Guidelines for diagnosing Alzheimer’s disease (AD) are undergoing the first major changes since they were developed 30 years ago. The National Institute on Aging (NIA) and the Alzheimer’s Association (AA) have established workgroups to revise guidelines that were written in 1984.¹

One of the major changes to these new guidelines is men­tion of research on biomarkers for diagnosing and monitoring progression of dementia in AD. This is an exciting and pro­vocative development, but the questions practitioners who diagnose and treat AD should be asking are whether such biomarkers have utility in clinical practice today, or whether their application is a distant promise of continuing research.

Principles put forward in the guidelines
The new AD guidelines set forth in 3 major papers by the workgroups created by the NIA and AA include a change in nomenclature of AD.² The workgroups have sought to define AD with specific stages that include:
   • a preclinical/prodromal phase, in which the pathophysi­ology responsible for future cognitive changes is ongoing but lacks clinical manifestations³
   • mild cognitive impairment, now considered a distinct entity from dementia and diagnosed when a person has early signs of AD; manifestations of impaired cognition in early dis­ease are not significant enough to affect daily functioning.⁴

These newly formulated stages of AD rely on clinical judg­ment, and AD remains a clinical diagnosis. However, the new diagnostic guidelines include the use of biomarkers to mea­sure disease progression.

Biomarkers of normal biologic function and pathology
The Biomarkers Definitions Working Group defines a biomarker as:
 
… a characteristic that is objectively measured and evaluated as an indica­tor of  normal biologic processes, patho­genic processes, or pharmacologic responses to a therapeutic intervention.⁵

These characteristics include imaging studies and body fluids, such as serum and cerebrospinal fluid (CSF).

In AD, biomarkers are meant to mea­sure the pathogenic processes of:
   • accumulation and deposition of amy­loid β _protein (Aβ42) plaques
   • neuronal degeneration characterized by an increase in phosphorylated tau protein and neurofibrillary tangles.⁶

The purpose of these biomarkers is to identify ongoing disease and help the cli­nician stage patients who display a spec­trum of symptoms.

Four classes of biomarkers (Table⁷)have been identified for use in the diagnosis of, and research on, AD:
   • neuroimaging
   • CSF
   • serum
   • genetic markers.

FDG PET identifies areas of the brain in which glucose metabolism is decreased. This finding is thought to represent syn­aptic dysfunction.⁸ The true clinical utility of FDG PET appears to be as an aid in dis­tinguishing cases of AD from frontotem­poral dementia, by identifying regions of metabolic dysfunction.⁹ (Note: Medicare will reimburse for FDG PET only if 1) the patient has met diagnostic criteria for both AD and frontotemporal dementia for at least 6 months and 2) the cause of symp­toms is uncertain.¹⁰)

FDG PET also can be useful in patients with mild cognitive impairment by identi­fying hypometabolism in the temporal and parietal regions of the brain years before clin­ical AD develops.In addition to FDG, 2 other imag­ing probes—Pittsburgh compound and 2-(1-{6-[(2-[fluorine-18]fluoroethyl)(methyl) amino]-2-naphthyl}-ethylidene) malononi­trile (more commonly, FDDNP)—have been used with PET as research tools to demon­strate evidence of AD.¹¹

MRI has been used to measure hippocampal atrophy and cortical thinning that occurs as a patient progresses from normal cognitive function or mild cognitive impairment to full dementia.⁵ The degree of atrophy has not been well correlated with the degree of func­tional impairment.

Florbetapir F 18 Injection was approved by the FDA in October 2013, under the brand name AMYViD, for measuring the quan­tity of Aβ42 deposition in the brain. When injected, this radiopharmaceutical binds to Aβ42 and can be detected on PET.¹² Use cri­teria for AMYViD PET recently were devel­oped¹³; the technique is indicated as an additional diagnostic tool for ruling out AD.

A negative AMYViD scan indicates sparse or no Aβ42 plaques, and is inconsistent with AD. However, a positive AMYViD scan does not establish a diagnosis of AD or other cognitive disorder.¹⁴ This lack of specificity decreases the potential utility of the scan in clinical practice.

Use of AMYViD PET in general practice also is constrained by cost, which varies by location, based on the fee for the PET scan ($1,000 to $3,000)¹⁵; to that, add the cost of a dose of AMYViD ($1,600, wholesale).¹⁶ The technique is not reimbursable, and the total out-of-pocket expense can be as much as $5,000—making an AMYViD PET prohibitive.

Cerebrospinal fluid markers
CSF biomarkers used in the evaluation of AD are Aβ42, t-tau protein, and p-tau protein.^6,17 It is generally thought that the level of Aβ42 in CSF decreases in AD—indicative of Aβ42 being deposited in the brain.8 Tau proteins are elevated in CSF as neurons are destroyed. P-tau is associated with the neurofibrillary tangles of AD; its presence in CSF is thought to represent an increase in those tangles. The combination of a low level of Aβ42 and an elevated level of p-tau in CSF is considered the signature CSF biomarker of AD.⁶

Serum markers
The search for reliable serum biomarkers of AD is the area of greatest research interest because a blood test is a less invasive form of screening. Regrettably, the utility of serum biomarkers for clinical practice has not been established.

Aβ42 can be measured in serum, but levels do not correlate well with CSF levels.¹⁸ Other serum markers that have been evaluated for clinical utility include measures of lipid metabolism, oxidation, and inflammation. With none of these is there clear correlation between the level of protein and AD.¹⁸

Fourth front: Genetics
Several alleles are associated with AD. Mutations in amyloid precursor protein, presenilin 1, and presenilin 2 have been shown to cause a change in the process­ing of Aβ42 and thus lead to AD.¹⁹ These mutations are inherited in an autosomal-dominant fashion and are detected in early-onset (age <65) AD.

Mutations in apolipoprotein 4-β4 also has been the subject of much research; this allele usually is associated with increased risk of the more common, later-onset AD.²⁰ Some evidence suggests that apolipoprotein 4-β4 carriers who develop AD might be at risk of earlier onset of symptoms, compared to non­carriers,²¹ but the clinical significance of that increased risk has not been established.

What utility do biomarkers have?
As we said at the beginning of this article, the question that clinicians should be asking is: “What is the current clinical utility of these sophisticated biomarkers and genetic testing?”

The answer is “little utility.” Diagnosing AD is a clinical enterprise, with, as we’ve out­lined, specific and narrow exceptions.

Recently, researchers demonstrated bio­marker evidence of AD before symptom onset in patients who have known autosomal-dominant gene mutations for AD.¹⁹ There is no evidence, however, that these biomarkers are useful for screening the general popula­tion to identify people who 1) are at risk of, or who have, AD and 2) do not have AD.

That being said, CSF and imaging bio­markers of AD are being used in clinical settings in some European countries to aid investigation of cognitive decline.

In conclusion
Here are key points to take away from this discussion of biomarkers of AD:
   • The utility of these biomarkers today is in research—although some of them might, on occasion, be useful to distinguish demen­tia caused by AD from other dementias.
   • The ultimate goal of research is to uncover a serum biomarker that can iden­tify patients in the preclinical/prodromal stage of AD, so that disease-modifying therapies and preventive measures can be initiated before symptoms manifest.
   • Science is a long way from making this goal a reality, but recent changes in the diagnostic criteria for AD will encourage research in this area of study.

Bottom Line
Researchers are working to uncover biomarkers that will identify patients in the preclinical or prodromal stage of Alzheimer’s disease, but diagnosis remains clinical. Recent changes to diagnostic criteria will encourage research in this area.

Related Resources
• Blennow K, Dubois B, Fagan AM, et al. Clinical utility of cere­brospinal fluid biomarkers in the diagnosis of early Alzheimer’s disease [published online May 5, 2014]. Alzheimers Dement. doi: 10.1016/j.jalz.2014.02.004.
• Chase A. Alzheimer disease: Advances in imaging of AD biomarkers could aid early diagnosis. Nat Rev Neurol. 2014;10(5):239.
• De Riva V, Galloni E, Marcon M, et al. Analysis of combined CSF biomarkers in AD diagnosis. Clin Lab. 2014;60(4):629-634.
• Kristofikova Z, Ricny J, Kolarova M, et al. Interactions between amyloid-β and tau in cerebrospinal fluid of people with mild cognitive impairment and Alzheimer’s disease [pub­lished online March 26, 2014]. J Alzheimers Dis. doi: 10.3233/ JAD-132393.

Drug Brand Name
Florbetapir F 18 Injection • AMYViD

Disclosures
The authors report no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Guidelines for diagnosing Alzheimer’s disease (AD) are undergoing the first major changes since they were developed 30 years ago. The National Institute on Aging (NIA) and the Alzheimer’s Association (AA) have established workgroups to revise guidelines that were written in 1984.¹

One of the major changes to these new guidelines is men­tion of research on biomarkers for diagnosing and monitoring progression of dementia in AD. This is an exciting and pro­vocative development, but the questions practitioners who diagnose and treat AD should be asking are whether such biomarkers have utility in clinical practice today, or whether their application is a distant promise of continuing research.

Principles put forward in the guidelines
The new AD guidelines set forth in 3 major papers by the workgroups created by the NIA and AA include a change in nomenclature of AD.² The workgroups have sought to define AD with specific stages that include:
   • a preclinical/prodromal phase, in which the pathophysi­ology responsible for future cognitive changes is ongoing but lacks clinical manifestations³
   • mild cognitive impairment, now considered a distinct entity from dementia and diagnosed when a person has early signs of AD; manifestations of impaired cognition in early dis­ease are not significant enough to affect daily functioning.⁴

These newly formulated stages of AD rely on clinical judg­ment, and AD remains a clinical diagnosis. However, the new diagnostic guidelines include the use of biomarkers to mea­sure disease progression.

Biomarkers of normal biologic function and pathology
The Biomarkers Definitions Working Group defines a biomarker as:
 
… a characteristic that is objectively measured and evaluated as an indica­tor of  normal biologic processes, patho­genic processes, or pharmacologic responses to a therapeutic intervention.⁵

These characteristics include imaging studies and body fluids, such as serum and cerebrospinal fluid (CSF).

In AD, biomarkers are meant to mea­sure the pathogenic processes of:
   • accumulation and deposition of amy­loid β _protein (Aβ42) plaques
   • neuronal degeneration characterized by an increase in phosphorylated tau protein and neurofibrillary tangles.⁶

The purpose of these biomarkers is to identify ongoing disease and help the cli­nician stage patients who display a spec­trum of symptoms.

Four classes of biomarkers (Table⁷)have been identified for use in the diagnosis of, and research on, AD:
   • neuroimaging
   • CSF
   • serum
   • genetic markers.

FDG PET identifies areas of the brain in which glucose metabolism is decreased. This finding is thought to represent syn­aptic dysfunction.⁸ The true clinical utility of FDG PET appears to be as an aid in dis­tinguishing cases of AD from frontotem­poral dementia, by identifying regions of metabolic dysfunction.⁹ (Note: Medicare will reimburse for FDG PET only if 1) the patient has met diagnostic criteria for both AD and frontotemporal dementia for at least 6 months and 2) the cause of symp­toms is uncertain.¹⁰)

FDG PET also can be useful in patients with mild cognitive impairment by identi­fying hypometabolism in the temporal and parietal regions of the brain years before clin­ical AD develops.In addition to FDG, 2 other imag­ing probes—Pittsburgh compound and 2-(1-{6-[(2-[fluorine-18]fluoroethyl)(methyl) amino]-2-naphthyl}-ethylidene) malononi­trile (more commonly, FDDNP)—have been used with PET as research tools to demon­strate evidence of AD.¹¹

MRI has been used to measure hippocampal atrophy and cortical thinning that occurs as a patient progresses from normal cognitive function or mild cognitive impairment to full dementia.⁵ The degree of atrophy has not been well correlated with the degree of func­tional impairment.

Florbetapir F 18 Injection was approved by the FDA in October 2013, under the brand name AMYViD, for measuring the quan­tity of Aβ42 deposition in the brain. When injected, this radiopharmaceutical binds to Aβ42 and can be detected on PET.¹² Use cri­teria for AMYViD PET recently were devel­oped¹³; the technique is indicated as an additional diagnostic tool for ruling out AD.

A negative AMYViD scan indicates sparse or no Aβ42 plaques, and is inconsistent with AD. However, a positive AMYViD scan does not establish a diagnosis of AD or other cognitive disorder.¹⁴ This lack of specificity decreases the potential utility of the scan in clinical practice.

Use of AMYViD PET in general practice also is constrained by cost, which varies by location, based on the fee for the PET scan ($1,000 to $3,000)¹⁵; to that, add the cost of a dose of AMYViD ($1,600, wholesale).¹⁶ The technique is not reimbursable, and the total out-of-pocket expense can be as much as $5,000—making an AMYViD PET prohibitive.

Cerebrospinal fluid markers
CSF biomarkers used in the evaluation of AD are Aβ42, t-tau protein, and p-tau protein.^6,17 It is generally thought that the level of Aβ42 in CSF decreases in AD—indicative of Aβ42 being deposited in the brain.8 Tau proteins are elevated in CSF as neurons are destroyed. P-tau is associated with the neurofibrillary tangles of AD; its presence in CSF is thought to represent an increase in those tangles. The combination of a low level of Aβ42 and an elevated level of p-tau in CSF is considered the signature CSF biomarker of AD.⁶

Serum markers
The search for reliable serum biomarkers of AD is the area of greatest research interest because a blood test is a less invasive form of screening. Regrettably, the utility of serum biomarkers for clinical practice has not been established.

Aβ42 can be measured in serum, but levels do not correlate well with CSF levels.¹⁸ Other serum markers that have been evaluated for clinical utility include measures of lipid metabolism, oxidation, and inflammation. With none of these is there clear correlation between the level of protein and AD.¹⁸

Fourth front: Genetics
Several alleles are associated with AD. Mutations in amyloid precursor protein, presenilin 1, and presenilin 2 have been shown to cause a change in the process­ing of Aβ42 and thus lead to AD.¹⁹ These mutations are inherited in an autosomal-dominant fashion and are detected in early-onset (age <65) AD.

Mutations in apolipoprotein 4-β4 also has been the subject of much research; this allele usually is associated with increased risk of the more common, later-onset AD.²⁰ Some evidence suggests that apolipoprotein 4-β4 carriers who develop AD might be at risk of earlier onset of symptoms, compared to non­carriers,²¹ but the clinical significance of that increased risk has not been established.

What utility do biomarkers have?
As we said at the beginning of this article, the question that clinicians should be asking is: “What is the current clinical utility of these sophisticated biomarkers and genetic testing?”

The answer is “little utility.” Diagnosing AD is a clinical enterprise, with, as we’ve out­lined, specific and narrow exceptions.

Recently, researchers demonstrated bio­marker evidence of AD before symptom onset in patients who have known autosomal-dominant gene mutations for AD.¹⁹ There is no evidence, however, that these biomarkers are useful for screening the general popula­tion to identify people who 1) are at risk of, or who have, AD and 2) do not have AD.

That being said, CSF and imaging bio­markers of AD are being used in clinical settings in some European countries to aid investigation of cognitive decline.

In conclusion
Here are key points to take away from this discussion of biomarkers of AD:
   • The utility of these biomarkers today is in research—although some of them might, on occasion, be useful to distinguish demen­tia caused by AD from other dementias.
   • The ultimate goal of research is to uncover a serum biomarker that can iden­tify patients in the preclinical/prodromal stage of AD, so that disease-modifying therapies and preventive measures can be initiated before symptoms manifest.
   • Science is a long way from making this goal a reality, but recent changes in the diagnostic criteria for AD will encourage research in this area of study.

Bottom Line
Researchers are working to uncover biomarkers that will identify patients in the preclinical or prodromal stage of Alzheimer’s disease, but diagnosis remains clinical. Recent changes to diagnostic criteria will encourage research in this area.

Related Resources
• Blennow K, Dubois B, Fagan AM, et al. Clinical utility of cere­brospinal fluid biomarkers in the diagnosis of early Alzheimer’s disease [published online May 5, 2014]. Alzheimers Dement. doi: 10.1016/j.jalz.2014.02.004.
• Chase A. Alzheimer disease: Advances in imaging of AD biomarkers could aid early diagnosis. Nat Rev Neurol. 2014;10(5):239.
• De Riva V, Galloni E, Marcon M, et al. Analysis of combined CSF biomarkers in AD diagnosis. Clin Lab. 2014;60(4):629-634.
• Kristofikova Z, Ricny J, Kolarova M, et al. Interactions between amyloid-β and tau in cerebrospinal fluid of people with mild cognitive impairment and Alzheimer’s disease [pub­lished online March 26, 2014]. J Alzheimers Dis. doi: 10.3233/ JAD-132393.

Drug Brand Name
Florbetapir F 18 Injection • AMYViD

Disclosures
The authors report no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Guidelines for diagnosing Alzheimer’s disease (AD) are undergoing the first major changes since they were developed 30 years ago. The National Institute on Aging (NIA) and the Alzheimer’s Association (AA) have established workgroups to revise guidelines that were written in 1984.¹

One of the major changes to these new guidelines is men­tion of research on biomarkers for diagnosing and monitoring progression of dementia in AD. This is an exciting and pro­vocative development, but the questions practitioners who diagnose and treat AD should be asking are whether such biomarkers have utility in clinical practice today, or whether their application is a distant promise of continuing research.

Principles put forward in the guidelines
The new AD guidelines set forth in 3 major papers by the workgroups created by the NIA and AA include a change in nomenclature of AD.² The workgroups have sought to define AD with specific stages that include:
   • a preclinical/prodromal phase, in which the pathophysi­ology responsible for future cognitive changes is ongoing but lacks clinical manifestations³
   • mild cognitive impairment, now considered a distinct entity from dementia and diagnosed when a person has early signs of AD; manifestations of impaired cognition in early dis­ease are not significant enough to affect daily functioning.⁴

These newly formulated stages of AD rely on clinical judg­ment, and AD remains a clinical diagnosis. However, the new diagnostic guidelines include the use of biomarkers to mea­sure disease progression.

Biomarkers of normal biologic function and pathology
The Biomarkers Definitions Working Group defines a biomarker as:
 
… a characteristic that is objectively measured and evaluated as an indica­tor of  normal biologic processes, patho­genic processes, or pharmacologic responses to a therapeutic intervention.⁵

These characteristics include imaging studies and body fluids, such as serum and cerebrospinal fluid (CSF).

In AD, biomarkers are meant to mea­sure the pathogenic processes of:
   • accumulation and deposition of amy­loid β _protein (Aβ42) plaques
   • neuronal degeneration characterized by an increase in phosphorylated tau protein and neurofibrillary tangles.⁶

The purpose of these biomarkers is to identify ongoing disease and help the cli­nician stage patients who display a spec­trum of symptoms.

Four classes of biomarkers (Table⁷)have been identified for use in the diagnosis of, and research on, AD:
   • neuroimaging
   • CSF
   • serum
   • genetic markers.

FDG PET identifies areas of the brain in which glucose metabolism is decreased. This finding is thought to represent syn­aptic dysfunction.⁸ The true clinical utility of FDG PET appears to be as an aid in dis­tinguishing cases of AD from frontotem­poral dementia, by identifying regions of metabolic dysfunction.⁹ (Note: Medicare will reimburse for FDG PET only if 1) the patient has met diagnostic criteria for both AD and frontotemporal dementia for at least 6 months and 2) the cause of symp­toms is uncertain.¹⁰)

FDG PET also can be useful in patients with mild cognitive impairment by identi­fying hypometabolism in the temporal and parietal regions of the brain years before clin­ical AD develops.In addition to FDG, 2 other imag­ing probes—Pittsburgh compound and 2-(1-{6-[(2-[fluorine-18]fluoroethyl)(methyl) amino]-2-naphthyl}-ethylidene) malononi­trile (more commonly, FDDNP)—have been used with PET as research tools to demon­strate evidence of AD.¹¹

MRI has been used to measure hippocampal atrophy and cortical thinning that occurs as a patient progresses from normal cognitive function or mild cognitive impairment to full dementia.⁵ The degree of atrophy has not been well correlated with the degree of func­tional impairment.

Florbetapir F 18 Injection was approved by the FDA in October 2013, under the brand name AMYViD, for measuring the quan­tity of Aβ42 deposition in the brain. When injected, this radiopharmaceutical binds to Aβ42 and can be detected on PET.¹² Use cri­teria for AMYViD PET recently were devel­oped¹³; the technique is indicated as an additional diagnostic tool for ruling out AD.

A negative AMYViD scan indicates sparse or no Aβ42 plaques, and is inconsistent with AD. However, a positive AMYViD scan does not establish a diagnosis of AD or other cognitive disorder.¹⁴ This lack of specificity decreases the potential utility of the scan in clinical practice.

Use of AMYViD PET in general practice also is constrained by cost, which varies by location, based on the fee for the PET scan ($1,000 to $3,000)¹⁵; to that, add the cost of a dose of AMYViD ($1,600, wholesale).¹⁶ The technique is not reimbursable, and the total out-of-pocket expense can be as much as $5,000—making an AMYViD PET prohibitive.

Cerebrospinal fluid markers
CSF biomarkers used in the evaluation of AD are Aβ42, t-tau protein, and p-tau protein.^6,17 It is generally thought that the level of Aβ42 in CSF decreases in AD—indicative of Aβ42 being deposited in the brain.8 Tau proteins are elevated in CSF as neurons are destroyed. P-tau is associated with the neurofibrillary tangles of AD; its presence in CSF is thought to represent an increase in those tangles. The combination of a low level of Aβ42 and an elevated level of p-tau in CSF is considered the signature CSF biomarker of AD.⁶

Serum markers
The search for reliable serum biomarkers of AD is the area of greatest research interest because a blood test is a less invasive form of screening. Regrettably, the utility of serum biomarkers for clinical practice has not been established.

Aβ42 can be measured in serum, but levels do not correlate well with CSF levels.¹⁸ Other serum markers that have been evaluated for clinical utility include measures of lipid metabolism, oxidation, and inflammation. With none of these is there clear correlation between the level of protein and AD.¹⁸

Fourth front: Genetics
Several alleles are associated with AD. Mutations in amyloid precursor protein, presenilin 1, and presenilin 2 have been shown to cause a change in the process­ing of Aβ42 and thus lead to AD.¹⁹ These mutations are inherited in an autosomal-dominant fashion and are detected in early-onset (age <65) AD.

Mutations in apolipoprotein 4-β4 also has been the subject of much research; this allele usually is associated with increased risk of the more common, later-onset AD.²⁰ Some evidence suggests that apolipoprotein 4-β4 carriers who develop AD might be at risk of earlier onset of symptoms, compared to non­carriers,²¹ but the clinical significance of that increased risk has not been established.

What utility do biomarkers have?
As we said at the beginning of this article, the question that clinicians should be asking is: “What is the current clinical utility of these sophisticated biomarkers and genetic testing?”

The answer is “little utility.” Diagnosing AD is a clinical enterprise, with, as we’ve out­lined, specific and narrow exceptions.

Recently, researchers demonstrated bio­marker evidence of AD before symptom onset in patients who have known autosomal-dominant gene mutations for AD.¹⁹ There is no evidence, however, that these biomarkers are useful for screening the general popula­tion to identify people who 1) are at risk of, or who have, AD and 2) do not have AD.

That being said, CSF and imaging bio­markers of AD are being used in clinical settings in some European countries to aid investigation of cognitive decline.

In conclusion
Here are key points to take away from this discussion of biomarkers of AD:
   • The utility of these biomarkers today is in research—although some of them might, on occasion, be useful to distinguish demen­tia caused by AD from other dementias.
   • The ultimate goal of research is to uncover a serum biomarker that can iden­tify patients in the preclinical/prodromal stage of AD, so that disease-modifying therapies and preventive measures can be initiated before symptoms manifest.
   • Science is a long way from making this goal a reality, but recent changes in the diagnostic criteria for AD will encourage research in this area of study.

Bottom Line
Researchers are working to uncover biomarkers that will identify patients in the preclinical or prodromal stage of Alzheimer’s disease, but diagnosis remains clinical. Recent changes to diagnostic criteria will encourage research in this area.

Related Resources
• Blennow K, Dubois B, Fagan AM, et al. Clinical utility of cere­brospinal fluid biomarkers in the diagnosis of early Alzheimer’s disease [published online May 5, 2014]. Alzheimers Dement. doi: 10.1016/j.jalz.2014.02.004.
• Chase A. Alzheimer disease: Advances in imaging of AD biomarkers could aid early diagnosis. Nat Rev Neurol. 2014;10(5):239.
• De Riva V, Galloni E, Marcon M, et al. Analysis of combined CSF biomarkers in AD diagnosis. Clin Lab. 2014;60(4):629-634.
• Kristofikova Z, Ricny J, Kolarova M, et al. Interactions between amyloid-β and tau in cerebrospinal fluid of people with mild cognitive impairment and Alzheimer’s disease [pub­lished online March 26, 2014]. J Alzheimers Dis. doi: 10.3233/ JAD-132393.

Drug Brand Name
Florbetapir F 18 Injection • AMYViD

Disclosures
The authors report no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Dr. Henry A. Nasrallah’s recent Editorial on borderline personality disorder (BPD) (Current Psychiatry, From the Editor, April 2014, p. 19-20, 32 [http://bit.ly/1e8yAwE]) describes BPD as a heritable brain disease. I have been arguing this point for many years, often finding support from my colleague, Hagop Akiskal, MD, and opposition from my psychoanalytic colleagues.

In recent papers^1,2 on brain changes in BPD and the connection between BPD and bipolar disorders, I wrote that there often is a heritable aspect to the condition. There are exceptions to such heritability, as in the setting of a horrific environment (eg, father-daughter incest, parental brutality), where the same symptoms seen in BPD develop primarily from post-natal influences. Dr. Akiskal and I were discussing this a long time back, before MRI. Now I feel vindicated, with generous help from someone of Dr. Nasrallah’s prestige and influence.

There also is electrophysiological (including evoked potential) evi­dence for neural pathology in BPD, as well as data derived from single photon emission CT scanning. The burgeoning literature on MRI and functional MRI studies of BPD is in good agreement about the brain changes most relevant to BPD and that are found with regularity in this condition.

Particularly when BPD is diag­nosed in people (usually women) who do not have a history of neglect, sexual molestation, parental humili­ation or cruelty, or head injury, what else is there, if not genetically pre­disposed alterations in the fronto­limbic structures (and maybe the periaqueductal gray) that underlie the so-called “personality disorder,” and, not surprisingly, bipolar dis­orders, especially bipolar II disor­der, which often is the other side of the coin as BPD, and amenable to the same combination of medication and psychotherapy?

Michael H. Stone, MD
Professor of Clinical Psychiatry Columbia College of Physicians and Surgeons
New York, New York

----------------------------------------------------------------------------------------------------------

As a psychiatrist/psychoanalyst who works with BPD patients, I read Dr. Nasrallah’s April 2014 Editorial with great interest and enthusiasm. Over the past 10 years, I have been impressed with the number of patients with BPD whose nonverbal learning disorders and auditory and visual pro­cessing disorders have gone undiag­nosed. Recently, I lectured on this topic to the staff of a school for children with a range of neuropsychiatric disorders; the staff found my observations about such comorbidity consistent with their observations. These dysfunctions, or neurological variations—unknown to the parent and the child—interfere with early object-relation formation, attachment capacity, and learning. Neuropsychiatry and psychological development are, in fact, part of the same system.

An example: For 12 years, I have been treating a patient who has audi­tory processing and working memory problems, meaning that she could not process the connections among different ideas. This difficulty frus­trated her parents, who, in their frus­tration, criticized her for not paying attention. She was labeled “bad” and assumed the role of the “black sheep” in her family. Although she was intelligent, she was often wrong in her judgments and choices, and easily frustrated. In therapy, as I realized what part of her prob­lem was, I changed my technique.

When my patient asked me to tell her the sequence of understandings that we had just put together, I invited her to take my pad and write down her sense of it. As she described each part of that sequence to me, we would discuss it and I would remind her of lost fragments. Gradually, she learned to put ideas together; however, I also watched her struggle to hold these ideas in working memory and to use them.

Over time, she has improved and is more functional. After several years of dis­ability, she returned to work, although she still struggles interpersonally.

With many of such patients, I have had to modify traditional techniques of psy­chotherapy. I am fascinated by, and enjoy, such intensive psychotherapy. I am also amazed to see the impact of previously unknown neuropathologic variations on development. The more I learn about the impact of neuropsychiatry on psychologi­cal development, the more I can help my patients.

Howard Wishnie, MD
Cambridge, Massachusetts

Dr. Nasrallah responds
I appreciate Dr. Stone’s kind words and concur­rence with my thinking about BPD. It would have been appropriate to include discussion of neuro­physiological findings in my Editorial, but I opted to use my limited space to focus on structural and functional neuroimaging and genetics.

Henry A. Nasrallah, MD
Professor and Chairman Department of Neurology & Psychiatry
Saint Louis University School of Medicine
St. Louis, Missouri

Dr. Henry A. Nasrallah’s recent Editorial on borderline personality disorder (BPD) (Current Psychiatry, From the Editor, April 2014, p. 19-20, 32 [http://bit.ly/1e8yAwE]) describes BPD as a heritable brain disease. I have been arguing this point for many years, often finding support from my colleague, Hagop Akiskal, MD, and opposition from my psychoanalytic colleagues.

In recent papers^1,2 on brain changes in BPD and the connection between BPD and bipolar disorders, I wrote that there often is a heritable aspect to the condition. There are exceptions to such heritability, as in the setting of a horrific environment (eg, father-daughter incest, parental brutality), where the same symptoms seen in BPD develop primarily from post-natal influences. Dr. Akiskal and I were discussing this a long time back, before MRI. Now I feel vindicated, with generous help from someone of Dr. Nasrallah’s prestige and influence.

There also is electrophysiological (including evoked potential) evi­dence for neural pathology in BPD, as well as data derived from single photon emission CT scanning. The burgeoning literature on MRI and functional MRI studies of BPD is in good agreement about the brain changes most relevant to BPD and that are found with regularity in this condition.

Particularly when BPD is diag­nosed in people (usually women) who do not have a history of neglect, sexual molestation, parental humili­ation or cruelty, or head injury, what else is there, if not genetically pre­disposed alterations in the fronto­limbic structures (and maybe the periaqueductal gray) that underlie the so-called “personality disorder,” and, not surprisingly, bipolar dis­orders, especially bipolar II disor­der, which often is the other side of the coin as BPD, and amenable to the same combination of medication and psychotherapy?

Michael H. Stone, MD
Professor of Clinical Psychiatry Columbia College of Physicians and Surgeons
New York, New York

----------------------------------------------------------------------------------------------------------

As a psychiatrist/psychoanalyst who works with BPD patients, I read Dr. Nasrallah’s April 2014 Editorial with great interest and enthusiasm. Over the past 10 years, I have been impressed with the number of patients with BPD whose nonverbal learning disorders and auditory and visual pro­cessing disorders have gone undiag­nosed. Recently, I lectured on this topic to the staff of a school for children with a range of neuropsychiatric disorders; the staff found my observations about such comorbidity consistent with their observations. These dysfunctions, or neurological variations—unknown to the parent and the child—interfere with early object-relation formation, attachment capacity, and learning. Neuropsychiatry and psychological development are, in fact, part of the same system.

An example: For 12 years, I have been treating a patient who has audi­tory processing and working memory problems, meaning that she could not process the connections among different ideas. This difficulty frus­trated her parents, who, in their frus­tration, criticized her for not paying attention. She was labeled “bad” and assumed the role of the “black sheep” in her family. Although she was intelligent, she was often wrong in her judgments and choices, and easily frustrated. In therapy, as I realized what part of her prob­lem was, I changed my technique.

When my patient asked me to tell her the sequence of understandings that we had just put together, I invited her to take my pad and write down her sense of it. As she described each part of that sequence to me, we would discuss it and I would remind her of lost fragments. Gradually, she learned to put ideas together; however, I also watched her struggle to hold these ideas in working memory and to use them.

Over time, she has improved and is more functional. After several years of dis­ability, she returned to work, although she still struggles interpersonally.

With many of such patients, I have had to modify traditional techniques of psy­chotherapy. I am fascinated by, and enjoy, such intensive psychotherapy. I am also amazed to see the impact of previously unknown neuropathologic variations on development. The more I learn about the impact of neuropsychiatry on psychologi­cal development, the more I can help my patients.

Howard Wishnie, MD
Cambridge, Massachusetts

Dr. Nasrallah responds
I appreciate Dr. Stone’s kind words and concur­rence with my thinking about BPD. It would have been appropriate to include discussion of neuro­physiological findings in my Editorial, but I opted to use my limited space to focus on structural and functional neuroimaging and genetics.

Henry A. Nasrallah, MD
Professor and Chairman Department of Neurology & Psychiatry
Saint Louis University School of Medicine
St. Louis, Missouri

Dr. Henry A. Nasrallah’s recent Editorial on borderline personality disorder (BPD) (Current Psychiatry, From the Editor, April 2014, p. 19-20, 32 [http://bit.ly/1e8yAwE]) describes BPD as a heritable brain disease. I have been arguing this point for many years, often finding support from my colleague, Hagop Akiskal, MD, and opposition from my psychoanalytic colleagues.

In recent papers^1,2 on brain changes in BPD and the connection between BPD and bipolar disorders, I wrote that there often is a heritable aspect to the condition. There are exceptions to such heritability, as in the setting of a horrific environment (eg, father-daughter incest, parental brutality), where the same symptoms seen in BPD develop primarily from post-natal influences. Dr. Akiskal and I were discussing this a long time back, before MRI. Now I feel vindicated, with generous help from someone of Dr. Nasrallah’s prestige and influence.

There also is electrophysiological (including evoked potential) evi­dence for neural pathology in BPD, as well as data derived from single photon emission CT scanning. The burgeoning literature on MRI and functional MRI studies of BPD is in good agreement about the brain changes most relevant to BPD and that are found with regularity in this condition.

Particularly when BPD is diag­nosed in people (usually women) who do not have a history of neglect, sexual molestation, parental humili­ation or cruelty, or head injury, what else is there, if not genetically pre­disposed alterations in the fronto­limbic structures (and maybe the periaqueductal gray) that underlie the so-called “personality disorder,” and, not surprisingly, bipolar dis­orders, especially bipolar II disor­der, which often is the other side of the coin as BPD, and amenable to the same combination of medication and psychotherapy?

Michael H. Stone, MD
Professor of Clinical Psychiatry Columbia College of Physicians and Surgeons
New York, New York

----------------------------------------------------------------------------------------------------------

As a psychiatrist/psychoanalyst who works with BPD patients, I read Dr. Nasrallah’s April 2014 Editorial with great interest and enthusiasm. Over the past 10 years, I have been impressed with the number of patients with BPD whose nonverbal learning disorders and auditory and visual pro­cessing disorders have gone undiag­nosed. Recently, I lectured on this topic to the staff of a school for children with a range of neuropsychiatric disorders; the staff found my observations about such comorbidity consistent with their observations. These dysfunctions, or neurological variations—unknown to the parent and the child—interfere with early object-relation formation, attachment capacity, and learning. Neuropsychiatry and psychological development are, in fact, part of the same system.

An example: For 12 years, I have been treating a patient who has audi­tory processing and working memory problems, meaning that she could not process the connections among different ideas. This difficulty frus­trated her parents, who, in their frus­tration, criticized her for not paying attention. She was labeled “bad” and assumed the role of the “black sheep” in her family. Although she was intelligent, she was often wrong in her judgments and choices, and easily frustrated. In therapy, as I realized what part of her prob­lem was, I changed my technique.

When my patient asked me to tell her the sequence of understandings that we had just put together, I invited her to take my pad and write down her sense of it. As she described each part of that sequence to me, we would discuss it and I would remind her of lost fragments. Gradually, she learned to put ideas together; however, I also watched her struggle to hold these ideas in working memory and to use them.

Over time, she has improved and is more functional. After several years of dis­ability, she returned to work, although she still struggles interpersonally.

With many of such patients, I have had to modify traditional techniques of psy­chotherapy. I am fascinated by, and enjoy, such intensive psychotherapy. I am also amazed to see the impact of previously unknown neuropathologic variations on development. The more I learn about the impact of neuropsychiatry on psychologi­cal development, the more I can help my patients.

Howard Wishnie, MD
Cambridge, Massachusetts

Dr. Nasrallah responds
I appreciate Dr. Stone’s kind words and concur­rence with my thinking about BPD. It would have been appropriate to include discussion of neuro­physiological findings in my Editorial, but I opted to use my limited space to focus on structural and functional neuroimaging and genetics.

Henry A. Nasrallah, MD
Professor and Chairman Department of Neurology & Psychiatry
Saint Louis University School of Medicine
St. Louis, Missouri

Dear Dr. Mossman,

My patient, Ms. A, asked me to write a let­ter to her landlord (who has a “no pets” policy) stating that she needed to keep her dog in her apartment for “therapeutic” pur­poses—to provide comfort and reduce her posttraumatic stress (PTSD) and anxiety. I hesitated. Could my written statement make me liable if her dog bit someone?

Submitted by “Dr. B”

Studies showing that animals can help outpatients manage psychiatric condi­tions have received a lot of publicity lately. As a result, more patients are asking physicians to provide documentation to sup­port having pets in their apartments or letting their pets accompany them on planes and buses and at restaurants and shopping malls.

But sometimes, animals hurt people. The Centers for Disease Control and Prevention reports that dogs bite 4.5 million Americans each year and that one-fifth of dog bites cause injury that requires medical atten­tion; in 2012, more than 27,000 dog-bite vic­tims needed reconstructive surgery.¹ If Dr. B writes a letter to support letting Ms. A keep a dog in her apartment, how likely is Dr. B to incur professional liability?

To answer this question, let’s examine:
   • the history and background of “pet therapy”
   • types of assistance animals
   • potential liability for owners, land­lords, and clinicians.

History and background
Using animals to improve hospitalized patients’ mental well-being dates back to the 18th century.² In the late 1980s, medical pub­lications began to document systematically how service dogs whose primary role was to help physically disabled individuals to navigate independently also provided social and emotional benefits.^3-7 Since the 1990s, accessibility mandates in Title III of the Americans with Disabilities Act (ADA) (Table 1⁸) have led to the gradual acceptance of service animals in public places where their presence was previously frowned upon or prohibited.^9,10

If service dogs help people with physical problems feel better, it only makes sense that dogs and other animals might lessen emo­tional ailments, too.^11-13 Florence Nightingale and Sigmund Freud both recognized that involving pets in treatment reduced patients’ depression and anxiety,¹⁴ but credit for for­mally introducing animals into therapy usu­ally goes to psychologist Boris Levinson, whose 1969 book described how his dog Jingles helped troubled children communi­cate.¹⁵ Over the past decade, using animals— trained and untrained—for psychological assistance has become an increasingly popu­lar therapeutic maneuver for diverse mental disorders, including autism, anxiety, schizo­phrenia, and PTSD.^16-19

Terminology
Because animals can provide many types of assistance and support, a variety of terms are used to refer to them: service animals, com­panion animals, therapy pets, and so on. In certain situations (including the one described by Dr. B), carefully delineating animals’ roles and functions can reduce confusion and mis­interpretation by patients, health care profes­sionals, policy makers, and regulators.

Parenti et al²⁰ have proposed a “taxon­omy” for assistance dogs based on variables that include:
   • performing task related to a disability
   • the skill level required of the dog
   • who uses the dog
   • applicable training standards
   • legal protections for the dog and its handler.

Table 2²⁰ summarizes this classification system and key variables that differentiate types of assistance dogs.

Certification
Health care facilities often require that visit­ing dogs have some form of “certification” that they are well behaved, and the ADA and many state statutes require that service dogs and some other animals be “certified” to perform their roles. Yet no federal or state statutes lay out explicit training standards or requirements for certification. Therapy Dogs International²¹ and Pet Partners²² are 2 orga­nizations that provide certifications accepted by many agencies and organizations.

Assistance Dogs International, an assis­tance animal advocacy group, has proposed “minimum standards” for training and deployment of service dogs. These include responding to basic obedience commands from the client in public and at home, being able to perform at least 3 tasks to mitigate the client’s disability, teaching the client about dog training and health care, and scheduled follow-ups for skill maintenance. Dogs also should be spayed or neutered, properly vac­cinated, nonaggressive, clean, and continent in public places.²³

Liability laws
Most U.S. jurisdictions make owners liable for animal-caused injuries, including inju­ries caused by service dogs.²⁴ In many states (eg, Minnesota²⁵), an owner can be liable for dog-bite injury even if the owner did noth­ing wrong and had no reason to suspect from prior behavior that the dog might bite someone. Other jurisdictions require evi­dence of owner negligence, or they allow liability only when bites occur off the own­er’s premises²⁶ or if the owner let the dog run loose.²⁷ Many homeowners’ insurance policies include liability coverage for dog bites, and a few companies offer a special canine liability policy.

Landlords often try to bar tenants from having a dog, partly to avoid liability for dog bites. Most states have case law stating that, if a tenant’s apparently friendly dog bites someone, the landlord is not liable for the injury^28,29; landlords can be liable only if they know about a dangerous dog and do nothing about it.³⁰ In a recent decision, however, the Kentucky Supreme Court made landlords statutory owners with potential liability for dog bites if they give tenants permission to have dogs “on or about” the rental premises.³¹

Clinicians and liability
Asking tenants to provide documentation about their need for therapeutic pets has become standard operating procedure for landlords in many states, so Ms. A’s request to Dr. B sounds reasonable. But could Dr. B’s written statement lead to liability if Ms. A’s dog bit and injured someone else?

The best answer is, “It’s conceivable, but really unlikely.” Donna Vanderpool, MBA, JD, an author and attorney who develops and implements risk management services for psychiatrists, has not seen any claims or case reports on litigation blaming mental health clinicians for injury caused by emotional support pets after the clinicians had written a letter for housing purposes (oral and written communications, April 7-13, 2014).

Dr. B might wonder whether writing a let­ter for Ms. A would imply that he had evalu­ated the dog and Ms. A’s ability to control it. Psychiatrists don’t usually discuss—let alone evaluate—the temperament or behavior of their patients’ pets; even if they did they aren’t experts on pet training. Recognizing this, Dr. B’s letter could include a statement to the effect that he was not vouching for the dog’s behavior, but only for how the dog would help Ms. A.

Dr. B also might talk with Ms. A about her need for the dog and whether she had obtained appropriate certification, as dis­cussed above. The ADA provisions pertain­ing to use and presence of service animals do not apply to dogs that are merely patients’ pets, notwithstanding the genuine emotional benefits that a dog’s companionship might provide. Stating that a patient needs an ani­mal to treat an illness might be fraud if the doctor knew the pet was just a buddy.

Bottom Line
Psychiatrists can expect that more and more patients will ask them for letters to support having pets accompany them at home or in public. Although liability seems unlikely, cautious psychiatrists can state in such letters that they have not evaluated the animal in question, only the potential benefits that the patient might derive from it.

Disclosure
Dr. Mossman reports no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing articles.

Dear Dr. Mossman,

My patient, Ms. A, asked me to write a let­ter to her landlord (who has a “no pets” policy) stating that she needed to keep her dog in her apartment for “therapeutic” pur­poses—to provide comfort and reduce her posttraumatic stress (PTSD) and anxiety. I hesitated. Could my written statement make me liable if her dog bit someone?

Submitted by “Dr. B”

Studies showing that animals can help outpatients manage psychiatric condi­tions have received a lot of publicity lately. As a result, more patients are asking physicians to provide documentation to sup­port having pets in their apartments or letting their pets accompany them on planes and buses and at restaurants and shopping malls.

But sometimes, animals hurt people. The Centers for Disease Control and Prevention reports that dogs bite 4.5 million Americans each year and that one-fifth of dog bites cause injury that requires medical atten­tion; in 2012, more than 27,000 dog-bite vic­tims needed reconstructive surgery.¹ If Dr. B writes a letter to support letting Ms. A keep a dog in her apartment, how likely is Dr. B to incur professional liability?

To answer this question, let’s examine:
   • the history and background of “pet therapy”
   • types of assistance animals
   • potential liability for owners, land­lords, and clinicians.

History and background
Using animals to improve hospitalized patients’ mental well-being dates back to the 18th century.² In the late 1980s, medical pub­lications began to document systematically how service dogs whose primary role was to help physically disabled individuals to navigate independently also provided social and emotional benefits.^3-7 Since the 1990s, accessibility mandates in Title III of the Americans with Disabilities Act (ADA) (Table 1⁸) have led to the gradual acceptance of service animals in public places where their presence was previously frowned upon or prohibited.^9,10

If service dogs help people with physical problems feel better, it only makes sense that dogs and other animals might lessen emo­tional ailments, too.^11-13 Florence Nightingale and Sigmund Freud both recognized that involving pets in treatment reduced patients’ depression and anxiety,¹⁴ but credit for for­mally introducing animals into therapy usu­ally goes to psychologist Boris Levinson, whose 1969 book described how his dog Jingles helped troubled children communi­cate.¹⁵ Over the past decade, using animals— trained and untrained—for psychological assistance has become an increasingly popu­lar therapeutic maneuver for diverse mental disorders, including autism, anxiety, schizo­phrenia, and PTSD.^16-19

Terminology
Because animals can provide many types of assistance and support, a variety of terms are used to refer to them: service animals, com­panion animals, therapy pets, and so on. In certain situations (including the one described by Dr. B), carefully delineating animals’ roles and functions can reduce confusion and mis­interpretation by patients, health care profes­sionals, policy makers, and regulators.

Parenti et al²⁰ have proposed a “taxon­omy” for assistance dogs based on variables that include:
   • performing task related to a disability
   • the skill level required of the dog
   • who uses the dog
   • applicable training standards
   • legal protections for the dog and its handler.

Table 2²⁰ summarizes this classification system and key variables that differentiate types of assistance dogs.

Certification
Health care facilities often require that visit­ing dogs have some form of “certification” that they are well behaved, and the ADA and many state statutes require that service dogs and some other animals be “certified” to perform their roles. Yet no federal or state statutes lay out explicit training standards or requirements for certification. Therapy Dogs International²¹ and Pet Partners²² are 2 orga­nizations that provide certifications accepted by many agencies and organizations.

Assistance Dogs International, an assis­tance animal advocacy group, has proposed “minimum standards” for training and deployment of service dogs. These include responding to basic obedience commands from the client in public and at home, being able to perform at least 3 tasks to mitigate the client’s disability, teaching the client about dog training and health care, and scheduled follow-ups for skill maintenance. Dogs also should be spayed or neutered, properly vac­cinated, nonaggressive, clean, and continent in public places.²³

Liability laws
Most U.S. jurisdictions make owners liable for animal-caused injuries, including inju­ries caused by service dogs.²⁴ In many states (eg, Minnesota²⁵), an owner can be liable for dog-bite injury even if the owner did noth­ing wrong and had no reason to suspect from prior behavior that the dog might bite someone. Other jurisdictions require evi­dence of owner negligence, or they allow liability only when bites occur off the own­er’s premises²⁶ or if the owner let the dog run loose.²⁷ Many homeowners’ insurance policies include liability coverage for dog bites, and a few companies offer a special canine liability policy.

Landlords often try to bar tenants from having a dog, partly to avoid liability for dog bites. Most states have case law stating that, if a tenant’s apparently friendly dog bites someone, the landlord is not liable for the injury^28,29; landlords can be liable only if they know about a dangerous dog and do nothing about it.³⁰ In a recent decision, however, the Kentucky Supreme Court made landlords statutory owners with potential liability for dog bites if they give tenants permission to have dogs “on or about” the rental premises.³¹

Clinicians and liability
Asking tenants to provide documentation about their need for therapeutic pets has become standard operating procedure for landlords in many states, so Ms. A’s request to Dr. B sounds reasonable. But could Dr. B’s written statement lead to liability if Ms. A’s dog bit and injured someone else?

The best answer is, “It’s conceivable, but really unlikely.” Donna Vanderpool, MBA, JD, an author and attorney who develops and implements risk management services for psychiatrists, has not seen any claims or case reports on litigation blaming mental health clinicians for injury caused by emotional support pets after the clinicians had written a letter for housing purposes (oral and written communications, April 7-13, 2014).

Dr. B might wonder whether writing a let­ter for Ms. A would imply that he had evalu­ated the dog and Ms. A’s ability to control it. Psychiatrists don’t usually discuss—let alone evaluate—the temperament or behavior of their patients’ pets; even if they did they aren’t experts on pet training. Recognizing this, Dr. B’s letter could include a statement to the effect that he was not vouching for the dog’s behavior, but only for how the dog would help Ms. A.

Dr. B also might talk with Ms. A about her need for the dog and whether she had obtained appropriate certification, as dis­cussed above. The ADA provisions pertain­ing to use and presence of service animals do not apply to dogs that are merely patients’ pets, notwithstanding the genuine emotional benefits that a dog’s companionship might provide. Stating that a patient needs an ani­mal to treat an illness might be fraud if the doctor knew the pet was just a buddy.

Bottom Line
Psychiatrists can expect that more and more patients will ask them for letters to support having pets accompany them at home or in public. Although liability seems unlikely, cautious psychiatrists can state in such letters that they have not evaluated the animal in question, only the potential benefits that the patient might derive from it.

Disclosure
Dr. Mossman reports no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing articles.

Dear Dr. Mossman,

My patient, Ms. A, asked me to write a let­ter to her landlord (who has a “no pets” policy) stating that she needed to keep her dog in her apartment for “therapeutic” pur­poses—to provide comfort and reduce her posttraumatic stress (PTSD) and anxiety. I hesitated. Could my written statement make me liable if her dog bit someone?

Submitted by “Dr. B”

Studies showing that animals can help outpatients manage psychiatric condi­tions have received a lot of publicity lately. As a result, more patients are asking physicians to provide documentation to sup­port having pets in their apartments or letting their pets accompany them on planes and buses and at restaurants and shopping malls.

But sometimes, animals hurt people. The Centers for Disease Control and Prevention reports that dogs bite 4.5 million Americans each year and that one-fifth of dog bites cause injury that requires medical atten­tion; in 2012, more than 27,000 dog-bite vic­tims needed reconstructive surgery.¹ If Dr. B writes a letter to support letting Ms. A keep a dog in her apartment, how likely is Dr. B to incur professional liability?

To answer this question, let’s examine:
   • the history and background of “pet therapy”
   • types of assistance animals
   • potential liability for owners, land­lords, and clinicians.

History and background
Using animals to improve hospitalized patients’ mental well-being dates back to the 18th century.² In the late 1980s, medical pub­lications began to document systematically how service dogs whose primary role was to help physically disabled individuals to navigate independently also provided social and emotional benefits.^3-7 Since the 1990s, accessibility mandates in Title III of the Americans with Disabilities Act (ADA) (Table 1⁸) have led to the gradual acceptance of service animals in public places where their presence was previously frowned upon or prohibited.^9,10

If service dogs help people with physical problems feel better, it only makes sense that dogs and other animals might lessen emo­tional ailments, too.^11-13 Florence Nightingale and Sigmund Freud both recognized that involving pets in treatment reduced patients’ depression and anxiety,¹⁴ but credit for for­mally introducing animals into therapy usu­ally goes to psychologist Boris Levinson, whose 1969 book described how his dog Jingles helped troubled children communi­cate.¹⁵ Over the past decade, using animals— trained and untrained—for psychological assistance has become an increasingly popu­lar therapeutic maneuver for diverse mental disorders, including autism, anxiety, schizo­phrenia, and PTSD.^16-19

Terminology
Because animals can provide many types of assistance and support, a variety of terms are used to refer to them: service animals, com­panion animals, therapy pets, and so on. In certain situations (including the one described by Dr. B), carefully delineating animals’ roles and functions can reduce confusion and mis­interpretation by patients, health care profes­sionals, policy makers, and regulators.

Parenti et al²⁰ have proposed a “taxon­omy” for assistance dogs based on variables that include:
   • performing task related to a disability
   • the skill level required of the dog
   • who uses the dog
   • applicable training standards
   • legal protections for the dog and its handler.

Table 2²⁰ summarizes this classification system and key variables that differentiate types of assistance dogs.

Certification
Health care facilities often require that visit­ing dogs have some form of “certification” that they are well behaved, and the ADA and many state statutes require that service dogs and some other animals be “certified” to perform their roles. Yet no federal or state statutes lay out explicit training standards or requirements for certification. Therapy Dogs International²¹ and Pet Partners²² are 2 orga­nizations that provide certifications accepted by many agencies and organizations.

Assistance Dogs International, an assis­tance animal advocacy group, has proposed “minimum standards” for training and deployment of service dogs. These include responding to basic obedience commands from the client in public and at home, being able to perform at least 3 tasks to mitigate the client’s disability, teaching the client about dog training and health care, and scheduled follow-ups for skill maintenance. Dogs also should be spayed or neutered, properly vac­cinated, nonaggressive, clean, and continent in public places.²³

Liability laws
Most U.S. jurisdictions make owners liable for animal-caused injuries, including inju­ries caused by service dogs.²⁴ In many states (eg, Minnesota²⁵), an owner can be liable for dog-bite injury even if the owner did noth­ing wrong and had no reason to suspect from prior behavior that the dog might bite someone. Other jurisdictions require evi­dence of owner negligence, or they allow liability only when bites occur off the own­er’s premises²⁶ or if the owner let the dog run loose.²⁷ Many homeowners’ insurance policies include liability coverage for dog bites, and a few companies offer a special canine liability policy.

Landlords often try to bar tenants from having a dog, partly to avoid liability for dog bites. Most states have case law stating that, if a tenant’s apparently friendly dog bites someone, the landlord is not liable for the injury^28,29; landlords can be liable only if they know about a dangerous dog and do nothing about it.³⁰ In a recent decision, however, the Kentucky Supreme Court made landlords statutory owners with potential liability for dog bites if they give tenants permission to have dogs “on or about” the rental premises.³¹

Clinicians and liability
Asking tenants to provide documentation about their need for therapeutic pets has become standard operating procedure for landlords in many states, so Ms. A’s request to Dr. B sounds reasonable. But could Dr. B’s written statement lead to liability if Ms. A’s dog bit and injured someone else?

The best answer is, “It’s conceivable, but really unlikely.” Donna Vanderpool, MBA, JD, an author and attorney who develops and implements risk management services for psychiatrists, has not seen any claims or case reports on litigation blaming mental health clinicians for injury caused by emotional support pets after the clinicians had written a letter for housing purposes (oral and written communications, April 7-13, 2014).

Dr. B might wonder whether writing a let­ter for Ms. A would imply that he had evalu­ated the dog and Ms. A’s ability to control it. Psychiatrists don’t usually discuss—let alone evaluate—the temperament or behavior of their patients’ pets; even if they did they aren’t experts on pet training. Recognizing this, Dr. B’s letter could include a statement to the effect that he was not vouching for the dog’s behavior, but only for how the dog would help Ms. A.

Dr. B also might talk with Ms. A about her need for the dog and whether she had obtained appropriate certification, as dis­cussed above. The ADA provisions pertain­ing to use and presence of service animals do not apply to dogs that are merely patients’ pets, notwithstanding the genuine emotional benefits that a dog’s companionship might provide. Stating that a patient needs an ani­mal to treat an illness might be fraud if the doctor knew the pet was just a buddy.

Bottom Line
Psychiatrists can expect that more and more patients will ask them for letters to support having pets accompany them at home or in public. Although liability seems unlikely, cautious psychiatrists can state in such letters that they have not evaluated the animal in question, only the potential benefits that the patient might derive from it.

Disclosure
Dr. Mossman reports no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing articles.

Prescription medications

Credit: CDC

Patients who delay filling a prescription of clopidogrel after coronary stenting may increase their risk of recurrent myocardial infarction (MI) and death, according to a study published in the Journal of the American Heart Association. 

Researchers analyzed records of more than 15,000 patients who received drug-eluting or bare metal stents.

Roughly 30% of patients in each group failed to fill their prescription of the anticoagulant clopidogrel within 3 days of hospital discharge.

And this roughly doubled the patients’ risk of death and recurrent MI, regardless of their stent type.

“It is very important that patients take clopidogrel after having a coronary stent implanted to prevent blood clots forming within the stent,” said study author Nicholas Cruden, MBChB, PhD, of the Royal Infirmary of Edinburgh in the UK.

He and his colleagues analyzed hospital administrative, community pharmacy, and cardiac revascularization data from all patients who received a coronary stent in British Columbia between 2004 and 2006, with follow-up out to 2 years.

Of the 15,629 patients, 3599 had received at least 1 drug-eluting stent (DES), and 12,030 had received a bare metal stent (BMS). Thirty percent (n=1064) of patients in the DES group, and 31% (n=3758) of patients in the BMS group failed to fill their prescription for clopidogrel within 3 days of hospital discharge.

And a delay of more than 3 days was predictive of mortality and recurrent MI, regardless of the stent type. The hazard ratios (HRs) for mortality were 2.4 for the DES group and 2.2 for the BMS group. The HRs for recurrent MI were 2.0 and 1.8, respectively.

The excess risk associated with a delay in filling the prescription was greatest in the immediate period after hospital discharge—up to 30 days. In all patients, the HRs were 5.5 for mortality and 3.1 for recurrent MI.

Delaying filling the prescription for more than 3 days remained an independent predictor of death and MI beyond 30 days from hospital discharge. The HRs were 2.1 and 2.0, respectively, for patients in the DES group and 2.0 and 1.8, respectively, for patients in the BMS group.

“This study highlights the importance of ensuring patients have access to medications as soon as they leave the hospital,” Dr Cruden said. “Even a delay of a day or 2 was associated with worse outcomes.”

Discharging patients from the hospital with enough medicine for the highest-risk period (the first month or so) could help, he added.

Prescription medications

Credit: CDC

Patients who delay filling a prescription of clopidogrel after coronary stenting may increase their risk of recurrent myocardial infarction (MI) and death, according to a study published in the Journal of the American Heart Association. 

Researchers analyzed records of more than 15,000 patients who received drug-eluting or bare metal stents.

Roughly 30% of patients in each group failed to fill their prescription of the anticoagulant clopidogrel within 3 days of hospital discharge.

And this roughly doubled the patients’ risk of death and recurrent MI, regardless of their stent type.

“It is very important that patients take clopidogrel after having a coronary stent implanted to prevent blood clots forming within the stent,” said study author Nicholas Cruden, MBChB, PhD, of the Royal Infirmary of Edinburgh in the UK.

He and his colleagues analyzed hospital administrative, community pharmacy, and cardiac revascularization data from all patients who received a coronary stent in British Columbia between 2004 and 2006, with follow-up out to 2 years.

Of the 15,629 patients, 3599 had received at least 1 drug-eluting stent (DES), and 12,030 had received a bare metal stent (BMS). Thirty percent (n=1064) of patients in the DES group, and 31% (n=3758) of patients in the BMS group failed to fill their prescription for clopidogrel within 3 days of hospital discharge.

And a delay of more than 3 days was predictive of mortality and recurrent MI, regardless of the stent type. The hazard ratios (HRs) for mortality were 2.4 for the DES group and 2.2 for the BMS group. The HRs for recurrent MI were 2.0 and 1.8, respectively.

The excess risk associated with a delay in filling the prescription was greatest in the immediate period after hospital discharge—up to 30 days. In all patients, the HRs were 5.5 for mortality and 3.1 for recurrent MI.

Delaying filling the prescription for more than 3 days remained an independent predictor of death and MI beyond 30 days from hospital discharge. The HRs were 2.1 and 2.0, respectively, for patients in the DES group and 2.0 and 1.8, respectively, for patients in the BMS group.

“This study highlights the importance of ensuring patients have access to medications as soon as they leave the hospital,” Dr Cruden said. “Even a delay of a day or 2 was associated with worse outcomes.”

Discharging patients from the hospital with enough medicine for the highest-risk period (the first month or so) could help, he added.

Prescription medications

Credit: CDC

Patients who delay filling a prescription of clopidogrel after coronary stenting may increase their risk of recurrent myocardial infarction (MI) and death, according to a study published in the Journal of the American Heart Association. 

Researchers analyzed records of more than 15,000 patients who received drug-eluting or bare metal stents.

Roughly 30% of patients in each group failed to fill their prescription of the anticoagulant clopidogrel within 3 days of hospital discharge.

And this roughly doubled the patients’ risk of death and recurrent MI, regardless of their stent type.

“It is very important that patients take clopidogrel after having a coronary stent implanted to prevent blood clots forming within the stent,” said study author Nicholas Cruden, MBChB, PhD, of the Royal Infirmary of Edinburgh in the UK.

He and his colleagues analyzed hospital administrative, community pharmacy, and cardiac revascularization data from all patients who received a coronary stent in British Columbia between 2004 and 2006, with follow-up out to 2 years.

Of the 15,629 patients, 3599 had received at least 1 drug-eluting stent (DES), and 12,030 had received a bare metal stent (BMS). Thirty percent (n=1064) of patients in the DES group, and 31% (n=3758) of patients in the BMS group failed to fill their prescription for clopidogrel within 3 days of hospital discharge.

And a delay of more than 3 days was predictive of mortality and recurrent MI, regardless of the stent type. The hazard ratios (HRs) for mortality were 2.4 for the DES group and 2.2 for the BMS group. The HRs for recurrent MI were 2.0 and 1.8, respectively.

The excess risk associated with a delay in filling the prescription was greatest in the immediate period after hospital discharge—up to 30 days. In all patients, the HRs were 5.5 for mortality and 3.1 for recurrent MI.

Delaying filling the prescription for more than 3 days remained an independent predictor of death and MI beyond 30 days from hospital discharge. The HRs were 2.1 and 2.0, respectively, for patients in the DES group and 2.0 and 1.8, respectively, for patients in the BMS group.

“This study highlights the importance of ensuring patients have access to medications as soon as they leave the hospital,” Dr Cruden said. “Even a delay of a day or 2 was associated with worse outcomes.”

Discharging patients from the hospital with enough medicine for the highest-risk period (the first month or so) could help, he added.

In November 2013, The American College of Obstetricians and Gynecologists (ACOG) published the results of its Task Force on Hypertension in Pregnancy.¹ The Task Force aimed to help clinicians become familiar with the state of research in hypertension during pregnancy as well as to assist us in standardizing management approaches to such patients.

The Task Force reported that, worldwide, hypertensive disorders complicate approximately 10% of pregnancies. In addition, in the United States, the past 20 years have brought a 25% increase in the incidence of preeclampsia. According to past ACOG President James N. Martin, Jr, MD, in the last 10 years, the pathophysiology of preeclampsia has become better understood, but the etiology remains unclear and evidence that has emerged to guide therapy has not translated into clinical practice.¹

Related articles:
The latest guidance from ACOG on hypertension in pregnancy. Jaimey E. Pauli, MD (Audiocast, January 2014)
Update on Obstetrics. Jaimey E. Pauli, MD, and John T. Repke, MD (January 2014)

The Task Force document contained 60 recommendations for the prevention, prediction, and management of hypertensive disorders of pregnancy, including preeclampsia, gestational hypertension, chronic hypertension, HELLP syndrome, and preeclampsia superimposed on an underlying hypertensive disorder (see box below). One recommendation was that women at high risk for preeclampsia, particularly those with a history of preeclampsia that required delivery before 34 weeks, could possibly benefit from taking aspirin (60–81 mg) daily starting at the end of the first trimester. They further noted that this benefit could include prevention of recurrent severe preeclampsia, or at least a reduction in recurrence risk.

The ACOG Task Force made its recommendation based on results of a meta-analysis of low-dose aspirin trials, involving more than 30,000 patients,² suggesting a small decrease in the risk of preeclampsia and associated morbidity. More precise risk reduction estimates were difficult to make due to the heterogeneity of the studies reviewed. And the Task Force further stated that this (low-dose aspirin) approach had no demonstrable acute adverse fetal effects, although long-term adverse effects could not be entirely excluded based on the current data.

Unfortunately, according to the ACOG document, the strength of the evidence supporting their recommendation was “moderate” and the strength of the recommendation was “qualified” so, not exactly a resounding endorsement of this approach, but a recommendation nonetheless.

OBSTETRIC PRACTICE CHANGERS 2014
Hypertension and pregnancy and preventing the first cesarean delivery

Data suggest aspirin for high-risk women could be reasonable
A recent study by Henderson and colleagues presented a systematic review for the US Preventive Services Task Force (USPSTF) on the potential for low-dose aspirin to prevent morbidity and mortality from preeclampsia.³ The design was a meta-analysis of 28 studies: two large, multisite, randomized clinical trials (RCTs); 13 smaller RCTs of high-risk women, of which eight were deemed “good quality”; and six RCTs and two observational studies of average-risk women, of which seven were deemed to be good quality.

The results essentially supported the notion that low-dose aspirin had a beneficial effect with respect to prevention of preeclampsia and perinatal morbidity in women at high risk for preeclampsia. Additionally, no harmful effects were identified, although the authors acknowledged potential rare or long-term harm could not be excluded.

Questions remain
While somewhat gratifying, the results of the USPSTF systematic review still leave many questions. First, the dose of aspirin used in the studies analyzed ranged from 50 mg/d to 150 mg/d. In the United States, “low-dose” aspirin is usually prescribed at 81 mg/d, so the applicability of this review’s findings to US clinical practice is not exact. Second, the authors acknowledged that the putative positive effects observed could be secondary to so-called “small study effects,” and that when only the larger studies were analyzed the effects were less impressive.

Related article: A stepwise approach to managing eclampsia and other hypertensive emergencies. Baha M. Sibai (October 2013)

In my opinion, both the USPSTF study and the recommendations from the ACOG Task Force provide some reassurance for clinicians that the use of daily, low-dose aspirin by women at high risk for preeclampsia probably does afford some benefit, and seems to be a safe approach—as we have known from the initial Maternal-Fetal Medicine Units (MFMU) trial published in 1993 on low-risk women⁴ and the follow-up MFMU study on high-risk women.⁵

The need for additional studies is clear, however. The idea that preeclampsia is the same in every patient would seem to make no more sense than thinking all cancer is the same, with the same risk factors, the same epidemiology and pathophysiology, and the same response to similar treatments. Fundamentally, we need to further explore the different pathways through which preeclampsia develops in women and then apply the strategy best suited to treating (or preventing) their form of the disease—a personalized medicine approach.

In the meantime, most patients who have delivered at 34 weeks or less because of preeclampsia and who are contemplating another pregnancy are really not interested in hearing us tell them that we cannot do anything to prevent recurrent preeclampsia because we are awaiting further studies. At least the ACOG recommendations and the results of the USPSTF’s systematic review provide us with a reasonable, although perhaps not yet optimal, therapeutic option.

Related article: 10 practical, evidence-based recommendations to improve outcomes in women who have eclampsia. Baha M. Sibai (November 2011)

The bottom line
In my own practice, I discuss the option of initiating low-dose aspirin (81 mg/d) as early as 12 weeks’ gestation for patients who had either prior early-onset preeclampsia requiring delivery before 34 weeks’ gestation or preeclampsia during more than one pregnancy.

QUICK POLL
Do you offer low-dose aspirin for preeclampsia prevention?
When faced with a patient with prior preeclampsia in more than one pregnancy or with preeclampsia that resulted in delivery prior to 34 weeks, do you offer low-dose aspirin as an option for preventing preeclampsia?
Visit the Quick Poll on the right column of the OBGManagement.com home page to register your answer and see how your colleagues voted.

WE WANT TO HEAR FROM YOU!
Share your thoughts on this article. Send your Letter to the Editor to: [email protected]

In November 2013, The American College of Obstetricians and Gynecologists (ACOG) published the results of its Task Force on Hypertension in Pregnancy.¹ The Task Force aimed to help clinicians become familiar with the state of research in hypertension during pregnancy as well as to assist us in standardizing management approaches to such patients.

The Task Force reported that, worldwide, hypertensive disorders complicate approximately 10% of pregnancies. In addition, in the United States, the past 20 years have brought a 25% increase in the incidence of preeclampsia. According to past ACOG President James N. Martin, Jr, MD, in the last 10 years, the pathophysiology of preeclampsia has become better understood, but the etiology remains unclear and evidence that has emerged to guide therapy has not translated into clinical practice.¹

Related articles:
The latest guidance from ACOG on hypertension in pregnancy. Jaimey E. Pauli, MD (Audiocast, January 2014)
Update on Obstetrics. Jaimey E. Pauli, MD, and John T. Repke, MD (January 2014)

The Task Force document contained 60 recommendations for the prevention, prediction, and management of hypertensive disorders of pregnancy, including preeclampsia, gestational hypertension, chronic hypertension, HELLP syndrome, and preeclampsia superimposed on an underlying hypertensive disorder (see box below). One recommendation was that women at high risk for preeclampsia, particularly those with a history of preeclampsia that required delivery before 34 weeks, could possibly benefit from taking aspirin (60–81 mg) daily starting at the end of the first trimester. They further noted that this benefit could include prevention of recurrent severe preeclampsia, or at least a reduction in recurrence risk.

The ACOG Task Force made its recommendation based on results of a meta-analysis of low-dose aspirin trials, involving more than 30,000 patients,² suggesting a small decrease in the risk of preeclampsia and associated morbidity. More precise risk reduction estimates were difficult to make due to the heterogeneity of the studies reviewed. And the Task Force further stated that this (low-dose aspirin) approach had no demonstrable acute adverse fetal effects, although long-term adverse effects could not be entirely excluded based on the current data.

Unfortunately, according to the ACOG document, the strength of the evidence supporting their recommendation was “moderate” and the strength of the recommendation was “qualified” so, not exactly a resounding endorsement of this approach, but a recommendation nonetheless.

OBSTETRIC PRACTICE CHANGERS 2014
Hypertension and pregnancy and preventing the first cesarean delivery

Data suggest aspirin for high-risk women could be reasonable
A recent study by Henderson and colleagues presented a systematic review for the US Preventive Services Task Force (USPSTF) on the potential for low-dose aspirin to prevent morbidity and mortality from preeclampsia.³ The design was a meta-analysis of 28 studies: two large, multisite, randomized clinical trials (RCTs); 13 smaller RCTs of high-risk women, of which eight were deemed “good quality”; and six RCTs and two observational studies of average-risk women, of which seven were deemed to be good quality.

The results essentially supported the notion that low-dose aspirin had a beneficial effect with respect to prevention of preeclampsia and perinatal morbidity in women at high risk for preeclampsia. Additionally, no harmful effects were identified, although the authors acknowledged potential rare or long-term harm could not be excluded.

Questions remain
While somewhat gratifying, the results of the USPSTF systematic review still leave many questions. First, the dose of aspirin used in the studies analyzed ranged from 50 mg/d to 150 mg/d. In the United States, “low-dose” aspirin is usually prescribed at 81 mg/d, so the applicability of this review’s findings to US clinical practice is not exact. Second, the authors acknowledged that the putative positive effects observed could be secondary to so-called “small study effects,” and that when only the larger studies were analyzed the effects were less impressive.

Related article: A stepwise approach to managing eclampsia and other hypertensive emergencies. Baha M. Sibai (October 2013)

In my opinion, both the USPSTF study and the recommendations from the ACOG Task Force provide some reassurance for clinicians that the use of daily, low-dose aspirin by women at high risk for preeclampsia probably does afford some benefit, and seems to be a safe approach—as we have known from the initial Maternal-Fetal Medicine Units (MFMU) trial published in 1993 on low-risk women⁴ and the follow-up MFMU study on high-risk women.⁵

The need for additional studies is clear, however. The idea that preeclampsia is the same in every patient would seem to make no more sense than thinking all cancer is the same, with the same risk factors, the same epidemiology and pathophysiology, and the same response to similar treatments. Fundamentally, we need to further explore the different pathways through which preeclampsia develops in women and then apply the strategy best suited to treating (or preventing) their form of the disease—a personalized medicine approach.

In the meantime, most patients who have delivered at 34 weeks or less because of preeclampsia and who are contemplating another pregnancy are really not interested in hearing us tell them that we cannot do anything to prevent recurrent preeclampsia because we are awaiting further studies. At least the ACOG recommendations and the results of the USPSTF’s systematic review provide us with a reasonable, although perhaps not yet optimal, therapeutic option.

Related article: 10 practical, evidence-based recommendations to improve outcomes in women who have eclampsia. Baha M. Sibai (November 2011)

The bottom line
In my own practice, I discuss the option of initiating low-dose aspirin (81 mg/d) as early as 12 weeks’ gestation for patients who had either prior early-onset preeclampsia requiring delivery before 34 weeks’ gestation or preeclampsia during more than one pregnancy.

QUICK POLL
Do you offer low-dose aspirin for preeclampsia prevention?
When faced with a patient with prior preeclampsia in more than one pregnancy or with preeclampsia that resulted in delivery prior to 34 weeks, do you offer low-dose aspirin as an option for preventing preeclampsia?
Visit the Quick Poll on the right column of the OBGManagement.com home page to register your answer and see how your colleagues voted.

WE WANT TO HEAR FROM YOU!
Share your thoughts on this article. Send your Letter to the Editor to: [email protected]

In November 2013, The American College of Obstetricians and Gynecologists (ACOG) published the results of its Task Force on Hypertension in Pregnancy.¹ The Task Force aimed to help clinicians become familiar with the state of research in hypertension during pregnancy as well as to assist us in standardizing management approaches to such patients.

The Task Force reported that, worldwide, hypertensive disorders complicate approximately 10% of pregnancies. In addition, in the United States, the past 20 years have brought a 25% increase in the incidence of preeclampsia. According to past ACOG President James N. Martin, Jr, MD, in the last 10 years, the pathophysiology of preeclampsia has become better understood, but the etiology remains unclear and evidence that has emerged to guide therapy has not translated into clinical practice.¹

Related articles:
The latest guidance from ACOG on hypertension in pregnancy. Jaimey E. Pauli, MD (Audiocast, January 2014)
Update on Obstetrics. Jaimey E. Pauli, MD, and John T. Repke, MD (January 2014)

The Task Force document contained 60 recommendations for the prevention, prediction, and management of hypertensive disorders of pregnancy, including preeclampsia, gestational hypertension, chronic hypertension, HELLP syndrome, and preeclampsia superimposed on an underlying hypertensive disorder (see box below). One recommendation was that women at high risk for preeclampsia, particularly those with a history of preeclampsia that required delivery before 34 weeks, could possibly benefit from taking aspirin (60–81 mg) daily starting at the end of the first trimester. They further noted that this benefit could include prevention of recurrent severe preeclampsia, or at least a reduction in recurrence risk.

The ACOG Task Force made its recommendation based on results of a meta-analysis of low-dose aspirin trials, involving more than 30,000 patients,² suggesting a small decrease in the risk of preeclampsia and associated morbidity. More precise risk reduction estimates were difficult to make due to the heterogeneity of the studies reviewed. And the Task Force further stated that this (low-dose aspirin) approach had no demonstrable acute adverse fetal effects, although long-term adverse effects could not be entirely excluded based on the current data.

Unfortunately, according to the ACOG document, the strength of the evidence supporting their recommendation was “moderate” and the strength of the recommendation was “qualified” so, not exactly a resounding endorsement of this approach, but a recommendation nonetheless.

OBSTETRIC PRACTICE CHANGERS 2014
Hypertension and pregnancy and preventing the first cesarean delivery

Data suggest aspirin for high-risk women could be reasonable
A recent study by Henderson and colleagues presented a systematic review for the US Preventive Services Task Force (USPSTF) on the potential for low-dose aspirin to prevent morbidity and mortality from preeclampsia.³ The design was a meta-analysis of 28 studies: two large, multisite, randomized clinical trials (RCTs); 13 smaller RCTs of high-risk women, of which eight were deemed “good quality”; and six RCTs and two observational studies of average-risk women, of which seven were deemed to be good quality.

The results essentially supported the notion that low-dose aspirin had a beneficial effect with respect to prevention of preeclampsia and perinatal morbidity in women at high risk for preeclampsia. Additionally, no harmful effects were identified, although the authors acknowledged potential rare or long-term harm could not be excluded.

Questions remain
While somewhat gratifying, the results of the USPSTF systematic review still leave many questions. First, the dose of aspirin used in the studies analyzed ranged from 50 mg/d to 150 mg/d. In the United States, “low-dose” aspirin is usually prescribed at 81 mg/d, so the applicability of this review’s findings to US clinical practice is not exact. Second, the authors acknowledged that the putative positive effects observed could be secondary to so-called “small study effects,” and that when only the larger studies were analyzed the effects were less impressive.

Related article: A stepwise approach to managing eclampsia and other hypertensive emergencies. Baha M. Sibai (October 2013)

In my opinion, both the USPSTF study and the recommendations from the ACOG Task Force provide some reassurance for clinicians that the use of daily, low-dose aspirin by women at high risk for preeclampsia probably does afford some benefit, and seems to be a safe approach—as we have known from the initial Maternal-Fetal Medicine Units (MFMU) trial published in 1993 on low-risk women⁴ and the follow-up MFMU study on high-risk women.⁵

The need for additional studies is clear, however. The idea that preeclampsia is the same in every patient would seem to make no more sense than thinking all cancer is the same, with the same risk factors, the same epidemiology and pathophysiology, and the same response to similar treatments. Fundamentally, we need to further explore the different pathways through which preeclampsia develops in women and then apply the strategy best suited to treating (or preventing) their form of the disease—a personalized medicine approach.

In the meantime, most patients who have delivered at 34 weeks or less because of preeclampsia and who are contemplating another pregnancy are really not interested in hearing us tell them that we cannot do anything to prevent recurrent preeclampsia because we are awaiting further studies. At least the ACOG recommendations and the results of the USPSTF’s systematic review provide us with a reasonable, although perhaps not yet optimal, therapeutic option.

Related article: 10 practical, evidence-based recommendations to improve outcomes in women who have eclampsia. Baha M. Sibai (November 2011)

The bottom line
In my own practice, I discuss the option of initiating low-dose aspirin (81 mg/d) as early as 12 weeks’ gestation for patients who had either prior early-onset preeclampsia requiring delivery before 34 weeks’ gestation or preeclampsia during more than one pregnancy.

QUICK POLL
Do you offer low-dose aspirin for preeclampsia prevention?
When faced with a patient with prior preeclampsia in more than one pregnancy or with preeclampsia that resulted in delivery prior to 34 weeks, do you offer low-dose aspirin as an option for preventing preeclampsia?
Visit the Quick Poll on the right column of the OBGManagement.com home page to register your answer and see how your colleagues voted.

WE WANT TO HEAR FROM YOU!
Share your thoughts on this article. Send your Letter to the Editor to: [email protected]