Perioperative chemotherapy versus adjuvant chemotherapy strategies in resectable gastric and gastroesophageal cancer: A Markov decision analysis.

Background: Perioperative chemotherapy has been shown to improve overall survival (OS) for operable gastric and gastroesophageal cancer. However, optimal sequence of surgery and chemotherapy has not been clearly identified. Markov models are useful for analyzing the outcomes of different treatment strategies in the absence of adequately powered randomized clinical trials. In this study, we use Markov decision analysis models to compare median OS (mOS), quality-adjusted mOS, life expectancy (LE), and quality-adjusted life expectancy (QALE) of perioperative chemotherapy with adjuvant chemotherapy strategies in resectable gastric and gastroesophageal cancer patients.

Methods: Markov models are constructed to compare two strategies: adjuvant chemotherapy after surgery and preoperative chemotherapy followed by cancer resection and postoperative chemotherapy. LE and QALE are calculated analytically, and mOS are obtained by simulation. Parameters used in the models are computed from prospective clinical trial data published in PUBMED from January 2000 to July 2020.

Results: Total of 8088 patients from 25 prospective studies were included in this analysis. Regardless of R0 resection ratio, the analyses of the models show a higher mOS for patients in the perioperative therapy arm compared to adjuvant chemotherapy. For R0 resected patients, the perioperative therapy arm provided an additional 11.0 mOS months (61.3 months vs. 50.3 months). For R1 resected patients, the perioperative therapy arm had mOS of 17.0 months vs. 10.7 months in adjuvant therapy.

Conclusions: The Markov models indicate that perioperative chemotherapy improves mOS, quality-adjusted mOS, LE, and QALE for resectable gastric and gastroesophageal cancer patients compared to adjuvant chemotherapy strategies.