A, Hematoxylin-eosin–stained section of vastus lateralis muscle showing numerous necrotic and regenerating fibers. Some necrotic fibers are invaded by macrophages (original magnification ×20). B, Acid phosphatase–stained section showing numerous necrotic fibers invaded by macrophages (original magnification ×20).
eTable. Treatment and Course
eFigure. Treatment Response
Kassardjian CD, Lennon VA, Alfugham NB, Mahler M, Milone M. Clinical Features and Treatment Outcomes of Necrotizing Autoimmune Myopathy. JAMA Neurol. 2015;72(9):996-1003. doi:10.1001/jamaneurol.2015.1207
Necrotizing autoimmune myopathy (NAM) is characterized pathologically by necrotic muscle fibers with absent or minimal inflammation. It is often accompanied by statin therapy, connective tissue diseases, cancer, and autoantibodies specific for signal recognition particle (SRP) or 3-hydroxy-3-methylglutaryl–coenzyme A reductase (HMGCR). Data are limited concerning differences among etiologic subgroups and treatment outcomes in NAM.
To describe the clinical, serologic, and electrophysiologic characteristics of NAM, compare patient subgroups, and determine clinical outcome predictors.
Design, Setting, and Participants
We conducted a retrospective review of medical records for 63 adult Mayo Clinic patients assigned the clinical and histopathologic diagnosis of NAM from January 1, 2004, through December 31, 2013. Patients were stratified by presumed cause and autoantibody status.
Main Outcomes and Measures
Clinical, electrophysiologic, and pathologic characteristics were collected and compared among patient subgroups. Predictors of response to treatment were identified by univariate logistic regression.
Lower extremity weakness predominated (46 [73%]). Distal weakness (26 [41%]), dysphagia (22 [35%]), and dyspnea (23 [37%]) were common. Twenty-two patients (35%) were receiving a statin medication at onset, 6 had cancer, and 3 had a connective tissue disease. The median creatine kinase level was 5326 U/L. In 13 patients (24%), SRP-IgG was detected, and in 17 patients (34%), HMGCR-IgG was detected (one-third of whom had not received statin medication). One patient was dual seropositive. Facial weakness was more common in SRP-IgG–positive patients. Myotonic discharges were more common in statin-associated NAM. Prednisone monotherapy was insufficient to control disease in most patients; 30 (90%) of 32 patients required 2 or more immunotherapeutic agents. Relapse occurred in 16 (55%) of 29 patients during immunosuppressant taper or discontinuation. Predictors of favorable outcome were male sex and use of 2 or more immunotherapeutic agents within 3 months of onset.
Conclusions and Relevance
Necrotizing autoimmune myopathy was idiopathic in half of this cohort with clinical and histopathologically defined disease. In the remainder, NAM was associated with statin medication, cancer, or connective tissue disease. One in 4 patients was SRP-IgG positive, and 1 in 3 was HMGCR-IgG positive. The disease was usually not controlled by corticosteroid monotherapy. Presentation, course, and outcomes did not differ significantly in seropositive, seronegative, and statin-associated cases. Early aggressive immunosuppressant therapy improved outcomes, and risk of relapse was high during medication dose reduction or withdrawal.
Necrotizing autoimmune myopathy (NAM) presents with subacute proximal limb muscle weakness and a high serum creatine kinase (CK) level.1- 3 In contrast to idiopathic inflammatory myopathies characterized histopathologically by an inflammatory exudate, biopsied muscle in NAM has prominent fiber necrosis and regeneration with minimal or no inflammation.1,3- 5
Risk factors recognized for NAM include statin exposure,6- 9 connective tissue disease (CTD),1,10 cancer,10,11 and, rarely, human immunodeficiency virus infection.12 Recognized serum autoantibody markers7,8,13 are specific for components of the endoplasmic reticulum–associated signal recognition particle (SRP) in the protein synthetic pathway14,15 or for 3-hydroxy-3-methylglutaryl–CoA reductase (HMGCR),7,8,13 the rate-controlling enzyme of the mevalonate pathway involved in cholesterol synthesis. The cause of NAM is presumed to be autoimmune on the basis of subacute onset, clinical response to immunotherapy, and serum autoantibody profile. Treatment strategies have not been evaluated prospectively and are based on case series and expert opinion. Generally, NAM is considered more refractory to immunotherapy than idiopathic inflammatory myopathies.1,6,14,16 The need for extended immunotherapy is debated.17
Most published studies of NAM have focused on patients with a specific serum autoantibody profile1,7,14,18 or patients included in descriptions of “idiopathic inflammatory myopathies” despite lacking histopathologic evidence of inflammation.19- 21 The frequency and phenotypic differences among patients with different presumed causes of NAM (statin medication, CTD, or cancer) are still not well characterized.18,22
The aim of this study was to document, in a large patient cohort, the clinical, electrophysiologic, serologic, and pathologic characteristics of NAM, as well as treatment strategies and outcomes. We attempted to identify differences among patients with NAM based on associated risk factors and potential determinants of outcome.
We searched the Mayo Clinic patient database from January 1, 2004, through December 31, 2013, using the following keywords: necrotizing autoimmune myopathy, immune-mediated necrotizing myopathy, paraneoplastic myopathy, overlap myopathy, and inflammatory myopathy. We adapted the inclusion criteria from the European Neuro Muscular Center (ENMC) International Workshop on Idiopathic Inflammatory Myopathies3,4: (1) acute or subacute onset of proximally predominant muscle weakness without rash, (2) elevated CK level, (3) abnormal spontaneous activity by electromyography (EMG) revealing fibrillation potentials and short-duration motor unit potentials, and (4) biopsy evidence of necrotic and regenerating myofibers with minimal or no inflammatory infiltrates. Exclusion criteria were not meeting all 4 inclusion criteria, thyroid dysfunction, normal muscle biopsy results or results suggestive of an alternative cause, and a recognized pathogenic genetic mutation or family history of myopathy. Three patients with axial weakness underwent genetic testing for fascioscapulohumeral muscular dystrophy, and 3 patients with prominent EMG myotonic discharges underwent genetic testing for myotonic dystrophy type 2; all results were negative. Five patients with muscle weakness for more than 6 months had normal immunohistochemical study results for caveolin-3, dysferlin, α-dystroglycan, and 3 epitopes of dystrophin. We reviewed history and neurologic findings, laboratory data, drug therapy, clinical response, and muscle histopathologic findings.
Patients were divided into 4 groups regardless of serum autoantibody profile: (1) idiopathic: no risk factor identified; (2) statin associated: receiving statin medication at symptom onset; (3) paraneoplastic: a cancer identified within 2 years of myopathy onset; and (4) CTD associated: CTD known or diagnosed after myopathy onset. The study was approved by the institutional review board at Mayo Clinic, Rochester, Minnesota, and informed consent was not required.
Weakness severity was based on the Medical Research Council (MRC) grade of the weakest muscle group: none (MRC grade 5), mild (MRC grade ≥4/5), moderate (MRC grade 3-4/5), and severe (MRC grade <3/5). Response to treatment was graded as no improvement, mild improvement (1 MRC grade in 1-2 muscle groups, persistently requiring assistance for ambulation and activities of daily living), moderate improvement (>1 MRC grade in multiple muscle groups, requiring minimal assistance with ambulation and with activities of daily living), marked improvement (symptoms and signs of mild weakness, but no functional limitation), and return to baseline (no symptoms or signs of weakness).
A clinically validated tissue-based indirect immunofluorescence assay detected SRP-IgG. Specificity was confirmed by recombinant SRP54 enzyme-linked immunosorbent assay, SRP72 immunoprecipitation, or both.23 Recombinant antigen enzyme-linked immunosorbent assay (QUANTA Lite HMGCR, Inova Diagnostics) detected HMGCR-IgG.24 Antinuclear antibodies were tested by Mayo Medical Laboratories’ Connective Tissue Diseases Cascade (ANA by ELISA).
Categorical variables were compared using the Fisher exact tests and continuous variables using the Wilcoxon rank sum test. Because of the small numbers in the paraneoplastic and CTD groups, comparisons were made between the idiopathic and statin groups. Univariate logistic regression was performed to determine variables predicting good outcome, but multivariate regression was not performed because of the small sample. The variables analyzed were age, sex, risk factor subgroup, duration of symptoms, weakness severity, dyspnea, presence of SRP- or HMGCR-IgG, and combination treatment with 2 or more agents within 3 months of symptom onset. False discovery rate was used to correct for multiple comparisons (false discovery rate set at 0.15),25 and P values less than the calculated q value were considered significant. Statistical analysis was performed using JMP software, version 10.0.0 (SAS Institute Inc).
We identified 63 patients with NAM (Table 1): 32 with idiopathic NAM, 22 with statin-associated NAM, 6 with paraneoplastic NAM, and 3 with CTD-associated NAM. Median age at symptom onset was 62.0 years (range, 31.0-84.0 years); 33 patients (52%) were women. Median symptom duration at presentation was 3.0 months (range, 0.5-14.0 months). The idiopathic group was significantly younger at onset (P < .001).
Muscle weakness at presentation was severe in 32 patients (51%) (Table 1) and generally worse in the lower limbs; distal weakness impairing foot dorsiflexion, finger abduction, or finger extension occurred in 26 (41%). Neck muscle weakness was common and was the predominant concern in 2 individuals. Dysphagia (22 patients [35%]) was more frequent in the idiopathic group (P = .03). Weight loss was reported in 18 patients (29%).
Statins used most commonly were atorvastatin (n = 10) and simvastatin (n = 7); 2 patients used pravastatin, 1 used rosuvastatin; the specific statin was not documented in 2 patients. The cancers diagnosed most commonly were gastrointestinal adenocarcinomas (2 colonic and 1 esophageal), with single cases of lung adenocarcinoma, ovarian adenocarcinoma, and thymoma. Cancer in 4 patients was detected on malignancy screening and followed the diagnosis of NAM by approximately 6 months. In 2 patients, cancer preceded the onset of the myopathy by 1.5 and 7 months. Of the 3 patients with CTD-associated NAM, 2 had scleroderma and 1 had Sjögren syndrome.
Antecedent events (noted within 2 weeks before weakness onset) were upper respiratory tract illness (n = 4), surgical procedure (n = 4), motor vehicle collision (n = 1), and stroke (n = 1). Coexisting diseases included single cases of psoriasis, autoimmune hepatitis, central nervous system demyelination, and 2 cases of autoimmune acetylcholine receptor myasthenia gravis (1 thymoma related).
The median CK value at presentation was 5326 U/L (to convert to microkatals per liter, multiply by 0.0167) (Table 2), and the level did not correlate with weakness severity. The serum troponin T level was elevated in 15 (94%) of 16 patients tested. The antinuclear antibody test result was positive in 14 (23%) of 62 patients tested. Only 1 patient was Jo-1 seropositive but lacked dermatologic or systemic manifestations of CTD or dermatomyositis. Human immunodeficiency virus antibodies were not detected (60% tested).
In 13 (24%) of 53 patients tested, SRP-IgG was detected. All cases were idiopathic (no risk factor) or statin associated (Table 2). The SRP-IgG–positive patients were more likely to have facial weakness (P = .02) and were younger than the SRP-IgG–negative patients (median age, 49.0 vs 64.0 years; P = .03).
In 17 (34%) of 50 patients tested, including 12 of 17 statin recipients and 5 of 24 in the idiopathic group, HMGCR-IgG was detected. One statin-recipient patient was dual positive for HMGCR-IgG and SRP-IgG. In the patients with paraneoplastic or CTD-associated NAM, HMGCR-IgG was not detected. The HMGCR-IgG–positive patients were more commonly statin recipients (P = .007) and more likely to be responsive to a single immunotherapeutic agent (P = .047).
All 63 patients had fibrillation potentials recorded by EMG, and 32 had myotonic discharges, which were more common in the statin-associated group independent of antibody status (P < .001). Follow-up EMG evaluation (11 patients) revealed improvement (reduction in the proportion of muscles with fibrillation potentials from 78% before treatment to 34% after treatment). Nine patients (14%) had electrophysiologic evidence of a coexisting chronic axonal peripheral neuropathy, without predilection for a specific subgroup.
Muscle biopsies were performed or slides reviewed in all cases at the time of Mayo Clinic evaluation. Four patients were receiving immunotherapy at the time of biopsy. No histopathologic differences were noted between the groups or among SRP-IgG–positive or HMGCR-IgG–positive patients. The predominant findings were fiber necrosis (100%) and regeneration (95%) (Figure). Fourteen patients (22%) had a mild inflammatory exudate that consisted of tiny perivascular collections of mononuclear cells in the perimysium. Invasion of nonnecrotic fibers was absent. Nonrimmed vacuoles were noted in 20 biopsy specimens. There were no congophilic deposits.
Twenty-three patients had dyspnea, and 5 required intubation. Pulmonary function tests documented a restrictive pattern consistent with neuromuscular respiratory weakness in 12 (32%) of 37 patients assessed (reduced total lung capacity, forced vital capacity, and maximal inspiratory and expiratory pressures). Abnormal oxygen diffusion capacity, corrected for hemoglobin, was noted in 6 (30%) of 20 patients tested and suggested impaired gas exchange. Chest computed tomography (completed in 44 cases) revealed reticular interstitial lung abnormalities suggestive of interstitial lung disease in 2 patients. Both patients were antinuclear antibody seropositive but Jo-1-IgG negative; 1 had N-type voltage-gated calcium channel antibody (0.12 nmol/L; reference range, 0.00-0.03 nmol/L). Neither patient had evidence of cancer. Overnight oximetry revealed that 9 (47%) of 19 patients had periodic oxygen desaturations consistent with sleep-disordered breathing.
Palpitations and chest pain were uncommon (6 patients). Bundle branch block (right, left, or fascicular) was evident on electrocardiograms of 11 (22%) of 50 patients. The QT interval was prolonged in 2 patients (no known pharmacologic cause), and new atrial fibrillation was detected in another patient. Echocardiogram abnormalities were noted in 25 (61%) of 41 patients, most commonly diastolic dysfunction (11 patients). Abnormal left ventricular wall motion was detected in 2 patients (without history of coronary artery disease). Cardiac disease or hypertension preexisted in 30% of patients with electrocardiographic abnormalities and in most patients with diastolic dysfunction.
Treatment plans were not standardized but depended on the individual’s clinical response. All patients received immunotherapy (eTable in the Supplement). The median interval between onset of weakness and initiation of therapy was 5 months (range, 1-37 months). Initial treatment, instituted within 3 months of assessment, consisted of oral or intravenous corticosteroids alone (n = 17), intravenous immune globulin (IVIG) alone (n = 4), methotrexate alone (n = 1), mycophenolate mofetil alone (n = 1), or combinations of 2 or more agents (n = 40). Combination therapy most commonly included corticosteroids and a steroid-sparing agent (n = 12), corticosteroids and IVIG (n = 6), and triple therapy with corticosteroids, IVIG, and a steroid-sparing agent (n = 17). The most common steroid-sparing medications were mycophenolate mofetil (n = 21), methotrexate (n = 7), and azathioprine (n = 4). In general, initial corticosteroid therapy included oral prednisone dosed at 1 mg/kg or intravenous methylprednisolone at 1 g weekly, whereas IVIG was dosed at 2 g/kg monthly. Adjustments in corticosteroid and IVIG dose and frequency were based on clinical response. Longitudinal treatment data were available for 32 patients followed up for 6 or more months (median, 13.5 months; mean, 26.4 months; range, 1-84 months). During the follow-up period, 18 patients (56%) needed 3 immunotherapeutic agents, 12 required 2 agents, and 2 were maintained with monotherapy (corticosteroids). Only 1 patient was able to discontinue immunotherapy because of marked improvement (close to normal).
Response to treatment was assessed at 3 months, 12 months, and at the time of last follow-up (eFigure, A, in the Supplement). At last-follow-up, 17 patients had marked improvement or had returned to baseline, 8 had moderate improvement, and 5 had mild improvement. Serum CK value decreased with treatment, paralleling the clinical improvement (eFigure, B, in the Supplement). Predictors of favorable outcome (defined as marked improvement or return to baseline at last follow-up) identified by univariate logistic regression analysis were male sex (odds ratio, 6.3) and the use of 2 or more immunosuppressants in the 3 months after presentation (odds ratio, 15.3). Antibody seropositivity (SRP-IgG or HMGCR-IgG) was not a significant predictor of outcome. Use of IVIG within the first 3 months (n = 10; nonrecipients, n = 11) was associated with a significantly better clinical response (more likely to attain marked improvement or return to baseline) at 6 months (P = .047). However, this difference was not sustained at last follow-up. The interval between onset and initiation of therapy did not seem to predict outcome, but the number of patients presenting beyond 6 months was too small to make a definite conclusion.
Follow-up data were sufficient in 29 patients to determine clinical course. The disease was monophasic (improvement sustained with immunotherapy) in 13 patients (45%). Weakness relapse and increase in serum CK level occurred in 16 patients (55%) whose immunotherapy was tapered or discontinued, requiring resumption of immunotherapy.
This retrospective study provides detailed treatment and outcome data for a large cohort of patients with NAM of different causes. In agreement with previously published reports,14,18,22,25 the weakness at presentation was subacute, relatively severe, and proximally predominant, and serum CK levels were markedly elevated. Weakness in some cases was more acute in onset (within ≤4 weeks), and in a few cases weakness evolved slowly (during 9-14 months), in agreement with a previous report26 of slowly progressive NAM occasionally mimicking muscular dystrophy. Distal weakness was common (41% of patients) and affected finger abduction, finger extension, and foot dorsiflexion most frequently. We also observed prominent neck weakness in 2 patients, a finding not previously described in NAM.
Although the predominant histopathologic finding was a necrotizing myopathy, minor inflammatory infiltrates were observed in approximately 20% of patients. These infiltrates were largely confined to perivascular sites in perimysium. No biopsy specimen revealed inflammatory cells invading nonnecrotic muscle fibers. A recent study27 characterized the composition of cellular infiltrates in HMGCR-IgG–associated NAM, demonstrating that macrophages are the predominant cells.
Half of the patients had no identifiable risk factors. The rarity of a CTD association and the association with scleroderma are consistent with previous reports.10,19 Systemic lupus erythematous has been reported in some cases,18 but we did not identify this association. Despite the rarity of a CTD association, patients with NAM should be evaluated for CTD because of management and prognostic implications. Cancer (most commonly gastrointestinal carcinoma) was found in approximately 10% of the cases, usually through malignancy screening after myopathy onset, as previously reported.11,19,28 Although some authors found no alteration in the risk of malignant tumors in patients with NAM compared with the general population, our recommendation is that patients with NAM be screened for malignant tumors by chest, abdominal, and pelvic imaging, mammography, and endoscopy.18 None of our SRP-IgG– and HMGCR-IgG–positive patients had cancer, despite reported cancer associations in the literature.23,27,29
The largest risk factor we identified for NAM was statin exposure, which was documented in one-third of our patients, 12 positive for HMGCR-IgG and 4 for SRP-IgG (1 patient was dual seropositive). As previously reported,6,9 atorvastatin and simvastatin were the most common statins, likely reflecting prescribing patterns. The myopathy persisted after discontinuation of statin therapy but improved with immunotherapy. The presumed autoimmune cause of statin-associated NAM and the pathogenicity of HMGCR-IgG remain uncertain. Indeed, almost one-third of our HMGCR-IgG–positive patients were statin naive (similar to 2 other large series7,8,22), and HMGCR-IgG has been detected in rare patients with a self-limited statin-associated myopathy.7 Independent of the presumed pathogenicity of these autoantibodies, our data suggest that NAM associated with HMGCR-IgG may have a relatively milder course. Patients in this group were less likely to require aggressive immunosuppressants, and 1 patient was able to discontinue immunotherapy.
We detected SRP-IgG mostly in the idiopathic group and in 4 patients with statin-associated NAM. This recognized serologic accompaniment of NAM22 is seldom encountered with other immune-mediated myopathies.30,31 One-fourth of the patients we tested were seropositive, a higher frequency than previously reported for NAM.18,22 Seropositive patients were younger, had more frequent facial weakness, and had fewer myotonic discharges on EMG compared with SRP-IgG–negative patients. Seropositivity for SRP-IgG did not predict clinical severity, course, or response to treatment. A report14 of a higher incidence of SRP-IgG–related NAM in the autumn prompted us to investigate possible triggers. Potential antecedent triggers were rare, with viral infections and surgical procedures each being reported in 4 patients.
We encountered frequent respiratory and cardiac abnormalities with NAM. Dyspnea and neuromuscular respiratory weakness were recorded in approximately one-third of cases. There were 2 cases of interstitial lung disease. Almost half of patients tested had findings of sleep-disordered breathing, suggesting the need for polysomnography. Troponin T levels were elevated in almost all tested patients without the short-term fluctuations anticipated in an acute coronary syndrome.32,33 Nevertheless, coronary artery disease should be excluded. A review33 of cardiac manifestations in immune-mediated myopathies (not specifically NAM) reported dysrhythmias as the most common electrocardiographic abnormality and diastolic dysfunction as the most common echocardiographic abnormality. Our data support this to be true for NAM. Cardiac conduction abnormalities not attributable to preexisting ischemic cardiac disease occurred in 22%, in line with older studies15,34 of SRP-IgG–positive patients and suggesting preferential involvement of the conduction system. Thus, ECG and echocardiography are justifiable considerations for all patients with NAM.
Although no randomized clinical trials exist to guide therapy in NAM, this entity is considered by many physicians to be refractory to conventional immunotherapy.6,7,14,15,22,35,36 Our study suggests that NAM is often refractory to corticosteroid monotherapy; indeed, almost all patients required at least 2 immunotherapeutic agents. Early treatment with IVIG increased the likelihood of strength improvement, consistent with studies26,37 indicating the possible preferential response of NAM to IVIG. Some reported long-term outcomes revealed persistent disability despite aggressive immunotherapy,7,14 and others reported favorable results.17,25,26,31,36 Our data suggest that more than half of patients receiving aggressive immunotherapy recover markedly or improve to normal; only 10% had little or no clinical improvement. However, the relapse rate is high (in 55% of cases) during medication taper or treatment discontinuation. Only one patient was able to discontinue immunotherapy, which is fewer than reported for polymyositis and dermatomyositis.19,20 Predictors of favorable clinical outcome included male sex and the use of at least 2 forms of immunotherapy early in the disease course. On the basis of our and other authors’ data and experience, it is reasonable to treat NAM with a combination of IVIG, corticosteroids, and a steroid-sparing immunosuppressant for at least 3 months, followed by long-term treatment with the steroid-sparing agent. However, treatment must be individualized.
A thorough search for differences among the 4 etiologic groups failed to reveal any difference in clinical severity, response to treatment, or outcome. Patients with statin-associated and paraneoplastic NAM were older, in agreement with previous studies,1,6 whereas patients with idiopathic NAM were more likely to have dysphagia. Myotonic discharges were significantly more common in statin-associated NAM. Therefore, the detection of myotonic discharges in a patient with subacute myopathy may be a valuable clue to the diagnosis of statin-associated NAM.38
A major strength of our study is that stringent inclusion and exclusion criteria ensured exclusion of other immune-mediated myopathies. By biopsying clinically weak muscles with electrophysiologic evidence of irritable myopathy, we avoided sampling error as a cause of lack of inflammation. Muscle magnetic resonance imaging could also be used to target an appropriate biopsy site. In only 4 patients was immunotherapy initiated before muscle biopsy. Thus, iatrogenic suppression of inflammation was avoided. A major limitation of this study is its retrospective design. Patients received similar but not consistent investigations or treatment, serologic data are limited, and functional scales were not applied to evaluate response to treatment. The small numbers also preclude multiple regression analysis for predictors of outcome. Despite these limitations, we provide a comprehensive review of clinical and laboratory data, treatment course, and exploratory data on predictors of outcome for a large cohort of patients with clinically and pathologically defined NAM. We recognize that prospective studies are needed to verify these findings.
In summary, NAM is a severe immune-mediated myopathy that requires prompt recognition because early aggressive management can lead to favorable outcomes. Statin medications are the most commonly identified risk factor, but malignant tumors should be excluded.
Accepted for Publication: April 30, 2015.
Corresponding Author: Margherita Milone, MD, PhD, Department of Neurology, Mayo Clinic, 200 First St SW, Rochester, MN 55905 (firstname.lastname@example.org).
Published Online: July 20, 2015. doi:10.1001/jamaneurol.2015.1207.
Author Contributions: Drs Kassardjian and Milone had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Kassardjian, Milone.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Kassardjian.
Critical revision of the manuscript for important intellectual content: Kassardjian, Lennon, Alfugham, Mahler, Milone.
Statistical analysis: Kassardjian, Mahler.
Administrative, technical, or material support: Lennon, Alfugham.
Study supervision: Milone.
Conflict of Interest Disclosures: Dr Mahler reported working for Inova Diagnostics but holds no shares of the company. No other disclosures were reported.