Algorithm to improve the diagnosis of paraneoplastic neurological syndromes associated with SOX1 antibodies.

SOX1 antibodies (SOX1-abs) are associated with paraneoplastic neurological syndromes (PNS) and small cell lung cancer (SCLC). In many clinical laboratories SOX1-abs are determined by commercial line blots without confirmation by cell-based assay (CBA) with HEK293 cells expressing SOX1. However, the diagnostic yield of commercial line blots is low and the accessibility to the CBA, that is not commercially available, limited. Here, we evaluated if the addition of the band intensity data of the line blot and the immunoreactivity in a tissue-based assay (TBA) improve the diagnostic performance of the line blot. We examined serum of 34 consecutive patients with adequate clinical information that tested positive for SOX1-abs in a commercial line blot. Samples were also assessed by TBA and CBA. SOX1-abs were confirmed by CBA in 17 (50%) patients, all (100%) had lung cancer (SCLC in 16) and 15/17 (88%) had a PNS. In the remaining 17 patients the CBA was negative and none had PNS associated with lung cancer. TBA was assessable in 30/34 patients and SOX1-abs reactivity was detected in 15/17 (88%) with positive and in 0/13 (0%) with negative CBA. Only 2 (13%) of the 15 TBA-negative patients were CBA-positive. The frequency of TBA-negative but CBA-positive increased from 10% (1/10) when the band intensity of the line blot was weak to 20% (1/5) in patients with a moderate or strong intensity band. Confirmation by CBA should be mandatory for samples (56% in this series) not assessable (4/34; 12%) or negative in the TBA (15/34; 44%).