Autoimmune Encephalitis Misdiagnosis in Adults; A Multicenter Observational Study of Outpatient Subspecialty Clinics

Abstract

Objective To determine the diseases misdiagnosed as AE and potential reasons for misdiagnosis. Background Misdiagnosis of autoimmune encephalitis (AE) may harm patients. Design/Methods Patients with AE misdiagnosis were identified (1/1/2014-12/31/2020) from outpatient AE subspecialty clinics including: Mayo Clinic (n = 44); Oxford (n = 18); UT Southwestern (n = 18); UCSF (n = 17); Washington University (n = 6); University of Utah (n = 4). Inclusion criteria were adults (=18 years) with: 1) A prior diagnosis of AE; and 2) An alternative diagnosis made at a participating center. We collected data on clinical features, investigations, fulfillment of possible AE criteria, alternative diagnoses, and potential contributors to misdiagnosis. Results We identified 107 patients misdiagnosed with AE. Thirty (28%) fulfilled diagnostic criteria for “possible AE”. Median onset age was 48 years (inter-quartile range, 35.5-60.5) and 65 (61%) were female. Correct diagnoses included: functional neurologic disorder, 27 (25%); neurodegenerative disease, 22 (21%); primary psychiatric disease, 19 (18%); cognitive deficits from comorbidities, 11 (10%); cerebral neoplasm, 10 (9%); and other, 18 (17%). Onset was insidious (>3 months) in 51 (48%). MRI brain was suggestive of encephalitis in 19/104 (18%) and CSF pleocytosis occurred in 16/84 (19%). Thyroid-peroxidase antibodies were elevated in 24/62 (39%). Positive neural autoantibodies were more frequent in serum (48/105[46%]) than CSF (7/91[8%]; p<0.001) and serum antibodies included: GAD65, 14; voltage-gated-potassium-channel-complex [LGI1, CASPR2 negative], 10; NMDA-receptor by cell-based assay only, 10 (6 negative in CSF); and other, 18. Immunotherapy adverse effects were observed in 17/84 (20%). Potential contributors to misdiagnosis included: over-interpretation of a non-specific positive serum antibody, 53 (50%); misinterpretation of functional, psychiatric, or non-specific cognitive dysfunction as encephalopathy, 41 (38%). Conclusions Red flags suggesting alternative diagnoses to AE include lack of fulfillment of “possible autoimmune encephalitis” criteria, positive non-specific serum antibody, and insidious onset. Avoiding AE misdiagnosis will prevent morbidity from unnecessary immunotherapies and delayed treatment of the correct diagnosis.