Pancreatoduodenectomy for suspected malignancy: nonmalignant histology confers increased risk of serious morbidity.

Background/aims: A tissue diagnosis is not always obtained prior to pancreatoduodenectomy (PD) and a proportion of patients are found to have noncancerous histology postoperatively. It is unknown if these patients have different outcomes when compared with those who have malignancy confirmed.

Methods: A retrospective paired case matched control study was undertaken. Patients who underwent PD for suspected malignancy but ultimately had nonmalignant histology were identified. Each was matched to a confirmed malignant control using the following criteria: age, gender, body mass index, American Society of Anesthesiologists grade, neoadjuvant treatment, preoperative serum bilirubin, preoperative biliary stenting and type of pancreatic anastomosis. Matching was blinded to the measured outcomes, which included perioperative morbidity and mortality.

Results: Forty-five cases were compared with 45 well-matched controls. There was no difference in 30- or 90-day mortality, or length of stay. While overall morbidity rates were the same, patients with nonmalignant disease were more likely to experience major (Clavien-Dindo grade III-IV) morbidity (40.0% versus 17.8%, = 0.0352). Independently, rates of clinically relevant pancreatic fistula (CR-POPF) were higher in the nonmalignant group (22.2% versus 4.44%, = 0.0131).

Conclusions: In our study, PD patients with nonmalignant histology had significantly higher incidence of major morbidity and CR-POPF when compared with those who had malignancy confirmed. This should be considered when planning the management of patients with known or presumed benign/premalignant disease.