Adjuvant therapy in the treatment of resected nonmetastatic gallbladder cancer of stage II-IV: A generalized propensity score analysis.

Abstract

Listen

Translate article

471 Background: The benefit of adjuvant therapy for gallbladder cancer (GBC) remains controversial due to the limited evidence. The current study was aimed to explore the efficacy of adjuvant therapy in patients with resected non-metastatic GBC and to establish a prognosis model to predict the survival benefit of GCB patients with different adjuvant therapies. Methods: Patients with resected non-metastatic GBC of stage II-IV were selected from the Surveillance, Epidemiology, and End Results database and divided into non-radiotherapy and chemotherapy (NCRT) group, chemotherapy (CT) group, and chemoradiotherapy (CRT) group. Generalized propensity score (GPS) and inverse probability of treatment weighting (IPTW) was used to reduce the imbalances between groups. Nomogram was constructed based on Cox proportional hazard model, and the model was validated for discrimination and calibration. Results: Among the 2689 enrolled patients, 1193 (44.4%) patients were classified as stage II, 1371 (51.0%) as stage III, and 125 (4.6%) as stage IV according to the 8th American Joint Commotion Cancer staging manual. Patients in NCRT, CT, and CRT groups were 1703, 444, and 542 respectively. After the IPTW, absolute standardized differences of baseline characteristics were less than 0.1 both in patients with stage II tumors and with stage III-IV tumors. In patients with GBC of stage II, no significant difference in OS was observed between the three treatment groups (P > 0.05). In patients with GBC of stage III-IV, CT group has a superior OS compared with the NCRT group (P < 0.001), and CRT group has a superior OS compared with the CT(P < 0.001) and NCRT (P < 0.001) group. Sensitivity analysis showed consistent results. A nomogram was constructed for patients with stage III-IV tumors to predict the survival benefit of adjuvant therapies. The C-index was 0.673 (95% CI: 0.654-0.692) in the validation using the training set (diagnosed at 2004-2012) and was 0.707 (95% CI: 0.677-0·739) in the internal validation using the validation set (diagnosed at 2013-2015). The calibration curves indicated that the predicted probability closely corresponded to the actual observation OS. Conclusions: Patients with GBC of stage II could not benefit from adjuvant therapy. Patients with GBC of stage III-IV could benefit from chemotherapy and chemoradiotherapy while chemoradiotherapy provide a superior OS. A nomogram was built to predict the survival benefit of different adjuvant therapies in patients with GBC of stage III-IV.