Cleveland Clinic Journal of Medicine

A 59-year-old woman presented with drooling out of the left side of her mouth and inability to close her left eye. She had no ear pain, hearing loss, or skin rash. The facial palsy affected all branches of the left facial nerve. This explained her inability to close her left eyelid and the generalized weakness of the left side of the face, including her forehead and angle of the mouth. No other signs of pontine dysfunction were noted.

See related editorial

The symptoms had begun 2 months earlier, and computed tomography (CT) of the head performed at a nearby clinic 3 days after the onset of symptoms showed no abnormalities. She was given a diagnosis of incomplete Bell palsy and was prescribed prednisolone and valacyclovir. However, her symptoms had not improved after 2 months of treatment, and so she presented to our hospital.

Physical examination revealed moderate nerve dysfunction (House-Brackmann grade III, with grade I normal and grade VI total paralysis) and generalized weakness on the left side of her face including her forehead.¹ She had no loss in facial sensation or hearing and no ataxia or ocular motility disorders.

BELL PALSY

Peripheral facial nerve palsy is classified either as Bell palsy, which is idiopathic, or as secondary facial nerve palsy. Because Bell palsy accounts for 60% to 70% of all cases,² treatment with oral steroids is indicated when no abnormal findings other than lateral peripheral facial nerve palsy are observed. Antiviral drugs may provide added benefit, although academic societies do not currently recommend combined therapy.³ However, 85% of patients with Bell palsy improve within 3 weeks without treatment, and 94% of patients with incomplete Bell palsy—defined by normal to severe dysfunction, ie, not total paralysis, based on House-Brackmann score—eventually achieve complete remission.²

Therefore, although progression of symptoms or lack of improvement at 2 months does not rule out Bell palsy, it should prompt a detailed imaging evaluation to rule out an underlying condition such as tumor (in the pons, cerebellopontine angle, parotid gland, middle ear, or petrosal bone), infection (herpes simplex, varicella zoster, Ramsey-Hunt syndrome, or otitis media), trauma, or systemic disease (diabetes mellitus, multiple sclerosis, sarcoidosis, or systemic lupus erythematosus).⁴

According to a review of common causes of facial nerve palsy, the most common finding in 224 patients misdiagnosed with Bell palsy was tumor (38%).⁵ This indicates the value of magnetic resonance imaging of the head rather than CT when secondary facial nerve palsy is suspected, as CT is not sensitive to brainstem lesions.

A 59-year-old woman presented with drooling out of the left side of her mouth and inability to close her left eye. She had no ear pain, hearing loss, or skin rash. The facial palsy affected all branches of the left facial nerve. This explained her inability to close her left eyelid and the generalized weakness of the left side of the face, including her forehead and angle of the mouth. No other signs of pontine dysfunction were noted.

See related editorial

The symptoms had begun 2 months earlier, and computed tomography (CT) of the head performed at a nearby clinic 3 days after the onset of symptoms showed no abnormalities. She was given a diagnosis of incomplete Bell palsy and was prescribed prednisolone and valacyclovir. However, her symptoms had not improved after 2 months of treatment, and so she presented to our hospital.

Physical examination revealed moderate nerve dysfunction (House-Brackmann grade III, with grade I normal and grade VI total paralysis) and generalized weakness on the left side of her face including her forehead.¹ She had no loss in facial sensation or hearing and no ataxia or ocular motility disorders.

BELL PALSY

Peripheral facial nerve palsy is classified either as Bell palsy, which is idiopathic, or as secondary facial nerve palsy. Because Bell palsy accounts for 60% to 70% of all cases,² treatment with oral steroids is indicated when no abnormal findings other than lateral peripheral facial nerve palsy are observed. Antiviral drugs may provide added benefit, although academic societies do not currently recommend combined therapy.³ However, 85% of patients with Bell palsy improve within 3 weeks without treatment, and 94% of patients with incomplete Bell palsy—defined by normal to severe dysfunction, ie, not total paralysis, based on House-Brackmann score—eventually achieve complete remission.²

Therefore, although progression of symptoms or lack of improvement at 2 months does not rule out Bell palsy, it should prompt a detailed imaging evaluation to rule out an underlying condition such as tumor (in the pons, cerebellopontine angle, parotid gland, middle ear, or petrosal bone), infection (herpes simplex, varicella zoster, Ramsey-Hunt syndrome, or otitis media), trauma, or systemic disease (diabetes mellitus, multiple sclerosis, sarcoidosis, or systemic lupus erythematosus).⁴

According to a review of common causes of facial nerve palsy, the most common finding in 224 patients misdiagnosed with Bell palsy was tumor (38%).⁵ This indicates the value of magnetic resonance imaging of the head rather than CT when secondary facial nerve palsy is suspected, as CT is not sensitive to brainstem lesions.

A 59-year-old woman presented with drooling out of the left side of her mouth and inability to close her left eye. She had no ear pain, hearing loss, or skin rash. The facial palsy affected all branches of the left facial nerve. This explained her inability to close her left eyelid and the generalized weakness of the left side of the face, including her forehead and angle of the mouth. No other signs of pontine dysfunction were noted.

See related editorial

The symptoms had begun 2 months earlier, and computed tomography (CT) of the head performed at a nearby clinic 3 days after the onset of symptoms showed no abnormalities. She was given a diagnosis of incomplete Bell palsy and was prescribed prednisolone and valacyclovir. However, her symptoms had not improved after 2 months of treatment, and so she presented to our hospital.

Physical examination revealed moderate nerve dysfunction (House-Brackmann grade III, with grade I normal and grade VI total paralysis) and generalized weakness on the left side of her face including her forehead.¹ She had no loss in facial sensation or hearing and no ataxia or ocular motility disorders.

BELL PALSY

Peripheral facial nerve palsy is classified either as Bell palsy, which is idiopathic, or as secondary facial nerve palsy. Because Bell palsy accounts for 60% to 70% of all cases,² treatment with oral steroids is indicated when no abnormal findings other than lateral peripheral facial nerve palsy are observed. Antiviral drugs may provide added benefit, although academic societies do not currently recommend combined therapy.³ However, 85% of patients with Bell palsy improve within 3 weeks without treatment, and 94% of patients with incomplete Bell palsy—defined by normal to severe dysfunction, ie, not total paralysis, based on House-Brackmann score—eventually achieve complete remission.²

Therefore, although progression of symptoms or lack of improvement at 2 months does not rule out Bell palsy, it should prompt a detailed imaging evaluation to rule out an underlying condition such as tumor (in the pons, cerebellopontine angle, parotid gland, middle ear, or petrosal bone), infection (herpes simplex, varicella zoster, Ramsey-Hunt syndrome, or otitis media), trauma, or systemic disease (diabetes mellitus, multiple sclerosis, sarcoidosis, or systemic lupus erythematosus).⁴

According to a review of common causes of facial nerve palsy, the most common finding in 224 patients misdiagnosed with Bell palsy was tumor (38%).⁵ This indicates the value of magnetic resonance imaging of the head rather than CT when secondary facial nerve palsy is suspected, as CT is not sensitive to brainstem lesions.

The article in this issue by Shikino et al¹ on a mimic of Bell palsy gives us an opportunity to discuss the question posed by the title of this editorial. The obvious short answer is “no.”

See related article

Any experienced clinician will acknowledge that the extent of the physical examination and the extent of information obtained during the history should be determined by the problem being evaluated at the time and by the setting in which it takes place. The difficulty, of course, is that this relies on the judgment of the clinician, and this may or may not pass the test of hindsight.

Verghese et al² have eloquently emphasized the hazards of an incomplete or inadequate physical examination. Their study was not designed to determine the prevalence of deficient physical examination, either in its extent or its accuracy. Their purpose was to promote the necessity of proper teaching and performance of examination technique.

The neurologic examination is one of the last bastions of physical assessment.³ Despite remarkable advances in imaging and physiologic techniques, the neurologic physical assessment remains critical for diagnosis and management of the neurologic patient. One of my mentors in neurology used to urge residents to examine patients and record the results of the examination as if every patient would subsequently be the subject of a clinicopathologic conference. Anyone who has reviewed a case for a conference or a case report can identify with that sentiment, wishing that some missing piece of information were available. Yet everyone also recognizes the difficulties, if not the impossibility, of achieving that ideal result.

But recording information obtained during the history or physical examination is important even in the course of a daily routine evaluation. I find myself wishing that a previous examiner had commented on whether the muscle stretch reflexes were somewhat hypoactive (eg, “1+”) or on the brisk side (“3+”) rather than “physiologic.” Was the right leg actually globally weak (“4/5”), or was there a discrepancy between proximal and distal muscles or between the physiologic flexors and the extensors?

This can make a big difference in following a patient’s neurologic progress, even over a short time span. It might tell us whether we are dealing with weakness from a peripheral neuromuscular disorder (eg, Guillain-Barré syndrome) or from a myelopathy due to impending spinal cord compression.

It should be mentioned that although Guillain-Barré syndrome is characterized as an ascending paralysis, ie, beginning distally and spreading rostrally, it is one of the few peripheral neuropathies that can present with predominant proximal weakness. It is, in fact, a radiculoneuropathy. But spinal cord (upper motor neuron) disorders preferentially weaken the physiologic flexors of the lower limbs (hamstrings and ankle dorsiflexors), leading to the characteristic extensor posture of the spastic leg. Other findings that can help differential peripheral vs spinal cord disorders include distal sensory loss and hypoactive or absent muscle stretch reflexes in a peripheral neuropathy, compared with dissociated sensory loss (eg, impaired pain and temperature sensation in one leg with reduced vibration perception and proprioception in the other) along with hyperreflexia with cord lesions.

Therefore, a careful neurologic examination may tell us whether magnetic resonance imaging of the spine or an electrodiagnostic study should be the next step.

Shikino et al describe a patient who presented with what looked like idiopathic facial palsy (Bell palsy) but turned out to be the result of a primary central nervous system (CNS) cause. Would a more detailed neurologic examination have identified this as a CNS disorder? Would more specific information about the degree and distribution of facial paresis have facilitated earlier recognition of a progressive process, making idiopathic facial palsy less likely? How much elevation of the eyebrow occurred with voluntary activation, how many millimeters of sclera were visible with gentle eyelid closure? How much space remained between the lips on attempted lip closure?

Upper facial muscle weakness is typically not seen in CNS disorders, although facial nerve or nucleus involvement at the pontine level can impair eyelid and frontalis function. Such lesions would usually be accompanied by “neighborhood” signs such as subtle ipsilateral lateral rectus or abducens palsy, involvement of the vestibular nuclei with vertigo, or facial sensory impairment from disruption of the descending trigeminal nucleus and tract. These would be “pertinent negatives” for excluding a brainstem lesion, and ipsilateral motor, sensory, or “higher cortical” functions would obviously signal a supratentorial CNS disorder.

In the case described by Shikino et al, observation and recording of the amount of facial motor function at the initial visit, 3 days after onset, could facilitate recognition of an aberrant course even a few days later and prompt further investigation at an early follow-up visit (idiopathic palsy is almost invariably maximal by 72 hours). I would assume that no additional clinical information was available to the subsequent examiner in this case, 2 months later, rather than suggesting that such information was omitted for the sake of parsimony.

Would any of this have made a difference? Probably not, but we need all the help we can get in medicine. Remember that every bit of information you obtain from your history or physical examination that you do not record disappears with you and is irretrievably lost.

The article in this issue by Shikino et al¹ on a mimic of Bell palsy gives us an opportunity to discuss the question posed by the title of this editorial. The obvious short answer is “no.”

See related article

Any experienced clinician will acknowledge that the extent of the physical examination and the extent of information obtained during the history should be determined by the problem being evaluated at the time and by the setting in which it takes place. The difficulty, of course, is that this relies on the judgment of the clinician, and this may or may not pass the test of hindsight.

Verghese et al² have eloquently emphasized the hazards of an incomplete or inadequate physical examination. Their study was not designed to determine the prevalence of deficient physical examination, either in its extent or its accuracy. Their purpose was to promote the necessity of proper teaching and performance of examination technique.

The neurologic examination is one of the last bastions of physical assessment.³ Despite remarkable advances in imaging and physiologic techniques, the neurologic physical assessment remains critical for diagnosis and management of the neurologic patient. One of my mentors in neurology used to urge residents to examine patients and record the results of the examination as if every patient would subsequently be the subject of a clinicopathologic conference. Anyone who has reviewed a case for a conference or a case report can identify with that sentiment, wishing that some missing piece of information were available. Yet everyone also recognizes the difficulties, if not the impossibility, of achieving that ideal result.

But recording information obtained during the history or physical examination is important even in the course of a daily routine evaluation. I find myself wishing that a previous examiner had commented on whether the muscle stretch reflexes were somewhat hypoactive (eg, “1+”) or on the brisk side (“3+”) rather than “physiologic.” Was the right leg actually globally weak (“4/5”), or was there a discrepancy between proximal and distal muscles or between the physiologic flexors and the extensors?

This can make a big difference in following a patient’s neurologic progress, even over a short time span. It might tell us whether we are dealing with weakness from a peripheral neuromuscular disorder (eg, Guillain-Barré syndrome) or from a myelopathy due to impending spinal cord compression.

It should be mentioned that although Guillain-Barré syndrome is characterized as an ascending paralysis, ie, beginning distally and spreading rostrally, it is one of the few peripheral neuropathies that can present with predominant proximal weakness. It is, in fact, a radiculoneuropathy. But spinal cord (upper motor neuron) disorders preferentially weaken the physiologic flexors of the lower limbs (hamstrings and ankle dorsiflexors), leading to the characteristic extensor posture of the spastic leg. Other findings that can help differential peripheral vs spinal cord disorders include distal sensory loss and hypoactive or absent muscle stretch reflexes in a peripheral neuropathy, compared with dissociated sensory loss (eg, impaired pain and temperature sensation in one leg with reduced vibration perception and proprioception in the other) along with hyperreflexia with cord lesions.

Therefore, a careful neurologic examination may tell us whether magnetic resonance imaging of the spine or an electrodiagnostic study should be the next step.

Shikino et al describe a patient who presented with what looked like idiopathic facial palsy (Bell palsy) but turned out to be the result of a primary central nervous system (CNS) cause. Would a more detailed neurologic examination have identified this as a CNS disorder? Would more specific information about the degree and distribution of facial paresis have facilitated earlier recognition of a progressive process, making idiopathic facial palsy less likely? How much elevation of the eyebrow occurred with voluntary activation, how many millimeters of sclera were visible with gentle eyelid closure? How much space remained between the lips on attempted lip closure?

Upper facial muscle weakness is typically not seen in CNS disorders, although facial nerve or nucleus involvement at the pontine level can impair eyelid and frontalis function. Such lesions would usually be accompanied by “neighborhood” signs such as subtle ipsilateral lateral rectus or abducens palsy, involvement of the vestibular nuclei with vertigo, or facial sensory impairment from disruption of the descending trigeminal nucleus and tract. These would be “pertinent negatives” for excluding a brainstem lesion, and ipsilateral motor, sensory, or “higher cortical” functions would obviously signal a supratentorial CNS disorder.

In the case described by Shikino et al, observation and recording of the amount of facial motor function at the initial visit, 3 days after onset, could facilitate recognition of an aberrant course even a few days later and prompt further investigation at an early follow-up visit (idiopathic palsy is almost invariably maximal by 72 hours). I would assume that no additional clinical information was available to the subsequent examiner in this case, 2 months later, rather than suggesting that such information was omitted for the sake of parsimony.

Would any of this have made a difference? Probably not, but we need all the help we can get in medicine. Remember that every bit of information you obtain from your history or physical examination that you do not record disappears with you and is irretrievably lost.

The article in this issue by Shikino et al¹ on a mimic of Bell palsy gives us an opportunity to discuss the question posed by the title of this editorial. The obvious short answer is “no.”

See related article

Any experienced clinician will acknowledge that the extent of the physical examination and the extent of information obtained during the history should be determined by the problem being evaluated at the time and by the setting in which it takes place. The difficulty, of course, is that this relies on the judgment of the clinician, and this may or may not pass the test of hindsight.

Verghese et al² have eloquently emphasized the hazards of an incomplete or inadequate physical examination. Their study was not designed to determine the prevalence of deficient physical examination, either in its extent or its accuracy. Their purpose was to promote the necessity of proper teaching and performance of examination technique.

The neurologic examination is one of the last bastions of physical assessment.³ Despite remarkable advances in imaging and physiologic techniques, the neurologic physical assessment remains critical for diagnosis and management of the neurologic patient. One of my mentors in neurology used to urge residents to examine patients and record the results of the examination as if every patient would subsequently be the subject of a clinicopathologic conference. Anyone who has reviewed a case for a conference or a case report can identify with that sentiment, wishing that some missing piece of information were available. Yet everyone also recognizes the difficulties, if not the impossibility, of achieving that ideal result.

But recording information obtained during the history or physical examination is important even in the course of a daily routine evaluation. I find myself wishing that a previous examiner had commented on whether the muscle stretch reflexes were somewhat hypoactive (eg, “1+”) or on the brisk side (“3+”) rather than “physiologic.” Was the right leg actually globally weak (“4/5”), or was there a discrepancy between proximal and distal muscles or between the physiologic flexors and the extensors?

This can make a big difference in following a patient’s neurologic progress, even over a short time span. It might tell us whether we are dealing with weakness from a peripheral neuromuscular disorder (eg, Guillain-Barré syndrome) or from a myelopathy due to impending spinal cord compression.

It should be mentioned that although Guillain-Barré syndrome is characterized as an ascending paralysis, ie, beginning distally and spreading rostrally, it is one of the few peripheral neuropathies that can present with predominant proximal weakness. It is, in fact, a radiculoneuropathy. But spinal cord (upper motor neuron) disorders preferentially weaken the physiologic flexors of the lower limbs (hamstrings and ankle dorsiflexors), leading to the characteristic extensor posture of the spastic leg. Other findings that can help differential peripheral vs spinal cord disorders include distal sensory loss and hypoactive or absent muscle stretch reflexes in a peripheral neuropathy, compared with dissociated sensory loss (eg, impaired pain and temperature sensation in one leg with reduced vibration perception and proprioception in the other) along with hyperreflexia with cord lesions.

Therefore, a careful neurologic examination may tell us whether magnetic resonance imaging of the spine or an electrodiagnostic study should be the next step.

Shikino et al describe a patient who presented with what looked like idiopathic facial palsy (Bell palsy) but turned out to be the result of a primary central nervous system (CNS) cause. Would a more detailed neurologic examination have identified this as a CNS disorder? Would more specific information about the degree and distribution of facial paresis have facilitated earlier recognition of a progressive process, making idiopathic facial palsy less likely? How much elevation of the eyebrow occurred with voluntary activation, how many millimeters of sclera were visible with gentle eyelid closure? How much space remained between the lips on attempted lip closure?

Upper facial muscle weakness is typically not seen in CNS disorders, although facial nerve or nucleus involvement at the pontine level can impair eyelid and frontalis function. Such lesions would usually be accompanied by “neighborhood” signs such as subtle ipsilateral lateral rectus or abducens palsy, involvement of the vestibular nuclei with vertigo, or facial sensory impairment from disruption of the descending trigeminal nucleus and tract. These would be “pertinent negatives” for excluding a brainstem lesion, and ipsilateral motor, sensory, or “higher cortical” functions would obviously signal a supratentorial CNS disorder.

In the case described by Shikino et al, observation and recording of the amount of facial motor function at the initial visit, 3 days after onset, could facilitate recognition of an aberrant course even a few days later and prompt further investigation at an early follow-up visit (idiopathic palsy is almost invariably maximal by 72 hours). I would assume that no additional clinical information was available to the subsequent examiner in this case, 2 months later, rather than suggesting that such information was omitted for the sake of parsimony.

Would any of this have made a difference? Probably not, but we need all the help we can get in medicine. Remember that every bit of information you obtain from your history or physical examination that you do not record disappears with you and is irretrievably lost.

A 57-year-old man was transferred to our  hospital with leg pain and confusion. His family reported that he had injured his leg while launching a motorized personal watercraft at the North Carolina seashore 2 days before. He had a history of cirrhosis secondary to hepatitis C and alcohol abuse.

Blood and wound cultures eventually grew V vulnificus, and surgical pathology confirmed the diagnosis of necrotizing fasciitis (Figure 2).

RISE IN V VULNIFICUS INFECTIONS IS ATTRIBUTED TO GLOBAL WARMING

V vulnificus infection occurs most commonly from consuming raw shellfish, especially oysters, but it also occurs after exposure of an open wound to contaminated salt water. The pathogen is a gram-negative bacterium that resides in coastal waters worldwide, but in the United States it is usually seen on the Pacific and Gulf coasts¹ during the summer.²

Although only 58 cases of V vulnificus infection were reported to the US Centers for Disease Control and Prevention in 1997, the number more than doubled to 124 in 2014.¹ This rise is suspected to be due in part to warmer coastal waters associated with global warming.²

Various marine pathogens can cause wound infections, but V vulnificus is most commonly implicated in deaths and hospitalizations.² Immunocompromised patients and those with liver disease are at particularly high risk of rapid progression to septic shock.

First-line antibiotics are doxycycline plus a third-generation cephalosporin.³ Studies have shown a direct correlation between delay of antibiotics and death,⁴ and early surgery is critical.⁵

Given the rising incidence of V vulnificus infection, it is increasingly important for providers across the United States to be aware of this infection.

A 57-year-old man was transferred to our  hospital with leg pain and confusion. His family reported that he had injured his leg while launching a motorized personal watercraft at the North Carolina seashore 2 days before. He had a history of cirrhosis secondary to hepatitis C and alcohol abuse.

Blood and wound cultures eventually grew V vulnificus, and surgical pathology confirmed the diagnosis of necrotizing fasciitis (Figure 2).

RISE IN V VULNIFICUS INFECTIONS IS ATTRIBUTED TO GLOBAL WARMING

V vulnificus infection occurs most commonly from consuming raw shellfish, especially oysters, but it also occurs after exposure of an open wound to contaminated salt water. The pathogen is a gram-negative bacterium that resides in coastal waters worldwide, but in the United States it is usually seen on the Pacific and Gulf coasts¹ during the summer.²

Although only 58 cases of V vulnificus infection were reported to the US Centers for Disease Control and Prevention in 1997, the number more than doubled to 124 in 2014.¹ This rise is suspected to be due in part to warmer coastal waters associated with global warming.²

Various marine pathogens can cause wound infections, but V vulnificus is most commonly implicated in deaths and hospitalizations.² Immunocompromised patients and those with liver disease are at particularly high risk of rapid progression to septic shock.

First-line antibiotics are doxycycline plus a third-generation cephalosporin.³ Studies have shown a direct correlation between delay of antibiotics and death,⁴ and early surgery is critical.⁵

Given the rising incidence of V vulnificus infection, it is increasingly important for providers across the United States to be aware of this infection.

A 57-year-old man was transferred to our  hospital with leg pain and confusion. His family reported that he had injured his leg while launching a motorized personal watercraft at the North Carolina seashore 2 days before. He had a history of cirrhosis secondary to hepatitis C and alcohol abuse.

Blood and wound cultures eventually grew V vulnificus, and surgical pathology confirmed the diagnosis of necrotizing fasciitis (Figure 2).

RISE IN V VULNIFICUS INFECTIONS IS ATTRIBUTED TO GLOBAL WARMING

V vulnificus infection occurs most commonly from consuming raw shellfish, especially oysters, but it also occurs after exposure of an open wound to contaminated salt water. The pathogen is a gram-negative bacterium that resides in coastal waters worldwide, but in the United States it is usually seen on the Pacific and Gulf coasts¹ during the summer.²

Although only 58 cases of V vulnificus infection were reported to the US Centers for Disease Control and Prevention in 1997, the number more than doubled to 124 in 2014.¹ This rise is suspected to be due in part to warmer coastal waters associated with global warming.²

Various marine pathogens can cause wound infections, but V vulnificus is most commonly implicated in deaths and hospitalizations.² Immunocompromised patients and those with liver disease are at particularly high risk of rapid progression to septic shock.

First-line antibiotics are doxycycline plus a third-generation cephalosporin.³ Studies have shown a direct correlation between delay of antibiotics and death,⁴ and early surgery is critical.⁵

Given the rising incidence of V vulnificus infection, it is increasingly important for providers across the United States to be aware of this infection.

Four decades after reverse T₃ (3,3´5´-triiodothyronine) was discovered, its physiologic and clinical relevance remains unclear and is still being studied. But scientific uncertainty has not stopped writers in the consumer press and on the Internet from making unsubstantiated claims about this hormone. Many patients believe their hypothyroid symptoms are due to high levels of reverse T₃ and want to be tested for it, and some even bring in test results from independent laboratories.  

HOW THYROID HORMONES WERE DISCOVERED

In 1970, Braverman et al⁹ showed that T₄ is converted to T₃ in athyreotic humans, and Sterling et al¹⁰ demonstrated the same in healthy humans. During that decade, techniques for measuring T₄ were refined,¹¹ and a specific radioimmunoassay for reverse T₃ allowed a glimpse of its physiologic role.¹² In 1975, Chopra et al¹³ noted reciprocal changes in the levels of T₃ and reverse T₃ in systemic illnesses—ie, when people are sick, their T₃ levels go down and their reverse T₃ levels go up.

The end of the 70s was marked by a surge of interest in T₄ metabolites, including the development of a radioimmunoassay for 3,3´-diiodothyronine (3-3´ T₂).¹⁸

The observed reciprocal changes in serum levels of T₃ and reverse T₃ suggested that T₄ degradation is regulated into activating (T₃) or inactivating (reverse T₃) pathways, and that these changes are a presumed homeostatic process of energy conservation.¹⁹

HOW THYROID HORMONES ARE METABOLIZED

In the thyroid gland, for thyroid hormones to be synthesized, iodide must be oxidized and incorporated into the precursors 3-monoiodotyrosine (MIT) and 3,5-diiodotyrosine (DIT). This process is mediated by the enzyme thyroid peroxidase in the presence of hydrogen peroxide.²⁰

The thyroid can make T₄ and some T₃

T₄ is the main iodothyronine produced by the thyroid gland, at a rate of 80 to 100 µg per day.²¹ It is synthesized from the fusion of 2 DIT molecules.

The thyroid can also make T₃ by fusing 1 DIT and 1 MIT molecule, but this process accounts for no more than 20% of the circulating T₃ in humans. The rest of T₃, and 95% to 98% of all reverse T₃, is derived from peripheral conversion of T₄ through deiodination.

T₄ is converted to T₃ or reverse T₃

The metabolic transformation of thyroid hormones in peripheral tissues determines their biologic potency and regulates their biologic effects.

The number 4 in T₄ means it has 4 iodine atoms. It can lose 1 of them, yielding either T₃ or reverse T₃, depending on which iodine atom it loses (Figure 3). Loss of iodine from the five-prime (5´) position on its outer ring yields T₃, the most potent thyroid hormone, produced at a rate of 30 to 40 µg per day.²¹ On the other hand, when T₄ loses an iodine atom from the five (5) position on its inner ring it yields reverse T₃, produced at a rate slightly less than that of T₃, 28 to 40 µg per day.²¹ Reverse T₃ is inactive.

Both T₃ and reverse T₃ can shed more iodine atoms, forming in turn various isomers of T₂, T₁, and ultimately T₀. Other pathways for thyroid hormone metabolism include glucuronidation, sulfation, oxidative deamination, and ether bond cleavage.^20–22

D1 and D2 catalyze T₃, D3 catalyzes reverse T₃

Three types of enzymes that mediate deiodination have been identified and designated D1, D2, and D3. In humans they are expressed in variable amounts throughout the body:

• D1 mainly in the liver, kidneys, thyroid, and pituitary, but notably absent in the central nervous system
• D2 in the central nervous system, pituitary, brown adipose tissue, thyroid, placenta, skeletal muscle, and heart
• D3 in the central nervous system, skin, hemangiomas, fetal liver, placenta, and fetal tissues.²³

D1 and D2 are responsible for converting T₄ to T₃, and D3 is responsible for converting T₄ to reverse T₃.

Plasma concentrations of free T₄ and free T₃ are relatively constant; however, tissue concentrations of free T₃ vary in different tissues according to the amount of hormone transported and the activity of local deiodinases.²³ Most thyroid hormone actions are initiated after T₃ binds to its nuclear receptor. In this setting, deiodinases play a critical role in maintaining tissue and cellular thyroid hormone levels, so that thyroid hormone signaling can change irrespective of serum hormonal concentrations.^22–24 For example, in the central nervous system, production of T₃ by local D2 is significantly relevant for T₃ homeostasis.^22,23

Deiodinases also modulate the tissue-specific concentrations of T₃ in response to iodine deficiency and to changes in thyroid state.²³ During iodine deficiency and hypothyroidism, tissues that express D2, especially brain tissues, increase the activity of this enzyme in order to increase local conversion of T₄ to T₃. In hyperthyroidism, D1 overexpression contributes to the relative excess of T₃ production, while D3 up-regulation in the brain protects the central nervous system from excessive amounts of thyroid hormone.²³

D3 is the main physiologic inactivator of thyroid hormones. This enzyme plays a central role in protecting tissues from an excess of thyroid hormone.^23,24 This mechanism is crucial for fetal development and explains the high expression of D3 in the human placenta and fetal tissues.

In adult tissues, the importance of D3 in the regulation of thyroid hormone homeostasis becomes apparent under certain pathophysiologic conditions, such as nonthyroidal illness and malnutrition.

Whenever a reduction in metabolism is homeostatically desirable, such as in critically ill patients or during starvation, conversion to T₃ is reduced and, alternatively, conversion to reverse T₃ is increased. This pathway represents a metabolic adaptation that may protect the tissues from the catabolic effects of thyroid hormone that could otherwise worsen the patient’s basic clinical condition.

Euthyroid sick syndrome or hypothyroid?

In a variety of systemic illnesses, some patients with low T₃, low or normal T₄, and normal thyroid-stimulating hormone (TSH) levels could in fact be “sick euthyroid” rather than hypothyroid. The first reports of the euthyroid sick syndrome or low T₃ syndrome date back to about 1976, and even though assays for reverse T₃ were not widely available, some authors linked the syndrome to high levels of reverse T₃.^15,16 The syndrome is also known as nonthyroidal illness syndrome.

Advances in techniques for measuring T₃, reverse T₃, and other iodothyronines filled a gap in the understanding of the alterations that occur in thyroid hormone economy during severe nonthyroidal diseases. In 1982, Wartofsky and Burman²⁵ reviewed the alterations in thyroid function in patients with systemic illness and discussed other factors that may alter thyroid economy, such as age, stress, and diverse drugs.

More recently, the low-T₃ syndrome was revisited with a generalized concept regarding the role of D3 in the syndrome.²⁶ D3, normally undetectable in mature tissues, is reactivated in diverse cell types in response to injury and is responsible for a fall in serum T₃ levels. Hypoxia induces D3 activity and mRNA in vitro and in vivo.²⁷ Recent studies have focused on the role of cytokines in the low T₃ syndrome. For instance, interleukin 6 reduces D1 and D2 activity and increases D3 activity in vitro.²⁸

In the outpatient setting, diverse conditions may affect thyroid hormone homeostasis, compatible with mild or atypical forms of low-T₃ syndrome, including caloric deprivation, heart failure, and human immunodeficiency virus infection.²⁹

POSSIBLE CLINICAL UTILITY OF MEASURING REVERSE T₃

In inpatients

Unfortunately, measuring serum reverse T₃ levels has not, in general, proven clinically useful for the diagnosis of hypothyroidism in systemically ill patients. Burmeister³⁰ demonstrated, in a retrospective study, that when illness complicates the interpretation of thyroid function tests, serum reverse T₃ measurements do not reliably distinguish the hypothyroid sick patient from the euthyroid sick patient. The best way to make the diagnosis, Burmeister suggested, is by clinical assessment, combined use of free T₄ and TSH measurements, and patient follow-up.

In the outpatient setting, the utility of reverse T3 measurements is controversial. In intensive care units, the differential diagnosis between hypothyroidism and nonthyroidal illness syndrome can sometimes be difficult. Reverse T₃ levels can be low, normal, or high regardless of the thyroidal state of the patient.³⁰ Moreover, endogenous changes in the hypothalamic-pituitary-thyroid axis may be further complicated by medications commonly used in intensive care units, such as dopamine and glucocorticoids. Changes in thyroid function should be evaluated in the context of the patient’s clinical condition (Table 1).²⁰ But regardless of the T₃ level, treatment with T₃ or T₄ should not be started without taking into consideration the patient’s general clinical context; controlled trials have not shown such therapy to be beneficial.²⁰

In outpatients

In noncritical conditions that may be associated with mild forms of low T₃ syndrome, patients generally present with low T₃ concentrations concurrently with low or normal TSH. Not infrequently, however, patients present with a serum reverse T₃ measurement and impute their symptoms of hypothyroidism to “abnormal” reverse T₃ levels, in spite of normal TSH levels.

There is no rationale for measuring reverse T₃ to initiate or to adjust levothyroxine therapy—the single test relevant for these purposes is the TSH measurement. The risks of basing treatment decisions on reverse T₃ levels include the use of excessive doses of levothyroxine that may lead to a state of subclinical or even clinical hyperthyroidism.

TAKE-HOME MESSAGE

The existence of an inactivating pathway of thyroid hormones represents a homeostatic mechanism, and in selected circumstances measuring serum reverse T₃ may be useful, such as in euthyroid sick patients. The discovery of the molecular mechanisms that lead to the reactivation of D3 in illness is an important field of research.

Four decades after reverse T₃ (3,3´5´-triiodothyronine) was discovered, its physiologic and clinical relevance remains unclear and is still being studied. But scientific uncertainty has not stopped writers in the consumer press and on the Internet from making unsubstantiated claims about this hormone. Many patients believe their hypothyroid symptoms are due to high levels of reverse T₃ and want to be tested for it, and some even bring in test results from independent laboratories.  

HOW THYROID HORMONES WERE DISCOVERED

In 1970, Braverman et al⁹ showed that T₄ is converted to T₃ in athyreotic humans, and Sterling et al¹⁰ demonstrated the same in healthy humans. During that decade, techniques for measuring T₄ were refined,¹¹ and a specific radioimmunoassay for reverse T₃ allowed a glimpse of its physiologic role.¹² In 1975, Chopra et al¹³ noted reciprocal changes in the levels of T₃ and reverse T₃ in systemic illnesses—ie, when people are sick, their T₃ levels go down and their reverse T₃ levels go up.

The end of the 70s was marked by a surge of interest in T₄ metabolites, including the development of a radioimmunoassay for 3,3´-diiodothyronine (3-3´ T₂).¹⁸

The observed reciprocal changes in serum levels of T₃ and reverse T₃ suggested that T₄ degradation is regulated into activating (T₃) or inactivating (reverse T₃) pathways, and that these changes are a presumed homeostatic process of energy conservation.¹⁹

HOW THYROID HORMONES ARE METABOLIZED

In the thyroid gland, for thyroid hormones to be synthesized, iodide must be oxidized and incorporated into the precursors 3-monoiodotyrosine (MIT) and 3,5-diiodotyrosine (DIT). This process is mediated by the enzyme thyroid peroxidase in the presence of hydrogen peroxide.²⁰

The thyroid can make T₄ and some T₃

T₄ is the main iodothyronine produced by the thyroid gland, at a rate of 80 to 100 µg per day.²¹ It is synthesized from the fusion of 2 DIT molecules.

The thyroid can also make T₃ by fusing 1 DIT and 1 MIT molecule, but this process accounts for no more than 20% of the circulating T₃ in humans. The rest of T₃, and 95% to 98% of all reverse T₃, is derived from peripheral conversion of T₄ through deiodination.

T₄ is converted to T₃ or reverse T₃

The metabolic transformation of thyroid hormones in peripheral tissues determines their biologic potency and regulates their biologic effects.

The number 4 in T₄ means it has 4 iodine atoms. It can lose 1 of them, yielding either T₃ or reverse T₃, depending on which iodine atom it loses (Figure 3). Loss of iodine from the five-prime (5´) position on its outer ring yields T₃, the most potent thyroid hormone, produced at a rate of 30 to 40 µg per day.²¹ On the other hand, when T₄ loses an iodine atom from the five (5) position on its inner ring it yields reverse T₃, produced at a rate slightly less than that of T₃, 28 to 40 µg per day.²¹ Reverse T₃ is inactive.

Both T₃ and reverse T₃ can shed more iodine atoms, forming in turn various isomers of T₂, T₁, and ultimately T₀. Other pathways for thyroid hormone metabolism include glucuronidation, sulfation, oxidative deamination, and ether bond cleavage.^20–22

D1 and D2 catalyze T₃, D3 catalyzes reverse T₃

Three types of enzymes that mediate deiodination have been identified and designated D1, D2, and D3. In humans they are expressed in variable amounts throughout the body:

• D1 mainly in the liver, kidneys, thyroid, and pituitary, but notably absent in the central nervous system
• D2 in the central nervous system, pituitary, brown adipose tissue, thyroid, placenta, skeletal muscle, and heart
• D3 in the central nervous system, skin, hemangiomas, fetal liver, placenta, and fetal tissues.²³

D1 and D2 are responsible for converting T₄ to T₃, and D3 is responsible for converting T₄ to reverse T₃.

Plasma concentrations of free T₄ and free T₃ are relatively constant; however, tissue concentrations of free T₃ vary in different tissues according to the amount of hormone transported and the activity of local deiodinases.²³ Most thyroid hormone actions are initiated after T₃ binds to its nuclear receptor. In this setting, deiodinases play a critical role in maintaining tissue and cellular thyroid hormone levels, so that thyroid hormone signaling can change irrespective of serum hormonal concentrations.^22–24 For example, in the central nervous system, production of T₃ by local D2 is significantly relevant for T₃ homeostasis.^22,23

Deiodinases also modulate the tissue-specific concentrations of T₃ in response to iodine deficiency and to changes in thyroid state.²³ During iodine deficiency and hypothyroidism, tissues that express D2, especially brain tissues, increase the activity of this enzyme in order to increase local conversion of T₄ to T₃. In hyperthyroidism, D1 overexpression contributes to the relative excess of T₃ production, while D3 up-regulation in the brain protects the central nervous system from excessive amounts of thyroid hormone.²³

D3 is the main physiologic inactivator of thyroid hormones. This enzyme plays a central role in protecting tissues from an excess of thyroid hormone.^23,24 This mechanism is crucial for fetal development and explains the high expression of D3 in the human placenta and fetal tissues.

In adult tissues, the importance of D3 in the regulation of thyroid hormone homeostasis becomes apparent under certain pathophysiologic conditions, such as nonthyroidal illness and malnutrition.

Whenever a reduction in metabolism is homeostatically desirable, such as in critically ill patients or during starvation, conversion to T₃ is reduced and, alternatively, conversion to reverse T₃ is increased. This pathway represents a metabolic adaptation that may protect the tissues from the catabolic effects of thyroid hormone that could otherwise worsen the patient’s basic clinical condition.

Euthyroid sick syndrome or hypothyroid?

In a variety of systemic illnesses, some patients with low T₃, low or normal T₄, and normal thyroid-stimulating hormone (TSH) levels could in fact be “sick euthyroid” rather than hypothyroid. The first reports of the euthyroid sick syndrome or low T₃ syndrome date back to about 1976, and even though assays for reverse T₃ were not widely available, some authors linked the syndrome to high levels of reverse T₃.^15,16 The syndrome is also known as nonthyroidal illness syndrome.

Advances in techniques for measuring T₃, reverse T₃, and other iodothyronines filled a gap in the understanding of the alterations that occur in thyroid hormone economy during severe nonthyroidal diseases. In 1982, Wartofsky and Burman²⁵ reviewed the alterations in thyroid function in patients with systemic illness and discussed other factors that may alter thyroid economy, such as age, stress, and diverse drugs.

More recently, the low-T₃ syndrome was revisited with a generalized concept regarding the role of D3 in the syndrome.²⁶ D3, normally undetectable in mature tissues, is reactivated in diverse cell types in response to injury and is responsible for a fall in serum T₃ levels. Hypoxia induces D3 activity and mRNA in vitro and in vivo.²⁷ Recent studies have focused on the role of cytokines in the low T₃ syndrome. For instance, interleukin 6 reduces D1 and D2 activity and increases D3 activity in vitro.²⁸

In the outpatient setting, diverse conditions may affect thyroid hormone homeostasis, compatible with mild or atypical forms of low-T₃ syndrome, including caloric deprivation, heart failure, and human immunodeficiency virus infection.²⁹

POSSIBLE CLINICAL UTILITY OF MEASURING REVERSE T₃

In inpatients

Unfortunately, measuring serum reverse T₃ levels has not, in general, proven clinically useful for the diagnosis of hypothyroidism in systemically ill patients. Burmeister³⁰ demonstrated, in a retrospective study, that when illness complicates the interpretation of thyroid function tests, serum reverse T₃ measurements do not reliably distinguish the hypothyroid sick patient from the euthyroid sick patient. The best way to make the diagnosis, Burmeister suggested, is by clinical assessment, combined use of free T₄ and TSH measurements, and patient follow-up.

In the outpatient setting, the utility of reverse T3 measurements is controversial. In intensive care units, the differential diagnosis between hypothyroidism and nonthyroidal illness syndrome can sometimes be difficult. Reverse T₃ levels can be low, normal, or high regardless of the thyroidal state of the patient.³⁰ Moreover, endogenous changes in the hypothalamic-pituitary-thyroid axis may be further complicated by medications commonly used in intensive care units, such as dopamine and glucocorticoids. Changes in thyroid function should be evaluated in the context of the patient’s clinical condition (Table 1).²⁰ But regardless of the T₃ level, treatment with T₃ or T₄ should not be started without taking into consideration the patient’s general clinical context; controlled trials have not shown such therapy to be beneficial.²⁰

In outpatients

In noncritical conditions that may be associated with mild forms of low T₃ syndrome, patients generally present with low T₃ concentrations concurrently with low or normal TSH. Not infrequently, however, patients present with a serum reverse T₃ measurement and impute their symptoms of hypothyroidism to “abnormal” reverse T₃ levels, in spite of normal TSH levels.

There is no rationale for measuring reverse T₃ to initiate or to adjust levothyroxine therapy—the single test relevant for these purposes is the TSH measurement. The risks of basing treatment decisions on reverse T₃ levels include the use of excessive doses of levothyroxine that may lead to a state of subclinical or even clinical hyperthyroidism.

TAKE-HOME MESSAGE

The existence of an inactivating pathway of thyroid hormones represents a homeostatic mechanism, and in selected circumstances measuring serum reverse T₃ may be useful, such as in euthyroid sick patients. The discovery of the molecular mechanisms that lead to the reactivation of D3 in illness is an important field of research.

Four decades after reverse T₃ (3,3´5´-triiodothyronine) was discovered, its physiologic and clinical relevance remains unclear and is still being studied. But scientific uncertainty has not stopped writers in the consumer press and on the Internet from making unsubstantiated claims about this hormone. Many patients believe their hypothyroid symptoms are due to high levels of reverse T₃ and want to be tested for it, and some even bring in test results from independent laboratories.  

HOW THYROID HORMONES WERE DISCOVERED

In 1970, Braverman et al⁹ showed that T₄ is converted to T₃ in athyreotic humans, and Sterling et al¹⁰ demonstrated the same in healthy humans. During that decade, techniques for measuring T₄ were refined,¹¹ and a specific radioimmunoassay for reverse T₃ allowed a glimpse of its physiologic role.¹² In 1975, Chopra et al¹³ noted reciprocal changes in the levels of T₃ and reverse T₃ in systemic illnesses—ie, when people are sick, their T₃ levels go down and their reverse T₃ levels go up.

The end of the 70s was marked by a surge of interest in T₄ metabolites, including the development of a radioimmunoassay for 3,3´-diiodothyronine (3-3´ T₂).¹⁸

The observed reciprocal changes in serum levels of T₃ and reverse T₃ suggested that T₄ degradation is regulated into activating (T₃) or inactivating (reverse T₃) pathways, and that these changes are a presumed homeostatic process of energy conservation.¹⁹

HOW THYROID HORMONES ARE METABOLIZED

In the thyroid gland, for thyroid hormones to be synthesized, iodide must be oxidized and incorporated into the precursors 3-monoiodotyrosine (MIT) and 3,5-diiodotyrosine (DIT). This process is mediated by the enzyme thyroid peroxidase in the presence of hydrogen peroxide.²⁰

The thyroid can make T₄ and some T₃

T₄ is the main iodothyronine produced by the thyroid gland, at a rate of 80 to 100 µg per day.²¹ It is synthesized from the fusion of 2 DIT molecules.

The thyroid can also make T₃ by fusing 1 DIT and 1 MIT molecule, but this process accounts for no more than 20% of the circulating T₃ in humans. The rest of T₃, and 95% to 98% of all reverse T₃, is derived from peripheral conversion of T₄ through deiodination.

T₄ is converted to T₃ or reverse T₃

The metabolic transformation of thyroid hormones in peripheral tissues determines their biologic potency and regulates their biologic effects.

The number 4 in T₄ means it has 4 iodine atoms. It can lose 1 of them, yielding either T₃ or reverse T₃, depending on which iodine atom it loses (Figure 3). Loss of iodine from the five-prime (5´) position on its outer ring yields T₃, the most potent thyroid hormone, produced at a rate of 30 to 40 µg per day.²¹ On the other hand, when T₄ loses an iodine atom from the five (5) position on its inner ring it yields reverse T₃, produced at a rate slightly less than that of T₃, 28 to 40 µg per day.²¹ Reverse T₃ is inactive.

Both T₃ and reverse T₃ can shed more iodine atoms, forming in turn various isomers of T₂, T₁, and ultimately T₀. Other pathways for thyroid hormone metabolism include glucuronidation, sulfation, oxidative deamination, and ether bond cleavage.^20–22

D1 and D2 catalyze T₃, D3 catalyzes reverse T₃

Three types of enzymes that mediate deiodination have been identified and designated D1, D2, and D3. In humans they are expressed in variable amounts throughout the body:

• D1 mainly in the liver, kidneys, thyroid, and pituitary, but notably absent in the central nervous system
• D2 in the central nervous system, pituitary, brown adipose tissue, thyroid, placenta, skeletal muscle, and heart
• D3 in the central nervous system, skin, hemangiomas, fetal liver, placenta, and fetal tissues.²³

D1 and D2 are responsible for converting T₄ to T₃, and D3 is responsible for converting T₄ to reverse T₃.

Plasma concentrations of free T₄ and free T₃ are relatively constant; however, tissue concentrations of free T₃ vary in different tissues according to the amount of hormone transported and the activity of local deiodinases.²³ Most thyroid hormone actions are initiated after T₃ binds to its nuclear receptor. In this setting, deiodinases play a critical role in maintaining tissue and cellular thyroid hormone levels, so that thyroid hormone signaling can change irrespective of serum hormonal concentrations.^22–24 For example, in the central nervous system, production of T₃ by local D2 is significantly relevant for T₃ homeostasis.^22,23

Deiodinases also modulate the tissue-specific concentrations of T₃ in response to iodine deficiency and to changes in thyroid state.²³ During iodine deficiency and hypothyroidism, tissues that express D2, especially brain tissues, increase the activity of this enzyme in order to increase local conversion of T₄ to T₃. In hyperthyroidism, D1 overexpression contributes to the relative excess of T₃ production, while D3 up-regulation in the brain protects the central nervous system from excessive amounts of thyroid hormone.²³

D3 is the main physiologic inactivator of thyroid hormones. This enzyme plays a central role in protecting tissues from an excess of thyroid hormone.^23,24 This mechanism is crucial for fetal development and explains the high expression of D3 in the human placenta and fetal tissues.

In adult tissues, the importance of D3 in the regulation of thyroid hormone homeostasis becomes apparent under certain pathophysiologic conditions, such as nonthyroidal illness and malnutrition.

Whenever a reduction in metabolism is homeostatically desirable, such as in critically ill patients or during starvation, conversion to T₃ is reduced and, alternatively, conversion to reverse T₃ is increased. This pathway represents a metabolic adaptation that may protect the tissues from the catabolic effects of thyroid hormone that could otherwise worsen the patient’s basic clinical condition.

Euthyroid sick syndrome or hypothyroid?

In a variety of systemic illnesses, some patients with low T₃, low or normal T₄, and normal thyroid-stimulating hormone (TSH) levels could in fact be “sick euthyroid” rather than hypothyroid. The first reports of the euthyroid sick syndrome or low T₃ syndrome date back to about 1976, and even though assays for reverse T₃ were not widely available, some authors linked the syndrome to high levels of reverse T₃.^15,16 The syndrome is also known as nonthyroidal illness syndrome.

Advances in techniques for measuring T₃, reverse T₃, and other iodothyronines filled a gap in the understanding of the alterations that occur in thyroid hormone economy during severe nonthyroidal diseases. In 1982, Wartofsky and Burman²⁵ reviewed the alterations in thyroid function in patients with systemic illness and discussed other factors that may alter thyroid economy, such as age, stress, and diverse drugs.

More recently, the low-T₃ syndrome was revisited with a generalized concept regarding the role of D3 in the syndrome.²⁶ D3, normally undetectable in mature tissues, is reactivated in diverse cell types in response to injury and is responsible for a fall in serum T₃ levels. Hypoxia induces D3 activity and mRNA in vitro and in vivo.²⁷ Recent studies have focused on the role of cytokines in the low T₃ syndrome. For instance, interleukin 6 reduces D1 and D2 activity and increases D3 activity in vitro.²⁸

In the outpatient setting, diverse conditions may affect thyroid hormone homeostasis, compatible with mild or atypical forms of low-T₃ syndrome, including caloric deprivation, heart failure, and human immunodeficiency virus infection.²⁹

POSSIBLE CLINICAL UTILITY OF MEASURING REVERSE T₃

In inpatients

Unfortunately, measuring serum reverse T₃ levels has not, in general, proven clinically useful for the diagnosis of hypothyroidism in systemically ill patients. Burmeister³⁰ demonstrated, in a retrospective study, that when illness complicates the interpretation of thyroid function tests, serum reverse T₃ measurements do not reliably distinguish the hypothyroid sick patient from the euthyroid sick patient. The best way to make the diagnosis, Burmeister suggested, is by clinical assessment, combined use of free T₄ and TSH measurements, and patient follow-up.

In the outpatient setting, the utility of reverse T3 measurements is controversial. In intensive care units, the differential diagnosis between hypothyroidism and nonthyroidal illness syndrome can sometimes be difficult. Reverse T₃ levels can be low, normal, or high regardless of the thyroidal state of the patient.³⁰ Moreover, endogenous changes in the hypothalamic-pituitary-thyroid axis may be further complicated by medications commonly used in intensive care units, such as dopamine and glucocorticoids. Changes in thyroid function should be evaluated in the context of the patient’s clinical condition (Table 1).²⁰ But regardless of the T₃ level, treatment with T₃ or T₄ should not be started without taking into consideration the patient’s general clinical context; controlled trials have not shown such therapy to be beneficial.²⁰

In outpatients

In noncritical conditions that may be associated with mild forms of low T₃ syndrome, patients generally present with low T₃ concentrations concurrently with low or normal TSH. Not infrequently, however, patients present with a serum reverse T₃ measurement and impute their symptoms of hypothyroidism to “abnormal” reverse T₃ levels, in spite of normal TSH levels.

There is no rationale for measuring reverse T₃ to initiate or to adjust levothyroxine therapy—the single test relevant for these purposes is the TSH measurement. The risks of basing treatment decisions on reverse T₃ levels include the use of excessive doses of levothyroxine that may lead to a state of subclinical or even clinical hyperthyroidism.

TAKE-HOME MESSAGE

The existence of an inactivating pathway of thyroid hormones represents a homeostatic mechanism, and in selected circumstances measuring serum reverse T₃ may be useful, such as in euthyroid sick patients. The discovery of the molecular mechanisms that lead to the reactivation of D3 in illness is an important field of research.

A 29-year-old African American woman presents with a photosensitive malar rash, fatigue, morning stiffness, and swelling in her hands. She is found to have elevated anti­nuclear antibody at a titer of 1:320. A complete blood cell count demonstrates leukopenia and thrombocytopenia. Results of renal function testing and urinalysis are within normal limits. She has no other medical problems and no history of blood clots or pregnancy loss.

Her arthritis and rash suggest systemic lupus erythematosus. She is counseled to avoid sun exposure, and treatment with hydroxychloroquine is considered.

WHAT IS HYDROXYCHLOROQUINE?

Hydroxychloroquine was developed to treat malaria but was later found to have immunomodulatory properties. It is now approved by the US Food and Drug Administration for treatment of discoid lupus, systemic lupus ery­thematosus, and rheumatoid arthritis. It is also approved to treat malaria; however, of the several malarial parasites, only Plasmodium falciparum can still be cured by hydroxychloroquine, and growing resistance limits the geographic locations where this drug can be used effectively.^1,2

HISTORICAL BACKGROUND

Antimalarial drugs were discovered shortly before World War II. Their production was industrialized during the war because malaria was a leading cause of disease among soldiers, especially those deployed to the South Pacific.³

Atabrine (quinacrine), the first antimalarial widely used, had numerous side effects including yellowing of the skin. Aggressive research efforts to develop an alternative led to field testing of one of its derivative compounds, chloroquine, by the US Army in 1943. Continued chemical modification would create hydroxychloroquine, introduced in 1955.

A serendipitous consequence of the mass use of antimalarials during World War II was the discovery that they could be used to treat inflammatory arthritis and lupus. Eight years after the war ended, Shee⁴ reported that chloroquine had a beneficial effect on lupus and rheumatoid arthritis in US soldiers. Hydroxychloroquine is now the most commonly prescribed antimalarial for treatment of autoimmune disease.

HOW HYDROXYCHLOROQUINE WORKS

The primary mechanism by which hydroxychloroquine modulates systemic lupus erythematosus is by suppressing activation of Toll-like receptors, which exist on the surface of endosomes and play a significant role in the innate immune response and in autoimmune disease. Their activation is necessary for the expression of interferon-regulated genes and production of tumor necrosis factor alpha, which are key in the cell-mediated inflammatory response.

Antimalarial drugs such as hydroxychlor­oquine prevent Toll-like receptor activation by binding directly to nucleic acids in the activation pathway.⁵ In vitro studies show that blocking this pathway blunts the body’s primary cell-mediated inflammatory response; in vivo studies show that use of hydroxychloroquine is strongly correlated with a reduction in interferon alpha levels.⁶ The powerful effect of hydroxychloroquine on the cell-mediated pattern of inflammation found in lupus is consistent with this theory.

It was previously hypothesized that the immune-modulating effects of hydroxychloroquine were associated with a more general dysregulation of cellular lysosomes through inhibition of proteolysis or changes in cellular pH.⁷ This theory has since been displaced by the more specific and elegant mechanism described above.⁵

HOW WELL DOES IT WORK?

Benefit in systemic lupus erythematosus

Hydroxychloroquine has consistently demonstrated significant and multifaceted benefit in patients with systemic lupus erythematosus.

A systematic review of 95 articles⁸ concluded that this drug decreases lupus flares and decreases mortality rates in lupus patients by at least 50%, with a high level of evidence. Beneficial effects that had a moderate level of evidence were an increase in bone mineral density, fewer thrombotic events, and fewer cases of irreversible organ damage.

The preventive effect of hydroxychlor­oquine on thrombosis in lupus patients has been consistently demonstrated and is one of the key reasons the drug is considered a cornerstone of therapy in this disease.⁹ A nested case-control study of patients with lupus and thromboembolism demonstrated an odds ratio of 0.31 and relative risk reduction of 68% for those using antimalarials.¹⁰

Benefit in antiphospholipid antibody syndrome

Hydroxychloroquine prevents thrombosis in other diseases as well. For example, it has been shown to reduce the incidence of thrombotic events in patients with primary antiphospholipid syndrome.

In a retrospective cohort study in 114 patients with this disease, hydroxychloroquine significantly reduced the incidence of arterial thrombotic events over 10 years of follow-up (recurrence incidence 0 in those treated with hydroxychloroquine vs 1.14% in those not treated).¹¹ The study also tracked levels of antiphospholipid antibodies and reported that hydroxychloroquine significantly reduced the levels of antibodies to cardiolipin and beta-2 glycoprotein 1, both implicated in the pathology of thrombosis.¹¹

In vitro studies have also demonstrated that hydroxychloroquine can modulate a dysregulated inflammatory system to reduce thrombosis. For example, it has been shown that hydroxychloroquine can reverse platelet activation by antiphospholipid antibodies, prevent linking of antibody complexes to cell membranes, and promote proper membrane protein expression, thereby reducing the thrombotic qualities of antiphospholipid antibodies and even improving clearance times of antiphospholipid-related thrombi.¹²

Though there is less evidence, hydroxychloroquine has also shown benefit in rheumatoid arthritis, where it can be used by itself in mild disease or as part of combination therapy with active arthritis. Compared with biologic therapy in patients with early aggressive rheumatoid arthritis, triple therapy with methotrexate, sulfasalazine, and hydroxychloroquine was nearly as effective in terms of quality of life, and it cost only one-third as much, saving $20,000 per year of therapy per patient.¹³

Hydroxychloroquine has also been compared directly with chloroquine, its closest relation, in a large study incorporating patients with rheumatoid arthritis and patients with systemic lupus erythematosus. Patients using chloroquine experienced significantly more side effects, though it did prove marginally more effective.¹⁴

No benefit shown in Sjögren syndrome

Unfortunately, despite widespread use, hydroxychloroquine has not demonstrated positive clinical effects when used to treat primary Sjögren syndrome. Most notably, a 2014 randomized controlled trial of hydroxychloroquine vs placebo in 120 Sjögren patients found no significant improvement in primary symptoms of dryness, pain, or fatigue after 6 months of therapy.¹⁵

Metabolic benefits

Unexpectedly, hydroxychloroquine is associated with multiple metabolic benefits including improved lipid profiles and lower blood glucose levels. These findings, in addition to a reduced incidence of thrombosis, were initially reported in the Baltimore Lupus Cohort in 1996.¹⁶ Specifically, longitudinal evaluation of a cohort of lupus patients showed that hydroxychloroquine use was associated with a 7.6% reduction in total cholesterol and a 13.7% reduction in low-density lipoprotein cholesterol (LDL-C) over 3 months of therapy.¹⁷

Similar findings, including a reduction in LDL-C and an increase in high-density lipoprotein cholesterol, were strongly associated with the addition of hydroxychloroquine to methotrexate or to methotrexate and etanercept in a large cohort of rheumatoid arthritis patients followed over 2 years of therapy.¹⁸

In nondiabetic women with systemic lupus erythematosus or rheumatoid arthritis, average blood glucose was significantly lower in those taking hydroxychloroquine than in nonusers. The incidence of insulin resistance was also lower, but the difference was not statistically significant.¹⁹

Some have suggested that hydroxychloroquine may prevent diabetes mellitus. In a retrospective case series, compared with rheumatoid arthritis patients not taking the drug, patients treated with hydroxychloroquine for more than 4 years had a 25% lower risk of developing diabetes mellitus.²⁰

In view of these metabolic benefits, especially regarding lipid regulation, and the above described antithrombotic properties of hydroxychloroquine, some researchers have recently hypothesized that hydroxychloroquine may be of benefit in patients with coronary artery disease.²¹ They suggested that the inflammatory contribution to the mechanism of coronary artery disease could be lessened by hydroxychloroquine even in patients without lupus erythematosus or rheumatoid arthritis.

PHARMACOLOGIC PROPERTIES

Understanding the pharmacologic properties of hydroxychloroquine is key to using it appropriately in clinical practice.

The half-life of elimination of hydroxychloroquine is 40 to 50 days, with half of the drug excreted renally in a concentration-dependent fashion.^22,23 The drug reaches 95% of its steady-state concentration by about 6 months of therapy. Shorter durations of therapy do not provide adequate time for the drug to achieve steady-state concentration and may not allow patients and providers time to see its full clinical results. Therefore, its manufacturers recommend a 6-month trial of therapy to adequately determine if the drug improves symptoms.¹

The oral bioavailability of hydroxychloroquine is about 75%, but pharmacokinetics vary among individuals.^22,23 It has been suggested that this variability affects the efficacy of hydroxychloroquine. In a study of 300 patients with cutaneous lupus erythematosus, those whose treatment failed had significantly lower blood concentrations of hydroxychloroquine, while those who achieved complete remission had significantly higher concentrations.²⁴

Another study found that titrating doses to target therapeutic blood concentrations can reduce disease activity in cutaneous lupus erythematosus.²⁵ Measuring the blood concentration of hydroxychloroquine is not common in clinical practice but may have a role in select patients in whom initial therapy using a standard dosing regimen does not produce the desired results.

HOW SAFE IS HYDROXYCHLOROQUINE?

Hydroxychloroquine has numerous adverse effects, necessitating vigilance on the part of the prescriber. Most commonly reported are retinopathy, hyperpigmentation, myopathy, and skin reactions.¹

Retinopathy

Retinopathy’s irreversibility—the threat of permanent vision loss—and its substantial prevalence in patients with a large drug exposure history, have marked retinopathy as the most concerning potential toxicity. The risk of ocular toxicity increases with the cumulative hydroxychloroquine dose. The prevalence of retinopathy in those using the drug less than 10 years is less than 2%; in contrast, the prevalence in patients with more than 20 years of exposure is reported to be as high as 20%.²⁶

The American Academy of Ophthalmology has long stated that retinopathy is a significant risk of hydroxychloroquine therapy and that patients taking hydroxychloroquine should therefore undergo routine retinal and visual field screening by an ophthalmologist.

Currently, initial screening followed by yearly screening beginning 5 years thereafter is recommended for patients at low risk of toxicity (Table 1).²⁷ Patients determined by an ophthalmologist to be at higher risk of retinopathy should be screened yearly. As identified by the American Academy of Ophthalmology, major risk factors for retinopathy include duration of use, concomitant tamoxifen exposure, significant renal disease, and preexisting retinal and macular disease.^26,28

Recommendations for hydroxychloroquine dosing and screening were recently revised, for 2 reasons. Initially, its manufacturers recommended that hydroxychloroquine dosage be no higher than 6.5 mg/kg of ideal body weight to prevent retinopathy.^1,29,30 However, it has recently been demonstrated that real body weight is a better predictor of risk of retinopathy than ideal body weight when dosing hydroxychloroquine, perhaps because of the increasing variance of real body weight in our patient population.²⁶

Further, an atypical pattern of retinopathy called pericentral retinopathy is more common in Asians. A study of about 200 patients with a history of hydroxychloroquine retinopathy, including 36 Asian patients, found that the pericentral pattern occurred in half the Asian patients but only 2% of the white patients.³¹ The mechanism for this finding is unclear, but because pericentral retinopathy spares the macula, it can be missed using standard screening methods. Therefore, the American Academy of Ophthalmology now recommends that the dose limit be reduced from 6.5 mg/kg of ideal body weight to no more than 5.0 mg/kg of real body weight (Table 2).²⁸

It is also recommended that screening methods such as automated visual fields and optical coherence tomography extend their fields beyond the macula in Asian patients to ensure that pericentral retinopathy is not missed.²⁸

Optical coherence tomography is a particularly useful tool in the ocular evaluation of patients taking hydroxychloroquine. It can detect subtle changes such as thinning of the foveal photoreceptor outer segment, thickening of the retinal pigment epithelium, and loss of the macular ganglion cell–inner plexiform layer before there are visible signs of retinopathy and before symptoms arise.³²

Currently, these guidelines are underutilized in clinical practice. Physician adherence to ophthalmologic guidelines is reported at about 50%.³³ This statistic is jarring, given the potential for permanent loss of vision in those with hydroxychloroquine-mediated retinopathy, and demonstrates the importance of reinforcing proper understanding of the use of hydroxychloroquine in clinical practice.

Cutaneous hyperpigmentation can occur with hydroxychloroquine use (Figure 1). The hyperpigmentation appears to be due to local bruising following deposition of iron in the soft tissue.

While the pigmentation may persist permanently and cause an undesirable cosmetic effect, it has not been associated with other adverse outcomes.

Myopathy is a rare adverse effect. In one case series, 3 of 214 patients treated with hydroxychloroquine developed hydroxychloroquine-induced myopathy.³⁵ Over the duration of their therapy, this was equivalent to an incidence of 1 case of myopathy in 100 patient-years of therapy. Myopathy improves with discontinuation of therapy, though it can persist for weeks, likely because of hydroxychloroquine’s prolonged elimination half-life.

Cardiomyopathy, specifically neurocardio­myopathy, is also an extremely rare adverse effect of hydroxychloroquine use. The mechanism is believed to be associated with the effect of hydroxychloroquine on lysosomal action, leading to an acquired lysosomal storage disorder with the typical cardiac hypertrophy and conduction abnormalities associated with this family of diseases.³⁶

Acute generalized exanthematous pustulosis is another rare complication of hydroxychloroquine therapy. The appearance of the reaction is similar to that of pustular psoriasis, with pustules overlying flaking and scaling skin. It usually resolves within 2 weeks after cessation of hydroxychloroquine therapy. In a select few cases, the reaction persists or waxes and wanes over a period of weeks to months, and longer durations of recovery are thought to be due to hydroxychloroquine’s long half-life, as in hydroxychloroquine-induced myopathy.³⁷

In view of this rare reaction, manufacturers of hydroxychloroquine recommend caution when using the drug in patients with psoriasis.¹

Hematologic abnormalities. In very rare cases, hydroxychloroquine is associated with hematologic abnormalities including agranulocytosis, anemia, aplastic anemia, leukopenia, and thrombocytopenia.¹

While no specific guidelines exist, caution is warranted when using hydroxychloroquine in patients with porphyria. Additionally, hydroxychloroquine and other antimalarials including primaquine have been associated with hemolysis in patients with glucose-6-phosphate dehydrogenase (G6PD) deficiency. The risk of hemolysis is generally considered low except at high hydroxychloroquine doses in patients with severe G6PD deficiency.³⁸

For the above reasons, manufacturers recommended baseline and routine blood counts, and some providers screen patients for G6PD deficiency when prescribing hydroxychloroquine (Table 3).

PREGNANCY

Hydroxychloroquine is in pregnancy category C. Information is limited, and in view of the risks, the manufacturer says that it should be avoided in pregnancy.¹ Nevertheless, it is generally considered safe during pregnancy, and its benefits may make it acceptable to continue in a patient who becomes pregnant, in spite of the possible risks.

We favor continuing hydroxychloroquine. This drug has been associated with improved maternal and fetal outcomes in lupus patients. Its use during pregnancy has not been associated with congenital malformations. The adverse visual effects of long-term hydroxychloroquine use, namely retinopathy, have never been reported in children as a consequence of exposure in utero.

In addition, hydroxychloroquine is transmitted only in minute quantities in breast milk.³⁹ In pregnant women with systemic lupus erythematosus, hydroxychloroquine was associated with a lower risk of adverse outcomes, including preterm delivery and intrauterine growth restriction.⁴⁰ However, hydroxychloroquine is far more toxic when ingested directly by infants than in adults.¹

Maternal outcomes are also improved with the use of hydroxychloroquine. Stopping hydroxychloroquine during pregnancy in women with systemic lupus erythematosus is associated with significantly higher disease activity—fully twice as high as in those who continue hydroxychloroquine.⁴¹ These study results were corroborated in a small randomized trial in which pregnant women with lupus on placebo had significantly higher lupus disease activity scores than those pregnant women who were given hydroxychloroquine.⁴² The women taking hydroxychloroquine experienced no severe lupus flares for the duration of their pregnancies.

These findings suggest not only that hydroxychloroquine is safe in pregnancy, but also that it should be continued in lupus patients during pregnancy to prevent disease flares and adverse fetal outcomes.

Benefit in preclinical lupus?

Hydroxychloroquine has a consistently profound effect on outcomes in systemic lupus erythematosus. These findings, in addition to the more widespread use of antibody screening, have led to suggestions that hydroxychloroquine could be of benefit even before systemic lupus erythematosus is diagnosed.

A study in US military personnel found that patients taking hydroxychloroquine experienced a significantly longer lag time between first reported clinical symptoms of lupus and official diagnosis compared with  matched controls who also went on to develop the disease, averaging 1.08 vs 0.29 years to disease classification.⁴³ Those who used hydroxychloroquine also had lower rates of autoantibody accumulation. Therefore, hydroxychloroquine could be of benefit in carefully selected candidates at high risk of developing systemic lupus erythematosus.

The beneficial effects of hydroxychloroquine on patients with lupus and rheumatoid arthritis, in terms of primary measures of disease activity and secondary outcomes, were discovered fortuitously and were not the original intended targets of the drug. Because of its versatility, there are numerous other disease states in which hydroxychloroquine has shown a degree of benefit or has shown a potential for benefit.

Antiviral activity?

It has been suggested that antimalarial drugs could serve as adjunctive therapies against filoviruses such as Marburg and Ebola. There is a small body of in vitro and in vivo evidence that hydroxychloroquine could temper severe systemic inflammatory responses to filoviruses both through dysregulation of lysosomes and lysosomal pH (filoviruses have a pH-dependent mechanism of action) and through decreased production of tumor necrosis factor alpha and interferons. Heavy burdens of interferons and tumor necrosis factor alpha are associated with increased mortality rates in those infected with filoviruses.⁴⁴

Antineoplastic activity?

Hydroxychloroquine has undergone in vitro testing as an adjunct to cancer therapies. There are several mechanisms by which it is theorized that hydroxychloroquine could target malignant cells, including inhibition of multidrug resistance pumps or autophagy, improvement of chemotherapy cell penetration, potentiation of presentation of major histocompatibility complexes, or even intercalation directly into DNA.^45,46 However, it can also impair natural anticancer immunity and may allow cancer cells better nutrient supply through vascular effects.

In vitro studies have shown tumoricidal effects in lymphoma and melanoma, and inhibition of growth in lung, colon, breast, cervix, larynx, liver, and prostate cancers. In vivo studies have shown that hydroxychloroquine in high doses can prolong survival in glioblastoma.⁴⁵

Unfortunately, all of these theorized or observed effects are dose-dependent and likely require doses that exceed currently recommended maximums.

Negative chronotropic effect?

Hydroxychloroquine has been found to decrease the resting heart rate in a cumulative dose-dependent fashion.⁴⁷ Further, hydroxychloroquine has been known to increase digoxin levels, and the medications should not be used in combination.¹

Whether the decrease in resting heart rate is associated with harm or benefit and whether the effect is significant enough to be considered when implementing therapy remain unanswered and deserve further investigation, as does the primary use of hydroxychloroquine for beneficial lipid and glucose reduction in patients who are otherwise healthy.

CASE CONCLUSION

The patient described at the beginning of this article was provided with information on the risks and benefits of hydroxychloroquine for treatment of her arthritis and rash suggestive of mild systemic lupus, and she opted to begin therapy. Her baseline eye screening was within normal limits. Based on her weight of 62 kg, she was started on 300 mg of hydroxychloroquine daily.

She had no significant adverse effects from the medication and reported slow improvement in her rash and joint complaints over the next 2 months. She remained on hydroxychloroquine over the next year without adverse effects or new evidence of autoimmune disease.

A 29-year-old African American woman presents with a photosensitive malar rash, fatigue, morning stiffness, and swelling in her hands. She is found to have elevated anti­nuclear antibody at a titer of 1:320. A complete blood cell count demonstrates leukopenia and thrombocytopenia. Results of renal function testing and urinalysis are within normal limits. She has no other medical problems and no history of blood clots or pregnancy loss.

Her arthritis and rash suggest systemic lupus erythematosus. She is counseled to avoid sun exposure, and treatment with hydroxychloroquine is considered.

WHAT IS HYDROXYCHLOROQUINE?

Hydroxychloroquine was developed to treat malaria but was later found to have immunomodulatory properties. It is now approved by the US Food and Drug Administration for treatment of discoid lupus, systemic lupus ery­thematosus, and rheumatoid arthritis. It is also approved to treat malaria; however, of the several malarial parasites, only Plasmodium falciparum can still be cured by hydroxychloroquine, and growing resistance limits the geographic locations where this drug can be used effectively.^1,2

HISTORICAL BACKGROUND

Antimalarial drugs were discovered shortly before World War II. Their production was industrialized during the war because malaria was a leading cause of disease among soldiers, especially those deployed to the South Pacific.³

Atabrine (quinacrine), the first antimalarial widely used, had numerous side effects including yellowing of the skin. Aggressive research efforts to develop an alternative led to field testing of one of its derivative compounds, chloroquine, by the US Army in 1943. Continued chemical modification would create hydroxychloroquine, introduced in 1955.

A serendipitous consequence of the mass use of antimalarials during World War II was the discovery that they could be used to treat inflammatory arthritis and lupus. Eight years after the war ended, Shee⁴ reported that chloroquine had a beneficial effect on lupus and rheumatoid arthritis in US soldiers. Hydroxychloroquine is now the most commonly prescribed antimalarial for treatment of autoimmune disease.

HOW HYDROXYCHLOROQUINE WORKS

The primary mechanism by which hydroxychloroquine modulates systemic lupus erythematosus is by suppressing activation of Toll-like receptors, which exist on the surface of endosomes and play a significant role in the innate immune response and in autoimmune disease. Their activation is necessary for the expression of interferon-regulated genes and production of tumor necrosis factor alpha, which are key in the cell-mediated inflammatory response.

Antimalarial drugs such as hydroxychlor­oquine prevent Toll-like receptor activation by binding directly to nucleic acids in the activation pathway.⁵ In vitro studies show that blocking this pathway blunts the body’s primary cell-mediated inflammatory response; in vivo studies show that use of hydroxychloroquine is strongly correlated with a reduction in interferon alpha levels.⁶ The powerful effect of hydroxychloroquine on the cell-mediated pattern of inflammation found in lupus is consistent with this theory.

It was previously hypothesized that the immune-modulating effects of hydroxychloroquine were associated with a more general dysregulation of cellular lysosomes through inhibition of proteolysis or changes in cellular pH.⁷ This theory has since been displaced by the more specific and elegant mechanism described above.⁵

HOW WELL DOES IT WORK?

Benefit in systemic lupus erythematosus

Hydroxychloroquine has consistently demonstrated significant and multifaceted benefit in patients with systemic lupus erythematosus.

A systematic review of 95 articles⁸ concluded that this drug decreases lupus flares and decreases mortality rates in lupus patients by at least 50%, with a high level of evidence. Beneficial effects that had a moderate level of evidence were an increase in bone mineral density, fewer thrombotic events, and fewer cases of irreversible organ damage.

The preventive effect of hydroxychlor­oquine on thrombosis in lupus patients has been consistently demonstrated and is one of the key reasons the drug is considered a cornerstone of therapy in this disease.⁹ A nested case-control study of patients with lupus and thromboembolism demonstrated an odds ratio of 0.31 and relative risk reduction of 68% for those using antimalarials.¹⁰

Benefit in antiphospholipid antibody syndrome

Hydroxychloroquine prevents thrombosis in other diseases as well. For example, it has been shown to reduce the incidence of thrombotic events in patients with primary antiphospholipid syndrome.

In a retrospective cohort study in 114 patients with this disease, hydroxychloroquine significantly reduced the incidence of arterial thrombotic events over 10 years of follow-up (recurrence incidence 0 in those treated with hydroxychloroquine vs 1.14% in those not treated).¹¹ The study also tracked levels of antiphospholipid antibodies and reported that hydroxychloroquine significantly reduced the levels of antibodies to cardiolipin and beta-2 glycoprotein 1, both implicated in the pathology of thrombosis.¹¹

In vitro studies have also demonstrated that hydroxychloroquine can modulate a dysregulated inflammatory system to reduce thrombosis. For example, it has been shown that hydroxychloroquine can reverse platelet activation by antiphospholipid antibodies, prevent linking of antibody complexes to cell membranes, and promote proper membrane protein expression, thereby reducing the thrombotic qualities of antiphospholipid antibodies and even improving clearance times of antiphospholipid-related thrombi.¹²

Though there is less evidence, hydroxychloroquine has also shown benefit in rheumatoid arthritis, where it can be used by itself in mild disease or as part of combination therapy with active arthritis. Compared with biologic therapy in patients with early aggressive rheumatoid arthritis, triple therapy with methotrexate, sulfasalazine, and hydroxychloroquine was nearly as effective in terms of quality of life, and it cost only one-third as much, saving $20,000 per year of therapy per patient.¹³

Hydroxychloroquine has also been compared directly with chloroquine, its closest relation, in a large study incorporating patients with rheumatoid arthritis and patients with systemic lupus erythematosus. Patients using chloroquine experienced significantly more side effects, though it did prove marginally more effective.¹⁴

No benefit shown in Sjögren syndrome

Unfortunately, despite widespread use, hydroxychloroquine has not demonstrated positive clinical effects when used to treat primary Sjögren syndrome. Most notably, a 2014 randomized controlled trial of hydroxychloroquine vs placebo in 120 Sjögren patients found no significant improvement in primary symptoms of dryness, pain, or fatigue after 6 months of therapy.¹⁵

Metabolic benefits

Unexpectedly, hydroxychloroquine is associated with multiple metabolic benefits including improved lipid profiles and lower blood glucose levels. These findings, in addition to a reduced incidence of thrombosis, were initially reported in the Baltimore Lupus Cohort in 1996.¹⁶ Specifically, longitudinal evaluation of a cohort of lupus patients showed that hydroxychloroquine use was associated with a 7.6% reduction in total cholesterol and a 13.7% reduction in low-density lipoprotein cholesterol (LDL-C) over 3 months of therapy.¹⁷

Similar findings, including a reduction in LDL-C and an increase in high-density lipoprotein cholesterol, were strongly associated with the addition of hydroxychloroquine to methotrexate or to methotrexate and etanercept in a large cohort of rheumatoid arthritis patients followed over 2 years of therapy.¹⁸

In nondiabetic women with systemic lupus erythematosus or rheumatoid arthritis, average blood glucose was significantly lower in those taking hydroxychloroquine than in nonusers. The incidence of insulin resistance was also lower, but the difference was not statistically significant.¹⁹

Some have suggested that hydroxychloroquine may prevent diabetes mellitus. In a retrospective case series, compared with rheumatoid arthritis patients not taking the drug, patients treated with hydroxychloroquine for more than 4 years had a 25% lower risk of developing diabetes mellitus.²⁰

In view of these metabolic benefits, especially regarding lipid regulation, and the above described antithrombotic properties of hydroxychloroquine, some researchers have recently hypothesized that hydroxychloroquine may be of benefit in patients with coronary artery disease.²¹ They suggested that the inflammatory contribution to the mechanism of coronary artery disease could be lessened by hydroxychloroquine even in patients without lupus erythematosus or rheumatoid arthritis.

PHARMACOLOGIC PROPERTIES

Understanding the pharmacologic properties of hydroxychloroquine is key to using it appropriately in clinical practice.

The half-life of elimination of hydroxychloroquine is 40 to 50 days, with half of the drug excreted renally in a concentration-dependent fashion.^22,23 The drug reaches 95% of its steady-state concentration by about 6 months of therapy. Shorter durations of therapy do not provide adequate time for the drug to achieve steady-state concentration and may not allow patients and providers time to see its full clinical results. Therefore, its manufacturers recommend a 6-month trial of therapy to adequately determine if the drug improves symptoms.¹

The oral bioavailability of hydroxychloroquine is about 75%, but pharmacokinetics vary among individuals.^22,23 It has been suggested that this variability affects the efficacy of hydroxychloroquine. In a study of 300 patients with cutaneous lupus erythematosus, those whose treatment failed had significantly lower blood concentrations of hydroxychloroquine, while those who achieved complete remission had significantly higher concentrations.²⁴

Another study found that titrating doses to target therapeutic blood concentrations can reduce disease activity in cutaneous lupus erythematosus.²⁵ Measuring the blood concentration of hydroxychloroquine is not common in clinical practice but may have a role in select patients in whom initial therapy using a standard dosing regimen does not produce the desired results.

HOW SAFE IS HYDROXYCHLOROQUINE?

Hydroxychloroquine has numerous adverse effects, necessitating vigilance on the part of the prescriber. Most commonly reported are retinopathy, hyperpigmentation, myopathy, and skin reactions.¹

Retinopathy

Retinopathy’s irreversibility—the threat of permanent vision loss—and its substantial prevalence in patients with a large drug exposure history, have marked retinopathy as the most concerning potential toxicity. The risk of ocular toxicity increases with the cumulative hydroxychloroquine dose. The prevalence of retinopathy in those using the drug less than 10 years is less than 2%; in contrast, the prevalence in patients with more than 20 years of exposure is reported to be as high as 20%.²⁶

The American Academy of Ophthalmology has long stated that retinopathy is a significant risk of hydroxychloroquine therapy and that patients taking hydroxychloroquine should therefore undergo routine retinal and visual field screening by an ophthalmologist.

Currently, initial screening followed by yearly screening beginning 5 years thereafter is recommended for patients at low risk of toxicity (Table 1).²⁷ Patients determined by an ophthalmologist to be at higher risk of retinopathy should be screened yearly. As identified by the American Academy of Ophthalmology, major risk factors for retinopathy include duration of use, concomitant tamoxifen exposure, significant renal disease, and preexisting retinal and macular disease.^26,28

Recommendations for hydroxychloroquine dosing and screening were recently revised, for 2 reasons. Initially, its manufacturers recommended that hydroxychloroquine dosage be no higher than 6.5 mg/kg of ideal body weight to prevent retinopathy.^1,29,30 However, it has recently been demonstrated that real body weight is a better predictor of risk of retinopathy than ideal body weight when dosing hydroxychloroquine, perhaps because of the increasing variance of real body weight in our patient population.²⁶

Further, an atypical pattern of retinopathy called pericentral retinopathy is more common in Asians. A study of about 200 patients with a history of hydroxychloroquine retinopathy, including 36 Asian patients, found that the pericentral pattern occurred in half the Asian patients but only 2% of the white patients.³¹ The mechanism for this finding is unclear, but because pericentral retinopathy spares the macula, it can be missed using standard screening methods. Therefore, the American Academy of Ophthalmology now recommends that the dose limit be reduced from 6.5 mg/kg of ideal body weight to no more than 5.0 mg/kg of real body weight (Table 2).²⁸

It is also recommended that screening methods such as automated visual fields and optical coherence tomography extend their fields beyond the macula in Asian patients to ensure that pericentral retinopathy is not missed.²⁸

Optical coherence tomography is a particularly useful tool in the ocular evaluation of patients taking hydroxychloroquine. It can detect subtle changes such as thinning of the foveal photoreceptor outer segment, thickening of the retinal pigment epithelium, and loss of the macular ganglion cell–inner plexiform layer before there are visible signs of retinopathy and before symptoms arise.³²

Currently, these guidelines are underutilized in clinical practice. Physician adherence to ophthalmologic guidelines is reported at about 50%.³³ This statistic is jarring, given the potential for permanent loss of vision in those with hydroxychloroquine-mediated retinopathy, and demonstrates the importance of reinforcing proper understanding of the use of hydroxychloroquine in clinical practice.

Cutaneous hyperpigmentation can occur with hydroxychloroquine use (Figure 1). The hyperpigmentation appears to be due to local bruising following deposition of iron in the soft tissue.

While the pigmentation may persist permanently and cause an undesirable cosmetic effect, it has not been associated with other adverse outcomes.

Myopathy is a rare adverse effect. In one case series, 3 of 214 patients treated with hydroxychloroquine developed hydroxychloroquine-induced myopathy.³⁵ Over the duration of their therapy, this was equivalent to an incidence of 1 case of myopathy in 100 patient-years of therapy. Myopathy improves with discontinuation of therapy, though it can persist for weeks, likely because of hydroxychloroquine’s prolonged elimination half-life.

Cardiomyopathy, specifically neurocardio­myopathy, is also an extremely rare adverse effect of hydroxychloroquine use. The mechanism is believed to be associated with the effect of hydroxychloroquine on lysosomal action, leading to an acquired lysosomal storage disorder with the typical cardiac hypertrophy and conduction abnormalities associated with this family of diseases.³⁶

Acute generalized exanthematous pustulosis is another rare complication of hydroxychloroquine therapy. The appearance of the reaction is similar to that of pustular psoriasis, with pustules overlying flaking and scaling skin. It usually resolves within 2 weeks after cessation of hydroxychloroquine therapy. In a select few cases, the reaction persists or waxes and wanes over a period of weeks to months, and longer durations of recovery are thought to be due to hydroxychloroquine’s long half-life, as in hydroxychloroquine-induced myopathy.³⁷

In view of this rare reaction, manufacturers of hydroxychloroquine recommend caution when using the drug in patients with psoriasis.¹

Hematologic abnormalities. In very rare cases, hydroxychloroquine is associated with hematologic abnormalities including agranulocytosis, anemia, aplastic anemia, leukopenia, and thrombocytopenia.¹

While no specific guidelines exist, caution is warranted when using hydroxychloroquine in patients with porphyria. Additionally, hydroxychloroquine and other antimalarials including primaquine have been associated with hemolysis in patients with glucose-6-phosphate dehydrogenase (G6PD) deficiency. The risk of hemolysis is generally considered low except at high hydroxychloroquine doses in patients with severe G6PD deficiency.³⁸

For the above reasons, manufacturers recommended baseline and routine blood counts, and some providers screen patients for G6PD deficiency when prescribing hydroxychloroquine (Table 3).

PREGNANCY

Hydroxychloroquine is in pregnancy category C. Information is limited, and in view of the risks, the manufacturer says that it should be avoided in pregnancy.¹ Nevertheless, it is generally considered safe during pregnancy, and its benefits may make it acceptable to continue in a patient who becomes pregnant, in spite of the possible risks.

We favor continuing hydroxychloroquine. This drug has been associated with improved maternal and fetal outcomes in lupus patients. Its use during pregnancy has not been associated with congenital malformations. The adverse visual effects of long-term hydroxychloroquine use, namely retinopathy, have never been reported in children as a consequence of exposure in utero.

In addition, hydroxychloroquine is transmitted only in minute quantities in breast milk.³⁹ In pregnant women with systemic lupus erythematosus, hydroxychloroquine was associated with a lower risk of adverse outcomes, including preterm delivery and intrauterine growth restriction.⁴⁰ However, hydroxychloroquine is far more toxic when ingested directly by infants than in adults.¹

Maternal outcomes are also improved with the use of hydroxychloroquine. Stopping hydroxychloroquine during pregnancy in women with systemic lupus erythematosus is associated with significantly higher disease activity—fully twice as high as in those who continue hydroxychloroquine.⁴¹ These study results were corroborated in a small randomized trial in which pregnant women with lupus on placebo had significantly higher lupus disease activity scores than those pregnant women who were given hydroxychloroquine.⁴² The women taking hydroxychloroquine experienced no severe lupus flares for the duration of their pregnancies.

These findings suggest not only that hydroxychloroquine is safe in pregnancy, but also that it should be continued in lupus patients during pregnancy to prevent disease flares and adverse fetal outcomes.

Benefit in preclinical lupus?

Hydroxychloroquine has a consistently profound effect on outcomes in systemic lupus erythematosus. These findings, in addition to the more widespread use of antibody screening, have led to suggestions that hydroxychloroquine could be of benefit even before systemic lupus erythematosus is diagnosed.

A study in US military personnel found that patients taking hydroxychloroquine experienced a significantly longer lag time between first reported clinical symptoms of lupus and official diagnosis compared with  matched controls who also went on to develop the disease, averaging 1.08 vs 0.29 years to disease classification.⁴³ Those who used hydroxychloroquine also had lower rates of autoantibody accumulation. Therefore, hydroxychloroquine could be of benefit in carefully selected candidates at high risk of developing systemic lupus erythematosus.

The beneficial effects of hydroxychloroquine on patients with lupus and rheumatoid arthritis, in terms of primary measures of disease activity and secondary outcomes, were discovered fortuitously and were not the original intended targets of the drug. Because of its versatility, there are numerous other disease states in which hydroxychloroquine has shown a degree of benefit or has shown a potential for benefit.

Antiviral activity?

It has been suggested that antimalarial drugs could serve as adjunctive therapies against filoviruses such as Marburg and Ebola. There is a small body of in vitro and in vivo evidence that hydroxychloroquine could temper severe systemic inflammatory responses to filoviruses both through dysregulation of lysosomes and lysosomal pH (filoviruses have a pH-dependent mechanism of action) and through decreased production of tumor necrosis factor alpha and interferons. Heavy burdens of interferons and tumor necrosis factor alpha are associated with increased mortality rates in those infected with filoviruses.⁴⁴

Antineoplastic activity?

Hydroxychloroquine has undergone in vitro testing as an adjunct to cancer therapies. There are several mechanisms by which it is theorized that hydroxychloroquine could target malignant cells, including inhibition of multidrug resistance pumps or autophagy, improvement of chemotherapy cell penetration, potentiation of presentation of major histocompatibility complexes, or even intercalation directly into DNA.^45,46 However, it can also impair natural anticancer immunity and may allow cancer cells better nutrient supply through vascular effects.

In vitro studies have shown tumoricidal effects in lymphoma and melanoma, and inhibition of growth in lung, colon, breast, cervix, larynx, liver, and prostate cancers. In vivo studies have shown that hydroxychloroquine in high doses can prolong survival in glioblastoma.⁴⁵

Unfortunately, all of these theorized or observed effects are dose-dependent and likely require doses that exceed currently recommended maximums.

Negative chronotropic effect?

Hydroxychloroquine has been found to decrease the resting heart rate in a cumulative dose-dependent fashion.⁴⁷ Further, hydroxychloroquine has been known to increase digoxin levels, and the medications should not be used in combination.¹

Whether the decrease in resting heart rate is associated with harm or benefit and whether the effect is significant enough to be considered when implementing therapy remain unanswered and deserve further investigation, as does the primary use of hydroxychloroquine for beneficial lipid and glucose reduction in patients who are otherwise healthy.

CASE CONCLUSION

The patient described at the beginning of this article was provided with information on the risks and benefits of hydroxychloroquine for treatment of her arthritis and rash suggestive of mild systemic lupus, and she opted to begin therapy. Her baseline eye screening was within normal limits. Based on her weight of 62 kg, she was started on 300 mg of hydroxychloroquine daily.

She had no significant adverse effects from the medication and reported slow improvement in her rash and joint complaints over the next 2 months. She remained on hydroxychloroquine over the next year without adverse effects or new evidence of autoimmune disease.

A 29-year-old African American woman presents with a photosensitive malar rash, fatigue, morning stiffness, and swelling in her hands. She is found to have elevated anti­nuclear antibody at a titer of 1:320. A complete blood cell count demonstrates leukopenia and thrombocytopenia. Results of renal function testing and urinalysis are within normal limits. She has no other medical problems and no history of blood clots or pregnancy loss.

Her arthritis and rash suggest systemic lupus erythematosus. She is counseled to avoid sun exposure, and treatment with hydroxychloroquine is considered.

WHAT IS HYDROXYCHLOROQUINE?

Hydroxychloroquine was developed to treat malaria but was later found to have immunomodulatory properties. It is now approved by the US Food and Drug Administration for treatment of discoid lupus, systemic lupus ery­thematosus, and rheumatoid arthritis. It is also approved to treat malaria; however, of the several malarial parasites, only Plasmodium falciparum can still be cured by hydroxychloroquine, and growing resistance limits the geographic locations where this drug can be used effectively.^1,2

HISTORICAL BACKGROUND

Antimalarial drugs were discovered shortly before World War II. Their production was industrialized during the war because malaria was a leading cause of disease among soldiers, especially those deployed to the South Pacific.³

Atabrine (quinacrine), the first antimalarial widely used, had numerous side effects including yellowing of the skin. Aggressive research efforts to develop an alternative led to field testing of one of its derivative compounds, chloroquine, by the US Army in 1943. Continued chemical modification would create hydroxychloroquine, introduced in 1955.

A serendipitous consequence of the mass use of antimalarials during World War II was the discovery that they could be used to treat inflammatory arthritis and lupus. Eight years after the war ended, Shee⁴ reported that chloroquine had a beneficial effect on lupus and rheumatoid arthritis in US soldiers. Hydroxychloroquine is now the most commonly prescribed antimalarial for treatment of autoimmune disease.

HOW HYDROXYCHLOROQUINE WORKS

The primary mechanism by which hydroxychloroquine modulates systemic lupus erythematosus is by suppressing activation of Toll-like receptors, which exist on the surface of endosomes and play a significant role in the innate immune response and in autoimmune disease. Their activation is necessary for the expression of interferon-regulated genes and production of tumor necrosis factor alpha, which are key in the cell-mediated inflammatory response.

Antimalarial drugs such as hydroxychlor­oquine prevent Toll-like receptor activation by binding directly to nucleic acids in the activation pathway.⁵ In vitro studies show that blocking this pathway blunts the body’s primary cell-mediated inflammatory response; in vivo studies show that use of hydroxychloroquine is strongly correlated with a reduction in interferon alpha levels.⁶ The powerful effect of hydroxychloroquine on the cell-mediated pattern of inflammation found in lupus is consistent with this theory.

It was previously hypothesized that the immune-modulating effects of hydroxychloroquine were associated with a more general dysregulation of cellular lysosomes through inhibition of proteolysis or changes in cellular pH.⁷ This theory has since been displaced by the more specific and elegant mechanism described above.⁵

HOW WELL DOES IT WORK?

Benefit in systemic lupus erythematosus

Hydroxychloroquine has consistently demonstrated significant and multifaceted benefit in patients with systemic lupus erythematosus.

A systematic review of 95 articles⁸ concluded that this drug decreases lupus flares and decreases mortality rates in lupus patients by at least 50%, with a high level of evidence. Beneficial effects that had a moderate level of evidence were an increase in bone mineral density, fewer thrombotic events, and fewer cases of irreversible organ damage.

The preventive effect of hydroxychlor­oquine on thrombosis in lupus patients has been consistently demonstrated and is one of the key reasons the drug is considered a cornerstone of therapy in this disease.⁹ A nested case-control study of patients with lupus and thromboembolism demonstrated an odds ratio of 0.31 and relative risk reduction of 68% for those using antimalarials.¹⁰

Benefit in antiphospholipid antibody syndrome

Hydroxychloroquine prevents thrombosis in other diseases as well. For example, it has been shown to reduce the incidence of thrombotic events in patients with primary antiphospholipid syndrome.

In a retrospective cohort study in 114 patients with this disease, hydroxychloroquine significantly reduced the incidence of arterial thrombotic events over 10 years of follow-up (recurrence incidence 0 in those treated with hydroxychloroquine vs 1.14% in those not treated).¹¹ The study also tracked levels of antiphospholipid antibodies and reported that hydroxychloroquine significantly reduced the levels of antibodies to cardiolipin and beta-2 glycoprotein 1, both implicated in the pathology of thrombosis.¹¹

In vitro studies have also demonstrated that hydroxychloroquine can modulate a dysregulated inflammatory system to reduce thrombosis. For example, it has been shown that hydroxychloroquine can reverse platelet activation by antiphospholipid antibodies, prevent linking of antibody complexes to cell membranes, and promote proper membrane protein expression, thereby reducing the thrombotic qualities of antiphospholipid antibodies and even improving clearance times of antiphospholipid-related thrombi.¹²

Though there is less evidence, hydroxychloroquine has also shown benefit in rheumatoid arthritis, where it can be used by itself in mild disease or as part of combination therapy with active arthritis. Compared with biologic therapy in patients with early aggressive rheumatoid arthritis, triple therapy with methotrexate, sulfasalazine, and hydroxychloroquine was nearly as effective in terms of quality of life, and it cost only one-third as much, saving $20,000 per year of therapy per patient.¹³

Hydroxychloroquine has also been compared directly with chloroquine, its closest relation, in a large study incorporating patients with rheumatoid arthritis and patients with systemic lupus erythematosus. Patients using chloroquine experienced significantly more side effects, though it did prove marginally more effective.¹⁴

No benefit shown in Sjögren syndrome

Unfortunately, despite widespread use, hydroxychloroquine has not demonstrated positive clinical effects when used to treat primary Sjögren syndrome. Most notably, a 2014 randomized controlled trial of hydroxychloroquine vs placebo in 120 Sjögren patients found no significant improvement in primary symptoms of dryness, pain, or fatigue after 6 months of therapy.¹⁵

Metabolic benefits

Unexpectedly, hydroxychloroquine is associated with multiple metabolic benefits including improved lipid profiles and lower blood glucose levels. These findings, in addition to a reduced incidence of thrombosis, were initially reported in the Baltimore Lupus Cohort in 1996.¹⁶ Specifically, longitudinal evaluation of a cohort of lupus patients showed that hydroxychloroquine use was associated with a 7.6% reduction in total cholesterol and a 13.7% reduction in low-density lipoprotein cholesterol (LDL-C) over 3 months of therapy.¹⁷

Similar findings, including a reduction in LDL-C and an increase in high-density lipoprotein cholesterol, were strongly associated with the addition of hydroxychloroquine to methotrexate or to methotrexate and etanercept in a large cohort of rheumatoid arthritis patients followed over 2 years of therapy.¹⁸

In nondiabetic women with systemic lupus erythematosus or rheumatoid arthritis, average blood glucose was significantly lower in those taking hydroxychloroquine than in nonusers. The incidence of insulin resistance was also lower, but the difference was not statistically significant.¹⁹

Some have suggested that hydroxychloroquine may prevent diabetes mellitus. In a retrospective case series, compared with rheumatoid arthritis patients not taking the drug, patients treated with hydroxychloroquine for more than 4 years had a 25% lower risk of developing diabetes mellitus.²⁰

In view of these metabolic benefits, especially regarding lipid regulation, and the above described antithrombotic properties of hydroxychloroquine, some researchers have recently hypothesized that hydroxychloroquine may be of benefit in patients with coronary artery disease.²¹ They suggested that the inflammatory contribution to the mechanism of coronary artery disease could be lessened by hydroxychloroquine even in patients without lupus erythematosus or rheumatoid arthritis.

PHARMACOLOGIC PROPERTIES

Understanding the pharmacologic properties of hydroxychloroquine is key to using it appropriately in clinical practice.

The half-life of elimination of hydroxychloroquine is 40 to 50 days, with half of the drug excreted renally in a concentration-dependent fashion.^22,23 The drug reaches 95% of its steady-state concentration by about 6 months of therapy. Shorter durations of therapy do not provide adequate time for the drug to achieve steady-state concentration and may not allow patients and providers time to see its full clinical results. Therefore, its manufacturers recommend a 6-month trial of therapy to adequately determine if the drug improves symptoms.¹

The oral bioavailability of hydroxychloroquine is about 75%, but pharmacokinetics vary among individuals.^22,23 It has been suggested that this variability affects the efficacy of hydroxychloroquine. In a study of 300 patients with cutaneous lupus erythematosus, those whose treatment failed had significantly lower blood concentrations of hydroxychloroquine, while those who achieved complete remission had significantly higher concentrations.²⁴

Another study found that titrating doses to target therapeutic blood concentrations can reduce disease activity in cutaneous lupus erythematosus.²⁵ Measuring the blood concentration of hydroxychloroquine is not common in clinical practice but may have a role in select patients in whom initial therapy using a standard dosing regimen does not produce the desired results.

HOW SAFE IS HYDROXYCHLOROQUINE?

Hydroxychloroquine has numerous adverse effects, necessitating vigilance on the part of the prescriber. Most commonly reported are retinopathy, hyperpigmentation, myopathy, and skin reactions.¹

Retinopathy

Retinopathy’s irreversibility—the threat of permanent vision loss—and its substantial prevalence in patients with a large drug exposure history, have marked retinopathy as the most concerning potential toxicity. The risk of ocular toxicity increases with the cumulative hydroxychloroquine dose. The prevalence of retinopathy in those using the drug less than 10 years is less than 2%; in contrast, the prevalence in patients with more than 20 years of exposure is reported to be as high as 20%.²⁶

The American Academy of Ophthalmology has long stated that retinopathy is a significant risk of hydroxychloroquine therapy and that patients taking hydroxychloroquine should therefore undergo routine retinal and visual field screening by an ophthalmologist.

Currently, initial screening followed by yearly screening beginning 5 years thereafter is recommended for patients at low risk of toxicity (Table 1).²⁷ Patients determined by an ophthalmologist to be at higher risk of retinopathy should be screened yearly. As identified by the American Academy of Ophthalmology, major risk factors for retinopathy include duration of use, concomitant tamoxifen exposure, significant renal disease, and preexisting retinal and macular disease.^26,28

Recommendations for hydroxychloroquine dosing and screening were recently revised, for 2 reasons. Initially, its manufacturers recommended that hydroxychloroquine dosage be no higher than 6.5 mg/kg of ideal body weight to prevent retinopathy.^1,29,30 However, it has recently been demonstrated that real body weight is a better predictor of risk of retinopathy than ideal body weight when dosing hydroxychloroquine, perhaps because of the increasing variance of real body weight in our patient population.²⁶

Further, an atypical pattern of retinopathy called pericentral retinopathy is more common in Asians. A study of about 200 patients with a history of hydroxychloroquine retinopathy, including 36 Asian patients, found that the pericentral pattern occurred in half the Asian patients but only 2% of the white patients.³¹ The mechanism for this finding is unclear, but because pericentral retinopathy spares the macula, it can be missed using standard screening methods. Therefore, the American Academy of Ophthalmology now recommends that the dose limit be reduced from 6.5 mg/kg of ideal body weight to no more than 5.0 mg/kg of real body weight (Table 2).²⁸

It is also recommended that screening methods such as automated visual fields and optical coherence tomography extend their fields beyond the macula in Asian patients to ensure that pericentral retinopathy is not missed.²⁸

Optical coherence tomography is a particularly useful tool in the ocular evaluation of patients taking hydroxychloroquine. It can detect subtle changes such as thinning of the foveal photoreceptor outer segment, thickening of the retinal pigment epithelium, and loss of the macular ganglion cell–inner plexiform layer before there are visible signs of retinopathy and before symptoms arise.³²

Currently, these guidelines are underutilized in clinical practice. Physician adherence to ophthalmologic guidelines is reported at about 50%.³³ This statistic is jarring, given the potential for permanent loss of vision in those with hydroxychloroquine-mediated retinopathy, and demonstrates the importance of reinforcing proper understanding of the use of hydroxychloroquine in clinical practice.

Cutaneous hyperpigmentation can occur with hydroxychloroquine use (Figure 1). The hyperpigmentation appears to be due to local bruising following deposition of iron in the soft tissue.

While the pigmentation may persist permanently and cause an undesirable cosmetic effect, it has not been associated with other adverse outcomes.

Myopathy is a rare adverse effect. In one case series, 3 of 214 patients treated with hydroxychloroquine developed hydroxychloroquine-induced myopathy.³⁵ Over the duration of their therapy, this was equivalent to an incidence of 1 case of myopathy in 100 patient-years of therapy. Myopathy improves with discontinuation of therapy, though it can persist for weeks, likely because of hydroxychloroquine’s prolonged elimination half-life.

Cardiomyopathy, specifically neurocardio­myopathy, is also an extremely rare adverse effect of hydroxychloroquine use. The mechanism is believed to be associated with the effect of hydroxychloroquine on lysosomal action, leading to an acquired lysosomal storage disorder with the typical cardiac hypertrophy and conduction abnormalities associated with this family of diseases.³⁶

Acute generalized exanthematous pustulosis is another rare complication of hydroxychloroquine therapy. The appearance of the reaction is similar to that of pustular psoriasis, with pustules overlying flaking and scaling skin. It usually resolves within 2 weeks after cessation of hydroxychloroquine therapy. In a select few cases, the reaction persists or waxes and wanes over a period of weeks to months, and longer durations of recovery are thought to be due to hydroxychloroquine’s long half-life, as in hydroxychloroquine-induced myopathy.³⁷

In view of this rare reaction, manufacturers of hydroxychloroquine recommend caution when using the drug in patients with psoriasis.¹

Hematologic abnormalities. In very rare cases, hydroxychloroquine is associated with hematologic abnormalities including agranulocytosis, anemia, aplastic anemia, leukopenia, and thrombocytopenia.¹

While no specific guidelines exist, caution is warranted when using hydroxychloroquine in patients with porphyria. Additionally, hydroxychloroquine and other antimalarials including primaquine have been associated with hemolysis in patients with glucose-6-phosphate dehydrogenase (G6PD) deficiency. The risk of hemolysis is generally considered low except at high hydroxychloroquine doses in patients with severe G6PD deficiency.³⁸

For the above reasons, manufacturers recommended baseline and routine blood counts, and some providers screen patients for G6PD deficiency when prescribing hydroxychloroquine (Table 3).

PREGNANCY

Hydroxychloroquine is in pregnancy category C. Information is limited, and in view of the risks, the manufacturer says that it should be avoided in pregnancy.¹ Nevertheless, it is generally considered safe during pregnancy, and its benefits may make it acceptable to continue in a patient who becomes pregnant, in spite of the possible risks.

We favor continuing hydroxychloroquine. This drug has been associated with improved maternal and fetal outcomes in lupus patients. Its use during pregnancy has not been associated with congenital malformations. The adverse visual effects of long-term hydroxychloroquine use, namely retinopathy, have never been reported in children as a consequence of exposure in utero.

In addition, hydroxychloroquine is transmitted only in minute quantities in breast milk.³⁹ In pregnant women with systemic lupus erythematosus, hydroxychloroquine was associated with a lower risk of adverse outcomes, including preterm delivery and intrauterine growth restriction.⁴⁰ However, hydroxychloroquine is far more toxic when ingested directly by infants than in adults.¹

Maternal outcomes are also improved with the use of hydroxychloroquine. Stopping hydroxychloroquine during pregnancy in women with systemic lupus erythematosus is associated with significantly higher disease activity—fully twice as high as in those who continue hydroxychloroquine.⁴¹ These study results were corroborated in a small randomized trial in which pregnant women with lupus on placebo had significantly higher lupus disease activity scores than those pregnant women who were given hydroxychloroquine.⁴² The women taking hydroxychloroquine experienced no severe lupus flares for the duration of their pregnancies.

These findings suggest not only that hydroxychloroquine is safe in pregnancy, but also that it should be continued in lupus patients during pregnancy to prevent disease flares and adverse fetal outcomes.

Benefit in preclinical lupus?

Hydroxychloroquine has a consistently profound effect on outcomes in systemic lupus erythematosus. These findings, in addition to the more widespread use of antibody screening, have led to suggestions that hydroxychloroquine could be of benefit even before systemic lupus erythematosus is diagnosed.

A study in US military personnel found that patients taking hydroxychloroquine experienced a significantly longer lag time between first reported clinical symptoms of lupus and official diagnosis compared with  matched controls who also went on to develop the disease, averaging 1.08 vs 0.29 years to disease classification.⁴³ Those who used hydroxychloroquine also had lower rates of autoantibody accumulation. Therefore, hydroxychloroquine could be of benefit in carefully selected candidates at high risk of developing systemic lupus erythematosus.

The beneficial effects of hydroxychloroquine on patients with lupus and rheumatoid arthritis, in terms of primary measures of disease activity and secondary outcomes, were discovered fortuitously and were not the original intended targets of the drug. Because of its versatility, there are numerous other disease states in which hydroxychloroquine has shown a degree of benefit or has shown a potential for benefit.

Antiviral activity?

It has been suggested that antimalarial drugs could serve as adjunctive therapies against filoviruses such as Marburg and Ebola. There is a small body of in vitro and in vivo evidence that hydroxychloroquine could temper severe systemic inflammatory responses to filoviruses both through dysregulation of lysosomes and lysosomal pH (filoviruses have a pH-dependent mechanism of action) and through decreased production of tumor necrosis factor alpha and interferons. Heavy burdens of interferons and tumor necrosis factor alpha are associated with increased mortality rates in those infected with filoviruses.⁴⁴

Antineoplastic activity?

Hydroxychloroquine has undergone in vitro testing as an adjunct to cancer therapies. There are several mechanisms by which it is theorized that hydroxychloroquine could target malignant cells, including inhibition of multidrug resistance pumps or autophagy, improvement of chemotherapy cell penetration, potentiation of presentation of major histocompatibility complexes, or even intercalation directly into DNA.^45,46 However, it can also impair natural anticancer immunity and may allow cancer cells better nutrient supply through vascular effects.

In vitro studies have shown tumoricidal effects in lymphoma and melanoma, and inhibition of growth in lung, colon, breast, cervix, larynx, liver, and prostate cancers. In vivo studies have shown that hydroxychloroquine in high doses can prolong survival in glioblastoma.⁴⁵

Unfortunately, all of these theorized or observed effects are dose-dependent and likely require doses that exceed currently recommended maximums.

Negative chronotropic effect?

Hydroxychloroquine has been found to decrease the resting heart rate in a cumulative dose-dependent fashion.⁴⁷ Further, hydroxychloroquine has been known to increase digoxin levels, and the medications should not be used in combination.¹

Whether the decrease in resting heart rate is associated with harm or benefit and whether the effect is significant enough to be considered when implementing therapy remain unanswered and deserve further investigation, as does the primary use of hydroxychloroquine for beneficial lipid and glucose reduction in patients who are otherwise healthy.

CASE CONCLUSION

The patient described at the beginning of this article was provided with information on the risks and benefits of hydroxychloroquine for treatment of her arthritis and rash suggestive of mild systemic lupus, and she opted to begin therapy. Her baseline eye screening was within normal limits. Based on her weight of 62 kg, she was started on 300 mg of hydroxychloroquine daily.

She had no significant adverse effects from the medication and reported slow improvement in her rash and joint complaints over the next 2 months. She remained on hydroxychloroquine over the next year without adverse effects or new evidence of autoimmune disease.