The magnitude of medication nonadherence and the consequent negative health impact are great and should create a sense of urgency among clinicians to address this serious problem.

The magnitude of medication nonadherence and the consequent negative health impact are great and should create a sense of urgency among clinicians to address this serious problem.

The magnitude of medication nonadherence and the consequent negative health impact are great and should create a sense of urgency among clinicians to address this serious problem.

Giant cell arteritis is the most common primary systemic vasculitis. The disease occurs almost exclusively in people over age 50, with an annual incidence of 15 to 25 per 100,000.¹ Incidence rates vary significantly depending on ethnicity. The highest rates are in whites, particularly those of North European descent.² Incidence rates progressively increase after age 50. The disease is more prevalent in women. Its cause is unknown; both genetic and environmental factors are thought to play a role.

INFLAMED ARTERIES

Giant cell arteritis is characterized by a granulomatous inflammatory infiltrate affecting large and medium-size arteries. Not all vessels are equally affected: the most susceptible are the cranial arteries, the aorta, and the aorta’s primary branches, particularly those in the upper extremities.

The disease is usually associated with an intense acute-phase response. Vessel wall inflammation results in intimal hyperplasia, luminal occlusion, and tissue ischemia. Typical histologic features include a mononuclear inflammatory infiltrate primarily composed of CD4+ T cells and activated macrophages. Multinucleated giant cells are seen in only about 50% of positive biopsies; therefore, their presence is not essential for the diagnosis.³

FOUR MAIN PHENOTYPES

Some of the possible symptoms of giant cell arteritis readily point to the correct diagnosis, eg, those due to cranial artery involvement, such as temporal headache, claudication of masticatory muscles, and visual changes. However, the clinical presentation can be quite varied.

There are four predominant clinical phenotypes, which may be present at the onset of disease or appear later as the disease progresses. Although they will be described separately in this review, these clinical presentations often overlap.

Cranial arteritis

Cranial arteritis is the clinical presentation most readily associated with giant cell arteritis. Clinical features result from involvement of branches of the external or internal carotid artery.

Headache, the most frequent symptom, is typically but not exclusively localized to the temporal areas.

Visual loss is due to involvement of the branches of the ophthalmic or posterior ciliary arteries, resulting in ischemia of the optic nerve or its tracts. It occurs in up to 20% of patients.^4,5

Other symptoms and complications from cranial arteritis include tenderness of the scalp and temporal areas, claudication of the tongue or jaw muscles, stroke, and more rarely, tongue infarction.

Polymyalgia rheumatica

Polymyalgia rheumatica is a clinical syndrome that can occur by itself or in conjunction with giant cell arteritis. It may occur independently of giant cell arteritis, but also occurs in about 40% of patients with giant cell arteritis. It may precede, develop simultaneously with, or develop later during the course of the giant cell arteritis.^6,7 It is a common clinical manifestation in relapses of giant cell arteritis, even in those who did not have symptoms of polymyalgia rheumatica at the time giant cell arteritis was diagnosed.

Polymyalgia rheumatica is characterized by aching of the shoulder and hip girdle, with morning stiffness. Fatigue and malaise are often present and may be severe. Some patients with polymyalgia rheumatica may also present with peripheral joint synovitis, which may be mistakenly diagnosed as rheumatoid arthritis.⁸ Muscle weakness and elevated muscle enzymes are not associated with polymyalgia rheumatica.

Polymyalgia rheumatica is a clinical diagnosis. Approximately 80% of patients with polymyalgia rheumatica have an elevated erythrocyte sedimentation rate or an elevated C-reactive protein level.⁹ When it occurs in the absence of giant cell arteritis, it is treated differently, with less intense doses of corticosteroids. All patients with polymyalgia rheumatica should be routinely questioned regarding symptoms of giant cell arteritis.

Nonspecific systemic inflammatory disease

Some patients with giant cell arteritis may present with a nonspecific systemic inflammatory disease characterized by some combination of fever, night sweats, fatigue, malaise, and weight loss. In these patients, the diagnosis may be delayed by the lack of localizing symptoms.

Laboratory tests typically show anemia, leukocytosis, and thrombocytosis. The erythrocyte sedimentation rate and the C-reactive protein level are usually very high.

Giant cell arteritis should be in the differential diagnosis when these signs and symptoms are found in patients over age 50.

Large-vessel vasculitis

Although thoracic aortic aneurysm and dissection have been described as late complications of giant cell arteritis, large-vessel vasculitis may precede or occur concomitantly with cranial arteritis early in the disease.^10,11

Population-based surveys have shown that large-vessel vasculitis is extremely frequent in patients with giant cell arteritis. In a postmortem study of 11 patients with giant cell arteritis, all of them had evidence of arteritis involving the subclavian artery, the carotid artery, and the aorta.¹²

Patients may have no symptoms or may present with symptoms or signs of tissue ischemia such as claudication of the extremities, carotid artery tenderness, decreased or absent pulses, and large-vessel bruits on physical examination.

CONSIDER THE DIAGNOSIS IN OLDER PATIENTS

Giant cell arteritis should always be considered in patients over age 50 who have any of the clinical features described above. It is therefore very important to be familiar with its symptoms and signs.

A complete and detailed history and a detailed but focused physical examination that includes a comprehensive vascular examination are the first and most important steps in establishing the diagnosis. The vascular examination includes measuring the blood pressure in all four extremities, palpating the peripheral pulses, listening for bruits, and palpating the temporal arteries.

Temporal artery biopsy: The gold standard

Confirming the diagnosis of giant cell arteritis requires histologic findings of inflammation in the temporal artery. Superficial temporal artery biopsy is recommended for diagnostic confirmation in patients who have cranial symptoms and other signs suggesting the disease.

The biopsy should be performed on the same side as the symptoms or abnormal findings on examination. Performing a biopsy in both temporal arteries may increase the diagnostic yield but may not need to be done routinely.¹³

Although some experts recommend temporal artery biopsy in all patients in whom giant cell arteritis is suspected, biopsy has a lower diagnostic yield in patients who have no cranial symptoms. Interestingly, 5% to 15% of temporal artery biopsies performed in patients who had isolated polymyalgia rheumatica were found to be positive.^14,15 Patients with polymyalgia rheumatica and no clinical symptoms to suggest giant cell arteritis generally are not biopsied.

The segmental nature of the inflammation involving the temporal artery in giant cell arteritis may result in negative biopsy results in patients with giant cell arteritis. A temporal artery biopsy length of 5 mm or less has a very low (8%) rate of positive results, whereas a length longer than 20 mm exceeds a 50% rate of positive results. Although the optimal length of a temporal artery specimen is still debated, a longer biopsy specimen should be obtained to increase the chance of arterial specimens showing inflammatory changes.^16,17

Laboratory studies: Acute-phase reactants may be elevated

High levels of acute-phase reactants should increase one’s suspicion of giant cell arteritis. Elevations in the erythrocyte sedimentation rate and C-reactive protein and interleukin 6 levels reflect the inflammatory process in this disease.¹⁹ However, not all patients with giant cell arteritis have a high sedimentation rate, and as many as 20% of patients with biopsy-proven giant cell arteritis have a normal sedimentation rate before therapy.²⁰ Therefore, a normal sedimentation rate does not exclude the diagnosis of giant cell arteritis and should not delay its diagnosis and treatment.

As a result of systemic inflammation, the patient may also present with normochromic normocytic anemia, leukocytosis, and thrombocytosis.

Imaging studies are controversial

Imaging studies are potentially useful diagnostic tools in large-vessel vasculitis but are still the subject of significant controversy.

Ultrasonography of the temporal artery has been a controversial subject in many studies.^21,22 Color duplex ultrasonography of the temporal artery has been reported to be helpful in the diagnosis of giant cell arteritis (showing a “halo” around the arterial lumen), but further studies are needed to establish its clinical utility.

At this time, temporal artery biopsy remains the gold standard diagnostic test for giant cell arteritis, and ultrasonography is neither a substitute for biopsy nor a screening test for this disease.²³ Some have suggested, however, that ultrasonography may help to identify the best site for biopsy of the temporal artery in some patients.

Arteriography is an accurate technique for evaluating the vessel lumen and allows for measuring central blood pressure and performing vascular interventions. However, because of potential complications, it has been largely replaced by noninvasive angiographic imaging to delineate vascular anatomy.

TREATMENT

Glucocorticoid therapy remains the standard of care

Once the diagnosis of giant cell arteritis is established, glucocorticoid treatment should be started. Glucocorticoids are the standard therapy, and they usually bring about a prompt clinical response. Although never evaluated in placebo-controlled trials, these drugs have been shown to prevent progression of visual loss in a retrospective study.²⁶

In patients with visual symptoms or imminent visual loss, therapy should be started promptly once suspicion of giant cell arteritis is raised; ie, one should not wait until the diagnosis is confirmed by biopsy.

Ideally, a glucocorticoid should be started after a temporal artery biopsy is done, but treatment should not be delayed, as it rapidly suppresses the inflammatory response and may prevent complications from tissue ischemia, such as blindness. Visual loss is usually irreversible.

There is still a role for obtaining a temporal artery biopsy up to several weeks after glucocorticoid therapy is started, as the pathological abnormalities of arteritis do not rapidly resolve.²⁷

Glucocorticoid therapy is highly effective in inducing disease remission in patients with giant cell arteritis. Nearly all patients respond to 1 mg/kg (40–60 mg) per day of prednisone or its equivalent.

The initial dosing is usually maintained for 4 weeks and then decreased slowly. The duration of therapy varies; most patients remain on therapy for at least 1 year, and some cannot stop it completely without recurrence of symptoms.

If a patient is about to lose his or her vision or has lost all or some vision in one eye, a higher initial dose of a glucocorticoid is usually used (ie, a pulse of 500 or 1,000 mg of intravenous methylprednisolone) and may be beneficial.²⁸

Although a rapid clinical response to therapy is usually seen within 48 hours, some patients may have a more delayed clinical improvement.

Alternate-day therapy was compared with daily therapy and was found to be less effective, and as a result it is not recommended.²⁹

Glucocorticoid therapy can cause significant toxicity in patients with giant cell arteritis, as they commonly must take these drugs for long periods. The rate of relapse in those who discontinue therapy is quite high—as high as 77% within 12 months.³⁰

Given the concern about glucocorticoid toxicity, several studies have evaluated alternative strategies and other immunosuppressive drugs. However, no study has concluded that other medications are effective in the treatment of giant cell arteritis.

Mazlumzadeh et al³¹ evaluated the initial use of intravenous pulse methylprednisolone therapy (15 mg/kg ideal body weight on 3 consecutive days) in an attempt to decrease the glucocorticoid requirement. Although the group receiving this therapy had a lower relapse rate than in the placebo group, and their cumulative dose of glucocorticoid was lower (all patients also received oral prednisone), there was no reduction in the rate of glucocorticoid-associated toxicity.³¹ Care must be taken to prevent and monitor for corticosteroid complications such as osteoporosis, glaucoma, diabetes mellitus, and hypertension.

Methotrexate: Mixed results in clinical trials

Methotrexate has been evaluated in three prospective randomized trials,^30,32,33 with mixed results.

Spiera et al³² enrolled 21 patients in a double-blind placebo-controlled trial: 12 patients received low-dose methotrexate (7.5 mg/week) and 9 received placebo. In addition, all 21 received a glucocorticoid. There was no significant difference between the methotrexate- and placebo-treated patients in the cumulative dose of glucocorticoid, duration of glucocorticoid therapy, time to taper off the glucocorticoid to less than 10 mg of prednisone per day, or glucocorticoidrelated adverse effects.

Jover et al,³³ in another double-blind placebo-controlled trial, studied 42 patients with giant cell arteritis, half of whom were randomized to receive methotrexate 10 mg/week, while the other half received placebo. All patients received prednisone. Patients in the methotrexate group had fewer relapses and a 25% lower cumulative dose of prednisone during follow-up. However, the incidence of adverse events was similar in both groups. Methotrexate was discontinued in 3 patients who developed drug-related adverse events.

Hoffman et al³⁰ randomized 98 patients to receive either methotrexate (up to 15 mg/week) or placebo in a double-blind fashion. All patients also received prednisone at an initial dose of 1 mg/kg/day (up to 60 mg/day). At completion of the study, no differences between the groups were noted in the rates of relapse or treatment-related morbidity or in the cumulative dose of glucocorticoid. However, treatment with methotrexate appeared to lower the rate of recurrence of isolated polymyalgia rheumatica in a small number of patients.³⁰

Comment. Differences in the results of these trials may be attributed to several factors, including different definitions of relapses and different glucocorticoid doses and tapering regimens.

A meta-analysis of these three trials³⁴ showed a reduction in the risk of relapse: 4 patients would have to be treated to prevent one first relapse, 5 would have to be treated to prevent one second relapse, and 11 would need to be treated to prevent one first relapse of cranial symptoms in the first 48 weeks. However, the main goal of methotrexate therapy is to decrease the frequency of adverse events from glucocorticoids, and this meta-analysis found no difference in rates of glucocorticoid-related adverse events in patients treated with methotrexate.

The study raises the question of whether methotrexate should be further evaluated in in different patient populations and at higher doses.³⁴

Infliximab is not recommended

In a prospective study, patients with giant cell arteritis were randomly assigned to receive either infliximab (Remicade) 5 mg/kg every 8 weeks or placebo, in addition to standard glucocorticoid therapy. The study showed no significant difference in the relapse rate and a higher rate of infection in the infliximab group (71%) than in the placebo group (56%). Given the lack of any benefit observed in this study, infliximab is not recommended in the treatment of patients with giant cell arteritis.³⁵

Aspirin is recommended

Daily low-dose aspirin therapy has been shown in several studies to be effective in preventing ischemic complications of giant cell arteritis, including stroke and visual loss. It is currently recommended that all patients with giant cell arteritis without a major contraindication take aspirin 81 mg daily.^36–38

Giant cell arteritis is the most common primary systemic vasculitis. The disease occurs almost exclusively in people over age 50, with an annual incidence of 15 to 25 per 100,000.¹ Incidence rates vary significantly depending on ethnicity. The highest rates are in whites, particularly those of North European descent.² Incidence rates progressively increase after age 50. The disease is more prevalent in women. Its cause is unknown; both genetic and environmental factors are thought to play a role.

INFLAMED ARTERIES

Giant cell arteritis is characterized by a granulomatous inflammatory infiltrate affecting large and medium-size arteries. Not all vessels are equally affected: the most susceptible are the cranial arteries, the aorta, and the aorta’s primary branches, particularly those in the upper extremities.

The disease is usually associated with an intense acute-phase response. Vessel wall inflammation results in intimal hyperplasia, luminal occlusion, and tissue ischemia. Typical histologic features include a mononuclear inflammatory infiltrate primarily composed of CD4+ T cells and activated macrophages. Multinucleated giant cells are seen in only about 50% of positive biopsies; therefore, their presence is not essential for the diagnosis.³

FOUR MAIN PHENOTYPES

Some of the possible symptoms of giant cell arteritis readily point to the correct diagnosis, eg, those due to cranial artery involvement, such as temporal headache, claudication of masticatory muscles, and visual changes. However, the clinical presentation can be quite varied.

There are four predominant clinical phenotypes, which may be present at the onset of disease or appear later as the disease progresses. Although they will be described separately in this review, these clinical presentations often overlap.

Cranial arteritis

Cranial arteritis is the clinical presentation most readily associated with giant cell arteritis. Clinical features result from involvement of branches of the external or internal carotid artery.

Headache, the most frequent symptom, is typically but not exclusively localized to the temporal areas.

Visual loss is due to involvement of the branches of the ophthalmic or posterior ciliary arteries, resulting in ischemia of the optic nerve or its tracts. It occurs in up to 20% of patients.^4,5

Other symptoms and complications from cranial arteritis include tenderness of the scalp and temporal areas, claudication of the tongue or jaw muscles, stroke, and more rarely, tongue infarction.

Polymyalgia rheumatica

Polymyalgia rheumatica is a clinical syndrome that can occur by itself or in conjunction with giant cell arteritis. It may occur independently of giant cell arteritis, but also occurs in about 40% of patients with giant cell arteritis. It may precede, develop simultaneously with, or develop later during the course of the giant cell arteritis.^6,7 It is a common clinical manifestation in relapses of giant cell arteritis, even in those who did not have symptoms of polymyalgia rheumatica at the time giant cell arteritis was diagnosed.

Polymyalgia rheumatica is characterized by aching of the shoulder and hip girdle, with morning stiffness. Fatigue and malaise are often present and may be severe. Some patients with polymyalgia rheumatica may also present with peripheral joint synovitis, which may be mistakenly diagnosed as rheumatoid arthritis.⁸ Muscle weakness and elevated muscle enzymes are not associated with polymyalgia rheumatica.

Polymyalgia rheumatica is a clinical diagnosis. Approximately 80% of patients with polymyalgia rheumatica have an elevated erythrocyte sedimentation rate or an elevated C-reactive protein level.⁹ When it occurs in the absence of giant cell arteritis, it is treated differently, with less intense doses of corticosteroids. All patients with polymyalgia rheumatica should be routinely questioned regarding symptoms of giant cell arteritis.

Nonspecific systemic inflammatory disease

Some patients with giant cell arteritis may present with a nonspecific systemic inflammatory disease characterized by some combination of fever, night sweats, fatigue, malaise, and weight loss. In these patients, the diagnosis may be delayed by the lack of localizing symptoms.

Laboratory tests typically show anemia, leukocytosis, and thrombocytosis. The erythrocyte sedimentation rate and the C-reactive protein level are usually very high.

Giant cell arteritis should be in the differential diagnosis when these signs and symptoms are found in patients over age 50.

Large-vessel vasculitis

Although thoracic aortic aneurysm and dissection have been described as late complications of giant cell arteritis, large-vessel vasculitis may precede or occur concomitantly with cranial arteritis early in the disease.^10,11

Population-based surveys have shown that large-vessel vasculitis is extremely frequent in patients with giant cell arteritis. In a postmortem study of 11 patients with giant cell arteritis, all of them had evidence of arteritis involving the subclavian artery, the carotid artery, and the aorta.¹²

Patients may have no symptoms or may present with symptoms or signs of tissue ischemia such as claudication of the extremities, carotid artery tenderness, decreased or absent pulses, and large-vessel bruits on physical examination.

CONSIDER THE DIAGNOSIS IN OLDER PATIENTS

Giant cell arteritis should always be considered in patients over age 50 who have any of the clinical features described above. It is therefore very important to be familiar with its symptoms and signs.

A complete and detailed history and a detailed but focused physical examination that includes a comprehensive vascular examination are the first and most important steps in establishing the diagnosis. The vascular examination includes measuring the blood pressure in all four extremities, palpating the peripheral pulses, listening for bruits, and palpating the temporal arteries.

Temporal artery biopsy: The gold standard

Confirming the diagnosis of giant cell arteritis requires histologic findings of inflammation in the temporal artery. Superficial temporal artery biopsy is recommended for diagnostic confirmation in patients who have cranial symptoms and other signs suggesting the disease.

The biopsy should be performed on the same side as the symptoms or abnormal findings on examination. Performing a biopsy in both temporal arteries may increase the diagnostic yield but may not need to be done routinely.¹³

Although some experts recommend temporal artery biopsy in all patients in whom giant cell arteritis is suspected, biopsy has a lower diagnostic yield in patients who have no cranial symptoms. Interestingly, 5% to 15% of temporal artery biopsies performed in patients who had isolated polymyalgia rheumatica were found to be positive.^14,15 Patients with polymyalgia rheumatica and no clinical symptoms to suggest giant cell arteritis generally are not biopsied.

The segmental nature of the inflammation involving the temporal artery in giant cell arteritis may result in negative biopsy results in patients with giant cell arteritis. A temporal artery biopsy length of 5 mm or less has a very low (8%) rate of positive results, whereas a length longer than 20 mm exceeds a 50% rate of positive results. Although the optimal length of a temporal artery specimen is still debated, a longer biopsy specimen should be obtained to increase the chance of arterial specimens showing inflammatory changes.^16,17

Laboratory studies: Acute-phase reactants may be elevated

High levels of acute-phase reactants should increase one’s suspicion of giant cell arteritis. Elevations in the erythrocyte sedimentation rate and C-reactive protein and interleukin 6 levels reflect the inflammatory process in this disease.¹⁹ However, not all patients with giant cell arteritis have a high sedimentation rate, and as many as 20% of patients with biopsy-proven giant cell arteritis have a normal sedimentation rate before therapy.²⁰ Therefore, a normal sedimentation rate does not exclude the diagnosis of giant cell arteritis and should not delay its diagnosis and treatment.

As a result of systemic inflammation, the patient may also present with normochromic normocytic anemia, leukocytosis, and thrombocytosis.

Imaging studies are controversial

Imaging studies are potentially useful diagnostic tools in large-vessel vasculitis but are still the subject of significant controversy.

Ultrasonography of the temporal artery has been a controversial subject in many studies.^21,22 Color duplex ultrasonography of the temporal artery has been reported to be helpful in the diagnosis of giant cell arteritis (showing a “halo” around the arterial lumen), but further studies are needed to establish its clinical utility.

At this time, temporal artery biopsy remains the gold standard diagnostic test for giant cell arteritis, and ultrasonography is neither a substitute for biopsy nor a screening test for this disease.²³ Some have suggested, however, that ultrasonography may help to identify the best site for biopsy of the temporal artery in some patients.

Arteriography is an accurate technique for evaluating the vessel lumen and allows for measuring central blood pressure and performing vascular interventions. However, because of potential complications, it has been largely replaced by noninvasive angiographic imaging to delineate vascular anatomy.

TREATMENT

Glucocorticoid therapy remains the standard of care

Once the diagnosis of giant cell arteritis is established, glucocorticoid treatment should be started. Glucocorticoids are the standard therapy, and they usually bring about a prompt clinical response. Although never evaluated in placebo-controlled trials, these drugs have been shown to prevent progression of visual loss in a retrospective study.²⁶

In patients with visual symptoms or imminent visual loss, therapy should be started promptly once suspicion of giant cell arteritis is raised; ie, one should not wait until the diagnosis is confirmed by biopsy.

Ideally, a glucocorticoid should be started after a temporal artery biopsy is done, but treatment should not be delayed, as it rapidly suppresses the inflammatory response and may prevent complications from tissue ischemia, such as blindness. Visual loss is usually irreversible.

There is still a role for obtaining a temporal artery biopsy up to several weeks after glucocorticoid therapy is started, as the pathological abnormalities of arteritis do not rapidly resolve.²⁷

Glucocorticoid therapy is highly effective in inducing disease remission in patients with giant cell arteritis. Nearly all patients respond to 1 mg/kg (40–60 mg) per day of prednisone or its equivalent.

The initial dosing is usually maintained for 4 weeks and then decreased slowly. The duration of therapy varies; most patients remain on therapy for at least 1 year, and some cannot stop it completely without recurrence of symptoms.

If a patient is about to lose his or her vision or has lost all or some vision in one eye, a higher initial dose of a glucocorticoid is usually used (ie, a pulse of 500 or 1,000 mg of intravenous methylprednisolone) and may be beneficial.²⁸

Although a rapid clinical response to therapy is usually seen within 48 hours, some patients may have a more delayed clinical improvement.

Alternate-day therapy was compared with daily therapy and was found to be less effective, and as a result it is not recommended.²⁹

Glucocorticoid therapy can cause significant toxicity in patients with giant cell arteritis, as they commonly must take these drugs for long periods. The rate of relapse in those who discontinue therapy is quite high—as high as 77% within 12 months.³⁰

Given the concern about glucocorticoid toxicity, several studies have evaluated alternative strategies and other immunosuppressive drugs. However, no study has concluded that other medications are effective in the treatment of giant cell arteritis.

Mazlumzadeh et al³¹ evaluated the initial use of intravenous pulse methylprednisolone therapy (15 mg/kg ideal body weight on 3 consecutive days) in an attempt to decrease the glucocorticoid requirement. Although the group receiving this therapy had a lower relapse rate than in the placebo group, and their cumulative dose of glucocorticoid was lower (all patients also received oral prednisone), there was no reduction in the rate of glucocorticoid-associated toxicity.³¹ Care must be taken to prevent and monitor for corticosteroid complications such as osteoporosis, glaucoma, diabetes mellitus, and hypertension.

Methotrexate: Mixed results in clinical trials

Methotrexate has been evaluated in three prospective randomized trials,^30,32,33 with mixed results.

Spiera et al³² enrolled 21 patients in a double-blind placebo-controlled trial: 12 patients received low-dose methotrexate (7.5 mg/week) and 9 received placebo. In addition, all 21 received a glucocorticoid. There was no significant difference between the methotrexate- and placebo-treated patients in the cumulative dose of glucocorticoid, duration of glucocorticoid therapy, time to taper off the glucocorticoid to less than 10 mg of prednisone per day, or glucocorticoidrelated adverse effects.

Jover et al,³³ in another double-blind placebo-controlled trial, studied 42 patients with giant cell arteritis, half of whom were randomized to receive methotrexate 10 mg/week, while the other half received placebo. All patients received prednisone. Patients in the methotrexate group had fewer relapses and a 25% lower cumulative dose of prednisone during follow-up. However, the incidence of adverse events was similar in both groups. Methotrexate was discontinued in 3 patients who developed drug-related adverse events.

Hoffman et al³⁰ randomized 98 patients to receive either methotrexate (up to 15 mg/week) or placebo in a double-blind fashion. All patients also received prednisone at an initial dose of 1 mg/kg/day (up to 60 mg/day). At completion of the study, no differences between the groups were noted in the rates of relapse or treatment-related morbidity or in the cumulative dose of glucocorticoid. However, treatment with methotrexate appeared to lower the rate of recurrence of isolated polymyalgia rheumatica in a small number of patients.³⁰

Comment. Differences in the results of these trials may be attributed to several factors, including different definitions of relapses and different glucocorticoid doses and tapering regimens.

A meta-analysis of these three trials³⁴ showed a reduction in the risk of relapse: 4 patients would have to be treated to prevent one first relapse, 5 would have to be treated to prevent one second relapse, and 11 would need to be treated to prevent one first relapse of cranial symptoms in the first 48 weeks. However, the main goal of methotrexate therapy is to decrease the frequency of adverse events from glucocorticoids, and this meta-analysis found no difference in rates of glucocorticoid-related adverse events in patients treated with methotrexate.

The study raises the question of whether methotrexate should be further evaluated in in different patient populations and at higher doses.³⁴

Infliximab is not recommended

In a prospective study, patients with giant cell arteritis were randomly assigned to receive either infliximab (Remicade) 5 mg/kg every 8 weeks or placebo, in addition to standard glucocorticoid therapy. The study showed no significant difference in the relapse rate and a higher rate of infection in the infliximab group (71%) than in the placebo group (56%). Given the lack of any benefit observed in this study, infliximab is not recommended in the treatment of patients with giant cell arteritis.³⁵

Aspirin is recommended

Daily low-dose aspirin therapy has been shown in several studies to be effective in preventing ischemic complications of giant cell arteritis, including stroke and visual loss. It is currently recommended that all patients with giant cell arteritis without a major contraindication take aspirin 81 mg daily.^36–38

Giant cell arteritis is the most common primary systemic vasculitis. The disease occurs almost exclusively in people over age 50, with an annual incidence of 15 to 25 per 100,000.¹ Incidence rates vary significantly depending on ethnicity. The highest rates are in whites, particularly those of North European descent.² Incidence rates progressively increase after age 50. The disease is more prevalent in women. Its cause is unknown; both genetic and environmental factors are thought to play a role.

INFLAMED ARTERIES

Giant cell arteritis is characterized by a granulomatous inflammatory infiltrate affecting large and medium-size arteries. Not all vessels are equally affected: the most susceptible are the cranial arteries, the aorta, and the aorta’s primary branches, particularly those in the upper extremities.

The disease is usually associated with an intense acute-phase response. Vessel wall inflammation results in intimal hyperplasia, luminal occlusion, and tissue ischemia. Typical histologic features include a mononuclear inflammatory infiltrate primarily composed of CD4+ T cells and activated macrophages. Multinucleated giant cells are seen in only about 50% of positive biopsies; therefore, their presence is not essential for the diagnosis.³

FOUR MAIN PHENOTYPES

Some of the possible symptoms of giant cell arteritis readily point to the correct diagnosis, eg, those due to cranial artery involvement, such as temporal headache, claudication of masticatory muscles, and visual changes. However, the clinical presentation can be quite varied.

There are four predominant clinical phenotypes, which may be present at the onset of disease or appear later as the disease progresses. Although they will be described separately in this review, these clinical presentations often overlap.

Cranial arteritis

Cranial arteritis is the clinical presentation most readily associated with giant cell arteritis. Clinical features result from involvement of branches of the external or internal carotid artery.

Headache, the most frequent symptom, is typically but not exclusively localized to the temporal areas.

Visual loss is due to involvement of the branches of the ophthalmic or posterior ciliary arteries, resulting in ischemia of the optic nerve or its tracts. It occurs in up to 20% of patients.^4,5

Other symptoms and complications from cranial arteritis include tenderness of the scalp and temporal areas, claudication of the tongue or jaw muscles, stroke, and more rarely, tongue infarction.

Polymyalgia rheumatica

Polymyalgia rheumatica is a clinical syndrome that can occur by itself or in conjunction with giant cell arteritis. It may occur independently of giant cell arteritis, but also occurs in about 40% of patients with giant cell arteritis. It may precede, develop simultaneously with, or develop later during the course of the giant cell arteritis.^6,7 It is a common clinical manifestation in relapses of giant cell arteritis, even in those who did not have symptoms of polymyalgia rheumatica at the time giant cell arteritis was diagnosed.

Polymyalgia rheumatica is characterized by aching of the shoulder and hip girdle, with morning stiffness. Fatigue and malaise are often present and may be severe. Some patients with polymyalgia rheumatica may also present with peripheral joint synovitis, which may be mistakenly diagnosed as rheumatoid arthritis.⁸ Muscle weakness and elevated muscle enzymes are not associated with polymyalgia rheumatica.

Polymyalgia rheumatica is a clinical diagnosis. Approximately 80% of patients with polymyalgia rheumatica have an elevated erythrocyte sedimentation rate or an elevated C-reactive protein level.⁹ When it occurs in the absence of giant cell arteritis, it is treated differently, with less intense doses of corticosteroids. All patients with polymyalgia rheumatica should be routinely questioned regarding symptoms of giant cell arteritis.

Nonspecific systemic inflammatory disease

Some patients with giant cell arteritis may present with a nonspecific systemic inflammatory disease characterized by some combination of fever, night sweats, fatigue, malaise, and weight loss. In these patients, the diagnosis may be delayed by the lack of localizing symptoms.

Laboratory tests typically show anemia, leukocytosis, and thrombocytosis. The erythrocyte sedimentation rate and the C-reactive protein level are usually very high.

Giant cell arteritis should be in the differential diagnosis when these signs and symptoms are found in patients over age 50.

Large-vessel vasculitis

Although thoracic aortic aneurysm and dissection have been described as late complications of giant cell arteritis, large-vessel vasculitis may precede or occur concomitantly with cranial arteritis early in the disease.^10,11

Population-based surveys have shown that large-vessel vasculitis is extremely frequent in patients with giant cell arteritis. In a postmortem study of 11 patients with giant cell arteritis, all of them had evidence of arteritis involving the subclavian artery, the carotid artery, and the aorta.¹²

Patients may have no symptoms or may present with symptoms or signs of tissue ischemia such as claudication of the extremities, carotid artery tenderness, decreased or absent pulses, and large-vessel bruits on physical examination.

CONSIDER THE DIAGNOSIS IN OLDER PATIENTS

Giant cell arteritis should always be considered in patients over age 50 who have any of the clinical features described above. It is therefore very important to be familiar with its symptoms and signs.

A complete and detailed history and a detailed but focused physical examination that includes a comprehensive vascular examination are the first and most important steps in establishing the diagnosis. The vascular examination includes measuring the blood pressure in all four extremities, palpating the peripheral pulses, listening for bruits, and palpating the temporal arteries.

Temporal artery biopsy: The gold standard

Confirming the diagnosis of giant cell arteritis requires histologic findings of inflammation in the temporal artery. Superficial temporal artery biopsy is recommended for diagnostic confirmation in patients who have cranial symptoms and other signs suggesting the disease.

The biopsy should be performed on the same side as the symptoms or abnormal findings on examination. Performing a biopsy in both temporal arteries may increase the diagnostic yield but may not need to be done routinely.¹³

Although some experts recommend temporal artery biopsy in all patients in whom giant cell arteritis is suspected, biopsy has a lower diagnostic yield in patients who have no cranial symptoms. Interestingly, 5% to 15% of temporal artery biopsies performed in patients who had isolated polymyalgia rheumatica were found to be positive.^14,15 Patients with polymyalgia rheumatica and no clinical symptoms to suggest giant cell arteritis generally are not biopsied.

The segmental nature of the inflammation involving the temporal artery in giant cell arteritis may result in negative biopsy results in patients with giant cell arteritis. A temporal artery biopsy length of 5 mm or less has a very low (8%) rate of positive results, whereas a length longer than 20 mm exceeds a 50% rate of positive results. Although the optimal length of a temporal artery specimen is still debated, a longer biopsy specimen should be obtained to increase the chance of arterial specimens showing inflammatory changes.^16,17

Laboratory studies: Acute-phase reactants may be elevated

High levels of acute-phase reactants should increase one’s suspicion of giant cell arteritis. Elevations in the erythrocyte sedimentation rate and C-reactive protein and interleukin 6 levels reflect the inflammatory process in this disease.¹⁹ However, not all patients with giant cell arteritis have a high sedimentation rate, and as many as 20% of patients with biopsy-proven giant cell arteritis have a normal sedimentation rate before therapy.²⁰ Therefore, a normal sedimentation rate does not exclude the diagnosis of giant cell arteritis and should not delay its diagnosis and treatment.

As a result of systemic inflammation, the patient may also present with normochromic normocytic anemia, leukocytosis, and thrombocytosis.

Imaging studies are controversial

Imaging studies are potentially useful diagnostic tools in large-vessel vasculitis but are still the subject of significant controversy.

Ultrasonography of the temporal artery has been a controversial subject in many studies.^21,22 Color duplex ultrasonography of the temporal artery has been reported to be helpful in the diagnosis of giant cell arteritis (showing a “halo” around the arterial lumen), but further studies are needed to establish its clinical utility.

At this time, temporal artery biopsy remains the gold standard diagnostic test for giant cell arteritis, and ultrasonography is neither a substitute for biopsy nor a screening test for this disease.²³ Some have suggested, however, that ultrasonography may help to identify the best site for biopsy of the temporal artery in some patients.

Arteriography is an accurate technique for evaluating the vessel lumen and allows for measuring central blood pressure and performing vascular interventions. However, because of potential complications, it has been largely replaced by noninvasive angiographic imaging to delineate vascular anatomy.

TREATMENT

Glucocorticoid therapy remains the standard of care

Once the diagnosis of giant cell arteritis is established, glucocorticoid treatment should be started. Glucocorticoids are the standard therapy, and they usually bring about a prompt clinical response. Although never evaluated in placebo-controlled trials, these drugs have been shown to prevent progression of visual loss in a retrospective study.²⁶

In patients with visual symptoms or imminent visual loss, therapy should be started promptly once suspicion of giant cell arteritis is raised; ie, one should not wait until the diagnosis is confirmed by biopsy.

Ideally, a glucocorticoid should be started after a temporal artery biopsy is done, but treatment should not be delayed, as it rapidly suppresses the inflammatory response and may prevent complications from tissue ischemia, such as blindness. Visual loss is usually irreversible.

There is still a role for obtaining a temporal artery biopsy up to several weeks after glucocorticoid therapy is started, as the pathological abnormalities of arteritis do not rapidly resolve.²⁷

Glucocorticoid therapy is highly effective in inducing disease remission in patients with giant cell arteritis. Nearly all patients respond to 1 mg/kg (40–60 mg) per day of prednisone or its equivalent.

The initial dosing is usually maintained for 4 weeks and then decreased slowly. The duration of therapy varies; most patients remain on therapy for at least 1 year, and some cannot stop it completely without recurrence of symptoms.

If a patient is about to lose his or her vision or has lost all or some vision in one eye, a higher initial dose of a glucocorticoid is usually used (ie, a pulse of 500 or 1,000 mg of intravenous methylprednisolone) and may be beneficial.²⁸

Although a rapid clinical response to therapy is usually seen within 48 hours, some patients may have a more delayed clinical improvement.

Alternate-day therapy was compared with daily therapy and was found to be less effective, and as a result it is not recommended.²⁹

Glucocorticoid therapy can cause significant toxicity in patients with giant cell arteritis, as they commonly must take these drugs for long periods. The rate of relapse in those who discontinue therapy is quite high—as high as 77% within 12 months.³⁰

Given the concern about glucocorticoid toxicity, several studies have evaluated alternative strategies and other immunosuppressive drugs. However, no study has concluded that other medications are effective in the treatment of giant cell arteritis.

Mazlumzadeh et al³¹ evaluated the initial use of intravenous pulse methylprednisolone therapy (15 mg/kg ideal body weight on 3 consecutive days) in an attempt to decrease the glucocorticoid requirement. Although the group receiving this therapy had a lower relapse rate than in the placebo group, and their cumulative dose of glucocorticoid was lower (all patients also received oral prednisone), there was no reduction in the rate of glucocorticoid-associated toxicity.³¹ Care must be taken to prevent and monitor for corticosteroid complications such as osteoporosis, glaucoma, diabetes mellitus, and hypertension.

Methotrexate: Mixed results in clinical trials

Methotrexate has been evaluated in three prospective randomized trials,^30,32,33 with mixed results.

Spiera et al³² enrolled 21 patients in a double-blind placebo-controlled trial: 12 patients received low-dose methotrexate (7.5 mg/week) and 9 received placebo. In addition, all 21 received a glucocorticoid. There was no significant difference between the methotrexate- and placebo-treated patients in the cumulative dose of glucocorticoid, duration of glucocorticoid therapy, time to taper off the glucocorticoid to less than 10 mg of prednisone per day, or glucocorticoidrelated adverse effects.

Jover et al,³³ in another double-blind placebo-controlled trial, studied 42 patients with giant cell arteritis, half of whom were randomized to receive methotrexate 10 mg/week, while the other half received placebo. All patients received prednisone. Patients in the methotrexate group had fewer relapses and a 25% lower cumulative dose of prednisone during follow-up. However, the incidence of adverse events was similar in both groups. Methotrexate was discontinued in 3 patients who developed drug-related adverse events.

Hoffman et al³⁰ randomized 98 patients to receive either methotrexate (up to 15 mg/week) or placebo in a double-blind fashion. All patients also received prednisone at an initial dose of 1 mg/kg/day (up to 60 mg/day). At completion of the study, no differences between the groups were noted in the rates of relapse or treatment-related morbidity or in the cumulative dose of glucocorticoid. However, treatment with methotrexate appeared to lower the rate of recurrence of isolated polymyalgia rheumatica in a small number of patients.³⁰

Comment. Differences in the results of these trials may be attributed to several factors, including different definitions of relapses and different glucocorticoid doses and tapering regimens.

A meta-analysis of these three trials³⁴ showed a reduction in the risk of relapse: 4 patients would have to be treated to prevent one first relapse, 5 would have to be treated to prevent one second relapse, and 11 would need to be treated to prevent one first relapse of cranial symptoms in the first 48 weeks. However, the main goal of methotrexate therapy is to decrease the frequency of adverse events from glucocorticoids, and this meta-analysis found no difference in rates of glucocorticoid-related adverse events in patients treated with methotrexate.

The study raises the question of whether methotrexate should be further evaluated in in different patient populations and at higher doses.³⁴

Infliximab is not recommended

In a prospective study, patients with giant cell arteritis were randomly assigned to receive either infliximab (Remicade) 5 mg/kg every 8 weeks or placebo, in addition to standard glucocorticoid therapy. The study showed no significant difference in the relapse rate and a higher rate of infection in the infliximab group (71%) than in the placebo group (56%). Given the lack of any benefit observed in this study, infliximab is not recommended in the treatment of patients with giant cell arteritis.³⁵

Aspirin is recommended

Daily low-dose aspirin therapy has been shown in several studies to be effective in preventing ischemic complications of giant cell arteritis, including stroke and visual loss. It is currently recommended that all patients with giant cell arteritis without a major contraindication take aspirin 81 mg daily.^36–38

A 50-year-old woman presents to the emergency department because of repeated episodes of vomiting over the past 12 hours. She reports eight episodes of non-bloody, nonbilious emesis associated with palpitations and feelings of anxiety, but with no fever or diarrhea. She has not traveled recently and does not have any sick contacts.

She reports that she never had health problems until 6 months ago, when she began having panic attacks that woke her from sleep. The episodes first occurred once or twice per week, usually at night, and involved palpitations and feelings of anxiety that lasted 2 to 4 hours, but no other associated symptoms. After a month, the episodes began to occur more regularly during the day and were accompanied by a pounding headache that began in the back of her neck and extended up and over her head. Her primary care physician prescribed sertraline (Zoloft) and referred her to a neurologist to evaluate the headaches. The neurologic workup included brain magnetic resonance imaging and electroencephalography, both of which were normal.

After 8 weeks on sertraline, the episodes continued to increase in frequency and severity, and her physician switched her to paroxetine (Paxil) and added lorazepam (Ativan), which did not improve her symptoms. Over the past 2 months, during which time she has not been taking any medications, the episodes began to involve nausea and, more recently, vomiting, with episodes occurring as often as once or twice daily, and with intermittent symptom-free days. None of the prior episodes was accompanied by symptoms as severe as those she is currently experiencing.

She is otherwise healthy with no chronic diseases. Her surgical history includes resection of an angiolipoma from her right arm and dilation and curettage for endometrial polyps. She has no personal or family history of psychiatric illness.

PHYSICAL EXAMINATION

The patient is slender and tremulous but does not appear diaphoretic. Her blood pressure is 176/92 mm Hg, pulse 98, temperature 36.5°C (97.7°F), and respiratory rate 20 per minute. Oxygen saturation by pulse oximetry is 98% on room air. She has dry mucus membranes and orthostatic hypotension, but her physical examination is otherwise normal. Electrocardiography (ECG) shows a normal sinus rhythm with a prolonged QTc of 571 ms and peaked P and T waves.

LABORATORY VALUES

• Hemoglobin 15.6 g/dL (reference range 11.5–15.5)
• Hematocrit 47.2% (36.0–46.0)
• Platelet count 448 × 10⁹/L (150–400)
• White cell count 18.65 × 10⁹/L (3.70–11.00)
• Potassium 2.5 mmol/L (3.5–4.0)
• Chloride 97 mmol/L (98–110)
• Bicarbonate 21 mmol/L (23–32)
• Anion gap 20 mmol/L (0–15)
• Glucose 233 mg/dL (65–100).

Sodium, blood urea nitrogen, and creatinine levels are all within normal limits. Urinalysis suggests a urinary tract infection.

IS THIS A PANIC ATTACK?

1. Which of the following is not characteristic of a panic attack?

• Nausea and vomiting
• Onset during sleep
• Palpitations
• Chest pain or discomfort
• Headache
• Trembling or shaking

According to the Diagnostic and Statistical Manual of Mental Disorders (Fourth Edition) (DSM-IV), the diagnosis of panic attack requires the presence of intense fear or discomfort and four or more other symptoms that may come from any of six domains:

• Cardiovascular: palpitations, pounding heart, tachycardia, and chest pain or discomfort
• Autonomic: sweating, chills or hot flushes, and trembling or shaking
• Pulmonary: shortness of breath or a smothering sensation
• Neurologic: dizziness or light-headedness and paresthesias
• Gastrointestinal: choking and nausea or abdominal distress
• Psychological: compass derealization, depersonalization, and the fear of losing control or “going crazy.”¹

Two aspects of the patient’s history may be misinterpreted by those unfamiliar with the symptomatology of panic attack. First, although panic disorder carries an increased risk of many comorbidities, including migraine, headache is not typically associated with the panic attacks themselves.² Second, while not a part of the diagnostic criteria, sleep disturbances are common in patients with panic disorder, and 30% to 45% of patients with the disorder experience recurrent nocturnal panic attacks.³ Therefore, the correct answer is headache.

THE DIFFERENTIAL DIAGNOSIS

When considering a diagnosis of panic attack or panic disorder, the DSM-IV mandates that medical causes of the symptoms must be excluded. Common conditions causing a similar spectrum of symptoms include hyperthyroidism, caffeine and stimulant use or abuse, asthma, cardiac arrhythmias, alcohol withdrawal, and, more rarely, complex partial seizures and pheochromocytoma.^2,4 Many of these conditions can be ruled out by the history alone in a reliable patient.

Our patient’s electrocardiogram showed no evidence of ischemia or arrhythmias. Also, her recent negative neurologic workup makes seizure activity less likely.

Many of this patient’s laboratory abnormalities are easily explained by her repeated bouts of vomiting. Specifically, her elevated hemoglobin level and hematocrit are likely secondary to volume contraction, while hypochloremia is seen following losses of HCl with emesis. Typically, however, patients with vomiting have a hypochloremic metabolic alkalosis, and her low serum bicarbonate level is inconsistent with the history.

Three factors might be contributing to this patient’s hypokalemia. First, in a volume-depleted state, the cortical collecting tubules secrete potassium in exchange for increased sodium reabsorption in an attempt to correct volume status. Second, the alkalotic state caused by losses of acid with vomiting results in a transcellular shift of potassium ions into cells in exchange for hydrogen ions. Third, increased levels of epinephrine also cause a shift of potassium ions into cells.⁵ Potassium is not lost directly through nausea and vomiting.

A state of catecholamine excess, such as during a severe panic attack or in the presence of a catecholamine-secreting tumor, could explain many of her abnormalities. In addition to causing hypokalemia, epinephrine has a gluconeogenic effect, whereas norepinephrine inhibits insulin release, providing a potential explanation for hyperglycemia in a patient with no risk factors for diabetes. Finally, catecholamine excess contributes to lactic acidosis, which could help to explain the low serum bicarbonate level and the elevated anion gap, but unless we take arterial blood gas measurements, the patient’s acid-base status cannot be determined.

While panic attacks do stimulate the sympathetic nervous system, certain elements of her history raise the clinical suspicion for another process. First, the severity of the electrolyte abnormalities is suspicious. Second, a typical panic attack peaks at 10 minutes and begins to subside, whereas this woman’s symptoms have persisted for 12 hours. Finally, the clinical history, in particular the prominence of headaches associated with the symptoms, is inconsistent with classic panic attack. Consequently, an alternative diagnosis, such as pheochromocytoma, deserves more careful evaluation.

Whenever laboratory results do not fit with the clinical scenario or patient, however, one final possibility should always be considered—laboratory error. Errors can be preanalytical (eg, patient misidentification), analytical, or postanalytical. In aggregate, the frequency of errors in laboratory results is 1 in 214 to 8,316.⁶ Given that even the more conservative estimates show an incidence higher than that of many of the rare diseases for which clinicians may be testing, laboratory error always deserves consideration.

COULD THIS BE PHEOCHROMOCYTOMA?

Pheochromocytoma is a neuroendocrine tumor most commonly arising from the chromaffin cells of the adrenal medulla. However, extra-adrenal pheochromocytoma, generally paraganglioma, accounts for 15% to 20% of these tumors. Although the condition is generally considered very rare, autopsy studies have demonstrated a prevalence of 0.05%, suggesting that many tumors are either missed or are not clinically significant.

The diagnosis is most often sought in hypertensive patients, a population in which pheochromocytoma has a prevalence of 0.1% to 0.6%.⁷

2. What is the most common presenting symptom of pheochromocytoma?

• Paroxysmal hypertension
• Sustained hypertension
• Nausea
• Cardiomyopathy
• Headache
• Hemorrhagic shock
• Psychological symptoms such as anxiety or panic

Although 50% to 60% of patients with pheochromocytoma have sustained hypertension, it may be absent in patients with primarily epinephrine-secreting tumors or large tumors that degrade catecholamines, leading to normal or low blood pressure.

Cardiomyopathy is a rare consequence of untreated pheochromocytoma, caused by the effects of excess circulating catecholamines over a long period of time.⁸ As seen in this patient, a prolonged QTc on ECG associated with elevated levels of norepinephrine and normetanephrine may be the only red flag.⁹

Pheochromocytoma is typically an extremely well-vascularized tumor, and rupture or hemorrhage is a rare but often fatal complication.

IMPORTANT FAMILY HISTORY

The classic “rule of 10s” suggests that 10% of pheochromocytomas are hereditary, but in fact the number may be higher. In a large cohort of patients with apparently sporadic pheochromocytoma, 25% were found to have germ-line mutations.¹⁰ This finding highlights the importance not only of obtaining a thorough family history, but also of genetic testing and counseling once the diagnosis has been made.

3. Which hereditary syndrome is not associated with pheochromocytoma?

• Von Hippel-Lindau syndrome
• Neurofibromatosis type 1
• Neurofibromatosis type 2
• Multiple endocrine neoplasia type 2
• Paraganglioma syndromes

Germ-line mutations in five genes related to three hereditary syndromes (von Hippel-Lindau, neurofibromatosis type 1, and multiple endocrine neoplasia type 2) and in two genes related to paraganglioma syndromes are known to be associated with pheochromocytoma.⁷

Von Hippel-Lindau syndrome

Von Hippel-Lindau syndrome affects 1 in 36,000 live births. It is caused by a mutation of the von Hippel-Lindau gene on chromosome 3, and 10% to 20% of patients with the syndrome have pheochromocytoma. Other associated problems include renal clear-cell carcinomas and cysts, central nervous system and retinal hemangioblastomas, pancreatic tumors and cysts, endolymphatic tumors, and epididymal cysts.

Neurofibromatosis type 1

Neurofibromatosis type 1 affects 1 in 2,500 to 3,000 individuals and is caused by a mutation of the neurofibromatosis type 1 gene on chromosome 17. The disease is diagnosed by the presence of café-au-lait macules, axillary or inguinal freckling (or both), dermal or plexiform neurofibromas, Lisch nodules, or osseous lesions, but the condition is associated with many other pathologic findings, including optic pathway gliomas, cardiovascular abnormalities, and, in up to 5.7% of patients, pheochromocytoma.¹¹

Neurofibromatosis type 2

Neurofibromatosis type 2 affects 1 in 25,000 live births and is caused by a mutation of the neurofibromatosis type 2 gene on chromosome 22. Patients often develop nervous system tumors, ophthalmologic pathology, and cutaneous lesions, but the condition is not associated with pheochromocytoma.¹²

Multiple endocrine neoplasia type 2

Multiple endocrine neoplasia type 2 affects 1 in 35,000 individuals and is caused by an activating mutation of the RET proto-oncogene on chromosome 21. The syndrome is most worrisome because of the 95% lifetime risk of medullary thyroid carcinoma in affected patients, but it is also associated with a 50% risk of pheochromocytoma and a 20% to 30% risk of primary hyperparathyroidism. Pheochromocytoma is the presenting clinical problem in 10% to 30% of patients.¹³

Paraganglioma syndromes

Paraganglioma syndromes are caused by mutations in the three genes encoding subunits of the succinate dehydrogenase enzyme. These mutations affect 1 in 30,000 to 100,000 individuals and incur a 70% lifetime risk of developing pheochromocytoma or paraganglioma.¹⁴

TESTING FOR AND MANAGING PHEOCHROMOCYTOMA

The consequences of untreated pheochromocytoma are potentially devastating and include progression to metastatic disease, hypertensive crises, cardiomyopathy, and adrenal hemorrhage. Nevertheless, the average patient goes 3 years before receiving the correct diagnosis.⁷ Consequently, heightened suspicion and tests with both high sensitivity and specificity are needed.

4. Which test for pheochromocytoma has the highest sensitivity?

• Plasma free metanephrines
• Plasma catecholamines
• Urine total metanephrines
• Urine fractionated metanephrines
• Urine catecholamines
• Urine vanillylmandelic acid

Some researchers have also examined plasma total metanephrines and found that any one of these three biochemical markers at a value two times greater than the upper limit of normal provides specificity of around 95%.¹⁶

Further laboratory tests in our patient

• Serum dopamine 70 pg/mL (reference range 0–20)
• Norepinephrine 2,018 pg/mL (80–520)
• Epinephrine 2,479 pg/mL (10–200)
• Free normetanephrine 12 pg/mL (< 0.9)
• Free metanephrine 17.8 pg/mL (< 0.5).

VALUE OF IMAGING STUDIES

Computed tomography has a sensitivity of up to 95% for detecting adrenal tumors and is able to detect tumors larger than 0.5 cm, but its specificity may be as low as 50%.¹⁷ Studies utilizing modern imaging equipment report a prevalence of adrenal incidentaloma of 4%, of which only 1.5% to 11% are pheochromocytoma.¹⁸ Thus, while the simultaneous occurrence of pheochromocytoma-like symptoms and an incidentaloma is not common, the potential for unnecessary surgery precludes diagnosis and treatment based on symptoms and imaging alone.

Magnetic resonance imaging has similar sensitivity and specificity but can better characterize the tumor’s blood supply and relationship to other structures.

Iodine 131 metaiodobenzylguanidine (MIBG) scanning is a physiologic study that uses a radiolabeled amine. Since it can identify pheochromocytoma regardless of location, MIBG scanning is typically used when pheochromocytoma is diagnosed by biochemical testing but CT and MRI fail to locate the lesion, or as a follow-up test in patients in whom recurrence or metastasis is suspected or documented.

The specificity of MIBG scanning is 95% to 100%, but the need to protect the thyroid from ablation and the potential need to repeat scans for up to 72 hours make it a poor choice for the initial evaluation.¹⁷

5. What is the next best step in our patient’s management?

• Treat her hypertension with a beta-blocker
• Begin a course of alpha-blockade
• Urgent surgery
• Observation

Because of the high concentration of circulating catecholamines and the instability of the tumor to physical manipulation, appropriate medical management before surgical resection is of paramount importance.

Beta-blockade can lead to malignant hypertension due to the unopposed alpha stimulation and must not be begun until alpha-blockade has been started. The standard of care is to give an alpha-blocker or calcium channel blocker 10 to 14 days before surgery. Typically, oral phenoxybenzamine (Dibenzyline) 10 mg twice daily is started and titrated upward daily by 10 to 20 mg until a target seated blood pressure of 120/80 mm Hg is obtained. Selective alpha-1 blockers such as prazosin (Minipress) and terazosin (Hytrin) have also been used and have the benefit of a preserved alpha-2 catecholamine reuptake mechanism.¹⁷

After several days, a beta-blocker may be added, particularly for patients with arrhythmias.⁷ In patients with refractory hypertension, metyrosine (Demser) can be useful.

During surgery, the patient’s hemodynamic stability and glucose levels can fluctuate rapidly from sudden releases of catecholamines during manipulation of the tumor, as well as from the sudden loss of catecholamines after ligation of draining vessels. Advances in medical care have reduced the perioperative death rate from 50% to less than 3%.^7,19

CASE CONCLUSION AND FOLLOW-UP

Two months after her initial presentation, the patient underwent open surgery and had the mass removed without complications. She reports that the “panic attacks” have ceased completely.

The recurrence rate of pheochromocytoma is 13% in patients with sporadic disease and 33% in patients with familial syndromes. The overall recurrence rate with long-term follow-up is 17%, half of recurrences being malignant disease. All patients should therefore be followed in the clinic annually for at least 10 years to identify and treat recurrences early,⁷ and many experts recommend lifelong follow-up, even for patients without hereditary syndromes.¹⁷

Nearly every diagnosis in the DSM-IV includes the caveat that medical causes of disease must be excluded before psychiatric labels can be applied. Although panic disorder and panic attack are far more common than pheochromocytoma, just as essential hypertension is far more common than pheochromocytoma, physicians need to remember that pheochromocytoma can cause symptoms common to both illnesses. Thus, while rare conditions are rare, atypical presentations of common conditions may deserve a second glance.

A 50-year-old woman presents to the emergency department because of repeated episodes of vomiting over the past 12 hours. She reports eight episodes of non-bloody, nonbilious emesis associated with palpitations and feelings of anxiety, but with no fever or diarrhea. She has not traveled recently and does not have any sick contacts.

She reports that she never had health problems until 6 months ago, when she began having panic attacks that woke her from sleep. The episodes first occurred once or twice per week, usually at night, and involved palpitations and feelings of anxiety that lasted 2 to 4 hours, but no other associated symptoms. After a month, the episodes began to occur more regularly during the day and were accompanied by a pounding headache that began in the back of her neck and extended up and over her head. Her primary care physician prescribed sertraline (Zoloft) and referred her to a neurologist to evaluate the headaches. The neurologic workup included brain magnetic resonance imaging and electroencephalography, both of which were normal.

After 8 weeks on sertraline, the episodes continued to increase in frequency and severity, and her physician switched her to paroxetine (Paxil) and added lorazepam (Ativan), which did not improve her symptoms. Over the past 2 months, during which time she has not been taking any medications, the episodes began to involve nausea and, more recently, vomiting, with episodes occurring as often as once or twice daily, and with intermittent symptom-free days. None of the prior episodes was accompanied by symptoms as severe as those she is currently experiencing.

She is otherwise healthy with no chronic diseases. Her surgical history includes resection of an angiolipoma from her right arm and dilation and curettage for endometrial polyps. She has no personal or family history of psychiatric illness.

PHYSICAL EXAMINATION

The patient is slender and tremulous but does not appear diaphoretic. Her blood pressure is 176/92 mm Hg, pulse 98, temperature 36.5°C (97.7°F), and respiratory rate 20 per minute. Oxygen saturation by pulse oximetry is 98% on room air. She has dry mucus membranes and orthostatic hypotension, but her physical examination is otherwise normal. Electrocardiography (ECG) shows a normal sinus rhythm with a prolonged QTc of 571 ms and peaked P and T waves.

LABORATORY VALUES

• Hemoglobin 15.6 g/dL (reference range 11.5–15.5)
• Hematocrit 47.2% (36.0–46.0)
• Platelet count 448 × 10⁹/L (150–400)
• White cell count 18.65 × 10⁹/L (3.70–11.00)
• Potassium 2.5 mmol/L (3.5–4.0)
• Chloride 97 mmol/L (98–110)
• Bicarbonate 21 mmol/L (23–32)
• Anion gap 20 mmol/L (0–15)
• Glucose 233 mg/dL (65–100).

Sodium, blood urea nitrogen, and creatinine levels are all within normal limits. Urinalysis suggests a urinary tract infection.

IS THIS A PANIC ATTACK?

1. Which of the following is not characteristic of a panic attack?

• Nausea and vomiting
• Onset during sleep
• Palpitations
• Chest pain or discomfort
• Headache
• Trembling or shaking

According to the Diagnostic and Statistical Manual of Mental Disorders (Fourth Edition) (DSM-IV), the diagnosis of panic attack requires the presence of intense fear or discomfort and four or more other symptoms that may come from any of six domains:

• Cardiovascular: palpitations, pounding heart, tachycardia, and chest pain or discomfort
• Autonomic: sweating, chills or hot flushes, and trembling or shaking
• Pulmonary: shortness of breath or a smothering sensation
• Neurologic: dizziness or light-headedness and paresthesias
• Gastrointestinal: choking and nausea or abdominal distress
• Psychological: compass derealization, depersonalization, and the fear of losing control or “going crazy.”¹

Two aspects of the patient’s history may be misinterpreted by those unfamiliar with the symptomatology of panic attack. First, although panic disorder carries an increased risk of many comorbidities, including migraine, headache is not typically associated with the panic attacks themselves.² Second, while not a part of the diagnostic criteria, sleep disturbances are common in patients with panic disorder, and 30% to 45% of patients with the disorder experience recurrent nocturnal panic attacks.³ Therefore, the correct answer is headache.

THE DIFFERENTIAL DIAGNOSIS

When considering a diagnosis of panic attack or panic disorder, the DSM-IV mandates that medical causes of the symptoms must be excluded. Common conditions causing a similar spectrum of symptoms include hyperthyroidism, caffeine and stimulant use or abuse, asthma, cardiac arrhythmias, alcohol withdrawal, and, more rarely, complex partial seizures and pheochromocytoma.^2,4 Many of these conditions can be ruled out by the history alone in a reliable patient.

Our patient’s electrocardiogram showed no evidence of ischemia or arrhythmias. Also, her recent negative neurologic workup makes seizure activity less likely.

Many of this patient’s laboratory abnormalities are easily explained by her repeated bouts of vomiting. Specifically, her elevated hemoglobin level and hematocrit are likely secondary to volume contraction, while hypochloremia is seen following losses of HCl with emesis. Typically, however, patients with vomiting have a hypochloremic metabolic alkalosis, and her low serum bicarbonate level is inconsistent with the history.

Three factors might be contributing to this patient’s hypokalemia. First, in a volume-depleted state, the cortical collecting tubules secrete potassium in exchange for increased sodium reabsorption in an attempt to correct volume status. Second, the alkalotic state caused by losses of acid with vomiting results in a transcellular shift of potassium ions into cells in exchange for hydrogen ions. Third, increased levels of epinephrine also cause a shift of potassium ions into cells.⁵ Potassium is not lost directly through nausea and vomiting.

A state of catecholamine excess, such as during a severe panic attack or in the presence of a catecholamine-secreting tumor, could explain many of her abnormalities. In addition to causing hypokalemia, epinephrine has a gluconeogenic effect, whereas norepinephrine inhibits insulin release, providing a potential explanation for hyperglycemia in a patient with no risk factors for diabetes. Finally, catecholamine excess contributes to lactic acidosis, which could help to explain the low serum bicarbonate level and the elevated anion gap, but unless we take arterial blood gas measurements, the patient’s acid-base status cannot be determined.

While panic attacks do stimulate the sympathetic nervous system, certain elements of her history raise the clinical suspicion for another process. First, the severity of the electrolyte abnormalities is suspicious. Second, a typical panic attack peaks at 10 minutes and begins to subside, whereas this woman’s symptoms have persisted for 12 hours. Finally, the clinical history, in particular the prominence of headaches associated with the symptoms, is inconsistent with classic panic attack. Consequently, an alternative diagnosis, such as pheochromocytoma, deserves more careful evaluation.

Whenever laboratory results do not fit with the clinical scenario or patient, however, one final possibility should always be considered—laboratory error. Errors can be preanalytical (eg, patient misidentification), analytical, or postanalytical. In aggregate, the frequency of errors in laboratory results is 1 in 214 to 8,316.⁶ Given that even the more conservative estimates show an incidence higher than that of many of the rare diseases for which clinicians may be testing, laboratory error always deserves consideration.

COULD THIS BE PHEOCHROMOCYTOMA?

Pheochromocytoma is a neuroendocrine tumor most commonly arising from the chromaffin cells of the adrenal medulla. However, extra-adrenal pheochromocytoma, generally paraganglioma, accounts for 15% to 20% of these tumors. Although the condition is generally considered very rare, autopsy studies have demonstrated a prevalence of 0.05%, suggesting that many tumors are either missed or are not clinically significant.

The diagnosis is most often sought in hypertensive patients, a population in which pheochromocytoma has a prevalence of 0.1% to 0.6%.⁷

2. What is the most common presenting symptom of pheochromocytoma?

• Paroxysmal hypertension
• Sustained hypertension
• Nausea
• Cardiomyopathy
• Headache
• Hemorrhagic shock
• Psychological symptoms such as anxiety or panic

Although 50% to 60% of patients with pheochromocytoma have sustained hypertension, it may be absent in patients with primarily epinephrine-secreting tumors or large tumors that degrade catecholamines, leading to normal or low blood pressure.

Cardiomyopathy is a rare consequence of untreated pheochromocytoma, caused by the effects of excess circulating catecholamines over a long period of time.⁸ As seen in this patient, a prolonged QTc on ECG associated with elevated levels of norepinephrine and normetanephrine may be the only red flag.⁹

Pheochromocytoma is typically an extremely well-vascularized tumor, and rupture or hemorrhage is a rare but often fatal complication.

IMPORTANT FAMILY HISTORY

The classic “rule of 10s” suggests that 10% of pheochromocytomas are hereditary, but in fact the number may be higher. In a large cohort of patients with apparently sporadic pheochromocytoma, 25% were found to have germ-line mutations.¹⁰ This finding highlights the importance not only of obtaining a thorough family history, but also of genetic testing and counseling once the diagnosis has been made.

3. Which hereditary syndrome is not associated with pheochromocytoma?

• Von Hippel-Lindau syndrome
• Neurofibromatosis type 1
• Neurofibromatosis type 2
• Multiple endocrine neoplasia type 2
• Paraganglioma syndromes

Germ-line mutations in five genes related to three hereditary syndromes (von Hippel-Lindau, neurofibromatosis type 1, and multiple endocrine neoplasia type 2) and in two genes related to paraganglioma syndromes are known to be associated with pheochromocytoma.⁷

Von Hippel-Lindau syndrome

Von Hippel-Lindau syndrome affects 1 in 36,000 live births. It is caused by a mutation of the von Hippel-Lindau gene on chromosome 3, and 10% to 20% of patients with the syndrome have pheochromocytoma. Other associated problems include renal clear-cell carcinomas and cysts, central nervous system and retinal hemangioblastomas, pancreatic tumors and cysts, endolymphatic tumors, and epididymal cysts.

Neurofibromatosis type 1

Neurofibromatosis type 1 affects 1 in 2,500 to 3,000 individuals and is caused by a mutation of the neurofibromatosis type 1 gene on chromosome 17. The disease is diagnosed by the presence of café-au-lait macules, axillary or inguinal freckling (or both), dermal or plexiform neurofibromas, Lisch nodules, or osseous lesions, but the condition is associated with many other pathologic findings, including optic pathway gliomas, cardiovascular abnormalities, and, in up to 5.7% of patients, pheochromocytoma.¹¹

Neurofibromatosis type 2

Neurofibromatosis type 2 affects 1 in 25,000 live births and is caused by a mutation of the neurofibromatosis type 2 gene on chromosome 22. Patients often develop nervous system tumors, ophthalmologic pathology, and cutaneous lesions, but the condition is not associated with pheochromocytoma.¹²

Multiple endocrine neoplasia type 2

Multiple endocrine neoplasia type 2 affects 1 in 35,000 individuals and is caused by an activating mutation of the RET proto-oncogene on chromosome 21. The syndrome is most worrisome because of the 95% lifetime risk of medullary thyroid carcinoma in affected patients, but it is also associated with a 50% risk of pheochromocytoma and a 20% to 30% risk of primary hyperparathyroidism. Pheochromocytoma is the presenting clinical problem in 10% to 30% of patients.¹³

Paraganglioma syndromes

Paraganglioma syndromes are caused by mutations in the three genes encoding subunits of the succinate dehydrogenase enzyme. These mutations affect 1 in 30,000 to 100,000 individuals and incur a 70% lifetime risk of developing pheochromocytoma or paraganglioma.¹⁴

TESTING FOR AND MANAGING PHEOCHROMOCYTOMA

The consequences of untreated pheochromocytoma are potentially devastating and include progression to metastatic disease, hypertensive crises, cardiomyopathy, and adrenal hemorrhage. Nevertheless, the average patient goes 3 years before receiving the correct diagnosis.⁷ Consequently, heightened suspicion and tests with both high sensitivity and specificity are needed.

4. Which test for pheochromocytoma has the highest sensitivity?

• Plasma free metanephrines
• Plasma catecholamines
• Urine total metanephrines
• Urine fractionated metanephrines
• Urine catecholamines
• Urine vanillylmandelic acid

Some researchers have also examined plasma total metanephrines and found that any one of these three biochemical markers at a value two times greater than the upper limit of normal provides specificity of around 95%.¹⁶

Further laboratory tests in our patient

• Serum dopamine 70 pg/mL (reference range 0–20)
• Norepinephrine 2,018 pg/mL (80–520)
• Epinephrine 2,479 pg/mL (10–200)
• Free normetanephrine 12 pg/mL (< 0.9)
• Free metanephrine 17.8 pg/mL (< 0.5).

VALUE OF IMAGING STUDIES

Computed tomography has a sensitivity of up to 95% for detecting adrenal tumors and is able to detect tumors larger than 0.5 cm, but its specificity may be as low as 50%.¹⁷ Studies utilizing modern imaging equipment report a prevalence of adrenal incidentaloma of 4%, of which only 1.5% to 11% are pheochromocytoma.¹⁸ Thus, while the simultaneous occurrence of pheochromocytoma-like symptoms and an incidentaloma is not common, the potential for unnecessary surgery precludes diagnosis and treatment based on symptoms and imaging alone.

Magnetic resonance imaging has similar sensitivity and specificity but can better characterize the tumor’s blood supply and relationship to other structures.

Iodine 131 metaiodobenzylguanidine (MIBG) scanning is a physiologic study that uses a radiolabeled amine. Since it can identify pheochromocytoma regardless of location, MIBG scanning is typically used when pheochromocytoma is diagnosed by biochemical testing but CT and MRI fail to locate the lesion, or as a follow-up test in patients in whom recurrence or metastasis is suspected or documented.

The specificity of MIBG scanning is 95% to 100%, but the need to protect the thyroid from ablation and the potential need to repeat scans for up to 72 hours make it a poor choice for the initial evaluation.¹⁷

5. What is the next best step in our patient’s management?

• Treat her hypertension with a beta-blocker
• Begin a course of alpha-blockade
• Urgent surgery
• Observation

Because of the high concentration of circulating catecholamines and the instability of the tumor to physical manipulation, appropriate medical management before surgical resection is of paramount importance.

Beta-blockade can lead to malignant hypertension due to the unopposed alpha stimulation and must not be begun until alpha-blockade has been started. The standard of care is to give an alpha-blocker or calcium channel blocker 10 to 14 days before surgery. Typically, oral phenoxybenzamine (Dibenzyline) 10 mg twice daily is started and titrated upward daily by 10 to 20 mg until a target seated blood pressure of 120/80 mm Hg is obtained. Selective alpha-1 blockers such as prazosin (Minipress) and terazosin (Hytrin) have also been used and have the benefit of a preserved alpha-2 catecholamine reuptake mechanism.¹⁷

After several days, a beta-blocker may be added, particularly for patients with arrhythmias.⁷ In patients with refractory hypertension, metyrosine (Demser) can be useful.

During surgery, the patient’s hemodynamic stability and glucose levels can fluctuate rapidly from sudden releases of catecholamines during manipulation of the tumor, as well as from the sudden loss of catecholamines after ligation of draining vessels. Advances in medical care have reduced the perioperative death rate from 50% to less than 3%.^7,19

CASE CONCLUSION AND FOLLOW-UP

Two months after her initial presentation, the patient underwent open surgery and had the mass removed without complications. She reports that the “panic attacks” have ceased completely.

The recurrence rate of pheochromocytoma is 13% in patients with sporadic disease and 33% in patients with familial syndromes. The overall recurrence rate with long-term follow-up is 17%, half of recurrences being malignant disease. All patients should therefore be followed in the clinic annually for at least 10 years to identify and treat recurrences early,⁷ and many experts recommend lifelong follow-up, even for patients without hereditary syndromes.¹⁷

Nearly every diagnosis in the DSM-IV includes the caveat that medical causes of disease must be excluded before psychiatric labels can be applied. Although panic disorder and panic attack are far more common than pheochromocytoma, just as essential hypertension is far more common than pheochromocytoma, physicians need to remember that pheochromocytoma can cause symptoms common to both illnesses. Thus, while rare conditions are rare, atypical presentations of common conditions may deserve a second glance.

A 50-year-old woman presents to the emergency department because of repeated episodes of vomiting over the past 12 hours. She reports eight episodes of non-bloody, nonbilious emesis associated with palpitations and feelings of anxiety, but with no fever or diarrhea. She has not traveled recently and does not have any sick contacts.

She reports that she never had health problems until 6 months ago, when she began having panic attacks that woke her from sleep. The episodes first occurred once or twice per week, usually at night, and involved palpitations and feelings of anxiety that lasted 2 to 4 hours, but no other associated symptoms. After a month, the episodes began to occur more regularly during the day and were accompanied by a pounding headache that began in the back of her neck and extended up and over her head. Her primary care physician prescribed sertraline (Zoloft) and referred her to a neurologist to evaluate the headaches. The neurologic workup included brain magnetic resonance imaging and electroencephalography, both of which were normal.

After 8 weeks on sertraline, the episodes continued to increase in frequency and severity, and her physician switched her to paroxetine (Paxil) and added lorazepam (Ativan), which did not improve her symptoms. Over the past 2 months, during which time she has not been taking any medications, the episodes began to involve nausea and, more recently, vomiting, with episodes occurring as often as once or twice daily, and with intermittent symptom-free days. None of the prior episodes was accompanied by symptoms as severe as those she is currently experiencing.

She is otherwise healthy with no chronic diseases. Her surgical history includes resection of an angiolipoma from her right arm and dilation and curettage for endometrial polyps. She has no personal or family history of psychiatric illness.

PHYSICAL EXAMINATION

The patient is slender and tremulous but does not appear diaphoretic. Her blood pressure is 176/92 mm Hg, pulse 98, temperature 36.5°C (97.7°F), and respiratory rate 20 per minute. Oxygen saturation by pulse oximetry is 98% on room air. She has dry mucus membranes and orthostatic hypotension, but her physical examination is otherwise normal. Electrocardiography (ECG) shows a normal sinus rhythm with a prolonged QTc of 571 ms and peaked P and T waves.

LABORATORY VALUES

• Hemoglobin 15.6 g/dL (reference range 11.5–15.5)
• Hematocrit 47.2% (36.0–46.0)
• Platelet count 448 × 10⁹/L (150–400)
• White cell count 18.65 × 10⁹/L (3.70–11.00)
• Potassium 2.5 mmol/L (3.5–4.0)
• Chloride 97 mmol/L (98–110)
• Bicarbonate 21 mmol/L (23–32)
• Anion gap 20 mmol/L (0–15)
• Glucose 233 mg/dL (65–100).

Sodium, blood urea nitrogen, and creatinine levels are all within normal limits. Urinalysis suggests a urinary tract infection.

IS THIS A PANIC ATTACK?

1. Which of the following is not characteristic of a panic attack?

• Nausea and vomiting
• Onset during sleep
• Palpitations
• Chest pain or discomfort
• Headache
• Trembling or shaking

According to the Diagnostic and Statistical Manual of Mental Disorders (Fourth Edition) (DSM-IV), the diagnosis of panic attack requires the presence of intense fear or discomfort and four or more other symptoms that may come from any of six domains:

• Cardiovascular: palpitations, pounding heart, tachycardia, and chest pain or discomfort
• Autonomic: sweating, chills or hot flushes, and trembling or shaking
• Pulmonary: shortness of breath or a smothering sensation
• Neurologic: dizziness or light-headedness and paresthesias
• Gastrointestinal: choking and nausea or abdominal distress
• Psychological: compass derealization, depersonalization, and the fear of losing control or “going crazy.”¹

Two aspects of the patient’s history may be misinterpreted by those unfamiliar with the symptomatology of panic attack. First, although panic disorder carries an increased risk of many comorbidities, including migraine, headache is not typically associated with the panic attacks themselves.² Second, while not a part of the diagnostic criteria, sleep disturbances are common in patients with panic disorder, and 30% to 45% of patients with the disorder experience recurrent nocturnal panic attacks.³ Therefore, the correct answer is headache.

THE DIFFERENTIAL DIAGNOSIS

When considering a diagnosis of panic attack or panic disorder, the DSM-IV mandates that medical causes of the symptoms must be excluded. Common conditions causing a similar spectrum of symptoms include hyperthyroidism, caffeine and stimulant use or abuse, asthma, cardiac arrhythmias, alcohol withdrawal, and, more rarely, complex partial seizures and pheochromocytoma.^2,4 Many of these conditions can be ruled out by the history alone in a reliable patient.

Our patient’s electrocardiogram showed no evidence of ischemia or arrhythmias. Also, her recent negative neurologic workup makes seizure activity less likely.

Many of this patient’s laboratory abnormalities are easily explained by her repeated bouts of vomiting. Specifically, her elevated hemoglobin level and hematocrit are likely secondary to volume contraction, while hypochloremia is seen following losses of HCl with emesis. Typically, however, patients with vomiting have a hypochloremic metabolic alkalosis, and her low serum bicarbonate level is inconsistent with the history.

Three factors might be contributing to this patient’s hypokalemia. First, in a volume-depleted state, the cortical collecting tubules secrete potassium in exchange for increased sodium reabsorption in an attempt to correct volume status. Second, the alkalotic state caused by losses of acid with vomiting results in a transcellular shift of potassium ions into cells in exchange for hydrogen ions. Third, increased levels of epinephrine also cause a shift of potassium ions into cells.⁵ Potassium is not lost directly through nausea and vomiting.

A state of catecholamine excess, such as during a severe panic attack or in the presence of a catecholamine-secreting tumor, could explain many of her abnormalities. In addition to causing hypokalemia, epinephrine has a gluconeogenic effect, whereas norepinephrine inhibits insulin release, providing a potential explanation for hyperglycemia in a patient with no risk factors for diabetes. Finally, catecholamine excess contributes to lactic acidosis, which could help to explain the low serum bicarbonate level and the elevated anion gap, but unless we take arterial blood gas measurements, the patient’s acid-base status cannot be determined.

While panic attacks do stimulate the sympathetic nervous system, certain elements of her history raise the clinical suspicion for another process. First, the severity of the electrolyte abnormalities is suspicious. Second, a typical panic attack peaks at 10 minutes and begins to subside, whereas this woman’s symptoms have persisted for 12 hours. Finally, the clinical history, in particular the prominence of headaches associated with the symptoms, is inconsistent with classic panic attack. Consequently, an alternative diagnosis, such as pheochromocytoma, deserves more careful evaluation.

Whenever laboratory results do not fit with the clinical scenario or patient, however, one final possibility should always be considered—laboratory error. Errors can be preanalytical (eg, patient misidentification), analytical, or postanalytical. In aggregate, the frequency of errors in laboratory results is 1 in 214 to 8,316.⁶ Given that even the more conservative estimates show an incidence higher than that of many of the rare diseases for which clinicians may be testing, laboratory error always deserves consideration.

COULD THIS BE PHEOCHROMOCYTOMA?

Pheochromocytoma is a neuroendocrine tumor most commonly arising from the chromaffin cells of the adrenal medulla. However, extra-adrenal pheochromocytoma, generally paraganglioma, accounts for 15% to 20% of these tumors. Although the condition is generally considered very rare, autopsy studies have demonstrated a prevalence of 0.05%, suggesting that many tumors are either missed or are not clinically significant.

The diagnosis is most often sought in hypertensive patients, a population in which pheochromocytoma has a prevalence of 0.1% to 0.6%.⁷

2. What is the most common presenting symptom of pheochromocytoma?

• Paroxysmal hypertension
• Sustained hypertension
• Nausea
• Cardiomyopathy
• Headache
• Hemorrhagic shock
• Psychological symptoms such as anxiety or panic

Although 50% to 60% of patients with pheochromocytoma have sustained hypertension, it may be absent in patients with primarily epinephrine-secreting tumors or large tumors that degrade catecholamines, leading to normal or low blood pressure.

Cardiomyopathy is a rare consequence of untreated pheochromocytoma, caused by the effects of excess circulating catecholamines over a long period of time.⁸ As seen in this patient, a prolonged QTc on ECG associated with elevated levels of norepinephrine and normetanephrine may be the only red flag.⁹

Pheochromocytoma is typically an extremely well-vascularized tumor, and rupture or hemorrhage is a rare but often fatal complication.

IMPORTANT FAMILY HISTORY

The classic “rule of 10s” suggests that 10% of pheochromocytomas are hereditary, but in fact the number may be higher. In a large cohort of patients with apparently sporadic pheochromocytoma, 25% were found to have germ-line mutations.¹⁰ This finding highlights the importance not only of obtaining a thorough family history, but also of genetic testing and counseling once the diagnosis has been made.

3. Which hereditary syndrome is not associated with pheochromocytoma?

• Von Hippel-Lindau syndrome
• Neurofibromatosis type 1
• Neurofibromatosis type 2
• Multiple endocrine neoplasia type 2
• Paraganglioma syndromes

Germ-line mutations in five genes related to three hereditary syndromes (von Hippel-Lindau, neurofibromatosis type 1, and multiple endocrine neoplasia type 2) and in two genes related to paraganglioma syndromes are known to be associated with pheochromocytoma.⁷

Von Hippel-Lindau syndrome

Von Hippel-Lindau syndrome affects 1 in 36,000 live births. It is caused by a mutation of the von Hippel-Lindau gene on chromosome 3, and 10% to 20% of patients with the syndrome have pheochromocytoma. Other associated problems include renal clear-cell carcinomas and cysts, central nervous system and retinal hemangioblastomas, pancreatic tumors and cysts, endolymphatic tumors, and epididymal cysts.

Neurofibromatosis type 1

Neurofibromatosis type 1 affects 1 in 2,500 to 3,000 individuals and is caused by a mutation of the neurofibromatosis type 1 gene on chromosome 17. The disease is diagnosed by the presence of café-au-lait macules, axillary or inguinal freckling (or both), dermal or plexiform neurofibromas, Lisch nodules, or osseous lesions, but the condition is associated with many other pathologic findings, including optic pathway gliomas, cardiovascular abnormalities, and, in up to 5.7% of patients, pheochromocytoma.¹¹

Neurofibromatosis type 2

Neurofibromatosis type 2 affects 1 in 25,000 live births and is caused by a mutation of the neurofibromatosis type 2 gene on chromosome 22. Patients often develop nervous system tumors, ophthalmologic pathology, and cutaneous lesions, but the condition is not associated with pheochromocytoma.¹²

Multiple endocrine neoplasia type 2

Multiple endocrine neoplasia type 2 affects 1 in 35,000 individuals and is caused by an activating mutation of the RET proto-oncogene on chromosome 21. The syndrome is most worrisome because of the 95% lifetime risk of medullary thyroid carcinoma in affected patients, but it is also associated with a 50% risk of pheochromocytoma and a 20% to 30% risk of primary hyperparathyroidism. Pheochromocytoma is the presenting clinical problem in 10% to 30% of patients.¹³

Paraganglioma syndromes

Paraganglioma syndromes are caused by mutations in the three genes encoding subunits of the succinate dehydrogenase enzyme. These mutations affect 1 in 30,000 to 100,000 individuals and incur a 70% lifetime risk of developing pheochromocytoma or paraganglioma.¹⁴

TESTING FOR AND MANAGING PHEOCHROMOCYTOMA

The consequences of untreated pheochromocytoma are potentially devastating and include progression to metastatic disease, hypertensive crises, cardiomyopathy, and adrenal hemorrhage. Nevertheless, the average patient goes 3 years before receiving the correct diagnosis.⁷ Consequently, heightened suspicion and tests with both high sensitivity and specificity are needed.

4. Which test for pheochromocytoma has the highest sensitivity?

• Plasma free metanephrines
• Plasma catecholamines
• Urine total metanephrines
• Urine fractionated metanephrines
• Urine catecholamines
• Urine vanillylmandelic acid

Some researchers have also examined plasma total metanephrines and found that any one of these three biochemical markers at a value two times greater than the upper limit of normal provides specificity of around 95%.¹⁶

Further laboratory tests in our patient

• Serum dopamine 70 pg/mL (reference range 0–20)
• Norepinephrine 2,018 pg/mL (80–520)
• Epinephrine 2,479 pg/mL (10–200)
• Free normetanephrine 12 pg/mL (< 0.9)
• Free metanephrine 17.8 pg/mL (< 0.5).

VALUE OF IMAGING STUDIES

Computed tomography has a sensitivity of up to 95% for detecting adrenal tumors and is able to detect tumors larger than 0.5 cm, but its specificity may be as low as 50%.¹⁷ Studies utilizing modern imaging equipment report a prevalence of adrenal incidentaloma of 4%, of which only 1.5% to 11% are pheochromocytoma.¹⁸ Thus, while the simultaneous occurrence of pheochromocytoma-like symptoms and an incidentaloma is not common, the potential for unnecessary surgery precludes diagnosis and treatment based on symptoms and imaging alone.

Magnetic resonance imaging has similar sensitivity and specificity but can better characterize the tumor’s blood supply and relationship to other structures.

Iodine 131 metaiodobenzylguanidine (MIBG) scanning is a physiologic study that uses a radiolabeled amine. Since it can identify pheochromocytoma regardless of location, MIBG scanning is typically used when pheochromocytoma is diagnosed by biochemical testing but CT and MRI fail to locate the lesion, or as a follow-up test in patients in whom recurrence or metastasis is suspected or documented.

The specificity of MIBG scanning is 95% to 100%, but the need to protect the thyroid from ablation and the potential need to repeat scans for up to 72 hours make it a poor choice for the initial evaluation.¹⁷

5. What is the next best step in our patient’s management?

• Treat her hypertension with a beta-blocker
• Begin a course of alpha-blockade
• Urgent surgery
• Observation

Because of the high concentration of circulating catecholamines and the instability of the tumor to physical manipulation, appropriate medical management before surgical resection is of paramount importance.

Beta-blockade can lead to malignant hypertension due to the unopposed alpha stimulation and must not be begun until alpha-blockade has been started. The standard of care is to give an alpha-blocker or calcium channel blocker 10 to 14 days before surgery. Typically, oral phenoxybenzamine (Dibenzyline) 10 mg twice daily is started and titrated upward daily by 10 to 20 mg until a target seated blood pressure of 120/80 mm Hg is obtained. Selective alpha-1 blockers such as prazosin (Minipress) and terazosin (Hytrin) have also been used and have the benefit of a preserved alpha-2 catecholamine reuptake mechanism.¹⁷

After several days, a beta-blocker may be added, particularly for patients with arrhythmias.⁷ In patients with refractory hypertension, metyrosine (Demser) can be useful.

During surgery, the patient’s hemodynamic stability and glucose levels can fluctuate rapidly from sudden releases of catecholamines during manipulation of the tumor, as well as from the sudden loss of catecholamines after ligation of draining vessels. Advances in medical care have reduced the perioperative death rate from 50% to less than 3%.^7,19

CASE CONCLUSION AND FOLLOW-UP

Two months after her initial presentation, the patient underwent open surgery and had the mass removed without complications. She reports that the “panic attacks” have ceased completely.

The recurrence rate of pheochromocytoma is 13% in patients with sporadic disease and 33% in patients with familial syndromes. The overall recurrence rate with long-term follow-up is 17%, half of recurrences being malignant disease. All patients should therefore be followed in the clinic annually for at least 10 years to identify and treat recurrences early,⁷ and many experts recommend lifelong follow-up, even for patients without hereditary syndromes.¹⁷

Nearly every diagnosis in the DSM-IV includes the caveat that medical causes of disease must be excluded before psychiatric labels can be applied. Although panic disorder and panic attack are far more common than pheochromocytoma, just as essential hypertension is far more common than pheochromocytoma, physicians need to remember that pheochromocytoma can cause symptoms common to both illnesses. Thus, while rare conditions are rare, atypical presentations of common conditions may deserve a second glance.

Q: What is the most likely diagnosis?

• Leukemia cutis
• Drug reaction
• Sweet syndrome
• Erythema multiforme
• Urticaria

A: The correct answer is leukemia cutis, defined as a cutaneous infiltration by neoplastic leukocytes.¹ When the leukocytes are primarily granulocytic precursors, the terms myeloid sarcoma, granulocytic sarcoma, chloroma, and primary extramedullary leukemia have been used.² The term monoblastic sarcoma has been used when the cutaneous infiltrate is composed of neoplastic monocytic precursors.²

Leukemia cutis most commonly manifests as erythematous papules and nodules, single or multiple, of varying sizes, and afflicting one or more body sites; it typically is asymptomatic.³ It occurs in 10% to 15% of patients with acute myeloid leukemia⁴ and is itself a poor prognostic sign.⁵ The cutaneous changes may pre-date the hematologic manifestations and may even herald a relapse.⁶

In patients presenting with extramedullary leukemia and no bone marrow or blood involvement, the importance of preemptive chemotherapy for acute myelogenous leukemia has recently been emphasized.⁷

CASE CONTINUED

The patient undergoes further testing with bone marrow aspiration and trephination, which are diagnostic of acute myeloid leukemia with myelodysplastic changes: the studies reveal a clear excess of myeloblasts (accounting for 40% to 50% of nucleated cells) and clearly dysplastic erythropoiesis and myelopoiesis. Bone marrow cytogenetic analysis reveals a complex abnormal female karyotype with multiple numerical and structural abnormalities, in particular deletion of the long arm of chromosome 5, suggestive of a poor prognosis.

Skin biopsy reveals a normal epidermis but dermal perivascular involvement with a reactive T-lymphocyte infiltrate associated with immature myeloid elements, characterized by positive staining to myeloperoxidase, in keeping with leukemia cutis. It should be noted that histopathologic confirmation of leukemia cutis can be challenging, as the condition can adopt a variety of patterns, and clinicopathologic correlation is often warranted.

The patient is treated with a cycle of cytarabine-based chemotherapy, and her skin eruption transiently improves. However, her clinical condition subsequently deteriorates; she has a relapse of leukemia, with the rash returning more florid and angry-looking than previously. She is subsequently managed palliatively and passes away 3 weeks later.

THE OTHER DIAGNOSTIC CHOICES

Sweet syndrome or acute febrile neutrophilic dermatosis is often seen in association with hematologic malignancies, but the lesions are typically tender, and histopathology reveals an intense dermal neutrophilic infiltrate.⁶

Erythema multiforme is associated with malignancy, but its characteristic concentric “target” lesions are typically acral and symmetrical in their distribution; their histopathology is inflammatory.⁸

The patient had not been on any regular medications and her rash could not have been medication-induced.

Urticaria presents with pruritic evanescent wheals, which rarely last more than 12 hours.⁹ Our patient had a fixed and entirely asymptomatic rash, which in addition did not have the histopathologic features of urticaria—namely, dermal edema involved with an infiltrate made of lymphocytes and eosinophils.⁹

Q: What is the most likely diagnosis?

• Leukemia cutis
• Drug reaction
• Sweet syndrome
• Erythema multiforme
• Urticaria

A: The correct answer is leukemia cutis, defined as a cutaneous infiltration by neoplastic leukocytes.¹ When the leukocytes are primarily granulocytic precursors, the terms myeloid sarcoma, granulocytic sarcoma, chloroma, and primary extramedullary leukemia have been used.² The term monoblastic sarcoma has been used when the cutaneous infiltrate is composed of neoplastic monocytic precursors.²

Leukemia cutis most commonly manifests as erythematous papules and nodules, single or multiple, of varying sizes, and afflicting one or more body sites; it typically is asymptomatic.³ It occurs in 10% to 15% of patients with acute myeloid leukemia⁴ and is itself a poor prognostic sign.⁵ The cutaneous changes may pre-date the hematologic manifestations and may even herald a relapse.⁶

In patients presenting with extramedullary leukemia and no bone marrow or blood involvement, the importance of preemptive chemotherapy for acute myelogenous leukemia has recently been emphasized.⁷

CASE CONTINUED

The patient undergoes further testing with bone marrow aspiration and trephination, which are diagnostic of acute myeloid leukemia with myelodysplastic changes: the studies reveal a clear excess of myeloblasts (accounting for 40% to 50% of nucleated cells) and clearly dysplastic erythropoiesis and myelopoiesis. Bone marrow cytogenetic analysis reveals a complex abnormal female karyotype with multiple numerical and structural abnormalities, in particular deletion of the long arm of chromosome 5, suggestive of a poor prognosis.

Skin biopsy reveals a normal epidermis but dermal perivascular involvement with a reactive T-lymphocyte infiltrate associated with immature myeloid elements, characterized by positive staining to myeloperoxidase, in keeping with leukemia cutis. It should be noted that histopathologic confirmation of leukemia cutis can be challenging, as the condition can adopt a variety of patterns, and clinicopathologic correlation is often warranted.

The patient is treated with a cycle of cytarabine-based chemotherapy, and her skin eruption transiently improves. However, her clinical condition subsequently deteriorates; she has a relapse of leukemia, with the rash returning more florid and angry-looking than previously. She is subsequently managed palliatively and passes away 3 weeks later.

THE OTHER DIAGNOSTIC CHOICES

Sweet syndrome or acute febrile neutrophilic dermatosis is often seen in association with hematologic malignancies, but the lesions are typically tender, and histopathology reveals an intense dermal neutrophilic infiltrate.⁶

Erythema multiforme is associated with malignancy, but its characteristic concentric “target” lesions are typically acral and symmetrical in their distribution; their histopathology is inflammatory.⁸

The patient had not been on any regular medications and her rash could not have been medication-induced.

Urticaria presents with pruritic evanescent wheals, which rarely last more than 12 hours.⁹ Our patient had a fixed and entirely asymptomatic rash, which in addition did not have the histopathologic features of urticaria—namely, dermal edema involved with an infiltrate made of lymphocytes and eosinophils.⁹

Q: What is the most likely diagnosis?

• Leukemia cutis
• Drug reaction
• Sweet syndrome
• Erythema multiforme
• Urticaria

A: The correct answer is leukemia cutis, defined as a cutaneous infiltration by neoplastic leukocytes.¹ When the leukocytes are primarily granulocytic precursors, the terms myeloid sarcoma, granulocytic sarcoma, chloroma, and primary extramedullary leukemia have been used.² The term monoblastic sarcoma has been used when the cutaneous infiltrate is composed of neoplastic monocytic precursors.²

Leukemia cutis most commonly manifests as erythematous papules and nodules, single or multiple, of varying sizes, and afflicting one or more body sites; it typically is asymptomatic.³ It occurs in 10% to 15% of patients with acute myeloid leukemia⁴ and is itself a poor prognostic sign.⁵ The cutaneous changes may pre-date the hematologic manifestations and may even herald a relapse.⁶

In patients presenting with extramedullary leukemia and no bone marrow or blood involvement, the importance of preemptive chemotherapy for acute myelogenous leukemia has recently been emphasized.⁷

CASE CONTINUED

The patient undergoes further testing with bone marrow aspiration and trephination, which are diagnostic of acute myeloid leukemia with myelodysplastic changes: the studies reveal a clear excess of myeloblasts (accounting for 40% to 50% of nucleated cells) and clearly dysplastic erythropoiesis and myelopoiesis. Bone marrow cytogenetic analysis reveals a complex abnormal female karyotype with multiple numerical and structural abnormalities, in particular deletion of the long arm of chromosome 5, suggestive of a poor prognosis.

Skin biopsy reveals a normal epidermis but dermal perivascular involvement with a reactive T-lymphocyte infiltrate associated with immature myeloid elements, characterized by positive staining to myeloperoxidase, in keeping with leukemia cutis. It should be noted that histopathologic confirmation of leukemia cutis can be challenging, as the condition can adopt a variety of patterns, and clinicopathologic correlation is often warranted.

The patient is treated with a cycle of cytarabine-based chemotherapy, and her skin eruption transiently improves. However, her clinical condition subsequently deteriorates; she has a relapse of leukemia, with the rash returning more florid and angry-looking than previously. She is subsequently managed palliatively and passes away 3 weeks later.

THE OTHER DIAGNOSTIC CHOICES

Sweet syndrome or acute febrile neutrophilic dermatosis is often seen in association with hematologic malignancies, but the lesions are typically tender, and histopathology reveals an intense dermal neutrophilic infiltrate.⁶

Erythema multiforme is associated with malignancy, but its characteristic concentric “target” lesions are typically acral and symmetrical in their distribution; their histopathology is inflammatory.⁸

The patient had not been on any regular medications and her rash could not have been medication-induced.

Urticaria presents with pruritic evanescent wheals, which rarely last more than 12 hours.⁹ Our patient had a fixed and entirely asymptomatic rash, which in addition did not have the histopathologic features of urticaria—namely, dermal edema involved with an infiltrate made of lymphocytes and eosinophils.⁹

In speaking about lesbian, gay, bisexual, and transgender (LGBT) health, it is not uncommon for me to be asked what is so unique about the health care needs of lesbians, gay men, bisexuals, and transgender individuals that it warrants focused attention in the training of health professionals and while providing care.¹ Although it is true that most health issues affecting LGBT individuals parallel those of the general population, people who are LGBT have been shown to have unique health needs and to experience disparities in care.

There is a growing if limited number of good studies of LGBT health. The Institute of Medicine² reported on lesbian health in 1999, concluding that enough evidence of disparities exists to support more research and to develop better methods of conducting the research. Healthy People 2020 actually recognizes significant health care disparities.³ Finally, the Institute of Medicine recently formed a committee on LGBT health issues to identify gaps in our knowledge and priorities for research. Their findings were expected to be published in late March 2011, after this article went to press.

MAKING A DIFFERENCE

While this article will not attempt to discuss all the disparities, the focus will be on how physicians can take the first critical step to helping LGBT individuals feel comfortable seeking care, ie, by being proactive in taking a history that includes discussion of sexual orientation and gender identity. Only by knowing this about patients will clinicians appropriately care for specific health needs, and will patients feel comfortable discussing their concerns in clinical settings.

While some feel this is relevant only in select areas of the country, recent data show that the LGBT population is both spread throughout the country and diverse in how they might present themselves in clinical settings.^1,4 In the United States, 1.4% to 4.1% of people identify themselves as lesbian, gay, or bisexual.⁵ About 3% of women and 4% of men say they have had a same-sex sexual contact in the last year, and 4% to 11% of women and 6% to 9% of men report having ever had one.

Everyone who practices clinical medicine needs to understand whether patients are LGBT and how to engage in conversation about sexual orientation and gender identity.

GETTING TO KNOW LGBT PATIENTS

What questions should a clinician ask to get this information? In thinking about what to ask, it helps to realize that patients generally do not mind being questioned about personal matters if the provider approaches the topic and the patient with genuine respect, empathy, and even curiosity.

On the other hand, providers often feel ill-prepared to discuss intimate issues, or feel uncomfortable doing so. Successfully achieving a change in clinical practice involves learning an approach to doing so and becoming comfortable with discussions that may follow. One question to consider is how you will feel and how you will follow up if a patient tells you that he or she is LGBT.

The core comprehensive history for LGBT patients is the same as for all patients, keeping in mind the unique LGBT health risks and issues. Clinicians may begin by getting to know each patient as a person (eg, ask about partners, children, and jobs). I like to begin a session with a patient who is otherwise in good health with an open-ended question such as “Tell me a bit about yourself.” This provides an opportunity for patients to raise a range of issues without any additional focused questions being asked. In this context, if a patient brings up issues regarding sexual orientation or gender identity, ask permission to include this information in the medical record and assure the patient of its importance and that it will be confidential.

If these issues do not come up in response to general questions, they can be embedded in the sexual history, which should be more than a history of risk behaviors and should include a discussion of sexual health, sexual orientation (including identity, behavior, and desire), and gender identity. One can start by simply asking, “Do you have any concerns or questions about your sexuality, sexual orientation, or sexual desires?”

When it is necessary to ask more directed questions, it helps to provide some context so patients do not wonder why you are asking questions they may never have been asked by a physician before. It is best to explain that these are questions you ask all patients, as the information can be important in providing quality care. Patients should be told that discussion of sexual identity, behavior, and desire, as well as gender identity, is routine and confidential. For example, you might say: “I am going to ask you some questions about your sexual health and sexuality that I ask all my patients. The answers to these questions are important for me to know to help keep you healthy. Like the rest of this visit, this information is strictly confidential.”

One usually need not be too probing to get answers; people are often very forthcoming. During such conversations, patients often tell me that it is the first time a doctor has shown any interest in talking about these topics.

In having these conversations, initially it is best to use gender-neutral terms and pronouns when referring to partners until you know which to use: for example, “Do you have a partner or a spouse?” “Are you currently in a relationship?” “What do you call your partner?” Even if you make an incorrect assumption, and the patient corrects you, you can always apologize if a mistake is made and ask which term the patient prefers. Once you know it, use the pronoun that matches a person’s gender identity.

In order to get more information from the patient, the physician can engage in a series of questions, such as:

• Are you sexually active?
• When was the last time you had sex?
• When you have sex, do you do so with men, women, or both?
• How many sexual partners have you had during the last year?
• Do you have any desires regarding sexual intimacy that you would like to discuss?

In general, it is best to mirror the patient’s language. If patients use the term “gay” or “lesbian” to describe themselves, it would be off-putting to the patient to use a more clinical term, such as homosexual, in response. Some patients may use terms such as “queer” to indicate that they do not choose to identify as gay or straight. If terms like this are unclear to you, you may simply ask what this term means to the patient.

ASSESS SEXUAL BEHAVIOR TO DETERMINE RISK

In taking a history, it is important to distinguish sexual identity from sexual behavior. Physicians need to discuss sexual behavior with patients regardless of their sexual identity in order to do a risk-assessment, ascertaining what activities they engage in and to learn what they do to prevent transmission of sexually transmitted disease. In a 2006 study of more than 4,000 men in New York City,⁴ 9.4% of those who identified themselves as straight had had sex with a man in the previous year. These men were more likely to be either foreign-born or from minority racial and ethnic groups with lower socioeconomic status. They were also less likely to have used a condom. A study of lesbians reported that 77% to 91% had at least one prior sexual experience with men, and 8% reported having had sex with a man in the previous year.⁶

Once you understand more about a patient’s sexual behavior, it is important to ask how patients protect themselves from human immunodeficiency virus (HIV) and other sexually transmitted diseases. If they use condoms or latex dams, they should be asked whether they do so consistently. Many patients have the misconception that they are practicing safe sex by only engaging in oral sex and do not realize that although it is probably protective against HIV infection, it does not protect against gonorrhea, syphilis, and other sexually transmitted diseases. Although most sexually transmitted diseases are treatable, their presence increases the risk of transmission of HIV.

Counseling on safer sex should include behavioral risk-reduction approaches. Depending on what behaviors a patient already engages in and what counseling he or she would be willing to accept, one could counsel abstinence, monogamy with an uninfected partner, reducing the number of partners, low-risk sexual practices, consistent and correct use of barrier methods, ceasing to engage in at least one high-risk activity, and avoiding excessive substance abuse. Physicians should advise patients to have a proactive plan to protect themselves and their partners. Patients should also be counseled on the correct use of barrier protection and on what is available for prophylaxis in case of high-risk HIV exposure (eg, a condom breaking or postcoital HIV disclosure). Another important question is, “Do you use alcohol or drugs when you have sex, and does your partner?” because these behaviors are often associated with unsafe sexual practices.

A new dimension of care will be biomedical prevention. While there are many ongoing studies of vaginal and anal microbicides to prevent HIV infection, there are also ongoing studies of antiretroviral therapies to do so.

One important new study demonstrated the effectiveness a biomedical intervention using antiretroviral therapy to prevent HIV infection in high-risk individuals.⁷ The study showed that men who were assigned to take a combination antiretroviral medication orally on a daily basis decreased their HIV risk by almost half compared with those assigned to take a placebo. The therapy was given along with intensive behavioral counseling. While this study was done in men who have sex with men, it is a major breakthrough and suggests there will be many new approaches to preventing HIV in the future.

A guide for clinicians has not been published by any government agency at this point, but guidance for clinicians is available from the Fenway Institute at www.fenwayhealth.org.

ASSESS GENDER-IDENTITY ISSUES

One should also routinely ask about whether patients are transgender or have gender-identity concerns. Psychologists start the conversation with the following example, which can also be used by general clinicians:

“Because so many people are impacted by gender issues, I have begun to ask everyone if they have any concerns about their gender. Anything you say about gender issues will be kept confidential. If this topic isn’t relevant to you, tell me and I’ll move on.”⁸

It is important to open the door to conversation, because many transgender people see a doctor for years and the topic never comes up. When they realize that they want to change their life, no one has ever helped them deal with the issues.

If appropriate, one can also say:

“Out of respect for my clients’ right to self-identify, I ask all clients what gender pronoun they’d prefer I use for them. What pronoun would you like me to use for you?”

Once these issues have been raised, it is important to support transgender people and help them explore a number of choices, including whether they wish to undergo hormone treatment, cosmetic surgery, and genital surgery. This may not be easy for many clinicians, so it will be important to learn about resources to care for transgender individuals in your community. Resources that can be very helpful for primary care clinicians include the following:

• The World Professional Association for Transgender Health (www.wpath.org) is the oldest and most traditional source for establishing standards of care.
• Vancouver Coastal Health published a series of monographs online (http://transhealth.vch.ca) that were developed by the University of British Columbia so that transgender people could be cared for in the community by primary care clinicians.
• The Endocrine Society in the United States published guidelines in 2009.⁹

PROVIDE SUPPORT FOR ‘COMING OUT’

We should also be understanding of people’s desires and support those who are “coming out.” The desire to reveal sexual orientation to others can happen at any age, including in childhood and among those who appear to have a traditional life because they are married and have children. Sometimes people do not know how to come out and would like to discuss such issues with their doctor.

MENTAL HEALTH CONCERNS

Given the marginalization and stigma that LGBT people face throughout their lives, it is not surprising that mental health problems are more prevalent in this population than in the general population. Gay and bisexual men have more depression, panic attacks, suicidal ideation, psychological distress, and body image and eating disorders than do heterosexual men. Lesbian and bisexual women are at greater risk of generalized anxiety disorder, depression, antidepressant use, and psychological distress.¹⁰ Care providers should screen for mental health disorders, assess comfort with sexual identity, and ask about social support.

FAMILY LIFE

Gays and lesbians increasingly want to discuss commitment, marriage, having children, parenting, and legal issues. A lot of research is being conducted on the sexual orientation of children raised by gay parents, and evidence shows that they are not more likely to be gay or lesbian than children raised by straight parents.

Elderly same-sex couples face special difficulties. They are less likely to feel comfortable “out of the closet” than are younger people. Fewer family and community supports are available to them, and they are often unable to live together in an assisted living facility. They particularly need to have advanced directives because they do not have the legal protections of other couples.

JUST A BEGINNING

While the points made above are relatively straightforward, they will open the door for many patients to have more meaningful conversations about their lives with their health care providers. It may only be a first step, but it can make a world of difference helping LGBT people feel comfortable accessing health care and receiving appropriate preventive care and treatment. Beyond the interaction with clinicians, health care providers should consider their overall environment and ensure that it is welcoming to LGBT individuals who come there for care.¹¹

RESOURCES

Family Acceptance Project. familyproject.sfsu.edu

Gay & Lesbian Medical Association. www.glma.org

Human Rights Campaign. HRC.org

Parents, Families and Friends of Lesbians and Gays. PFLAG.org

World Professional Association for Transgender Health. www.wpath.org

Youth Resource (website by and for LGBT youth). Youthresource.com

In speaking about lesbian, gay, bisexual, and transgender (LGBT) health, it is not uncommon for me to be asked what is so unique about the health care needs of lesbians, gay men, bisexuals, and transgender individuals that it warrants focused attention in the training of health professionals and while providing care.¹ Although it is true that most health issues affecting LGBT individuals parallel those of the general population, people who are LGBT have been shown to have unique health needs and to experience disparities in care.

There is a growing if limited number of good studies of LGBT health. The Institute of Medicine² reported on lesbian health in 1999, concluding that enough evidence of disparities exists to support more research and to develop better methods of conducting the research. Healthy People 2020 actually recognizes significant health care disparities.³ Finally, the Institute of Medicine recently formed a committee on LGBT health issues to identify gaps in our knowledge and priorities for research. Their findings were expected to be published in late March 2011, after this article went to press.

MAKING A DIFFERENCE

While this article will not attempt to discuss all the disparities, the focus will be on how physicians can take the first critical step to helping LGBT individuals feel comfortable seeking care, ie, by being proactive in taking a history that includes discussion of sexual orientation and gender identity. Only by knowing this about patients will clinicians appropriately care for specific health needs, and will patients feel comfortable discussing their concerns in clinical settings.

While some feel this is relevant only in select areas of the country, recent data show that the LGBT population is both spread throughout the country and diverse in how they might present themselves in clinical settings.^1,4 In the United States, 1.4% to 4.1% of people identify themselves as lesbian, gay, or bisexual.⁵ About 3% of women and 4% of men say they have had a same-sex sexual contact in the last year, and 4% to 11% of women and 6% to 9% of men report having ever had one.

Everyone who practices clinical medicine needs to understand whether patients are LGBT and how to engage in conversation about sexual orientation and gender identity.

GETTING TO KNOW LGBT PATIENTS

What questions should a clinician ask to get this information? In thinking about what to ask, it helps to realize that patients generally do not mind being questioned about personal matters if the provider approaches the topic and the patient with genuine respect, empathy, and even curiosity.

On the other hand, providers often feel ill-prepared to discuss intimate issues, or feel uncomfortable doing so. Successfully achieving a change in clinical practice involves learning an approach to doing so and becoming comfortable with discussions that may follow. One question to consider is how you will feel and how you will follow up if a patient tells you that he or she is LGBT.

The core comprehensive history for LGBT patients is the same as for all patients, keeping in mind the unique LGBT health risks and issues. Clinicians may begin by getting to know each patient as a person (eg, ask about partners, children, and jobs). I like to begin a session with a patient who is otherwise in good health with an open-ended question such as “Tell me a bit about yourself.” This provides an opportunity for patients to raise a range of issues without any additional focused questions being asked. In this context, if a patient brings up issues regarding sexual orientation or gender identity, ask permission to include this information in the medical record and assure the patient of its importance and that it will be confidential.

If these issues do not come up in response to general questions, they can be embedded in the sexual history, which should be more than a history of risk behaviors and should include a discussion of sexual health, sexual orientation (including identity, behavior, and desire), and gender identity. One can start by simply asking, “Do you have any concerns or questions about your sexuality, sexual orientation, or sexual desires?”

When it is necessary to ask more directed questions, it helps to provide some context so patients do not wonder why you are asking questions they may never have been asked by a physician before. It is best to explain that these are questions you ask all patients, as the information can be important in providing quality care. Patients should be told that discussion of sexual identity, behavior, and desire, as well as gender identity, is routine and confidential. For example, you might say: “I am going to ask you some questions about your sexual health and sexuality that I ask all my patients. The answers to these questions are important for me to know to help keep you healthy. Like the rest of this visit, this information is strictly confidential.”

One usually need not be too probing to get answers; people are often very forthcoming. During such conversations, patients often tell me that it is the first time a doctor has shown any interest in talking about these topics.

In having these conversations, initially it is best to use gender-neutral terms and pronouns when referring to partners until you know which to use: for example, “Do you have a partner or a spouse?” “Are you currently in a relationship?” “What do you call your partner?” Even if you make an incorrect assumption, and the patient corrects you, you can always apologize if a mistake is made and ask which term the patient prefers. Once you know it, use the pronoun that matches a person’s gender identity.

In order to get more information from the patient, the physician can engage in a series of questions, such as:

• Are you sexually active?
• When was the last time you had sex?
• When you have sex, do you do so with men, women, or both?
• How many sexual partners have you had during the last year?
• Do you have any desires regarding sexual intimacy that you would like to discuss?

In general, it is best to mirror the patient’s language. If patients use the term “gay” or “lesbian” to describe themselves, it would be off-putting to the patient to use a more clinical term, such as homosexual, in response. Some patients may use terms such as “queer” to indicate that they do not choose to identify as gay or straight. If terms like this are unclear to you, you may simply ask what this term means to the patient.

ASSESS SEXUAL BEHAVIOR TO DETERMINE RISK

In taking a history, it is important to distinguish sexual identity from sexual behavior. Physicians need to discuss sexual behavior with patients regardless of their sexual identity in order to do a risk-assessment, ascertaining what activities they engage in and to learn what they do to prevent transmission of sexually transmitted disease. In a 2006 study of more than 4,000 men in New York City,⁴ 9.4% of those who identified themselves as straight had had sex with a man in the previous year. These men were more likely to be either foreign-born or from minority racial and ethnic groups with lower socioeconomic status. They were also less likely to have used a condom. A study of lesbians reported that 77% to 91% had at least one prior sexual experience with men, and 8% reported having had sex with a man in the previous year.⁶

Once you understand more about a patient’s sexual behavior, it is important to ask how patients protect themselves from human immunodeficiency virus (HIV) and other sexually transmitted diseases. If they use condoms or latex dams, they should be asked whether they do so consistently. Many patients have the misconception that they are practicing safe sex by only engaging in oral sex and do not realize that although it is probably protective against HIV infection, it does not protect against gonorrhea, syphilis, and other sexually transmitted diseases. Although most sexually transmitted diseases are treatable, their presence increases the risk of transmission of HIV.

Counseling on safer sex should include behavioral risk-reduction approaches. Depending on what behaviors a patient already engages in and what counseling he or she would be willing to accept, one could counsel abstinence, monogamy with an uninfected partner, reducing the number of partners, low-risk sexual practices, consistent and correct use of barrier methods, ceasing to engage in at least one high-risk activity, and avoiding excessive substance abuse. Physicians should advise patients to have a proactive plan to protect themselves and their partners. Patients should also be counseled on the correct use of barrier protection and on what is available for prophylaxis in case of high-risk HIV exposure (eg, a condom breaking or postcoital HIV disclosure). Another important question is, “Do you use alcohol or drugs when you have sex, and does your partner?” because these behaviors are often associated with unsafe sexual practices.

A new dimension of care will be biomedical prevention. While there are many ongoing studies of vaginal and anal microbicides to prevent HIV infection, there are also ongoing studies of antiretroviral therapies to do so.

One important new study demonstrated the effectiveness a biomedical intervention using antiretroviral therapy to prevent HIV infection in high-risk individuals.⁷ The study showed that men who were assigned to take a combination antiretroviral medication orally on a daily basis decreased their HIV risk by almost half compared with those assigned to take a placebo. The therapy was given along with intensive behavioral counseling. While this study was done in men who have sex with men, it is a major breakthrough and suggests there will be many new approaches to preventing HIV in the future.

A guide for clinicians has not been published by any government agency at this point, but guidance for clinicians is available from the Fenway Institute at www.fenwayhealth.org.

ASSESS GENDER-IDENTITY ISSUES

One should also routinely ask about whether patients are transgender or have gender-identity concerns. Psychologists start the conversation with the following example, which can also be used by general clinicians:

“Because so many people are impacted by gender issues, I have begun to ask everyone if they have any concerns about their gender. Anything you say about gender issues will be kept confidential. If this topic isn’t relevant to you, tell me and I’ll move on.”⁸

It is important to open the door to conversation, because many transgender people see a doctor for years and the topic never comes up. When they realize that they want to change their life, no one has ever helped them deal with the issues.

If appropriate, one can also say:

“Out of respect for my clients’ right to self-identify, I ask all clients what gender pronoun they’d prefer I use for them. What pronoun would you like me to use for you?”

Once these issues have been raised, it is important to support transgender people and help them explore a number of choices, including whether they wish to undergo hormone treatment, cosmetic surgery, and genital surgery. This may not be easy for many clinicians, so it will be important to learn about resources to care for transgender individuals in your community. Resources that can be very helpful for primary care clinicians include the following:

• The World Professional Association for Transgender Health (www.wpath.org) is the oldest and most traditional source for establishing standards of care.
• Vancouver Coastal Health published a series of monographs online (http://transhealth.vch.ca) that were developed by the University of British Columbia so that transgender people could be cared for in the community by primary care clinicians.
• The Endocrine Society in the United States published guidelines in 2009.⁹

PROVIDE SUPPORT FOR ‘COMING OUT’

We should also be understanding of people’s desires and support those who are “coming out.” The desire to reveal sexual orientation to others can happen at any age, including in childhood and among those who appear to have a traditional life because they are married and have children. Sometimes people do not know how to come out and would like to discuss such issues with their doctor.

MENTAL HEALTH CONCERNS

Given the marginalization and stigma that LGBT people face throughout their lives, it is not surprising that mental health problems are more prevalent in this population than in the general population. Gay and bisexual men have more depression, panic attacks, suicidal ideation, psychological distress, and body image and eating disorders than do heterosexual men. Lesbian and bisexual women are at greater risk of generalized anxiety disorder, depression, antidepressant use, and psychological distress.¹⁰ Care providers should screen for mental health disorders, assess comfort with sexual identity, and ask about social support.

FAMILY LIFE

Gays and lesbians increasingly want to discuss commitment, marriage, having children, parenting, and legal issues. A lot of research is being conducted on the sexual orientation of children raised by gay parents, and evidence shows that they are not more likely to be gay or lesbian than children raised by straight parents.

Elderly same-sex couples face special difficulties. They are less likely to feel comfortable “out of the closet” than are younger people. Fewer family and community supports are available to them, and they are often unable to live together in an assisted living facility. They particularly need to have advanced directives because they do not have the legal protections of other couples.

JUST A BEGINNING

While the points made above are relatively straightforward, they will open the door for many patients to have more meaningful conversations about their lives with their health care providers. It may only be a first step, but it can make a world of difference helping LGBT people feel comfortable accessing health care and receiving appropriate preventive care and treatment. Beyond the interaction with clinicians, health care providers should consider their overall environment and ensure that it is welcoming to LGBT individuals who come there for care.¹¹

RESOURCES

Family Acceptance Project. familyproject.sfsu.edu

Gay & Lesbian Medical Association. www.glma.org

Human Rights Campaign. HRC.org

Parents, Families and Friends of Lesbians and Gays. PFLAG.org

World Professional Association for Transgender Health. www.wpath.org

Youth Resource (website by and for LGBT youth). Youthresource.com

In speaking about lesbian, gay, bisexual, and transgender (LGBT) health, it is not uncommon for me to be asked what is so unique about the health care needs of lesbians, gay men, bisexuals, and transgender individuals that it warrants focused attention in the training of health professionals and while providing care.¹ Although it is true that most health issues affecting LGBT individuals parallel those of the general population, people who are LGBT have been shown to have unique health needs and to experience disparities in care.

There is a growing if limited number of good studies of LGBT health. The Institute of Medicine² reported on lesbian health in 1999, concluding that enough evidence of disparities exists to support more research and to develop better methods of conducting the research. Healthy People 2020 actually recognizes significant health care disparities.³ Finally, the Institute of Medicine recently formed a committee on LGBT health issues to identify gaps in our knowledge and priorities for research. Their findings were expected to be published in late March 2011, after this article went to press.

MAKING A DIFFERENCE

While this article will not attempt to discuss all the disparities, the focus will be on how physicians can take the first critical step to helping LGBT individuals feel comfortable seeking care, ie, by being proactive in taking a history that includes discussion of sexual orientation and gender identity. Only by knowing this about patients will clinicians appropriately care for specific health needs, and will patients feel comfortable discussing their concerns in clinical settings.

While some feel this is relevant only in select areas of the country, recent data show that the LGBT population is both spread throughout the country and diverse in how they might present themselves in clinical settings.^1,4 In the United States, 1.4% to 4.1% of people identify themselves as lesbian, gay, or bisexual.⁵ About 3% of women and 4% of men say they have had a same-sex sexual contact in the last year, and 4% to 11% of women and 6% to 9% of men report having ever had one.

Everyone who practices clinical medicine needs to understand whether patients are LGBT and how to engage in conversation about sexual orientation and gender identity.

GETTING TO KNOW LGBT PATIENTS

What questions should a clinician ask to get this information? In thinking about what to ask, it helps to realize that patients generally do not mind being questioned about personal matters if the provider approaches the topic and the patient with genuine respect, empathy, and even curiosity.

On the other hand, providers often feel ill-prepared to discuss intimate issues, or feel uncomfortable doing so. Successfully achieving a change in clinical practice involves learning an approach to doing so and becoming comfortable with discussions that may follow. One question to consider is how you will feel and how you will follow up if a patient tells you that he or she is LGBT.

The core comprehensive history for LGBT patients is the same as for all patients, keeping in mind the unique LGBT health risks and issues. Clinicians may begin by getting to know each patient as a person (eg, ask about partners, children, and jobs). I like to begin a session with a patient who is otherwise in good health with an open-ended question such as “Tell me a bit about yourself.” This provides an opportunity for patients to raise a range of issues without any additional focused questions being asked. In this context, if a patient brings up issues regarding sexual orientation or gender identity, ask permission to include this information in the medical record and assure the patient of its importance and that it will be confidential.

If these issues do not come up in response to general questions, they can be embedded in the sexual history, which should be more than a history of risk behaviors and should include a discussion of sexual health, sexual orientation (including identity, behavior, and desire), and gender identity. One can start by simply asking, “Do you have any concerns or questions about your sexuality, sexual orientation, or sexual desires?”

When it is necessary to ask more directed questions, it helps to provide some context so patients do not wonder why you are asking questions they may never have been asked by a physician before. It is best to explain that these are questions you ask all patients, as the information can be important in providing quality care. Patients should be told that discussion of sexual identity, behavior, and desire, as well as gender identity, is routine and confidential. For example, you might say: “I am going to ask you some questions about your sexual health and sexuality that I ask all my patients. The answers to these questions are important for me to know to help keep you healthy. Like the rest of this visit, this information is strictly confidential.”

One usually need not be too probing to get answers; people are often very forthcoming. During such conversations, patients often tell me that it is the first time a doctor has shown any interest in talking about these topics.

In having these conversations, initially it is best to use gender-neutral terms and pronouns when referring to partners until you know which to use: for example, “Do you have a partner or a spouse?” “Are you currently in a relationship?” “What do you call your partner?” Even if you make an incorrect assumption, and the patient corrects you, you can always apologize if a mistake is made and ask which term the patient prefers. Once you know it, use the pronoun that matches a person’s gender identity.

In order to get more information from the patient, the physician can engage in a series of questions, such as:

• Are you sexually active?
• When was the last time you had sex?
• When you have sex, do you do so with men, women, or both?
• How many sexual partners have you had during the last year?
• Do you have any desires regarding sexual intimacy that you would like to discuss?

In general, it is best to mirror the patient’s language. If patients use the term “gay” or “lesbian” to describe themselves, it would be off-putting to the patient to use a more clinical term, such as homosexual, in response. Some patients may use terms such as “queer” to indicate that they do not choose to identify as gay or straight. If terms like this are unclear to you, you may simply ask what this term means to the patient.

ASSESS SEXUAL BEHAVIOR TO DETERMINE RISK

In taking a history, it is important to distinguish sexual identity from sexual behavior. Physicians need to discuss sexual behavior with patients regardless of their sexual identity in order to do a risk-assessment, ascertaining what activities they engage in and to learn what they do to prevent transmission of sexually transmitted disease. In a 2006 study of more than 4,000 men in New York City,⁴ 9.4% of those who identified themselves as straight had had sex with a man in the previous year. These men were more likely to be either foreign-born or from minority racial and ethnic groups with lower socioeconomic status. They were also less likely to have used a condom. A study of lesbians reported that 77% to 91% had at least one prior sexual experience with men, and 8% reported having had sex with a man in the previous year.⁶

Once you understand more about a patient’s sexual behavior, it is important to ask how patients protect themselves from human immunodeficiency virus (HIV) and other sexually transmitted diseases. If they use condoms or latex dams, they should be asked whether they do so consistently. Many patients have the misconception that they are practicing safe sex by only engaging in oral sex and do not realize that although it is probably protective against HIV infection, it does not protect against gonorrhea, syphilis, and other sexually transmitted diseases. Although most sexually transmitted diseases are treatable, their presence increases the risk of transmission of HIV.

Counseling on safer sex should include behavioral risk-reduction approaches. Depending on what behaviors a patient already engages in and what counseling he or she would be willing to accept, one could counsel abstinence, monogamy with an uninfected partner, reducing the number of partners, low-risk sexual practices, consistent and correct use of barrier methods, ceasing to engage in at least one high-risk activity, and avoiding excessive substance abuse. Physicians should advise patients to have a proactive plan to protect themselves and their partners. Patients should also be counseled on the correct use of barrier protection and on what is available for prophylaxis in case of high-risk HIV exposure (eg, a condom breaking or postcoital HIV disclosure). Another important question is, “Do you use alcohol or drugs when you have sex, and does your partner?” because these behaviors are often associated with unsafe sexual practices.

A new dimension of care will be biomedical prevention. While there are many ongoing studies of vaginal and anal microbicides to prevent HIV infection, there are also ongoing studies of antiretroviral therapies to do so.

One important new study demonstrated the effectiveness a biomedical intervention using antiretroviral therapy to prevent HIV infection in high-risk individuals.⁷ The study showed that men who were assigned to take a combination antiretroviral medication orally on a daily basis decreased their HIV risk by almost half compared with those assigned to take a placebo. The therapy was given along with intensive behavioral counseling. While this study was done in men who have sex with men, it is a major breakthrough and suggests there will be many new approaches to preventing HIV in the future.

A guide for clinicians has not been published by any government agency at this point, but guidance for clinicians is available from the Fenway Institute at www.fenwayhealth.org.

ASSESS GENDER-IDENTITY ISSUES

One should also routinely ask about whether patients are transgender or have gender-identity concerns. Psychologists start the conversation with the following example, which can also be used by general clinicians:

“Because so many people are impacted by gender issues, I have begun to ask everyone if they have any concerns about their gender. Anything you say about gender issues will be kept confidential. If this topic isn’t relevant to you, tell me and I’ll move on.”⁸

It is important to open the door to conversation, because many transgender people see a doctor for years and the topic never comes up. When they realize that they want to change their life, no one has ever helped them deal with the issues.

If appropriate, one can also say:

“Out of respect for my clients’ right to self-identify, I ask all clients what gender pronoun they’d prefer I use for them. What pronoun would you like me to use for you?”

Once these issues have been raised, it is important to support transgender people and help them explore a number of choices, including whether they wish to undergo hormone treatment, cosmetic surgery, and genital surgery. This may not be easy for many clinicians, so it will be important to learn about resources to care for transgender individuals in your community. Resources that can be very helpful for primary care clinicians include the following:

• The World Professional Association for Transgender Health (www.wpath.org) is the oldest and most traditional source for establishing standards of care.
• Vancouver Coastal Health published a series of monographs online (http://transhealth.vch.ca) that were developed by the University of British Columbia so that transgender people could be cared for in the community by primary care clinicians.
• The Endocrine Society in the United States published guidelines in 2009.⁹

PROVIDE SUPPORT FOR ‘COMING OUT’

We should also be understanding of people’s desires and support those who are “coming out.” The desire to reveal sexual orientation to others can happen at any age, including in childhood and among those who appear to have a traditional life because they are married and have children. Sometimes people do not know how to come out and would like to discuss such issues with their doctor.

MENTAL HEALTH CONCERNS

Given the marginalization and stigma that LGBT people face throughout their lives, it is not surprising that mental health problems are more prevalent in this population than in the general population. Gay and bisexual men have more depression, panic attacks, suicidal ideation, psychological distress, and body image and eating disorders than do heterosexual men. Lesbian and bisexual women are at greater risk of generalized anxiety disorder, depression, antidepressant use, and psychological distress.¹⁰ Care providers should screen for mental health disorders, assess comfort with sexual identity, and ask about social support.

FAMILY LIFE

Gays and lesbians increasingly want to discuss commitment, marriage, having children, parenting, and legal issues. A lot of research is being conducted on the sexual orientation of children raised by gay parents, and evidence shows that they are not more likely to be gay or lesbian than children raised by straight parents.

Elderly same-sex couples face special difficulties. They are less likely to feel comfortable “out of the closet” than are younger people. Fewer family and community supports are available to them, and they are often unable to live together in an assisted living facility. They particularly need to have advanced directives because they do not have the legal protections of other couples.

JUST A BEGINNING

While the points made above are relatively straightforward, they will open the door for many patients to have more meaningful conversations about their lives with their health care providers. It may only be a first step, but it can make a world of difference helping LGBT people feel comfortable accessing health care and receiving appropriate preventive care and treatment. Beyond the interaction with clinicians, health care providers should consider their overall environment and ensure that it is welcoming to LGBT individuals who come there for care.¹¹

RESOURCES

Family Acceptance Project. familyproject.sfsu.edu

Gay & Lesbian Medical Association. www.glma.org

Human Rights Campaign. HRC.org

Parents, Families and Friends of Lesbians and Gays. PFLAG.org

World Professional Association for Transgender Health. www.wpath.org

Youth Resource (website by and for LGBT youth). Youthresource.com

For some drugs there is a key step in metabolism, often in a rate-limiting pathway, with an enzyme that has known and detectable polymorphisms that differ dramatically in their ability to affect the drug’s degradation. In theory, by determining the patient’s specific genotype ahead of time, the initial dose of the drug can be determined more rationally. In this issue of the Journal, Kitzmiller et al describe several drugs for which this may be true.

However, for this approach to be practical and cost-effective, several conditions should be met. The drug must be one that needs to be dosed to its therapeutic level rapidly: if there is time to titrate slowly, then there is little need for the extra expense associated with genotyping in order to titrate it more rapidly. Also, it should be proven that dosing based on advance knowledge of the genotype of the target actually results in safer or more efficacious dosing.

For carbamazepine (Tegretol, Equetro) and allopurinol (Zyloprim), specific human leukocyte antigen haplotypes are associated with a strikingly increased frequency of serious hypersensitivity reactions. In some patients, these should be checked before giving the drug.

But the concept of pharmacogenomics is broad, and it may yet explain many vagaries of drug-responsiveness in individual patients. Polymorphisms in renal anion transporters may dictate the level of anionic drugs. Drug-receptor polymorphisms may determine the affinity of a drug for its target and, hence, its efficacy. Cell-membrane transporters, which may have functionally different stable alleles or polymorphisms, may regulate intracellular drug levels by pumping the drug into or out of cells with different efficiencies.

As the entire human genome is dissected and analyzed, and as more and more genes (with their polymorphisms) are linked to specific functions readily detectable in specific patients, we will have more opportunities to match the right drug and dose to the right patient. We are not there yet, but that day is coming.

For some drugs there is a key step in metabolism, often in a rate-limiting pathway, with an enzyme that has known and detectable polymorphisms that differ dramatically in their ability to affect the drug’s degradation. In theory, by determining the patient’s specific genotype ahead of time, the initial dose of the drug can be determined more rationally. In this issue of the Journal, Kitzmiller et al describe several drugs for which this may be true.

However, for this approach to be practical and cost-effective, several conditions should be met. The drug must be one that needs to be dosed to its therapeutic level rapidly: if there is time to titrate slowly, then there is little need for the extra expense associated with genotyping in order to titrate it more rapidly. Also, it should be proven that dosing based on advance knowledge of the genotype of the target actually results in safer or more efficacious dosing.

For carbamazepine (Tegretol, Equetro) and allopurinol (Zyloprim), specific human leukocyte antigen haplotypes are associated with a strikingly increased frequency of serious hypersensitivity reactions. In some patients, these should be checked before giving the drug.

But the concept of pharmacogenomics is broad, and it may yet explain many vagaries of drug-responsiveness in individual patients. Polymorphisms in renal anion transporters may dictate the level of anionic drugs. Drug-receptor polymorphisms may determine the affinity of a drug for its target and, hence, its efficacy. Cell-membrane transporters, which may have functionally different stable alleles or polymorphisms, may regulate intracellular drug levels by pumping the drug into or out of cells with different efficiencies.

As the entire human genome is dissected and analyzed, and as more and more genes (with their polymorphisms) are linked to specific functions readily detectable in specific patients, we will have more opportunities to match the right drug and dose to the right patient. We are not there yet, but that day is coming.

For some drugs there is a key step in metabolism, often in a rate-limiting pathway, with an enzyme that has known and detectable polymorphisms that differ dramatically in their ability to affect the drug’s degradation. In theory, by determining the patient’s specific genotype ahead of time, the initial dose of the drug can be determined more rationally. In this issue of the Journal, Kitzmiller et al describe several drugs for which this may be true.

However, for this approach to be practical and cost-effective, several conditions should be met. The drug must be one that needs to be dosed to its therapeutic level rapidly: if there is time to titrate slowly, then there is little need for the extra expense associated with genotyping in order to titrate it more rapidly. Also, it should be proven that dosing based on advance knowledge of the genotype of the target actually results in safer or more efficacious dosing.

For carbamazepine (Tegretol, Equetro) and allopurinol (Zyloprim), specific human leukocyte antigen haplotypes are associated with a strikingly increased frequency of serious hypersensitivity reactions. In some patients, these should be checked before giving the drug.

But the concept of pharmacogenomics is broad, and it may yet explain many vagaries of drug-responsiveness in individual patients. Polymorphisms in renal anion transporters may dictate the level of anionic drugs. Drug-receptor polymorphisms may determine the affinity of a drug for its target and, hence, its efficacy. Cell-membrane transporters, which may have functionally different stable alleles or polymorphisms, may regulate intracellular drug levels by pumping the drug into or out of cells with different efficiencies.

As the entire human genome is dissected and analyzed, and as more and more genes (with their polymorphisms) are linked to specific functions readily detectable in specific patients, we will have more opportunities to match the right drug and dose to the right patient. We are not there yet, but that day is coming.

Since the human genome was sequenced in 2000, the American public has continued to hold hope that our growing understanding of genetics will revolutionize the practice of medicine.

See related article

One way genetics promises to improve the quality and value of health care is in personalized medicine, by helping us tailor treatment to a person’s individual genetic makeup. One such approach is called pharmacogenomics.

Pharmacogenomics uses knowledge of a person’s genetics to understand how a particular drug will work, or not work, in his or her body. For instance, some people might carry genes that make them more sensitive than average to a drug, and therefore they would require a lower dose. Others might have genes that make them resistant to the drug, meaning the drug is ineffective unless they receive a higher dose.

Adverse drug reactions are a leading cause of death in hospitalized patients in the United States and are responsible for billions of dollars in health care costs.^1,2 Our current practice of prescribing for adult patients is largely trial-and-error, with the same dose given to all patients, in many cases with little regard even to sex, height, or weight.

Pharmacogenomics promises to change this way of prescribing to a customized approach that uses genetic information to predict an individual’s response to medications. It is one piece of an overall initiative to personalize patient care based on the patient’s individual characteristics and preferences.

OVERCOMING BARRIERS TO USING PHARMACOGENOMICS IN PRACTICE

If personalized medicine has promised to improve the quality and value of health care for our patients, why have we been so slow to adopt this information in clinical practice?

The usual barriers to clinical adoption certainly exist. We need further studies to determine whether genetic-based prescribing is truly valid, and for which patient populations. We need to determine whether this approach is cost-effective and better than the current standard of care. We need to work on payment options.

However, one of the largest barriers for busy primary care physicians is the lack of time to keep up with new information. Many practicing physicians were taught little about formal genetics in medical school. The body of scientific literature on pharmacogenomics is fragmented, and it crosses disease states and specialties, making it difficult to unite. Given the breadth of diseases treated and drugs prescribed by primary care physicians, it is unrealistic for most to keep track of the vast body of literature of pharmacogenomic testing and to decipher how to apply this to clinical practice.

In this issue of the Journal, Kitzmiller et al³ provide one solution to this problem, giving an overview of pharmacogenomic applications that might be pertinent to practicing physicians. However, as we try to make pharmacogenomics accessible to busy physicians, we need other solutions to integrate pharmacogenomic information efficiently into the clinical work flow. One approach might be to build pharmacogenomics into the electronic medical record. We can also store the integrated information in research databases and provide clinical recommendations on Internet sites such as www.pharmgkb.org, and we can develop applications to run on cell phones and iPads.

QUESTIONS REMAIN

Kitzmiller et al discuss an important step in this process, highlighting several key questions:

Should we seek genetics-based information to personalize drug selection? Based on the information presented in the literature and in the Kitzmiller paper, there may be circumstances when it is appropriate to consider doing so. While the evidence is not yet compelling to order these tests on a regular basis in clinical practice, this information might be helpful in some situations, such as for patients who have had adverse effects from minimal doses of antidepressants.

For now, clinicians should not abandon their current practice of personalizing patient care on the basis of personal, cultural, and economic preferences. Rather, they should consider pharmacogenomic information an additional piece of information when selecting drug therapy. We should also encourage health care systems and interested providers to be early adopters and to study how their outcomes compare with the standard of care.

Once we have this information, what is our obligation to use it? An increasing number of patients already have genetic information in their health record, either ordered by or provided to their physicians. However, there is little in the scientific literature to guide us in this arena.

Yet most of us would agree that if we have information (genetic or otherwise) that can help to select a drug type or dose or reduce adverse events or costs, we should consider this information in our decision-making. Several circumstances are documented in this paper and in the literature in which prior knowledge about drug metabolism can help in selecting a dose of medication. One example would be the 50% recommended reduction in tricyclic antidepressant dose if the patient is a CYP2D6 poor metabolizer.⁴

MOVING FORWARD AS A TEAM

In summary, Kitzmiller et al bring to light the promise and the uncertainties that currently exist in the field of pharmacogenomics. While it is unclear if we should incorporate pharmacogenomic tests into standard medical practice at this time, it is clear that this information is becoming more readily available, whether or not we have requested it. Some would argue that, once we have the information, we have an obligation to use it, just as we use other information in our clinical decision-making. This means we need to develop tools and resources to help practitioners evaluate pharmacogenomic data and incorporate it into clinical care in an efficient manner.

The authors also highlight the need for more education about drug metabolism in general, and they cite several instances in which knowledge of drug interactions and metabolism can clearly influence decision-making. An example is paroxetine (Paxil) inhibition of tamoxifen (Nolvadex).⁵

Lastly, regardless of our personal feelings about the clinical usefulness of genetic testing in large populations, we need to work together to determine clinical utility and validity and to develop efficient ways to put into practice findings that could affect patient care. As we move forward, we need to work as a team, utilizing our clinical partners—pharmacists, pharmacologists, metabolism and health information technology experts, and medical geneticists. Working as a team, pooling our resources and tools, we move closer to providing world-class personalized health care.

Since the human genome was sequenced in 2000, the American public has continued to hold hope that our growing understanding of genetics will revolutionize the practice of medicine.

See related article

One way genetics promises to improve the quality and value of health care is in personalized medicine, by helping us tailor treatment to a person’s individual genetic makeup. One such approach is called pharmacogenomics.

Pharmacogenomics uses knowledge of a person’s genetics to understand how a particular drug will work, or not work, in his or her body. For instance, some people might carry genes that make them more sensitive than average to a drug, and therefore they would require a lower dose. Others might have genes that make them resistant to the drug, meaning the drug is ineffective unless they receive a higher dose.

Adverse drug reactions are a leading cause of death in hospitalized patients in the United States and are responsible for billions of dollars in health care costs.^1,2 Our current practice of prescribing for adult patients is largely trial-and-error, with the same dose given to all patients, in many cases with little regard even to sex, height, or weight.

Pharmacogenomics promises to change this way of prescribing to a customized approach that uses genetic information to predict an individual’s response to medications. It is one piece of an overall initiative to personalize patient care based on the patient’s individual characteristics and preferences.

OVERCOMING BARRIERS TO USING PHARMACOGENOMICS IN PRACTICE

If personalized medicine has promised to improve the quality and value of health care for our patients, why have we been so slow to adopt this information in clinical practice?

The usual barriers to clinical adoption certainly exist. We need further studies to determine whether genetic-based prescribing is truly valid, and for which patient populations. We need to determine whether this approach is cost-effective and better than the current standard of care. We need to work on payment options.

However, one of the largest barriers for busy primary care physicians is the lack of time to keep up with new information. Many practicing physicians were taught little about formal genetics in medical school. The body of scientific literature on pharmacogenomics is fragmented, and it crosses disease states and specialties, making it difficult to unite. Given the breadth of diseases treated and drugs prescribed by primary care physicians, it is unrealistic for most to keep track of the vast body of literature of pharmacogenomic testing and to decipher how to apply this to clinical practice.

In this issue of the Journal, Kitzmiller et al³ provide one solution to this problem, giving an overview of pharmacogenomic applications that might be pertinent to practicing physicians. However, as we try to make pharmacogenomics accessible to busy physicians, we need other solutions to integrate pharmacogenomic information efficiently into the clinical work flow. One approach might be to build pharmacogenomics into the electronic medical record. We can also store the integrated information in research databases and provide clinical recommendations on Internet sites such as www.pharmgkb.org, and we can develop applications to run on cell phones and iPads.

QUESTIONS REMAIN

Kitzmiller et al discuss an important step in this process, highlighting several key questions:

Should we seek genetics-based information to personalize drug selection? Based on the information presented in the literature and in the Kitzmiller paper, there may be circumstances when it is appropriate to consider doing so. While the evidence is not yet compelling to order these tests on a regular basis in clinical practice, this information might be helpful in some situations, such as for patients who have had adverse effects from minimal doses of antidepressants.

For now, clinicians should not abandon their current practice of personalizing patient care on the basis of personal, cultural, and economic preferences. Rather, they should consider pharmacogenomic information an additional piece of information when selecting drug therapy. We should also encourage health care systems and interested providers to be early adopters and to study how their outcomes compare with the standard of care.

Once we have this information, what is our obligation to use it? An increasing number of patients already have genetic information in their health record, either ordered by or provided to their physicians. However, there is little in the scientific literature to guide us in this arena.

Yet most of us would agree that if we have information (genetic or otherwise) that can help to select a drug type or dose or reduce adverse events or costs, we should consider this information in our decision-making. Several circumstances are documented in this paper and in the literature in which prior knowledge about drug metabolism can help in selecting a dose of medication. One example would be the 50% recommended reduction in tricyclic antidepressant dose if the patient is a CYP2D6 poor metabolizer.⁴

MOVING FORWARD AS A TEAM

In summary, Kitzmiller et al bring to light the promise and the uncertainties that currently exist in the field of pharmacogenomics. While it is unclear if we should incorporate pharmacogenomic tests into standard medical practice at this time, it is clear that this information is becoming more readily available, whether or not we have requested it. Some would argue that, once we have the information, we have an obligation to use it, just as we use other information in our clinical decision-making. This means we need to develop tools and resources to help practitioners evaluate pharmacogenomic data and incorporate it into clinical care in an efficient manner.

The authors also highlight the need for more education about drug metabolism in general, and they cite several instances in which knowledge of drug interactions and metabolism can clearly influence decision-making. An example is paroxetine (Paxil) inhibition of tamoxifen (Nolvadex).⁵

Lastly, regardless of our personal feelings about the clinical usefulness of genetic testing in large populations, we need to work together to determine clinical utility and validity and to develop efficient ways to put into practice findings that could affect patient care. As we move forward, we need to work as a team, utilizing our clinical partners—pharmacists, pharmacologists, metabolism and health information technology experts, and medical geneticists. Working as a team, pooling our resources and tools, we move closer to providing world-class personalized health care.

Since the human genome was sequenced in 2000, the American public has continued to hold hope that our growing understanding of genetics will revolutionize the practice of medicine.

See related article

One way genetics promises to improve the quality and value of health care is in personalized medicine, by helping us tailor treatment to a person’s individual genetic makeup. One such approach is called pharmacogenomics.

Pharmacogenomics uses knowledge of a person’s genetics to understand how a particular drug will work, or not work, in his or her body. For instance, some people might carry genes that make them more sensitive than average to a drug, and therefore they would require a lower dose. Others might have genes that make them resistant to the drug, meaning the drug is ineffective unless they receive a higher dose.

Adverse drug reactions are a leading cause of death in hospitalized patients in the United States and are responsible for billions of dollars in health care costs.^1,2 Our current practice of prescribing for adult patients is largely trial-and-error, with the same dose given to all patients, in many cases with little regard even to sex, height, or weight.

Pharmacogenomics promises to change this way of prescribing to a customized approach that uses genetic information to predict an individual’s response to medications. It is one piece of an overall initiative to personalize patient care based on the patient’s individual characteristics and preferences.

OVERCOMING BARRIERS TO USING PHARMACOGENOMICS IN PRACTICE

If personalized medicine has promised to improve the quality and value of health care for our patients, why have we been so slow to adopt this information in clinical practice?

The usual barriers to clinical adoption certainly exist. We need further studies to determine whether genetic-based prescribing is truly valid, and for which patient populations. We need to determine whether this approach is cost-effective and better than the current standard of care. We need to work on payment options.

However, one of the largest barriers for busy primary care physicians is the lack of time to keep up with new information. Many practicing physicians were taught little about formal genetics in medical school. The body of scientific literature on pharmacogenomics is fragmented, and it crosses disease states and specialties, making it difficult to unite. Given the breadth of diseases treated and drugs prescribed by primary care physicians, it is unrealistic for most to keep track of the vast body of literature of pharmacogenomic testing and to decipher how to apply this to clinical practice.

In this issue of the Journal, Kitzmiller et al³ provide one solution to this problem, giving an overview of pharmacogenomic applications that might be pertinent to practicing physicians. However, as we try to make pharmacogenomics accessible to busy physicians, we need other solutions to integrate pharmacogenomic information efficiently into the clinical work flow. One approach might be to build pharmacogenomics into the electronic medical record. We can also store the integrated information in research databases and provide clinical recommendations on Internet sites such as www.pharmgkb.org, and we can develop applications to run on cell phones and iPads.

QUESTIONS REMAIN

Kitzmiller et al discuss an important step in this process, highlighting several key questions:

Should we seek genetics-based information to personalize drug selection? Based on the information presented in the literature and in the Kitzmiller paper, there may be circumstances when it is appropriate to consider doing so. While the evidence is not yet compelling to order these tests on a regular basis in clinical practice, this information might be helpful in some situations, such as for patients who have had adverse effects from minimal doses of antidepressants.

For now, clinicians should not abandon their current practice of personalizing patient care on the basis of personal, cultural, and economic preferences. Rather, they should consider pharmacogenomic information an additional piece of information when selecting drug therapy. We should also encourage health care systems and interested providers to be early adopters and to study how their outcomes compare with the standard of care.

Once we have this information, what is our obligation to use it? An increasing number of patients already have genetic information in their health record, either ordered by or provided to their physicians. However, there is little in the scientific literature to guide us in this arena.

Yet most of us would agree that if we have information (genetic or otherwise) that can help to select a drug type or dose or reduce adverse events or costs, we should consider this information in our decision-making. Several circumstances are documented in this paper and in the literature in which prior knowledge about drug metabolism can help in selecting a dose of medication. One example would be the 50% recommended reduction in tricyclic antidepressant dose if the patient is a CYP2D6 poor metabolizer.⁴

MOVING FORWARD AS A TEAM

In summary, Kitzmiller et al bring to light the promise and the uncertainties that currently exist in the field of pharmacogenomics. While it is unclear if we should incorporate pharmacogenomic tests into standard medical practice at this time, it is clear that this information is becoming more readily available, whether or not we have requested it. Some would argue that, once we have the information, we have an obligation to use it, just as we use other information in our clinical decision-making. This means we need to develop tools and resources to help practitioners evaluate pharmacogenomic data and incorporate it into clinical care in an efficient manner.

The authors also highlight the need for more education about drug metabolism in general, and they cite several instances in which knowledge of drug interactions and metabolism can clearly influence decision-making. An example is paroxetine (Paxil) inhibition of tamoxifen (Nolvadex).⁵

Lastly, regardless of our personal feelings about the clinical usefulness of genetic testing in large populations, we need to work together to determine clinical utility and validity and to develop efficient ways to put into practice findings that could affect patient care. As we move forward, we need to work as a team, utilizing our clinical partners—pharmacists, pharmacologists, metabolism and health information technology experts, and medical geneticists. Working as a team, pooling our resources and tools, we move closer to providing world-class personalized health care.

In many patients, certain drugs do not work as well as expected, whereas in other patients they cause toxic effects, even at lower doses. For some patients, the reason may be genetic.

Sizeable minorities of the population carry genetic variants—polymorphisms— that affect their response to various drugs. Thanks to genetic research, our understanding of the variability of drug response has advanced markedly in the last decade. Many relevant polymorphisms have been identified, and tests for some of them are available.

See related editorial

Armed with the knowledge of their patients’ genetic status, physicians could predict their response to certain drugs, leading to better efficacy, fewer adverse drug reactions, and a better cost-benefit ratio.

The possible impact is substantial, since more than half of the drugs most commonly involved in adverse drug reactions are metabolized by polymorphic enzymes.¹ Adverse drug reactions remain a significant detriment to public health, having a substantial impact on rates of morbidity and death and on healthcare costs.^2–8 In the United States, adverse drug reactions are a leading cause of death in hospitalized patients⁴ and are annually responsible for hundreds of thousands of deaths and hundreds of billions of dollars in added costs.^2,4,6–8

In the meantime, physicians can educate their patients and promote efforts to incorporate genomic information into standard clinical decision-making.

This article offers an overview of pharmacogenomic testing, discussing implications and limitations of a few validated tests. Specifically, we will discuss testing that is relevant when using warfarin (Coumadin), clopidogrel (Plavix), statins, tamoxifen (Nolvadex), codeine, and psychotropic medications, as well as the future role of pharmacogenomic testing in medicine.

WHAT IS PHARMACOGENOMICS?

Pharmacogenomics is the study of how genetic factors relate to interindividual variability of drug response.

Many clinicians may not be familiar with the background and terminology used in the pharmacogenomic literature. Below, a brief review of the terminology is followed by a schematic describing the various stages of research involved in pharmacogenomics and the advancement of a test into standard practice.

The review and schematic may be helpful for evaluating the clinical significance of pharmacogenomics-related articles.

From genotype to phenotype

Genotype refers to the coding sequence of DNA base pairs for a particular gene, and phenotype (eg, disease or drug response) refers to a trait resulting from the protein product encoded by the gene. The name of a gene often refers to its protein product and is italicized (eg, the CYP3A4 gene encodes for the CYP3A4 enzyme).

Two alleles per autosomal gene (one paternal and one maternal) form the genotype. Heterozygotes possess two different alleles, and homozygotes possess two of the same alleles. The most common allele in a population is referred to as the wild type, and allele frequencies can vary greatly in different populations.⁹

Most sequence variations are single nucleotide polymorphisms (SNPs, pronounced “snips”), a single DNA base pair substitution that may result in a different gene product. SNPs can be classified as structural RNA polymorphisms (srSNPs), regulatory polymorphisms (rSNPs), or polymorphisms in coding regions (cSNPs)¹⁰: srSNPs alter mRNA processing and translation, rSNPs alter transcription, and cSNPs alter protein sequence and function.

Recently, genetic associations with a phenotype have been done on a large scale, with millions of SNPs measured in each of many subjects. This approach, called a genomewide association study or GWAS, has revealed countless candidate genes for clinical traits, but only a few have resulted in a practical clinical application. SNPs may by themselves exert a pharmacokinetic effect (ie, how the body processes the drug), a pharmacodynamic effect (ie, how the drug affects the body), or both, or they may act in concert with other genetic factors. Pharmacodynamic effects can result from a pharmacokinetic effect or can result from variations in a pharmacologic target.

Establishing a genotype-phenotype association can involve clinical studies, animal transgenic studies, or molecular and cellular functional assays.

Clinical applications are emerging

WARFARIN: IMPORTANCE OF CYP2C9, VKORC1

Warfarin is used for the long-term treatment and prevention of thromboembolic events.

This drug has a narrow therapeutic window and shows substantial interpatient dose variability. The start of warfarin therapy is associated with one of the highest rates of adverse events and emergency room visits of any single drug.¹² More than 2 million patients start warfarin each year in the United States alone,¹³ and about 20% of them are hospitalized within the first 6 months because of bleeding due to overanticoagulation.¹⁴

The findings from a recent study suggest that pharmacogenomic testing may eventually allow more patients to safely benefit from warfarin therapy. In this large, nationwide, prospective study, hospitalization rates were 30% lower when pharmacogenomic testing was used.¹⁴ However, no reduction was seen in the time needed to reach the target international normalized ratio (INR) or in the need for INR checks at 6 months. Furthermore, this study used historical control data, and some or all of the reduction in hospitalization rates may be attributed to more frequent INR checks in the patients who underwent testing than in the historical control group.

A relationship between warfarin dose requirements and the genetic status of CYP2C9, which encodes a major drug-metabolizing enzyme, has been demonstrated in retrospective and prospective studies.^15–17

S-warfarin is metabolized by CYP2C9, which is polymorphic

Warfarin contains equal amounts of two isomers, designated S and R. S-warfarin, which is more potent, is metabolized principally by CYP2C9, while R-warfarin is metabolized by CYP1A2, CYP2C19, and CYP3A4.

People who possess two copies of the wild type CYP2C9 gene CYP2C9*1 metabolize warfarin very well and so are called “extensive warfarin metabolizers.” Carriers of the allelic variants CYP2C9*2 and CYP2C9*3 (which have point mutations in exons 3 and 7 of CYP2C9, respectively), have less capacity. Compared with those who are homozygous for the wild-type gene, homozygous carriers of CYP2C9* 3 clear S-warfarin at a rate that is 90% lower, and those with the CYP2C9*1/*3, CYP2C9* 1/*2, CYP2C9*2/*2, or CYP2C9*2/*3 genotypes clear it at a rate 50% to 75% lower. A meta-analysis of 12 studies found that the CYP2C9 genotype accounted for 12% of the interindividual variability of warfarin dose requirements.¹⁸

About 8% of whites carry at least one copy of CYP2C9*2, as do 1% of African Americans; the allele is rare in Asian populations. The frequency of CYP2C9*3 is 6% in whites, 1% in African Americans, and 3% in Asians.^19,20 People with CYP2C9*4 or CYP2C9*5 have a diminished capacity to clear warfarin; however, these variants occur so infrequently that their clinical relevance may be minimal.

Warfarin’s target, VKOR, is also polymorphic

Genetic variation in warfarin’s pharmacologic target, vitamin K 2,3-epoxide reductase (VKOR), also influences dose requirements. Warfarin decreases the synthesis of vitamin-K-dependent clotting factors by inhibiting VKOR. This inhibition depends on the patient’s C1 subunit gene, VKORC1. Patients with a guanine-to-adenine SNP 1,639 bases upstream of VKORC1 (−1639G>A) need lower warfarin doses—an average of 25% lower in those with the GA genotype (ie, one allele has guanine in the −1639 position and the other allele has adenine in that position) and 50% lower in those with the AA genotype compared with the wild-type genotype GG.²¹ This promoter SNP, positioned upstream (ie, before the gene-coding region), greatly influences VKORC1 expression.

A meta-analysis of 10 studies found that the VKORC1 polymorphism accounts for 25% of the interindividual variation in warfarin dose.¹⁸ In one study, the frequency of the AA genotype in a white population was 14%, AG 47%, and GG 39%; in a Chinese population the frequency of AA was 82%, AG 18%, and GG 0.35%.²²

CYP4F2 and GGCX also affect warfarin’s dose requirements

Genetic variations in the enzymes CYP4F2 and gamma-glutamyl carboxylase (GGCX) also influence warfarin dose requirements. Although the data are limited and the effects are smaller than those of CYP2C9 and VKORC1, people with a SNP in CYP4F2 need 8% higher doses of warfarin, while those with a SNP in GGCX need 6% lower doses.²³

CYP2C9 and VKORC1 testing is available

Currently, the clinical pharmacogenetic tests relevant for warfarin use are for CYP2C9 and VKORC1.¹⁰

The FDA has approved four warfarin pharmacogenetic test kits, but most third-party payers are reluctant to reimburse for testing because it is not currently considered a standard of care. Testing typically costs a few hundred dollars, but it should become less expensive as it becomes more commonplace. The current FDA-approved product label for warfarin does not recommend routine pharmacogenomic testing for determining initial or maintenance doses, but it does acknowledge that dose requirements are influenced by CYP2C9 and VKORC1 and states that genotype information, when available, can assist in selecting the starting dose.²⁴

Clinical trials of warfarin pharmacogenomic testing are under way

Although genetic status can greatly influence an individual patient’s warfarin dosing requirement, routine prospective pharmacogenomic testing is not endorsed by the FDA or by other expert panels²⁶ because there is currently insufficient evidence to recommend for or against it.

Several large prospective trials are under way. For example, the National Heart, Lung, and Blood Institute began a prospective trial in about 1,200 patients to evaluate the use of clinical plus genetic information to guide the initiation of warfarin therapy and to improve anticoagulation control for patients.²⁷ The results, expected in September 2011, and those of other large prospective trials should provide adequate evidence for making recommendations about the clinical utility of routine pharmacogenetic testing for guiding warfarin therapy.

Several recent cost-utility and cost-effectiveness studies have attempted to quantify the value of pharmacogenomic testing for warfarin therapy,^28–30 but their analyses are largely limited because the benefit (clinical utility) is yet to be sufficiently characterized.

The relevance of such analyses may soon be drastically diminished, as several non-vitamin-K-dependent blood thinners such as rivaroxaban (Xarelto), dabigatran (Pradaxa), and apixaban are poised to enter clinical practice.³¹

CLOPIDOGREL IS ACTIVATED BY CYP2C19

Clopidogrel, taken by about 40 million patients worldwide, is used to prevent atherothrombotic events and cardiac stent thrombosis when given along with aspirin.

Studies of clopidogrel pharmacogenomics

A recent genome-wide association study conducted in a cohort of 429 healthy Amish persons revealed a SNP in CYP2C19 to be associated with a diminished response to clopidogrel and to account for 12% of the variation in drug response.³³ Traditional factors (the patient’s age, body-mass index, and cholesterol level) combined accounted for less than 10% of the variation.

Findings were similar in a subsequent investigation in 227 cardiac patients receiving clopidogrel: 21% of those with the variant had a cardiovascular ischemic event or died during a 1-year follow-up period compared with 10% of those without the variant (hazard ratio 2.42, P = .02).³³

A 12-year prospective study investigating clopidogrel efficacy in 300 cardiac patients under the age of 45 used cardiovascular death, nonfatal myocardial infarction, and urgent coronary revascularization as end points. It concluded that the only independent predictor of these events was the patient’s CYP2C19 status.³⁴

A study in 2,200 patients with recent myocardial infarction examined whether any of the known allelic variations that modulate clopidogrel’s absorption (ABCB1), metabolic activation (CYP3A4/5 and CYP2C19), or biologic activity (P2RY12 and ITGB3) was associated with a higher rate of the combined end point of all-cause mortality, nonfatal myocardial infarction, or stroke. None of the SNPs in CYP3A4/5, P2RY12, or ITGB3 that were evaluated was associated with a higher risk at 1 year. However, the allelic variations modulating clopidogrel’s absorption (ABCB1) and metabolism (CYP2C19) were associated with higher event rates. Patients with two variant ABCB1 alleles had a higher adjusted hazard ratio (95% confidence interval [CI] 1.2–2.47) than those with the wild-type allele. Patients who had one or two CYP2C19 loss-of-function alleles had a higher event rate than those with two wild-type alleles (95% CI 1.10–3.58 and 1.71–7.51, respectively).³⁵

Conversely, researchers from the Population Health Research Institute found no association between poor-metabolizer status and treatment outcomes when CYP2C19 analysis was retrospectively added to the findings of two large clinical trials (combined N > 5,000). However, patients with acute coronary syndrome benefited more from clopidogrel treatment if they were ultra-rapid metabolizers (possessing the gain-of-function allele CYP2C19*17).³⁶

Current status of clopidogrel testing: Uncertain

A current FDA boxed warning states that poor CYP2C19 metabolizers may not benefit from clopidogrel and recommends that prescribers consider alternative treatment for patients in this category.³⁷ However, routine CYP2C19 testing is not recommended, and no firm recommendations have been established regarding dose adjustments for CYP2C19 status.

In 2010, the American College of Cardiology Foundation Task Force on Clinical Expert Consensus Documents and the American Heart Association collectively pronounced the current evidence base insufficient for recommending routine pharmacogenomic testing.³⁹

Needed are large-scale studies examining the cost-effectiveness and clinical utility of genotype-guided clopidogrel therapy compared with other therapy options such as prasugrel (Effient), an analogue not metabolized by CYP2C19. One such study, sponsored by Medco Health Solutions, plans to enroll 14,600 cardiac patients and has an estimated completion date in June 2011.⁴⁰ The expectation that clopidogrel will be available in generic form in 2012 adds to the uncertainty regarding the cost-effectiveness of CYP2C19 testing for clopidogrel therapy.

STATINS: SLC01B1*5 INCREASES MYOPATHY RISK

Statins lower the concentration of low-density lipoprotein cholesterol (LDL-C), resulting in a relative-risk reduction of about 20% for each 1 mmol/L (39 mg/dL) decrement in LDL-C.⁴¹ They are one of the most commonly prescribed classes of drugs, but their side effects can limit their appeal: statin-induced myopathy occurs in about 1:1,000 to 1:10,000 patients and is difficult to predict.

SLC01B1. The Study of the Effectiveness of Additional Reductions in Cholesterol and Homocysteine (SEARCH), a genome-wide association study, recently found a SNP (SLCO1B1* 5) in the SLC01B1 gene to be associated with a higher risk of statin-induced myopathy in cardiac patients receiving simvastatin (Zocor) 40 or 80 mg daily.⁴² The SLC01B1 gene, located on chromosome 12, influences the extent of the drug’s hepatic uptake and its serum concentration. Only the SLC01B1*5 SNP emerged as a predictor of statin-induced myopathy across the entire genome.⁴²

The authors believe the findings are likely to apply to other statins. The mechanisms leading to statin-induced myopathy and the impact of statin pharmacogenomics are still unclear.⁴³

CYP3A4. Other genetic variants may play a vital role in determining response to statin therapy. Carriers of a newly identified CYP3A4 polymorphism (intron 6 SNP, rs35599367, C>T) required significantly lower statin doses (0.2–0.6 times less) for optimal lipid control. The analyses included atorvastatin (Lipitor), simvastatin, and lovastatin (Mevacor), and the association was robust (P = .019).⁴⁴

Statin pharmacogenomic testing is not routinely recommended

Routine pharmacogenomic testing for statin therapy is not recommended. Additional studies are needed to determine the clinical utility and cost-effectiveness of pharmacogenomic testing (involving a combination of several polymorphisms) in various patient populations delineated by type of statin, dose, and concomitant use of other drugs.

TAMOXIFEN IS ACTIVATED BY CYP2D6

Tamoxifen is prescribed to prevent the recurrence of estrogen-receptor-positive breast cancer, to treat metastatic breast cancer, to prevent cancer in high-risk populations, and to treat ductal carcinoma in situ.

Tamoxifen is metabolized to form endoxifen, which has much higher potency and higher systemic levels than tamoxifen.⁴⁵ Both CYP2D6 and CYP3A4/5 are required to produce endoxifen via two intermediates, but CYP2D6 catalyzes the critical step leading to metabolic activation.

The CYP2D6 gene is highly polymorphic, with more than 75 allelic variants identified. Extensive literature is available describing the influence of CYP2D6 polymorphisms on tamoxifen metabolism and therapy outcomes.^46–52 Several CYP2D6 variants result in reduced or no enzyme activity, and people who have more than two normally functioning alleles have exaggerated enzyme activity (gene amplification).

Classification of CYP2D6 status

Several systems have been developed to categorize the phenotypic activity of CYP2D6 based on genotype.

A genetic basis for the observed diversity in the metabolism of cytochrome P450 substrates was recognized more than 30 years ago. People were categorized as either extensive or poor metabolizers, reflecting normal vs impaired ability to metabolize the CYP2D6 substrates sparteine and debrisoquine. Later work expanded this system to include categories for intermediate (between poor and extensive) and ultra-rapid (better than extensive) metabolizers.

The genetic basis for these categories includes homozygosity for dysfunctional variants (the poor-metabolizer group) or extra copies of normal functioning variants (the ultra-rapid-metabolizer group).

Newer systems have been described for characterizing the CYP2D6 activity phenotype whereby CYP2D6 variants are assigned activity scores.^53–56 The various scoring systems have been reviewed by Kirchheiner.⁵⁷

A recent version of the activity scoring system also takes into consideration the many drugs that inhibit CYP2D6, such as amiodarone (Cordarone) and fluoxetine (Prozac) that can reduce the action of tamoxifen if given with it (Table 4).⁵⁸ For example, the tamoxifen exposure (as predicted by the CYP2D6-activity score) experienced by a CYP2D6 extensive metabolizer taking a CYP2D6-inhibiting drug may be similar to the exposure experienced by a CYP2D6 poor metabolizer receiving the same tamoxifen dose but not taking a CYP2D6-inhibiting drug.

Likewise, the activity score of a CYP2D6 intermediate metabolizer taking a CYP2D6-inducing drug may be similar to that of a CYP2D6 ultra-rapid metabolizer not taking a CYP2D6-inducing drug. Examples of CYP2D6 inducers are dexamethasone, rifampin, and hyperforin (St. John’s wort).

While the newer systems are reported to provide better correlations between genotype and phenotype scores, the older scoring systems and the categorical names are still widely used (eg, in the FDA-approved AmpliChip CYP450 test from Roche,⁵⁹ which includes genotype data for CYP2D6 and CYP2C19).

No firm recommendations for CYP2D6 testing in tamoxifen users

The different genotypes and phenotypes vary in prevalence in different ethnic groups, and significantly different activity levels for endoxifen formation are observed. Tamoxifen lacks efficacy in those who are poor CYP2D6 metabolizers—ie, about 7% of the white population.

However, the FDA has not made firm recommendations about CYP2D6 testing for prescribing tamoxifen because the evidence of benefit, although suggestive, has been considered insufficient.

Clinicians should be aware that tamoxifen’s efficacy is greatly reduced by concomitant therapy with CYP2D6-inhibiting drugs (Table 4).

Other genes affecting tamoxifen: CYP3A4/5, SULT1A1, and UGT2B15

Some investigators propose that polymorphisms in additional genes encoding enzymes in the tamoxifen metabolic and elimination pathways (eg, CYP3A4/5, SULT1A1, and UGT2B15) also need to be considered to account adequately for interindividual variation in drug response.

For example, CYP3A4 and CYP3A5 are also polymorphic, and large interindividual variation exists in their enzyme activities. These enzymes have overlapping substrate specificities, represent the most abundant drug-metabolizing enzymes in the human liver, and are involved in the biotransformation of a broad range of endogenous substrates and most drugs.⁶⁰

Clinical studies evaluating the impact of CYP3A4/5 polymorphisms have been inconsistent in their conclusions, which is generally attributed to the relatively low functional impact or the low prevalence of the SNPs evaluated. Many of the nearly 100 CYP3A4/5 polymorphisms identified have not yet been characterized regarding their functional impact on enzyme expression or activity. CYP-3A4/5 enzyme activity is highly variable between individuals and warrants further study of its role in outcomes of tamoxifen therapy. Ongoing and future prospective clinical trials evaluating CYP2D6, CYP3A4/5, and other relevant polymorphisms are necessary to define their clinical relevance before routine genetic testing for tamoxifen can be justified.

CODEINE IS ALSO ACTIVATED BY CYP2D6

Codeine also depends on the CYP2D6 gene, as it must be activated to its more potent opioid metabolites, including morphine. Poor CYP2D6 metabolizers do not benefit from codeine therapy.

The pharmacogenomics of codeine has become a hot topic, especially regarding breast-feeding mothers. The debate was ignited with the publication in 2006 of a case report of an infant’s death, apparently the result of metabolic polymorphisms.⁶¹ The evolution of this debate and the outcome of the case may be noteworthy to clinicians, as they illustrate the gravity of public and patient interest in pharmacogenomic testing. In this case, the breast-feeding mother had taken codeine regularly for about 14 days when her 13-day-old infant died from toxic levels of morphine. Unknown to her and the prescriber, both the mother and infant were ultra-rapid CYP2D6 metabolizers, resulting in a more rapid and extensive conversion of codeine to morphine.

A logical strategy for preventing similar deaths would be routine CYP2D6 genotyping when prescribing codeine to breast-feeding mothers. However, after several investigations examined the metabolic and excretion pathways of codeine in their entirety, the FDA did not recommend routine CYP2D6 testing when prescribing codeine to breastfeeding mothers because several other factors, including rare genetic variations of other enzymes, proved necessary for reaching the opioid toxicity leading to the infant’s death.⁶²

PHARMACOGENOMICS OF PSYCHOTROPIC DRUGS

Pharmacogenomic testing has clinical utility for some psychotropic drugs.

HLA-B and carbamazepine

Considered a standard of care, HLA-B genotyping is appropriate before prescribing carbamazepine (Tegretol, Equetro) to patients in populations in which HLAB*1502 is likely to be present, such as Asians. Carriers of HLAB* 1502 are at higher risk of life-threatening skin reactions such as Stevens-Johnson syndrome.¹¹

Several other pharmacogenomic applications for psychotropic medications have been suggested, but routine testing has not been recommended by the FDA or endorsed by any expert panel because sufficient clinical utility and cost-effectiveness have not been demonstrated. A brief summary of study findings and a few practical suggestions follow.

Polymorphisms in metabolizing enzymes have been investigated in patients receiving psychotropic drugs.

CYP2D6 and antidepressants

Many antidepressants show significant differences in plasma drug levels with CYP2D6 polymorphisms (in descending order of influence)⁵⁵:

• Imipramine (Tofranil)
• Doxepin (Adapin, Silenor, Sinequan)
• Maprotiline (Deprilept, Ludiomil, Psymion)
• Trimipramine (Surmontil)
• Desipramine (Noraprim)
• Nortriptyline (Aventyl, Pamelor)
• Clomipramine (Anafranil)
• Paroxetine (Paxil)
• Venlafaxine (Effexor)
• Amitriptyline (Elavil)
• Mianserin
• Trazadone (Desyrel)
• Bupropion (Wellbutrin)
• Nefazodone (Serzone)
• Citalopram (Celexa)
• Sertraline (Zoloft).

CYP2D6 and antipsychotics

Several antipsychotics are also influenced by CYP2D6 polymorphisms (also in descending order of influence)⁵⁵:

• Perphenazine (Trilafon)
• Thioridazine (Mellaril)
• Olanzapine (Zyprexa)
• Zuclopenthixol (Cisordinol, Clopixol, Acuphase)
• Aripiprazole (Abilify)
• Flupentixol (Depixol, Fluanxol)
• Haloperidol (Haldol)
• Perazine (Taxilan)
• Risperidone (Risperdal)
• Pimozide (Orap).

CYP2C19 and antidepressants

CYP2C19 polymorphisms are likewise associated with differences in drug metabolism for many antidepressants, such as (in descending order of CYP2C19-mediated influence)⁵⁵:

• Trimipramine
• Doxepin
• Amitriptyline
• Imipramine
• Citalopram (Celexa)
• Clomipramine
• Moclobemide (Aurorix, Manerix)
• Sertraline
• Fluvoxamine (Luvox).

Clinical relevance of CYP2D6 and CYP2C19

Several studies have demonstrated that poor and intermediate CYP2D6 metabolizers have a higher incidence of adverse effects when taking CYP2D6-dependent antidepressants^63–68; however, an almost equal number of studies did not find statistically significant associations.^69–72 Likewise, several studies have found an association between ultra-rapid CYP2D6 metabolizer status and diminished response to antidepressants,^65,73,74 but no association was found in a larger retrospective study.⁷⁵

Routine CYP2D6 and CYP2C19 screening is not recommended when prescribing psychotropic drugs. However, reviews of the pharmacokinetic data have suggested a few practical applications when genetic status is already known. In general, clinicians can consider reducing the dose of tricyclic antidepressants by about 50% when prescribing to CYP2D6-poor-metabolizers.^55,76–78

Genes that affect serotonin metabolism

Several genes in the serotonin pathway have been investigated to determine whether they influence patients’ susceptibility to depression and adverse effects and response to psychotropic medications.

SLC6A4. Polymorphisms in the promoter region of the serotonin transporter gene SLC6A4 appear to influence the treatment response and side-effect profiles of selective serotonin reuptake inhibitors (SSRIs). Carriers of the SLC6A4 5-HTTLPR L alleles have fewer side effects⁷⁹ and better response to SSRI treatment, and carriers of the S allele have a higher incidence of antidepressant-induced mania⁸⁰ and poorer response to SSRI treatment.⁸¹

5-HT. Polymorphisms in serotonin receptors (2A and 2C subtypes) appear to influence SSRI response and side effects. Carriers of 5-HT 2A C alleles had more severe adverse effects from paroxetine,⁷¹ but another 5-HT 2A polymorphism common to Asians is associated with better response to antidepressant therapy.⁸² A 5-HT 2C polymorphism was associated with a lower incidence of antipsychotic-induced weight gain.⁸³

Although the understanding of these relationships is incomplete and routine pharmacogenomic testing is not currently recommended, reviews of the pharmacodynamic data have suggested a few practical applications when a patient’s genetic status is already known. One should consider:

• Selecting treatments other than SSRIs for depressed patients known to possess the SLC6A4 variant
• Selecting citalopram for depressed patients known to carry the 5-HT 2A polymorphism
• Avoiding treatment with antipsychotic drugs for patients known to possess the 5-HT 2C polymorphism.

THE FUTURE OF PHARMACOGENOMIC TESTING

The examples discussed in this article provide some insight about how pharmacogenomic testing is maturing and slowly being integrated into the practice of medicine. They also illustrate the complexity of the multiple stages of research that pharmacogenomic applications must go through in order to be adopted as standard practice.

In the future, pharmacogenomic data will continue to accumulate, and the clinical utility of many other pharmacogenomic tests may be uncovered. The FDA provides information on emerging pharmacogenomic tests at its Web site, www.fda.gov.¹¹ Its up-to-date “Table of Valid Genomic Biomarkers in the Context of Approved Drug Labels” includes boxed warnings, recommendations, research outcomes, and relevant population genetics.

If the FDA continues its current policy, prospective randomized trials that show improvement in patient outcomes will remain the gold standard for determining the clinical significance of a pharmacogenomic test. Furthermore, cost-benefit analyses are likely to continue dictating policy regarding pharmacogenomic testing, and cost-benefit profiles should improve as technology advances and as information gathered from a single test becomes applicable to multiple medications and clinical scenarios.

In the meantime, physicians should become familiar with the terms used in medical genetics and pharmacogenomics and begin to understand genetic contributions to the outcomes of drug therapy. For example, understanding the consequences of metabolizer status and the frequency of variants in a given population can be tremendously helpful when advising our patients about anticipating potential problems when taking specific medications and about making informed decisions about pharmacogenomic testing.

This exchange of information alone may go a long way in improving therapy outcomes even when prospective pharmacogenomic testing is not routinely performed. Furthermore, an increasing number of patients will already have genotyping information available when they come to us, and clinicians need to be aware of the many pharmacogenomic applications recommended by the FDA when genetic status is known.¹⁰

In many patients, certain drugs do not work as well as expected, whereas in other patients they cause toxic effects, even at lower doses. For some patients, the reason may be genetic.

Sizeable minorities of the population carry genetic variants—polymorphisms— that affect their response to various drugs. Thanks to genetic research, our understanding of the variability of drug response has advanced markedly in the last decade. Many relevant polymorphisms have been identified, and tests for some of them are available.

See related editorial

Armed with the knowledge of their patients’ genetic status, physicians could predict their response to certain drugs, leading to better efficacy, fewer adverse drug reactions, and a better cost-benefit ratio.

The possible impact is substantial, since more than half of the drugs most commonly involved in adverse drug reactions are metabolized by polymorphic enzymes.¹ Adverse drug reactions remain a significant detriment to public health, having a substantial impact on rates of morbidity and death and on healthcare costs.^2–8 In the United States, adverse drug reactions are a leading cause of death in hospitalized patients⁴ and are annually responsible for hundreds of thousands of deaths and hundreds of billions of dollars in added costs.^2,4,6–8

In the meantime, physicians can educate their patients and promote efforts to incorporate genomic information into standard clinical decision-making.

This article offers an overview of pharmacogenomic testing, discussing implications and limitations of a few validated tests. Specifically, we will discuss testing that is relevant when using warfarin (Coumadin), clopidogrel (Plavix), statins, tamoxifen (Nolvadex), codeine, and psychotropic medications, as well as the future role of pharmacogenomic testing in medicine.

WHAT IS PHARMACOGENOMICS?

Pharmacogenomics is the study of how genetic factors relate to interindividual variability of drug response.

Many clinicians may not be familiar with the background and terminology used in the pharmacogenomic literature. Below, a brief review of the terminology is followed by a schematic describing the various stages of research involved in pharmacogenomics and the advancement of a test into standard practice.

The review and schematic may be helpful for evaluating the clinical significance of pharmacogenomics-related articles.

From genotype to phenotype

Genotype refers to the coding sequence of DNA base pairs for a particular gene, and phenotype (eg, disease or drug response) refers to a trait resulting from the protein product encoded by the gene. The name of a gene often refers to its protein product and is italicized (eg, the CYP3A4 gene encodes for the CYP3A4 enzyme).

Two alleles per autosomal gene (one paternal and one maternal) form the genotype. Heterozygotes possess two different alleles, and homozygotes possess two of the same alleles. The most common allele in a population is referred to as the wild type, and allele frequencies can vary greatly in different populations.⁹

Most sequence variations are single nucleotide polymorphisms (SNPs, pronounced “snips”), a single DNA base pair substitution that may result in a different gene product. SNPs can be classified as structural RNA polymorphisms (srSNPs), regulatory polymorphisms (rSNPs), or polymorphisms in coding regions (cSNPs)¹⁰: srSNPs alter mRNA processing and translation, rSNPs alter transcription, and cSNPs alter protein sequence and function.

Recently, genetic associations with a phenotype have been done on a large scale, with millions of SNPs measured in each of many subjects. This approach, called a genomewide association study or GWAS, has revealed countless candidate genes for clinical traits, but only a few have resulted in a practical clinical application. SNPs may by themselves exert a pharmacokinetic effect (ie, how the body processes the drug), a pharmacodynamic effect (ie, how the drug affects the body), or both, or they may act in concert with other genetic factors. Pharmacodynamic effects can result from a pharmacokinetic effect or can result from variations in a pharmacologic target.

Establishing a genotype-phenotype association can involve clinical studies, animal transgenic studies, or molecular and cellular functional assays.

Clinical applications are emerging

WARFARIN: IMPORTANCE OF CYP2C9, VKORC1

Warfarin is used for the long-term treatment and prevention of thromboembolic events.

This drug has a narrow therapeutic window and shows substantial interpatient dose variability. The start of warfarin therapy is associated with one of the highest rates of adverse events and emergency room visits of any single drug.¹² More than 2 million patients start warfarin each year in the United States alone,¹³ and about 20% of them are hospitalized within the first 6 months because of bleeding due to overanticoagulation.¹⁴

The findings from a recent study suggest that pharmacogenomic testing may eventually allow more patients to safely benefit from warfarin therapy. In this large, nationwide, prospective study, hospitalization rates were 30% lower when pharmacogenomic testing was used.¹⁴ However, no reduction was seen in the time needed to reach the target international normalized ratio (INR) or in the need for INR checks at 6 months. Furthermore, this study used historical control data, and some or all of the reduction in hospitalization rates may be attributed to more frequent INR checks in the patients who underwent testing than in the historical control group.

A relationship between warfarin dose requirements and the genetic status of CYP2C9, which encodes a major drug-metabolizing enzyme, has been demonstrated in retrospective and prospective studies.^15–17

S-warfarin is metabolized by CYP2C9, which is polymorphic

Warfarin contains equal amounts of two isomers, designated S and R. S-warfarin, which is more potent, is metabolized principally by CYP2C9, while R-warfarin is metabolized by CYP1A2, CYP2C19, and CYP3A4.

People who possess two copies of the wild type CYP2C9 gene CYP2C9*1 metabolize warfarin very well and so are called “extensive warfarin metabolizers.” Carriers of the allelic variants CYP2C9*2 and CYP2C9*3 (which have point mutations in exons 3 and 7 of CYP2C9, respectively), have less capacity. Compared with those who are homozygous for the wild-type gene, homozygous carriers of CYP2C9* 3 clear S-warfarin at a rate that is 90% lower, and those with the CYP2C9*1/*3, CYP2C9* 1/*2, CYP2C9*2/*2, or CYP2C9*2/*3 genotypes clear it at a rate 50% to 75% lower. A meta-analysis of 12 studies found that the CYP2C9 genotype accounted for 12% of the interindividual variability of warfarin dose requirements.¹⁸

About 8% of whites carry at least one copy of CYP2C9*2, as do 1% of African Americans; the allele is rare in Asian populations. The frequency of CYP2C9*3 is 6% in whites, 1% in African Americans, and 3% in Asians.^19,20 People with CYP2C9*4 or CYP2C9*5 have a diminished capacity to clear warfarin; however, these variants occur so infrequently that their clinical relevance may be minimal.

Warfarin’s target, VKOR, is also polymorphic

Genetic variation in warfarin’s pharmacologic target, vitamin K 2,3-epoxide reductase (VKOR), also influences dose requirements. Warfarin decreases the synthesis of vitamin-K-dependent clotting factors by inhibiting VKOR. This inhibition depends on the patient’s C1 subunit gene, VKORC1. Patients with a guanine-to-adenine SNP 1,639 bases upstream of VKORC1 (−1639G>A) need lower warfarin doses—an average of 25% lower in those with the GA genotype (ie, one allele has guanine in the −1639 position and the other allele has adenine in that position) and 50% lower in those with the AA genotype compared with the wild-type genotype GG.²¹ This promoter SNP, positioned upstream (ie, before the gene-coding region), greatly influences VKORC1 expression.

A meta-analysis of 10 studies found that the VKORC1 polymorphism accounts for 25% of the interindividual variation in warfarin dose.¹⁸ In one study, the frequency of the AA genotype in a white population was 14%, AG 47%, and GG 39%; in a Chinese population the frequency of AA was 82%, AG 18%, and GG 0.35%.²²

CYP4F2 and GGCX also affect warfarin’s dose requirements

Genetic variations in the enzymes CYP4F2 and gamma-glutamyl carboxylase (GGCX) also influence warfarin dose requirements. Although the data are limited and the effects are smaller than those of CYP2C9 and VKORC1, people with a SNP in CYP4F2 need 8% higher doses of warfarin, while those with a SNP in GGCX need 6% lower doses.²³

CYP2C9 and VKORC1 testing is available

Currently, the clinical pharmacogenetic tests relevant for warfarin use are for CYP2C9 and VKORC1.¹⁰

The FDA has approved four warfarin pharmacogenetic test kits, but most third-party payers are reluctant to reimburse for testing because it is not currently considered a standard of care. Testing typically costs a few hundred dollars, but it should become less expensive as it becomes more commonplace. The current FDA-approved product label for warfarin does not recommend routine pharmacogenomic testing for determining initial or maintenance doses, but it does acknowledge that dose requirements are influenced by CYP2C9 and VKORC1 and states that genotype information, when available, can assist in selecting the starting dose.²⁴

Clinical trials of warfarin pharmacogenomic testing are under way

Although genetic status can greatly influence an individual patient’s warfarin dosing requirement, routine prospective pharmacogenomic testing is not endorsed by the FDA or by other expert panels²⁶ because there is currently insufficient evidence to recommend for or against it.

Several large prospective trials are under way. For example, the National Heart, Lung, and Blood Institute began a prospective trial in about 1,200 patients to evaluate the use of clinical plus genetic information to guide the initiation of warfarin therapy and to improve anticoagulation control for patients.²⁷ The results, expected in September 2011, and those of other large prospective trials should provide adequate evidence for making recommendations about the clinical utility of routine pharmacogenetic testing for guiding warfarin therapy.

Several recent cost-utility and cost-effectiveness studies have attempted to quantify the value of pharmacogenomic testing for warfarin therapy,^28–30 but their analyses are largely limited because the benefit (clinical utility) is yet to be sufficiently characterized.

The relevance of such analyses may soon be drastically diminished, as several non-vitamin-K-dependent blood thinners such as rivaroxaban (Xarelto), dabigatran (Pradaxa), and apixaban are poised to enter clinical practice.³¹

CLOPIDOGREL IS ACTIVATED BY CYP2C19

Clopidogrel, taken by about 40 million patients worldwide, is used to prevent atherothrombotic events and cardiac stent thrombosis when given along with aspirin.

Studies of clopidogrel pharmacogenomics

A recent genome-wide association study conducted in a cohort of 429 healthy Amish persons revealed a SNP in CYP2C19 to be associated with a diminished response to clopidogrel and to account for 12% of the variation in drug response.³³ Traditional factors (the patient’s age, body-mass index, and cholesterol level) combined accounted for less than 10% of the variation.

Findings were similar in a subsequent investigation in 227 cardiac patients receiving clopidogrel: 21% of those with the variant had a cardiovascular ischemic event or died during a 1-year follow-up period compared with 10% of those without the variant (hazard ratio 2.42, P = .02).³³

A 12-year prospective study investigating clopidogrel efficacy in 300 cardiac patients under the age of 45 used cardiovascular death, nonfatal myocardial infarction, and urgent coronary revascularization as end points. It concluded that the only independent predictor of these events was the patient’s CYP2C19 status.³⁴

A study in 2,200 patients with recent myocardial infarction examined whether any of the known allelic variations that modulate clopidogrel’s absorption (ABCB1), metabolic activation (CYP3A4/5 and CYP2C19), or biologic activity (P2RY12 and ITGB3) was associated with a higher rate of the combined end point of all-cause mortality, nonfatal myocardial infarction, or stroke. None of the SNPs in CYP3A4/5, P2RY12, or ITGB3 that were evaluated was associated with a higher risk at 1 year. However, the allelic variations modulating clopidogrel’s absorption (ABCB1) and metabolism (CYP2C19) were associated with higher event rates. Patients with two variant ABCB1 alleles had a higher adjusted hazard ratio (95% confidence interval [CI] 1.2–2.47) than those with the wild-type allele. Patients who had one or two CYP2C19 loss-of-function alleles had a higher event rate than those with two wild-type alleles (95% CI 1.10–3.58 and 1.71–7.51, respectively).³⁵

Conversely, researchers from the Population Health Research Institute found no association between poor-metabolizer status and treatment outcomes when CYP2C19 analysis was retrospectively added to the findings of two large clinical trials (combined N > 5,000). However, patients with acute coronary syndrome benefited more from clopidogrel treatment if they were ultra-rapid metabolizers (possessing the gain-of-function allele CYP2C19*17).³⁶

Current status of clopidogrel testing: Uncertain

A current FDA boxed warning states that poor CYP2C19 metabolizers may not benefit from clopidogrel and recommends that prescribers consider alternative treatment for patients in this category.³⁷ However, routine CYP2C19 testing is not recommended, and no firm recommendations have been established regarding dose adjustments for CYP2C19 status.

In 2010, the American College of Cardiology Foundation Task Force on Clinical Expert Consensus Documents and the American Heart Association collectively pronounced the current evidence base insufficient for recommending routine pharmacogenomic testing.³⁹

Needed are large-scale studies examining the cost-effectiveness and clinical utility of genotype-guided clopidogrel therapy compared with other therapy options such as prasugrel (Effient), an analogue not metabolized by CYP2C19. One such study, sponsored by Medco Health Solutions, plans to enroll 14,600 cardiac patients and has an estimated completion date in June 2011.⁴⁰ The expectation that clopidogrel will be available in generic form in 2012 adds to the uncertainty regarding the cost-effectiveness of CYP2C19 testing for clopidogrel therapy.

STATINS: SLC01B1*5 INCREASES MYOPATHY RISK

Statins lower the concentration of low-density lipoprotein cholesterol (LDL-C), resulting in a relative-risk reduction of about 20% for each 1 mmol/L (39 mg/dL) decrement in LDL-C.⁴¹ They are one of the most commonly prescribed classes of drugs, but their side effects can limit their appeal: statin-induced myopathy occurs in about 1:1,000 to 1:10,000 patients and is difficult to predict.

SLC01B1. The Study of the Effectiveness of Additional Reductions in Cholesterol and Homocysteine (SEARCH), a genome-wide association study, recently found a SNP (SLCO1B1* 5) in the SLC01B1 gene to be associated with a higher risk of statin-induced myopathy in cardiac patients receiving simvastatin (Zocor) 40 or 80 mg daily.⁴² The SLC01B1 gene, located on chromosome 12, influences the extent of the drug’s hepatic uptake and its serum concentration. Only the SLC01B1*5 SNP emerged as a predictor of statin-induced myopathy across the entire genome.⁴²

The authors believe the findings are likely to apply to other statins. The mechanisms leading to statin-induced myopathy and the impact of statin pharmacogenomics are still unclear.⁴³

CYP3A4. Other genetic variants may play a vital role in determining response to statin therapy. Carriers of a newly identified CYP3A4 polymorphism (intron 6 SNP, rs35599367, C>T) required significantly lower statin doses (0.2–0.6 times less) for optimal lipid control. The analyses included atorvastatin (Lipitor), simvastatin, and lovastatin (Mevacor), and the association was robust (P = .019).⁴⁴

Statin pharmacogenomic testing is not routinely recommended

Routine pharmacogenomic testing for statin therapy is not recommended. Additional studies are needed to determine the clinical utility and cost-effectiveness of pharmacogenomic testing (involving a combination of several polymorphisms) in various patient populations delineated by type of statin, dose, and concomitant use of other drugs.

TAMOXIFEN IS ACTIVATED BY CYP2D6

Tamoxifen is prescribed to prevent the recurrence of estrogen-receptor-positive breast cancer, to treat metastatic breast cancer, to prevent cancer in high-risk populations, and to treat ductal carcinoma in situ.

Tamoxifen is metabolized to form endoxifen, which has much higher potency and higher systemic levels than tamoxifen.⁴⁵ Both CYP2D6 and CYP3A4/5 are required to produce endoxifen via two intermediates, but CYP2D6 catalyzes the critical step leading to metabolic activation.

The CYP2D6 gene is highly polymorphic, with more than 75 allelic variants identified. Extensive literature is available describing the influence of CYP2D6 polymorphisms on tamoxifen metabolism and therapy outcomes.^46–52 Several CYP2D6 variants result in reduced or no enzyme activity, and people who have more than two normally functioning alleles have exaggerated enzyme activity (gene amplification).

Classification of CYP2D6 status

Several systems have been developed to categorize the phenotypic activity of CYP2D6 based on genotype.

A genetic basis for the observed diversity in the metabolism of cytochrome P450 substrates was recognized more than 30 years ago. People were categorized as either extensive or poor metabolizers, reflecting normal vs impaired ability to metabolize the CYP2D6 substrates sparteine and debrisoquine. Later work expanded this system to include categories for intermediate (between poor and extensive) and ultra-rapid (better than extensive) metabolizers.

The genetic basis for these categories includes homozygosity for dysfunctional variants (the poor-metabolizer group) or extra copies of normal functioning variants (the ultra-rapid-metabolizer group).

Newer systems have been described for characterizing the CYP2D6 activity phenotype whereby CYP2D6 variants are assigned activity scores.^53–56 The various scoring systems have been reviewed by Kirchheiner.⁵⁷

A recent version of the activity scoring system also takes into consideration the many drugs that inhibit CYP2D6, such as amiodarone (Cordarone) and fluoxetine (Prozac) that can reduce the action of tamoxifen if given with it (Table 4).⁵⁸ For example, the tamoxifen exposure (as predicted by the CYP2D6-activity score) experienced by a CYP2D6 extensive metabolizer taking a CYP2D6-inhibiting drug may be similar to the exposure experienced by a CYP2D6 poor metabolizer receiving the same tamoxifen dose but not taking a CYP2D6-inhibiting drug.

Likewise, the activity score of a CYP2D6 intermediate metabolizer taking a CYP2D6-inducing drug may be similar to that of a CYP2D6 ultra-rapid metabolizer not taking a CYP2D6-inducing drug. Examples of CYP2D6 inducers are dexamethasone, rifampin, and hyperforin (St. John’s wort).

While the newer systems are reported to provide better correlations between genotype and phenotype scores, the older scoring systems and the categorical names are still widely used (eg, in the FDA-approved AmpliChip CYP450 test from Roche,⁵⁹ which includes genotype data for CYP2D6 and CYP2C19).

No firm recommendations for CYP2D6 testing in tamoxifen users

The different genotypes and phenotypes vary in prevalence in different ethnic groups, and significantly different activity levels for endoxifen formation are observed. Tamoxifen lacks efficacy in those who are poor CYP2D6 metabolizers—ie, about 7% of the white population.

However, the FDA has not made firm recommendations about CYP2D6 testing for prescribing tamoxifen because the evidence of benefit, although suggestive, has been considered insufficient.

Clinicians should be aware that tamoxifen’s efficacy is greatly reduced by concomitant therapy with CYP2D6-inhibiting drugs (Table 4).

Other genes affecting tamoxifen: CYP3A4/5, SULT1A1, and UGT2B15

Some investigators propose that polymorphisms in additional genes encoding enzymes in the tamoxifen metabolic and elimination pathways (eg, CYP3A4/5, SULT1A1, and UGT2B15) also need to be considered to account adequately for interindividual variation in drug response.

For example, CYP3A4 and CYP3A5 are also polymorphic, and large interindividual variation exists in their enzyme activities. These enzymes have overlapping substrate specificities, represent the most abundant drug-metabolizing enzymes in the human liver, and are involved in the biotransformation of a broad range of endogenous substrates and most drugs.⁶⁰

Clinical studies evaluating the impact of CYP3A4/5 polymorphisms have been inconsistent in their conclusions, which is generally attributed to the relatively low functional impact or the low prevalence of the SNPs evaluated. Many of the nearly 100 CYP3A4/5 polymorphisms identified have not yet been characterized regarding their functional impact on enzyme expression or activity. CYP-3A4/5 enzyme activity is highly variable between individuals and warrants further study of its role in outcomes of tamoxifen therapy. Ongoing and future prospective clinical trials evaluating CYP2D6, CYP3A4/5, and other relevant polymorphisms are necessary to define their clinical relevance before routine genetic testing for tamoxifen can be justified.

CODEINE IS ALSO ACTIVATED BY CYP2D6

Codeine also depends on the CYP2D6 gene, as it must be activated to its more potent opioid metabolites, including morphine. Poor CYP2D6 metabolizers do not benefit from codeine therapy.

The pharmacogenomics of codeine has become a hot topic, especially regarding breast-feeding mothers. The debate was ignited with the publication in 2006 of a case report of an infant’s death, apparently the result of metabolic polymorphisms.⁶¹ The evolution of this debate and the outcome of the case may be noteworthy to clinicians, as they illustrate the gravity of public and patient interest in pharmacogenomic testing. In this case, the breast-feeding mother had taken codeine regularly for about 14 days when her 13-day-old infant died from toxic levels of morphine. Unknown to her and the prescriber, both the mother and infant were ultra-rapid CYP2D6 metabolizers, resulting in a more rapid and extensive conversion of codeine to morphine.

A logical strategy for preventing similar deaths would be routine CYP2D6 genotyping when prescribing codeine to breast-feeding mothers. However, after several investigations examined the metabolic and excretion pathways of codeine in their entirety, the FDA did not recommend routine CYP2D6 testing when prescribing codeine to breastfeeding mothers because several other factors, including rare genetic variations of other enzymes, proved necessary for reaching the opioid toxicity leading to the infant’s death.⁶²

PHARMACOGENOMICS OF PSYCHOTROPIC DRUGS

Pharmacogenomic testing has clinical utility for some psychotropic drugs.

HLA-B and carbamazepine

Considered a standard of care, HLA-B genotyping is appropriate before prescribing carbamazepine (Tegretol, Equetro) to patients in populations in which HLAB*1502 is likely to be present, such as Asians. Carriers of HLAB* 1502 are at higher risk of life-threatening skin reactions such as Stevens-Johnson syndrome.¹¹

Several other pharmacogenomic applications for psychotropic medications have been suggested, but routine testing has not been recommended by the FDA or endorsed by any expert panel because sufficient clinical utility and cost-effectiveness have not been demonstrated. A brief summary of study findings and a few practical suggestions follow.

Polymorphisms in metabolizing enzymes have been investigated in patients receiving psychotropic drugs.

CYP2D6 and antidepressants

Many antidepressants show significant differences in plasma drug levels with CYP2D6 polymorphisms (in descending order of influence)⁵⁵:

• Imipramine (Tofranil)
• Doxepin (Adapin, Silenor, Sinequan)
• Maprotiline (Deprilept, Ludiomil, Psymion)
• Trimipramine (Surmontil)
• Desipramine (Noraprim)
• Nortriptyline (Aventyl, Pamelor)
• Clomipramine (Anafranil)
• Paroxetine (Paxil)
• Venlafaxine (Effexor)
• Amitriptyline (Elavil)
• Mianserin
• Trazadone (Desyrel)
• Bupropion (Wellbutrin)
• Nefazodone (Serzone)
• Citalopram (Celexa)
• Sertraline (Zoloft).

CYP2D6 and antipsychotics

Several antipsychotics are also influenced by CYP2D6 polymorphisms (also in descending order of influence)⁵⁵:

• Perphenazine (Trilafon)
• Thioridazine (Mellaril)
• Olanzapine (Zyprexa)
• Zuclopenthixol (Cisordinol, Clopixol, Acuphase)
• Aripiprazole (Abilify)
• Flupentixol (Depixol, Fluanxol)
• Haloperidol (Haldol)
• Perazine (Taxilan)
• Risperidone (Risperdal)
• Pimozide (Orap).

CYP2C19 and antidepressants

CYP2C19 polymorphisms are likewise associated with differences in drug metabolism for many antidepressants, such as (in descending order of CYP2C19-mediated influence)⁵⁵:

• Trimipramine
• Doxepin
• Amitriptyline
• Imipramine
• Citalopram (Celexa)
• Clomipramine
• Moclobemide (Aurorix, Manerix)
• Sertraline
• Fluvoxamine (Luvox).

Clinical relevance of CYP2D6 and CYP2C19

Several studies have demonstrated that poor and intermediate CYP2D6 metabolizers have a higher incidence of adverse effects when taking CYP2D6-dependent antidepressants^63–68; however, an almost equal number of studies did not find statistically significant associations.^69–72 Likewise, several studies have found an association between ultra-rapid CYP2D6 metabolizer status and diminished response to antidepressants,^65,73,74 but no association was found in a larger retrospective study.⁷⁵

Routine CYP2D6 and CYP2C19 screening is not recommended when prescribing psychotropic drugs. However, reviews of the pharmacokinetic data have suggested a few practical applications when genetic status is already known. In general, clinicians can consider reducing the dose of tricyclic antidepressants by about 50% when prescribing to CYP2D6-poor-metabolizers.^55,76–78

Genes that affect serotonin metabolism

Several genes in the serotonin pathway have been investigated to determine whether they influence patients’ susceptibility to depression and adverse effects and response to psychotropic medications.

SLC6A4. Polymorphisms in the promoter region of the serotonin transporter gene SLC6A4 appear to influence the treatment response and side-effect profiles of selective serotonin reuptake inhibitors (SSRIs). Carriers of the SLC6A4 5-HTTLPR L alleles have fewer side effects⁷⁹ and better response to SSRI treatment, and carriers of the S allele have a higher incidence of antidepressant-induced mania⁸⁰ and poorer response to SSRI treatment.⁸¹

5-HT. Polymorphisms in serotonin receptors (2A and 2C subtypes) appear to influence SSRI response and side effects. Carriers of 5-HT 2A C alleles had more severe adverse effects from paroxetine,⁷¹ but another 5-HT 2A polymorphism common to Asians is associated with better response to antidepressant therapy.⁸² A 5-HT 2C polymorphism was associated with a lower incidence of antipsychotic-induced weight gain.⁸³

Although the understanding of these relationships is incomplete and routine pharmacogenomic testing is not currently recommended, reviews of the pharmacodynamic data have suggested a few practical applications when a patient’s genetic status is already known. One should consider:

• Selecting treatments other than SSRIs for depressed patients known to possess the SLC6A4 variant
• Selecting citalopram for depressed patients known to carry the 5-HT 2A polymorphism
• Avoiding treatment with antipsychotic drugs for patients known to possess the 5-HT 2C polymorphism.

THE FUTURE OF PHARMACOGENOMIC TESTING

The examples discussed in this article provide some insight about how pharmacogenomic testing is maturing and slowly being integrated into the practice of medicine. They also illustrate the complexity of the multiple stages of research that pharmacogenomic applications must go through in order to be adopted as standard practice.

In the future, pharmacogenomic data will continue to accumulate, and the clinical utility of many other pharmacogenomic tests may be uncovered. The FDA provides information on emerging pharmacogenomic tests at its Web site, www.fda.gov.¹¹ Its up-to-date “Table of Valid Genomic Biomarkers in the Context of Approved Drug Labels” includes boxed warnings, recommendations, research outcomes, and relevant population genetics.

If the FDA continues its current policy, prospective randomized trials that show improvement in patient outcomes will remain the gold standard for determining the clinical significance of a pharmacogenomic test. Furthermore, cost-benefit analyses are likely to continue dictating policy regarding pharmacogenomic testing, and cost-benefit profiles should improve as technology advances and as information gathered from a single test becomes applicable to multiple medications and clinical scenarios.

In the meantime, physicians should become familiar with the terms used in medical genetics and pharmacogenomics and begin to understand genetic contributions to the outcomes of drug therapy. For example, understanding the consequences of metabolizer status and the frequency of variants in a given population can be tremendously helpful when advising our patients about anticipating potential problems when taking specific medications and about making informed decisions about pharmacogenomic testing.

This exchange of information alone may go a long way in improving therapy outcomes even when prospective pharmacogenomic testing is not routinely performed. Furthermore, an increasing number of patients will already have genotyping information available when they come to us, and clinicians need to be aware of the many pharmacogenomic applications recommended by the FDA when genetic status is known.¹⁰

In many patients, certain drugs do not work as well as expected, whereas in other patients they cause toxic effects, even at lower doses. For some patients, the reason may be genetic.

Sizeable minorities of the population carry genetic variants—polymorphisms— that affect their response to various drugs. Thanks to genetic research, our understanding of the variability of drug response has advanced markedly in the last decade. Many relevant polymorphisms have been identified, and tests for some of them are available.

See related editorial

Armed with the knowledge of their patients’ genetic status, physicians could predict their response to certain drugs, leading to better efficacy, fewer adverse drug reactions, and a better cost-benefit ratio.

The possible impact is substantial, since more than half of the drugs most commonly involved in adverse drug reactions are metabolized by polymorphic enzymes.¹ Adverse drug reactions remain a significant detriment to public health, having a substantial impact on rates of morbidity and death and on healthcare costs.^2–8 In the United States, adverse drug reactions are a leading cause of death in hospitalized patients⁴ and are annually responsible for hundreds of thousands of deaths and hundreds of billions of dollars in added costs.^2,4,6–8