Immunology

A 25-year-old man is evaluated for angioedema (swelling of lips and tongue) after eating paella at a Spanish restaurant. He has no history of allergies, but he says he had never eaten such a large variety of seafood before, especially shellfish.

He suspects that he is allergic to shellfish and asks the attending physician to order blood tests for seafood allergies, as he heard from a friend that blood tests are superior to other types of tests for allergy. The physician requests a serum immunoglobulin E (IgE) food panel test for this patient.

SERUM ALLERGEN-SPECIFIC IgE TESTING

Many methods of testing for allergy are available, including the skin-prick test, double-blind and single-blind placebo-controlled food challenges, open food challenges, inhalant challenges, drug challenges, and serum IgE tests. In clinical practice, these tests are often used in combination because when used individually, few of them are both highly sensitive and specific (Table 1).^1–6

Skin-prick testing is generally the method of choice for the preliminary evaluation of IgE-mediated allergies because it is more sensitive and requires less time to get a result.¹ But it is not the preferred test if the patient is at risk of a systemic reaction or has widespread dermatitis, nor is it useful if the patient is taking drugs that suppress the histamine response, such as antihistamines or tricyclic antidepressants.⁶ Moreover, skin-prick testing is more invasive and time-consuming than serum IgE testing.

Serum IgE testing is an attractive alternative, and it is more convenient because it requires only a single blood draw and poses a lower risk of adverse effects.

NOT A RELIABLE DIAGNOSTIC TOOL

As serum IgE testing has gained popularity, researchers have tried to improve its diagnostic power (ie, maximize its sensitivity and specificity) by determining the best cutoff values for IgE against specific antigens. Unfortunately, these values are difficult to determine because of confounding factors such as the lack of a reference standard, population diversity, patient atopy, and the overwhelming number of allergens that must be examined.

In addition, some researchers have used positive and negative predictive values to evaluate diagnostic cutoffs for serum antigen-specific IgE values. But these are not the most suitable performance measure to evaluate because they depend on disease prevalence and population characteristics.

Despite these efforts, results are still conflicting, and serum antigen-specific IgE testing is not a reliable diagnostic tool.

In an effort to gain insight from the available research data, we evaluated the clinical usefulness of 89 antigen-specific IgE tests, using an approach of summing their sensitivity and specificity. Previously, Wians⁷ proposed that a test is likely to be clinically useful if the sum of its sensitivity and specificity is equal to or greater than 170. Figure 1 shows the 89 tests, grouped into categories, and their summed sensitivities and specificities. The dashed line indicates a cutoff of 170; any bar that touches or crosses that line indicates that the test may be clinically useful, according to Wians.⁷

Only 7 of the 89 tests (cow, buckwheat, hazelnut, latex, Alternaria alternata, honey bee venom, and Johnson grass) satisfied this criterion. This suggests that a significant number of serum antigen-specific IgE tests perform suboptimally, and we are left with the question of why they are so commonly ordered.

Inappropriate use can lead to false-positive results, a situation in which patients may be subjected to unnecessary food avoidance that can result in nutritional deficiencies and decreased quality of life. It can also lead to false-negative results, when life-threatening diagnoses are missed and further excessive downstream testing is required—all leading; to negative outcomes for both patients and healthcare providers.

CHOOSING WISELY

The Choosing Wisely campaign in the United States has partnered with the American Academy of Allergy, Asthma, and Immunology to advocate against indiscriminate IgE testing in evaluating allergy.⁸ Allergy diagnosis and evaluation should be based on a combination of clinical history and judicious ordering of specific IgE tests, whether through skin or blood testing. Ordering of serum allergen-specific IgE tests for food allergies should be consistent with a clinical history of potential IgE-mediated food allergy⁸ and not food intolerance (Table 2).^4,5

Some jurisdictions in Canada have followed suit by restricting the number of serum IgE tests each physician is allowed to order per patient, to encourage more responsible ordering and to lower the number of potential false-positive results, which can lead to increased downstream costs as well as unnecessary patient worry and lifestyle modification.

CLINICAL BOTTOM LINE

Ordering diagnostic tests that have little clinical utility has long-term detrimental effects on both patient safety and healthcare sustainability.

In the case of the 25-year-old evaluated for shellfish allergy, the clinician should first explain that the swelling of the lips and tongue (angioedema) does suggest an IgE-mediated allergic reaction and not a non–IgE-mediated allergic reaction or a food intolerance. Non–IgE-mediated food allergies and food intolerances are marked by symptoms relating mainly to nonimmune aspects of the digestive system, whereas IgE-mediated food allergies affect the immune system and can involve a multitude of organs, including the skin and the respiratory and digestive systems (Table 2).

However, clinicians should avoid indiscriminately ordering food allergen IgE panels and instead should focus on foods likely to be the culprits based on the clinical history.⁹ Indiscriminate testing can lead to false-positive results and unnecessary food avoidance.

Since the patient developed symptoms of angioedema when he was exposed to his allergen, he may be apprehensive about a skin- prick test and the possibility of being subjected to the same discomfort. Therefore, in this situation, it may be best to perform serum IgE tests, but on a few targeted seafoods rather than the food panel the physician had ordered. A patient can be sensitized to an allergen (possess IgE antibodies) but not experience symptoms when exposed to it (ie, have tolerance).⁵ Also, false-negative results may occur, so a negative serum allergen-specific IgE test should likewise be interpreted in light of the pretest probability of allergy to a specific antigen.

If the history and the results of testing are not clear and congruent, the patient should be referred to an allergist for diagnosis or for management. The allergist can provide management techniques and periodic assessment as to the progression and resolution of the allergy. Table 2 highlights symptoms that differentiate an IgE-mediated from a non–IgE-mediated food allergy.^10,11Table 1 presents clinical indications and suggested diagnostic methods to the five most common allergen groups and the diagnostically invalid tests.^1–6

The bottom line is that we must consider the poor performance of serum allergen-specific IgE tests when diagnosing and treating suspected type I allergies and avoid ordering food allergen IgE panels whenever possible.

A 25-year-old man is evaluated for angioedema (swelling of lips and tongue) after eating paella at a Spanish restaurant. He has no history of allergies, but he says he had never eaten such a large variety of seafood before, especially shellfish.

He suspects that he is allergic to shellfish and asks the attending physician to order blood tests for seafood allergies, as he heard from a friend that blood tests are superior to other types of tests for allergy. The physician requests a serum immunoglobulin E (IgE) food panel test for this patient.

SERUM ALLERGEN-SPECIFIC IgE TESTING

Many methods of testing for allergy are available, including the skin-prick test, double-blind and single-blind placebo-controlled food challenges, open food challenges, inhalant challenges, drug challenges, and serum IgE tests. In clinical practice, these tests are often used in combination because when used individually, few of them are both highly sensitive and specific (Table 1).^1–6

Skin-prick testing is generally the method of choice for the preliminary evaluation of IgE-mediated allergies because it is more sensitive and requires less time to get a result.¹ But it is not the preferred test if the patient is at risk of a systemic reaction or has widespread dermatitis, nor is it useful if the patient is taking drugs that suppress the histamine response, such as antihistamines or tricyclic antidepressants.⁶ Moreover, skin-prick testing is more invasive and time-consuming than serum IgE testing.

Serum IgE testing is an attractive alternative, and it is more convenient because it requires only a single blood draw and poses a lower risk of adverse effects.

NOT A RELIABLE DIAGNOSTIC TOOL

As serum IgE testing has gained popularity, researchers have tried to improve its diagnostic power (ie, maximize its sensitivity and specificity) by determining the best cutoff values for IgE against specific antigens. Unfortunately, these values are difficult to determine because of confounding factors such as the lack of a reference standard, population diversity, patient atopy, and the overwhelming number of allergens that must be examined.

In addition, some researchers have used positive and negative predictive values to evaluate diagnostic cutoffs for serum antigen-specific IgE values. But these are not the most suitable performance measure to evaluate because they depend on disease prevalence and population characteristics.

Despite these efforts, results are still conflicting, and serum antigen-specific IgE testing is not a reliable diagnostic tool.

In an effort to gain insight from the available research data, we evaluated the clinical usefulness of 89 antigen-specific IgE tests, using an approach of summing their sensitivity and specificity. Previously, Wians⁷ proposed that a test is likely to be clinically useful if the sum of its sensitivity and specificity is equal to or greater than 170. Figure 1 shows the 89 tests, grouped into categories, and their summed sensitivities and specificities. The dashed line indicates a cutoff of 170; any bar that touches or crosses that line indicates that the test may be clinically useful, according to Wians.⁷

Only 7 of the 89 tests (cow, buckwheat, hazelnut, latex, Alternaria alternata, honey bee venom, and Johnson grass) satisfied this criterion. This suggests that a significant number of serum antigen-specific IgE tests perform suboptimally, and we are left with the question of why they are so commonly ordered.

Inappropriate use can lead to false-positive results, a situation in which patients may be subjected to unnecessary food avoidance that can result in nutritional deficiencies and decreased quality of life. It can also lead to false-negative results, when life-threatening diagnoses are missed and further excessive downstream testing is required—all leading; to negative outcomes for both patients and healthcare providers.

CHOOSING WISELY

The Choosing Wisely campaign in the United States has partnered with the American Academy of Allergy, Asthma, and Immunology to advocate against indiscriminate IgE testing in evaluating allergy.⁸ Allergy diagnosis and evaluation should be based on a combination of clinical history and judicious ordering of specific IgE tests, whether through skin or blood testing. Ordering of serum allergen-specific IgE tests for food allergies should be consistent with a clinical history of potential IgE-mediated food allergy⁸ and not food intolerance (Table 2).^4,5

Some jurisdictions in Canada have followed suit by restricting the number of serum IgE tests each physician is allowed to order per patient, to encourage more responsible ordering and to lower the number of potential false-positive results, which can lead to increased downstream costs as well as unnecessary patient worry and lifestyle modification.

CLINICAL BOTTOM LINE

Ordering diagnostic tests that have little clinical utility has long-term detrimental effects on both patient safety and healthcare sustainability.

In the case of the 25-year-old evaluated for shellfish allergy, the clinician should first explain that the swelling of the lips and tongue (angioedema) does suggest an IgE-mediated allergic reaction and not a non–IgE-mediated allergic reaction or a food intolerance. Non–IgE-mediated food allergies and food intolerances are marked by symptoms relating mainly to nonimmune aspects of the digestive system, whereas IgE-mediated food allergies affect the immune system and can involve a multitude of organs, including the skin and the respiratory and digestive systems (Table 2).

However, clinicians should avoid indiscriminately ordering food allergen IgE panels and instead should focus on foods likely to be the culprits based on the clinical history.⁹ Indiscriminate testing can lead to false-positive results and unnecessary food avoidance.

Since the patient developed symptoms of angioedema when he was exposed to his allergen, he may be apprehensive about a skin- prick test and the possibility of being subjected to the same discomfort. Therefore, in this situation, it may be best to perform serum IgE tests, but on a few targeted seafoods rather than the food panel the physician had ordered. A patient can be sensitized to an allergen (possess IgE antibodies) but not experience symptoms when exposed to it (ie, have tolerance).⁵ Also, false-negative results may occur, so a negative serum allergen-specific IgE test should likewise be interpreted in light of the pretest probability of allergy to a specific antigen.

If the history and the results of testing are not clear and congruent, the patient should be referred to an allergist for diagnosis or for management. The allergist can provide management techniques and periodic assessment as to the progression and resolution of the allergy. Table 2 highlights symptoms that differentiate an IgE-mediated from a non–IgE-mediated food allergy.^10,11Table 1 presents clinical indications and suggested diagnostic methods to the five most common allergen groups and the diagnostically invalid tests.^1–6

The bottom line is that we must consider the poor performance of serum allergen-specific IgE tests when diagnosing and treating suspected type I allergies and avoid ordering food allergen IgE panels whenever possible.

A 25-year-old man is evaluated for angioedema (swelling of lips and tongue) after eating paella at a Spanish restaurant. He has no history of allergies, but he says he had never eaten such a large variety of seafood before, especially shellfish.

He suspects that he is allergic to shellfish and asks the attending physician to order blood tests for seafood allergies, as he heard from a friend that blood tests are superior to other types of tests for allergy. The physician requests a serum immunoglobulin E (IgE) food panel test for this patient.

SERUM ALLERGEN-SPECIFIC IgE TESTING

Many methods of testing for allergy are available, including the skin-prick test, double-blind and single-blind placebo-controlled food challenges, open food challenges, inhalant challenges, drug challenges, and serum IgE tests. In clinical practice, these tests are often used in combination because when used individually, few of them are both highly sensitive and specific (Table 1).^1–6

Skin-prick testing is generally the method of choice for the preliminary evaluation of IgE-mediated allergies because it is more sensitive and requires less time to get a result.¹ But it is not the preferred test if the patient is at risk of a systemic reaction or has widespread dermatitis, nor is it useful if the patient is taking drugs that suppress the histamine response, such as antihistamines or tricyclic antidepressants.⁶ Moreover, skin-prick testing is more invasive and time-consuming than serum IgE testing.

Serum IgE testing is an attractive alternative, and it is more convenient because it requires only a single blood draw and poses a lower risk of adverse effects.

NOT A RELIABLE DIAGNOSTIC TOOL

As serum IgE testing has gained popularity, researchers have tried to improve its diagnostic power (ie, maximize its sensitivity and specificity) by determining the best cutoff values for IgE against specific antigens. Unfortunately, these values are difficult to determine because of confounding factors such as the lack of a reference standard, population diversity, patient atopy, and the overwhelming number of allergens that must be examined.

In addition, some researchers have used positive and negative predictive values to evaluate diagnostic cutoffs for serum antigen-specific IgE values. But these are not the most suitable performance measure to evaluate because they depend on disease prevalence and population characteristics.

Despite these efforts, results are still conflicting, and serum antigen-specific IgE testing is not a reliable diagnostic tool.

In an effort to gain insight from the available research data, we evaluated the clinical usefulness of 89 antigen-specific IgE tests, using an approach of summing their sensitivity and specificity. Previously, Wians⁷ proposed that a test is likely to be clinically useful if the sum of its sensitivity and specificity is equal to or greater than 170. Figure 1 shows the 89 tests, grouped into categories, and their summed sensitivities and specificities. The dashed line indicates a cutoff of 170; any bar that touches or crosses that line indicates that the test may be clinically useful, according to Wians.⁷

Only 7 of the 89 tests (cow, buckwheat, hazelnut, latex, Alternaria alternata, honey bee venom, and Johnson grass) satisfied this criterion. This suggests that a significant number of serum antigen-specific IgE tests perform suboptimally, and we are left with the question of why they are so commonly ordered.

Inappropriate use can lead to false-positive results, a situation in which patients may be subjected to unnecessary food avoidance that can result in nutritional deficiencies and decreased quality of life. It can also lead to false-negative results, when life-threatening diagnoses are missed and further excessive downstream testing is required—all leading; to negative outcomes for both patients and healthcare providers.

CHOOSING WISELY

The Choosing Wisely campaign in the United States has partnered with the American Academy of Allergy, Asthma, and Immunology to advocate against indiscriminate IgE testing in evaluating allergy.⁸ Allergy diagnosis and evaluation should be based on a combination of clinical history and judicious ordering of specific IgE tests, whether through skin or blood testing. Ordering of serum allergen-specific IgE tests for food allergies should be consistent with a clinical history of potential IgE-mediated food allergy⁸ and not food intolerance (Table 2).^4,5

Some jurisdictions in Canada have followed suit by restricting the number of serum IgE tests each physician is allowed to order per patient, to encourage more responsible ordering and to lower the number of potential false-positive results, which can lead to increased downstream costs as well as unnecessary patient worry and lifestyle modification.

CLINICAL BOTTOM LINE

Ordering diagnostic tests that have little clinical utility has long-term detrimental effects on both patient safety and healthcare sustainability.

In the case of the 25-year-old evaluated for shellfish allergy, the clinician should first explain that the swelling of the lips and tongue (angioedema) does suggest an IgE-mediated allergic reaction and not a non–IgE-mediated allergic reaction or a food intolerance. Non–IgE-mediated food allergies and food intolerances are marked by symptoms relating mainly to nonimmune aspects of the digestive system, whereas IgE-mediated food allergies affect the immune system and can involve a multitude of organs, including the skin and the respiratory and digestive systems (Table 2).

However, clinicians should avoid indiscriminately ordering food allergen IgE panels and instead should focus on foods likely to be the culprits based on the clinical history.⁹ Indiscriminate testing can lead to false-positive results and unnecessary food avoidance.

Since the patient developed symptoms of angioedema when he was exposed to his allergen, he may be apprehensive about a skin- prick test and the possibility of being subjected to the same discomfort. Therefore, in this situation, it may be best to perform serum IgE tests, but on a few targeted seafoods rather than the food panel the physician had ordered. A patient can be sensitized to an allergen (possess IgE antibodies) but not experience symptoms when exposed to it (ie, have tolerance).⁵ Also, false-negative results may occur, so a negative serum allergen-specific IgE test should likewise be interpreted in light of the pretest probability of allergy to a specific antigen.

If the history and the results of testing are not clear and congruent, the patient should be referred to an allergist for diagnosis or for management. The allergist can provide management techniques and periodic assessment as to the progression and resolution of the allergy. Table 2 highlights symptoms that differentiate an IgE-mediated from a non–IgE-mediated food allergy.^10,11Table 1 presents clinical indications and suggested diagnostic methods to the five most common allergen groups and the diagnostically invalid tests.^1–6

The bottom line is that we must consider the poor performance of serum allergen-specific IgE tests when diagnosing and treating suspected type I allergies and avoid ordering food allergen IgE panels whenever possible.

Measuring serum creatine kinase (CK) is an important part of the evaluation of patients with muscle weakness or myalgia, and of assessing patients with myopathies or rhabdomyolysis. But elevated CK sometimes is an incidental finding in a patient without muscle-related symptoms or with only minimal  nonspecific muscle symptoms (eg, cramps, spasms, fatigue) that do not significantly interfere with activities of daily living. This condition is sometimes referred to as “asymptomatic hyper-CK-emia.” Four other muscle enzymes that may also be elevated are aspartate aminotransferase, alanine aminotransferase, lactate dehydrogenase, and aldolase.

This review focuses on the evaluation of patients with elevated CK without significant muscle-related symptoms and proposes an algorithm for this purpose (Figure 1).

CURRENT THRESHOLDS MAY BE LOW

What appears to be an elevated CK level may in fact be normal, and it is important to determine in the initial assessment whether a CK value is truly abnormal.

Most laboratories use the central 95% of observations in white people as a reference range for serum CK, assuming that levels have a gaussian (bell-shaped) distribution, which is usually about 0 to 200 IU/L. Using these parameters, an abnormal CK level was observed in 19% of men and 5% of women in a study of nearly 1,000 healthy young people,¹ leading to overdiagnosis.

The actual distribution of serum CK levels in a healthy population is markedly skewed toward higher values and is nongaussian.^1–3 A 97.5% normal threshold is associated with a much lower false-positive rate and is recommended by the European Federation of Neurological Societies (now the European Academy of Neurology).⁴ This group also recommends pursuing further investigation only for patients whose level is at least 1.5 times the upper limit of normal; this threshold results in only a small reduction in sensitivity.

CK levels vary significantly by sex and race.⁵ Possible reasons include differences in muscle mass or total body mass and inherited differences in the permeability of the sarcolemma to CK.⁶ There is also a small reduction in CK levels as people age.²

The European Federation of Neurological Societies suggests redefining elevated CK as values 1.5 times beyond the upper limit of normal. Based on a 97.5% threshold and normal values determined by Brewster et al³ for black and white men and women, the following thresholds can be used to help decide whether to pursue further evaluation⁴:

• White women—325 IU/L
• White men—504 IU/L
• Black women—621 IU/L
• Black men—1,200 IU/L

PHYSICAL ACTIVITY RAISES CK

CK levels transiently rise after exercise or heavy manual labor. Serum CK levels may increase to as much as 30 times the upper limit of normal within 24 hours of strenuous physical activity, then slowly decline over the next 7 days. The degree of CK elevation depends on the type and duration of exercise, with greater elevation in those who are untrained.^2,4

In assessing asymptomatic or minimally symptomatic CK elevation, the test should be repeated after 7 days without exercise. A large community study in Norway found that repeat CK levels in people with incidentally discovered elevated CK were normal after 3 days of rest in 70% of cases.²

NONNEUROMUSCULAR CAUSES
NEED TO BE INVESTIGATED

Asymptomatic or minimally symptomatic elevated CK can be due to a primary neuromuscular disease or a variety of nonneuromuscular causes.

Patients who still have elevated CK after taking into account the 97.5% threshold, repeat testing after a week of rest, and a level more than 1.5 times the upper limit of normal for sex and race should first be evaluated for the many nonneuromuscular conditions that can cause elevated CK (Table 1).^7–9

Cardiac causes should be evaluated by history and physical examination, electrocardiography, and possibly testing for cardiac troponins.

Drugs commonly elevate CK

Prescription drugs and supplements are an important and common cause of CK elevation, so it is important to carefully review medications the patient is taking.

Statins can cause myalgia, muscle weakness, and rhabdomyolysis. Up to 5% of users develop CK elevation, typically 2 to 10 times the upper limit of normal.¹⁰ CK usually drops after stopping statins but may require weeks to months to normalize. Rarely, statin users develop a serious immune-mediated necrotizing myopathy.^11–13

The diversity of response to statin therapy appears to have a genetic basis. The SEARCH Collaborative Group¹⁴ conducted a genome-wide association study of 300,000 markers in 85 patients with definite or incipient myopathy and in 90 controls, all of whom were taking simvastatin 80 mg daily. They identified a single-nucleotide polymorphism in the SLCO1B1 gene on chromosome 12 that was strongly associated with a higher risk of statin-induced myopathy.

Patients with statin-related myopathy seem to have a higher frequency of occult metabolic muscle disease than the general population, also suggesting genetic susceptibility, although ascertainment bias could be a factor.¹⁴

Mechanisms of CK elevation in response to statins include increased muscle membrane fragility due to decreased cholesterol content, inhibition of isoprenoid production (a necessary step in the synthesis of membrane proteins), and depletion of ubiquinone, leading to mitochondrial dysfunction.

Macro CK: An abnormal enzyme complex

About 4% of patients with asymptomatic or minimally symptomatic elevated CK have “macro CK,” an enzyme complex with an atypically high molecular mass and reduced clearance, resulting in abnormally high blood levels of CK. Macro CK type 1 is more common and is found in up to 1.2% of the general population: complexes are composed of CK and immunoglobulin and are associated with autoimmune diseases.^9,15 Macro CK type 2 complexes consist of CK and an undetermined protein and are associated with malignancies.

CK electrophoresis is required to detect macro CK. Types 1 and 2 can be distinguished by protein G affinity chromatography.^9,15 

Endocrine disorders

Muscle involvement in endocrine disorders often presents with muscle weakness in addition to muscle enzyme abnormalities.

Hypothyroidism often causes weakness, cramps, myalgia, and a mild to moderate serum CK elevation.¹⁶ Severe CK elevation has been reported to occur after vigorous exercise.¹⁷ Thyroid replacement usually results in normalization of serum CK levels in 1 to 2 months.¹⁸

Hyperthyroidism is typically associated with normal serum CK concentrations, but in rare cases it can cause rhabdomyolysis.¹⁹

NEUROMUSCULAR CAUSES ARE NOT ALWAYS WORTH PURSUING

Only after the nonneuromuscular causes of elevated CK have been ruled out should neuromuscular disorders be considered (Table 2). Evaluation with electromyography, nerve conduction studies, and muscle biopsy may lead to the diagnosis of a specific neuromuscular disorder: patients may be in the presymptomatic stage of disease and may or may not eventually develop muscle weakness or other symptoms.^20,21

Is testing needed?

Most adult dystrophies and metabolic myopathies have no available treatment and their course is often benign, particularly if they present only with asymptomatic elevated CK. The value of a potentially extensive, expensive, and invasive evaluation for a specific neuromuscular cause should be weighed against the limited yield and treatment options. Moreover, specialized testing such as biochemical muscle enzyme analysis, sarcolemmal protein staining, and genetic testing are not available at all centers.

The European Federation of Neurological Societies guidelines recommend biopsy for  patients with asymptomatic elevated CK who also have any of the following:

• Abnormal (myopathic) findings on electromyography
• CK more than three times the upper limit of normal
• Age less than 25
• Exercise intolerance.⁴

Idiopathic inflammatory myopathies rarely present with asymptomatic elevated CK.^22–26 In one study,²⁷ they were found in just 5% of patients with asymptomatic elevated CK.

Hypomyopathic dermatomyositis and inclusion body myositis can present with mild CK elevations with normal muscle strength, especially early in the disease course. A myositis subset of antisynthetase syndrome can present with mildly elevated CK and interstitial lung disease.²⁷ Many of the inflammatory myopathies respond to treatment so are worth investigating.

In view of complexities in diagnosis of these conditions, one should proceed with testing only after discussing it with patients. Referral to a rheumatology specialist is preferred.

MUSCLE BIOPSY, ELECTROMYOGRAPHY, AND NERVE CONDUCTION STUDIES

Electromyography, nerve conduction studies, or muscle biopsy, or a combination of these tests, is usually needed to investigate neuromuscular causes of elevated CK.

Muscle biopsy abnormalities are found in about two-thirds of cases of asymptomatic elevated CK, but most abnormalities include nonspecific myopathic changes that are not diagnostic. A muscle biopsy that may include special stains for sarcolemmal proteins for muscular dystrophy and biochemical muscle enzyme analysis for metabolic myopathies is diagnostic in only 20% to 25% cases of asymptomatic elevated CK on average, with a variation between different series of 0% to 79%.^7,21,27–33

Electromyography and nerve conduction studies alone add little to the workup of asymp­tomatic elevated CK apart from a modest negative predictive value and as a guide for muscle biopsy. For a very few neuromuscular disorders causing an elevated CK (eg, motor neuron disease, Charcot-Marie-Tooth disease, myotonic dystrophy), electromyography and nerve conduction studies could suffice to make the diagnosis. 

Electromyography and nerve conduction studies detect abnormalities in nearly half of cases of asymptomatic CK elevation,^{7,21,27,28,30,31,33} but, as with biopsy, most changes are nonspecific. Although electromyography and nerve conduction studies can help distinguish primary neuropathic from myopathic disorders, the sensitivity and specificity are low for diagnosis. Normal studies do not rule out a condition, and abnormal studies are not diagnostic of a particular condition, although completely normal studies provide strong evidence against a severe neuromuscular disorder.

Combined testing

Using combined muscle biopsy, electromyography, and nerve conduction studies, the likelihood of making a diagnosis in patients with asymptomatic elevated CK is 28% on average (range of studies 4%–79%),^{2,7,21,26–28,30–32} and findings are nonspecific in 30% to 40% of cases. Findings are normal in about 30% to 40% of cases, which are thus diagnosed as idiopathic asymptomatic elevated CK.^28–31,34

Prelle et al³¹ retrospectively reviewed the cases of 114 patients, ages 3 to 70, with incidentally discovered elevated CK and few or no symptoms, who underwent muscle biopsy, electromyography, and nerve conduction studies after nonneuromuscular causes were ruled out. Although muscle biopsy findings were abnormal in 39% of cases, a diagnosis was established in only 18% of cases after an extensive workup: the diagnosis was definitive in only 10% and included dystrophinopathies, metabolic myopathies, and rare noninflammatory myopathies. For the remaining 8%, the diagnosis was probable and included four cases of partial carnitine palmitoyl transferase deficiency, three cases of malignant hyperthermia, and two rare inherited disorders.

DNA testing

In women with a serum CK less than three times the upper limit of normal who have a family history of Duchenne or Becker muscular dystrophy, DNA analysis of blood lymphocytes identifies 70% of carriers.⁴

IDIOPATHIC ELEVATED SERUM CK

Rowland et al³⁵ first coined the term “idiopathic hyper-CK-emia” and defined it as persistent elevation of serum CK despite a normal neurologic examination and testing, including electromyography, nerve conduction studies, and muscle biopsy.^35,36 To receive this diagnosis, patients must also have no family history or clinical evidence of neuromuscular disease.

Idiopathic elevated serum CK is sometimes familial. In one study,³⁷ elevated CK was found in family members of 13 of 28 unrelated probands. In the 13 families, 41 individuals had elevated CK. Genetic studies revealed that the condition is genetically heterogeneous and autosomal dominant in at least 60% of cases, with higher penetrance  in men.

D’Adda et al26 followed 55 people with idiopathic elevated CK for 7 years. Ten percent were eventually diagnosed with a neuromuscular disorder, 10% developed malignancy, and the remaining 80% developed no new condition. The CK level normalized or decreased in many patients, but most continued to have persistent CK elevations with minimal or no symptoms.

Measuring serum creatine kinase (CK) is an important part of the evaluation of patients with muscle weakness or myalgia, and of assessing patients with myopathies or rhabdomyolysis. But elevated CK sometimes is an incidental finding in a patient without muscle-related symptoms or with only minimal  nonspecific muscle symptoms (eg, cramps, spasms, fatigue) that do not significantly interfere with activities of daily living. This condition is sometimes referred to as “asymptomatic hyper-CK-emia.” Four other muscle enzymes that may also be elevated are aspartate aminotransferase, alanine aminotransferase, lactate dehydrogenase, and aldolase.

This review focuses on the evaluation of patients with elevated CK without significant muscle-related symptoms and proposes an algorithm for this purpose (Figure 1).

CURRENT THRESHOLDS MAY BE LOW

What appears to be an elevated CK level may in fact be normal, and it is important to determine in the initial assessment whether a CK value is truly abnormal.

Most laboratories use the central 95% of observations in white people as a reference range for serum CK, assuming that levels have a gaussian (bell-shaped) distribution, which is usually about 0 to 200 IU/L. Using these parameters, an abnormal CK level was observed in 19% of men and 5% of women in a study of nearly 1,000 healthy young people,¹ leading to overdiagnosis.

The actual distribution of serum CK levels in a healthy population is markedly skewed toward higher values and is nongaussian.^1–3 A 97.5% normal threshold is associated with a much lower false-positive rate and is recommended by the European Federation of Neurological Societies (now the European Academy of Neurology).⁴ This group also recommends pursuing further investigation only for patients whose level is at least 1.5 times the upper limit of normal; this threshold results in only a small reduction in sensitivity.

CK levels vary significantly by sex and race.⁵ Possible reasons include differences in muscle mass or total body mass and inherited differences in the permeability of the sarcolemma to CK.⁶ There is also a small reduction in CK levels as people age.²

The European Federation of Neurological Societies suggests redefining elevated CK as values 1.5 times beyond the upper limit of normal. Based on a 97.5% threshold and normal values determined by Brewster et al³ for black and white men and women, the following thresholds can be used to help decide whether to pursue further evaluation⁴:

• White women—325 IU/L
• White men—504 IU/L
• Black women—621 IU/L
• Black men—1,200 IU/L

PHYSICAL ACTIVITY RAISES CK

CK levels transiently rise after exercise or heavy manual labor. Serum CK levels may increase to as much as 30 times the upper limit of normal within 24 hours of strenuous physical activity, then slowly decline over the next 7 days. The degree of CK elevation depends on the type and duration of exercise, with greater elevation in those who are untrained.^2,4

In assessing asymptomatic or minimally symptomatic CK elevation, the test should be repeated after 7 days without exercise. A large community study in Norway found that repeat CK levels in people with incidentally discovered elevated CK were normal after 3 days of rest in 70% of cases.²

NONNEUROMUSCULAR CAUSES
NEED TO BE INVESTIGATED

Asymptomatic or minimally symptomatic elevated CK can be due to a primary neuromuscular disease or a variety of nonneuromuscular causes.

Patients who still have elevated CK after taking into account the 97.5% threshold, repeat testing after a week of rest, and a level more than 1.5 times the upper limit of normal for sex and race should first be evaluated for the many nonneuromuscular conditions that can cause elevated CK (Table 1).^7–9

Cardiac causes should be evaluated by history and physical examination, electrocardiography, and possibly testing for cardiac troponins.

Drugs commonly elevate CK

Prescription drugs and supplements are an important and common cause of CK elevation, so it is important to carefully review medications the patient is taking.

Statins can cause myalgia, muscle weakness, and rhabdomyolysis. Up to 5% of users develop CK elevation, typically 2 to 10 times the upper limit of normal.¹⁰ CK usually drops after stopping statins but may require weeks to months to normalize. Rarely, statin users develop a serious immune-mediated necrotizing myopathy.^11–13

The diversity of response to statin therapy appears to have a genetic basis. The SEARCH Collaborative Group¹⁴ conducted a genome-wide association study of 300,000 markers in 85 patients with definite or incipient myopathy and in 90 controls, all of whom were taking simvastatin 80 mg daily. They identified a single-nucleotide polymorphism in the SLCO1B1 gene on chromosome 12 that was strongly associated with a higher risk of statin-induced myopathy.

Patients with statin-related myopathy seem to have a higher frequency of occult metabolic muscle disease than the general population, also suggesting genetic susceptibility, although ascertainment bias could be a factor.¹⁴

Mechanisms of CK elevation in response to statins include increased muscle membrane fragility due to decreased cholesterol content, inhibition of isoprenoid production (a necessary step in the synthesis of membrane proteins), and depletion of ubiquinone, leading to mitochondrial dysfunction.

Macro CK: An abnormal enzyme complex

About 4% of patients with asymptomatic or minimally symptomatic elevated CK have “macro CK,” an enzyme complex with an atypically high molecular mass and reduced clearance, resulting in abnormally high blood levels of CK. Macro CK type 1 is more common and is found in up to 1.2% of the general population: complexes are composed of CK and immunoglobulin and are associated with autoimmune diseases.^9,15 Macro CK type 2 complexes consist of CK and an undetermined protein and are associated with malignancies.

CK electrophoresis is required to detect macro CK. Types 1 and 2 can be distinguished by protein G affinity chromatography.^9,15 

Endocrine disorders

Muscle involvement in endocrine disorders often presents with muscle weakness in addition to muscle enzyme abnormalities.

Hypothyroidism often causes weakness, cramps, myalgia, and a mild to moderate serum CK elevation.¹⁶ Severe CK elevation has been reported to occur after vigorous exercise.¹⁷ Thyroid replacement usually results in normalization of serum CK levels in 1 to 2 months.¹⁸

Hyperthyroidism is typically associated with normal serum CK concentrations, but in rare cases it can cause rhabdomyolysis.¹⁹

NEUROMUSCULAR CAUSES ARE NOT ALWAYS WORTH PURSUING

Only after the nonneuromuscular causes of elevated CK have been ruled out should neuromuscular disorders be considered (Table 2). Evaluation with electromyography, nerve conduction studies, and muscle biopsy may lead to the diagnosis of a specific neuromuscular disorder: patients may be in the presymptomatic stage of disease and may or may not eventually develop muscle weakness or other symptoms.^20,21

Is testing needed?

Most adult dystrophies and metabolic myopathies have no available treatment and their course is often benign, particularly if they present only with asymptomatic elevated CK. The value of a potentially extensive, expensive, and invasive evaluation for a specific neuromuscular cause should be weighed against the limited yield and treatment options. Moreover, specialized testing such as biochemical muscle enzyme analysis, sarcolemmal protein staining, and genetic testing are not available at all centers.

The European Federation of Neurological Societies guidelines recommend biopsy for  patients with asymptomatic elevated CK who also have any of the following:

• Abnormal (myopathic) findings on electromyography
• CK more than three times the upper limit of normal
• Age less than 25
• Exercise intolerance.⁴

Idiopathic inflammatory myopathies rarely present with asymptomatic elevated CK.^22–26 In one study,²⁷ they were found in just 5% of patients with asymptomatic elevated CK.

Hypomyopathic dermatomyositis and inclusion body myositis can present with mild CK elevations with normal muscle strength, especially early in the disease course. A myositis subset of antisynthetase syndrome can present with mildly elevated CK and interstitial lung disease.²⁷ Many of the inflammatory myopathies respond to treatment so are worth investigating.

In view of complexities in diagnosis of these conditions, one should proceed with testing only after discussing it with patients. Referral to a rheumatology specialist is preferred.

MUSCLE BIOPSY, ELECTROMYOGRAPHY, AND NERVE CONDUCTION STUDIES

Electromyography, nerve conduction studies, or muscle biopsy, or a combination of these tests, is usually needed to investigate neuromuscular causes of elevated CK.

Muscle biopsy abnormalities are found in about two-thirds of cases of asymptomatic elevated CK, but most abnormalities include nonspecific myopathic changes that are not diagnostic. A muscle biopsy that may include special stains for sarcolemmal proteins for muscular dystrophy and biochemical muscle enzyme analysis for metabolic myopathies is diagnostic in only 20% to 25% cases of asymptomatic elevated CK on average, with a variation between different series of 0% to 79%.^7,21,27–33

Electromyography and nerve conduction studies alone add little to the workup of asymp­tomatic elevated CK apart from a modest negative predictive value and as a guide for muscle biopsy. For a very few neuromuscular disorders causing an elevated CK (eg, motor neuron disease, Charcot-Marie-Tooth disease, myotonic dystrophy), electromyography and nerve conduction studies could suffice to make the diagnosis. 

Electromyography and nerve conduction studies detect abnormalities in nearly half of cases of asymptomatic CK elevation,^{7,21,27,28,30,31,33} but, as with biopsy, most changes are nonspecific. Although electromyography and nerve conduction studies can help distinguish primary neuropathic from myopathic disorders, the sensitivity and specificity are low for diagnosis. Normal studies do not rule out a condition, and abnormal studies are not diagnostic of a particular condition, although completely normal studies provide strong evidence against a severe neuromuscular disorder.

Combined testing

Using combined muscle biopsy, electromyography, and nerve conduction studies, the likelihood of making a diagnosis in patients with asymptomatic elevated CK is 28% on average (range of studies 4%–79%),^{2,7,21,26–28,30–32} and findings are nonspecific in 30% to 40% of cases. Findings are normal in about 30% to 40% of cases, which are thus diagnosed as idiopathic asymptomatic elevated CK.^28–31,34

Prelle et al³¹ retrospectively reviewed the cases of 114 patients, ages 3 to 70, with incidentally discovered elevated CK and few or no symptoms, who underwent muscle biopsy, electromyography, and nerve conduction studies after nonneuromuscular causes were ruled out. Although muscle biopsy findings were abnormal in 39% of cases, a diagnosis was established in only 18% of cases after an extensive workup: the diagnosis was definitive in only 10% and included dystrophinopathies, metabolic myopathies, and rare noninflammatory myopathies. For the remaining 8%, the diagnosis was probable and included four cases of partial carnitine palmitoyl transferase deficiency, three cases of malignant hyperthermia, and two rare inherited disorders.

DNA testing

In women with a serum CK less than three times the upper limit of normal who have a family history of Duchenne or Becker muscular dystrophy, DNA analysis of blood lymphocytes identifies 70% of carriers.⁴

IDIOPATHIC ELEVATED SERUM CK

Rowland et al³⁵ first coined the term “idiopathic hyper-CK-emia” and defined it as persistent elevation of serum CK despite a normal neurologic examination and testing, including electromyography, nerve conduction studies, and muscle biopsy.^35,36 To receive this diagnosis, patients must also have no family history or clinical evidence of neuromuscular disease.

Idiopathic elevated serum CK is sometimes familial. In one study,³⁷ elevated CK was found in family members of 13 of 28 unrelated probands. In the 13 families, 41 individuals had elevated CK. Genetic studies revealed that the condition is genetically heterogeneous and autosomal dominant in at least 60% of cases, with higher penetrance  in men.

D’Adda et al26 followed 55 people with idiopathic elevated CK for 7 years. Ten percent were eventually diagnosed with a neuromuscular disorder, 10% developed malignancy, and the remaining 80% developed no new condition. The CK level normalized or decreased in many patients, but most continued to have persistent CK elevations with minimal or no symptoms.

Measuring serum creatine kinase (CK) is an important part of the evaluation of patients with muscle weakness or myalgia, and of assessing patients with myopathies or rhabdomyolysis. But elevated CK sometimes is an incidental finding in a patient without muscle-related symptoms or with only minimal  nonspecific muscle symptoms (eg, cramps, spasms, fatigue) that do not significantly interfere with activities of daily living. This condition is sometimes referred to as “asymptomatic hyper-CK-emia.” Four other muscle enzymes that may also be elevated are aspartate aminotransferase, alanine aminotransferase, lactate dehydrogenase, and aldolase.

This review focuses on the evaluation of patients with elevated CK without significant muscle-related symptoms and proposes an algorithm for this purpose (Figure 1).

CURRENT THRESHOLDS MAY BE LOW

What appears to be an elevated CK level may in fact be normal, and it is important to determine in the initial assessment whether a CK value is truly abnormal.

Most laboratories use the central 95% of observations in white people as a reference range for serum CK, assuming that levels have a gaussian (bell-shaped) distribution, which is usually about 0 to 200 IU/L. Using these parameters, an abnormal CK level was observed in 19% of men and 5% of women in a study of nearly 1,000 healthy young people,¹ leading to overdiagnosis.

The actual distribution of serum CK levels in a healthy population is markedly skewed toward higher values and is nongaussian.^1–3 A 97.5% normal threshold is associated with a much lower false-positive rate and is recommended by the European Federation of Neurological Societies (now the European Academy of Neurology).⁴ This group also recommends pursuing further investigation only for patients whose level is at least 1.5 times the upper limit of normal; this threshold results in only a small reduction in sensitivity.

CK levels vary significantly by sex and race.⁵ Possible reasons include differences in muscle mass or total body mass and inherited differences in the permeability of the sarcolemma to CK.⁶ There is also a small reduction in CK levels as people age.²

The European Federation of Neurological Societies suggests redefining elevated CK as values 1.5 times beyond the upper limit of normal. Based on a 97.5% threshold and normal values determined by Brewster et al³ for black and white men and women, the following thresholds can be used to help decide whether to pursue further evaluation⁴:

• White women—325 IU/L
• White men—504 IU/L
• Black women—621 IU/L
• Black men—1,200 IU/L

PHYSICAL ACTIVITY RAISES CK

CK levels transiently rise after exercise or heavy manual labor. Serum CK levels may increase to as much as 30 times the upper limit of normal within 24 hours of strenuous physical activity, then slowly decline over the next 7 days. The degree of CK elevation depends on the type and duration of exercise, with greater elevation in those who are untrained.^2,4

In assessing asymptomatic or minimally symptomatic CK elevation, the test should be repeated after 7 days without exercise. A large community study in Norway found that repeat CK levels in people with incidentally discovered elevated CK were normal after 3 days of rest in 70% of cases.²

NONNEUROMUSCULAR CAUSES
NEED TO BE INVESTIGATED

Asymptomatic or minimally symptomatic elevated CK can be due to a primary neuromuscular disease or a variety of nonneuromuscular causes.

Patients who still have elevated CK after taking into account the 97.5% threshold, repeat testing after a week of rest, and a level more than 1.5 times the upper limit of normal for sex and race should first be evaluated for the many nonneuromuscular conditions that can cause elevated CK (Table 1).^7–9

Cardiac causes should be evaluated by history and physical examination, electrocardiography, and possibly testing for cardiac troponins.

Drugs commonly elevate CK

Prescription drugs and supplements are an important and common cause of CK elevation, so it is important to carefully review medications the patient is taking.

Statins can cause myalgia, muscle weakness, and rhabdomyolysis. Up to 5% of users develop CK elevation, typically 2 to 10 times the upper limit of normal.¹⁰ CK usually drops after stopping statins but may require weeks to months to normalize. Rarely, statin users develop a serious immune-mediated necrotizing myopathy.^11–13

The diversity of response to statin therapy appears to have a genetic basis. The SEARCH Collaborative Group¹⁴ conducted a genome-wide association study of 300,000 markers in 85 patients with definite or incipient myopathy and in 90 controls, all of whom were taking simvastatin 80 mg daily. They identified a single-nucleotide polymorphism in the SLCO1B1 gene on chromosome 12 that was strongly associated with a higher risk of statin-induced myopathy.

Patients with statin-related myopathy seem to have a higher frequency of occult metabolic muscle disease than the general population, also suggesting genetic susceptibility, although ascertainment bias could be a factor.¹⁴

Mechanisms of CK elevation in response to statins include increased muscle membrane fragility due to decreased cholesterol content, inhibition of isoprenoid production (a necessary step in the synthesis of membrane proteins), and depletion of ubiquinone, leading to mitochondrial dysfunction.

Macro CK: An abnormal enzyme complex

About 4% of patients with asymptomatic or minimally symptomatic elevated CK have “macro CK,” an enzyme complex with an atypically high molecular mass and reduced clearance, resulting in abnormally high blood levels of CK. Macro CK type 1 is more common and is found in up to 1.2% of the general population: complexes are composed of CK and immunoglobulin and are associated with autoimmune diseases.^9,15 Macro CK type 2 complexes consist of CK and an undetermined protein and are associated with malignancies.

CK electrophoresis is required to detect macro CK. Types 1 and 2 can be distinguished by protein G affinity chromatography.^9,15 

Endocrine disorders

Muscle involvement in endocrine disorders often presents with muscle weakness in addition to muscle enzyme abnormalities.

Hypothyroidism often causes weakness, cramps, myalgia, and a mild to moderate serum CK elevation.¹⁶ Severe CK elevation has been reported to occur after vigorous exercise.¹⁷ Thyroid replacement usually results in normalization of serum CK levels in 1 to 2 months.¹⁸

Hyperthyroidism is typically associated with normal serum CK concentrations, but in rare cases it can cause rhabdomyolysis.¹⁹

NEUROMUSCULAR CAUSES ARE NOT ALWAYS WORTH PURSUING

Only after the nonneuromuscular causes of elevated CK have been ruled out should neuromuscular disorders be considered (Table 2). Evaluation with electromyography, nerve conduction studies, and muscle biopsy may lead to the diagnosis of a specific neuromuscular disorder: patients may be in the presymptomatic stage of disease and may or may not eventually develop muscle weakness or other symptoms.^20,21

Is testing needed?

Most adult dystrophies and metabolic myopathies have no available treatment and their course is often benign, particularly if they present only with asymptomatic elevated CK. The value of a potentially extensive, expensive, and invasive evaluation for a specific neuromuscular cause should be weighed against the limited yield and treatment options. Moreover, specialized testing such as biochemical muscle enzyme analysis, sarcolemmal protein staining, and genetic testing are not available at all centers.

The European Federation of Neurological Societies guidelines recommend biopsy for  patients with asymptomatic elevated CK who also have any of the following:

• Abnormal (myopathic) findings on electromyography
• CK more than three times the upper limit of normal
• Age less than 25
• Exercise intolerance.⁴

Idiopathic inflammatory myopathies rarely present with asymptomatic elevated CK.^22–26 In one study,²⁷ they were found in just 5% of patients with asymptomatic elevated CK.

Hypomyopathic dermatomyositis and inclusion body myositis can present with mild CK elevations with normal muscle strength, especially early in the disease course. A myositis subset of antisynthetase syndrome can present with mildly elevated CK and interstitial lung disease.²⁷ Many of the inflammatory myopathies respond to treatment so are worth investigating.

In view of complexities in diagnosis of these conditions, one should proceed with testing only after discussing it with patients. Referral to a rheumatology specialist is preferred.

MUSCLE BIOPSY, ELECTROMYOGRAPHY, AND NERVE CONDUCTION STUDIES

Electromyography, nerve conduction studies, or muscle biopsy, or a combination of these tests, is usually needed to investigate neuromuscular causes of elevated CK.

Muscle biopsy abnormalities are found in about two-thirds of cases of asymptomatic elevated CK, but most abnormalities include nonspecific myopathic changes that are not diagnostic. A muscle biopsy that may include special stains for sarcolemmal proteins for muscular dystrophy and biochemical muscle enzyme analysis for metabolic myopathies is diagnostic in only 20% to 25% cases of asymptomatic elevated CK on average, with a variation between different series of 0% to 79%.^7,21,27–33

Electromyography and nerve conduction studies alone add little to the workup of asymp­tomatic elevated CK apart from a modest negative predictive value and as a guide for muscle biopsy. For a very few neuromuscular disorders causing an elevated CK (eg, motor neuron disease, Charcot-Marie-Tooth disease, myotonic dystrophy), electromyography and nerve conduction studies could suffice to make the diagnosis. 

Electromyography and nerve conduction studies detect abnormalities in nearly half of cases of asymptomatic CK elevation,^{7,21,27,28,30,31,33} but, as with biopsy, most changes are nonspecific. Although electromyography and nerve conduction studies can help distinguish primary neuropathic from myopathic disorders, the sensitivity and specificity are low for diagnosis. Normal studies do not rule out a condition, and abnormal studies are not diagnostic of a particular condition, although completely normal studies provide strong evidence against a severe neuromuscular disorder.

Combined testing

Using combined muscle biopsy, electromyography, and nerve conduction studies, the likelihood of making a diagnosis in patients with asymptomatic elevated CK is 28% on average (range of studies 4%–79%),^{2,7,21,26–28,30–32} and findings are nonspecific in 30% to 40% of cases. Findings are normal in about 30% to 40% of cases, which are thus diagnosed as idiopathic asymptomatic elevated CK.^28–31,34

Prelle et al³¹ retrospectively reviewed the cases of 114 patients, ages 3 to 70, with incidentally discovered elevated CK and few or no symptoms, who underwent muscle biopsy, electromyography, and nerve conduction studies after nonneuromuscular causes were ruled out. Although muscle biopsy findings were abnormal in 39% of cases, a diagnosis was established in only 18% of cases after an extensive workup: the diagnosis was definitive in only 10% and included dystrophinopathies, metabolic myopathies, and rare noninflammatory myopathies. For the remaining 8%, the diagnosis was probable and included four cases of partial carnitine palmitoyl transferase deficiency, three cases of malignant hyperthermia, and two rare inherited disorders.

DNA testing

In women with a serum CK less than three times the upper limit of normal who have a family history of Duchenne or Becker muscular dystrophy, DNA analysis of blood lymphocytes identifies 70% of carriers.⁴

IDIOPATHIC ELEVATED SERUM CK

Rowland et al³⁵ first coined the term “idiopathic hyper-CK-emia” and defined it as persistent elevation of serum CK despite a normal neurologic examination and testing, including electromyography, nerve conduction studies, and muscle biopsy.^35,36 To receive this diagnosis, patients must also have no family history or clinical evidence of neuromuscular disease.

Idiopathic elevated serum CK is sometimes familial. In one study,³⁷ elevated CK was found in family members of 13 of 28 unrelated probands. In the 13 families, 41 individuals had elevated CK. Genetic studies revealed that the condition is genetically heterogeneous and autosomal dominant in at least 60% of cases, with higher penetrance  in men.

D’Adda et al26 followed 55 people with idiopathic elevated CK for 7 years. Ten percent were eventually diagnosed with a neuromuscular disorder, 10% developed malignancy, and the remaining 80% developed no new condition. The CK level normalized or decreased in many patients, but most continued to have persistent CK elevations with minimal or no symptoms.

A 79-year-old woman with a history of breast cancer in remission and hypertension presented to a local emergency department because of subacute memory loss and compulsive shopping. Her serum sodium concentration was 127 mmol/L (reference range 132–148). Computed tomography (CT) and magnetic resonance imaging (MRI) of the brain were normal, and she was sent home.

Three days later, she experienced a generalized tonic-clonic seizure that evolved into status epilepticus. She was intubated and admitted to the intensive care unit. Cerebrospinal fluid analysis was normal, and infectious causes of encephalitis were ruled out. MRI showed increased signal in both hippocampi (Figure 1). Her seizures were refractory to treatment, and she was given pentobarbital to induce a coma.

Serum evaluation of neuronal antibodies revealed elevated titers of the voltage-gated potassium channel (VGKC) complex antibody, with subsequent subtyping confirming the leucine-rich glioma-inactivated protein 1 (LGI1) protein as the antigenic target.

She received a 5-day course of intravenous immunoglobulin and methylprednisolone, pentobarbital was withdrawn, and the seizures did not recur, but weeks later she remained comatose. Positron emission tomography (PET) of the brain revealed hypermetabolism in the medial and anterior aspects of both temporal lobes. She underwent five sessions of plasma exchange, after which she began to improve and follow commands. She was ultimately discharged to an acute rehabilitation facility after a 4-week hospital stay.

She received infusions of intravenous immunoglobulin twice a month for 6 months.  At her last follow-up visit, she was seizure-free and neurologically intact except for mild inattention.

NEWLY RECOGNIZED DISEASES

Although autoantibody-mediated encephalitic syndromes were first described more than 50 years ago,^1,2 their autoimmune basis was not recognized until the early 1980s.³ In the past 10 years, a flood of novel clinical syndromes associated with neuronal autoantibodies has been described that may be markedly improved or even completely resolved with immunotherapy. In cases of unexplained seizure, encephalitis, or acute-onset psychiatric syndromes, suspecting these syndromes can lead to diagnosis, treatment, and a good outcome.

This review describes the key clinical autoantibody-mediated encephalitic syndromes, explains the better-characterized antibody associations, and discusses their diagnosis and treatment. 

CLASSIFIED ANATOMICALLY, IMMUNOLOGICALLY, OR EPONYMOUSLY

Autoantibody-mediated encephalitis is also known as autoimmune-mediated encephalitis, autoimmune-mediated limbic encephalitis, and autoimmune synaptic encephalitis.

How to categorize these syndromes is still in flux: they can be listed by the area of the brain affected, the antibody involved, or the name of the discoverer (eg, Morvan syndrome).

Autoantibodies identified in autoimmune encephalitis fall under two broad categories:

• Those targeting intracellular (intranuclear or intracytoplasmic) antigens; the syndromes they cause are more likely to be paraneoplastic and less responsive to immunotherapy
• Those targeting antigens on the neuronal surface: the syndromes they cause are less likely to be paraneoplastic and are more responsive to immunotherapy.⁴

SYNDROMES DEFINED BY BRAIN AREA AFFECTED

Below, we provide examples of neurologic syndromes of autoantibody-mediated encephalitis according to the region of the brain most affected, ie, the limbic system, the brainstem, or the cerebellum (Figure 2).

LIMBIC ENCEPHALITIS

Memory loss, behavioral changes, seizures

Patients with limbic encephalitis (such as the patient described in the vignette above) present with symptoms attributed to dysfunction of mesial temporal lobe structures, most notably the hippocampus. Prominent symptoms include short-term memory loss, behavioral disturbances such as agitation and confusion, and psychiatric problems such as depression and psychosis. Recurrent seizures are a salient feature and, not uncommonly, progress to status epilepticus.

Antibodies are not all cancer-associated

Cerebrospinal fluid analysis can be normal or show abnormalities suggesting immune activation, eg, slight pleocytosis, elevated protein, increased immunoglobulin G synthesis, and oligoclonal banding.⁵

In many cases, an autoantibody is found in the blood or in the cerebrospinal fluid. Some patients may express more than one autoantibody, so the traditional view of “one antibody, one syndrome” is incorrect.

Although initially identified as a rare paraneoplastic disorder, limbic encephalitis sometimes occurs in the absence of malignancy.

Multiple antibodies have been linked to the syndrome (Table 1).^6–9 The “classic” antibodies initially found in paraneoplastic forms are now generally viewed as nonpathogenic, in part because they are directed against intracellular antigens. Neuronal injury in paraneo­plastic limbic encephalitis is believed to be mediated by cytotoxic T lymphocytes, with neuronal autoantibodies being produced after the injury.⁴ Recently defined antibodies, such as those targeting the N-methyl-d-aspartate (NMDA) receptor⁶ and the LGI1 protein,⁷ are now understood to be common causes of limbic encephalitis. 

Imaging usually shows limbic focal changes

Structural MRI or functional fluorodeoxyglucose (FDG)-PET imaging may show focal changes in limbic system structures, such as the mesial temporal lobes.  It is now recognized that other cortical areas may be involved, and the term “limbic encephalitis” may give way to “cortical” or “focal encephalitis.”

In about 60% of patients, MRI shows hyperintense fluid-attenuated inversion recovery (FLAIR) or T2 signal changes in the mesial temporal lobes, likely reflecting inflammatory changes.^4,10,11 On FDG-PET, hypermetabolism may be observed in the mesial temporal lobes early in the disease despite normal findings on MRI.¹² Hypometabolism, either diffuse or localized to the mesial temporal lobes, eventually sets in, likely reflecting cytotoxic injury in the aftermath of prolonged inflammation or seizures.

Consider other causes

Before diagnosing limbic encephalitis, it is essential to evaluate for infectious meningoencephalitis, especially herpes simplex viral encephalitis. Thiamine deficiency (Wernicke encephalopathy), drug intoxication, prion disease, Hashimoto encephalopathy, tumor, and subclinical status epilepticus should also be considered. Some of these conditions are associated with the same neuronal autoantibodies detected in limbic encephalitis. Further complicating the picture, case reports have shown the presence of serum neuronal autoantibodies—VGKC complex^13–15 and NMDA-receptor antibodies^16,17—in confirmed cases of prion disease. In addition, adequately treated herpes simplex viral encephalitis can precipitate the production of NMDA-receptor antibodies and their characteristic syndrome.^18–20

BRAINSTEM ENCEPHALITIS

The brainstem—the midbrain, pons, and medulla—can be affected, either in isolation or more commonly as part of a more widespread autoantibody-mediated encephalitis. Symptoms and signs include eye movement abnormalities, ptosis, dysphagia, dysarthria, ataxia, facial palsy, vertigo, hearing impairment, reduced consciousness, and hypoventilation.²¹

Anti-Hu, anti-Ri, and anti-Ma2 antibodies are most commonly associated with brainstem encephalitis (Table 2). Anti-Ma2-associated encephalitis may improve after a combination of immunotherapy and tumor removal²¹; the others have a poor prognosis.

Neuromyelitis optica spectrum disorders

Neuromyelitis optica spectrum disorders most commonly involve demyelination affecting the optic nerves and spinal cord, leading to  unilateral or bilateral optic neuritis and transverse myelitis spanning three or more vertebral segments.²² The initial clinical manifestation may be an encephalitic pattern, affecting predominantly the brainstem in a restricted fashion,²² or the central nervous system in a more diffuse pattern, mimicking either acute disseminated encephalomyelitis or, in less severe cases, posterior reversible encephalopathy syndrome.²³

Testing for antiaquaporin-4 antibody, also known as neuromyelitis optica immunoglobulin G, is the single most decisive laboratory test for diagnosing neuromyelitis optica spectrum disorders, so serum and cerebrospinal fluid evaluation for this autoantibody should be considered when caring for a patient whose clinical picture suggests brainstem encephalitis.²²

Bickerstaff brainstem encephalitis

Bickerstaff brainstem encephalitis was first described more than half a century ago in patients with postinfectious ataxia, ophthalmoparesis, and altered consciousness. This rare disease was later found to be associated with antiganglioside GQ1b (anti-GQ1b) autoantibody. MRI is normal in about 90% of cases, so recognizing the clinical presentation and analyzing anti-GQ1b serum titers are critical to diagnosis.

Recovery is usually spontaneous and complete and can be hastened by immunotherapy, especially intravenous immunoglobulin.²⁴

Other causes of brainstem encephalitis

The differential diagnosis of a presentation of brainstem encephalitis includes:

• Infectious causes, the most common being Listeria species followed by enterovirus 71 and herpes simplex virus.²⁵ Tuberculosis, brucellosis, and Whipple disease should also be considered.
• Primary central nervous system inflammatory and demyelinating conditions, eg, multiple sclerosis and acute disseminated encephalomyelitis.
• Systemic inflammatory conditions, eg, Beh­çet disease, systemic lupus erythematosus, and sarcoidosis.
• Direct brainstem neoplastic involvement, as might occur in primary central nervous system lymphoma or leptomeningeal carcinomatosis.

CEREBELLAR SYNDROME

Patients with autoantibody-mediated encephalitis localized predominantly to the cerebellum typically present with dizziness, vertigo, and unsteady gait, progressing eventually to limb and gait ataxia.⁴ Symptoms are often subacute, progressing over weeks.

Multiple neuronal autoantibodies have been found to occur with cerebellar encephalitis (Table 2). In most cases, they are paraneo­plastic and considered not to be pathogenic, given the intracellular location of their target antigen.⁴ In such cases, the syndrome is more accurately described as autoantibody-associated rather than autoantibody-mediated. Only in a minority of cases have neuronal autoantibodies been demonstrated to be directly pathogenic, ie, antimetabotropic glutamate receptor type 1 (anti-mGluR1) antibody-associated cerebellitis²⁶ and antiglutamic acid decarboxylase (anti-GAD)-associated cerebellar ataxia.²⁷

Differential diagnosis of cerebellar syndromes

The differential diagnosis of autoantibody-associated cerebellar syndromes is broad and includes:

• Alcohol-induced atrophy
• Drug-induced cerebellar atrophy (eg, from lithium, phenytoin, gabapentin, metronidazole, amiodarone, carbamazepine)
• Vitamin B₁ and E deficiency
• Hypothyroidism, hypoparathyroidism
• Neurodegenerative disease (eg, prion disease, multiple system atrophy)
• Parainfectious causes (eg, after infection with Epstein-Barr virus)
• Immune-mediated diseases (Miller-Fisher syndrome, associated with anti-GQ1b antibodies, and antigliadin-associated ataxia, which can occur in isolation or as part of celiac disease).⁴

SYNDROMES ASSOCIATED WITH SPECIFIC ANTIBODIES

A few of the autoantibody-mediated encephalitic syndromes have specific antibody associations and characteristic clinical presentations. The most prominent of these syndromes are VGKC complex antibody encephalitis (as in the patient described at the beginning of this article) and anti-NMDA receptor encephalitis.

VGKC COMPLEX ANTIBODY-MEDIATED LIMBIC ENCEPHALITIS

VGKC complex antibodies, initially reported to be associated with the peripheral nerve hyperexcitability disorder neuromyotonia, were subsequently found in Morvan syndrome.^28,29 Patients with this syndrome often present with autonomic dysfunction and peripheral nerve hyperexcitability but also develop insomnia, confusion, hallucinations, and memory loss. Drawing on the clinical overlap between Morvan syndrome and limbic encephalitis, Buckley et al³⁰ were the first to report VGKC complex antibodies in two cases of limbic encephalitis.

VGKC complex antibodies are now understood to be associated with a wide variety of neurologic conditions, including chronic idiopathic pain, epilepsy,³¹ movement disorders, cranial nerve abnormalities, autonomic dysfunction,³² and gut dysmotility.³³ In contrast, these antibodies are rare in healthy people.³⁴ Limbic encephalitis associated with VGKC complex antibody usually lacks cerebellar and brainstem dysfunction, which may help distinguish it from other types of autoantibody-mediated limbic encephalitis.¹²

VGKC complex antibody does not bind to the potassium channel itself. Instead it recognizes other constituents of the channel complex, most notably LGI1 and contactin-associated protein 2 (CASPR2). LGI1 antibody is more commonly associated with limbic encephalitis—as illustrated in our case study—in addition to a distinctive type of seizure affecting the arm and face (faciobrachial dystonic seizure).³⁴ The CASPR2 antibody, on the other hand, more often correlates with peripheral nerve manifestations and Morvan syndrome.²⁹ Hyponatremia is commonly seen on serum chemical analysis and provides a clue that these syndromes are present.¹²

Good response to immunotherapy

A critical change in therapy came as clinicians realized that seizures were often refractory to standard antiepileptic drugs but responded well to immunotherapies. On the basis of these observations, sera of patients with long-standing epilepsy have been reanalyzed to look for neuronal autoantibodies.³¹ These antibodies should be checked in cases of new-onset refractory status epilepticus of unknown origin that does not respond to antiepileptic medications.

About half of patients with VGKC complex antibody-mediated limbic encephalitis have normal findings on brain MRI.⁵ Seven of 10 patients who were prospectively followed for VGKC complex antibody-mediated faciobrachial dystonic seizures had normal brain MRIs.³⁵

VGKC complex antibody-mediated limbic encephalitis does not usually recur.³⁶ Most cases are nonparaneoplastic, as evidenced by failure to detect a single active tumor in 64 patients after a median follow-up of 3 years. The prognosis is generally favorable except in cases with coexisting tumors.¹²

ANTI-NMDA RECEPTOR ENCEPHALITIS

Often associated with ovarian teratoma

Anti-NMDA receptor encephalitis typically affects women in their 20s and 30s, and about half of patients have an ovarian teratoma. It can also occur in younger patients and in men, in whom it is less likely to be associated with a neoplasm.³⁷

Typical initial symptoms include striking and often stereotyped neuropsychiatric disturbances manifesting as psychosis, confusion, seizures, and amnesia. After 1 to 2 weeks, new symptoms set in, including reduced consciousness, movement disorders (ranging from orolingualfacial dyskinesia to rigidity and choreoathetosis), autonomic dysfunction, and hypoventilation, often prompting admission to the intensive care unit.³⁸

Although the outcome is favorable in most cases, recovery, in contrast to VGKC complex antibody-mediated limbic encephalitis, is slow and may take longer than 1 year. Up to a quarter of patients have a relapse, underscoring the importance of maintenance immunotherapy.

It is important to undertake an intensive search for possible ovarian and extraovarian teratomas in young women with this syndrome—including CT of the pelvis, vaginal ultrasonography, and PET imaging—as removal of the teratoma may be curative.³⁷

DIAGNOSIS OF AUTOANTIBODY-MEDIATED ENCEPHALITIS

Critical to diagnosing autoantibody-mediated encephalitis is awareness of these disorders. Since antibody testing may be very specific and is not usually part of the standard batteries of tests, a high level of suspicion is needed. Patients may present to different specialists in different settings; therefore, clinicians in pediatrics, rheumatology, psychiatry, and intensive care medicine need to be aware of these syndromes to avoid delay and misdiagnosis.

Clinical features suggesting autoantibody-mediated encephalitis include:

• Acute or subacute onset of a neurologic syndrome
• New-onset refractory status epilepticus of unknown etiology
• Acute or subacute psychiatric illness with unexpected progression to neurologic symptoms or delirium
• Unusual movement disorders not conforming to standard syndromes
• Cognitive impairment, psychosis, or behavioral or language disorders with atypical findings on imaging or cerebrospinal fluid analysis.

Imaging. Diagnosis of autoantibody-mediated encephalitis focuses on evidence suggesting an inflammatory central nervous system syndrome. MRI may show hyperintense signals on T2, FLAIR, or diffusion-weighted imaging changes in various brain regions. In many cases, however, MRI is negative despite severe clinical symptoms. In a study of 72 patients suspected of having autoimmune dementia of various etiologies, including but not restricted to antineuronal surface antibody-mediated causes, Flanagan et al³⁹ identified atypical neuroimaging findings in only 29%. PET imaging may show hypermetabolism in certain brain areas correlating to clinical syndromes but is often difficult to obtain in a timely fashion.

Cerebrospinal fluid is often abnormal, showing elevated protein, increased immunoglobulin G synthesis, or oligoclonal banding. As with imaging studies, the cerebrospinal fluid may be normal despite severe clinical manifestations.

Electroencephalography may show focal slowing or seizure activity. Neuropsychologic testing may show different patterns of abnormalities.

Antibody testing. None of these tests can be used in isolation, and the diagnosis of autoantibody-mediated encephalitis hinges on recognizing a clinical syndrome and ordering supportive testing. Specific antibodies are more likely in different clinical syndromes and should be sought (Table 3).

Patients who have autoantibody-mediated encephalitis may test negative for autoantibodies for many possible reasons:

• Blood testing for antibodies may be less sensitive than cerebrospinal fluid testing
• Antibody titers may vary in the course of the disease
• The patient may be expressing an antibody that is less often tested for (eg, anti-AMPA receptor or antigamma-aminobutyric acid B) or one that has not yet been isolated.

Evaluating for malignancy is recommended in all cases of autoantibody-mediated encephalitis. The initial workup may involve CT of the chest, abdomen, and pelvis, as well as mammography in women and serum prostate-specific antigen testing and testicular ultrasonography in men. Ordering FDG-PET in cases in which CT is negative or inconclusive increases cancer detection.⁴⁰ If no cancer is found, close tumor surveillance—every 3 to 6 months—is recommended for at least 2 years.⁴¹

TREATMENT

Owing in large part to the rarity of autoantibody-mediated encephalitides, no randomized trials of therapy have been performed. Treatment at present is guided mostly by case series and expert consensus, which suggest first-line therapy with intravenous immunoglobulin, high-dose corticosteroids, plasmapheresis, or a combination.

Different syndromes and antibody-related disorders respond differently to therapy. Syndromes associated with antibodies against intracellular antigens tend to be more resistant to immune therapy than cell surface antigen-related syndromes.⁴

Tiered approach

Combined treatment with intravenous immunoglobulin and high-dose corticosteroids may be superior to treatment with steroids alone for LGI1-antibody mediated limbic encephalitis.⁴²

In cases refractory to first-line (“tier 1”) therapy, second-line immunotherapy with drugs affecting B-cell populations (eg, rituximab, cyclophosphamide, and mycophenolate mofetil) has been used.

A tiered approach has been most extensively studied for anti-NMDA-receptor encephalitis, with better outcomes found using second-line therapy.⁴³

Treatment strategies for these disorders will likely evolve over time with additional experience.

Outpatient management

Once the patient is discharged from the hospital, a multidisciplinary approach to care is recommended, including physical rehabilitation, speech therapy, neuropsychiatric and neuroimmunologic follow-up, and annual surveillance for malignancies.

A 79-year-old woman with a history of breast cancer in remission and hypertension presented to a local emergency department because of subacute memory loss and compulsive shopping. Her serum sodium concentration was 127 mmol/L (reference range 132–148). Computed tomography (CT) and magnetic resonance imaging (MRI) of the brain were normal, and she was sent home.

Three days later, she experienced a generalized tonic-clonic seizure that evolved into status epilepticus. She was intubated and admitted to the intensive care unit. Cerebrospinal fluid analysis was normal, and infectious causes of encephalitis were ruled out. MRI showed increased signal in both hippocampi (Figure 1). Her seizures were refractory to treatment, and she was given pentobarbital to induce a coma.

Serum evaluation of neuronal antibodies revealed elevated titers of the voltage-gated potassium channel (VGKC) complex antibody, with subsequent subtyping confirming the leucine-rich glioma-inactivated protein 1 (LGI1) protein as the antigenic target.

She received a 5-day course of intravenous immunoglobulin and methylprednisolone, pentobarbital was withdrawn, and the seizures did not recur, but weeks later she remained comatose. Positron emission tomography (PET) of the brain revealed hypermetabolism in the medial and anterior aspects of both temporal lobes. She underwent five sessions of plasma exchange, after which she began to improve and follow commands. She was ultimately discharged to an acute rehabilitation facility after a 4-week hospital stay.

She received infusions of intravenous immunoglobulin twice a month for 6 months.  At her last follow-up visit, she was seizure-free and neurologically intact except for mild inattention.

NEWLY RECOGNIZED DISEASES

Although autoantibody-mediated encephalitic syndromes were first described more than 50 years ago,^1,2 their autoimmune basis was not recognized until the early 1980s.³ In the past 10 years, a flood of novel clinical syndromes associated with neuronal autoantibodies has been described that may be markedly improved or even completely resolved with immunotherapy. In cases of unexplained seizure, encephalitis, or acute-onset psychiatric syndromes, suspecting these syndromes can lead to diagnosis, treatment, and a good outcome.

This review describes the key clinical autoantibody-mediated encephalitic syndromes, explains the better-characterized antibody associations, and discusses their diagnosis and treatment. 

CLASSIFIED ANATOMICALLY, IMMUNOLOGICALLY, OR EPONYMOUSLY

Autoantibody-mediated encephalitis is also known as autoimmune-mediated encephalitis, autoimmune-mediated limbic encephalitis, and autoimmune synaptic encephalitis.

How to categorize these syndromes is still in flux: they can be listed by the area of the brain affected, the antibody involved, or the name of the discoverer (eg, Morvan syndrome).

Autoantibodies identified in autoimmune encephalitis fall under two broad categories:

• Those targeting intracellular (intranuclear or intracytoplasmic) antigens; the syndromes they cause are more likely to be paraneoplastic and less responsive to immunotherapy
• Those targeting antigens on the neuronal surface: the syndromes they cause are less likely to be paraneoplastic and are more responsive to immunotherapy.⁴

SYNDROMES DEFINED BY BRAIN AREA AFFECTED

Below, we provide examples of neurologic syndromes of autoantibody-mediated encephalitis according to the region of the brain most affected, ie, the limbic system, the brainstem, or the cerebellum (Figure 2).

LIMBIC ENCEPHALITIS

Memory loss, behavioral changes, seizures

Patients with limbic encephalitis (such as the patient described in the vignette above) present with symptoms attributed to dysfunction of mesial temporal lobe structures, most notably the hippocampus. Prominent symptoms include short-term memory loss, behavioral disturbances such as agitation and confusion, and psychiatric problems such as depression and psychosis. Recurrent seizures are a salient feature and, not uncommonly, progress to status epilepticus.

Antibodies are not all cancer-associated

Cerebrospinal fluid analysis can be normal or show abnormalities suggesting immune activation, eg, slight pleocytosis, elevated protein, increased immunoglobulin G synthesis, and oligoclonal banding.⁵

In many cases, an autoantibody is found in the blood or in the cerebrospinal fluid. Some patients may express more than one autoantibody, so the traditional view of “one antibody, one syndrome” is incorrect.

Although initially identified as a rare paraneoplastic disorder, limbic encephalitis sometimes occurs in the absence of malignancy.

Multiple antibodies have been linked to the syndrome (Table 1).^6–9 The “classic” antibodies initially found in paraneoplastic forms are now generally viewed as nonpathogenic, in part because they are directed against intracellular antigens. Neuronal injury in paraneo­plastic limbic encephalitis is believed to be mediated by cytotoxic T lymphocytes, with neuronal autoantibodies being produced after the injury.⁴ Recently defined antibodies, such as those targeting the N-methyl-d-aspartate (NMDA) receptor⁶ and the LGI1 protein,⁷ are now understood to be common causes of limbic encephalitis. 

Imaging usually shows limbic focal changes

Structural MRI or functional fluorodeoxyglucose (FDG)-PET imaging may show focal changes in limbic system structures, such as the mesial temporal lobes.  It is now recognized that other cortical areas may be involved, and the term “limbic encephalitis” may give way to “cortical” or “focal encephalitis.”

In about 60% of patients, MRI shows hyperintense fluid-attenuated inversion recovery (FLAIR) or T2 signal changes in the mesial temporal lobes, likely reflecting inflammatory changes.^4,10,11 On FDG-PET, hypermetabolism may be observed in the mesial temporal lobes early in the disease despite normal findings on MRI.¹² Hypometabolism, either diffuse or localized to the mesial temporal lobes, eventually sets in, likely reflecting cytotoxic injury in the aftermath of prolonged inflammation or seizures.

Consider other causes

Before diagnosing limbic encephalitis, it is essential to evaluate for infectious meningoencephalitis, especially herpes simplex viral encephalitis. Thiamine deficiency (Wernicke encephalopathy), drug intoxication, prion disease, Hashimoto encephalopathy, tumor, and subclinical status epilepticus should also be considered. Some of these conditions are associated with the same neuronal autoantibodies detected in limbic encephalitis. Further complicating the picture, case reports have shown the presence of serum neuronal autoantibodies—VGKC complex^13–15 and NMDA-receptor antibodies^16,17—in confirmed cases of prion disease. In addition, adequately treated herpes simplex viral encephalitis can precipitate the production of NMDA-receptor antibodies and their characteristic syndrome.^18–20

BRAINSTEM ENCEPHALITIS

The brainstem—the midbrain, pons, and medulla—can be affected, either in isolation or more commonly as part of a more widespread autoantibody-mediated encephalitis. Symptoms and signs include eye movement abnormalities, ptosis, dysphagia, dysarthria, ataxia, facial palsy, vertigo, hearing impairment, reduced consciousness, and hypoventilation.²¹

Anti-Hu, anti-Ri, and anti-Ma2 antibodies are most commonly associated with brainstem encephalitis (Table 2). Anti-Ma2-associated encephalitis may improve after a combination of immunotherapy and tumor removal²¹; the others have a poor prognosis.

Neuromyelitis optica spectrum disorders

Neuromyelitis optica spectrum disorders most commonly involve demyelination affecting the optic nerves and spinal cord, leading to  unilateral or bilateral optic neuritis and transverse myelitis spanning three or more vertebral segments.²² The initial clinical manifestation may be an encephalitic pattern, affecting predominantly the brainstem in a restricted fashion,²² or the central nervous system in a more diffuse pattern, mimicking either acute disseminated encephalomyelitis or, in less severe cases, posterior reversible encephalopathy syndrome.²³

Testing for antiaquaporin-4 antibody, also known as neuromyelitis optica immunoglobulin G, is the single most decisive laboratory test for diagnosing neuromyelitis optica spectrum disorders, so serum and cerebrospinal fluid evaluation for this autoantibody should be considered when caring for a patient whose clinical picture suggests brainstem encephalitis.²²

Bickerstaff brainstem encephalitis

Bickerstaff brainstem encephalitis was first described more than half a century ago in patients with postinfectious ataxia, ophthalmoparesis, and altered consciousness. This rare disease was later found to be associated with antiganglioside GQ1b (anti-GQ1b) autoantibody. MRI is normal in about 90% of cases, so recognizing the clinical presentation and analyzing anti-GQ1b serum titers are critical to diagnosis.

Recovery is usually spontaneous and complete and can be hastened by immunotherapy, especially intravenous immunoglobulin.²⁴

Other causes of brainstem encephalitis

The differential diagnosis of a presentation of brainstem encephalitis includes:

• Infectious causes, the most common being Listeria species followed by enterovirus 71 and herpes simplex virus.²⁵ Tuberculosis, brucellosis, and Whipple disease should also be considered.
• Primary central nervous system inflammatory and demyelinating conditions, eg, multiple sclerosis and acute disseminated encephalomyelitis.
• Systemic inflammatory conditions, eg, Beh­çet disease, systemic lupus erythematosus, and sarcoidosis.
• Direct brainstem neoplastic involvement, as might occur in primary central nervous system lymphoma or leptomeningeal carcinomatosis.

CEREBELLAR SYNDROME

Patients with autoantibody-mediated encephalitis localized predominantly to the cerebellum typically present with dizziness, vertigo, and unsteady gait, progressing eventually to limb and gait ataxia.⁴ Symptoms are often subacute, progressing over weeks.

Multiple neuronal autoantibodies have been found to occur with cerebellar encephalitis (Table 2). In most cases, they are paraneo­plastic and considered not to be pathogenic, given the intracellular location of their target antigen.⁴ In such cases, the syndrome is more accurately described as autoantibody-associated rather than autoantibody-mediated. Only in a minority of cases have neuronal autoantibodies been demonstrated to be directly pathogenic, ie, antimetabotropic glutamate receptor type 1 (anti-mGluR1) antibody-associated cerebellitis²⁶ and antiglutamic acid decarboxylase (anti-GAD)-associated cerebellar ataxia.²⁷

Differential diagnosis of cerebellar syndromes

The differential diagnosis of autoantibody-associated cerebellar syndromes is broad and includes:

• Alcohol-induced atrophy
• Drug-induced cerebellar atrophy (eg, from lithium, phenytoin, gabapentin, metronidazole, amiodarone, carbamazepine)
• Vitamin B₁ and E deficiency
• Hypothyroidism, hypoparathyroidism
• Neurodegenerative disease (eg, prion disease, multiple system atrophy)
• Parainfectious causes (eg, after infection with Epstein-Barr virus)
• Immune-mediated diseases (Miller-Fisher syndrome, associated with anti-GQ1b antibodies, and antigliadin-associated ataxia, which can occur in isolation or as part of celiac disease).⁴

SYNDROMES ASSOCIATED WITH SPECIFIC ANTIBODIES

A few of the autoantibody-mediated encephalitic syndromes have specific antibody associations and characteristic clinical presentations. The most prominent of these syndromes are VGKC complex antibody encephalitis (as in the patient described at the beginning of this article) and anti-NMDA receptor encephalitis.

VGKC COMPLEX ANTIBODY-MEDIATED LIMBIC ENCEPHALITIS

VGKC complex antibodies, initially reported to be associated with the peripheral nerve hyperexcitability disorder neuromyotonia, were subsequently found in Morvan syndrome.^28,29 Patients with this syndrome often present with autonomic dysfunction and peripheral nerve hyperexcitability but also develop insomnia, confusion, hallucinations, and memory loss. Drawing on the clinical overlap between Morvan syndrome and limbic encephalitis, Buckley et al³⁰ were the first to report VGKC complex antibodies in two cases of limbic encephalitis.

VGKC complex antibodies are now understood to be associated with a wide variety of neurologic conditions, including chronic idiopathic pain, epilepsy,³¹ movement disorders, cranial nerve abnormalities, autonomic dysfunction,³² and gut dysmotility.³³ In contrast, these antibodies are rare in healthy people.³⁴ Limbic encephalitis associated with VGKC complex antibody usually lacks cerebellar and brainstem dysfunction, which may help distinguish it from other types of autoantibody-mediated limbic encephalitis.¹²

VGKC complex antibody does not bind to the potassium channel itself. Instead it recognizes other constituents of the channel complex, most notably LGI1 and contactin-associated protein 2 (CASPR2). LGI1 antibody is more commonly associated with limbic encephalitis—as illustrated in our case study—in addition to a distinctive type of seizure affecting the arm and face (faciobrachial dystonic seizure).³⁴ The CASPR2 antibody, on the other hand, more often correlates with peripheral nerve manifestations and Morvan syndrome.²⁹ Hyponatremia is commonly seen on serum chemical analysis and provides a clue that these syndromes are present.¹²

Good response to immunotherapy

A critical change in therapy came as clinicians realized that seizures were often refractory to standard antiepileptic drugs but responded well to immunotherapies. On the basis of these observations, sera of patients with long-standing epilepsy have been reanalyzed to look for neuronal autoantibodies.³¹ These antibodies should be checked in cases of new-onset refractory status epilepticus of unknown origin that does not respond to antiepileptic medications.

About half of patients with VGKC complex antibody-mediated limbic encephalitis have normal findings on brain MRI.⁵ Seven of 10 patients who were prospectively followed for VGKC complex antibody-mediated faciobrachial dystonic seizures had normal brain MRIs.³⁵

VGKC complex antibody-mediated limbic encephalitis does not usually recur.³⁶ Most cases are nonparaneoplastic, as evidenced by failure to detect a single active tumor in 64 patients after a median follow-up of 3 years. The prognosis is generally favorable except in cases with coexisting tumors.¹²

ANTI-NMDA RECEPTOR ENCEPHALITIS

Often associated with ovarian teratoma

Anti-NMDA receptor encephalitis typically affects women in their 20s and 30s, and about half of patients have an ovarian teratoma. It can also occur in younger patients and in men, in whom it is less likely to be associated with a neoplasm.³⁷

Typical initial symptoms include striking and often stereotyped neuropsychiatric disturbances manifesting as psychosis, confusion, seizures, and amnesia. After 1 to 2 weeks, new symptoms set in, including reduced consciousness, movement disorders (ranging from orolingualfacial dyskinesia to rigidity and choreoathetosis), autonomic dysfunction, and hypoventilation, often prompting admission to the intensive care unit.³⁸

Although the outcome is favorable in most cases, recovery, in contrast to VGKC complex antibody-mediated limbic encephalitis, is slow and may take longer than 1 year. Up to a quarter of patients have a relapse, underscoring the importance of maintenance immunotherapy.

It is important to undertake an intensive search for possible ovarian and extraovarian teratomas in young women with this syndrome—including CT of the pelvis, vaginal ultrasonography, and PET imaging—as removal of the teratoma may be curative.³⁷

DIAGNOSIS OF AUTOANTIBODY-MEDIATED ENCEPHALITIS

Critical to diagnosing autoantibody-mediated encephalitis is awareness of these disorders. Since antibody testing may be very specific and is not usually part of the standard batteries of tests, a high level of suspicion is needed. Patients may present to different specialists in different settings; therefore, clinicians in pediatrics, rheumatology, psychiatry, and intensive care medicine need to be aware of these syndromes to avoid delay and misdiagnosis.

Clinical features suggesting autoantibody-mediated encephalitis include:

• Acute or subacute onset of a neurologic syndrome
• New-onset refractory status epilepticus of unknown etiology
• Acute or subacute psychiatric illness with unexpected progression to neurologic symptoms or delirium
• Unusual movement disorders not conforming to standard syndromes
• Cognitive impairment, psychosis, or behavioral or language disorders with atypical findings on imaging or cerebrospinal fluid analysis.

Imaging. Diagnosis of autoantibody-mediated encephalitis focuses on evidence suggesting an inflammatory central nervous system syndrome. MRI may show hyperintense signals on T2, FLAIR, or diffusion-weighted imaging changes in various brain regions. In many cases, however, MRI is negative despite severe clinical symptoms. In a study of 72 patients suspected of having autoimmune dementia of various etiologies, including but not restricted to antineuronal surface antibody-mediated causes, Flanagan et al³⁹ identified atypical neuroimaging findings in only 29%. PET imaging may show hypermetabolism in certain brain areas correlating to clinical syndromes but is often difficult to obtain in a timely fashion.

Cerebrospinal fluid is often abnormal, showing elevated protein, increased immunoglobulin G synthesis, or oligoclonal banding. As with imaging studies, the cerebrospinal fluid may be normal despite severe clinical manifestations.

Electroencephalography may show focal slowing or seizure activity. Neuropsychologic testing may show different patterns of abnormalities.

Antibody testing. None of these tests can be used in isolation, and the diagnosis of autoantibody-mediated encephalitis hinges on recognizing a clinical syndrome and ordering supportive testing. Specific antibodies are more likely in different clinical syndromes and should be sought (Table 3).

Patients who have autoantibody-mediated encephalitis may test negative for autoantibodies for many possible reasons:

• Blood testing for antibodies may be less sensitive than cerebrospinal fluid testing
• Antibody titers may vary in the course of the disease
• The patient may be expressing an antibody that is less often tested for (eg, anti-AMPA receptor or antigamma-aminobutyric acid B) or one that has not yet been isolated.

Evaluating for malignancy is recommended in all cases of autoantibody-mediated encephalitis. The initial workup may involve CT of the chest, abdomen, and pelvis, as well as mammography in women and serum prostate-specific antigen testing and testicular ultrasonography in men. Ordering FDG-PET in cases in which CT is negative or inconclusive increases cancer detection.⁴⁰ If no cancer is found, close tumor surveillance—every 3 to 6 months—is recommended for at least 2 years.⁴¹

TREATMENT

Owing in large part to the rarity of autoantibody-mediated encephalitides, no randomized trials of therapy have been performed. Treatment at present is guided mostly by case series and expert consensus, which suggest first-line therapy with intravenous immunoglobulin, high-dose corticosteroids, plasmapheresis, or a combination.

Different syndromes and antibody-related disorders respond differently to therapy. Syndromes associated with antibodies against intracellular antigens tend to be more resistant to immune therapy than cell surface antigen-related syndromes.⁴

Tiered approach

Combined treatment with intravenous immunoglobulin and high-dose corticosteroids may be superior to treatment with steroids alone for LGI1-antibody mediated limbic encephalitis.⁴²

In cases refractory to first-line (“tier 1”) therapy, second-line immunotherapy with drugs affecting B-cell populations (eg, rituximab, cyclophosphamide, and mycophenolate mofetil) has been used.

A tiered approach has been most extensively studied for anti-NMDA-receptor encephalitis, with better outcomes found using second-line therapy.⁴³

Treatment strategies for these disorders will likely evolve over time with additional experience.

Outpatient management

Once the patient is discharged from the hospital, a multidisciplinary approach to care is recommended, including physical rehabilitation, speech therapy, neuropsychiatric and neuroimmunologic follow-up, and annual surveillance for malignancies.

A 79-year-old woman with a history of breast cancer in remission and hypertension presented to a local emergency department because of subacute memory loss and compulsive shopping. Her serum sodium concentration was 127 mmol/L (reference range 132–148). Computed tomography (CT) and magnetic resonance imaging (MRI) of the brain were normal, and she was sent home.

Three days later, she experienced a generalized tonic-clonic seizure that evolved into status epilepticus. She was intubated and admitted to the intensive care unit. Cerebrospinal fluid analysis was normal, and infectious causes of encephalitis were ruled out. MRI showed increased signal in both hippocampi (Figure 1). Her seizures were refractory to treatment, and she was given pentobarbital to induce a coma.

Serum evaluation of neuronal antibodies revealed elevated titers of the voltage-gated potassium channel (VGKC) complex antibody, with subsequent subtyping confirming the leucine-rich glioma-inactivated protein 1 (LGI1) protein as the antigenic target.

She received a 5-day course of intravenous immunoglobulin and methylprednisolone, pentobarbital was withdrawn, and the seizures did not recur, but weeks later she remained comatose. Positron emission tomography (PET) of the brain revealed hypermetabolism in the medial and anterior aspects of both temporal lobes. She underwent five sessions of plasma exchange, after which she began to improve and follow commands. She was ultimately discharged to an acute rehabilitation facility after a 4-week hospital stay.

She received infusions of intravenous immunoglobulin twice a month for 6 months.  At her last follow-up visit, she was seizure-free and neurologically intact except for mild inattention.

NEWLY RECOGNIZED DISEASES

Although autoantibody-mediated encephalitic syndromes were first described more than 50 years ago,^1,2 their autoimmune basis was not recognized until the early 1980s.³ In the past 10 years, a flood of novel clinical syndromes associated with neuronal autoantibodies has been described that may be markedly improved or even completely resolved with immunotherapy. In cases of unexplained seizure, encephalitis, or acute-onset psychiatric syndromes, suspecting these syndromes can lead to diagnosis, treatment, and a good outcome.

This review describes the key clinical autoantibody-mediated encephalitic syndromes, explains the better-characterized antibody associations, and discusses their diagnosis and treatment. 

CLASSIFIED ANATOMICALLY, IMMUNOLOGICALLY, OR EPONYMOUSLY

Autoantibody-mediated encephalitis is also known as autoimmune-mediated encephalitis, autoimmune-mediated limbic encephalitis, and autoimmune synaptic encephalitis.

How to categorize these syndromes is still in flux: they can be listed by the area of the brain affected, the antibody involved, or the name of the discoverer (eg, Morvan syndrome).

Autoantibodies identified in autoimmune encephalitis fall under two broad categories:

• Those targeting intracellular (intranuclear or intracytoplasmic) antigens; the syndromes they cause are more likely to be paraneoplastic and less responsive to immunotherapy
• Those targeting antigens on the neuronal surface: the syndromes they cause are less likely to be paraneoplastic and are more responsive to immunotherapy.⁴

SYNDROMES DEFINED BY BRAIN AREA AFFECTED

Below, we provide examples of neurologic syndromes of autoantibody-mediated encephalitis according to the region of the brain most affected, ie, the limbic system, the brainstem, or the cerebellum (Figure 2).

LIMBIC ENCEPHALITIS

Memory loss, behavioral changes, seizures

Patients with limbic encephalitis (such as the patient described in the vignette above) present with symptoms attributed to dysfunction of mesial temporal lobe structures, most notably the hippocampus. Prominent symptoms include short-term memory loss, behavioral disturbances such as agitation and confusion, and psychiatric problems such as depression and psychosis. Recurrent seizures are a salient feature and, not uncommonly, progress to status epilepticus.

Antibodies are not all cancer-associated

Cerebrospinal fluid analysis can be normal or show abnormalities suggesting immune activation, eg, slight pleocytosis, elevated protein, increased immunoglobulin G synthesis, and oligoclonal banding.⁵

In many cases, an autoantibody is found in the blood or in the cerebrospinal fluid. Some patients may express more than one autoantibody, so the traditional view of “one antibody, one syndrome” is incorrect.

Although initially identified as a rare paraneoplastic disorder, limbic encephalitis sometimes occurs in the absence of malignancy.

Multiple antibodies have been linked to the syndrome (Table 1).^6–9 The “classic” antibodies initially found in paraneoplastic forms are now generally viewed as nonpathogenic, in part because they are directed against intracellular antigens. Neuronal injury in paraneo­plastic limbic encephalitis is believed to be mediated by cytotoxic T lymphocytes, with neuronal autoantibodies being produced after the injury.⁴ Recently defined antibodies, such as those targeting the N-methyl-d-aspartate (NMDA) receptor⁶ and the LGI1 protein,⁷ are now understood to be common causes of limbic encephalitis. 

Imaging usually shows limbic focal changes

Structural MRI or functional fluorodeoxyglucose (FDG)-PET imaging may show focal changes in limbic system structures, such as the mesial temporal lobes.  It is now recognized that other cortical areas may be involved, and the term “limbic encephalitis” may give way to “cortical” or “focal encephalitis.”

In about 60% of patients, MRI shows hyperintense fluid-attenuated inversion recovery (FLAIR) or T2 signal changes in the mesial temporal lobes, likely reflecting inflammatory changes.^4,10,11 On FDG-PET, hypermetabolism may be observed in the mesial temporal lobes early in the disease despite normal findings on MRI.¹² Hypometabolism, either diffuse or localized to the mesial temporal lobes, eventually sets in, likely reflecting cytotoxic injury in the aftermath of prolonged inflammation or seizures.

Consider other causes

Before diagnosing limbic encephalitis, it is essential to evaluate for infectious meningoencephalitis, especially herpes simplex viral encephalitis. Thiamine deficiency (Wernicke encephalopathy), drug intoxication, prion disease, Hashimoto encephalopathy, tumor, and subclinical status epilepticus should also be considered. Some of these conditions are associated with the same neuronal autoantibodies detected in limbic encephalitis. Further complicating the picture, case reports have shown the presence of serum neuronal autoantibodies—VGKC complex^13–15 and NMDA-receptor antibodies^16,17—in confirmed cases of prion disease. In addition, adequately treated herpes simplex viral encephalitis can precipitate the production of NMDA-receptor antibodies and their characteristic syndrome.^18–20

BRAINSTEM ENCEPHALITIS

The brainstem—the midbrain, pons, and medulla—can be affected, either in isolation or more commonly as part of a more widespread autoantibody-mediated encephalitis. Symptoms and signs include eye movement abnormalities, ptosis, dysphagia, dysarthria, ataxia, facial palsy, vertigo, hearing impairment, reduced consciousness, and hypoventilation.²¹

Anti-Hu, anti-Ri, and anti-Ma2 antibodies are most commonly associated with brainstem encephalitis (Table 2). Anti-Ma2-associated encephalitis may improve after a combination of immunotherapy and tumor removal²¹; the others have a poor prognosis.

Neuromyelitis optica spectrum disorders

Neuromyelitis optica spectrum disorders most commonly involve demyelination affecting the optic nerves and spinal cord, leading to  unilateral or bilateral optic neuritis and transverse myelitis spanning three or more vertebral segments.²² The initial clinical manifestation may be an encephalitic pattern, affecting predominantly the brainstem in a restricted fashion,²² or the central nervous system in a more diffuse pattern, mimicking either acute disseminated encephalomyelitis or, in less severe cases, posterior reversible encephalopathy syndrome.²³

Testing for antiaquaporin-4 antibody, also known as neuromyelitis optica immunoglobulin G, is the single most decisive laboratory test for diagnosing neuromyelitis optica spectrum disorders, so serum and cerebrospinal fluid evaluation for this autoantibody should be considered when caring for a patient whose clinical picture suggests brainstem encephalitis.²²

Bickerstaff brainstem encephalitis

Bickerstaff brainstem encephalitis was first described more than half a century ago in patients with postinfectious ataxia, ophthalmoparesis, and altered consciousness. This rare disease was later found to be associated with antiganglioside GQ1b (anti-GQ1b) autoantibody. MRI is normal in about 90% of cases, so recognizing the clinical presentation and analyzing anti-GQ1b serum titers are critical to diagnosis.

Recovery is usually spontaneous and complete and can be hastened by immunotherapy, especially intravenous immunoglobulin.²⁴

Other causes of brainstem encephalitis

The differential diagnosis of a presentation of brainstem encephalitis includes:

• Infectious causes, the most common being Listeria species followed by enterovirus 71 and herpes simplex virus.²⁵ Tuberculosis, brucellosis, and Whipple disease should also be considered.
• Primary central nervous system inflammatory and demyelinating conditions, eg, multiple sclerosis and acute disseminated encephalomyelitis.
• Systemic inflammatory conditions, eg, Beh­çet disease, systemic lupus erythematosus, and sarcoidosis.
• Direct brainstem neoplastic involvement, as might occur in primary central nervous system lymphoma or leptomeningeal carcinomatosis.

CEREBELLAR SYNDROME

Patients with autoantibody-mediated encephalitis localized predominantly to the cerebellum typically present with dizziness, vertigo, and unsteady gait, progressing eventually to limb and gait ataxia.⁴ Symptoms are often subacute, progressing over weeks.

Multiple neuronal autoantibodies have been found to occur with cerebellar encephalitis (Table 2). In most cases, they are paraneo­plastic and considered not to be pathogenic, given the intracellular location of their target antigen.⁴ In such cases, the syndrome is more accurately described as autoantibody-associated rather than autoantibody-mediated. Only in a minority of cases have neuronal autoantibodies been demonstrated to be directly pathogenic, ie, antimetabotropic glutamate receptor type 1 (anti-mGluR1) antibody-associated cerebellitis²⁶ and antiglutamic acid decarboxylase (anti-GAD)-associated cerebellar ataxia.²⁷

Differential diagnosis of cerebellar syndromes

The differential diagnosis of autoantibody-associated cerebellar syndromes is broad and includes:

• Alcohol-induced atrophy
• Drug-induced cerebellar atrophy (eg, from lithium, phenytoin, gabapentin, metronidazole, amiodarone, carbamazepine)
• Vitamin B₁ and E deficiency
• Hypothyroidism, hypoparathyroidism
• Neurodegenerative disease (eg, prion disease, multiple system atrophy)
• Parainfectious causes (eg, after infection with Epstein-Barr virus)
• Immune-mediated diseases (Miller-Fisher syndrome, associated with anti-GQ1b antibodies, and antigliadin-associated ataxia, which can occur in isolation or as part of celiac disease).⁴

SYNDROMES ASSOCIATED WITH SPECIFIC ANTIBODIES

A few of the autoantibody-mediated encephalitic syndromes have specific antibody associations and characteristic clinical presentations. The most prominent of these syndromes are VGKC complex antibody encephalitis (as in the patient described at the beginning of this article) and anti-NMDA receptor encephalitis.

VGKC COMPLEX ANTIBODY-MEDIATED LIMBIC ENCEPHALITIS

VGKC complex antibodies, initially reported to be associated with the peripheral nerve hyperexcitability disorder neuromyotonia, were subsequently found in Morvan syndrome.^28,29 Patients with this syndrome often present with autonomic dysfunction and peripheral nerve hyperexcitability but also develop insomnia, confusion, hallucinations, and memory loss. Drawing on the clinical overlap between Morvan syndrome and limbic encephalitis, Buckley et al³⁰ were the first to report VGKC complex antibodies in two cases of limbic encephalitis.

VGKC complex antibodies are now understood to be associated with a wide variety of neurologic conditions, including chronic idiopathic pain, epilepsy,³¹ movement disorders, cranial nerve abnormalities, autonomic dysfunction,³² and gut dysmotility.³³ In contrast, these antibodies are rare in healthy people.³⁴ Limbic encephalitis associated with VGKC complex antibody usually lacks cerebellar and brainstem dysfunction, which may help distinguish it from other types of autoantibody-mediated limbic encephalitis.¹²

VGKC complex antibody does not bind to the potassium channel itself. Instead it recognizes other constituents of the channel complex, most notably LGI1 and contactin-associated protein 2 (CASPR2). LGI1 antibody is more commonly associated with limbic encephalitis—as illustrated in our case study—in addition to a distinctive type of seizure affecting the arm and face (faciobrachial dystonic seizure).³⁴ The CASPR2 antibody, on the other hand, more often correlates with peripheral nerve manifestations and Morvan syndrome.²⁹ Hyponatremia is commonly seen on serum chemical analysis and provides a clue that these syndromes are present.¹²

Good response to immunotherapy

A critical change in therapy came as clinicians realized that seizures were often refractory to standard antiepileptic drugs but responded well to immunotherapies. On the basis of these observations, sera of patients with long-standing epilepsy have been reanalyzed to look for neuronal autoantibodies.³¹ These antibodies should be checked in cases of new-onset refractory status epilepticus of unknown origin that does not respond to antiepileptic medications.

About half of patients with VGKC complex antibody-mediated limbic encephalitis have normal findings on brain MRI.⁵ Seven of 10 patients who were prospectively followed for VGKC complex antibody-mediated faciobrachial dystonic seizures had normal brain MRIs.³⁵

VGKC complex antibody-mediated limbic encephalitis does not usually recur.³⁶ Most cases are nonparaneoplastic, as evidenced by failure to detect a single active tumor in 64 patients after a median follow-up of 3 years. The prognosis is generally favorable except in cases with coexisting tumors.¹²

ANTI-NMDA RECEPTOR ENCEPHALITIS

Often associated with ovarian teratoma

Anti-NMDA receptor encephalitis typically affects women in their 20s and 30s, and about half of patients have an ovarian teratoma. It can also occur in younger patients and in men, in whom it is less likely to be associated with a neoplasm.³⁷

Typical initial symptoms include striking and often stereotyped neuropsychiatric disturbances manifesting as psychosis, confusion, seizures, and amnesia. After 1 to 2 weeks, new symptoms set in, including reduced consciousness, movement disorders (ranging from orolingualfacial dyskinesia to rigidity and choreoathetosis), autonomic dysfunction, and hypoventilation, often prompting admission to the intensive care unit.³⁸

Although the outcome is favorable in most cases, recovery, in contrast to VGKC complex antibody-mediated limbic encephalitis, is slow and may take longer than 1 year. Up to a quarter of patients have a relapse, underscoring the importance of maintenance immunotherapy.

It is important to undertake an intensive search for possible ovarian and extraovarian teratomas in young women with this syndrome—including CT of the pelvis, vaginal ultrasonography, and PET imaging—as removal of the teratoma may be curative.³⁷

DIAGNOSIS OF AUTOANTIBODY-MEDIATED ENCEPHALITIS

Critical to diagnosing autoantibody-mediated encephalitis is awareness of these disorders. Since antibody testing may be very specific and is not usually part of the standard batteries of tests, a high level of suspicion is needed. Patients may present to different specialists in different settings; therefore, clinicians in pediatrics, rheumatology, psychiatry, and intensive care medicine need to be aware of these syndromes to avoid delay and misdiagnosis.

Clinical features suggesting autoantibody-mediated encephalitis include:

• Acute or subacute onset of a neurologic syndrome
• New-onset refractory status epilepticus of unknown etiology
• Acute or subacute psychiatric illness with unexpected progression to neurologic symptoms or delirium
• Unusual movement disorders not conforming to standard syndromes
• Cognitive impairment, psychosis, or behavioral or language disorders with atypical findings on imaging or cerebrospinal fluid analysis.

Imaging. Diagnosis of autoantibody-mediated encephalitis focuses on evidence suggesting an inflammatory central nervous system syndrome. MRI may show hyperintense signals on T2, FLAIR, or diffusion-weighted imaging changes in various brain regions. In many cases, however, MRI is negative despite severe clinical symptoms. In a study of 72 patients suspected of having autoimmune dementia of various etiologies, including but not restricted to antineuronal surface antibody-mediated causes, Flanagan et al³⁹ identified atypical neuroimaging findings in only 29%. PET imaging may show hypermetabolism in certain brain areas correlating to clinical syndromes but is often difficult to obtain in a timely fashion.

Cerebrospinal fluid is often abnormal, showing elevated protein, increased immunoglobulin G synthesis, or oligoclonal banding. As with imaging studies, the cerebrospinal fluid may be normal despite severe clinical manifestations.

Electroencephalography may show focal slowing or seizure activity. Neuropsychologic testing may show different patterns of abnormalities.

Antibody testing. None of these tests can be used in isolation, and the diagnosis of autoantibody-mediated encephalitis hinges on recognizing a clinical syndrome and ordering supportive testing. Specific antibodies are more likely in different clinical syndromes and should be sought (Table 3).

Patients who have autoantibody-mediated encephalitis may test negative for autoantibodies for many possible reasons:

• Blood testing for antibodies may be less sensitive than cerebrospinal fluid testing
• Antibody titers may vary in the course of the disease
• The patient may be expressing an antibody that is less often tested for (eg, anti-AMPA receptor or antigamma-aminobutyric acid B) or one that has not yet been isolated.

Evaluating for malignancy is recommended in all cases of autoantibody-mediated encephalitis. The initial workup may involve CT of the chest, abdomen, and pelvis, as well as mammography in women and serum prostate-specific antigen testing and testicular ultrasonography in men. Ordering FDG-PET in cases in which CT is negative or inconclusive increases cancer detection.⁴⁰ If no cancer is found, close tumor surveillance—every 3 to 6 months—is recommended for at least 2 years.⁴¹

TREATMENT

Owing in large part to the rarity of autoantibody-mediated encephalitides, no randomized trials of therapy have been performed. Treatment at present is guided mostly by case series and expert consensus, which suggest first-line therapy with intravenous immunoglobulin, high-dose corticosteroids, plasmapheresis, or a combination.

Different syndromes and antibody-related disorders respond differently to therapy. Syndromes associated with antibodies against intracellular antigens tend to be more resistant to immune therapy than cell surface antigen-related syndromes.⁴

Tiered approach

Combined treatment with intravenous immunoglobulin and high-dose corticosteroids may be superior to treatment with steroids alone for LGI1-antibody mediated limbic encephalitis.⁴²

In cases refractory to first-line (“tier 1”) therapy, second-line immunotherapy with drugs affecting B-cell populations (eg, rituximab, cyclophosphamide, and mycophenolate mofetil) has been used.

A tiered approach has been most extensively studied for anti-NMDA-receptor encephalitis, with better outcomes found using second-line therapy.⁴³

Treatment strategies for these disorders will likely evolve over time with additional experience.

Outpatient management

Once the patient is discharged from the hospital, a multidisciplinary approach to care is recommended, including physical rehabilitation, speech therapy, neuropsychiatric and neuroimmunologic follow-up, and annual surveillance for malignancies.

About half of patients with symptomatic esophageal eosinophilia achieved complete clinical and histologic remission on proton pump inhibitors (PPIs), according to a systematic review and meta-analysis of 33 studies.

“Our results support the concept of PPIs as first-line therapy in both children and adults,” Dr. Alfredo Lucendo of the Servicio de Salud de Castillo–La Mancha, Hospital General de Tomelloso, Albacete, Spain, and his associates wrote in the January issue of Clinical Gastroenterology and Hepatology. “Other effective alternatives, such as dietary or topical steroid therapy, likely might be set aside as second-line treatment, owing to long-term safety concerns (topical steroid therapy) and impairment of quality of life and nutritional inadequacy (dietary interventions).”

©Magnus Manske/Wikimedia Commons/CC SA 3.0

The study also confirmed that esophageal pH monitoring does not accurately predict therapeutic response to PPI therapy. “The performance of this test before histologic reevaluation on PPI therapy should be discouraged,” according to the researchers (Clin Gastroenterol Hepatol. 2015. [doi:10.1016/j.cgh.2015.07.041]). Eosinophilic esophagitis was first described as a distinct disorder about 20 years ago, but only recently was understood to be the most common cause of chronic esophageal symptoms among children and young adults. Some cases are now known to respond to PPI therapy, but reported remission rates have varied depending on study design and patient population, the investigators said. In addition, no one had systematically reviewed studies of PPI-responsive esophageal eosinophilia for quality or to determine the optimal type of PPI, dose, or treatment duration.

Therefore, the investigators searched MEDLINE, EMBASE, SCOPUS, and abstracts from the annual meetings of the American Gastroenterological Association, the American College of Gastroenterology, and United European Gastroenterology, identifying 33 studies of 619 patients with symptomatic esophageal eosinophilia. Eleven of the studies were prospective, of which only two were randomized controlled trials. The researchers defined a histologic response as less than 15 eosinophils per high-powered frame after PPI therapy. “Missing data regarding PPI therapy were common and prevented us from drawing conclusions on the most effective PPI drug and doses,” they said. In addition, most studies lacked structured or objective survey tools or other measures of clinical improvement, making it impossible to rule out self-adapted coping strategies as a main cause of improvement over time.

With those caveats in mind, about 61% of patients in the pooled analysis had a clinical response to PPI therapy (95% confidence interval, 48%-72%), and half achieved clinical and histologic remission (95% CI, 42%-59%), the investigators reported. Therapy was somewhat more effective when administered twice a day instead of once daily, when clinicians used esophageal pH monitoring, and when studies were prospective instead of retrospective, but the differences were not significant. Nor did therapeutic response significantly differ based on the age of patients, type of report, or quality of the study.

The overall findings “should be interpreted with caution because of poor-quality evidence, heterogeneity, and publication bias,” the researchers said. Prospective studies are needed to examine the best PPI, dose, and dosing interval to use in an initial trial in the clinic; to clarify long-term effects and dosing strategies; to assess the ability of PPIs to reverse fibrotic esophageal remodeling; and to examine the effects of the CYP2C19 genotype on clinical and histologic response, they added. “More quality evidence on pediatric PPI-responsive eosinophilic esophagitis is needed urgently,” they emphasized.

The authors reported no funding sources and had no disclosures.

About half of patients with symptomatic esophageal eosinophilia achieved complete clinical and histologic remission on proton pump inhibitors (PPIs), according to a systematic review and meta-analysis of 33 studies.

“Our results support the concept of PPIs as first-line therapy in both children and adults,” Dr. Alfredo Lucendo of the Servicio de Salud de Castillo–La Mancha, Hospital General de Tomelloso, Albacete, Spain, and his associates wrote in the January issue of Clinical Gastroenterology and Hepatology. “Other effective alternatives, such as dietary or topical steroid therapy, likely might be set aside as second-line treatment, owing to long-term safety concerns (topical steroid therapy) and impairment of quality of life and nutritional inadequacy (dietary interventions).”

©Magnus Manske/Wikimedia Commons/CC SA 3.0

The study also confirmed that esophageal pH monitoring does not accurately predict therapeutic response to PPI therapy. “The performance of this test before histologic reevaluation on PPI therapy should be discouraged,” according to the researchers (Clin Gastroenterol Hepatol. 2015. [doi:10.1016/j.cgh.2015.07.041]). Eosinophilic esophagitis was first described as a distinct disorder about 20 years ago, but only recently was understood to be the most common cause of chronic esophageal symptoms among children and young adults. Some cases are now known to respond to PPI therapy, but reported remission rates have varied depending on study design and patient population, the investigators said. In addition, no one had systematically reviewed studies of PPI-responsive esophageal eosinophilia for quality or to determine the optimal type of PPI, dose, or treatment duration.

Therefore, the investigators searched MEDLINE, EMBASE, SCOPUS, and abstracts from the annual meetings of the American Gastroenterological Association, the American College of Gastroenterology, and United European Gastroenterology, identifying 33 studies of 619 patients with symptomatic esophageal eosinophilia. Eleven of the studies were prospective, of which only two were randomized controlled trials. The researchers defined a histologic response as less than 15 eosinophils per high-powered frame after PPI therapy. “Missing data regarding PPI therapy were common and prevented us from drawing conclusions on the most effective PPI drug and doses,” they said. In addition, most studies lacked structured or objective survey tools or other measures of clinical improvement, making it impossible to rule out self-adapted coping strategies as a main cause of improvement over time.

With those caveats in mind, about 61% of patients in the pooled analysis had a clinical response to PPI therapy (95% confidence interval, 48%-72%), and half achieved clinical and histologic remission (95% CI, 42%-59%), the investigators reported. Therapy was somewhat more effective when administered twice a day instead of once daily, when clinicians used esophageal pH monitoring, and when studies were prospective instead of retrospective, but the differences were not significant. Nor did therapeutic response significantly differ based on the age of patients, type of report, or quality of the study.

The overall findings “should be interpreted with caution because of poor-quality evidence, heterogeneity, and publication bias,” the researchers said. Prospective studies are needed to examine the best PPI, dose, and dosing interval to use in an initial trial in the clinic; to clarify long-term effects and dosing strategies; to assess the ability of PPIs to reverse fibrotic esophageal remodeling; and to examine the effects of the CYP2C19 genotype on clinical and histologic response, they added. “More quality evidence on pediatric PPI-responsive eosinophilic esophagitis is needed urgently,” they emphasized.

The authors reported no funding sources and had no disclosures.

About half of patients with symptomatic esophageal eosinophilia achieved complete clinical and histologic remission on proton pump inhibitors (PPIs), according to a systematic review and meta-analysis of 33 studies.

“Our results support the concept of PPIs as first-line therapy in both children and adults,” Dr. Alfredo Lucendo of the Servicio de Salud de Castillo–La Mancha, Hospital General de Tomelloso, Albacete, Spain, and his associates wrote in the January issue of Clinical Gastroenterology and Hepatology. “Other effective alternatives, such as dietary or topical steroid therapy, likely might be set aside as second-line treatment, owing to long-term safety concerns (topical steroid therapy) and impairment of quality of life and nutritional inadequacy (dietary interventions).”

©Magnus Manske/Wikimedia Commons/CC SA 3.0

The study also confirmed that esophageal pH monitoring does not accurately predict therapeutic response to PPI therapy. “The performance of this test before histologic reevaluation on PPI therapy should be discouraged,” according to the researchers (Clin Gastroenterol Hepatol. 2015. [doi:10.1016/j.cgh.2015.07.041]). Eosinophilic esophagitis was first described as a distinct disorder about 20 years ago, but only recently was understood to be the most common cause of chronic esophageal symptoms among children and young adults. Some cases are now known to respond to PPI therapy, but reported remission rates have varied depending on study design and patient population, the investigators said. In addition, no one had systematically reviewed studies of PPI-responsive esophageal eosinophilia for quality or to determine the optimal type of PPI, dose, or treatment duration.

Therefore, the investigators searched MEDLINE, EMBASE, SCOPUS, and abstracts from the annual meetings of the American Gastroenterological Association, the American College of Gastroenterology, and United European Gastroenterology, identifying 33 studies of 619 patients with symptomatic esophageal eosinophilia. Eleven of the studies were prospective, of which only two were randomized controlled trials. The researchers defined a histologic response as less than 15 eosinophils per high-powered frame after PPI therapy. “Missing data regarding PPI therapy were common and prevented us from drawing conclusions on the most effective PPI drug and doses,” they said. In addition, most studies lacked structured or objective survey tools or other measures of clinical improvement, making it impossible to rule out self-adapted coping strategies as a main cause of improvement over time.

With those caveats in mind, about 61% of patients in the pooled analysis had a clinical response to PPI therapy (95% confidence interval, 48%-72%), and half achieved clinical and histologic remission (95% CI, 42%-59%), the investigators reported. Therapy was somewhat more effective when administered twice a day instead of once daily, when clinicians used esophageal pH monitoring, and when studies were prospective instead of retrospective, but the differences were not significant. Nor did therapeutic response significantly differ based on the age of patients, type of report, or quality of the study.

The overall findings “should be interpreted with caution because of poor-quality evidence, heterogeneity, and publication bias,” the researchers said. Prospective studies are needed to examine the best PPI, dose, and dosing interval to use in an initial trial in the clinic; to clarify long-term effects and dosing strategies; to assess the ability of PPIs to reverse fibrotic esophageal remodeling; and to examine the effects of the CYP2C19 genotype on clinical and histologic response, they added. “More quality evidence on pediatric PPI-responsive eosinophilic esophagitis is needed urgently,” they emphasized.

The authors reported no funding sources and had no disclosures.