The Role of Chemotherapy in Well-Differentiated Gastroenteropancreatic Neuroendocrine Tumors.

Even though the neuroendocrine tumor (NET) field has entered the era of 'targeted therapy', the role of cytotoxic chemotherapy continues to be debated. High response rates, ranging from 30 to 70% depending on the line of therapy, are consistently observed in the treatment of pancreatic NETs, with lesser evidence of activity in other foregut tumors. Activity in midgut carcinoid tumors appears to be negligible. Unfortunately, placebo-controlled randomized controlled trials using modern response criteria are lacking: the bulk of the literature consists of small phase II trials and retrospective series. There are also no completed trials comparing modern chemotherapy regimens, and therefore little data exist to favor the use of streptozocin- versus temozolomide- versus oxaliplatin-based therapies. Due to the absence of high-level evidence, it is difficult to generate data-based guidelines on the appropriate sequencing of cytotoxic drugs versus targeted agents. Although conventional wisdom holds that targeted agents such as everolimus or sunitinib are more tolerable than cytotoxic drugs, there is no evidence to support this perception. As a general principle, chemotherapy may be more appropriate as early-line therapy in patients with bulky and/or symptomatic and/or rapidly progressive tumors, particularly of pancreatic origin. In patients with low-volume disease or slow-growing tumors, noncytotoxic drugs may be preferable in early lines of therapy, reserving chemotherapy for the salvage setting. Validation of predictive factors is imperative in order to appropriately match patients with optimal treatment. Methyl-guanine-methyl-transferase (MGMT) deficiency is likely to be a positive predictive factor for alkylating agents, but needs to be evaluated prospectively. It is also unclear whether immunostaining for MGMT expression, which can be somewhat subjective, is superior to PCR-based techniques, which assess MGMT methylation status. Other basic predictive factors, such as ki-67 index, need to be evaluated in prospective clinical trials. It will also be important to evaluate whether newly discovered mutations in pancreatic NETs, such as DAXX/ATRX, are associated with response to chemotherapy.