Leichman, Lawrence P.

4175 Background: Patients (pts) with pancreatic adenocarcinoma (PDAC) who are candidates for surgery are anatomically segregated into two groups: those with resectable tumors (R-PDAC), and those with borderline resectable tumors (BR-PDAC). For each group, complete surgical resection with no visible or microscopic disease left behind (R0 resection) is generally accepted as the path to cure. Neoadjuvant therapy (NAT) has been used to improve the R0 resection rate. However, the optimal role and type of NAT that will improve R0 resection rate and overall survival (OS) remains controversial. The Academic Gastrointestinal Cancer Consortium (AGICC) evaluated a perioperative strategy with combination chemotherapy and preoperative stereotactic radiation therapy (SBRT) to improve the R0 resection rate. Methods: This prospective, multicenter open label, single arm phase 2 clinical trial conducted among 5 member AGICC institutions enrolled pts with R-PDAC and BR-PDAC. Treatment included 3 cycles of gemcitabine (1000mg/m2) and nab-paclitaxel (125 mg/m2) (GEM/NAB) on days 1, 8 and 15 of a 28-day cycle prior to SBRT (33Gy: 6.6Gy daily x 5) followed by surgery and 3 cycles of adjuvant GEM/NAB. Accrual goal was 20 R-PDAC and 30 BR-PDAC pts selected by central radiology review. Primary objective was to estimate the R0 resection rate for each cohort. Secondary objectives included safety and tolerability of neoadjuvant treatment and OS. Results: Eighty-six pts signed consent and after radiologic screening, 20 R-PDAC patients and 29 BR-PDAC patients were found eligible to begin treatment. One R-PDAC and two BR-PDAC pts did not initiate therapy. 14 (70%) and 19 (66%) R-PDAC and BR-PDAC pts, respectively, completed all neoadjuvant chemotherapy.14 R-PDAC (70%) and 17 BR-PDAC (59%) pts completed SBRT. 11 R-PDAC pts (55%) and 11 (38%) BR-PDAC, underwent surgery. 8 R-PDAC pts (40%) had an R0 resection. One R-PDAC pt had a complete pathologic response (pCR). 8 BR-PDAC pts (28%) had an R0 resection. 19% (3/16) pts undergoing an R0 resection are alive (47-73+ mos). 16 pts (32%) had at least one grade ≥ 3 non-hematologic adverse event (AE); 25 patients (51%) had at least one grade ≥ 3 hematologic AE. 8 pts, 4 R-PDAC and 4 BR-PDAC were hospitalized prior to surgery. One pt (2%) died during NAT. There were no deaths due to surgery. Conclusions: R0 resection rates and survival in R-PDAC and BR-PDAC were low despite a perioperative approach of systemic therapy and SBRT. The lack of survival benefit of GEM/NAB as an adjuvant regimen, as well as SBRT pre-operatively (both shown in studies subsequent to this one) may explain the disappointing results. While an R0 resection for localized PDAC is recognized as essential, our results demonstrate that an R0 resection is not sufficient to cure the majority of PDAC pts who receive NAT with GEM/NAB + SBRT. Future NAT therapy trials should aim to markedly improve the pCR rates associated with R0 resection. Clinical trial information: NCT02723331 .

4500 Background: Trastuzumab is a monoclonal antibody against human epidermal growth factor receptor 2 (HER2). The primary objective of RTOG 1010 was to determine if trastuzumab increases disease-free survival (DFS) when combined with trimodality treatment for patients with HER2 overexpressing esophageal adenocarcinoma. Methods: This open label, randomized phase III trial included patients with newly diagnosed stage T1N1-2, T2-3N0-2 adenocarcinoma of the esophagus involving the mid, distal, or esophagogastric junction and up to 5cm of the stomach. All patients received chemotherapy (C) of paclitaxel, 50mg/m2 and carboplatin AUC = 2, weekly for 6 weeks, with radiation (XRT: 3D-CRT or IMRT, 50.4 Gy in 28 fractions) followed by surgery. Patients were randomized 1:1 to receive weekly trastuzumab 4mg/kg week 1 then 2mg/kg/weekly x 5 during CXRT then 6 mg/kg for 1 dose prior to surgery and 6mg/kg every 3 weeks for 13 treatments after surgery. HER2 status was determined by IHC and gene amplification by FISH. With a 2-sided alpha of 0.05, 162 DFS events provide 90% power to detect a signal for an increase in median DFS from 15 to 25 months. DFS and overall survival (OS) were estimated by the Kaplan-Meier method. and arms were compared using the log rank test. The Cox proportional hazards model was used to analyze treatment effect. Results: 571 patients were entered for assessment of HER2 expression, 203 HER2+ patients randomized. The median follow-up for alive patients is 5.0 years. The estimated 2, 3, and 4-year DFS (95% CI) for the CXRT +trastuzumab arm were 41.8% (31.8%, 51.7%), 34.3% (24.7%, 43.9%), and 33.1% (23.6%, 42.7%), respectively, and for the CXRT arm were 40.0% (30.0%, 49.9%), 33.4% (23.8%, 43.0%), and 30.1% (20.7%, 39.4%), respectively; log-rank p = 0.85. The median DFS time is 19.6 months (13.5-26.2) for the CXRT +trastuzumab arm compared to 14.2 months (10.5-23.0) for the CXRT arm. The hazard ratio (95% CI) comparing the DFS of CXRT+trastuzumab arm to the CXRT arm was 0.97 (0.69, 1.36). The median OS time was 38.5 months (26.2-70.4) for the CXRT+trastuzumab arm compared to 38.9 months (29.0-64.5) for the CXRT arm, hazard ratio (95% CI): 1.01 (0.69, 1.47). There was no statistically significant increase in treatment-related toxicities with the addition of trastuzumab including no increase in cardiac events. Conclusions: The addition of trastuzumab to trimodality treatment did not improve DFS for patients with HER2 overexpressing esophageal adenocarcinoma. Supported by NCI grants U10CA180868, UG1CA189867, U10CA180822 and Genentech. Clinical trial information: NCT01196390 .