Outcomes of patients with borderline resectable and resectable pancreatic adenocarcinoma treated with neoadjuvant three-week course chemoradiotherapy using capecitabine-based versus gemcitabine-based concurrent chemotherapy

BACKGROUND: Neoadjuvant chemoradiotherapy can provide downstaging and improve margin negativity for borderline resectable and resectable pancreatic adenocarcinoma [(B)RPC]. Little is known about the relative efficacy of capecitabine (CAPE)-based vs. gemcitabine (GEM)-based 3-week chemoradiation (3WCRT) with 36 Gy in 15 fractions. This study aimed to compare the odds of achieving surgical resection, time to progression (TTP), and overall survival (OS) of patients treated with 3WCRT with concurrent CAPE versus GEM. METHODS: A retrospective cohort study was conducted, examining medical records from a single center for patients with (B)RPC treated with 3WCRT between 1/2009–12/2020. Odd ratios (OR) of achieving surgical resection were estimated using logistic regression for univariable and multivariable analyses. Median TTP (mTTP) and median OS (mOS) were estimated using the Kaplan-Meier method. Cox proportional hazards analysis was conducted to estimate hazard ratios (HR) of progression and survival in univariable and multivariable analyses. RESULTS: Thirty-one patients were included in the analysis. Twenty-two (71%) patients were treated with CAPE, while 9 (29%) were treated with GEM. All patients in the GEM group were borderline resectable, vs. 18 (82%) patients in the CAPE group, P=0.30. Nineteen (86%) patients in the CAPE group were treated with neoadjuvant FOLFIRINOX, vs. 4 (44%) patients in the GEM group, P=0.03. The CAPE group had higher odds of achieving surgical resection [OR =9.33; 95% confidence interval (CI): 1.50–58.20]. Adjusting for covariates, the odds of achieving surgical resection were still statistically higher in the CAPE group vs. the GEM group (OR =25.34; 95% CI: 1.14–563.72). The CAPE group had superior mTTP compared to the GEM group (15.4 months, 95% CI: 4.9–71.1 vs. 4.0 months, 95% CI: 0.4–14.5; P=0.01), corresponding to a hazard ratio of 0.33 (95% CI: 0.14–0.81). Adjusting for covariates this effect persisted; the adjusted hazard ratio (AHR) for progression was 0.24 (95% CI: 0.08–0.77). Cox proportional hazards analysis also demonstrated that the CAPE group had superior OS compared to the GEM group in unadjusted (HR =0.13; 95% CI: 0.04–0.40) and adjusted models (HR =0.13, 95% CI: 0.03–0.52). CONCLUSIONS: For neoadjuvant 3WCRT, this hypothesis-generating study suggests concurrent CAPE may be a more effective radiosensitizer than GEM for patients with (B)RPC.