NCMP-26. PATIENTS WITH ANTI-HU/ANNA1 PARANEOPLASTIC NEUROLOGICAL DISORDERS DEVELOP PROGRESSIVE DISABILITY BUT HAVE INCREASED SURVIVAL

Abstract ANNA-1/Hu-IgG paraneoplastic neurological disorders (PND) are rare and factors associated with disability and survival outcomes are unknown. We aimed to 1) examine disability and survival in ANNA1/Hu-IgG PND and 2) compare survival to a control group of small cell lung cancer (SCLC) without PND. METHODS This is a retrospective cohort study of ANNA1/Hu-IgG+ with PND and a control cohort of SCLC without PND (n=1513). Cox proportional hazard models examined clinical predictors of disability and survival. Kaplan-Meier curves examined time to mortality and mRS>2. Cox proportional hazard models compared survival between ANNA1/Hu-IgG PND and SCLC without PND, adjusting for age, sex, cancer stage, and brain metastases. RESULTS Forty-five patients with ANNA/Hu1-IgG PND were included (73% female, mean age=63 years [SD=13]). Phenotypes included; neuropathy (16, 36%), ataxia (11, 24%), encephalitis (10, 22%), myasthenic syndromes (3, 7%), myelopathy (2, 4%), enteric neuropathy (2, 4%), myoclonus (1, 2%). Cancers were identified in 78%; SCLC (26, 74%), non-SCLC (4, 11%), other (5, 14%). Immunosuppressive therapy for PND was given in 53%. At 5-year follow-up, 74% had mRS>2. Limbic encephalitis had greater hazard of mRS>2 at last follow-up (HR=7.76, [95% CI=2.20-27.39], p=0.001). At 5-year follow-up, 58% had died. Survival was not influenced by phenotype nor immunosuppressive treatment. Patients with SCLC and ANNA1/Hu-IgG (n=26) were compared to SCLC without PND (n=1513) and had 54% lower hazard of death (adjusted HR =0.46, [95% CI = 0.28–0.76], p=0.003). In ANNA-1/Hu-IgG PND, most patients experienced disability and greater disability was associated with limbic encephalitis, suggesting a potential modifiable factor. However, patients with SCLC and ANNA1/Hu-IgG PND survived longer than patients with SCLC without ANNA1/Hu-IgG PND, suggesting a possible survival advantage due to earlier cancer detection (lead time bias) and/or increased antitumor immunity.