Abstract
PURPOSE: Favourable small cell lung carcinoma (SCLC) survival outcomes have been reported in patients with paraneoplastic neurological disorders (PNDs) associated with neuronal antibodies (Neur-Abs), but the presence of a PND might have expedited diagnosis. Our aim was to establish whether neuronal antibodies, independent of clinical neurological features, correlate with SCLC survival. EXPERIMENTAL DESIGN: 262 consecutive SCLC patients were examined: of these, 24 with neurological disease were excluded from this study. The remaining 238 were tested for a broad array of Neur-Abs at the time of cancer diagnosis; survival time was established from follow-up clinical data. RESULTS: Median survival of the non-PND cohort (n = 238) was 9.5 months. 103 patients (43%) had one or more antigen-defined Neur-Abs. We found significantly longer median survival in 23 patients (10%) with HuD/anti-neuronal nuclear antibody type 1 (ANNA-1, 13.0 months P = 0.037), but not with any of the other antigen-defined antibodies, including the PND-related SOX2 (n = 56, 24%). An additional 28 patients (12%) had uncharacterised anti-neuronal nuclear antibodies (ANNA-U); their median survival time was longer still (15.0 months, P = 0.0048), contrasting with the survival time in patients with non-neuronal anti-nuclear antibodies (detected using HEp-2 cells, n = 23 (10%), 9.25 months). In multivariate analyses, both ANNA-1 and ANNA-U independently reduced the mortality hazard by a ratio of 0.532 (P = 0.01) and 0.430 (P<0.001) respectively. CONCLUSIONS: ANNAs, including the newly described ANNA-U, may be key components of the SCLC immunome and have a potential role in predicting SCLC survival; screening for them could add prognostic value that is similar in magnitude to that of limited staging at diagnosis.