Prophylactic abdominal drainage after distal pancreatectomy (PANDORINA): an international, multicentre, open-label, randomised controlled, non-inferiority trial
Eduard A van Bodegraven,Alberto Balduzzi,Tess M E van Ramshorst,Giuseppe Malleo,Frederique L Vissers,Jony van Hilst,Sebastiaan Festen,Mohammad Abu Hilal,Horacio J Asbun,Nynke Michiels,Bas Groot Koerkamp,Olivier R C Busch,Freek Daams,Misha D P Luyer,Marco Ramera,Giovanni Marchegiani,Joost M Klaase,I Quintus Molenaar,Matteo de Pastena,Gabriella Lionetto,Pier Giuseppe Vacca,Hjalmar C van Santvoort,Martijn W J Stommel,Daan J Lips,Mariëlle M E Coolsen,J Sven D Mieog,Roberto Salvia,Casper H J van Eijck,Marc G Besselink,Dutch Pancreatic Cancer Group
DOI: https://doi.org/10.1016/s2468-1253(24)00037-2
IF: 45.042
2024-03-17
The Lancet Gastroenterology & Hepatology
Abstract:Summary Background Prophylactic passive abdominal drainage is standard practice after distal pancreatectomy. 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 distal pancreatectomy. Methods In this international, multicentre, open-label, randomised controlled, non-inferiority trial, we recruited patients aged 18 years or older undergoing open or minimally invasive elective distal pancreatectomy for all indications in 12 centres in the Netherlands and Italy. We excluded patients with an American Society of Anesthesiology (ASA) physical status of 4–5 or WHO performance status of 3–4, added by amendment following the death of a patient with ASA 4 due to a pre-existing cardiac condition. Patients were randomly assigned (1:1) intraoperatively by permuted blocks (size four to eight) to either no drain or prophylactic passive drain placement, stratified by annual centre volume (<40 or ≥40 distal pancreatectomies) and low risk or high risk of grade B or C POPF. High-risk was defined as a pancreatic duct of more than 3 mm in diameter, a pancreatic thickness at the neck of more than 19 mm, or both, based on the Distal Pancreatectomy Fistula Risk Score. Other patients were considered low-risk. The primary outcome was the rate of major morbidity (Clavien–Dindo score ≥III), and the most relevant secondary outcome was grade B or C POPF, grading per the International Study Group for Pancreatic Surgery. Outcomes were assessed up to 90 days postoperatively and analysed in the intention-to-treat population and per-protocol population, which only included patients who received the allocated treatment. A prespecified non-inferiority margin of 8% was compared with the upper limit of the two-sided 95% CI (Wald) of unadjusted risk difference to assess non-inferiority. This trial is closed and registered in the Netherlands Trial Registry, NL9116. Findings Between Oct 3, 2020, and April 28, 2023, 376 patients were screened for eligibility and 282 patients were randomly assigned to the no-drain group (n=138; 75 [54%] women and 63 [46%] men) or the drain group (n=144; 73 [51%] women and 71 [49%] men). Seven patients in the no-drain group received a drain intraoperatively; consequently, the per-protocol population included 131 patients in the no-drain group and 144 patients in the drain group. The rate of major morbidity was non-inferior in the no-drain group compared with the drain group in the intention-to-treat analysis (21 [15%] vs 29 [20%]; risk difference –4·9 percentage points [95% CI –13·8 to 4·0]; p non-inferiority =0·0022) and the per-protocol analysis (21 [16%] vs 29 [20%]; risk difference –4·1 percentage points [–13·2 to 5·0]; p non-inferiority =0·0045). Grade B or C POPF was observed in 16 (12%) patients in the no-drain group and in 39 (27%) patients in the drain group (risk difference –15·5 percentage points [95% CI –24·5 to –6·5]; p non-inferiority <0·0001) in the intention-to-treat analysis. Three patients in the no-drain group died within 90 days; the cause of death in two was not considered related to the trial. The third death was a patient with an ASA score of 4 who died after sepsis and a watershed cerebral infarction at second admission, leading to multiple organ failure. No patients in the drain group died within 90 days. Interpretation A no-drain policy is safe in terms of major morbidity and reduced the detection of grade B or C POPF, and should be the new standard approach in eligible patients undergoing distal pancreatectomy. Funding Ethicon UK (Johnson & Johnson Medical, Edinburgh, UK).
gastroenterology & hepatology