Multi-agent neoadjuvant chemotherapy improves survival in early-stage pancreatic cancer: A National Cancer Database analysis.

PURPOSE:To compare overall survival (OS) in patients who underwent surgery for early-stage pancreatic adenocarcinoma (rPca) based on sequence (NAT, neoadjuvant therapy and/or AT, adjuvant therapy) and type (SA, single-agent or MA, multi-agent) of chemotherapy received. METHODS:Using the National Cancer Database, patients with clinical stage I/II rPca diagnosed between 2010 and 2014 were identified and five comparison matches (1: NAT vs. upfront resection (UR); 2: multi-agent neoadjuvant (MA NAT) vs. single-agent adjuvant therapy (SA AT), single-agent neoadjuvant therapy (SA NAT), multi-agent adjuvant therapy (MA AT); 3: MA NAT vs. MA AT; 4: NAT + AT vs NAT; 5: NAT + AT vs AT) were constructed using minimum distance matching strategy. Median OS (mOS) was analysed using Kaplan-Meier method, log-rank test and Cox proportional hazard model. RESULTS:A total of 18,470 patients with stage I/II rPca were eligible for analysis. NAT showed a 5 month (mo.) improved OS compared with UR (3271 patients/group, 28.1 vs 23.2 mo. P < 0.0001 hazard ratio [HR]: 0.79). MA-NAT was shown to be superior to other chemotherapy approaches SA AT, SA NAT, and MA AT (1349 patients/group: 30 vs. 25.9 mo., P = 0.0001 [HR: 0.82]). MA NAT showed a survival advantage over MA-AT (1349 patients/group, 30 vs 26.1 mo., P = 0.0008 [HR: 0.86]). The combination of NAT and AT showed a better outcome when compared with NAT alone (1128 patients/group, 31.6 vs 27.4 mo., P = 0.0011 [HR: 0.81]) or AT alone (1128 patients/group, 31.6 vs. 25.2 mo., P < 0.0001 [HR: 0.76]). CONCLUSIONS:In patients with stage I/II rPca, MA NAT showed improved mOS compared to UR and all other chemotherapy sequences except both NAT plus AT. These findings support the use of MA NAT in stage I/II rPca patients and warrant prospective trials evaluating MA NAT and post-resection maintenance therapies.