NET-02: a randomised, non-comparative, phase II trial of nal-IRI/5-FU or docetaxel as second-line therapy in patients with progressive poorly differentiated extra-pulmonary neuroendocrine carcinomaResearch in context

Summary: Background: The prognosis for patients with poorly-differentiated extra-pulmonary neuroendocrine carcinoma (PD-EP-NEC) is poor. A recognised first-line (1L) treatment for advanced disease is etoposide/platinum-based chemotherapy with no standard second-line (2L) treatment. Methods: Patients with histologically-confirmed PD-EP-NEC (Ki-67 > 20%; Grade 3) received IV liposomal irinotecan (nal-IRI) (70 mg/m2 free base)/5-FU (2400 mg/m2)/folinic acid, Q14 days (ARM A), or IV docetaxel (75 mg/m2), Q21 days (ARM B), as 2L therapy. Primary endpoint was 6-month progression-free survival (PFS) rate (80% power to demonstrate one-sided 95% lower confidence interval excluded 15% (target level of efficacy: 30%)). Secondary endpoints: objective response rate (ORR), median PFS, overall survival (OS), toxicity and patient-reported quality-of-life (QoL) (ClinicalTrials.gov: NCT03837977). Findings: Of 58 patients (29 each arm); 57% male, 90% ECOG PS 0/1, 10% PS 2, 89.7% Ki-67 ≥ 55%, primary site: 70.7%-gastrointestinal, 18.9%-other, 10.3%-unknown, 91.4%/6.9%/1.7% were resistant/sensitive/intolerant to 1L platinum-based treatment, respectively. The primary end-point of 6-month PFS rate was met by ARM A: 29.6% (lower 95% Confidence-Limit (CL) 15.7), but not by ARM B: 13.8% (lower 95%CL:4.9). ORR, median PFS and OS were 11.1% (95%CI:2.4–29.2) and 10.3% (95%CI:2.2–27.4%); 3 months (95%CI:2–6) and 2 months (95%CI:2-2); and 6 months (95%CI:3–10) and 6 months (95%CI:3–9) in ARMS A and B, respectively. Adverse events ≥ grade 3 occurred in 51.7% and 55.2% (1 and 6 discontinuations due to toxicity in ARMS A and B), respectively. QoL was maintained in ARM A, but not ARM B. Interpretation: nal-IRI/5-FU/folinic acid, but not docetaxel, met the primary endpoint, with manageable toxicity and maintained QoL, with no difference in OS. ORR and median PFS were similar in both arms. This study provides prospective efficacy, toxicity and QoL data in the 2L setting in a disease group of unmet need, and represents some of the strongest evidence available to recommend systemic treatment to these patients. Funding: Servier.