PATIENTS AND METHODS: We compared a prospectively collected group of 48 patients undergoing oxaliplatin/irinotecan-based perioperative systemic chemotherapy (s-CT) with targeted agents, and cytoreductive surgery (CRS) (no-HIPEC group) with 48 controls undergoing the same perioperative s-CT and CRS/HIPEC (HIPEC group). Patients were matched (1:1) according to the Peritoneal Surface Disease Severity Score, completeness of cytoreduction, history of extraperitoneal disease (EPD), and Peritoneal Cancer Index.

RESULTS: The groups were comparable, except for a higher number of patients in the HIPEC group with World Health Organization performance status 0, pN2 stage primary tumor, and treated with preoperative s-CT. Forty-one patients in the no-HIPEC group and 43 patients in the HIPEC group had optimal comprehensive treatment (P = 0.759), defined as complete cytoreduction of PM and margin-negative EPD resection. Median follow-up was 31.6 months in the no-HIPEC group and 39.9 months in the HIPEC group. Median overall survival was 39.3 months in the no-HIPEC group and 34.8 months in the HIPEC group (P = 0.702). In the two groups, severe morbidity occurred in 14 (29.2%) and 13 (27.1%) patients, respectively (P = 1.000), with no operative deaths. On multivariate analysis, left-sided primary and curative treatment independently correlated with better survival while HIPEC did not (hazard ratio 0.73; 95% confidence interval 0.47-1.15; P = 0.178).

METHODS: In this open-label, phase 3, multicentre randomised trial, patients aged 21-80 years with cT3 or cT4 gastric cancer undergoing curative resection were enrolled at 22 centres from South Korea, China, Japan, Malaysia, Hong Kong, and Singapore. Patients were randomly assigned to receive surgery and EIPL (EIPL group) or surgery alone (standard surgery group) via a web-based programme in random permuted blocks in varying block sizes of four and six, assuming equal allocation between treatment groups. Randomisation was stratified according to study site and the sequence was generated using a computer program and concealed until the interventions were assigned. After surgery in the EIPL group, peritoneal lavage was done with 1 L of warm (42°C) normal 0·9% saline followed by complete aspiration; this procedure was repeated ten times. The primary endpoint was overall survival. All analyses were done assuming intention to treat. This trial is registered with ClinicalTrials.gov, NCT02140034.

FINDINGS: Between Sept 16, 2012, and Aug 3, 2018, 800 patients were randomly assigned to the EIPL group (n=398) or the standard surgery group (n=402). Two patients in the EIPL group and one in the standard surgery group withdrew from the trial immediately after randomisation and were excluded from the intention-to-treat analysis. At the third interim analysis on Aug 28, 2019, the predictive probability of overall survival being significantly higher in the EIPL group was less than 0·5%; therefore, the trial was terminated on the basis of futility. With a median follow-up of 2·4 years (IQR 1·5-3·0), the two groups were similar in terms of overall survival (hazard ratio 1·09 [95% CI 0·78-1·52; p=0·62). 3-year overall survival was 77·0% (95% CI 71·4-81·6) for the EIPL group and 76·7% (71·0-81·5) for the standard surgery group. 60 adverse events were reported in the EIPL group and 41 were reported in the standard surgery group. The most common adverse events included anastomotic leak (ten [3%] of 346 patients in the EIPL group vs six [2%] of 362 patients in the standard surgery group), bleeding (six [2%] vs six [2%]), intra-abdominal abscess (four [1%] vs five [1%]), superficial wound infection (seven [2%] vs one [<1%]), and abnormal liver function (six [2%] vs one [<1%]). Ten of the reported adverse events (eight in the EIPL group and two in the standard surgery group) resulted in death.

INTERPRETATION: EIPL and surgery did not have a survival benefit compared with surgery alone and is not recommended for patients undergoing curative gastrectomy for gastric cancer.