Pancreaticojejunostomy and clinically relevant pancreatic fistula: a modified dunking technique and single-center experience

Abstract Background Pancreaticojejunostomy (PJ) anastomosis is the “Achilles heel” of the pancreatoduodenectomy (PD). Clinicaly relevant postoperative pancreatic fistula (CR-POPF) remains a significantly morbid complication after pancreatic resection. Methods This retrospective study compares patients who underwent PD utilizing our modified dunking technique for PJ anastomosis to patients who underwent the standard PD technique. The studied outcome was CR-POPF. Patient demographics and clinical characteristics were summarized and compared. The association between surgical technique and CR-POPF was assessed using univariate logistic regression and four multivariable logistic models, each assessing technique while adjusting for gland texture, intraoperative blood loss, receipt of neoadjuvant therapy, or pathology. A supplementary inverse probability of treatment weighting (IPTW) analysis among the subset of patients with complete fistula risk score components was also performed. Results A total of 208 patients who underwent the modified dunking technique were compared with 159 patients who underwent the standard technique. CR-POPF occurred in 8.2% of modified-technique patients compared with 15.7% of standard-technique patients ( p = 0.024), with 42 events (11.4%) overall. Overall morbidity and Clavien-Dindo grade ≥ III complications were lower after the modified technique (12% vs. 27%, p = 0.002), as were reoperation rates (6.4% vs. 18.3%, p = 0.002) and length of stay (mean 9.5 days vs. 13.1 days p = 0.025). The modified-technique was associated with later year of surgery ( p < 0.001) and a higher proportion of firm glands ( p = 0.002). Modified technique was associated with significantly lower rates of CR-POPF in univariate analysis (OR 0.48, p = 0.027), while none of the four multivariable models showed a significant difference (gland texture OR 1.27, p = 0.646; intraoperative blood loss OR 0.66, p = 0.341; receipt of neoadjuvant therapy OR 0.65, p = 0.215; pathology OR 0.54, p = 0.07). In a supplementary IPTW analysis of 238 patients with complete fistula risk score components, the modified technique no significant difference was found in odds of CR-POPF (OR 3.48, 95% CI 0.90–13.5, p = 0.07). Conclusions Unadjusted analysis showed significantly improved fistula rates among patients receiving the modified technique while adjusted analyses showed no statistically significant difference. The technique is reproducible and safe, and initial results are encouraging, but its true effect on CR-POPF warrants prospective evaluation.