Prophylactic abdominal drainage after distal pancreatectomy (PANDORINA): an international, multicentre, randomised controlled trial

Background: Prophylactic abdominal drainage is standard practice after distal pancreatectomy (DP). This approach aims to mitigate the consequences of postoperative pancreatic fistula (POPF) but its added value, especially in patients at low risk of POPF, is currently being debated. We aimed to assess the non-inferiority of a no-drain policy in patients after DP. Methods: In this international, multicentre, randomised controlled non-inferiority trial, we recruited patients undergoing elective DP in 12 centres in the Netherlands and Italy. Patient were randomly assigned to either no drain or drain placement. Stratification was performed for patients at low or high risk of POPF, based on the DP fistula risk score (D-FRS). Primary outcome was major morbidity (Clavien-Dindo score ≥3) and the most relevant secondary outcome was grade B/C POPF. Analyses were performed by intention-to-treat. The predefined non-inferiority margin of 8% was compared with the upper limit of the two-sided 95% confidence interval (CI) of absolute difference in both the primary and most relevant secondary outcome. Subgroup analyses were based on the D-FRS. This trial is registered with the Netherlands Trial Registry (NL9116). Findings: Between October 3, 2020 and April 28, 2023, 282 patients undergoing DP were randomised: 138 in the no-drain group and 144 in the drain group. Major morbidity was comparable between groups (21 [15·2%] vs 29 [20·1%], difference -4·9%, 95% CI -13·77 to 3·97, pnon-inferiority = 0·002). The rates of grade B/C POPF (16 [11·6%] vs 39 [27·1%], difference -15·5%, 95% CI -24·51 to -6·49, psuperiority <0·001) and overall complications (46 patients, 33·3% vs 73 patients, 50·7%, p=0·003) were both lower in the no-drain group. The rates of postoperative radiological and endoscopic interventions (14 patients, 10·1% vs 24 patients, 16·7%, p=0·109) and reoperations (6 patients, 4·4% vs 4 patients, 2·8%, p=0·476) were comparable between groups. In the low-risk POPF group (n=81), the no-drain group had a lower major morbidity rate (2 [4·5%] vs 7 [18·9%], difference -14·4, 95% CI -28·42 to -0·38, p=0·040). In the intermediate- and high-risk POPF groups, major morbidity did not differ between the groups. Interpretation: A no-drain policy after DP is non-inferior to drain placement in terms of major morbidity and superior in terms of grade B/C POPF which was reduced by over 50%.