Background: The combination of capecitabine (CAP) with temozolomide (TEM) chemotherapy in advanced pancreatic neuroendocrine tumors (PanNET) relies on limited evidence. We compared TEM-CAP to TEM alone in patients with advanced PanNET. Methods: Consecutive patients with advanced PanNET treated with TEM or TEM-CAP between 2004 and 2017 in three expert centers were included. Progression-free survival (PFS), tolerance, tumor response, and overall survival were compared between the two groups. Propensity-based analyses were performed to reduce confounding bias due to the nonrandomized setting. Results: TEM and TEM-CAP were administered to 38 patients and 100 patients, respectively, with a median age of 58 years. The patients in the TEM group more often had hormonal syndromes (p = 0.03), a longer median delay to diagnosis (p = 0.001), and a higher number of pretreatment lines (p < 0.001). The performance status was 0 in 58% versus 65% of the patients, and tumor’s median Ki-67 index was 8% versus 11%, respectively. Tolerance was similar, except that there were more cases of asthenia in the TEM group (p = 0.017) and more cases of hand-foot syndrome in the TEM-CAP group (p = 0.025). The objective response rate was 34% versus 51% (p = 0.088). The raw median PFS was similar with TEM and with TEM-CAP (21.4 vs. 19.8 months, p = 0.84). Although CAP tended to decrease the risk of progression in Cox multivariate analysis (HR 0.65, p = 0.12), it had no effect after adjustment for the propensity score (HR 1.06, p = 0.80). Conclusions: TEM-CAP might not prolong PFS but might achieve a higher response rate than TEM alone. Hence, TEM-CAP might be preferred when tumor shrinkage is the main therapeutic objective. Otherwise, TEM might be adequate for patients with an impaired performance status or in case of extrahepatic metastases.

Introduction: Although chemotherapy combining 5-fluorouracil (5FU)-dacarbazine (DTIC) or temozolomide (TEM)-capecitabine (CAP) is extensively used in patients with neuroendocrine tumors (NET), they were never compared. We compared their tolerance and efficacy in advanced NET. Methods: We evaluated the records of consecutive patients with pancreatic or small-intestine advanced NET who received 5FU-DTIC or TEM-CAP between July 2004 and December 2017 in 5 French centers. Tolerance, tumor response and progression-free survival (PFS) were compared. Factors associated with PFS were analyzed using Cox multivariate regression model. To reduce the confounding bias of the nonrandomized design, PFS was compared using propensity score analyses. Results: Ninety-four (5FU-DTIC) patients and 153 (TEM-CAP) patients were included. Pancreatic NET represented 82.3% of cases and 17.1, 61.8 and 10.9% of patients had G1, G2 or G3 NET respectively. Progression at baseline was reported in 92.7% of patients with available data. Grades 3–4 adverse events occurred in 24.7 and 8.5% of TEM-CAP and 5FU-DTIC patients respectively (p = 0.002). The overall response rate was 38.3 and 39.2% respectively (p = 0.596). Median PFS on raw analysis was similar to 5FU-DTIC and TEM-CAP (13.9 vs. 18.3 months, respectively p = 0.86). TEM-CAP was associated with an increased risk of progression on the raw multivariate analysis (hazard ratio [HR] 1.90, 95% CI [1.32–2.73], p = 0.001) and when adjusted on propensity score (HR 1.65, 95% CI [1.18–2.31], p = 0.004). Conclusion: PFS may be longer with 5FU-DTIC than TEM-CAP in patients with advanced NET. Although patients often prefer oral chemotherapy, 5FU-DTIC is a relevant alternative. A randomized comparison is needed to confirm these results.

The existence of a notoriety bias has an impact on measures of disproportionality. The detection of pharmacovigilance signals might be delayed. It is advisable to list all drugs being taken when an adverse drug reaction occurs, and not only those known to be associated with the observed reaction.