A Phase II Trial of Ziv-Aflibercept in Patients with Advanced Pancreatic Neuroendocrine Tumors

Objectives Therapies for patients with advanced well-differentiated pancreatic neuroendocrine tumors (pNETs) are insufficient, with patients succumbing to disease despite recent treatment advances. Ziv-aflibercept is a fusion protein of portions of the vascular endothelial growth factor (VEGF) receptors 1 and 2, fused to the Fc portion of immunoglobulin G1, forming a VEGF trap. Perfusion computed tomography (CT) has previously defined hyperperfused neuroendocrine tumors, potentially predicting antiangiogenic benefit. Methods We performed a single-arm open-label study, using the Simon optimal 2-stage design, of 6 mg/kg ziv-aflibercept in patients with advanced pNETs. The primary end point was objective radiographic response, with a hierarchically tested co-primary end point of response prediction by baseline hyperperfusion, defined as blood volume greater than 5.25 mL/100 g and permeability surface area product greater than 25 mL/min per 100 g. Results Between July 3, 2014, and September 28, 2016, 21 patients were treated. Two patients (9.5%; 95% confidence interval, 1.1%-30.4%) demonstrated objective response, satisfying criteria to open the second stage, but the study was terminated for accrual. Perfusion CT could not be confirmed to predict radiographic response. Toxicities include 1 grade 5 gastrointestinal hemorrhage and 5 incidents of proteinuria requiring treatment discontinuation. Conclusions Responses with ziv-aflibercept were consistent with other VEGF inhibitors in pNET, but perfusion CT could not be confirmed to predict outcome.