Objective: To describe the clinical associations of SOX1 antibodies (SOX1-Abs), determine the accuracy of various detection techniques, and propose laboratory criteria to identify definite paraneoplastic neurological syndromes (PNS) associated with SOX1-Abs. Methods: Single-center, retrospective study of patients referred to the French Reference Center between 2009 and 2019 for confirmation of SOX1-Ab positivity, without concurrent neural antibodies. Patients were classified according to the updated diagnostic PNS criteria; biological samples were systematically retested with three distinct techniques (line blot, cell-based assay, indirect immunofluorescence). Results: Among 77 patients with isolated SOX1-Ab positivity, 23 (29.9%) fulfilled the criteria for definite PNS; all of them had lung cancer (mostly small-cell) and presented mainly with Lambert-Eaton myasthenic syndrome (10/23) and rapidly progressive cerebellar ataxia (6/23). SOX1-Ab positivity varied depending on the laboratory methods which were used, and a single technique was not sufficient to draw conclusions about the PNS diagnosis. The combination of an antigen-specific test (line blot and/or cell-based assay) and immunofluorescence showed the highest accuracy (81.5%, 95% CI 70.0–90.1) in identifying definite PNS. Moreover, when the PNS-Care score was recalculated assigning three points at the laboratory-level only to patients with positive “antigenic-specific test + immunofluorescence” and 0 points to the remaining cases, a higher certainty for definite and non-PNS was achieved (from 41/77, 53.2%, to 60/77, 77.9%; p < 0.001). Conclusion: SOX1-Abs should be considered high-risk antibodies only when detected with a positive antigenic-specific test and immunofluorescence. Other laboratory results and clinical associations different from Lambert-Eaton myasthenic syndrome and rapidly progressive cerebellar ataxia should be carefully reassessed to rule out false positivity and alternative diagnoses.
