Carboplatin (CB) combined with oral or intravenous (IV) etoposide (ET) for advanced extra-pulmonary (EP) poorly differentiated (PD) neuroendocrine carcinoma (NEC); real-world findings from two European Neuroendocrine Tumour Society Centres of Excellence

Background Carboplatin-Etoposide is a 1st-line (1L) option for patients (pts) with advanced EP-PD-NEC. Schedules with oral or IV ET are used in clinical practice. Data from randomised trials are lacking.MethodsRecords of pts diagnosed with advanced EP-PD-NEC and treated with CB/oral-ET and CB/IV-ET were reviewed retrospectively (09/96-02/17). First-line survival/activity/toxicity data are reported. Results One-hundred-thirteen pts were identified: median (med) follow-up was 11.5 months (m); med age was 65.8 years (range 24-88); male=64%; ECOG performance status 0-1=81%; no/mild comorbidities=81%; gastro-entero-pancreatic origin=54%; stage IV = 90% (53% liver metastases). Median Ki-67=70% (95%CI 60-80%), Ki-67>55%=59%. A total of 123 courses of CB-ET were administered: 106 (86%) 1L, 16 (13%) 2nd-line (2L) and 1 (1%) 3rd-line; med cycles/line=4; oral-ET 45%, IV-ET 55%. Median CB-ET dose-intensity (available for 82 courses): 96% (1L), 90% (2L). Median 1L-progression free survival (PFS) was 5.9m (95%CI 5.0-7.1): oral-ET 5.6m, IV-ET 6.2m, hazard ratio (HR)=0.76 (95%CI 0.51-1.14). Median 1L-overall survival (OS) was 11.5m (95%CI 8.9-13.6): oral-ET 8.9m, IV-ET 12.1m, HR = 0.68 (95%CI 0.45-1.03), p = 0.07. Radiological response (assessed for 95 pts), 1L-disease control rate was 75.8% (95%CI 67.1-84.6): oral-ET 69.8% (95%CI 55.5-84.1), IV-ET 80.8% (95%CI 69.7-91.8). Liver metastases were the only independent factor related to worse 1L-PFS on multivariable analysis, HR = 1.71 (95%CI 1.11-2.63). Commonest 1L-grade 3-4 adverse event (AE) was myelosuppression (47.2%); no significant differences between oral-ET and IV-ET AEs, except for venous thromboembolism; oral-ET 12.5%, IV-ET 1.7% (p = 0.04). Conclusions This is one of the largest series of pts with advanced EP-PD-NEC treated with CB-ET in the current literature. Oral-ET and IV-ET schedules are associated with comparable 1L-PFS/activity/toxicity data. There is a trend towards better 1L-OS for IV-ET schedules; this, however, may be driven by differences in patient selection between the two subgroups.