CMET-44. PREDICTORS OF SURVIVAL IN NEURO-METASTATIC MERKEL CELL CARCINOMA

BACKGROUND: Merkel cell carcinoma (MCC) is a rare cutaneous malignancy of neuroendocrine origin, with about 30 cases of brain metastasis (BM) reported in the literature. Historically, the treatment of neuro-metastatic MCC has largely included chemotherapy (CT) and radiotherapy (RT). The aim was to investigate predictors of overall survival (OS) in neuro-metastatic MCC. METHODS: In this retrospective study, we surveyed institutional databases and additionally conducted a systematic review of the literature to identify cases reporting on management of distant MCC BM. A pooled analysis was performed on the institutional and literature cases to assess predictors of OS. Survival analysis was done on R (ver 3.4.0) using a Log Rank statistic and cox proportional hazard ratio. RESULTS: Forty cases were included for analysis, describing operative (14) and non-operative (26) management. Median time from initial MCC diagnosis to CNS involvement was 17.0-mos (IQR 10.5- 26.5), and most patients had a single BM (62.5%). Management of intracranial disease included RT (84.2%), systemic therapy (59.5%) and surgical resection (35%). Operative management was associated with a lower intracranial burden of disease (BoD, single BM: op 92.9% vs. non-op 46.2%, p=0.004), but similar systemic BoD. Median OS was longer in patients treated with neurosurgery (73-mos, 95%CI:31–115 vs. 25-mos, 95%CI:17–44, p ‹ 0.001). Both neurosurgery (HR 0.18, 95%CI:0.06–0.54, p=0.002) and having a single BM (multiple BM or leptomeningeal disease: HR 2.51, 95%CI:1.12–5.6, p=0.03) conferred an OS benefit on risk-unadjusted analysis. On multivariable analysis, only neurosurgical resection was an independent predictor of OS (HR 0.16, 95%CI:0.04–0.59, p=0.006), controlling for age, BoD and RT. Systemic therapy and RT were not associated with OS. CONCLUSIONS: Resection of MCC BM may confer a survival benefit relative to non-operative management given appropriate patient selection. Prospective investigation of multimodal management of neurometastatic MCC is warranted, especially given the promise of new immunotherapy agents in treating MCC.