Radical antegrade modular pancreatosplenectomy: Myth or reality? A systematic review and trial sequential...

The superiority of radical antegrade modular pancreatosplenectomy versus standard distal pancreatectomy has never been demonstrated.

A systematic review was performed to identify all comparative studies about radical antegrade modular pancreatosplenectomy versus standard distal pancreatectomy. Random-effects analysis was performed, and hazard ratios, odds ratios, and mean differences were calculated. Using trial sequential analysis, type I and II errors were evaluated by comparing the accrued sample size with the required sample size. When the required sample size is superior to the accrued sample size, type I or II errors can be hypothesized. The critical endpoint was overall survival. Secondary endpoints were disease-free survival, R0 resection rate, major morbidity and mortality rate, clinically relevant postoperative pancreatic fistula, delayed gastric emptying, postpancreatectomy hemorrhage, length of stay, and operative time.

The accrued sample size and required sample size were 1,172 and 176 for the primary endpoint, respectively. The overall survival was similar between the 2 groups, with a hazard ratio of 1.33 (95% confidence interval: 0.89-2.00). The required sample size reached, and false-negative equivalence can be excluded. Disease-free survival, R0 resection rate, major morbidity and mortality rate, clinically relevant postoperative pancreatic fistula, delayed gastric emptying, postpancreatectomy hemorrhage, and operative time are similar and reached required sample sizes, suggesting that false equivalence can be excluded. Length of stay was shorter in radical antegrade modular pancreatosplenectomy than in standard distal pancreatectomy (-3.48 days; -6.66 to -0.31 days). The accrued sample size was 826, and the required sample size was not reached. False-positive results cannot be excluded.

Radical antegrade modular pancreatosplenectomy was not superior in guaranteeing a better overall survival and disease-free survival. The data are robust, and further retrospective comparative studies are unnecessary.