for neuroendoscopic intraventricular procedures proved this laser to be a valuable and useful tool with safe applicability for endoscopic intracranial procedures in patients of all ages.³⁴

Parallel to the launch of video screens for other uses with higher image definition, the image quality on the 2D endoscope cameras has been constantly improving over the last years. At the same time, the introduction of modern 3D endoscopic monitors is promising. However, 3D endoscopes have some disadvantages compared with the 2D endoscopes. First, the smallest 3D endoscopes are 4 mm in diameter, compared with 2.7 mm for 2D endoscopes. Moreover, the field of view with the 3D endoscope is less than half of that with conventional 2D endoscopes.³⁴ When working in and around a region with critical neurovascular structures in close proximity, this loss of field of view can result in an increase in iatrogenic injury from the endoscope. In addition, 3D endoscopes require special glasses, generating a potential obstacle to the seamless integration of visual information from the microscope and endoscope. Finally, some surgeons experience vertigo when looking at the 3D picture through the glasses, which limits its universal applicability.

CONCLUSIONS

Using the endoscope and microscope as complementary and not competing tools allows surgeons to benefit from both technologies at the same time.^35,36 The intraoperative combination of these 2 powerful visualization tools expands the effectiveness of microsurgical procedures and has the potential to further improve surgical results and reduce surgical risks. With endoscope-assisted microsurgery, visualization is often far superior to surgical maneuverability.

Endoscopic neurosurgery will likely be influenced by further innovations in optical physics, electronics, and robotics. Specific implementations in endoscopic systems are likely to pave the way for remarkable progress in minimally invasive surgery, such as robotic surgical technology, further miniaturization of devices, improvements in 3D endoscopy, multiport endoscopy, and new designs for surgical instruments. Future progress in flexible endoscopes and wireless capsule or camera technology may reduce our dependence on rigid rod lens systems. Rigid variable-view endoscopes will bring endoscopes closer to ideal attributes utilizing newer instrumentation that is tailored to specific indications and techniques.^37,38 Extension of the visual field by the feature of a movable optic lens may allow the neurosurgeon to use tailored keyhole approaches to treat pathologies in smaller surgical corridors with less trauma and greater efficacy.

Over the last 3 decades, the endoscope has become a highly valued visualization tool in neurosurgery, applicable to a broad range of neurosurgical procedures. Following the pace of technological innovations, the quality of the instrumentation has greatly improved along with the status of endoscopy in the neurosurgical field. The use of the endoscope in interdisciplinary extended transnasal approaches revolutionized skull-base surgery.¹ Transcranial neurosurgery took advantage of the endoscope for inspection, endoscope-assisted, and endoscope-controlled procedures, although the main visualization tool during these interventions remains the operating microscope.

At present, endoscopy has applications in a variety of neurosurgical procedures including transnasal approaches for pituitary and other skull-base tumors, third ventriculostomy, and resection of intraventricular tumors. The range of application is expanding to include extracranial procedures such as peripheral nerve and spine surgery.

From Li KW, Nelson C, Suk I, Jallo GI. Neuroendoscopy: past, present, and future. Neurosurg Focus 2005; 19(6):E1. Figure used with permission.

CURRENT CONCEPTS

Hopf and Perneczky² defined the terminology regarding endoscopic procedures and divided them into 3 categories:

Pure endoscopic neurosurgery, ie, procedures performed through working channels under complete endoscopic visualization and with endoscopic instrumentation (Figure 1).³

Endoscope-controlled microsurgery, ie, operations performed with standard microsurgical instruments under endoscopic visualization—the microscope is not used (Figure 2).

Endoscope-assisted neurosurgery, ie, the use of both microscope and endoscope during the same intervention. In endoscopic inspection the endoscope is solely used as an adjunctive tool for visualization and not for surgical manipulations.

Enhanced area and surgical dissection

Technical innovations are probably the major reason for the growing role of endoscopy in neurosurgery over the last 3 decades.⁴ High-definition imaging, neuronavigation, new instruments, an interdisciplinary approach mostly with ear, nose, and throat (ENT) surgeons, and detailed anatomic studies led to the breakthrough of endoscopic endonasal extended approaches in skull-base surgery.⁵

These endoscopic techniques allow the neurosurgeon to optimize tumor resection, increasing the area of surgical dissection without increasing the size of the surgical approach, thereby limiting perioperative morbidity due to surgical manipulation of eloquent brain structures. Endoscopy offers direct illumination of the operative field, magnification, and the ability to look around corners with angled optics.

However, while angled endoscopic optics provide various visual perspectives, the surgical issue is not only to see but also to work on and around remote structures. Microsurgical endoscope-assisted manipulations require optimal working angles that are guaranteed only by a sufficiently large craniotomy. As an example, a dissection study by Chaynes et al⁶ highlights that a craniotomy that is too narrow often hinders a sufficient exploration of the entire cerebellopontine angle. Most neuro­surgeons are familiar with the operating microscope. The microscopic field of inspection is 3-dimensional (3D) and of high quality. However, the light stream is straight and thus limited in the narrow and angled corridor of the cerebellopontine angle or in the perimesencephalic cisterns. In these situations, the angled optic of the endoscope offers the advantage of being able to look around the corner with the appropriate amount of direct illumination.⁷

Minimally invasive endoscopic approaches are also being used in peripheral nerve surgery, especially carpal tunnel decompression. The first carpal tunnel release treated endoscopically was performed by Okutsu et al in the late 1980s.⁸ Since that time, endoscopic carpal tunnel decompression has become very common and is the preferred method for many surgeons, using either single-portal or dual-portal techniques. Although the superiority of endoscopic over conventional minimally invasive microsurgical peripheral nerve surgeries has not been proven, large series of endoscopic carpal tunnel decompressions have reported low complication rates and excellent success rates with high patient satisfaction scores.^8,9

Visualization of the spinal canal

Expanding the use of the endoscope to spine surgery, endoscopic explorations of the interlaminar spaces after having completed open surgical laminectomies have been reported since the early 1980s,¹⁰ while endoscope-assisted interlaminar procedures started in the late 1990s.^11–13 The development of fully endoscopic transforaminal or interlaminar approaches for lumbar stenosis or lumbar disk herniation has been ongoing in the last 2 decades. The rationale for direct endoscopic visualization of the spinal canal is to reduce scarring of the epidural space, which might affect the outcome of possible revision surgeries (recurrent disk herniation), and to reduce injury to the paraspinal muscles, which may reduce postoperative incisional pain and length of hospital stay. Major limiting factors for fully endoscopic spine surgeries such as the narrow working channels (which are limited by the osseous perimeter of the neuroforamina, as well as the pelvis and abdominal structures) and the learning curve for the surgeons are, however, still matters of debate and restrict the use of endoscopy to very carefully selected cases.^14,15

Pediatric craniosynostosis

Recently, the use of the endoscope has extended to treatment of craniosynostosis in pediatric patients, historically treated with large and occasionally staged craniotomic approaches. A meta-analysis of the literature showed statistically significant reductions in blood loss and rates of perioperative complications, reoperation, and transfusion compared with open approaches.¹⁶

Technical limitations

While neurosurgeons increasingly advocate the use of the endoscope in their practice, the development of instruments for endoscopic surgery does not always follow the same pace. There are technical problems with current rigid endoscopes and ergonomic limitations of the endoscope-assisted techniques in trans­cranial neurosurgery. The endoscope itself occupies space in an already limited surgical corridor like the posterior fossa, the parasellar space, or the intraventricular region. The ideal endoscope is thin and sturdy, does not generate heat, and provides high-resolution images. In addition, a self-irrigating feature could minimize the need to remove and reinsert the endoscope for cleaning. Finally, most intracranial surgery is extremely delicate and requires bimanual dissection. The ideal endoscope should also be easily integrated with a holder that allows the surgeon to easily transition between static and dynamic endoscope movements.

Newer flexible fiberscopes with even smaller diameters are likely to be launched on the market in the near future. When working in a surgical corridor less than 10 mm wide, this difference could be substantial.

In addition, surgical instruments specifically designed for endoscopic endonasal procedures are needed for microdissection in these regions, which were previously only visible but not reachable endoscopically. These include tools such as malleable suctions and curettes, rotatable back-biting microscissors, and malleable bipolar instruments (Figure 3).

IMPACT OF NEUROENDOSCOPY IN CURRENT CLINICAL PRACTICE

The introduction of endoscopy in neurosurgery changed many treatment paradigms and had an important impact on morbidity and outcomes. In this section, we discuss the specific indications, contraindications, and expected benefit of endoscopic vs open surgical approaches applied to neurosurgical pathology at the present time.

Skull-base tumors and CSF leaks

The use of the endoscope in skull-base surgery was originally applied to purely midline intrasellar tumors without suprasellar or lateral extension beyond the carotid cave. Ideal cases were intrasellar pituitary microadenomas not responding to medical treatment or Rathke cleft cysts.

These pathologies were traditionally addressed via microscopic craniotomic approaches and later through sublabial or transnasal transsphenoidal approaches. Traditional transsphenoidal approaches were highly invasive for the oral mucosa, causing delayed healing, oral dysesthesia, and, in some cases, loss of the superior dental arch (sublabial) or limited visualization and surgical maneuverability (microscopic endonasal).

The endoscope offered better visualization and surgical freedom, thus allowing higher resection rates to be achieved. Resection of purely intrasellar pathology with preservation of the diaphragma sellae as a barrier to the subarachnoid cysterns and third ventricle guaranteed a lower incidence of cerebrospinal fluid (CSF) leaks.

New endoscope optics with varied angles, together with dedicated long surgical instruments with low steric volume, offered a large variety of new endonasal surgical corridors, so-called expanded endonasal approaches on the sagittal and coronal planes, as discussed in detail by Kassam et al.^17–19 These allowed endoscopic treatment of invasive tumors extending on the coronary plane into the suprasellar region or invading the cavernous sinuses (pituitary macroadenomas, craniopharyngiomas).

Highly specialized centers with expertise in endoscopic skull-base surgery can now also offer pure endoscopic treatment for some selected cases of lesions located far laterally to the cavernous sinus, such as trigeminal schwannomas, or along the sagittal plane like olfactory groove or tuberculum sellae meningiomas and clival lesions (chordomas, chondrosarcomas).

As one might expect, the increase in surgical complexity corresponded to an increase in complication rates. For example, the incidence of CSF leaks varied from 5% for standard midline transsphenoidal approaches to 11% for expanded endonasal approaches.^20,21 The consolidation of the use of the endoscope and the cooperation with ENT surgeons led to the development of surgical strategies to prevent and reduce the incidence of CSF leaks, such as the use of “rescue flaps,” nasoseptal flaps, or temporoparietal fascia flaps.^21–23

The development of such techniques allowed endoscopic endonasal approaches to be used in treatment of other pathologies, such as spontaneous CSF leaks, treated in the past with large transcranial repairs that carried high morbidity rates due to the surgical frontal lobe retraction and injury to the olfactory mucosa.^24,25 Progress in the field of neuroendoscopy therefore led to the creation of specialized endoscopic skull-base surgery centers, including neurosurgery, ENT, ophthalmology, and endocrinology services.

In clinical practice, when evaluating a patient with intracranial skull-base pathology amenable to endoscopic resection, one should consider referring the patient not only to a neurosurgeon, but also to an ENT surgeon for preoperative assessment of the sinonasal cavities. The same concept applies to postsurgical follow-up, which is mostly performed by the ENT physician to assess nasal mucosa healing and nasal hygiene.

The introduction of endoscopic third ventriculostomy created the opportunity to offer a more physiologic treatment in selected patients with obstructive hydrocephalus by creating an internal CSF diversion through the basal cisterns. Two advantages of this procedure are that it does not create dependence on a CSF shunt, and it eliminates the related risks of shunt infection and malfunction. Its drawback is the recurrence rate of hydrocephalus (around 58% at 2 years of follow-up) due to formation of scarring in the perforated Lilie­quist membrane, which may require repeat surgery or conversion to CSF shunting.^26,27

Neuroendoscopic approaches are also used in cases of purely intraventricular pathology such as colloid cyst or choroid plexus papillomas. The concept behind neuroendoscopy is to achieve maximal resection in a minimally invasive way, using the natural cavity of the cerebral ventricles and reducing the need for brain retraction and, in particular, the risk of injury of the fornix (therefore causing memory deficits) of open transventricular approaches and of the corpus callosum necessary in inter­hemispheric approaches. Large tumor size and inability to tolerate a longer surgical procedure can be relative contraindications to a pure endoscopic approach to these lesions.

Degenerative spine disease

In recent years there has been a growing interest in the use of endoscopy for selected cases of degenerative lumbar spondylosis (generally, lateral disk herniation above the L5-S1 level or spinal canal stenosis). This approach has been shown to reduce postoperative incisional pain, scarring of the epidural space affecting the outcome of possible revision surgeries (recurrent disc herniation), and length of hospital stay.^14,15 Information on surgical nuances should be provided when consulting on selected patients with lumbar degenerative disease resistant to conservative treatment.

Carpal tunnel syndrome

Although endoscopic carpal tunnel release is controversial, its supporters report smaller incision size and lower recurrence rates due to better visualization of the entire carpal ligament compared with open surgery, with high patient satisfaction scores.^8,9,28

Craniosynostosis

Increasing data from specialized centers show that early endoscopic suturectomy is an effective treatment option alone or when combined with open surgeries for patients with syndromic and nonsyndromic craniosynostosis. The aesthetic advantage of small incisions (which can also be achieved with some open techniques) is accompanied by significant reductions in blood loss (median 162.4 mL), operative time (median 112.38 minutes), length of stay (median 2.56 days), and rates of perioperative complications (odds ratio 0.58), reoperation (odds ratio 0.37), and transfusion (odds ratio 0.09) compared with open approaches.¹⁶

SURGICAL TRAINING

Today’s patients expect high-quality healthcare, and they approach their surgeons with an enormous amount of information collected through unlimited Web-based access or peer-group blogs. In this respect, the pressure on young surgeons to achieve excellent results is high and growing from the very beginning of their careers.

Residency training programs differ in each country, and surgical standards usually focus on open microscopic procedures rather than newly developed endoscopic techniques. Endoscopic pituitary adenoma surgery, the most frequent neuroendoscopic procedure, is still performed mostly by experienced neurosurgeons, not trainees. Moreover, many training institutions might not offer pediatric neurosurgery care, limiting exposure to endoscopic third ventriculostomy procedures. The European Union of Medical Specialists, responsible for harmonizing and improving the quality of training of medical specialists in Europe, set low neuroendoscopic surgical requirements for trainees to complete their residency programs (minimum of 0 to optimum of 5 total transcranial or transsphenoidal pituitary adenoma resections as first operator, 10 procedures as assistant, and a minimum of 2 to an optimum of 4 endoscopic third ventriculostomies as first operator).²⁹

The need to develop training programs in neuroendoscopy is especially urgent because endoscopic surgery has a steeper learning curve than conventional microneurosurgery. In particular, endoscopy requires a good deal of dexterity and hand-eye coordination, which surgeons consider the main pitfall of neuroendoscopy. For such reasons, many accredited clinical fellowship programs have been developed inside and outside North America that offer intensive training in endoscopic skull-base surgery and pediatric neurosurgery after residency.

Some clinical studies have shown that the complication rate of neuroendoscopy is 15% to 18%.^27,30 In view of this statistic, it is ethically questionable to perform a randomized study to prospectively compare microscopic and endoscopic procedures. Surgeons specialize in one technique or the other, experience their own learning curve, and do not randomly decide which tool to use. Furthermore, every intracranial surgical exploration is unique and somewhat difficult to compare with each other without the risk of bias.

Multivariable rigid endoscopes like the EndoCAMeleon (Karl Storz, Tuttlingen, Germany) or the EndActive (Karl Storz, Tuttlingen, Germany) for cerebellopontine angle surgery represent a starting point to overcome some of the aforementioned limitations.^31,32 They are inserted in the surgical field with a direct 0° angulation view into the operative site beyond neurovascular structures that need to be preserved and that obstruct the microscopic view. Once the final position is reached, the field of view is directed toward the region of interest without moving the endoscope tip.

The EndoCAMeleon is a rigid rod-lens endoscope, steerable in one plane from –10° to +120° by a fine optomechanical mechanism. Anatomic laboratory testing found it to be superior in terms of usability and visualization compared with rigid fixed-angle endoscopes.^₃₁ The first clinical experiences have been promising; however, ergonomics and the limited perspective of a single plane of rotation leave room for improvement.

The EndActive endoscope might overcome such limitations.³³ This device is a rigid videoendoscope connected to a laptop (video data) and USB port (control and power supply); thus, it weighs less and can be held in one hand like a microsurgical instrument. The endoscopic imaging system allows the operator to simultaneously see a 160° wide-angle view of the site and an inset of a specific region of interest. The surgeon can hold the device like a microsurgical instrument in one hand and control movements precisely due to its reduced weight and ergonomic shape.

The multiplanar variable-view rigid endoscope has proven to be useful for working on diverse anatomic structures such as intracranial vessels and cranial nerves. The device is effective in narrow working spaces where even small movements can jeopardize the delicate surrounding structures. The multiplanar variable-viewing mechanism in a compact device offers advantages in terms of safety and ergonomics. Improving the usability will probably optimize the applicability of those endoscopic devices in neurosurgery. A major drawback of the current prototype is poor image resolution, which will probably soon be overcome with the ongoing progress in electronic microchip technology.

The addition of laser technology to endoscopic techniques offers a huge potential to neurosurgery but has achieved little acceptance to date. The reasons include concern regarding heat production, uncontrollable and distant penetration, and tissue interaction. Experiences with a 2-micron continuous­- wave laser (RevoLix Jr, LISA Laser Products, Katlenburg-Lindau, Germany) for neuroendoscopic intraventricular procedures proved this laser to be a valuable and useful tool with safe applicability for endoscopic intracranial procedures in patients of all ages.³⁴

Parallel to the launch of video screens for other uses with higher image definition, the image quality on the 2D endoscope cameras has been constantly improving over the last years. At the same time, the introduction of modern 3D endoscopic monitors is promising. However, 3D endoscopes have some disadvantages compared with the 2D endoscopes. First, the smallest 3D endoscopes are 4 mm in diameter, compared with 2.7 mm for 2D endoscopes. Moreover, the field of view with the 3D endoscope is less than half of that with conventional 2D endoscopes.³⁴ When working in and around a region with critical neurovascular structures in close proximity, this loss of field of view can result in an increase in iatrogenic injury from the endoscope. In addition, 3D endoscopes require special glasses, generating a potential obstacle to the seamless integration of visual information from the microscope and endoscope. Finally, some surgeons experience vertigo when looking at the 3D picture through the glasses, which limits its universal applicability.

CONCLUSIONS

Using the endoscope and microscope as complementary and not competing tools allows surgeons to benefit from both technologies at the same time.^35,36 The intraoperative combination of these 2 powerful visualization tools expands the effectiveness of microsurgical procedures and has the potential to further improve surgical results and reduce surgical risks. With endoscope-assisted microsurgery, visualization is often far superior to surgical maneuverability.

Endoscopic neurosurgery will likely be influenced by further innovations in optical physics, electronics, and robotics. Specific implementations in endoscopic systems are likely to pave the way for remarkable progress in minimally invasive surgery, such as robotic surgical technology, further miniaturization of devices, improvements in 3D endoscopy, multiport endoscopy, and new designs for surgical instruments. Future progress in flexible endoscopes and wireless capsule or camera technology may reduce our dependence on rigid rod lens systems. Rigid variable-view endoscopes will bring endoscopes closer to ideal attributes utilizing newer instrumentation that is tailored to specific indications and techniques.^37,38 Extension of the visual field by the feature of a movable optic lens may allow the neurosurgeon to use tailored keyhole approaches to treat pathologies in smaller surgical corridors with less trauma and greater efficacy.

We need to recognize the diverse problems that patients with potential multisystem disease can develop, lobby when necessary for them to be seen promptly by the relevant specialists, and initiate appropriate diagnostic testing and management in less-urgent scenarios. Most of us need frequent refreshers on the clinical manifestations of these disorders so that we can recognize them when they appear unannounced in our exam rooms.

The caregiver with a prepared mind is more likely to experience the diagnostic epiphany, and then use point-of-care references to hone in on the details. With many patients and clinical conundrums, the basics matter.

Dr. Chester Oddis, in this issue of the Journal, reviews the basics of several primary muscle disorders. He discusses, in a case-based format extracted from his recent Medicine Grand Rounds presentation at Cleveland Clinic, nuances of specific diagnoses and the clinical progression of diseases that are critical to be aware of in order to recognize and manage them, and expeditiously refer the patient to our appropriate subspecialty colleagues.

Major challenges exist in recognizing the inflammatory myopathies and their mimics early in their course. These are serious but uncommon entities, and in part because patients and physicians often attribute their early symptoms to more-common causes, diagnosis can be elusive—until the possibility is considered. We hope that Dr. Oddis’s article will make it easier to rapidly recognize these muscle disorders.

Patients often struggle to explain their symptoms of early muscle dysfunction. Since patients often verbalize their fatigue as “feeling weak,” we often misconstrue complaints of true muscle weakness (like difficulty walking up steps) as being due to fatigue. Add in some anemia from chronic inflammation and some “liver test” abnormalities, and it is easy to see how the recognition of true muscle weakness can be delayed.

We can tease muscle weakness from fatigue or dyspnea by asking the patient to specifically and functionally describe their “weakness,” and then by asking pointed questions: “Do you have difficulty getting up from the toilet without using your arms? Do you have trouble brushing your hair or teeth?” Physical examination can clearly help here, but without routine examination of muscle strength in normal fragile elderly patients, the degree of muscle weakness can be difficult to assess. Likewise challenging is detecting the early onset of weakness by examination in a 280-lb power-lifter.

Obtaining an accurate functional and behavioral history is often critical to the early recognition of muscle disease. Muscle pain, as Dr. Oddis notes, is not a characteristic feature of many myopathies, whereas, paradoxically, the coexistence of new-onset symmetrical small-joint pain (especially with arthritis) along with muscle weakness can be a powerful clue to the diagnosis of an inflammatory myopathy.

An elevated creatine kinase (CK) level generally points directly to a muscle disease, although some neurologic disorders are associated with elevations in CK, and the entity of benign “hyperCKemia” must be recognized and not overmanaged. The latter becomes a problem when laboratory tests are allowed to drive the diagnostic evaluation in a vacuum of clinical details.

A more common scenario is the misinterpretation of common laboratory test abnormalities in the setting of a patient with “fatigue” or generalized weakness who has elevations in aspartate aminotransferase (AST) and alanine aminotransferase (ALT). Although AST and ALT are often called “liver function tests,” these enzymes are also abundant in skeletal muscle, and since they are included on routine biochemical panels, their elevation often leads to liver imaging and sometimes even biopsy before anyone recognizes muscle disease as the cause of the patient’s symptoms and laboratory test abnormalities. Hence, a muscle source (or hemolysis) should at least be considered when AST and ALT are elevated in the absence of elevated alkaline phosphatase or gamma-glutamyl transferase.

When evaluating innumerable clinical scenarios, experienced clinicians can most certainly generate similar principles of diagnostic reasoning, based on having a few fundamental facts at their fingertips. Increasing the chances of having a prepared mind when confronted with a patient with a less-than-straightforward set of symptoms is one of my major arguments in support of continuing to read and generate internal medicine teaching literature and to attend and participate in clinical teaching conferences such as Medicine Grand Rounds. It is also why we will continue to appreciate and publish presentations like this one in the Journal.

I don’t expect to retain all the details from these and similar papers, and I know we all carry virtually infinite databases in our pockets. But keeping a few clinical pearls outside of my specialty in my head comes in handy. Having a prepared mind makes it much easier to converse with patients, to promptly initiate appropriate testing, plans, and consultations, and to then decide what to search for on my smartphone between patients.

We need to recognize the diverse problems that patients with potential multisystem disease can develop, lobby when necessary for them to be seen promptly by the relevant specialists, and initiate appropriate diagnostic testing and management in less-urgent scenarios. Most of us need frequent refreshers on the clinical manifestations of these disorders so that we can recognize them when they appear unannounced in our exam rooms.

The caregiver with a prepared mind is more likely to experience the diagnostic epiphany, and then use point-of-care references to hone in on the details. With many patients and clinical conundrums, the basics matter.

Dr. Chester Oddis, in this issue of the Journal, reviews the basics of several primary muscle disorders. He discusses, in a case-based format extracted from his recent Medicine Grand Rounds presentation at Cleveland Clinic, nuances of specific diagnoses and the clinical progression of diseases that are critical to be aware of in order to recognize and manage them, and expeditiously refer the patient to our appropriate subspecialty colleagues.

Major challenges exist in recognizing the inflammatory myopathies and their mimics early in their course. These are serious but uncommon entities, and in part because patients and physicians often attribute their early symptoms to more-common causes, diagnosis can be elusive—until the possibility is considered. We hope that Dr. Oddis’s article will make it easier to rapidly recognize these muscle disorders.

Patients often struggle to explain their symptoms of early muscle dysfunction. Since patients often verbalize their fatigue as “feeling weak,” we often misconstrue complaints of true muscle weakness (like difficulty walking up steps) as being due to fatigue. Add in some anemia from chronic inflammation and some “liver test” abnormalities, and it is easy to see how the recognition of true muscle weakness can be delayed.

We can tease muscle weakness from fatigue or dyspnea by asking the patient to specifically and functionally describe their “weakness,” and then by asking pointed questions: “Do you have difficulty getting up from the toilet without using your arms? Do you have trouble brushing your hair or teeth?” Physical examination can clearly help here, but without routine examination of muscle strength in normal fragile elderly patients, the degree of muscle weakness can be difficult to assess. Likewise challenging is detecting the early onset of weakness by examination in a 280-lb power-lifter.

Obtaining an accurate functional and behavioral history is often critical to the early recognition of muscle disease. Muscle pain, as Dr. Oddis notes, is not a characteristic feature of many myopathies, whereas, paradoxically, the coexistence of new-onset symmetrical small-joint pain (especially with arthritis) along with muscle weakness can be a powerful clue to the diagnosis of an inflammatory myopathy.

An elevated creatine kinase (CK) level generally points directly to a muscle disease, although some neurologic disorders are associated with elevations in CK, and the entity of benign “hyperCKemia” must be recognized and not overmanaged. The latter becomes a problem when laboratory tests are allowed to drive the diagnostic evaluation in a vacuum of clinical details.

A more common scenario is the misinterpretation of common laboratory test abnormalities in the setting of a patient with “fatigue” or generalized weakness who has elevations in aspartate aminotransferase (AST) and alanine aminotransferase (ALT). Although AST and ALT are often called “liver function tests,” these enzymes are also abundant in skeletal muscle, and since they are included on routine biochemical panels, their elevation often leads to liver imaging and sometimes even biopsy before anyone recognizes muscle disease as the cause of the patient’s symptoms and laboratory test abnormalities. Hence, a muscle source (or hemolysis) should at least be considered when AST and ALT are elevated in the absence of elevated alkaline phosphatase or gamma-glutamyl transferase.

When evaluating innumerable clinical scenarios, experienced clinicians can most certainly generate similar principles of diagnostic reasoning, based on having a few fundamental facts at their fingertips. Increasing the chances of having a prepared mind when confronted with a patient with a less-than-straightforward set of symptoms is one of my major arguments in support of continuing to read and generate internal medicine teaching literature and to attend and participate in clinical teaching conferences such as Medicine Grand Rounds. It is also why we will continue to appreciate and publish presentations like this one in the Journal.

I don’t expect to retain all the details from these and similar papers, and I know we all carry virtually infinite databases in our pockets. But keeping a few clinical pearls outside of my specialty in my head comes in handy. Having a prepared mind makes it much easier to converse with patients, to promptly initiate appropriate testing, plans, and consultations, and to then decide what to search for on my smartphone between patients.

We need to recognize the diverse problems that patients with potential multisystem disease can develop, lobby when necessary for them to be seen promptly by the relevant specialists, and initiate appropriate diagnostic testing and management in less-urgent scenarios. Most of us need frequent refreshers on the clinical manifestations of these disorders so that we can recognize them when they appear unannounced in our exam rooms.

The caregiver with a prepared mind is more likely to experience the diagnostic epiphany, and then use point-of-care references to hone in on the details. With many patients and clinical conundrums, the basics matter.

Dr. Chester Oddis, in this issue of the Journal, reviews the basics of several primary muscle disorders. He discusses, in a case-based format extracted from his recent Medicine Grand Rounds presentation at Cleveland Clinic, nuances of specific diagnoses and the clinical progression of diseases that are critical to be aware of in order to recognize and manage them, and expeditiously refer the patient to our appropriate subspecialty colleagues.

Major challenges exist in recognizing the inflammatory myopathies and their mimics early in their course. These are serious but uncommon entities, and in part because patients and physicians often attribute their early symptoms to more-common causes, diagnosis can be elusive—until the possibility is considered. We hope that Dr. Oddis’s article will make it easier to rapidly recognize these muscle disorders.

Patients often struggle to explain their symptoms of early muscle dysfunction. Since patients often verbalize their fatigue as “feeling weak,” we often misconstrue complaints of true muscle weakness (like difficulty walking up steps) as being due to fatigue. Add in some anemia from chronic inflammation and some “liver test” abnormalities, and it is easy to see how the recognition of true muscle weakness can be delayed.

We can tease muscle weakness from fatigue or dyspnea by asking the patient to specifically and functionally describe their “weakness,” and then by asking pointed questions: “Do you have difficulty getting up from the toilet without using your arms? Do you have trouble brushing your hair or teeth?” Physical examination can clearly help here, but without routine examination of muscle strength in normal fragile elderly patients, the degree of muscle weakness can be difficult to assess. Likewise challenging is detecting the early onset of weakness by examination in a 280-lb power-lifter.

Obtaining an accurate functional and behavioral history is often critical to the early recognition of muscle disease. Muscle pain, as Dr. Oddis notes, is not a characteristic feature of many myopathies, whereas, paradoxically, the coexistence of new-onset symmetrical small-joint pain (especially with arthritis) along with muscle weakness can be a powerful clue to the diagnosis of an inflammatory myopathy.

An elevated creatine kinase (CK) level generally points directly to a muscle disease, although some neurologic disorders are associated with elevations in CK, and the entity of benign “hyperCKemia” must be recognized and not overmanaged. The latter becomes a problem when laboratory tests are allowed to drive the diagnostic evaluation in a vacuum of clinical details.

A more common scenario is the misinterpretation of common laboratory test abnormalities in the setting of a patient with “fatigue” or generalized weakness who has elevations in aspartate aminotransferase (AST) and alanine aminotransferase (ALT). Although AST and ALT are often called “liver function tests,” these enzymes are also abundant in skeletal muscle, and since they are included on routine biochemical panels, their elevation often leads to liver imaging and sometimes even biopsy before anyone recognizes muscle disease as the cause of the patient’s symptoms and laboratory test abnormalities. Hence, a muscle source (or hemolysis) should at least be considered when AST and ALT are elevated in the absence of elevated alkaline phosphatase or gamma-glutamyl transferase.

When evaluating innumerable clinical scenarios, experienced clinicians can most certainly generate similar principles of diagnostic reasoning, based on having a few fundamental facts at their fingertips. Increasing the chances of having a prepared mind when confronted with a patient with a less-than-straightforward set of symptoms is one of my major arguments in support of continuing to read and generate internal medicine teaching literature and to attend and participate in clinical teaching conferences such as Medicine Grand Rounds. It is also why we will continue to appreciate and publish presentations like this one in the Journal.

I don’t expect to retain all the details from these and similar papers, and I know we all carry virtually infinite databases in our pockets. But keeping a few clinical pearls outside of my specialty in my head comes in handy. Having a prepared mind makes it much easier to converse with patients, to promptly initiate appropriate testing, plans, and consultations, and to then decide what to search for on my smartphone between patients.

A 37-year-old woman presented with recurring painful swelling and erythema of the vulva over the last year. Despite a series of negative vaginal cultures, she was prescribed multiple courses of antifungal and antibacterial treatments, while her symptoms continued to worsen. She had no other relevant medical history except for occasional diarrhea and abdominal cramping, which were attributed to irritable bowel syndrome.

CROHN DISEASE OUTSIDE THE GASTROINTESTINAL TRACT

Crohn disease primarily affects the gastrointestinal tract but is associated with extraintestinal manifestations (in the oral cavity, eyes, skin, and joints) in up to 45% of patients.¹

The most common mucocutaneous manifestations are granulomatous lesions that extend directly from the gastrointestinal tract, including perianal and peristomal skin tags, fistulas, and perineal ulcerations. In most cases, the onset of cutaneous manifestations follows intestinal disease, but vulvar Crohn disease may precede gastrointestinal symptoms in approximately 25% of patients, with the average age at onset in the mid-30s.¹

The pathogenesis of vulvar Crohn disease remains unclear. One theory involves production of immune complexes from the gastrointestinal tract and a possible T-lymphocyte-mediated type IV hypersensitivity reaction.²

The diagnosis of vulvar Crohn disease should be considered in a patient who has vulvar pain, edema, and ulcerations not otherwise explained, whether or not gastrointestinal Crohn disease is present. The diagnosis is established with clinical history and characteristic histopathology on biopsy. Multiple biopsies may be needed, and early endoscopy is recommended to establish the diagnosis. The histologic features include noncaseating and nonnecrotizing granulomatous dermatitis or vulvitis with occasional reports of eosinophilic infiltrates and necrobiosis.^5,6 An imaging study such as ultrasonography is sometimes used to differentiate between a specific cutaneous manifestation of Crohn disease and its complications such as perianal fistula or abscess.

Clinical vulvar lesions are nonspecific, and those of Crohn disease are frequently mistaken for infectious, inflammatory, or traumatic vulvitis. Diagnostic biopsy for histologic analysis is warranted.

A 37-year-old woman presented with recurring painful swelling and erythema of the vulva over the last year. Despite a series of negative vaginal cultures, she was prescribed multiple courses of antifungal and antibacterial treatments, while her symptoms continued to worsen. She had no other relevant medical history except for occasional diarrhea and abdominal cramping, which were attributed to irritable bowel syndrome.

CROHN DISEASE OUTSIDE THE GASTROINTESTINAL TRACT

Crohn disease primarily affects the gastrointestinal tract but is associated with extraintestinal manifestations (in the oral cavity, eyes, skin, and joints) in up to 45% of patients.¹

The most common mucocutaneous manifestations are granulomatous lesions that extend directly from the gastrointestinal tract, including perianal and peristomal skin tags, fistulas, and perineal ulcerations. In most cases, the onset of cutaneous manifestations follows intestinal disease, but vulvar Crohn disease may precede gastrointestinal symptoms in approximately 25% of patients, with the average age at onset in the mid-30s.¹

The pathogenesis of vulvar Crohn disease remains unclear. One theory involves production of immune complexes from the gastrointestinal tract and a possible T-lymphocyte-mediated type IV hypersensitivity reaction.²

The diagnosis of vulvar Crohn disease should be considered in a patient who has vulvar pain, edema, and ulcerations not otherwise explained, whether or not gastrointestinal Crohn disease is present. The diagnosis is established with clinical history and characteristic histopathology on biopsy. Multiple biopsies may be needed, and early endoscopy is recommended to establish the diagnosis. The histologic features include noncaseating and nonnecrotizing granulomatous dermatitis or vulvitis with occasional reports of eosinophilic infiltrates and necrobiosis.^5,6 An imaging study such as ultrasonography is sometimes used to differentiate between a specific cutaneous manifestation of Crohn disease and its complications such as perianal fistula or abscess.

Clinical vulvar lesions are nonspecific, and those of Crohn disease are frequently mistaken for infectious, inflammatory, or traumatic vulvitis. Diagnostic biopsy for histologic analysis is warranted.

A 37-year-old woman presented with recurring painful swelling and erythema of the vulva over the last year. Despite a series of negative vaginal cultures, she was prescribed multiple courses of antifungal and antibacterial treatments, while her symptoms continued to worsen. She had no other relevant medical history except for occasional diarrhea and abdominal cramping, which were attributed to irritable bowel syndrome.

CROHN DISEASE OUTSIDE THE GASTROINTESTINAL TRACT

Crohn disease primarily affects the gastrointestinal tract but is associated with extraintestinal manifestations (in the oral cavity, eyes, skin, and joints) in up to 45% of patients.¹

The most common mucocutaneous manifestations are granulomatous lesions that extend directly from the gastrointestinal tract, including perianal and peristomal skin tags, fistulas, and perineal ulcerations. In most cases, the onset of cutaneous manifestations follows intestinal disease, but vulvar Crohn disease may precede gastrointestinal symptoms in approximately 25% of patients, with the average age at onset in the mid-30s.¹

The pathogenesis of vulvar Crohn disease remains unclear. One theory involves production of immune complexes from the gastrointestinal tract and a possible T-lymphocyte-mediated type IV hypersensitivity reaction.²

The diagnosis of vulvar Crohn disease should be considered in a patient who has vulvar pain, edema, and ulcerations not otherwise explained, whether or not gastrointestinal Crohn disease is present. The diagnosis is established with clinical history and characteristic histopathology on biopsy. Multiple biopsies may be needed, and early endoscopy is recommended to establish the diagnosis. The histologic features include noncaseating and nonnecrotizing granulomatous dermatitis or vulvitis with occasional reports of eosinophilic infiltrates and necrobiosis.^5,6 An imaging study such as ultrasonography is sometimes used to differentiate between a specific cutaneous manifestation of Crohn disease and its complications such as perianal fistula or abscess.

Clinical vulvar lesions are nonspecific, and those of Crohn disease are frequently mistaken for infectious, inflammatory, or traumatic vulvitis. Diagnostic biopsy for histologic analysis is warranted.

A 78-year-old woman presented to the emergency department with shortness of breath and palpitations and was found to have atrial fibrillation with rapid ventricular response. Medical therapy with drug therapy and cardioversion proved ineffective. She then underwent atrioventricular node ablation and placement of a pacemaker.

At the time of admission, anticoagulation was started with full-dose enoxaparin, injected subcutaneously on the left side of the abdominal wall, as her CHA₂DS₂-VASc score (http://chadvasc.org) was 5, due to age, female sex, and history of heart failure and hypertension.

Four days after admission, she reported lower abdominal pain, and her urine output was minimal. A bladder scan showed more than 500 mL of residual urine. She was hemodynamically stable, but physical examination revealed mild abdominal distention and tenderness in the suprapubic region. Laboratory testing showed a sharp rise in serum creatinine and a drop in hematocrit.

The patient was initially managed conservatively with serial physical examinations, monitoring of the hematocrit, serial imaging studies, and discontinuation of anticoagulation, but the pain and anuria persisted. Repeat computed tomography 15 days after admission showed that the hematoma had expanded, and she now had hydronephrosis on the right side as well, requiring urologic intervention with bilateral nephrostomy tube placement.

The size of the hematoma was evaluated with serial abdominal and pelvic examinations. After several days, her urine output had improved, the nephrostomy tubes were removed, and she was discharged.

RECTUS SHEATH HEMATOMA

Our patient had a giant pelvic hematoma, probably arising from the rectus sheath. This uncommon problem can arise from trauma, anticoagulation, or increased intra-abdominal pressure, but it can also occur spontaneously.¹

In rectus sheath hematoma, a branch of the inferior epigastric artery is injured at its insertion into the rectus abdominis muscle. Symptoms arise if bleeding does not stop spontaneously from a tamponade effect.²

We speculate that in our patient, deep injection of enoxaparin into the abdominal wall injured the inferior epigastric artery, which started the hematoma, and the bleeding was exacerbated by the anticoagulation effect of the enoxaparin.

Another form of pelvic hematoma is retro­peritoneal. It is most commonly caused by trauma but can occur due to rupture of the aorta, compression from tumors, or, infrequently, anticoagulation therapy.³

The role of anticoagulation

Spontaneous pelvic hematoma is usually missed as a cause of abdominal pain in patients on anticoagulation therapy and is mistaken for common acute conditions such as ulcer, diverticulitis, appendicitis, ovarian cyst torsion, and tumor.⁴ It usually develops within 5 days of starting anticoagulation therapy. Symptoms vary depending on the location of the hematoma and are best diagnosed with abdominal computed tomography, with sensitivity as high as 100%.

MANAGEMENT

Conservative management, reserved for patients in stable condition, includes temporarily stopping and reevaluating the risks and benefits of anticoagulation and antiplatelet agents, giving blood transfusions, and controlling pain. If conservative measures fail, options are arterial embolization, stent grafting, and blood vessel ligation.⁵ If these measures fail, patients should undergo surgical evacuation of the hematoma and ligation of bleeding vessels.⁶

TAKE-HOME MESSAGE

Subcutaneous injections, especially of anticoagulants, into the abdominal wall can increase the risk of hematoma. Other risk factors are older age, female sex, and thin body habitus with less abdominal fat.⁷ Healthcare professionals should avoid deep injections into the abdomen and should counsel patients and their caregivers about this, as well. The deltoid region could be a safer alternative.

A 78-year-old woman presented to the emergency department with shortness of breath and palpitations and was found to have atrial fibrillation with rapid ventricular response. Medical therapy with drug therapy and cardioversion proved ineffective. She then underwent atrioventricular node ablation and placement of a pacemaker.

At the time of admission, anticoagulation was started with full-dose enoxaparin, injected subcutaneously on the left side of the abdominal wall, as her CHA₂DS₂-VASc score (http://chadvasc.org) was 5, due to age, female sex, and history of heart failure and hypertension.

Four days after admission, she reported lower abdominal pain, and her urine output was minimal. A bladder scan showed more than 500 mL of residual urine. She was hemodynamically stable, but physical examination revealed mild abdominal distention and tenderness in the suprapubic region. Laboratory testing showed a sharp rise in serum creatinine and a drop in hematocrit.

The patient was initially managed conservatively with serial physical examinations, monitoring of the hematocrit, serial imaging studies, and discontinuation of anticoagulation, but the pain and anuria persisted. Repeat computed tomography 15 days after admission showed that the hematoma had expanded, and she now had hydronephrosis on the right side as well, requiring urologic intervention with bilateral nephrostomy tube placement.

The size of the hematoma was evaluated with serial abdominal and pelvic examinations. After several days, her urine output had improved, the nephrostomy tubes were removed, and she was discharged.

RECTUS SHEATH HEMATOMA

Our patient had a giant pelvic hematoma, probably arising from the rectus sheath. This uncommon problem can arise from trauma, anticoagulation, or increased intra-abdominal pressure, but it can also occur spontaneously.¹

In rectus sheath hematoma, a branch of the inferior epigastric artery is injured at its insertion into the rectus abdominis muscle. Symptoms arise if bleeding does not stop spontaneously from a tamponade effect.²

We speculate that in our patient, deep injection of enoxaparin into the abdominal wall injured the inferior epigastric artery, which started the hematoma, and the bleeding was exacerbated by the anticoagulation effect of the enoxaparin.

Another form of pelvic hematoma is retro­peritoneal. It is most commonly caused by trauma but can occur due to rupture of the aorta, compression from tumors, or, infrequently, anticoagulation therapy.³

The role of anticoagulation

Spontaneous pelvic hematoma is usually missed as a cause of abdominal pain in patients on anticoagulation therapy and is mistaken for common acute conditions such as ulcer, diverticulitis, appendicitis, ovarian cyst torsion, and tumor.⁴ It usually develops within 5 days of starting anticoagulation therapy. Symptoms vary depending on the location of the hematoma and are best diagnosed with abdominal computed tomography, with sensitivity as high as 100%.

MANAGEMENT

Conservative management, reserved for patients in stable condition, includes temporarily stopping and reevaluating the risks and benefits of anticoagulation and antiplatelet agents, giving blood transfusions, and controlling pain. If conservative measures fail, options are arterial embolization, stent grafting, and blood vessel ligation.⁵ If these measures fail, patients should undergo surgical evacuation of the hematoma and ligation of bleeding vessels.⁶

TAKE-HOME MESSAGE

Subcutaneous injections, especially of anticoagulants, into the abdominal wall can increase the risk of hematoma. Other risk factors are older age, female sex, and thin body habitus with less abdominal fat.⁷ Healthcare professionals should avoid deep injections into the abdomen and should counsel patients and their caregivers about this, as well. The deltoid region could be a safer alternative.

A 78-year-old woman presented to the emergency department with shortness of breath and palpitations and was found to have atrial fibrillation with rapid ventricular response. Medical therapy with drug therapy and cardioversion proved ineffective. She then underwent atrioventricular node ablation and placement of a pacemaker.

At the time of admission, anticoagulation was started with full-dose enoxaparin, injected subcutaneously on the left side of the abdominal wall, as her CHA₂DS₂-VASc score (http://chadvasc.org) was 5, due to age, female sex, and history of heart failure and hypertension.

Four days after admission, she reported lower abdominal pain, and her urine output was minimal. A bladder scan showed more than 500 mL of residual urine. She was hemodynamically stable, but physical examination revealed mild abdominal distention and tenderness in the suprapubic region. Laboratory testing showed a sharp rise in serum creatinine and a drop in hematocrit.

The patient was initially managed conservatively with serial physical examinations, monitoring of the hematocrit, serial imaging studies, and discontinuation of anticoagulation, but the pain and anuria persisted. Repeat computed tomography 15 days after admission showed that the hematoma had expanded, and she now had hydronephrosis on the right side as well, requiring urologic intervention with bilateral nephrostomy tube placement.

The size of the hematoma was evaluated with serial abdominal and pelvic examinations. After several days, her urine output had improved, the nephrostomy tubes were removed, and she was discharged.

RECTUS SHEATH HEMATOMA

Our patient had a giant pelvic hematoma, probably arising from the rectus sheath. This uncommon problem can arise from trauma, anticoagulation, or increased intra-abdominal pressure, but it can also occur spontaneously.¹

In rectus sheath hematoma, a branch of the inferior epigastric artery is injured at its insertion into the rectus abdominis muscle. Symptoms arise if bleeding does not stop spontaneously from a tamponade effect.²

We speculate that in our patient, deep injection of enoxaparin into the abdominal wall injured the inferior epigastric artery, which started the hematoma, and the bleeding was exacerbated by the anticoagulation effect of the enoxaparin.

Another form of pelvic hematoma is retro­peritoneal. It is most commonly caused by trauma but can occur due to rupture of the aorta, compression from tumors, or, infrequently, anticoagulation therapy.³

The role of anticoagulation

Spontaneous pelvic hematoma is usually missed as a cause of abdominal pain in patients on anticoagulation therapy and is mistaken for common acute conditions such as ulcer, diverticulitis, appendicitis, ovarian cyst torsion, and tumor.⁴ It usually develops within 5 days of starting anticoagulation therapy. Symptoms vary depending on the location of the hematoma and are best diagnosed with abdominal computed tomography, with sensitivity as high as 100%.

MANAGEMENT

Conservative management, reserved for patients in stable condition, includes temporarily stopping and reevaluating the risks and benefits of anticoagulation and antiplatelet agents, giving blood transfusions, and controlling pain. If conservative measures fail, options are arterial embolization, stent grafting, and blood vessel ligation.⁵ If these measures fail, patients should undergo surgical evacuation of the hematoma and ligation of bleeding vessels.⁶

TAKE-HOME MESSAGE

Subcutaneous injections, especially of anticoagulants, into the abdominal wall can increase the risk of hematoma. Other risk factors are older age, female sex, and thin body habitus with less abdominal fat.⁷ Healthcare professionals should avoid deep injections into the abdomen and should counsel patients and their caregivers about this, as well. The deltoid region could be a safer alternative.

Hospital-acquired infections related to construction and renovation activities account for more than 5,000 deaths per year across the United States.¹

Hospital construction, renovation, and demolition projects ultimately serve the interests of patients, but they also can put inpatients at risk of mold infection, Legionnaires disease, sleep deprivation, exacerbation of lung disease, and in rare cases, physical injury.

Hospitals are in a continuous state of transformation to meet the needs of medical and technologic advances and an increasing patient population,¹ and in the last 10 years, more than $200 billion has been spent on construction projects at US healthcare facilities. Therefore, constant attention is needed to reduce the risks to the health of hospitalized patients during these projects.

HOSPITAL-ACQUIRED INFECTIONS

Mold infections

Construction can cause substantial dust contamination and scatter large amounts of fungal spores. An analysis conducted during a period of excavation at a hospital campus showed a significant association between excavation activities and hospital-acquired mold infections (hazard ratio [HR] 2.8, P = .01) but not yeast infections (HR 0.75, P = .78).²

Aspergillus species have been the organisms most commonly involved in hospital-acquired mold infection. In a review of 53 studies including 458 patients,³A fumigatus was identified in 154 patients, and A flavus was identified in 101 patients. A niger, A terreus, A nidulans, Zygomycetes, and other fungi were also identified, but to a much lesser extent. Hematologic malignancies were the predominant underlying morbidity in 299 patients. Half of the sources of healthcare-associated Aspergillus outbreaks were estimated to result from construction and renovation activities within or surrounding the hospital.³

Heavy demolition and transportation of wreckage have been found to cause the greatest concentrations of Aspergillus species,¹ but even small concentrations may be sufficient to cause infection in high-risk hospitalized patients.³ Invasive pulmonary aspergillosis is the mold infection most commonly associated with these activities, particularly in immunocompromised and critically ill patients. It is characterized by invasion of lung tissue by Aspergillus hyphae. Hematogenous dissemination occurs in about 25% of patients, and the death rate often exceeds 50%.⁴

A review of cases of fungal infection during hospital construction, renovation, and demolition projects from 1976 to 2014 identified 372 infected patients, of whom 180 died.⁵ The majority of infections were due to Aspergillus. Other fungi included Rhizopus, Candida, and Fusarium. Infections occurred mainly in patients with hematologic malignancies and patients who had undergone stem cell transplant (76%), followed by patients with other malignancies or transplant (19%). Rarely affected were patients in the intensive care unit or patients with rheumatologic diseases or on hemodialysis.⁵

Legionnaires disease

Legionnaires disease is a form of atypical pneumonia caused by the bacterium Legionella, often associated with differing degrees of gastrointestinal symptoms. Legionella species are the bacteria most often associated with construction in hospitals, as construction and demolition often result in collections of stagnant water.

The primary mode of transmission is inhalation of contaminated mist or aerosols. Legionella species can also colonize newly constructed hospital buildings within weeks of installation of water fixtures.

In a large university-affiliated hospital, 2 cases of nosocomial legionellosis were identified during a period of major construction.⁶ An epidemiologic investigation traced the source to a widespread contamination of potable water within the hospital. One patient’s isolate was similar to that of a water sample from the faucet in his room, and an association between Legionnaires disease and construction was postulated.

Another institution’s newly constructed hematology-oncology unit identified 10 cases of Legionnaires disease over a 12-week period in patients and visitors with exposure to the unit during and within the incubation period.⁷ A clinical and environmental assessment found 3 clinical isolates of Legionella identical to environmental isolates found from the unit, strongly implicating the potable water system as the likely source.⁷

In Ohio, 11 cases of hospital-acquired Legionnaires disease were identified in patients moved to a newly constructed 12-story addition to a hospital, and 1 of those died.⁸

Legionella infections appear to be less common than mold infections when reviewing the available literature on patients exposed to hospital construction, renovation, or demolition activities. Yet unlike mold infections, which occur mostly in immunocompromised patients, Legionella also affects people with normal immunity.¹

Sleep deprivation

Noise in hospitals has been linked to sleep disturbances in inpatients. A study using noise dosimeters in a university hospital found a mean continuous noise level of 63.5 dBA (A-weighting of decibels indicates risk of hearing loss) over a 24-hour period, a level more than 2 times higher than the recommended 30 dBA.⁹ The same study also found a significant correlation between sleep disturbance in inpatients and increasing noise levels, in a dose-response manner.

Common sources of noise during construction may include power generators, welding and cutting equipment, and transport of materials. While construction activities themselves have yet to be directly linked to sleep deprivation in patients, construction is inevitably accompanied by noise.

Noise is the most common factor interfering with sleep reported by hospitalized patients. Other effects of noise on patients include a rise in heart rate and blood pressure, increased cholesterol and triglyceride levels, increased use of sedatives, and longer length of stay.^9,10 Although construction is rarely done at night, patients generally take naps during the day, so the noise is disruptive.

Physical injuries

Hospitalized patients rarely suffer injuries related to hospital construction. However, these incidents may be underreported. Few cases of physical injury in patients exposed to construction or renovation in healthcare facilities can be found through a Web search.^11,12

Exacerbation of lung disease

Inhalation of indoor air pollutants exposed during renovation can directly trigger an inflammatory response and cause exacerbation in patients with chronic lung diseases such as asthma and chronic obstructive pulmonary disease. No study has specifically examined the effect of hospital construction or renovation on exacerbation of chronic lung diseases in hospitalized patients. Nevertheless, dust and indoor air pollutants from building renovation have often been reported as agents associated with work-related asthma.¹³

THE MESSAGE

Although the risks to inpatients during hospital construction projects appear minimal, their effect can at times be detrimental, especially to the immunocompromised. Hospitals should adhere to infection control risk assessment protocols during construction events. The small number of outbreaks of construction-related infections can make the diagnosis of nosocomial origin of these infections challenging; a high index of suspicion is needed.

Currently in the United States, there is no standard regarding acceptable levels of airborne mold concentrations, and data to support routine hospital air sampling or validation of available air samplers are inadequate. This remains an area for future research.^14,15

Certain measures have been shown to significantly decrease the risk of mold infections and other nosocomial infections during construction projects, including¹⁶:

• Effective dust control through containment units and barriers
• Consistent use of high-efficiency particulate air filters in hospital units that care for immunocompromised and critically ill patients
• Routine surveillance.

Noise and vibration can be reduced by temporary walls and careful tool selection and scheduling. Similarly, temporary walls and other barriers help protect healthcare employees and patients from the risk of direct physical injury.

Preconstruction risk assessments that address infection control, safety, noise, and air quality are crucial, and the Joint Commission generally requires such assessments. Further, education of hospital staff and members of the construction team about the potential detrimental effects of hospital construction and renovation is essential to secure a safe environment.        

Hospital-acquired infections related to construction and renovation activities account for more than 5,000 deaths per year across the United States.¹

Hospital construction, renovation, and demolition projects ultimately serve the interests of patients, but they also can put inpatients at risk of mold infection, Legionnaires disease, sleep deprivation, exacerbation of lung disease, and in rare cases, physical injury.

Hospitals are in a continuous state of transformation to meet the needs of medical and technologic advances and an increasing patient population,¹ and in the last 10 years, more than $200 billion has been spent on construction projects at US healthcare facilities. Therefore, constant attention is needed to reduce the risks to the health of hospitalized patients during these projects.

HOSPITAL-ACQUIRED INFECTIONS

Mold infections

Construction can cause substantial dust contamination and scatter large amounts of fungal spores. An analysis conducted during a period of excavation at a hospital campus showed a significant association between excavation activities and hospital-acquired mold infections (hazard ratio [HR] 2.8, P = .01) but not yeast infections (HR 0.75, P = .78).²

Aspergillus species have been the organisms most commonly involved in hospital-acquired mold infection. In a review of 53 studies including 458 patients,³A fumigatus was identified in 154 patients, and A flavus was identified in 101 patients. A niger, A terreus, A nidulans, Zygomycetes, and other fungi were also identified, but to a much lesser extent. Hematologic malignancies were the predominant underlying morbidity in 299 patients. Half of the sources of healthcare-associated Aspergillus outbreaks were estimated to result from construction and renovation activities within or surrounding the hospital.³

Heavy demolition and transportation of wreckage have been found to cause the greatest concentrations of Aspergillus species,¹ but even small concentrations may be sufficient to cause infection in high-risk hospitalized patients.³ Invasive pulmonary aspergillosis is the mold infection most commonly associated with these activities, particularly in immunocompromised and critically ill patients. It is characterized by invasion of lung tissue by Aspergillus hyphae. Hematogenous dissemination occurs in about 25% of patients, and the death rate often exceeds 50%.⁴

A review of cases of fungal infection during hospital construction, renovation, and demolition projects from 1976 to 2014 identified 372 infected patients, of whom 180 died.⁵ The majority of infections were due to Aspergillus. Other fungi included Rhizopus, Candida, and Fusarium. Infections occurred mainly in patients with hematologic malignancies and patients who had undergone stem cell transplant (76%), followed by patients with other malignancies or transplant (19%). Rarely affected were patients in the intensive care unit or patients with rheumatologic diseases or on hemodialysis.⁵

Legionnaires disease

Legionnaires disease is a form of atypical pneumonia caused by the bacterium Legionella, often associated with differing degrees of gastrointestinal symptoms. Legionella species are the bacteria most often associated with construction in hospitals, as construction and demolition often result in collections of stagnant water.

The primary mode of transmission is inhalation of contaminated mist or aerosols. Legionella species can also colonize newly constructed hospital buildings within weeks of installation of water fixtures.

In a large university-affiliated hospital, 2 cases of nosocomial legionellosis were identified during a period of major construction.⁶ An epidemiologic investigation traced the source to a widespread contamination of potable water within the hospital. One patient’s isolate was similar to that of a water sample from the faucet in his room, and an association between Legionnaires disease and construction was postulated.

Another institution’s newly constructed hematology-oncology unit identified 10 cases of Legionnaires disease over a 12-week period in patients and visitors with exposure to the unit during and within the incubation period.⁷ A clinical and environmental assessment found 3 clinical isolates of Legionella identical to environmental isolates found from the unit, strongly implicating the potable water system as the likely source.⁷

In Ohio, 11 cases of hospital-acquired Legionnaires disease were identified in patients moved to a newly constructed 12-story addition to a hospital, and 1 of those died.⁸

Legionella infections appear to be less common than mold infections when reviewing the available literature on patients exposed to hospital construction, renovation, or demolition activities. Yet unlike mold infections, which occur mostly in immunocompromised patients, Legionella also affects people with normal immunity.¹

Sleep deprivation

Noise in hospitals has been linked to sleep disturbances in inpatients. A study using noise dosimeters in a university hospital found a mean continuous noise level of 63.5 dBA (A-weighting of decibels indicates risk of hearing loss) over a 24-hour period, a level more than 2 times higher than the recommended 30 dBA.⁹ The same study also found a significant correlation between sleep disturbance in inpatients and increasing noise levels, in a dose-response manner.

Common sources of noise during construction may include power generators, welding and cutting equipment, and transport of materials. While construction activities themselves have yet to be directly linked to sleep deprivation in patients, construction is inevitably accompanied by noise.

Noise is the most common factor interfering with sleep reported by hospitalized patients. Other effects of noise on patients include a rise in heart rate and blood pressure, increased cholesterol and triglyceride levels, increased use of sedatives, and longer length of stay.^9,10 Although construction is rarely done at night, patients generally take naps during the day, so the noise is disruptive.

Physical injuries

Hospitalized patients rarely suffer injuries related to hospital construction. However, these incidents may be underreported. Few cases of physical injury in patients exposed to construction or renovation in healthcare facilities can be found through a Web search.^11,12

Exacerbation of lung disease

Inhalation of indoor air pollutants exposed during renovation can directly trigger an inflammatory response and cause exacerbation in patients with chronic lung diseases such as asthma and chronic obstructive pulmonary disease. No study has specifically examined the effect of hospital construction or renovation on exacerbation of chronic lung diseases in hospitalized patients. Nevertheless, dust and indoor air pollutants from building renovation have often been reported as agents associated with work-related asthma.¹³

THE MESSAGE

Although the risks to inpatients during hospital construction projects appear minimal, their effect can at times be detrimental, especially to the immunocompromised. Hospitals should adhere to infection control risk assessment protocols during construction events. The small number of outbreaks of construction-related infections can make the diagnosis of nosocomial origin of these infections challenging; a high index of suspicion is needed.

Currently in the United States, there is no standard regarding acceptable levels of airborne mold concentrations, and data to support routine hospital air sampling or validation of available air samplers are inadequate. This remains an area for future research.^14,15

Certain measures have been shown to significantly decrease the risk of mold infections and other nosocomial infections during construction projects, including¹⁶:

• Effective dust control through containment units and barriers
• Consistent use of high-efficiency particulate air filters in hospital units that care for immunocompromised and critically ill patients
• Routine surveillance.

Noise and vibration can be reduced by temporary walls and careful tool selection and scheduling. Similarly, temporary walls and other barriers help protect healthcare employees and patients from the risk of direct physical injury.

Preconstruction risk assessments that address infection control, safety, noise, and air quality are crucial, and the Joint Commission generally requires such assessments. Further, education of hospital staff and members of the construction team about the potential detrimental effects of hospital construction and renovation is essential to secure a safe environment.        

Hospital-acquired infections related to construction and renovation activities account for more than 5,000 deaths per year across the United States.¹

Hospital construction, renovation, and demolition projects ultimately serve the interests of patients, but they also can put inpatients at risk of mold infection, Legionnaires disease, sleep deprivation, exacerbation of lung disease, and in rare cases, physical injury.

Hospitals are in a continuous state of transformation to meet the needs of medical and technologic advances and an increasing patient population,¹ and in the last 10 years, more than $200 billion has been spent on construction projects at US healthcare facilities. Therefore, constant attention is needed to reduce the risks to the health of hospitalized patients during these projects.

HOSPITAL-ACQUIRED INFECTIONS

Mold infections

Construction can cause substantial dust contamination and scatter large amounts of fungal spores. An analysis conducted during a period of excavation at a hospital campus showed a significant association between excavation activities and hospital-acquired mold infections (hazard ratio [HR] 2.8, P = .01) but not yeast infections (HR 0.75, P = .78).²

Aspergillus species have been the organisms most commonly involved in hospital-acquired mold infection. In a review of 53 studies including 458 patients,³A fumigatus was identified in 154 patients, and A flavus was identified in 101 patients. A niger, A terreus, A nidulans, Zygomycetes, and other fungi were also identified, but to a much lesser extent. Hematologic malignancies were the predominant underlying morbidity in 299 patients. Half of the sources of healthcare-associated Aspergillus outbreaks were estimated to result from construction and renovation activities within or surrounding the hospital.³

Heavy demolition and transportation of wreckage have been found to cause the greatest concentrations of Aspergillus species,¹ but even small concentrations may be sufficient to cause infection in high-risk hospitalized patients.³ Invasive pulmonary aspergillosis is the mold infection most commonly associated with these activities, particularly in immunocompromised and critically ill patients. It is characterized by invasion of lung tissue by Aspergillus hyphae. Hematogenous dissemination occurs in about 25% of patients, and the death rate often exceeds 50%.⁴

A review of cases of fungal infection during hospital construction, renovation, and demolition projects from 1976 to 2014 identified 372 infected patients, of whom 180 died.⁵ The majority of infections were due to Aspergillus. Other fungi included Rhizopus, Candida, and Fusarium. Infections occurred mainly in patients with hematologic malignancies and patients who had undergone stem cell transplant (76%), followed by patients with other malignancies or transplant (19%). Rarely affected were patients in the intensive care unit or patients with rheumatologic diseases or on hemodialysis.⁵

Legionnaires disease

Legionnaires disease is a form of atypical pneumonia caused by the bacterium Legionella, often associated with differing degrees of gastrointestinal symptoms. Legionella species are the bacteria most often associated with construction in hospitals, as construction and demolition often result in collections of stagnant water.

The primary mode of transmission is inhalation of contaminated mist or aerosols. Legionella species can also colonize newly constructed hospital buildings within weeks of installation of water fixtures.

In a large university-affiliated hospital, 2 cases of nosocomial legionellosis were identified during a period of major construction.⁶ An epidemiologic investigation traced the source to a widespread contamination of potable water within the hospital. One patient’s isolate was similar to that of a water sample from the faucet in his room, and an association between Legionnaires disease and construction was postulated.

Another institution’s newly constructed hematology-oncology unit identified 10 cases of Legionnaires disease over a 12-week period in patients and visitors with exposure to the unit during and within the incubation period.⁷ A clinical and environmental assessment found 3 clinical isolates of Legionella identical to environmental isolates found from the unit, strongly implicating the potable water system as the likely source.⁷

In Ohio, 11 cases of hospital-acquired Legionnaires disease were identified in patients moved to a newly constructed 12-story addition to a hospital, and 1 of those died.⁸

Legionella infections appear to be less common than mold infections when reviewing the available literature on patients exposed to hospital construction, renovation, or demolition activities. Yet unlike mold infections, which occur mostly in immunocompromised patients, Legionella also affects people with normal immunity.¹

Sleep deprivation

Noise in hospitals has been linked to sleep disturbances in inpatients. A study using noise dosimeters in a university hospital found a mean continuous noise level of 63.5 dBA (A-weighting of decibels indicates risk of hearing loss) over a 24-hour period, a level more than 2 times higher than the recommended 30 dBA.⁹ The same study also found a significant correlation between sleep disturbance in inpatients and increasing noise levels, in a dose-response manner.

Common sources of noise during construction may include power generators, welding and cutting equipment, and transport of materials. While construction activities themselves have yet to be directly linked to sleep deprivation in patients, construction is inevitably accompanied by noise.

Noise is the most common factor interfering with sleep reported by hospitalized patients. Other effects of noise on patients include a rise in heart rate and blood pressure, increased cholesterol and triglyceride levels, increased use of sedatives, and longer length of stay.^9,10 Although construction is rarely done at night, patients generally take naps during the day, so the noise is disruptive.

Physical injuries

Hospitalized patients rarely suffer injuries related to hospital construction. However, these incidents may be underreported. Few cases of physical injury in patients exposed to construction or renovation in healthcare facilities can be found through a Web search.^11,12

Exacerbation of lung disease

Inhalation of indoor air pollutants exposed during renovation can directly trigger an inflammatory response and cause exacerbation in patients with chronic lung diseases such as asthma and chronic obstructive pulmonary disease. No study has specifically examined the effect of hospital construction or renovation on exacerbation of chronic lung diseases in hospitalized patients. Nevertheless, dust and indoor air pollutants from building renovation have often been reported as agents associated with work-related asthma.¹³

THE MESSAGE

Although the risks to inpatients during hospital construction projects appear minimal, their effect can at times be detrimental, especially to the immunocompromised. Hospitals should adhere to infection control risk assessment protocols during construction events. The small number of outbreaks of construction-related infections can make the diagnosis of nosocomial origin of these infections challenging; a high index of suspicion is needed.

Currently in the United States, there is no standard regarding acceptable levels of airborne mold concentrations, and data to support routine hospital air sampling or validation of available air samplers are inadequate. This remains an area for future research.^14,15

Certain measures have been shown to significantly decrease the risk of mold infections and other nosocomial infections during construction projects, including¹⁶:

• Effective dust control through containment units and barriers
• Consistent use of high-efficiency particulate air filters in hospital units that care for immunocompromised and critically ill patients
• Routine surveillance.

Noise and vibration can be reduced by temporary walls and careful tool selection and scheduling. Similarly, temporary walls and other barriers help protect healthcare employees and patients from the risk of direct physical injury.

Preconstruction risk assessments that address infection control, safety, noise, and air quality are crucial, and the Joint Commission generally requires such assessments. Further, education of hospital staff and members of the construction team about the potential detrimental effects of hospital construction and renovation is essential to secure a safe environment.        

My adult nonacutely ill patient, weighing 70 kg with a glomerular filtration rate (GFR) greater than 60 mL/min/1.73 m², is admitted to the general medical service. She is to receive nothing by mouth for at least the next 24 hours for testing. Do I need to provide maintenance fluids intravenously?

The question seems like it should have an easy answer. However, there is no consensus either on the type of fluids or the need for them at all.

Mortiz and Ayus¹ have described the role of maintenance intravenous (IV) fluids in acutely ill patients and made the case for isotonic saline (0.9% NaCl) to minimize the risk of hyponatremia, while acknowledging that it provides 7 to 10 g of sodium per day.

Recommendations for IV fluids for nonacutely ill hospitalized patients range from isotonic solutions such as 0.9% NaCl and lactated Ringer’s, to hypotonic fluids such as 5% dextrose in water (D5W) in 0.45% NaCl and D5W in 0.2% NaCl.^2–5

The 2013 guidelines of the UK National Institute for Health and Care Excellence (NICE) recommend hypotonic fluids to provide 25 to 30 mL/kg/day of water with 1 mmol/kg/day of sodium. For a 70-kg patient (body surface area 1.7 m²), this would be 1,750 to 2,000 mL of water, with a maximum of 70 mEq/L of sodium (35 mEq/L).⁵ An option would be D5W in 0.2% NaCl, which has 34 mEq/L of sodium.

When choosing maintenance IV fluids, we need to consider the following questions:

• What is my patient’s volume status?
• What is the baseline serum sodium and renal function?
• Are there comorbid conditions that may affect antidiuretic hormone (ADH) status such as physiologic stimulation from volume depletion, drugs, pathologic medical conditions, or syndrome of inappropriate ADH stimulation?
• Will my patient be receiving strictly nothing by mouth?
• Are there unusual fluid losses?

SCENARIO 1: ‘USUAL’ MAINTENANCE

If the patient is euvolemic, with a normal serum osmolality, a GFR more than 60 mL/min/1.73 m², no stimuli for ADH secretion, and no unusual fluid losses, “usual” maintenance would be expected. The usual volume for this patient can be estimated by the following formulas:

• Maintenance volume: 2,550 mL (1,500 mL × 1.7 m² body surface area)
• Holliday-Segar method⁶: 2,500 mL (1,500 mL plus 20 mL/kg for every kilogram over 20 kg).

The usual sodium can be also estimated by the following formulas:

• 2 g Na/day = 2,000 mg/day = 87 mEq/day
• Holliday-Segar⁶: 3 mEq Na/100 mL and 2 mEq K/100 mL of maintenance fluid.

Maintenance IV fluids for our nonacutely ill adult patient could be:

• NICE guideline⁵: D5W in 0.2% NaCl with 20 mEq KCl, to run at 75 mL/hour
• Holliday-Segar method⁶: D5W in 0.2% NaCl with 20 mEq KCl, to run at 100 mL/hour.

Twenty-four hours later, assuming no unusual fluid losses or stimulation of ADH secretion, our patient would weigh the same and would have no significant change in serum osmolality.

OTHER OPTIONS

What if I provide 0.9% NaCl instead?

Each 1 L of normal saline provides 154 mEq of sodium, equivalent to 3.5 g of sodium. Thus, for the 24 hours, with administration of 2 to 2.5 L, the patient would receive a sodium load of 7 to 8.75 g. The consequences of this can be debated, but for 24 hours, more than likely, nothing will happen or be noticeable. The kidneys have a wonderful ability to “dump” excess sodium ingested in the diet, as evidenced by the average Western diet with a sodium load in the range of 4 g per day.^7,8

What if I provide 0.45% NaCl instead?

Each liter provides 50% of the sodium load of 0.9% NaCl. With the 24-hour administration of 2 to 2.5 L of D5W in 0.45% NaCl, the sodium load would be 3.5 to 4.8 g, and the kidneys would dump the excess sodium.

What if I provide ‘catch-up’ fluids after 24 hours, not maintenance fluids?

Assuming only usual losses and no unusual ADH stimulation except for the physiologic stimuli from volume depletion for 24 hours, our patient would lose 2 kg (1 L fluid loss = 1 kg weight loss) and 87 mEq of sodium. This is approximately 4.5% dehydration; thus, other than increased thirst, no physical findings of volume depletion would be clinically evident.

However, serum osmolality and sodium would increase. After 24 hours of nothing by mouth with usual fluid losses, there would be a rise in serum osmolality of 13.5 mOsm/L (a rise in sodium of 6 to 7 mEq/L), which would stimulate ADH in an attempt to minimize further urinary losses. There would be an intracellular volume loss of 1.3 L (Table 1). Clinically, just as with the administration of 0.9% sodium, these changes would not likely be of any clinical consequence in the first 24 hours.

SCENARIO 2: IMPAIRED WATER EXCRETION, AND FLUIDS GIVEN

If the patient is euvolemic but has or is at risk for ADH stimulation,^1,9 providing maintenance IV fluids according to the NICE or Holliday-Segar recommendations (a total of 2 L of 0.2% NaCl = 34 mEq Na/L = 68 mOsm/L) would result in an excess of free water, as an increase in ADH secretion impairs free water clearance. A potential scenario with impaired water excretion is shown in Table 2.

After 24 hours, the patient’s serum osmolality would drop by about 7 mOsm/L, and the serum sodium would decrease by 3 or 4 mEq. The consequence of the intracellular fluid shift would be seen by the expansion of the intracellular volume from 28 to 28.7 L.

If this patient were to have received 2 L of 0.9% NaCl (308 mOsm/L × 2 L = 616 Osm) as suggested by Moritz and Ayus,¹ the result would be a serum osmolality of 284 mOsm/L, thus avoiding hyponatremia and intracellular fluid shifts.

THE BOTTOM LINE

Know your patient, answer the clinical questions noted above, and decide.

For a euvolemic patient with normal serum sodium, GFR greater than 60 mL/1.73 m², and no ADH stimulation, for 24 hours it probably doesn’t matter that much, but a daily reassessment of the continued need for and type of intravenous fluids is critical.

For patients not meeting the criteria noted above such as a patient with systolic or diastolic heart failure, advanced or end-stage renal disease puts the patient at risk for early potential complications of either hyponatremia or sodium overload. For these patients, maintenance intravenous fluids need to be chosen wisely. Daily weights, examinations, and laboratory testing will let you know if something is not right and will allow for early detection and treatment.

My adult nonacutely ill patient, weighing 70 kg with a glomerular filtration rate (GFR) greater than 60 mL/min/1.73 m², is admitted to the general medical service. She is to receive nothing by mouth for at least the next 24 hours for testing. Do I need to provide maintenance fluids intravenously?

The question seems like it should have an easy answer. However, there is no consensus either on the type of fluids or the need for them at all.

Mortiz and Ayus¹ have described the role of maintenance intravenous (IV) fluids in acutely ill patients and made the case for isotonic saline (0.9% NaCl) to minimize the risk of hyponatremia, while acknowledging that it provides 7 to 10 g of sodium per day.

Recommendations for IV fluids for nonacutely ill hospitalized patients range from isotonic solutions such as 0.9% NaCl and lactated Ringer’s, to hypotonic fluids such as 5% dextrose in water (D5W) in 0.45% NaCl and D5W in 0.2% NaCl.^2–5

The 2013 guidelines of the UK National Institute for Health and Care Excellence (NICE) recommend hypotonic fluids to provide 25 to 30 mL/kg/day of water with 1 mmol/kg/day of sodium. For a 70-kg patient (body surface area 1.7 m²), this would be 1,750 to 2,000 mL of water, with a maximum of 70 mEq/L of sodium (35 mEq/L).⁵ An option would be D5W in 0.2% NaCl, which has 34 mEq/L of sodium.

When choosing maintenance IV fluids, we need to consider the following questions:

• What is my patient’s volume status?
• What is the baseline serum sodium and renal function?
• Are there comorbid conditions that may affect antidiuretic hormone (ADH) status such as physiologic stimulation from volume depletion, drugs, pathologic medical conditions, or syndrome of inappropriate ADH stimulation?
• Will my patient be receiving strictly nothing by mouth?
• Are there unusual fluid losses?

SCENARIO 1: ‘USUAL’ MAINTENANCE

If the patient is euvolemic, with a normal serum osmolality, a GFR more than 60 mL/min/1.73 m², no stimuli for ADH secretion, and no unusual fluid losses, “usual” maintenance would be expected. The usual volume for this patient can be estimated by the following formulas:

• Maintenance volume: 2,550 mL (1,500 mL × 1.7 m² body surface area)
• Holliday-Segar method⁶: 2,500 mL (1,500 mL plus 20 mL/kg for every kilogram over 20 kg).

The usual sodium can be also estimated by the following formulas:

• 2 g Na/day = 2,000 mg/day = 87 mEq/day
• Holliday-Segar⁶: 3 mEq Na/100 mL and 2 mEq K/100 mL of maintenance fluid.

Maintenance IV fluids for our nonacutely ill adult patient could be:

• NICE guideline⁵: D5W in 0.2% NaCl with 20 mEq KCl, to run at 75 mL/hour
• Holliday-Segar method⁶: D5W in 0.2% NaCl with 20 mEq KCl, to run at 100 mL/hour.

Twenty-four hours later, assuming no unusual fluid losses or stimulation of ADH secretion, our patient would weigh the same and would have no significant change in serum osmolality.

OTHER OPTIONS

What if I provide 0.9% NaCl instead?

Each 1 L of normal saline provides 154 mEq of sodium, equivalent to 3.5 g of sodium. Thus, for the 24 hours, with administration of 2 to 2.5 L, the patient would receive a sodium load of 7 to 8.75 g. The consequences of this can be debated, but for 24 hours, more than likely, nothing will happen or be noticeable. The kidneys have a wonderful ability to “dump” excess sodium ingested in the diet, as evidenced by the average Western diet with a sodium load in the range of 4 g per day.^7,8

What if I provide 0.45% NaCl instead?

Each liter provides 50% of the sodium load of 0.9% NaCl. With the 24-hour administration of 2 to 2.5 L of D5W in 0.45% NaCl, the sodium load would be 3.5 to 4.8 g, and the kidneys would dump the excess sodium.

What if I provide ‘catch-up’ fluids after 24 hours, not maintenance fluids?

Assuming only usual losses and no unusual ADH stimulation except for the physiologic stimuli from volume depletion for 24 hours, our patient would lose 2 kg (1 L fluid loss = 1 kg weight loss) and 87 mEq of sodium. This is approximately 4.5% dehydration; thus, other than increased thirst, no physical findings of volume depletion would be clinically evident.

However, serum osmolality and sodium would increase. After 24 hours of nothing by mouth with usual fluid losses, there would be a rise in serum osmolality of 13.5 mOsm/L (a rise in sodium of 6 to 7 mEq/L), which would stimulate ADH in an attempt to minimize further urinary losses. There would be an intracellular volume loss of 1.3 L (Table 1). Clinically, just as with the administration of 0.9% sodium, these changes would not likely be of any clinical consequence in the first 24 hours.

SCENARIO 2: IMPAIRED WATER EXCRETION, AND FLUIDS GIVEN

If the patient is euvolemic but has or is at risk for ADH stimulation,^1,9 providing maintenance IV fluids according to the NICE or Holliday-Segar recommendations (a total of 2 L of 0.2% NaCl = 34 mEq Na/L = 68 mOsm/L) would result in an excess of free water, as an increase in ADH secretion impairs free water clearance. A potential scenario with impaired water excretion is shown in Table 2.

After 24 hours, the patient’s serum osmolality would drop by about 7 mOsm/L, and the serum sodium would decrease by 3 or 4 mEq. The consequence of the intracellular fluid shift would be seen by the expansion of the intracellular volume from 28 to 28.7 L.

If this patient were to have received 2 L of 0.9% NaCl (308 mOsm/L × 2 L = 616 Osm) as suggested by Moritz and Ayus,¹ the result would be a serum osmolality of 284 mOsm/L, thus avoiding hyponatremia and intracellular fluid shifts.

THE BOTTOM LINE

Know your patient, answer the clinical questions noted above, and decide.

For a euvolemic patient with normal serum sodium, GFR greater than 60 mL/1.73 m², and no ADH stimulation, for 24 hours it probably doesn’t matter that much, but a daily reassessment of the continued need for and type of intravenous fluids is critical.

For patients not meeting the criteria noted above such as a patient with systolic or diastolic heart failure, advanced or end-stage renal disease puts the patient at risk for early potential complications of either hyponatremia or sodium overload. For these patients, maintenance intravenous fluids need to be chosen wisely. Daily weights, examinations, and laboratory testing will let you know if something is not right and will allow for early detection and treatment.

My adult nonacutely ill patient, weighing 70 kg with a glomerular filtration rate (GFR) greater than 60 mL/min/1.73 m², is admitted to the general medical service. She is to receive nothing by mouth for at least the next 24 hours for testing. Do I need to provide maintenance fluids intravenously?

The question seems like it should have an easy answer. However, there is no consensus either on the type of fluids or the need for them at all.

Mortiz and Ayus¹ have described the role of maintenance intravenous (IV) fluids in acutely ill patients and made the case for isotonic saline (0.9% NaCl) to minimize the risk of hyponatremia, while acknowledging that it provides 7 to 10 g of sodium per day.

Recommendations for IV fluids for nonacutely ill hospitalized patients range from isotonic solutions such as 0.9% NaCl and lactated Ringer’s, to hypotonic fluids such as 5% dextrose in water (D5W) in 0.45% NaCl and D5W in 0.2% NaCl.^2–5

The 2013 guidelines of the UK National Institute for Health and Care Excellence (NICE) recommend hypotonic fluids to provide 25 to 30 mL/kg/day of water with 1 mmol/kg/day of sodium. For a 70-kg patient (body surface area 1.7 m²), this would be 1,750 to 2,000 mL of water, with a maximum of 70 mEq/L of sodium (35 mEq/L).⁵ An option would be D5W in 0.2% NaCl, which has 34 mEq/L of sodium.

When choosing maintenance IV fluids, we need to consider the following questions:

• What is my patient’s volume status?
• What is the baseline serum sodium and renal function?
• Are there comorbid conditions that may affect antidiuretic hormone (ADH) status such as physiologic stimulation from volume depletion, drugs, pathologic medical conditions, or syndrome of inappropriate ADH stimulation?
• Will my patient be receiving strictly nothing by mouth?
• Are there unusual fluid losses?

SCENARIO 1: ‘USUAL’ MAINTENANCE

If the patient is euvolemic, with a normal serum osmolality, a GFR more than 60 mL/min/1.73 m², no stimuli for ADH secretion, and no unusual fluid losses, “usual” maintenance would be expected. The usual volume for this patient can be estimated by the following formulas:

• Maintenance volume: 2,550 mL (1,500 mL × 1.7 m² body surface area)
• Holliday-Segar method⁶: 2,500 mL (1,500 mL plus 20 mL/kg for every kilogram over 20 kg).

The usual sodium can be also estimated by the following formulas:

• 2 g Na/day = 2,000 mg/day = 87 mEq/day
• Holliday-Segar⁶: 3 mEq Na/100 mL and 2 mEq K/100 mL of maintenance fluid.

Maintenance IV fluids for our nonacutely ill adult patient could be:

• NICE guideline⁵: D5W in 0.2% NaCl with 20 mEq KCl, to run at 75 mL/hour
• Holliday-Segar method⁶: D5W in 0.2% NaCl with 20 mEq KCl, to run at 100 mL/hour.

Twenty-four hours later, assuming no unusual fluid losses or stimulation of ADH secretion, our patient would weigh the same and would have no significant change in serum osmolality.

OTHER OPTIONS

What if I provide 0.9% NaCl instead?

Each 1 L of normal saline provides 154 mEq of sodium, equivalent to 3.5 g of sodium. Thus, for the 24 hours, with administration of 2 to 2.5 L, the patient would receive a sodium load of 7 to 8.75 g. The consequences of this can be debated, but for 24 hours, more than likely, nothing will happen or be noticeable. The kidneys have a wonderful ability to “dump” excess sodium ingested in the diet, as evidenced by the average Western diet with a sodium load in the range of 4 g per day.^7,8

What if I provide 0.45% NaCl instead?

Each liter provides 50% of the sodium load of 0.9% NaCl. With the 24-hour administration of 2 to 2.5 L of D5W in 0.45% NaCl, the sodium load would be 3.5 to 4.8 g, and the kidneys would dump the excess sodium.

What if I provide ‘catch-up’ fluids after 24 hours, not maintenance fluids?

Assuming only usual losses and no unusual ADH stimulation except for the physiologic stimuli from volume depletion for 24 hours, our patient would lose 2 kg (1 L fluid loss = 1 kg weight loss) and 87 mEq of sodium. This is approximately 4.5% dehydration; thus, other than increased thirst, no physical findings of volume depletion would be clinically evident.

However, serum osmolality and sodium would increase. After 24 hours of nothing by mouth with usual fluid losses, there would be a rise in serum osmolality of 13.5 mOsm/L (a rise in sodium of 6 to 7 mEq/L), which would stimulate ADH in an attempt to minimize further urinary losses. There would be an intracellular volume loss of 1.3 L (Table 1). Clinically, just as with the administration of 0.9% sodium, these changes would not likely be of any clinical consequence in the first 24 hours.

SCENARIO 2: IMPAIRED WATER EXCRETION, AND FLUIDS GIVEN

If the patient is euvolemic but has or is at risk for ADH stimulation,^1,9 providing maintenance IV fluids according to the NICE or Holliday-Segar recommendations (a total of 2 L of 0.2% NaCl = 34 mEq Na/L = 68 mOsm/L) would result in an excess of free water, as an increase in ADH secretion impairs free water clearance. A potential scenario with impaired water excretion is shown in Table 2.

After 24 hours, the patient’s serum osmolality would drop by about 7 mOsm/L, and the serum sodium would decrease by 3 or 4 mEq. The consequence of the intracellular fluid shift would be seen by the expansion of the intracellular volume from 28 to 28.7 L.

If this patient were to have received 2 L of 0.9% NaCl (308 mOsm/L × 2 L = 616 Osm) as suggested by Moritz and Ayus,¹ the result would be a serum osmolality of 284 mOsm/L, thus avoiding hyponatremia and intracellular fluid shifts.

THE BOTTOM LINE

Know your patient, answer the clinical questions noted above, and decide.

For a euvolemic patient with normal serum sodium, GFR greater than 60 mL/1.73 m², and no ADH stimulation, for 24 hours it probably doesn’t matter that much, but a daily reassessment of the continued need for and type of intravenous fluids is critical.

For patients not meeting the criteria noted above such as a patient with systolic or diastolic heart failure, advanced or end-stage renal disease puts the patient at risk for early potential complications of either hyponatremia or sodium overload. For these patients, maintenance intravenous fluids need to be chosen wisely. Daily weights, examinations, and laboratory testing will let you know if something is not right and will allow for early detection and treatment.

A 66-year-old man presented to the emergency department with increasing shortness of breath and productive cough, which had begun 5 days earlier. Three years previously, he had been diagnosed with chronic obstructive pulmonary disease (COPD).

One week before the current presentation, he developed a sore throat, rhinorrhea, and nasal congestion, and the shortness of breath had started 2 days after that. Although he could speak in sentences, he was breathless even at rest. His dyspnea was associated with noisy breathing and cough productive of yellowish sputum; there was no hemoptysis. He reported fever, but he had no chills, night sweats, chest pain, or paroxysmal nocturnal dyspnea. The review of other systems was unremarkable.

His COPD was known to be mild, in Global Initiative for Chronic Obstructive Lung Disease (GOLD) grade 1, group A. His postbronchodilator ratio of forced expiratory volume in 1 second (FEV₁) to forced vital capacity (FVC) was less than 0.70, and his FEV₁ was 84% of predicted. Apart from mild intermittent cough with white sputum, his COPD had been under good control with inhaled ipratropium 4 times daily and inhaled albuterol as needed. He said he did not have shortness of breath except when hurrying on level ground or walking up a slight hill (Modified Medical Research Council dyspnea scale grade 1; COPD Assessment Test score < 10). In the last 3 years, he had 2 exacerbations of COPD, 1 year apart, both requiring oral prednisone and antibiotic therapy.

Other relevant history included hypertension and dyslipidemia of 15-year duration, for which he was taking candesartan 16 mg twice daily and atorvastatin 20 mg daily. He was compliant with his medications.

Though he usually received an influenza vaccine every year, he did not get it the previous year. Also, 3 years previously, he received the 23-valent pneumococcal polysaccharide vaccine (PPSV23), and the year before that he received the pneumococcal conjugate vaccine (PCV13). In addition, he was immunized against herpes zoster and tetanus.

The patient had smoked 1 pack per day for the past 38 years. His primary care physician had advised him many times to quit smoking. He had enrolled in a smoking cessation program 2 years previously, in which he received varenicline in addition to behavioral counseling in the form of motivational interviewing and a telephone quit-line. Nevertheless, he continued to smoke.

He was a retired engineer. He did not drink alcohol or use illicit drugs.

PHYSICAL EXAMINATION

On physical examination, the patient was sitting up in bed, leaning forward. He was alert and oriented but was breathing rapidly and looked sick. He had no cyanosis, clubbing, pallor, or jaundice. His blood pressure was 145/90 mm Hg, heart rate 110 beats per minute and regular, respiratory rate 29 breaths per minute, and oral temperature 38.1°C (100.6°F). His oxygen saturation was 88% while breathing room air. His body mass index was 27.1 kg/m².

His throat was mildly congested. His neck veins were flat, and there were no carotid bruits. His thyroid examination was normal, without goiter, nodules, or tenderness.

Intercostal retractions were noted around the anterolateral costal margins. He had no chest wall deformities. Chest expansion was reduced bilaterally. There was hyperresonance bilaterally. Expiratory wheezes were heard over both lungs, without crackles.

His heart had no murmurs or added sounds. There was no lower-limb edema or swelling. The rest of his physical examination was unremarkable.

Results of initial laboratory testing are shown in Table 1.

Assessment: A 66-year-old man with GOLD grade 1, group A COPD, presenting with a severe exacerbation, most likely due to viral bronchitis.

The patient was given oxygen 28% by Venturi mask, and his oxygen saturation went up to 90%. He was started on nebulized albuterol 2.5 mg with ipratropium bromide 500 µg every 4 hours, prednisone 40 mg orally daily for 5 days, and ceftriaxone 1 g intravenously every 24 hours. The first dose of each medication was given in the emergency department.

The patient was then admitted to a progressive care unit, where he was placed on noninvasive positive pressure ventilation, continuous cardiac monitoring, and pulse oximetry. He was started on enoxaparin 40 mg subcutaneously daily to prevent venous thromboembolism, and the oral medications he had been taking at home were continued. Because he was receiving a glucocorticoid, his blood glucose was monitored in the fasting state, 2 hours after each meal, and as needed.

Two hours after he started noninvasive positive pressure ventilation, his arterial blood gases were remeasured and showed the following results:

• pH 7.35
• Partial pressure of carbon dioxide (Paco₂) 52 mm Hg
• Bicarbonate 28 mmol/L
• Partial pressure of oxygen (Pao₂) 60 mm Hg
• Oxygen saturation 90%.

HOSPITAL COURSE

On hospital day 3, his dyspnea had slightly improved. His respiratory rate was 26 to 28 breaths per minute. His oxygen saturation remained between 90% and 92%.

At 10:21 pm, his cardiac monitor showed an episode of focal atrial tachycardia at a rate of 129 beats per minute that lasted for 3 minutes and 21 seconds, terminating spontaneously. He denied any change in his clinical condition during the episode, with no chest pain, palpitation, or change in dyspnea. There was no change in his vital signs. He had another similar asymptomatic episode lasting 4 minutes and 9 seconds at 6:30 am of hospital day 4.

Because of these episodes, the attending physician ordered thyroid function tests.

THYROID FUNCTION TESTING

1. Which thyroid function test is most likely to be helpful in the assessment of this patient’s thyroid status?

• Serum thyroid-stimulating hormone (TSH) alone
• Serum TSH and total thyroxine (T₄)
• Serum TSH and total triiodothyronine (T₃)
• Serum TSH and free T₄
• Serum TSH and free T₃

There are several tests to assess thyroid function: the serum TSH, total T₄, free T₄, total T₃, and free T₃ concentrations.¹

In normal physiology, TSH from the pituitary stimulates the thyroid gland to produce and secrete T₄ and T₃, which in turn inhibit TSH secretion through negative feedback. A negative log-linear relation exists between serum free T₄ and TSH levels.² Thus, the serum free T₄ level can remain within the normal reference range even if the TSH level is high or low. 

TSH assays can have different detection limits. A third-generation TSH assay with a detection limit of 0.01 mU/L is recommended for use in clinical practice.³

TSH testing alone. Given its superior sensitivity and specificity, serum TSH measurement is considered the best single test for assessing thyroid function in most cases.⁴ Nevertheless, measurement of the serum TSH level alone could be misleading in several situations, eg, hypothalamic or pituitary disorders, recent treatment of thyrotoxicosis, impaired sensitivity to thyroid hormone, and acute nonthyroidal illness.⁴

Free vs total T₄ and T₃ levels

Serum total T₄ includes a fraction that is bound, mainly to thyroxin-binding globulin, and a very small unbound (free) fraction. The same applies to T₃. Only free thyroid hormones represent the “active” fraction available for interaction with their protein receptors in the nucleus.⁸ Patients with conditions that can affect the thyroid-binding protein concentrations usually have altered serum total T₄ and T₃ levels, whereas their free hormone concentrations remain normal. Accordingly, measurement of free hormone levels, especially free T₄, is usually recommended.

Although equilibrium dialysis is the method most likely to provide an accurate serum free T₄ measurement, it is not commonly used because of its limited availability and high cost. Thus, most commercial laboratories use “direct” free T₄ measurement or, to a lesser degree, the free T₄ index.⁹ However, none of the currently available free T₄ tests actually measure free T₄ directly; rather, they estimate it.¹⁰

Commercial laboratories can provide a direct free T₃ estimate, but it may be less reliable than total T₃. If serum T₃ measurement is indicated, serum total T₃ is usually measured. However, total T₃ measurement is rarely indicated for patients with hypothyroidism because it usually remains within the normal reference range.¹¹ Nevertheless, serum total T₃ measurement could be useful in patients with T₃ toxicosis and in those who are acutely ill.

Accordingly, in acutely ill hospitalized patients like ours, measuring serum TSH using a third-generation assay and free T₄ is essential to assess thyroid function. Many clinicians also measure serum total T₃.

The attending physician ordered serum TSH, free T₄, and total T₃ measurements, which yielded the following:

• TSH 0.1 mU/L (0.5–5.0)
• Total T₃ 55 ng/dL (80–180)
• Free T₄ 0.9 ng/dL (0.9–2.4).

2. Which best explains this patient’s abnormal thyroid test results?

• His acute illness
• Central hypothyroidism due to pituitary infarction
• His albuterol therapy
• Subclinical thyrotoxicosis
• Hashimoto thyroiditis

Since euthyroid patients with an acute illness may have abnormal thyroid test results (Table 2),^5–7 thyroid function testing is not recommended unless there is a strong indication for it, such as new-onset atrial fibrillation, atrial flutter, or focal atrial tachycardia.¹ In such patients, it is important to know whether the test abnormalities represent true thyroid disorder or are the result of a nonthyroidal illness.

Thyroid function testing in patients with nonthyroidal illness usually shows low serum total T₃, normal or low serum TSH, and normal, low, or high serum free T₄. However, transient mild serum TSH elevation can be seen in some patients during the recovery period.¹⁶ These abnormalities with their mechanisms are shown in Table 2.^5–7 In several commercial kits, serum direct free T₄ can be falsely decreased or increased.⁸

THE DIFFERENTIAL DIAGNOSIS

Our patient had low serum TSH, low-normal serum direct free T₄, and low serum total T₃. This profile could be caused by a nonthyroidal illness, “true” central hypothyroidism, or his glucocorticoid treatment. The reason we use the term “true” in this setting is that some experts suggest that the thyroid function test abnormalities in patients with acute nonthyroidal illness represent a transient central hypothyroidism.¹⁷ The clinical presentation is key in differentiating true central hypothyroidism from nonthyroidal illness.

In addition, measuring serum cortisol may help to differentiate between the 2 states, as it would be elevated in patients with nonthyroidal illness as part of a stress response but low in patients with true central hypothyroidism, since it is usually part of combined pituitary hormone deficiency.¹⁸ Of note, some critically ill patients have low serum cortisol because of transient central adrenal insufficiency.^19,20

The serum concentration of reverse T₃ has been suggested as a way to differentiate between hypothyroidism (low) and nonthyroidal illness (high); however, further studies showed that it does not reliably differentiate between the conditions.²¹

GLUCOCORTICOIDS AND THYROID FUNCTION TESTS

By inhibiting D1, glucocorticoids can decrease peripheral conversion of T₄ to T₃ and thus decrease serum total T₃. This effect depends on the type and dose of the glucocorticoid and the duration of therapy.

In one study,²² there was a significant reduction in serum total T₃ concentration 24 hours after a single oral dose of dexamethasone 12 mg in normal participants. This effect lasted 48 hours, after which serum total T₃ returned to its pretreatment level.

In another study,²³ a daily oral dose of betamethasone 1.5 mg for 5 days did not significantly reduce the serum total T₃ in healthy volunteers, but a daily dose of 3 mg did. This effect was more pronounced at a daily dose of 4.5 mg, whereas a dose of 6.0 mg had no further effect.

Long-term glucocorticoid therapy also decreases serum total T₄ and total T₃ by lowering serum thyroid-binding globulin.²⁴

Finally, glucocorticoids can decrease TSH secretion by directly inhibiting thyrotropin-releasing hormone.^25,26 However, chronic hypercortisolism, whether endogenous or exogenous, does not cause clinically central hypothyroidism, possibly because of the negative feedback mechanism of low thyroid hormones on the pituitary and the hypothalamus.²⁷

Other drugs including dopamine, dopamine agonists, dobutamine, and somatostatin analogues can suppress serum TSH. As with glucocorticoids, these drugs do not cause clinically evident central hypothyroidism.²⁸ Bexarotene, a retinoid X receptor ligand used in the treatment of cutaneous T-cell lymphoma, has been reported to cause clinically evident central hypothyroidism by suppressing TSH and increasing T₄ clearance.²⁹

While there is general agreement that beta-adrenergic antagonists (beta-blockers) do not affect the serum TSH concentration, conflicting data have been reported concerning their effect on other thyroid function tests. This may be due to several factors, including dose, duration of therapy, the patient’s thyroid status, and differences in laboratory methodology.³⁰

In studies of propranolol, serum total T₄ concentrations did not change or were increased with daily doses of 160 mg or more in both euthyroid participants and hyperthyroid patients^31–33; serum total T₃ concentrations did not change or were decreased with 40 mg or more daily³⁴; and serum reverse T₃ concentrations were increased with daily doses of 80 mg or more.³¹ It is most likely that propranolol exerts these changes by inhibiting D1 activity in peripheral tissues.

Furthermore, a significant decrease in serum total T₃ concentrations was observed in hyperthyroid patients treated with atenolol 100 mg daily, metoprolol 100 mg daily, and alprenolol 100 mg daily, but not with sotalol 80 mg daily or nadolol (up to 240 mg daily).^35,36

On the other hand, beta-adrenergic agonists have not been reported to cause significant changes in thyroid function tests.³⁷

SUBCLINICAL THYROTOXICOSIS OR HASHIMOTO THYROIDITIS?

Our patient’s thyroid function test results are more likely due to his nonthyroidal illness and glucocorticoid therapy, as there is no clinical evidence to point to a hypothalamic-pituitary disorder accounting for true central hypothyroidism.

The other options mentioned in question 2 are unlikely to explain our patient’s thyroid function test results.

Subclinical thyrotoxicosis is characterized by suppressed serum TSH, but both serum free T₄ and total T₃ remain within the normal reference ranges. In addition, the serum TSH level may help to differentiate between thyrotoxicosis and nonthyroidal illness. In the former, serum TSH is usually suppressed (< 0.01 mU/L), whereas in the latter it is usually low but detectable (0.05– 0.3 mU/L).^38,39

Hashimoto thyroiditis is a chronic autoimmune thyroid disease characterized by diffuse lymphocytic infiltration of the thyroid gland. Almost all patients with Hashimoto thyroiditis have elevated levels of antibodies to thyroid peroxidase or thyroglobulin.⁴⁰ Clinically, patients with Hashimoto thyroiditis can either be hypothyroid or have normal thyroid function, which is not the case in our patient.

CASE CONTINUED

An endocrinologist, consulted for a second opinion, agreed that the patient’s thyroid function test results were most likely due to his nonthyroidal illness and glucocorticoid therapy.

3. In view of the endocrinologist’s opinion, which should be the next step in the management of the patient’s thyroid condition?

• Start levothyroxine (T₄) therapy
• Start liothyronine (T₃) therapy
• Start N-acetylcysteine therapy
• Start thyrotropin-releasing hormone therapy
• Remeasure thyroid hormones after full recovery from his acute illness

It is not clear whether the changes in thyroid hormone levels during an acute illness are a pathologic alteration for which thyroid hormone therapy may be beneficial, or a physiologic adaptation for which such therapy would not be indicated.⁴¹

However, current data argue against thyroid hormone therapy using T₄ or T₃ for patients with nonthyroidal illness syndrome (also called euthyroid sick syndrome).⁴² Indeed, several randomized controlled trials showed that thyroid hormone therapy is not beneficial in such patients and may be detrimental.^41,43

Therapies other than thyroid hormone have been investigated to ameliorate thyroid hormone abnormalities in patients with nonthyroidal illness. These include N-acetylcysteine, thyrotropin-releasing hormone therapy, and nutritional support.

Some studies showed that giving N-acetyl­cysteine, an antioxidant, increased serum T₃ and decreased serum reverse T₃ concentrations in patients with acute myocardial infarction.⁴⁴ Nevertheless, the mortality rate and length of hospitalization were not affected. Further studies are needed to know whether N-acetylcysteine therapy is beneficial for such patients.

Similarly, a study using a thyrotropin-releasing hormone analogue along with growth hormone-releasing peptide 2 showed an increase in serum TSH, T₄, and T₃ levels in critically ill patients.⁴⁵ The benefit of this therapy has yet to be determined. On the other hand, early nutritional support was reported to prevent thyroid hormonal changes in patients postoperatively.⁴⁶

Measuring thyroid hormone levels after full recovery is the most appropriate next step in our patient, as the changes in thyroid hormone concentrations subside as the acute illness resolves.⁴⁷

The patient continued to improve. On hospital day 6, he was feeling better but still had mild respiratory distress. There had been no further episodes of arrhythmia since day 4. His blood pressure was 136/86 mm Hg, heart rate 88 beats per minute and regular, respiratory rate 18 breaths per minute, and oral temperature 37.1°C. His oxygen saturation was 92% on room air.

Before discharge, he was encouraged to quit smoking. He was offered behavioral counseling and medication therapy, but he only said that he would think about it. He was discharged on oral cefixime for 4 more days and was instructed to switch to a long-acting bronchodilator along with his other home medications and to return in 1 week to have his thyroid hormones checked.

One week later, his laboratory results were:

• TSH 11.2 mU/L (reference range 0.5–5.0)
• Free T₄ 1.2 ng/dL (0.9–2.4)
• Total T₃ 92 ng/dL (80–180).

Clinically, the patient was euthyroid, and examination of his thyroid was unremarkable.

4. Based on these last test results, which statement is correct?

• Levothyroxine therapy should be started
• His serum TSH elevation is most likely transient
• Thyroid ultrasonography is strongly indicated
• A radioactive iodine uptake study should be performed
• Measurement of thyroid-stimulating immunoglobulins is indicated

During recovery from nonthyroidal illness, some patients may have elevated serum TSH levels that are usually transient and modest (< 20 mU/L).⁴⁸ Normalization of the thyroid function tests including serum TSH may take several weeks⁴⁹ or months.⁵⁰ However, a systematic review found that the likelihood of permanent primary hypothyroidism is high in patients with serum TSH levels higher than 20 mU/L during the recovery phase of their nonthyroidal illness.⁵¹

Ultrasonography is useful for evaluating patients with thyroid nodules or goiter but is of little benefit for patients like ours, in whom the thyroid is normal on examination.

Similarly, a radioactive iodine uptake study is not indicated, as it is principally used to help differentiate between types of thyrotoxicosis. (Radioactive iodine is also used to treat differentiated thyroid cancer.)

Thyroid-stimulating immunoglobins are TSH receptor-stimulating antibodies that cause Graves disease. Nevertheless, measuring them is not routinely indicated for its diagnosis. However, their measurement is of significant help in the diagnosis of Graves disease if a radioactive iodine uptake study cannot be performed (as in pregnancy) and in atypical presentations such as euthyroid Graves ophthalmopathy.⁵² Other indications for thyroid-stimulating immunoglobin measurement are beyond the scope of the article. Our patient’s test results are not consistent with hyperthyroidism, so measuring thyroid-stimulating immunoglobins is not indicated.

CASE CONCLUSION: BETTER, BUT STILL SMOKING

The patient missed his 1-month clinic follow-up, but he visited the clinic for follow-up 3 months later. He was feeling well with no complaints. Test results including serum TSH, free T₄, and total T₃ were within normal ranges. His COPD was under control, with an FEV₁ 88% of predicted.

He was again encouraged to quit smoking and was offered drug therapy and behavioral counseling, but he declined. In addition, he was instructed to adhere to his annual influenza vaccination.

KEY POINTS

• In patients with acute illness, it is recommended that thyroid function not be assessed unless there is a strong indication.
• If thyroid function assessment is indicated for critically ill patients, serum TSH and free T₄ concentrations should be measured. Some clinicians also measure serum total T₃ level.
• Thyroid function testing in critically ill patients usually shows low serum total T₃, normal or low serum TSH, and normal or low serum free T₄.
• Many drugs can alter thyroid hormone levels.
• Thyroid hormone therapy is not recommended for critically ill patients with low T₃, low T₄, or both.
• During recovery from nonthyroidal illness, some patients may have mild elevation in serum TSH levels (< 20 mU/L).
• Thyroid hormone levels may take several weeks or months to return to normal after the acute illness.
• Patients with serum TSH levels higher than 20 mU/L during the recovery phase of their nonthyroidal illness are more likely to have permanent primary hypothyroidism.

A 66-year-old man presented to the emergency department with increasing shortness of breath and productive cough, which had begun 5 days earlier. Three years previously, he had been diagnosed with chronic obstructive pulmonary disease (COPD).

One week before the current presentation, he developed a sore throat, rhinorrhea, and nasal congestion, and the shortness of breath had started 2 days after that. Although he could speak in sentences, he was breathless even at rest. His dyspnea was associated with noisy breathing and cough productive of yellowish sputum; there was no hemoptysis. He reported fever, but he had no chills, night sweats, chest pain, or paroxysmal nocturnal dyspnea. The review of other systems was unremarkable.

His COPD was known to be mild, in Global Initiative for Chronic Obstructive Lung Disease (GOLD) grade 1, group A. His postbronchodilator ratio of forced expiratory volume in 1 second (FEV₁) to forced vital capacity (FVC) was less than 0.70, and his FEV₁ was 84% of predicted. Apart from mild intermittent cough with white sputum, his COPD had been under good control with inhaled ipratropium 4 times daily and inhaled albuterol as needed. He said he did not have shortness of breath except when hurrying on level ground or walking up a slight hill (Modified Medical Research Council dyspnea scale grade 1; COPD Assessment Test score < 10). In the last 3 years, he had 2 exacerbations of COPD, 1 year apart, both requiring oral prednisone and antibiotic therapy.

Other relevant history included hypertension and dyslipidemia of 15-year duration, for which he was taking candesartan 16 mg twice daily and atorvastatin 20 mg daily. He was compliant with his medications.

Though he usually received an influenza vaccine every year, he did not get it the previous year. Also, 3 years previously, he received the 23-valent pneumococcal polysaccharide vaccine (PPSV23), and the year before that he received the pneumococcal conjugate vaccine (PCV13). In addition, he was immunized against herpes zoster and tetanus.

The patient had smoked 1 pack per day for the past 38 years. His primary care physician had advised him many times to quit smoking. He had enrolled in a smoking cessation program 2 years previously, in which he received varenicline in addition to behavioral counseling in the form of motivational interviewing and a telephone quit-line. Nevertheless, he continued to smoke.

He was a retired engineer. He did not drink alcohol or use illicit drugs.

PHYSICAL EXAMINATION

On physical examination, the patient was sitting up in bed, leaning forward. He was alert and oriented but was breathing rapidly and looked sick. He had no cyanosis, clubbing, pallor, or jaundice. His blood pressure was 145/90 mm Hg, heart rate 110 beats per minute and regular, respiratory rate 29 breaths per minute, and oral temperature 38.1°C (100.6°F). His oxygen saturation was 88% while breathing room air. His body mass index was 27.1 kg/m².

His throat was mildly congested. His neck veins were flat, and there were no carotid bruits. His thyroid examination was normal, without goiter, nodules, or tenderness.

Intercostal retractions were noted around the anterolateral costal margins. He had no chest wall deformities. Chest expansion was reduced bilaterally. There was hyperresonance bilaterally. Expiratory wheezes were heard over both lungs, without crackles.

His heart had no murmurs or added sounds. There was no lower-limb edema or swelling. The rest of his physical examination was unremarkable.

Results of initial laboratory testing are shown in Table 1.

Assessment: A 66-year-old man with GOLD grade 1, group A COPD, presenting with a severe exacerbation, most likely due to viral bronchitis.

The patient was given oxygen 28% by Venturi mask, and his oxygen saturation went up to 90%. He was started on nebulized albuterol 2.5 mg with ipratropium bromide 500 µg every 4 hours, prednisone 40 mg orally daily for 5 days, and ceftriaxone 1 g intravenously every 24 hours. The first dose of each medication was given in the emergency department.

The patient was then admitted to a progressive care unit, where he was placed on noninvasive positive pressure ventilation, continuous cardiac monitoring, and pulse oximetry. He was started on enoxaparin 40 mg subcutaneously daily to prevent venous thromboembolism, and the oral medications he had been taking at home were continued. Because he was receiving a glucocorticoid, his blood glucose was monitored in the fasting state, 2 hours after each meal, and as needed.

Two hours after he started noninvasive positive pressure ventilation, his arterial blood gases were remeasured and showed the following results:

• pH 7.35
• Partial pressure of carbon dioxide (Paco₂) 52 mm Hg
• Bicarbonate 28 mmol/L
• Partial pressure of oxygen (Pao₂) 60 mm Hg
• Oxygen saturation 90%.

HOSPITAL COURSE

On hospital day 3, his dyspnea had slightly improved. His respiratory rate was 26 to 28 breaths per minute. His oxygen saturation remained between 90% and 92%.

At 10:21 pm, his cardiac monitor showed an episode of focal atrial tachycardia at a rate of 129 beats per minute that lasted for 3 minutes and 21 seconds, terminating spontaneously. He denied any change in his clinical condition during the episode, with no chest pain, palpitation, or change in dyspnea. There was no change in his vital signs. He had another similar asymptomatic episode lasting 4 minutes and 9 seconds at 6:30 am of hospital day 4.

Because of these episodes, the attending physician ordered thyroid function tests.

THYROID FUNCTION TESTING

1. Which thyroid function test is most likely to be helpful in the assessment of this patient’s thyroid status?

• Serum thyroid-stimulating hormone (TSH) alone
• Serum TSH and total thyroxine (T₄)
• Serum TSH and total triiodothyronine (T₃)
• Serum TSH and free T₄
• Serum TSH and free T₃

There are several tests to assess thyroid function: the serum TSH, total T₄, free T₄, total T₃, and free T₃ concentrations.¹

In normal physiology, TSH from the pituitary stimulates the thyroid gland to produce and secrete T₄ and T₃, which in turn inhibit TSH secretion through negative feedback. A negative log-linear relation exists between serum free T₄ and TSH levels.² Thus, the serum free T₄ level can remain within the normal reference range even if the TSH level is high or low. 

TSH assays can have different detection limits. A third-generation TSH assay with a detection limit of 0.01 mU/L is recommended for use in clinical practice.³

TSH testing alone. Given its superior sensitivity and specificity, serum TSH measurement is considered the best single test for assessing thyroid function in most cases.⁴ Nevertheless, measurement of the serum TSH level alone could be misleading in several situations, eg, hypothalamic or pituitary disorders, recent treatment of thyrotoxicosis, impaired sensitivity to thyroid hormone, and acute nonthyroidal illness.⁴

Free vs total T₄ and T₃ levels

Serum total T₄ includes a fraction that is bound, mainly to thyroxin-binding globulin, and a very small unbound (free) fraction. The same applies to T₃. Only free thyroid hormones represent the “active” fraction available for interaction with their protein receptors in the nucleus.⁸ Patients with conditions that can affect the thyroid-binding protein concentrations usually have altered serum total T₄ and T₃ levels, whereas their free hormone concentrations remain normal. Accordingly, measurement of free hormone levels, especially free T₄, is usually recommended.

Although equilibrium dialysis is the method most likely to provide an accurate serum free T₄ measurement, it is not commonly used because of its limited availability and high cost. Thus, most commercial laboratories use “direct” free T₄ measurement or, to a lesser degree, the free T₄ index.⁹ However, none of the currently available free T₄ tests actually measure free T₄ directly; rather, they estimate it.¹⁰

Commercial laboratories can provide a direct free T₃ estimate, but it may be less reliable than total T₃. If serum T₃ measurement is indicated, serum total T₃ is usually measured. However, total T₃ measurement is rarely indicated for patients with hypothyroidism because it usually remains within the normal reference range.¹¹ Nevertheless, serum total T₃ measurement could be useful in patients with T₃ toxicosis and in those who are acutely ill.

Accordingly, in acutely ill hospitalized patients like ours, measuring serum TSH using a third-generation assay and free T₄ is essential to assess thyroid function. Many clinicians also measure serum total T₃.

The attending physician ordered serum TSH, free T₄, and total T₃ measurements, which yielded the following:

• TSH 0.1 mU/L (0.5–5.0)
• Total T₃ 55 ng/dL (80–180)
• Free T₄ 0.9 ng/dL (0.9–2.4).

2. Which best explains this patient’s abnormal thyroid test results?

• His acute illness
• Central hypothyroidism due to pituitary infarction
• His albuterol therapy
• Subclinical thyrotoxicosis
• Hashimoto thyroiditis

Since euthyroid patients with an acute illness may have abnormal thyroid test results (Table 2),^5–7 thyroid function testing is not recommended unless there is a strong indication for it, such as new-onset atrial fibrillation, atrial flutter, or focal atrial tachycardia.¹ In such patients, it is important to know whether the test abnormalities represent true thyroid disorder or are the result of a nonthyroidal illness.

Thyroid function testing in patients with nonthyroidal illness usually shows low serum total T₃, normal or low serum TSH, and normal, low, or high serum free T₄. However, transient mild serum TSH elevation can be seen in some patients during the recovery period.¹⁶ These abnormalities with their mechanisms are shown in Table 2.^5–7 In several commercial kits, serum direct free T₄ can be falsely decreased or increased.⁸

THE DIFFERENTIAL DIAGNOSIS

Our patient had low serum TSH, low-normal serum direct free T₄, and low serum total T₃. This profile could be caused by a nonthyroidal illness, “true” central hypothyroidism, or his glucocorticoid treatment. The reason we use the term “true” in this setting is that some experts suggest that the thyroid function test abnormalities in patients with acute nonthyroidal illness represent a transient central hypothyroidism.¹⁷ The clinical presentation is key in differentiating true central hypothyroidism from nonthyroidal illness.

In addition, measuring serum cortisol may help to differentiate between the 2 states, as it would be elevated in patients with nonthyroidal illness as part of a stress response but low in patients with true central hypothyroidism, since it is usually part of combined pituitary hormone deficiency.¹⁸ Of note, some critically ill patients have low serum cortisol because of transient central adrenal insufficiency.^19,20

The serum concentration of reverse T₃ has been suggested as a way to differentiate between hypothyroidism (low) and nonthyroidal illness (high); however, further studies showed that it does not reliably differentiate between the conditions.²¹

GLUCOCORTICOIDS AND THYROID FUNCTION TESTS

By inhibiting D1, glucocorticoids can decrease peripheral conversion of T₄ to T₃ and thus decrease serum total T₃. This effect depends on the type and dose of the glucocorticoid and the duration of therapy.

In one study,²² there was a significant reduction in serum total T₃ concentration 24 hours after a single oral dose of dexamethasone 12 mg in normal participants. This effect lasted 48 hours, after which serum total T₃ returned to its pretreatment level.

In another study,²³ a daily oral dose of betamethasone 1.5 mg for 5 days did not significantly reduce the serum total T₃ in healthy volunteers, but a daily dose of 3 mg did. This effect was more pronounced at a daily dose of 4.5 mg, whereas a dose of 6.0 mg had no further effect.

Long-term glucocorticoid therapy also decreases serum total T₄ and total T₃ by lowering serum thyroid-binding globulin.²⁴

Finally, glucocorticoids can decrease TSH secretion by directly inhibiting thyrotropin-releasing hormone.^25,26 However, chronic hypercortisolism, whether endogenous or exogenous, does not cause clinically central hypothyroidism, possibly because of the negative feedback mechanism of low thyroid hormones on the pituitary and the hypothalamus.²⁷

Other drugs including dopamine, dopamine agonists, dobutamine, and somatostatin analogues can suppress serum TSH. As with glucocorticoids, these drugs do not cause clinically evident central hypothyroidism.²⁸ Bexarotene, a retinoid X receptor ligand used in the treatment of cutaneous T-cell lymphoma, has been reported to cause clinically evident central hypothyroidism by suppressing TSH and increasing T₄ clearance.²⁹

While there is general agreement that beta-adrenergic antagonists (beta-blockers) do not affect the serum TSH concentration, conflicting data have been reported concerning their effect on other thyroid function tests. This may be due to several factors, including dose, duration of therapy, the patient’s thyroid status, and differences in laboratory methodology.³⁰

In studies of propranolol, serum total T₄ concentrations did not change or were increased with daily doses of 160 mg or more in both euthyroid participants and hyperthyroid patients^31–33; serum total T₃ concentrations did not change or were decreased with 40 mg or more daily³⁴; and serum reverse T₃ concentrations were increased with daily doses of 80 mg or more.³¹ It is most likely that propranolol exerts these changes by inhibiting D1 activity in peripheral tissues.

Furthermore, a significant decrease in serum total T₃ concentrations was observed in hyperthyroid patients treated with atenolol 100 mg daily, metoprolol 100 mg daily, and alprenolol 100 mg daily, but not with sotalol 80 mg daily or nadolol (up to 240 mg daily).^35,36

On the other hand, beta-adrenergic agonists have not been reported to cause significant changes in thyroid function tests.³⁷

SUBCLINICAL THYROTOXICOSIS OR HASHIMOTO THYROIDITIS?

Our patient’s thyroid function test results are more likely due to his nonthyroidal illness and glucocorticoid therapy, as there is no clinical evidence to point to a hypothalamic-pituitary disorder accounting for true central hypothyroidism.

The other options mentioned in question 2 are unlikely to explain our patient’s thyroid function test results.

Subclinical thyrotoxicosis is characterized by suppressed serum TSH, but both serum free T₄ and total T₃ remain within the normal reference ranges. In addition, the serum TSH level may help to differentiate between thyrotoxicosis and nonthyroidal illness. In the former, serum TSH is usually suppressed (< 0.01 mU/L), whereas in the latter it is usually low but detectable (0.05– 0.3 mU/L).^38,39

Hashimoto thyroiditis is a chronic autoimmune thyroid disease characterized by diffuse lymphocytic infiltration of the thyroid gland. Almost all patients with Hashimoto thyroiditis have elevated levels of antibodies to thyroid peroxidase or thyroglobulin.⁴⁰ Clinically, patients with Hashimoto thyroiditis can either be hypothyroid or have normal thyroid function, which is not the case in our patient.

CASE CONTINUED

An endocrinologist, consulted for a second opinion, agreed that the patient’s thyroid function test results were most likely due to his nonthyroidal illness and glucocorticoid therapy.

3. In view of the endocrinologist’s opinion, which should be the next step in the management of the patient’s thyroid condition?

• Start levothyroxine (T₄) therapy
• Start liothyronine (T₃) therapy
• Start N-acetylcysteine therapy
• Start thyrotropin-releasing hormone therapy
• Remeasure thyroid hormones after full recovery from his acute illness

It is not clear whether the changes in thyroid hormone levels during an acute illness are a pathologic alteration for which thyroid hormone therapy may be beneficial, or a physiologic adaptation for which such therapy would not be indicated.⁴¹

However, current data argue against thyroid hormone therapy using T₄ or T₃ for patients with nonthyroidal illness syndrome (also called euthyroid sick syndrome).⁴² Indeed, several randomized controlled trials showed that thyroid hormone therapy is not beneficial in such patients and may be detrimental.^41,43

Therapies other than thyroid hormone have been investigated to ameliorate thyroid hormone abnormalities in patients with nonthyroidal illness. These include N-acetylcysteine, thyrotropin-releasing hormone therapy, and nutritional support.

Some studies showed that giving N-acetyl­cysteine, an antioxidant, increased serum T₃ and decreased serum reverse T₃ concentrations in patients with acute myocardial infarction.⁴⁴ Nevertheless, the mortality rate and length of hospitalization were not affected. Further studies are needed to know whether N-acetylcysteine therapy is beneficial for such patients.

Similarly, a study using a thyrotropin-releasing hormone analogue along with growth hormone-releasing peptide 2 showed an increase in serum TSH, T₄, and T₃ levels in critically ill patients.⁴⁵ The benefit of this therapy has yet to be determined. On the other hand, early nutritional support was reported to prevent thyroid hormonal changes in patients postoperatively.⁴⁶

Measuring thyroid hormone levels after full recovery is the most appropriate next step in our patient, as the changes in thyroid hormone concentrations subside as the acute illness resolves.⁴⁷

The patient continued to improve. On hospital day 6, he was feeling better but still had mild respiratory distress. There had been no further episodes of arrhythmia since day 4. His blood pressure was 136/86 mm Hg, heart rate 88 beats per minute and regular, respiratory rate 18 breaths per minute, and oral temperature 37.1°C. His oxygen saturation was 92% on room air.

Before discharge, he was encouraged to quit smoking. He was offered behavioral counseling and medication therapy, but he only said that he would think about it. He was discharged on oral cefixime for 4 more days and was instructed to switch to a long-acting bronchodilator along with his other home medications and to return in 1 week to have his thyroid hormones checked.

One week later, his laboratory results were:

• TSH 11.2 mU/L (reference range 0.5–5.0)
• Free T₄ 1.2 ng/dL (0.9–2.4)
• Total T₃ 92 ng/dL (80–180).

Clinically, the patient was euthyroid, and examination of his thyroid was unremarkable.

4. Based on these last test results, which statement is correct?

• Levothyroxine therapy should be started
• His serum TSH elevation is most likely transient
• Thyroid ultrasonography is strongly indicated
• A radioactive iodine uptake study should be performed
• Measurement of thyroid-stimulating immunoglobulins is indicated

During recovery from nonthyroidal illness, some patients may have elevated serum TSH levels that are usually transient and modest (< 20 mU/L).⁴⁸ Normalization of the thyroid function tests including serum TSH may take several weeks⁴⁹ or months.⁵⁰ However, a systematic review found that the likelihood of permanent primary hypothyroidism is high in patients with serum TSH levels higher than 20 mU/L during the recovery phase of their nonthyroidal illness.⁵¹

Ultrasonography is useful for evaluating patients with thyroid nodules or goiter but is of little benefit for patients like ours, in whom the thyroid is normal on examination.

Similarly, a radioactive iodine uptake study is not indicated, as it is principally used to help differentiate between types of thyrotoxicosis. (Radioactive iodine is also used to treat differentiated thyroid cancer.)

Thyroid-stimulating immunoglobins are TSH receptor-stimulating antibodies that cause Graves disease. Nevertheless, measuring them is not routinely indicated for its diagnosis. However, their measurement is of significant help in the diagnosis of Graves disease if a radioactive iodine uptake study cannot be performed (as in pregnancy) and in atypical presentations such as euthyroid Graves ophthalmopathy.⁵² Other indications for thyroid-stimulating immunoglobin measurement are beyond the scope of the article. Our patient’s test results are not consistent with hyperthyroidism, so measuring thyroid-stimulating immunoglobins is not indicated.

CASE CONCLUSION: BETTER, BUT STILL SMOKING

The patient missed his 1-month clinic follow-up, but he visited the clinic for follow-up 3 months later. He was feeling well with no complaints. Test results including serum TSH, free T₄, and total T₃ were within normal ranges. His COPD was under control, with an FEV₁ 88% of predicted.

He was again encouraged to quit smoking and was offered drug therapy and behavioral counseling, but he declined. In addition, he was instructed to adhere to his annual influenza vaccination.

KEY POINTS

• In patients with acute illness, it is recommended that thyroid function not be assessed unless there is a strong indication.
• If thyroid function assessment is indicated for critically ill patients, serum TSH and free T₄ concentrations should be measured. Some clinicians also measure serum total T₃ level.
• Thyroid function testing in critically ill patients usually shows low serum total T₃, normal or low serum TSH, and normal or low serum free T₄.
• Many drugs can alter thyroid hormone levels.
• Thyroid hormone therapy is not recommended for critically ill patients with low T₃, low T₄, or both.
• During recovery from nonthyroidal illness, some patients may have mild elevation in serum TSH levels (< 20 mU/L).
• Thyroid hormone levels may take several weeks or months to return to normal after the acute illness.
• Patients with serum TSH levels higher than 20 mU/L during the recovery phase of their nonthyroidal illness are more likely to have permanent primary hypothyroidism.

A 66-year-old man presented to the emergency department with increasing shortness of breath and productive cough, which had begun 5 days earlier. Three years previously, he had been diagnosed with chronic obstructive pulmonary disease (COPD).

One week before the current presentation, he developed a sore throat, rhinorrhea, and nasal congestion, and the shortness of breath had started 2 days after that. Although he could speak in sentences, he was breathless even at rest. His dyspnea was associated with noisy breathing and cough productive of yellowish sputum; there was no hemoptysis. He reported fever, but he had no chills, night sweats, chest pain, or paroxysmal nocturnal dyspnea. The review of other systems was unremarkable.

His COPD was known to be mild, in Global Initiative for Chronic Obstructive Lung Disease (GOLD) grade 1, group A. His postbronchodilator ratio of forced expiratory volume in 1 second (FEV₁) to forced vital capacity (FVC) was less than 0.70, and his FEV₁ was 84% of predicted. Apart from mild intermittent cough with white sputum, his COPD had been under good control with inhaled ipratropium 4 times daily and inhaled albuterol as needed. He said he did not have shortness of breath except when hurrying on level ground or walking up a slight hill (Modified Medical Research Council dyspnea scale grade 1; COPD Assessment Test score < 10). In the last 3 years, he had 2 exacerbations of COPD, 1 year apart, both requiring oral prednisone and antibiotic therapy.

Other relevant history included hypertension and dyslipidemia of 15-year duration, for which he was taking candesartan 16 mg twice daily and atorvastatin 20 mg daily. He was compliant with his medications.

Though he usually received an influenza vaccine every year, he did not get it the previous year. Also, 3 years previously, he received the 23-valent pneumococcal polysaccharide vaccine (PPSV23), and the year before that he received the pneumococcal conjugate vaccine (PCV13). In addition, he was immunized against herpes zoster and tetanus.

The patient had smoked 1 pack per day for the past 38 years. His primary care physician had advised him many times to quit smoking. He had enrolled in a smoking cessation program 2 years previously, in which he received varenicline in addition to behavioral counseling in the form of motivational interviewing and a telephone quit-line. Nevertheless, he continued to smoke.

He was a retired engineer. He did not drink alcohol or use illicit drugs.

PHYSICAL EXAMINATION

On physical examination, the patient was sitting up in bed, leaning forward. He was alert and oriented but was breathing rapidly and looked sick. He had no cyanosis, clubbing, pallor, or jaundice. His blood pressure was 145/90 mm Hg, heart rate 110 beats per minute and regular, respiratory rate 29 breaths per minute, and oral temperature 38.1°C (100.6°F). His oxygen saturation was 88% while breathing room air. His body mass index was 27.1 kg/m².

His throat was mildly congested. His neck veins were flat, and there were no carotid bruits. His thyroid examination was normal, without goiter, nodules, or tenderness.

Intercostal retractions were noted around the anterolateral costal margins. He had no chest wall deformities. Chest expansion was reduced bilaterally. There was hyperresonance bilaterally. Expiratory wheezes were heard over both lungs, without crackles.

His heart had no murmurs or added sounds. There was no lower-limb edema or swelling. The rest of his physical examination was unremarkable.

Results of initial laboratory testing are shown in Table 1.

Assessment: A 66-year-old man with GOLD grade 1, group A COPD, presenting with a severe exacerbation, most likely due to viral bronchitis.

The patient was given oxygen 28% by Venturi mask, and his oxygen saturation went up to 90%. He was started on nebulized albuterol 2.5 mg with ipratropium bromide 500 µg every 4 hours, prednisone 40 mg orally daily for 5 days, and ceftriaxone 1 g intravenously every 24 hours. The first dose of each medication was given in the emergency department.

The patient was then admitted to a progressive care unit, where he was placed on noninvasive positive pressure ventilation, continuous cardiac monitoring, and pulse oximetry. He was started on enoxaparin 40 mg subcutaneously daily to prevent venous thromboembolism, and the oral medications he had been taking at home were continued. Because he was receiving a glucocorticoid, his blood glucose was monitored in the fasting state, 2 hours after each meal, and as needed.

Two hours after he started noninvasive positive pressure ventilation, his arterial blood gases were remeasured and showed the following results:

• pH 7.35
• Partial pressure of carbon dioxide (Paco₂) 52 mm Hg
• Bicarbonate 28 mmol/L
• Partial pressure of oxygen (Pao₂) 60 mm Hg
• Oxygen saturation 90%.

HOSPITAL COURSE

On hospital day 3, his dyspnea had slightly improved. His respiratory rate was 26 to 28 breaths per minute. His oxygen saturation remained between 90% and 92%.

At 10:21 pm, his cardiac monitor showed an episode of focal atrial tachycardia at a rate of 129 beats per minute that lasted for 3 minutes and 21 seconds, terminating spontaneously. He denied any change in his clinical condition during the episode, with no chest pain, palpitation, or change in dyspnea. There was no change in his vital signs. He had another similar asymptomatic episode lasting 4 minutes and 9 seconds at 6:30 am of hospital day 4.

Because of these episodes, the attending physician ordered thyroid function tests.

THYROID FUNCTION TESTING

1. Which thyroid function test is most likely to be helpful in the assessment of this patient’s thyroid status?

• Serum thyroid-stimulating hormone (TSH) alone
• Serum TSH and total thyroxine (T₄)
• Serum TSH and total triiodothyronine (T₃)
• Serum TSH and free T₄
• Serum TSH and free T₃

There are several tests to assess thyroid function: the serum TSH, total T₄, free T₄, total T₃, and free T₃ concentrations.¹

In normal physiology, TSH from the pituitary stimulates the thyroid gland to produce and secrete T₄ and T₃, which in turn inhibit TSH secretion through negative feedback. A negative log-linear relation exists between serum free T₄ and TSH levels.² Thus, the serum free T₄ level can remain within the normal reference range even if the TSH level is high or low. 

TSH assays can have different detection limits. A third-generation TSH assay with a detection limit of 0.01 mU/L is recommended for use in clinical practice.³

TSH testing alone. Given its superior sensitivity and specificity, serum TSH measurement is considered the best single test for assessing thyroid function in most cases.⁴ Nevertheless, measurement of the serum TSH level alone could be misleading in several situations, eg, hypothalamic or pituitary disorders, recent treatment of thyrotoxicosis, impaired sensitivity to thyroid hormone, and acute nonthyroidal illness.⁴

Free vs total T₄ and T₃ levels

Serum total T₄ includes a fraction that is bound, mainly to thyroxin-binding globulin, and a very small unbound (free) fraction. The same applies to T₃. Only free thyroid hormones represent the “active” fraction available for interaction with their protein receptors in the nucleus.⁸ Patients with conditions that can affect the thyroid-binding protein concentrations usually have altered serum total T₄ and T₃ levels, whereas their free hormone concentrations remain normal. Accordingly, measurement of free hormone levels, especially free T₄, is usually recommended.

Although equilibrium dialysis is the method most likely to provide an accurate serum free T₄ measurement, it is not commonly used because of its limited availability and high cost. Thus, most commercial laboratories use “direct” free T₄ measurement or, to a lesser degree, the free T₄ index.⁹ However, none of the currently available free T₄ tests actually measure free T₄ directly; rather, they estimate it.¹⁰

Commercial laboratories can provide a direct free T₃ estimate, but it may be less reliable than total T₃. If serum T₃ measurement is indicated, serum total T₃ is usually measured. However, total T₃ measurement is rarely indicated for patients with hypothyroidism because it usually remains within the normal reference range.¹¹ Nevertheless, serum total T₃ measurement could be useful in patients with T₃ toxicosis and in those who are acutely ill.

Accordingly, in acutely ill hospitalized patients like ours, measuring serum TSH using a third-generation assay and free T₄ is essential to assess thyroid function. Many clinicians also measure serum total T₃.

The attending physician ordered serum TSH, free T₄, and total T₃ measurements, which yielded the following:

• TSH 0.1 mU/L (0.5–5.0)
• Total T₃ 55 ng/dL (80–180)
• Free T₄ 0.9 ng/dL (0.9–2.4).

2. Which best explains this patient’s abnormal thyroid test results?

• His acute illness
• Central hypothyroidism due to pituitary infarction
• His albuterol therapy
• Subclinical thyrotoxicosis
• Hashimoto thyroiditis

Since euthyroid patients with an acute illness may have abnormal thyroid test results (Table 2),^5–7 thyroid function testing is not recommended unless there is a strong indication for it, such as new-onset atrial fibrillation, atrial flutter, or focal atrial tachycardia.¹ In such patients, it is important to know whether the test abnormalities represent true thyroid disorder or are the result of a nonthyroidal illness.

Thyroid function testing in patients with nonthyroidal illness usually shows low serum total T₃, normal or low serum TSH, and normal, low, or high serum free T₄. However, transient mild serum TSH elevation can be seen in some patients during the recovery period.¹⁶ These abnormalities with their mechanisms are shown in Table 2.^5–7 In several commercial kits, serum direct free T₄ can be falsely decreased or increased.⁸

THE DIFFERENTIAL DIAGNOSIS

Our patient had low serum TSH, low-normal serum direct free T₄, and low serum total T₃. This profile could be caused by a nonthyroidal illness, “true” central hypothyroidism, or his glucocorticoid treatment. The reason we use the term “true” in this setting is that some experts suggest that the thyroid function test abnormalities in patients with acute nonthyroidal illness represent a transient central hypothyroidism.¹⁷ The clinical presentation is key in differentiating true central hypothyroidism from nonthyroidal illness.

In addition, measuring serum cortisol may help to differentiate between the 2 states, as it would be elevated in patients with nonthyroidal illness as part of a stress response but low in patients with true central hypothyroidism, since it is usually part of combined pituitary hormone deficiency.¹⁸ Of note, some critically ill patients have low serum cortisol because of transient central adrenal insufficiency.^19,20

The serum concentration of reverse T₃ has been suggested as a way to differentiate between hypothyroidism (low) and nonthyroidal illness (high); however, further studies showed that it does not reliably differentiate between the conditions.²¹

GLUCOCORTICOIDS AND THYROID FUNCTION TESTS

By inhibiting D1, glucocorticoids can decrease peripheral conversion of T₄ to T₃ and thus decrease serum total T₃. This effect depends on the type and dose of the glucocorticoid and the duration of therapy.

In one study,²² there was a significant reduction in serum total T₃ concentration 24 hours after a single oral dose of dexamethasone 12 mg in normal participants. This effect lasted 48 hours, after which serum total T₃ returned to its pretreatment level.

In another study,²³ a daily oral dose of betamethasone 1.5 mg for 5 days did not significantly reduce the serum total T₃ in healthy volunteers, but a daily dose of 3 mg did. This effect was more pronounced at a daily dose of 4.5 mg, whereas a dose of 6.0 mg had no further effect.

Long-term glucocorticoid therapy also decreases serum total T₄ and total T₃ by lowering serum thyroid-binding globulin.²⁴

Finally, glucocorticoids can decrease TSH secretion by directly inhibiting thyrotropin-releasing hormone.^25,26 However, chronic hypercortisolism, whether endogenous or exogenous, does not cause clinically central hypothyroidism, possibly because of the negative feedback mechanism of low thyroid hormones on the pituitary and the hypothalamus.²⁷

Other drugs including dopamine, dopamine agonists, dobutamine, and somatostatin analogues can suppress serum TSH. As with glucocorticoids, these drugs do not cause clinically evident central hypothyroidism.²⁸ Bexarotene, a retinoid X receptor ligand used in the treatment of cutaneous T-cell lymphoma, has been reported to cause clinically evident central hypothyroidism by suppressing TSH and increasing T₄ clearance.²⁹

While there is general agreement that beta-adrenergic antagonists (beta-blockers) do not affect the serum TSH concentration, conflicting data have been reported concerning their effect on other thyroid function tests. This may be due to several factors, including dose, duration of therapy, the patient’s thyroid status, and differences in laboratory methodology.³⁰

In studies of propranolol, serum total T₄ concentrations did not change or were increased with daily doses of 160 mg or more in both euthyroid participants and hyperthyroid patients^31–33; serum total T₃ concentrations did not change or were decreased with 40 mg or more daily³⁴; and serum reverse T₃ concentrations were increased with daily doses of 80 mg or more.³¹ It is most likely that propranolol exerts these changes by inhibiting D1 activity in peripheral tissues.

Furthermore, a significant decrease in serum total T₃ concentrations was observed in hyperthyroid patients treated with atenolol 100 mg daily, metoprolol 100 mg daily, and alprenolol 100 mg daily, but not with sotalol 80 mg daily or nadolol (up to 240 mg daily).^35,36

On the other hand, beta-adrenergic agonists have not been reported to cause significant changes in thyroid function tests.³⁷

SUBCLINICAL THYROTOXICOSIS OR HASHIMOTO THYROIDITIS?

Our patient’s thyroid function test results are more likely due to his nonthyroidal illness and glucocorticoid therapy, as there is no clinical evidence to point to a hypothalamic-pituitary disorder accounting for true central hypothyroidism.

The other options mentioned in question 2 are unlikely to explain our patient’s thyroid function test results.

Subclinical thyrotoxicosis is characterized by suppressed serum TSH, but both serum free T₄ and total T₃ remain within the normal reference ranges. In addition, the serum TSH level may help to differentiate between thyrotoxicosis and nonthyroidal illness. In the former, serum TSH is usually suppressed (< 0.01 mU/L), whereas in the latter it is usually low but detectable (0.05– 0.3 mU/L).^38,39

Hashimoto thyroiditis is a chronic autoimmune thyroid disease characterized by diffuse lymphocytic infiltration of the thyroid gland. Almost all patients with Hashimoto thyroiditis have elevated levels of antibodies to thyroid peroxidase or thyroglobulin.⁴⁰ Clinically, patients with Hashimoto thyroiditis can either be hypothyroid or have normal thyroid function, which is not the case in our patient.

CASE CONTINUED

An endocrinologist, consulted for a second opinion, agreed that the patient’s thyroid function test results were most likely due to his nonthyroidal illness and glucocorticoid therapy.

3. In view of the endocrinologist’s opinion, which should be the next step in the management of the patient’s thyroid condition?

• Start levothyroxine (T₄) therapy
• Start liothyronine (T₃) therapy
• Start N-acetylcysteine therapy
• Start thyrotropin-releasing hormone therapy
• Remeasure thyroid hormones after full recovery from his acute illness

It is not clear whether the changes in thyroid hormone levels during an acute illness are a pathologic alteration for which thyroid hormone therapy may be beneficial, or a physiologic adaptation for which such therapy would not be indicated.⁴¹

However, current data argue against thyroid hormone therapy using T₄ or T₃ for patients with nonthyroidal illness syndrome (also called euthyroid sick syndrome).⁴² Indeed, several randomized controlled trials showed that thyroid hormone therapy is not beneficial in such patients and may be detrimental.^41,43

Therapies other than thyroid hormone have been investigated to ameliorate thyroid hormone abnormalities in patients with nonthyroidal illness. These include N-acetylcysteine, thyrotropin-releasing hormone therapy, and nutritional support.

Some studies showed that giving N-acetyl­cysteine, an antioxidant, increased serum T₃ and decreased serum reverse T₃ concentrations in patients with acute myocardial infarction.⁴⁴ Nevertheless, the mortality rate and length of hospitalization were not affected. Further studies are needed to know whether N-acetylcysteine therapy is beneficial for such patients.

Similarly, a study using a thyrotropin-releasing hormone analogue along with growth hormone-releasing peptide 2 showed an increase in serum TSH, T₄, and T₃ levels in critically ill patients.⁴⁵ The benefit of this therapy has yet to be determined. On the other hand, early nutritional support was reported to prevent thyroid hormonal changes in patients postoperatively.⁴⁶

Measuring thyroid hormone levels after full recovery is the most appropriate next step in our patient, as the changes in thyroid hormone concentrations subside as the acute illness resolves.⁴⁷

The patient continued to improve. On hospital day 6, he was feeling better but still had mild respiratory distress. There had been no further episodes of arrhythmia since day 4. His blood pressure was 136/86 mm Hg, heart rate 88 beats per minute and regular, respiratory rate 18 breaths per minute, and oral temperature 37.1°C. His oxygen saturation was 92% on room air.

Before discharge, he was encouraged to quit smoking. He was offered behavioral counseling and medication therapy, but he only said that he would think about it. He was discharged on oral cefixime for 4 more days and was instructed to switch to a long-acting bronchodilator along with his other home medications and to return in 1 week to have his thyroid hormones checked.

One week later, his laboratory results were:

• TSH 11.2 mU/L (reference range 0.5–5.0)
• Free T₄ 1.2 ng/dL (0.9–2.4)
• Total T₃ 92 ng/dL (80–180).

Clinically, the patient was euthyroid, and examination of his thyroid was unremarkable.

4. Based on these last test results, which statement is correct?

• Levothyroxine therapy should be started
• His serum TSH elevation is most likely transient
• Thyroid ultrasonography is strongly indicated
• A radioactive iodine uptake study should be performed
• Measurement of thyroid-stimulating immunoglobulins is indicated

During recovery from nonthyroidal illness, some patients may have elevated serum TSH levels that are usually transient and modest (< 20 mU/L).⁴⁸ Normalization of the thyroid function tests including serum TSH may take several weeks⁴⁹ or months.⁵⁰ However, a systematic review found that the likelihood of permanent primary hypothyroidism is high in patients with serum TSH levels higher than 20 mU/L during the recovery phase of their nonthyroidal illness.⁵¹

Ultrasonography is useful for evaluating patients with thyroid nodules or goiter but is of little benefit for patients like ours, in whom the thyroid is normal on examination.

Similarly, a radioactive iodine uptake study is not indicated, as it is principally used to help differentiate between types of thyrotoxicosis. (Radioactive iodine is also used to treat differentiated thyroid cancer.)

Thyroid-stimulating immunoglobins are TSH receptor-stimulating antibodies that cause Graves disease. Nevertheless, measuring them is not routinely indicated for its diagnosis. However, their measurement is of significant help in the diagnosis of Graves disease if a radioactive iodine uptake study cannot be performed (as in pregnancy) and in atypical presentations such as euthyroid Graves ophthalmopathy.⁵² Other indications for thyroid-stimulating immunoglobin measurement are beyond the scope of the article. Our patient’s test results are not consistent with hyperthyroidism, so measuring thyroid-stimulating immunoglobins is not indicated.

CASE CONCLUSION: BETTER, BUT STILL SMOKING

The patient missed his 1-month clinic follow-up, but he visited the clinic for follow-up 3 months later. He was feeling well with no complaints. Test results including serum TSH, free T₄, and total T₃ were within normal ranges. His COPD was under control, with an FEV₁ 88% of predicted.

He was again encouraged to quit smoking and was offered drug therapy and behavioral counseling, but he declined. In addition, he was instructed to adhere to his annual influenza vaccination.

KEY POINTS

• In patients with acute illness, it is recommended that thyroid function not be assessed unless there is a strong indication.
• If thyroid function assessment is indicated for critically ill patients, serum TSH and free T₄ concentrations should be measured. Some clinicians also measure serum total T₃ level.
• Thyroid function testing in critically ill patients usually shows low serum total T₃, normal or low serum TSH, and normal or low serum free T₄.
• Many drugs can alter thyroid hormone levels.
• Thyroid hormone therapy is not recommended for critically ill patients with low T₃, low T₄, or both.
• During recovery from nonthyroidal illness, some patients may have mild elevation in serum TSH levels (< 20 mU/L).
• Thyroid hormone levels may take several weeks or months to return to normal after the acute illness.
• Patients with serum TSH levels higher than 20 mU/L during the recovery phase of their nonthyroidal illness are more likely to have permanent primary hypothyroidism.

Myopathies can present with a wide variety of symptoms, so patients with muscle weakness are often seen initially by a general practitioner. Nonrheumatologists should be able to evaluate a patient presenting with muscle weakness or myalgia and be aware of red flags indicating potentially dangerous syndromes that require a prompt, thorough investigation.

This article reviews selected causes of muscle weakness, such as statin-induced and autoimmune disorders, and systemic features of inflammatory myopathies beyond myositis, such as dermatologic and pulmonary manifestations.

FOCUSING THE EVALUATION

The evaluation of a patient presenting with muscle weakness should include several assessments: 

Temporal progression. Was the onset of symptoms rapid or insidious? Patterns of onset may give clues to etiology, including the possibility of an associated autoimmune condition.

Location of muscle weakness. Are symptoms global or localized? And if localized, are they proximal or distal? Proximal weakness can be manifested by difficulty rising from a chair (hip muscles) or combing one’s hair (shoulder muscles), whereas distal weakness can involve difficulty standing on toes (gastrocnemius and soleus muscles) or performing fine motor activities (intrinsic hand muscles).

Symmetry. A focal or asymmetric pattern often has a neurologic etiology, but this could also be consistent with inclusion body myositis.

Other symptoms. Arthritis, rash, and swallowing problems point to a possible underlying rheumatologic disease. Weight gain or loss may indicate a thyroid disorder.

Family history. Some patients report that others in their family have this pattern of weakness, indicating a likely genetic myopathy. If the patient reports a relative with multiple sclerosis, lupus erythematosus, rheumatoid arthritis, or another autoimmune disease, then an immune-mediated myopathy should be considered.

Medications should be reviewed, particularly statins.

CASE 1: SLOWLY PROGRESSIVE WEAKNESS

A 65-year-old man presented with the insidious onset of muscle weakness and episodes of falling. On review of his medical record, his serum creatine kinase (CK) levels were elevated at various periods at 2 to 4 times the upper limit of normal. Electromyography (EMG) previously showed a myopathic pattern, and a muscle biopsy was abnormal, consistent with endomysial inflammation (term is consistent with “polymyositis”). He was treated for polymyositis for several years with prednisone alone, with steroids plus methotrexate, and with combined immunosuppression including methotrexate and azathioprine, but with no improvement. Eventually, another muscle biopsy revealed inclusion bodies with rimmed vacuoles, consistent with inclusion body myositis.

Inclusion body myositis

Inclusion body myositis is the most common myopathy in middle-aged to elderly people, especially men. These patients are often told “You are just getting old,” but they have a defined condition. It should also be considered in patients failing to respond to treatment or with those with “refractory” polymyositis.

The onset of muscle weakness is insidious and painless, and the weakness progresses slowly. The pattern is distal and asymmetric (eg, foot drop), and muscle atrophy typically affects the forearm flexors, quadriceps, and intrinsic muscles of the hands.¹

Magnetic resonance imaging may show marked muscle atrophy. Unfortunately, no treatment has shown efficacy, and most neuromuscular and rheumatology experts do not treat inclusion body myositis with immunosuppressive drugs. 

CASE 2: MILD MYALGIA WITHOUT WEAKNESS

A black 52-year-old man was referred because of myalgia and a CK level of 862 U/L (reference range < 200). His physician wanted to start him on a statin but was hesitant to do so without first consulting a rheumatologist.

The patient had a long history of mild arthralgias and myalgias without muscle weakness. He had dyslipidemia and hypertension. He reported no family history of myopathy and no illicit drug use. He was formerly an athlete. Medications included a thiazide diuretic and a beta-blocker. On examination, his muscles were strong (rated 5 on a scale of 5) in the upper and lower extremities, without atrophy.

His records showed that his CK levels had risen and fallen repeatedly over the past few years, ranging from 600 to 1,100 U/L. On further questioning, he reported that when he had joined the army 30 years previously, a physician had recommended he undergo a liver biopsy in view of elevated liver function tests, but that he had refused because he felt fine.

Currently, his gamma-glutamyl transpeptidase levels were normal.

Idiopathic ‘hyperCKemia’

So-called idiopathic hyperCKemia is not a form of myositis but merely a laboratory result outside the “normal” range. Reference ranges are based predominantly on measurements in white people and on an assumption that the distribution is Gaussian (bell-shaped). A normal CK level is usually defined as less than 200 U/L. Using this standard, up to 20% of men and 5% of women have hyperCKemia.²

However, CK levels vary by sex and ethnicity, with mean levels highest in black men, followed by black women, white men, and white women. The mean level in black men is higher than the standard cutoff point for normal, and especially in this population, there is wide fluctuation around the mean, leading to hyperCKemia quite frequently in black men. Exercise and manual labor also drive up CK levels.^3–5 

Idiopathic hyperCKemia is benign. D’Adda et al⁶ followed 55 patients for a mean of 7.5 years. CK levels normalized in 12 patients or at least decreased in 24. Most remained symptom-free or had minimal symptoms. 

Idiopathic hyperCKemia: Bottom line

Before prescribing a statin, determine the baseline CK level. If slightly elevated (ie, up to 3 to 5 times the upper limit of normal, or even higher) in the setting of normal muscle strength, there is no need for electromyography or muscle biopsy, and the patient can certainly receive a statin. Most of these patients do not need to see a rheumatologist but can simply have their CK and muscle strength monitored.

Myositis (idiopathic inflammatory myopathy) is a heterogeneous group of autoimmune syndromes of unknown cause characterized by chronic muscle weakness and inflammation of striated muscle. These syndromes likely arise as a result of genetic predisposition and an environmental or infectious “hit.”

Myositis is rare, with an incidence of 5 to 10 cases per million per year and an estimated prevalence of 50 to 90 cases per million. It has 2 incidence peaks: 1 in childhood (age 5–15) and another in adult midlife (age 30–50). Women are affected 2 to 3 times more often than men, with black women most commonly affected.

Myositis is traditionally classified as follows:

• Adult polymyositis
• Adult dermatomyositis
• Juvenile myositis (dermatomyositis much more frequent than polymyositis)
• Malignancy-associated myositis (usually dermatomyositis)
• Myositis overlapping with another autoimmune disease
• Inclusion body myositis.

However, polymyositis is less common than we originally thought, and the term necrotizing myopathy is now used in many patients, as noted in the case studies below. Further, myositis overlap syndromes are being increasingly diagnosed, likely related to the emergence of autoantibodies and clinical “syndromes” associated with these autoantibody subsets (discussed in cases below).

Dermatomyositis

Dermatomyositis is characterized by muscle weakness and a rash that can be obvious or subtle. Classic skin lesions are Gottron papules, which are raised, flat-topped red or purplish lesions over the knuckles, elbows, or knees.

Lesions may be confused with those of psoriasis. There can also be a V-neck rash over the anterior chest or upper back (“shawl sign”) or a rash over the lateral thigh (“holster sign”). A facial rash may occur, but unlike lupus, dermatomyositis does not spare the nasolabial area. However, the V-neck rash can be similar to that seen in lupus.

Dermatomyositis may cause muscle pain, perhaps related to muscle ischemia, whereas polymyositis and necrotizing myopathy are often painless. However, pain is also associated with fibromyalgia, which may be seen in many autoimmune conditions. It is important not to overtreat rheumatologic diseases with immunosuppression to try to control pain if the pain is actually caused by fibromyalgia.

Polymyositis mimics

Hypothyroid myopathy can present as classic polymyositis. The serum CK may be elevated, and there may be myalgias, muscle hypertrophy with stiffness, weakness, cramps, and even features of a proximal myopathy, and rhabdomyolysis. The electromyogram can be normal or myopathic. Results of muscle biopsy are often normal but may show focal necrosis and mild inflammatory infiltrates, thus mimicking that seen with inflammatory myopathy.⁷

Drug-induced or toxic myopathies can also mimic polymyositis. Statins are among the most commonly prescribed drugs in the United States, with more than 35 million people taking them. Statins are generally well tolerated but have a broad spectrum of toxicity, ranging from myalgias to life-threatening rhabdomyolysis. Myalgias lead to about 5% to 10% of patients refusing to take a statin or stopping it on their own.

Myalgias affect up to 20% of statin users in clinical practice.^8,9 A small cross-sectional study¹⁰ of 1,000 patients in a primary care setting found that the risk of muscle complaints in statin users was 1.5 times higher than in nonstatin users, similar to findings in other studies.

My strategy for managing a patient with possible statin-induced myopathy is illustrated in Figure 1.

CASE 3: WEAKNESS, VERY HIGH CK ON A STATIN

In March 2010, a 67-year-old woman presented with muscle weakness. She had a history of hypertension, hyperlipidemia, and, more than 10 years previously, uterine cancer. In 2004, she was given atorvastatin for dyslipidemia. Four years later, she developed lower-extremity weakness, which her doctor attributed to normal aging. A year after that, she found it difficult to walk up steps and lift her arms overhead. In June 2009, she stopped taking the atorvastatin on her own, but the weakness did not improve.

In September 2009, she returned to her doctor, who found her CK level was 6,473 U/L but believed it to be an error, so the test was repeated, with a result of 9,375 U/L. She had no rash or joint involvement.

She was admitted to the hospital and underwent muscle biopsy, which showed myonecrosis with no inflammation or vasculitis.  She was treated with prednisone 60 mg/day, and her elevated CK level and weakness improved.

Immune-mediated necrotizing myopathy associated with statins

The hallmark of necrotizing myopathy is myonecrosis without significant inflammation.¹² This pattern contrasts with that of polymyositis, which is characterized by lymphocytic inflammation.

Although statins became available in the United States in 1987, immune-mediated necrotizing myopathy associated with statins was first described only in 2010. In that report, Grable-Esposito et al¹³ described 25 patients from 2 neuromuscular centers seen between 2000 and 2008 who had elevated CK and proximal weakness during or after statin use, both of which persisted despite stopping the statin. Patients improved with immunosuppressive agents but had a relapse when steroids were stopped or tapered, a pattern typical in autoimmune disease.

Also in 2010, Christopher-Stine et al¹⁴ reported an antibody associated with necrotizing myopathy. Of 38 patients with the condition, 16 were found to have an abnormal “doublet” autoantibody recognizing 200- and 100-kDa proteins. All patients had weakness and a high CK level, and 63% had statin exposure before the weakness (this percentage increased to 83% in patients older than 50). All responded to immunosuppressive therapy, and many had a relapse when it was withdrawn.

Statins lower cholesterol by inhibiting 3-hydroxy-3-methylglutaryl-Co A reductase (HMGCR), and paradoxically, they also upregulate it. HMGCR has a molecular weight of 97 kDa. Mammen et al¹⁵ identified HMGCR as the 100-kDa target of the identified antibody and developed an enzyme-linked immunosorbent assay for it. Of 750 patients presenting to one center, only 45 (6%) had anti-HMGCR autoantibodies, but all 16 patients who had the abnormal doublet antibody tested positive for anti-HMGCR. Regenerating muscle cells express high levels of HMGCR, which may sustain the immune response after statins are discontinued.

Case 3 continued: Intravenous immunoglobulin brings improvement

In March 2010, when the 67-year-old patient presented to our myositis center, her CK level was 5,800 U/L, which increased as prednisone was tapered. She still felt weak. On examination, her muscle strength findings were deltoids 4+/5, neck flexors 4/5, and iliopsoas 3+/5. She was treated with methotrexate and azathioprine without benefit. She was next treated with intravenous immunoglobulin, and after 3 months, her strength normalized for the first time in years. Her CK level decreased but did not normalize. Testing showed that she was positive for anti-HMGCR auto­antibody, as this test had become commercially available.

In 2015, Mammen and Tiniakou¹⁶ suggested using intravenous immunoglobulin as first-line therapy for statin-associated autoimmune necrotizing myopathy, based on experience at a single center with 3 patients who declined glucocorticoid treatment.

Necrotizing myopathy: Bottom line

Myositis overlap syndromes

Heterogeneity is the rule in myositis, and it can present with a wide variety of signs and symptoms as outlined in Table 2.

CASE 4: FEVER, NEW ‘RHEUMATOID ARTHRITIS,’ AND LUNG DISEASE

A 52-year-old woman with knee osteoarthritis saw her primary care physician in November 2013 for dyspnea and low-grade fever. The next month, she presented with polyarthritis, muscle weakness, and Raynaud phenomenon.

In January 2014, she developed acrocyanosis of her fingers. Examination revealed hyperkeratotic, cracked areas of her fingers. Her oxygen saturation by pulse oximetry was low. She was admitted to the hospital. Her doctor suspected new onset of rheumatoid arthritis, but blood tests revealed a negative antinuclear antibody, so an autoimmune condition was deemed unlikely. Her CK was mildly elevated at 350 U/L.

Because of her dyspnea, an open-lung biopsy was performed. High-resolution computed tomography (CT) revealed infiltrates and ground-glass opacities, leading to the diagnosis of nonspecific interstitial pneumonia. A rheumatologist was consulted and recommended pulse methylprednisolone, followed by prednisone 60 mg/day and mycophenolate mofetil. Testing for Jo-1 antibodies was positive.

Antisynthetase syndrome

The antisynthetase syndrome is a clinically heterogeneous condition that can occur with any or all of the following:

• Fever
• Myositis
• Arthritis (often misdiagnosed as rheumatoid arthritis)
• Raynaud phenomenon
• Mechanic’s hands (hyperkeratotic  roughness with fissures on the lateral aspects of the fingers and finger pads)
• Interstitial lung disease.

The skin rashes and myositis may be subtle, making the presentation “lung-dominant,” and nonrheumatologists should be aware of this syndrome. Although in our patient the condition developed in a classic manner, with all of the aforementioned features of the antisynthetase syndrome, some patients will manifest one or a few of the features.

Clinically, patients with the Jo-1 antisynthetase syndrome often present differently than those with non-Jo-1 antisynthetase autoantibodies. When we compared 122 patients with Jo-1 vs 80 patients with a non-Jo-1 antisynthetase autoantibody, patients with Jo-1 antibodies were more likely to have initially received a diagnosis of myositis (83%), while  myositis was the original diagnosis in only 17% of those possessing non-Jo-1 antisynthetase autoantibodies. In fact, many patients (approximately 50%) were diagnosed as having undifferentiated connective tissue disease or an overlap syndrome, and 13% had scleroderma as their first diagnosis.¹⁷

We also found that the survival rate was higher in patients with Jo-1 syndrome compared with patients with non-Jo-1 antisynthetase syndromes. We attributed the difference in survival rates to a delayed diagnosis in the non-Jo-1 group, perhaps due to their “nonclassic” presentations of the antisynthetase syndrome, delaying appropriate treatment. Patients received a diagnosis of Jo-1 antibody syndrome after a mean of 0.4 year (range 0.2–0.8), while those with a non-Jo-1 antisynthetase autoantibody had a delay in diagnosis of 1.0 year (range 0.4–5.1) (P < .01).¹⁷

In nearly half the cases in this cohort, pulmonary fibrosis was the cause of death, with primary pulmonary hypertension being the second leading cause (11%).

Antisynthetase syndrome: Bottom line

Antisynthetase syndrome is an often fatal disease that does not always present in a typical fashion with symptoms of myositis, as lung disease may be the predominant feature. A negative antinuclear antibody test result does not imply antibody negativity, as the autoantigen in these diseases is not located in the nucleus. Prompt diagnosis and appropriate immunosuppressive therapy are critical to improving outcomes.

CASE 5: FEVER, UNDIAGNOSED LUNG DISEASE, NO MYOSITIS

In January 2001, a 39-year-old woman was admitted to the hospital after 5 weeks of fever (temperatures 103°–104°F) and myalgias. An extensive workup was negative except for low-titer antinuclear antibody and for mild basilar fibrosis noted on chest radiography. She left the hospital against medical advice because of frustration with a lack of a specific diagnosis (“fever of unknown origin”).

Two months later, at a follow-up rheumatology consult, she reported more myalgias and arthralgias, as well as fever. Chest radiography now showed pleural effusions. Her fingers had color changes consistent with Raynaud phenomenon. At that time, I diagnosed an undifferentiated connective tissue disease and told her that I suspected an autoimmune condition that would need time to reveal itself. In the meantime, I treated her empirically with prednisone.

In April, she returned, much more short of breath and with more prominent diffuse pulmonary infiltrates. Physical examination revealed subtle Gottron changes. Testing revealed poor pulmonary function: forced vital capacity (FVC) 56%, forced expiratory volume in 1 second (FEV₁) 52%, and diffusing capacity for carbon monoxide (Dlco) 40%. Blood testing was positive for anti-PL-12 antibody, one of the non-Jo-1 antisynthetase antibodies. At this time, we treated her with glucocorticoids and tacrolimus.

More than 15 years later, this patient is doing well. Her skin rash, joint symptoms, and fever have not returned, and interestingly, she never developed myositis. Her Raynaud symptoms are mild. Her most recent pulmonary function test results (January 2018) were FVC  75%, FEV₁ 87%, and Dlco 78%. Although these results are not normal, they are much improved and allow her to be completely functional without supplemental oxygen. Echocardiography showed  normal pulmonary artery systolic pressure (25 mm Hg). She was still taking tacrolimus and prednisone. When we tried to stop tacrolimus after she had done well for many years, her condition flared.

Non-Jo-1 antisynthetase syndrome: Bottom line

Patients with a non-Jo-1 antisynthetase syndrome often present without myositis symptoms and may never manifest myositis symptoms. Likely because of this presentation, diagnosis of a specific connective tissue disorder is delayed, perhaps leading to increased mortality risk from pulmonary disease. Chronic immunosuppression is often required for these autoimmune conditions.

CASE 6: DERMATOMYOSITIS, RAPIDLY PROGRESSIVE INTERSTITIAL LUNG DISEASE

A 58-year-old woman presented in the summer of 2012 with a photosensitive rash. The following January, she returned with polyarthritis, mild muscle weakness, and a dermatomyositis-pattern rash. Her CK level was normal, and her antinuclear antibody and Sjögren syndrome antibody test results were negative. She improved on low-dose prednisone and methotrexate.

She was originally referred to me in May of that year for worsening rash and mild weakness. She denied pulmonary symptoms, but examination revealed faint basilar crackles. I increased her prednisone dosage to 20 mg/day and started mycophenolate mofetil mainly for the mild cutaneous and myositis features. I also recommended high-resolution CT of the lungs and pulmonary function tests, which she underwent in early June. High-resolution CT showed nonspecific mild infiltrates with minimal ground-glass opacities.

On July 1, she presented to her local emergency department with severe shortness of breath, requiring oxygen 12 L/min. She had a palmar rash. Repeat high-resolution CT showed dramatic worsening compared with the scan the previous month. Because of continued inadequate oxygenation, she was transferred to our center. A blood test later was positive for antimelanoma differentiation-associated gene 5 (MDA-5) autoantibody, previously known as anticlinically amyopathic dermatomyositis (anti-CADM)-140 antibody (based on immunoprecipitation results).

She died on the third day after transfer, just 2 months after I had originally seen her, at which time she had had no pulmonary symptoms.

Clinically amyopathic dermatomyositis

Anti-CADM-140, first reported from Asia,^18–20 is an autoantibody-associated disease but not an antisynthetase. It is associated with dermatomyositis; patients often have a “vasculopathy” with cutaneous ulcerations and palmar papules.

MDA-5 is a cytoplasmic protein that “senses” viral RNA and induces production of type 1 interferon. It is involved in the innate immune defense against viruses.

Anti-MDA-5 positivity is associated with a poor pulmonary outcome.²¹ In our cohort from the University of Pittsburgh, many patients died within 3 years, compared with about a 40% survival rate in patients with dermatomyositis who tested negative for this antibody. That being said, many patients with anti-MDA-5 do not develop rapidly progressive interstitial lung disease.

Autoimmune interstitial lung disease: Bottom line

Autoimmune interstitial lung disease is easy to miss, especially in the case of a non-Jo-1 syndrome, for 3 important reasons:

• The autoimmune features may initially be subtle (eg, Raynaud phenomena, mild dermatomyositis rash, undifferentiated connective tissue disease)
• Autoantibody testing is not often ordered, is not standardized, or may be unavailable 
• Providers are mistakenly reassured that a patient who tests negative for antinuclear antibody does not have an autoimmune condition.

To emphasize the last point, in a cohort of 202 patients who tested positive for an antisynthetase antibody, only half were antinuclear antibody-positive, but nearly three-quarters demonstrated anticytoplasmic staining on indirect immunofluorescence (due to the location of the autoantigen in the cytoplasm), making the latter a better screening test for an antisynthetase antibody. For scleroderma, 99% were antinculear antibody-positive, but for myositis, this test is much less sensitive.²²

Myopathies can present with a wide variety of symptoms, so patients with muscle weakness are often seen initially by a general practitioner. Nonrheumatologists should be able to evaluate a patient presenting with muscle weakness or myalgia and be aware of red flags indicating potentially dangerous syndromes that require a prompt, thorough investigation.

This article reviews selected causes of muscle weakness, such as statin-induced and autoimmune disorders, and systemic features of inflammatory myopathies beyond myositis, such as dermatologic and pulmonary manifestations.

FOCUSING THE EVALUATION

The evaluation of a patient presenting with muscle weakness should include several assessments: 

Temporal progression. Was the onset of symptoms rapid or insidious? Patterns of onset may give clues to etiology, including the possibility of an associated autoimmune condition.

Location of muscle weakness. Are symptoms global or localized? And if localized, are they proximal or distal? Proximal weakness can be manifested by difficulty rising from a chair (hip muscles) or combing one’s hair (shoulder muscles), whereas distal weakness can involve difficulty standing on toes (gastrocnemius and soleus muscles) or performing fine motor activities (intrinsic hand muscles).

Symmetry. A focal or asymmetric pattern often has a neurologic etiology, but this could also be consistent with inclusion body myositis.

Other symptoms. Arthritis, rash, and swallowing problems point to a possible underlying rheumatologic disease. Weight gain or loss may indicate a thyroid disorder.

Family history. Some patients report that others in their family have this pattern of weakness, indicating a likely genetic myopathy. If the patient reports a relative with multiple sclerosis, lupus erythematosus, rheumatoid arthritis, or another autoimmune disease, then an immune-mediated myopathy should be considered.

Medications should be reviewed, particularly statins.

CASE 1: SLOWLY PROGRESSIVE WEAKNESS

A 65-year-old man presented with the insidious onset of muscle weakness and episodes of falling. On review of his medical record, his serum creatine kinase (CK) levels were elevated at various periods at 2 to 4 times the upper limit of normal. Electromyography (EMG) previously showed a myopathic pattern, and a muscle biopsy was abnormal, consistent with endomysial inflammation (term is consistent with “polymyositis”). He was treated for polymyositis for several years with prednisone alone, with steroids plus methotrexate, and with combined immunosuppression including methotrexate and azathioprine, but with no improvement. Eventually, another muscle biopsy revealed inclusion bodies with rimmed vacuoles, consistent with inclusion body myositis.

Inclusion body myositis

Inclusion body myositis is the most common myopathy in middle-aged to elderly people, especially men. These patients are often told “You are just getting old,” but they have a defined condition. It should also be considered in patients failing to respond to treatment or with those with “refractory” polymyositis.

The onset of muscle weakness is insidious and painless, and the weakness progresses slowly. The pattern is distal and asymmetric (eg, foot drop), and muscle atrophy typically affects the forearm flexors, quadriceps, and intrinsic muscles of the hands.¹

Magnetic resonance imaging may show marked muscle atrophy. Unfortunately, no treatment has shown efficacy, and most neuromuscular and rheumatology experts do not treat inclusion body myositis with immunosuppressive drugs. 

CASE 2: MILD MYALGIA WITHOUT WEAKNESS

A black 52-year-old man was referred because of myalgia and a CK level of 862 U/L (reference range < 200). His physician wanted to start him on a statin but was hesitant to do so without first consulting a rheumatologist.

The patient had a long history of mild arthralgias and myalgias without muscle weakness. He had dyslipidemia and hypertension. He reported no family history of myopathy and no illicit drug use. He was formerly an athlete. Medications included a thiazide diuretic and a beta-blocker. On examination, his muscles were strong (rated 5 on a scale of 5) in the upper and lower extremities, without atrophy.

His records showed that his CK levels had risen and fallen repeatedly over the past few years, ranging from 600 to 1,100 U/L. On further questioning, he reported that when he had joined the army 30 years previously, a physician had recommended he undergo a liver biopsy in view of elevated liver function tests, but that he had refused because he felt fine.

Currently, his gamma-glutamyl transpeptidase levels were normal.

Idiopathic ‘hyperCKemia’

So-called idiopathic hyperCKemia is not a form of myositis but merely a laboratory result outside the “normal” range. Reference ranges are based predominantly on measurements in white people and on an assumption that the distribution is Gaussian (bell-shaped). A normal CK level is usually defined as less than 200 U/L. Using this standard, up to 20% of men and 5% of women have hyperCKemia.²

However, CK levels vary by sex and ethnicity, with mean levels highest in black men, followed by black women, white men, and white women. The mean level in black men is higher than the standard cutoff point for normal, and especially in this population, there is wide fluctuation around the mean, leading to hyperCKemia quite frequently in black men. Exercise and manual labor also drive up CK levels.^3–5 

Idiopathic hyperCKemia is benign. D’Adda et al⁶ followed 55 patients for a mean of 7.5 years. CK levels normalized in 12 patients or at least decreased in 24. Most remained symptom-free or had minimal symptoms. 

Idiopathic hyperCKemia: Bottom line

Before prescribing a statin, determine the baseline CK level. If slightly elevated (ie, up to 3 to 5 times the upper limit of normal, or even higher) in the setting of normal muscle strength, there is no need for electromyography or muscle biopsy, and the patient can certainly receive a statin. Most of these patients do not need to see a rheumatologist but can simply have their CK and muscle strength monitored.

Myositis (idiopathic inflammatory myopathy) is a heterogeneous group of autoimmune syndromes of unknown cause characterized by chronic muscle weakness and inflammation of striated muscle. These syndromes likely arise as a result of genetic predisposition and an environmental or infectious “hit.”

Myositis is rare, with an incidence of 5 to 10 cases per million per year and an estimated prevalence of 50 to 90 cases per million. It has 2 incidence peaks: 1 in childhood (age 5–15) and another in adult midlife (age 30–50). Women are affected 2 to 3 times more often than men, with black women most commonly affected.

Myositis is traditionally classified as follows:

• Adult polymyositis
• Adult dermatomyositis
• Juvenile myositis (dermatomyositis much more frequent than polymyositis)
• Malignancy-associated myositis (usually dermatomyositis)
• Myositis overlapping with another autoimmune disease
• Inclusion body myositis.

However, polymyositis is less common than we originally thought, and the term necrotizing myopathy is now used in many patients, as noted in the case studies below. Further, myositis overlap syndromes are being increasingly diagnosed, likely related to the emergence of autoantibodies and clinical “syndromes” associated with these autoantibody subsets (discussed in cases below).

Dermatomyositis

Dermatomyositis is characterized by muscle weakness and a rash that can be obvious or subtle. Classic skin lesions are Gottron papules, which are raised, flat-topped red or purplish lesions over the knuckles, elbows, or knees.

Lesions may be confused with those of psoriasis. There can also be a V-neck rash over the anterior chest or upper back (“shawl sign”) or a rash over the lateral thigh (“holster sign”). A facial rash may occur, but unlike lupus, dermatomyositis does not spare the nasolabial area. However, the V-neck rash can be similar to that seen in lupus.

Dermatomyositis may cause muscle pain, perhaps related to muscle ischemia, whereas polymyositis and necrotizing myopathy are often painless. However, pain is also associated with fibromyalgia, which may be seen in many autoimmune conditions. It is important not to overtreat rheumatologic diseases with immunosuppression to try to control pain if the pain is actually caused by fibromyalgia.

Polymyositis mimics

Hypothyroid myopathy can present as classic polymyositis. The serum CK may be elevated, and there may be myalgias, muscle hypertrophy with stiffness, weakness, cramps, and even features of a proximal myopathy, and rhabdomyolysis. The electromyogram can be normal or myopathic. Results of muscle biopsy are often normal but may show focal necrosis and mild inflammatory infiltrates, thus mimicking that seen with inflammatory myopathy.⁷

Drug-induced or toxic myopathies can also mimic polymyositis. Statins are among the most commonly prescribed drugs in the United States, with more than 35 million people taking them. Statins are generally well tolerated but have a broad spectrum of toxicity, ranging from myalgias to life-threatening rhabdomyolysis. Myalgias lead to about 5% to 10% of patients refusing to take a statin or stopping it on their own.

Myalgias affect up to 20% of statin users in clinical practice.^8,9 A small cross-sectional study¹⁰ of 1,000 patients in a primary care setting found that the risk of muscle complaints in statin users was 1.5 times higher than in nonstatin users, similar to findings in other studies.

My strategy for managing a patient with possible statin-induced myopathy is illustrated in Figure 1.

CASE 3: WEAKNESS, VERY HIGH CK ON A STATIN

In March 2010, a 67-year-old woman presented with muscle weakness. She had a history of hypertension, hyperlipidemia, and, more than 10 years previously, uterine cancer. In 2004, she was given atorvastatin for dyslipidemia. Four years later, she developed lower-extremity weakness, which her doctor attributed to normal aging. A year after that, she found it difficult to walk up steps and lift her arms overhead. In June 2009, she stopped taking the atorvastatin on her own, but the weakness did not improve.

In September 2009, she returned to her doctor, who found her CK level was 6,473 U/L but believed it to be an error, so the test was repeated, with a result of 9,375 U/L. She had no rash or joint involvement.

She was admitted to the hospital and underwent muscle biopsy, which showed myonecrosis with no inflammation or vasculitis.  She was treated with prednisone 60 mg/day, and her elevated CK level and weakness improved.

Immune-mediated necrotizing myopathy associated with statins

The hallmark of necrotizing myopathy is myonecrosis without significant inflammation.¹² This pattern contrasts with that of polymyositis, which is characterized by lymphocytic inflammation.

Although statins became available in the United States in 1987, immune-mediated necrotizing myopathy associated with statins was first described only in 2010. In that report, Grable-Esposito et al¹³ described 25 patients from 2 neuromuscular centers seen between 2000 and 2008 who had elevated CK and proximal weakness during or after statin use, both of which persisted despite stopping the statin. Patients improved with immunosuppressive agents but had a relapse when steroids were stopped or tapered, a pattern typical in autoimmune disease.

Also in 2010, Christopher-Stine et al¹⁴ reported an antibody associated with necrotizing myopathy. Of 38 patients with the condition, 16 were found to have an abnormal “doublet” autoantibody recognizing 200- and 100-kDa proteins. All patients had weakness and a high CK level, and 63% had statin exposure before the weakness (this percentage increased to 83% in patients older than 50). All responded to immunosuppressive therapy, and many had a relapse when it was withdrawn.

Statins lower cholesterol by inhibiting 3-hydroxy-3-methylglutaryl-Co A reductase (HMGCR), and paradoxically, they also upregulate it. HMGCR has a molecular weight of 97 kDa. Mammen et al¹⁵ identified HMGCR as the 100-kDa target of the identified antibody and developed an enzyme-linked immunosorbent assay for it. Of 750 patients presenting to one center, only 45 (6%) had anti-HMGCR autoantibodies, but all 16 patients who had the abnormal doublet antibody tested positive for anti-HMGCR. Regenerating muscle cells express high levels of HMGCR, which may sustain the immune response after statins are discontinued.

Case 3 continued: Intravenous immunoglobulin brings improvement

In March 2010, when the 67-year-old patient presented to our myositis center, her CK level was 5,800 U/L, which increased as prednisone was tapered. She still felt weak. On examination, her muscle strength findings were deltoids 4+/5, neck flexors 4/5, and iliopsoas 3+/5. She was treated with methotrexate and azathioprine without benefit. She was next treated with intravenous immunoglobulin, and after 3 months, her strength normalized for the first time in years. Her CK level decreased but did not normalize. Testing showed that she was positive for anti-HMGCR auto­antibody, as this test had become commercially available.

In 2015, Mammen and Tiniakou¹⁶ suggested using intravenous immunoglobulin as first-line therapy for statin-associated autoimmune necrotizing myopathy, based on experience at a single center with 3 patients who declined glucocorticoid treatment.

Necrotizing myopathy: Bottom line

Myositis overlap syndromes

Heterogeneity is the rule in myositis, and it can present with a wide variety of signs and symptoms as outlined in Table 2.

CASE 4: FEVER, NEW ‘RHEUMATOID ARTHRITIS,’ AND LUNG DISEASE

A 52-year-old woman with knee osteoarthritis saw her primary care physician in November 2013 for dyspnea and low-grade fever. The next month, she presented with polyarthritis, muscle weakness, and Raynaud phenomenon.

In January 2014, she developed acrocyanosis of her fingers. Examination revealed hyperkeratotic, cracked areas of her fingers. Her oxygen saturation by pulse oximetry was low. She was admitted to the hospital. Her doctor suspected new onset of rheumatoid arthritis, but blood tests revealed a negative antinuclear antibody, so an autoimmune condition was deemed unlikely. Her CK was mildly elevated at 350 U/L.

Because of her dyspnea, an open-lung biopsy was performed. High-resolution computed tomography (CT) revealed infiltrates and ground-glass opacities, leading to the diagnosis of nonspecific interstitial pneumonia. A rheumatologist was consulted and recommended pulse methylprednisolone, followed by prednisone 60 mg/day and mycophenolate mofetil. Testing for Jo-1 antibodies was positive.

Antisynthetase syndrome

The antisynthetase syndrome is a clinically heterogeneous condition that can occur with any or all of the following:

• Fever
• Myositis
• Arthritis (often misdiagnosed as rheumatoid arthritis)
• Raynaud phenomenon
• Mechanic’s hands (hyperkeratotic  roughness with fissures on the lateral aspects of the fingers and finger pads)
• Interstitial lung disease.

The skin rashes and myositis may be subtle, making the presentation “lung-dominant,” and nonrheumatologists should be aware of this syndrome. Although in our patient the condition developed in a classic manner, with all of the aforementioned features of the antisynthetase syndrome, some patients will manifest one or a few of the features.

Clinically, patients with the Jo-1 antisynthetase syndrome often present differently than those with non-Jo-1 antisynthetase autoantibodies. When we compared 122 patients with Jo-1 vs 80 patients with a non-Jo-1 antisynthetase autoantibody, patients with Jo-1 antibodies were more likely to have initially received a diagnosis of myositis (83%), while  myositis was the original diagnosis in only 17% of those possessing non-Jo-1 antisynthetase autoantibodies. In fact, many patients (approximately 50%) were diagnosed as having undifferentiated connective tissue disease or an overlap syndrome, and 13% had scleroderma as their first diagnosis.¹⁷

We also found that the survival rate was higher in patients with Jo-1 syndrome compared with patients with non-Jo-1 antisynthetase syndromes. We attributed the difference in survival rates to a delayed diagnosis in the non-Jo-1 group, perhaps due to their “nonclassic” presentations of the antisynthetase syndrome, delaying appropriate treatment. Patients received a diagnosis of Jo-1 antibody syndrome after a mean of 0.4 year (range 0.2–0.8), while those with a non-Jo-1 antisynthetase autoantibody had a delay in diagnosis of 1.0 year (range 0.4–5.1) (P < .01).¹⁷

In nearly half the cases in this cohort, pulmonary fibrosis was the cause of death, with primary pulmonary hypertension being the second leading cause (11%).

Antisynthetase syndrome: Bottom line

Antisynthetase syndrome is an often fatal disease that does not always present in a typical fashion with symptoms of myositis, as lung disease may be the predominant feature. A negative antinuclear antibody test result does not imply antibody negativity, as the autoantigen in these diseases is not located in the nucleus. Prompt diagnosis and appropriate immunosuppressive therapy are critical to improving outcomes.

CASE 5: FEVER, UNDIAGNOSED LUNG DISEASE, NO MYOSITIS

In January 2001, a 39-year-old woman was admitted to the hospital after 5 weeks of fever (temperatures 103°–104°F) and myalgias. An extensive workup was negative except for low-titer antinuclear antibody and for mild basilar fibrosis noted on chest radiography. She left the hospital against medical advice because of frustration with a lack of a specific diagnosis (“fever of unknown origin”).

Two months later, at a follow-up rheumatology consult, she reported more myalgias and arthralgias, as well as fever. Chest radiography now showed pleural effusions. Her fingers had color changes consistent with Raynaud phenomenon. At that time, I diagnosed an undifferentiated connective tissue disease and told her that I suspected an autoimmune condition that would need time to reveal itself. In the meantime, I treated her empirically with prednisone.

In April, she returned, much more short of breath and with more prominent diffuse pulmonary infiltrates. Physical examination revealed subtle Gottron changes. Testing revealed poor pulmonary function: forced vital capacity (FVC) 56%, forced expiratory volume in 1 second (FEV₁) 52%, and diffusing capacity for carbon monoxide (Dlco) 40%. Blood testing was positive for anti-PL-12 antibody, one of the non-Jo-1 antisynthetase antibodies. At this time, we treated her with glucocorticoids and tacrolimus.

More than 15 years later, this patient is doing well. Her skin rash, joint symptoms, and fever have not returned, and interestingly, she never developed myositis. Her Raynaud symptoms are mild. Her most recent pulmonary function test results (January 2018) were FVC  75%, FEV₁ 87%, and Dlco 78%. Although these results are not normal, they are much improved and allow her to be completely functional without supplemental oxygen. Echocardiography showed  normal pulmonary artery systolic pressure (25 mm Hg). She was still taking tacrolimus and prednisone. When we tried to stop tacrolimus after she had done well for many years, her condition flared.

Non-Jo-1 antisynthetase syndrome: Bottom line

Patients with a non-Jo-1 antisynthetase syndrome often present without myositis symptoms and may never manifest myositis symptoms. Likely because of this presentation, diagnosis of a specific connective tissue disorder is delayed, perhaps leading to increased mortality risk from pulmonary disease. Chronic immunosuppression is often required for these autoimmune conditions.

CASE 6: DERMATOMYOSITIS, RAPIDLY PROGRESSIVE INTERSTITIAL LUNG DISEASE

A 58-year-old woman presented in the summer of 2012 with a photosensitive rash. The following January, she returned with polyarthritis, mild muscle weakness, and a dermatomyositis-pattern rash. Her CK level was normal, and her antinuclear antibody and Sjögren syndrome antibody test results were negative. She improved on low-dose prednisone and methotrexate.

She was originally referred to me in May of that year for worsening rash and mild weakness. She denied pulmonary symptoms, but examination revealed faint basilar crackles. I increased her prednisone dosage to 20 mg/day and started mycophenolate mofetil mainly for the mild cutaneous and myositis features. I also recommended high-resolution CT of the lungs and pulmonary function tests, which she underwent in early June. High-resolution CT showed nonspecific mild infiltrates with minimal ground-glass opacities.

On July 1, she presented to her local emergency department with severe shortness of breath, requiring oxygen 12 L/min. She had a palmar rash. Repeat high-resolution CT showed dramatic worsening compared with the scan the previous month. Because of continued inadequate oxygenation, she was transferred to our center. A blood test later was positive for antimelanoma differentiation-associated gene 5 (MDA-5) autoantibody, previously known as anticlinically amyopathic dermatomyositis (anti-CADM)-140 antibody (based on immunoprecipitation results).

She died on the third day after transfer, just 2 months after I had originally seen her, at which time she had had no pulmonary symptoms.

Clinically amyopathic dermatomyositis

Anti-CADM-140, first reported from Asia,^18–20 is an autoantibody-associated disease but not an antisynthetase. It is associated with dermatomyositis; patients often have a “vasculopathy” with cutaneous ulcerations and palmar papules.

MDA-5 is a cytoplasmic protein that “senses” viral RNA and induces production of type 1 interferon. It is involved in the innate immune defense against viruses.

Anti-MDA-5 positivity is associated with a poor pulmonary outcome.²¹ In our cohort from the University of Pittsburgh, many patients died within 3 years, compared with about a 40% survival rate in patients with dermatomyositis who tested negative for this antibody. That being said, many patients with anti-MDA-5 do not develop rapidly progressive interstitial lung disease.

Autoimmune interstitial lung disease: Bottom line

Autoimmune interstitial lung disease is easy to miss, especially in the case of a non-Jo-1 syndrome, for 3 important reasons:

• The autoimmune features may initially be subtle (eg, Raynaud phenomena, mild dermatomyositis rash, undifferentiated connective tissue disease)
• Autoantibody testing is not often ordered, is not standardized, or may be unavailable 
• Providers are mistakenly reassured that a patient who tests negative for antinuclear antibody does not have an autoimmune condition.

To emphasize the last point, in a cohort of 202 patients who tested positive for an antisynthetase antibody, only half were antinuclear antibody-positive, but nearly three-quarters demonstrated anticytoplasmic staining on indirect immunofluorescence (due to the location of the autoantigen in the cytoplasm), making the latter a better screening test for an antisynthetase antibody. For scleroderma, 99% were antinculear antibody-positive, but for myositis, this test is much less sensitive.²²

Myopathies can present with a wide variety of symptoms, so patients with muscle weakness are often seen initially by a general practitioner. Nonrheumatologists should be able to evaluate a patient presenting with muscle weakness or myalgia and be aware of red flags indicating potentially dangerous syndromes that require a prompt, thorough investigation.

This article reviews selected causes of muscle weakness, such as statin-induced and autoimmune disorders, and systemic features of inflammatory myopathies beyond myositis, such as dermatologic and pulmonary manifestations.

FOCUSING THE EVALUATION

The evaluation of a patient presenting with muscle weakness should include several assessments: 

Temporal progression. Was the onset of symptoms rapid or insidious? Patterns of onset may give clues to etiology, including the possibility of an associated autoimmune condition.

Location of muscle weakness. Are symptoms global or localized? And if localized, are they proximal or distal? Proximal weakness can be manifested by difficulty rising from a chair (hip muscles) or combing one’s hair (shoulder muscles), whereas distal weakness can involve difficulty standing on toes (gastrocnemius and soleus muscles) or performing fine motor activities (intrinsic hand muscles).

Symmetry. A focal or asymmetric pattern often has a neurologic etiology, but this could also be consistent with inclusion body myositis.

Other symptoms. Arthritis, rash, and swallowing problems point to a possible underlying rheumatologic disease. Weight gain or loss may indicate a thyroid disorder.

Family history. Some patients report that others in their family have this pattern of weakness, indicating a likely genetic myopathy. If the patient reports a relative with multiple sclerosis, lupus erythematosus, rheumatoid arthritis, or another autoimmune disease, then an immune-mediated myopathy should be considered.

Medications should be reviewed, particularly statins.

CASE 1: SLOWLY PROGRESSIVE WEAKNESS

A 65-year-old man presented with the insidious onset of muscle weakness and episodes of falling. On review of his medical record, his serum creatine kinase (CK) levels were elevated at various periods at 2 to 4 times the upper limit of normal. Electromyography (EMG) previously showed a myopathic pattern, and a muscle biopsy was abnormal, consistent with endomysial inflammation (term is consistent with “polymyositis”). He was treated for polymyositis for several years with prednisone alone, with steroids plus methotrexate, and with combined immunosuppression including methotrexate and azathioprine, but with no improvement. Eventually, another muscle biopsy revealed inclusion bodies with rimmed vacuoles, consistent with inclusion body myositis.

Inclusion body myositis

Inclusion body myositis is the most common myopathy in middle-aged to elderly people, especially men. These patients are often told “You are just getting old,” but they have a defined condition. It should also be considered in patients failing to respond to treatment or with those with “refractory” polymyositis.

The onset of muscle weakness is insidious and painless, and the weakness progresses slowly. The pattern is distal and asymmetric (eg, foot drop), and muscle atrophy typically affects the forearm flexors, quadriceps, and intrinsic muscles of the hands.¹

Magnetic resonance imaging may show marked muscle atrophy. Unfortunately, no treatment has shown efficacy, and most neuromuscular and rheumatology experts do not treat inclusion body myositis with immunosuppressive drugs. 

CASE 2: MILD MYALGIA WITHOUT WEAKNESS

A black 52-year-old man was referred because of myalgia and a CK level of 862 U/L (reference range < 200). His physician wanted to start him on a statin but was hesitant to do so without first consulting a rheumatologist.

The patient had a long history of mild arthralgias and myalgias without muscle weakness. He had dyslipidemia and hypertension. He reported no family history of myopathy and no illicit drug use. He was formerly an athlete. Medications included a thiazide diuretic and a beta-blocker. On examination, his muscles were strong (rated 5 on a scale of 5) in the upper and lower extremities, without atrophy.

His records showed that his CK levels had risen and fallen repeatedly over the past few years, ranging from 600 to 1,100 U/L. On further questioning, he reported that when he had joined the army 30 years previously, a physician had recommended he undergo a liver biopsy in view of elevated liver function tests, but that he had refused because he felt fine.

Currently, his gamma-glutamyl transpeptidase levels were normal.

Idiopathic ‘hyperCKemia’

So-called idiopathic hyperCKemia is not a form of myositis but merely a laboratory result outside the “normal” range. Reference ranges are based predominantly on measurements in white people and on an assumption that the distribution is Gaussian (bell-shaped). A normal CK level is usually defined as less than 200 U/L. Using this standard, up to 20% of men and 5% of women have hyperCKemia.²

However, CK levels vary by sex and ethnicity, with mean levels highest in black men, followed by black women, white men, and white women. The mean level in black men is higher than the standard cutoff point for normal, and especially in this population, there is wide fluctuation around the mean, leading to hyperCKemia quite frequently in black men. Exercise and manual labor also drive up CK levels.^3–5 

Idiopathic hyperCKemia is benign. D’Adda et al⁶ followed 55 patients for a mean of 7.5 years. CK levels normalized in 12 patients or at least decreased in 24. Most remained symptom-free or had minimal symptoms. 

Idiopathic hyperCKemia: Bottom line

Before prescribing a statin, determine the baseline CK level. If slightly elevated (ie, up to 3 to 5 times the upper limit of normal, or even higher) in the setting of normal muscle strength, there is no need for electromyography or muscle biopsy, and the patient can certainly receive a statin. Most of these patients do not need to see a rheumatologist but can simply have their CK and muscle strength monitored.

Myositis (idiopathic inflammatory myopathy) is a heterogeneous group of autoimmune syndromes of unknown cause characterized by chronic muscle weakness and inflammation of striated muscle. These syndromes likely arise as a result of genetic predisposition and an environmental or infectious “hit.”

Myositis is rare, with an incidence of 5 to 10 cases per million per year and an estimated prevalence of 50 to 90 cases per million. It has 2 incidence peaks: 1 in childhood (age 5–15) and another in adult midlife (age 30–50). Women are affected 2 to 3 times more often than men, with black women most commonly affected.

Myositis is traditionally classified as follows:

• Adult polymyositis
• Adult dermatomyositis
• Juvenile myositis (dermatomyositis much more frequent than polymyositis)
• Malignancy-associated myositis (usually dermatomyositis)
• Myositis overlapping with another autoimmune disease
• Inclusion body myositis.

However, polymyositis is less common than we originally thought, and the term necrotizing myopathy is now used in many patients, as noted in the case studies below. Further, myositis overlap syndromes are being increasingly diagnosed, likely related to the emergence of autoantibodies and clinical “syndromes” associated with these autoantibody subsets (discussed in cases below).

Dermatomyositis

Dermatomyositis is characterized by muscle weakness and a rash that can be obvious or subtle. Classic skin lesions are Gottron papules, which are raised, flat-topped red or purplish lesions over the knuckles, elbows, or knees.

Lesions may be confused with those of psoriasis. There can also be a V-neck rash over the anterior chest or upper back (“shawl sign”) or a rash over the lateral thigh (“holster sign”). A facial rash may occur, but unlike lupus, dermatomyositis does not spare the nasolabial area. However, the V-neck rash can be similar to that seen in lupus.

Dermatomyositis may cause muscle pain, perhaps related to muscle ischemia, whereas polymyositis and necrotizing myopathy are often painless. However, pain is also associated with fibromyalgia, which may be seen in many autoimmune conditions. It is important not to overtreat rheumatologic diseases with immunosuppression to try to control pain if the pain is actually caused by fibromyalgia.

Polymyositis mimics

Hypothyroid myopathy can present as classic polymyositis. The serum CK may be elevated, and there may be myalgias, muscle hypertrophy with stiffness, weakness, cramps, and even features of a proximal myopathy, and rhabdomyolysis. The electromyogram can be normal or myopathic. Results of muscle biopsy are often normal but may show focal necrosis and mild inflammatory infiltrates, thus mimicking that seen with inflammatory myopathy.⁷

Drug-induced or toxic myopathies can also mimic polymyositis. Statins are among the most commonly prescribed drugs in the United States, with more than 35 million people taking them. Statins are generally well tolerated but have a broad spectrum of toxicity, ranging from myalgias to life-threatening rhabdomyolysis. Myalgias lead to about 5% to 10% of patients refusing to take a statin or stopping it on their own.

Myalgias affect up to 20% of statin users in clinical practice.^8,9 A small cross-sectional study¹⁰ of 1,000 patients in a primary care setting found that the risk of muscle complaints in statin users was 1.5 times higher than in nonstatin users, similar to findings in other studies.

My strategy for managing a patient with possible statin-induced myopathy is illustrated in Figure 1.

CASE 3: WEAKNESS, VERY HIGH CK ON A STATIN

In March 2010, a 67-year-old woman presented with muscle weakness. She had a history of hypertension, hyperlipidemia, and, more than 10 years previously, uterine cancer. In 2004, she was given atorvastatin for dyslipidemia. Four years later, she developed lower-extremity weakness, which her doctor attributed to normal aging. A year after that, she found it difficult to walk up steps and lift her arms overhead. In June 2009, she stopped taking the atorvastatin on her own, but the weakness did not improve.

In September 2009, she returned to her doctor, who found her CK level was 6,473 U/L but believed it to be an error, so the test was repeated, with a result of 9,375 U/L. She had no rash or joint involvement.

She was admitted to the hospital and underwent muscle biopsy, which showed myonecrosis with no inflammation or vasculitis.  She was treated with prednisone 60 mg/day, and her elevated CK level and weakness improved.

Immune-mediated necrotizing myopathy associated with statins

The hallmark of necrotizing myopathy is myonecrosis without significant inflammation.¹² This pattern contrasts with that of polymyositis, which is characterized by lymphocytic inflammation.

Although statins became available in the United States in 1987, immune-mediated necrotizing myopathy associated with statins was first described only in 2010. In that report, Grable-Esposito et al¹³ described 25 patients from 2 neuromuscular centers seen between 2000 and 2008 who had elevated CK and proximal weakness during or after statin use, both of which persisted despite stopping the statin. Patients improved with immunosuppressive agents but had a relapse when steroids were stopped or tapered, a pattern typical in autoimmune disease.

Also in 2010, Christopher-Stine et al¹⁴ reported an antibody associated with necrotizing myopathy. Of 38 patients with the condition, 16 were found to have an abnormal “doublet” autoantibody recognizing 200- and 100-kDa proteins. All patients had weakness and a high CK level, and 63% had statin exposure before the weakness (this percentage increased to 83% in patients older than 50). All responded to immunosuppressive therapy, and many had a relapse when it was withdrawn.

Statins lower cholesterol by inhibiting 3-hydroxy-3-methylglutaryl-Co A reductase (HMGCR), and paradoxically, they also upregulate it. HMGCR has a molecular weight of 97 kDa. Mammen et al¹⁵ identified HMGCR as the 100-kDa target of the identified antibody and developed an enzyme-linked immunosorbent assay for it. Of 750 patients presenting to one center, only 45 (6%) had anti-HMGCR autoantibodies, but all 16 patients who had the abnormal doublet antibody tested positive for anti-HMGCR. Regenerating muscle cells express high levels of HMGCR, which may sustain the immune response after statins are discontinued.

Case 3 continued: Intravenous immunoglobulin brings improvement

In March 2010, when the 67-year-old patient presented to our myositis center, her CK level was 5,800 U/L, which increased as prednisone was tapered. She still felt weak. On examination, her muscle strength findings were deltoids 4+/5, neck flexors 4/5, and iliopsoas 3+/5. She was treated with methotrexate and azathioprine without benefit. She was next treated with intravenous immunoglobulin, and after 3 months, her strength normalized for the first time in years. Her CK level decreased but did not normalize. Testing showed that she was positive for anti-HMGCR auto­antibody, as this test had become commercially available.

In 2015, Mammen and Tiniakou¹⁶ suggested using intravenous immunoglobulin as first-line therapy for statin-associated autoimmune necrotizing myopathy, based on experience at a single center with 3 patients who declined glucocorticoid treatment.

Necrotizing myopathy: Bottom line

Myositis overlap syndromes

Heterogeneity is the rule in myositis, and it can present with a wide variety of signs and symptoms as outlined in Table 2.

CASE 4: FEVER, NEW ‘RHEUMATOID ARTHRITIS,’ AND LUNG DISEASE

A 52-year-old woman with knee osteoarthritis saw her primary care physician in November 2013 for dyspnea and low-grade fever. The next month, she presented with polyarthritis, muscle weakness, and Raynaud phenomenon.

In January 2014, she developed acrocyanosis of her fingers. Examination revealed hyperkeratotic, cracked areas of her fingers. Her oxygen saturation by pulse oximetry was low. She was admitted to the hospital. Her doctor suspected new onset of rheumatoid arthritis, but blood tests revealed a negative antinuclear antibody, so an autoimmune condition was deemed unlikely. Her CK was mildly elevated at 350 U/L.

Because of her dyspnea, an open-lung biopsy was performed. High-resolution computed tomography (CT) revealed infiltrates and ground-glass opacities, leading to the diagnosis of nonspecific interstitial pneumonia. A rheumatologist was consulted and recommended pulse methylprednisolone, followed by prednisone 60 mg/day and mycophenolate mofetil. Testing for Jo-1 antibodies was positive.

Antisynthetase syndrome

The antisynthetase syndrome is a clinically heterogeneous condition that can occur with any or all of the following:

• Fever
• Myositis
• Arthritis (often misdiagnosed as rheumatoid arthritis)
• Raynaud phenomenon
• Mechanic’s hands (hyperkeratotic  roughness with fissures on the lateral aspects of the fingers and finger pads)
• Interstitial lung disease.

The skin rashes and myositis may be subtle, making the presentation “lung-dominant,” and nonrheumatologists should be aware of this syndrome. Although in our patient the condition developed in a classic manner, with all of the aforementioned features of the antisynthetase syndrome, some patients will manifest one or a few of the features.

Clinically, patients with the Jo-1 antisynthetase syndrome often present differently than those with non-Jo-1 antisynthetase autoantibodies. When we compared 122 patients with Jo-1 vs 80 patients with a non-Jo-1 antisynthetase autoantibody, patients with Jo-1 antibodies were more likely to have initially received a diagnosis of myositis (83%), while  myositis was the original diagnosis in only 17% of those possessing non-Jo-1 antisynthetase autoantibodies. In fact, many patients (approximately 50%) were diagnosed as having undifferentiated connective tissue disease or an overlap syndrome, and 13% had scleroderma as their first diagnosis.¹⁷

We also found that the survival rate was higher in patients with Jo-1 syndrome compared with patients with non-Jo-1 antisynthetase syndromes. We attributed the difference in survival rates to a delayed diagnosis in the non-Jo-1 group, perhaps due to their “nonclassic” presentations of the antisynthetase syndrome, delaying appropriate treatment. Patients received a diagnosis of Jo-1 antibody syndrome after a mean of 0.4 year (range 0.2–0.8), while those with a non-Jo-1 antisynthetase autoantibody had a delay in diagnosis of 1.0 year (range 0.4–5.1) (P < .01).¹⁷

In nearly half the cases in this cohort, pulmonary fibrosis was the cause of death, with primary pulmonary hypertension being the second leading cause (11%).

Antisynthetase syndrome: Bottom line

Antisynthetase syndrome is an often fatal disease that does not always present in a typical fashion with symptoms of myositis, as lung disease may be the predominant feature. A negative antinuclear antibody test result does not imply antibody negativity, as the autoantigen in these diseases is not located in the nucleus. Prompt diagnosis and appropriate immunosuppressive therapy are critical to improving outcomes.

CASE 5: FEVER, UNDIAGNOSED LUNG DISEASE, NO MYOSITIS

In January 2001, a 39-year-old woman was admitted to the hospital after 5 weeks of fever (temperatures 103°–104°F) and myalgias. An extensive workup was negative except for low-titer antinuclear antibody and for mild basilar fibrosis noted on chest radiography. She left the hospital against medical advice because of frustration with a lack of a specific diagnosis (“fever of unknown origin”).

Two months later, at a follow-up rheumatology consult, she reported more myalgias and arthralgias, as well as fever. Chest radiography now showed pleural effusions. Her fingers had color changes consistent with Raynaud phenomenon. At that time, I diagnosed an undifferentiated connective tissue disease and told her that I suspected an autoimmune condition that would need time to reveal itself. In the meantime, I treated her empirically with prednisone.

In April, she returned, much more short of breath and with more prominent diffuse pulmonary infiltrates. Physical examination revealed subtle Gottron changes. Testing revealed poor pulmonary function: forced vital capacity (FVC) 56%, forced expiratory volume in 1 second (FEV₁) 52%, and diffusing capacity for carbon monoxide (Dlco) 40%. Blood testing was positive for anti-PL-12 antibody, one of the non-Jo-1 antisynthetase antibodies. At this time, we treated her with glucocorticoids and tacrolimus.

More than 15 years later, this patient is doing well. Her skin rash, joint symptoms, and fever have not returned, and interestingly, she never developed myositis. Her Raynaud symptoms are mild. Her most recent pulmonary function test results (January 2018) were FVC  75%, FEV₁ 87%, and Dlco 78%. Although these results are not normal, they are much improved and allow her to be completely functional without supplemental oxygen. Echocardiography showed  normal pulmonary artery systolic pressure (25 mm Hg). She was still taking tacrolimus and prednisone. When we tried to stop tacrolimus after she had done well for many years, her condition flared.

Non-Jo-1 antisynthetase syndrome: Bottom line

Patients with a non-Jo-1 antisynthetase syndrome often present without myositis symptoms and may never manifest myositis symptoms. Likely because of this presentation, diagnosis of a specific connective tissue disorder is delayed, perhaps leading to increased mortality risk from pulmonary disease. Chronic immunosuppression is often required for these autoimmune conditions.

CASE 6: DERMATOMYOSITIS, RAPIDLY PROGRESSIVE INTERSTITIAL LUNG DISEASE

A 58-year-old woman presented in the summer of 2012 with a photosensitive rash. The following January, she returned with polyarthritis, mild muscle weakness, and a dermatomyositis-pattern rash. Her CK level was normal, and her antinuclear antibody and Sjögren syndrome antibody test results were negative. She improved on low-dose prednisone and methotrexate.

She was originally referred to me in May of that year for worsening rash and mild weakness. She denied pulmonary symptoms, but examination revealed faint basilar crackles. I increased her prednisone dosage to 20 mg/day and started mycophenolate mofetil mainly for the mild cutaneous and myositis features. I also recommended high-resolution CT of the lungs and pulmonary function tests, which she underwent in early June. High-resolution CT showed nonspecific mild infiltrates with minimal ground-glass opacities.

On July 1, she presented to her local emergency department with severe shortness of breath, requiring oxygen 12 L/min. She had a palmar rash. Repeat high-resolution CT showed dramatic worsening compared with the scan the previous month. Because of continued inadequate oxygenation, she was transferred to our center. A blood test later was positive for antimelanoma differentiation-associated gene 5 (MDA-5) autoantibody, previously known as anticlinically amyopathic dermatomyositis (anti-CADM)-140 antibody (based on immunoprecipitation results).

She died on the third day after transfer, just 2 months after I had originally seen her, at which time she had had no pulmonary symptoms.

Clinically amyopathic dermatomyositis

Anti-CADM-140, first reported from Asia,^18–20 is an autoantibody-associated disease but not an antisynthetase. It is associated with dermatomyositis; patients often have a “vasculopathy” with cutaneous ulcerations and palmar papules.

MDA-5 is a cytoplasmic protein that “senses” viral RNA and induces production of type 1 interferon. It is involved in the innate immune defense against viruses.

Anti-MDA-5 positivity is associated with a poor pulmonary outcome.²¹ In our cohort from the University of Pittsburgh, many patients died within 3 years, compared with about a 40% survival rate in patients with dermatomyositis who tested negative for this antibody. That being said, many patients with anti-MDA-5 do not develop rapidly progressive interstitial lung disease.

Autoimmune interstitial lung disease: Bottom line

Autoimmune interstitial lung disease is easy to miss, especially in the case of a non-Jo-1 syndrome, for 3 important reasons:

• The autoimmune features may initially be subtle (eg, Raynaud phenomena, mild dermatomyositis rash, undifferentiated connective tissue disease)
• Autoantibody testing is not often ordered, is not standardized, or may be unavailable 
• Providers are mistakenly reassured that a patient who tests negative for antinuclear antibody does not have an autoimmune condition.

To emphasize the last point, in a cohort of 202 patients who tested positive for an antisynthetase antibody, only half were antinuclear antibody-positive, but nearly three-quarters demonstrated anticytoplasmic staining on indirect immunofluorescence (due to the location of the autoantigen in the cytoplasm), making the latter a better screening test for an antisynthetase antibody. For scleroderma, 99% were antinculear antibody-positive, but for myositis, this test is much less sensitive.²²

Click here to read Hot Topics in Primary Care.

This supplement includes 4 CME credits (scroll down for more information). 

Topics include: 

• Chronic Kidney Disease in Type 2 Diabetes
• Heart Failure in Type 2 Diabetes
• Diabetes Management
• Naproxen vs Opioids
• Statin Selection
• Insomnia in Alzheimer's Disease
• Cluster Headache
• Irritable Bowel Syndrome

This supplement offers the opportunity to earn a total of 4 CME credits. 

Credit is awarded for scucesful completion of the online evaluations at the links below. These links may also be found within the supplement on the first page of each article. 

• Chronic Kidney Disease in Type 2 Diabetes: Optimizing Glucose-Lowering Therapy
  • To receive CME credit, please read the journal article and, upon completion, go to www.pceconsortium.org/DKD to complete the online post-test and receive your certificate of credit.
• Diabetes Management Update: Individualizing Treatment
  • To receive CME credit, please read the journal article and, upon completion, go to www.pceconsortium.org/DM to complete the online post-test and receive your certificate of credit.
• Identification and Management of Insomnia in Alzheimer's Disease
  • To receive CME credit, please read the journal article and, upon completion, go to www.pceconsortium.org/insomnia to complete the online post-test and receive your certificate of credit. 
• Patient-Centric Care of Diarrhea-Predominant Irritable Bowel Syndrome
  • To receive CME credit, please read the journal article and, upon completion, go to www.pceconsortium.org/IBSD to complete the online post-test and receive your certificate of credit.

Click here to read Hot Topics in Primary Care.

This supplement includes 4 CME credits (scroll down for more information). 

Topics include: 

• Chronic Kidney Disease in Type 2 Diabetes
• Heart Failure in Type 2 Diabetes
• Diabetes Management
• Naproxen vs Opioids
• Statin Selection
• Insomnia in Alzheimer's Disease
• Cluster Headache
• Irritable Bowel Syndrome

This supplement offers the opportunity to earn a total of 4 CME credits. 

Credit is awarded for scucesful completion of the online evaluations at the links below. These links may also be found within the supplement on the first page of each article. 

• Chronic Kidney Disease in Type 2 Diabetes: Optimizing Glucose-Lowering Therapy
  • To receive CME credit, please read the journal article and, upon completion, go to www.pceconsortium.org/DKD to complete the online post-test and receive your certificate of credit.
• Diabetes Management Update: Individualizing Treatment
  • To receive CME credit, please read the journal article and, upon completion, go to www.pceconsortium.org/DM to complete the online post-test and receive your certificate of credit.
• Identification and Management of Insomnia in Alzheimer's Disease
  • To receive CME credit, please read the journal article and, upon completion, go to www.pceconsortium.org/insomnia to complete the online post-test and receive your certificate of credit. 
• Patient-Centric Care of Diarrhea-Predominant Irritable Bowel Syndrome
  • To receive CME credit, please read the journal article and, upon completion, go to www.pceconsortium.org/IBSD to complete the online post-test and receive your certificate of credit.

Click here to read Hot Topics in Primary Care.

This supplement includes 4 CME credits (scroll down for more information). 

Topics include: 

• Chronic Kidney Disease in Type 2 Diabetes
• Heart Failure in Type 2 Diabetes
• Diabetes Management
• Naproxen vs Opioids
• Statin Selection
• Insomnia in Alzheimer's Disease
• Cluster Headache
• Irritable Bowel Syndrome

This supplement offers the opportunity to earn a total of 4 CME credits. 

Credit is awarded for scucesful completion of the online evaluations at the links below. These links may also be found within the supplement on the first page of each article. 

• Chronic Kidney Disease in Type 2 Diabetes: Optimizing Glucose-Lowering Therapy
  • To receive CME credit, please read the journal article and, upon completion, go to www.pceconsortium.org/DKD to complete the online post-test and receive your certificate of credit.
• Diabetes Management Update: Individualizing Treatment
  • To receive CME credit, please read the journal article and, upon completion, go to www.pceconsortium.org/DM to complete the online post-test and receive your certificate of credit.
• Identification and Management of Insomnia in Alzheimer's Disease
  • To receive CME credit, please read the journal article and, upon completion, go to www.pceconsortium.org/insomnia to complete the online post-test and receive your certificate of credit. 
• Patient-Centric Care of Diarrhea-Predominant Irritable Bowel Syndrome
  • To receive CME credit, please read the journal article and, upon completion, go to www.pceconsortium.org/IBSD to complete the online post-test and receive your certificate of credit.