Outcomes and Risk Score for Distal Pancreatectomy with Celiac Axis Resection (DP-CAR): An International Multicenter Analysis

Abstract

Background: Distal pancreatectomy with celiac axis resection (DP-CAR) is a treatment option for selected patients with pancreatic cancer involving the celiac axis. A recent multicenter European study reported a 90-day mortality rate of 16%, highlighting the importance of patient selection. The authors constructed a risk score to predict 90-day mortality and assessed oncologic outcomes.

Methods: This multicenter retrospective cohort study investigated patients undergoing DP-CAR at 20 European centers from 12 countries (model design 2000-2016) and three very-high-volume international centers in the United States and Japan (model validation 2004-2017). The area under receiver operator curve (AUC) and calibration plots were used for validation of the 90-day mortality risk model. Secondary outcomes included resection margin status, adjuvant therapy, and survival.

Results: For 191 DP-CAR patients, the 90-day mortality rate was 5.5% (95 confidence interval [CI], 2.2-11%) at 5 high-volume (≥ 1 DP-CAR/year) and 18% (95 CI, 9-30%) at 18 low-volume DP-CAR centers (P = 0.015). A risk score with age, sex, body mass index (BMI), American Society of Anesthesiologists (ASA) score, multivisceral resection, open versus minimally invasive surgery, and low- versus high-volume center performed well in both the design and validation cohorts (AUC, 0.79 vs 0.74; P = 0.642). For 174 patients with pancreatic ductal adenocarcinoma, the R0 resection rate was 60%, neoadjuvant and adjuvant therapies were applied for respectively 69% and 67% of the patients, and the median overall survival period was 19 months (95 CI, 15-25 months).

Conclusions: When performed for selected patients at high-volume centers, DP-CAR is associated with acceptable 90-day mortality and overall survival. The authors propose a 90-day mortality risk score to improve patient selection and outcomes, with DP-CAR volume as the dominant predictor.

Fig. 1

Study flow chart of data sources and year of inclusion. Two cases were lost to follow-up evaluation within 90 days after surgery. E-AHPBA, European-African Hepato-Pancreato-Biliary Association; JHH, Johns Hopkins Hospital; UPMC, University of Pittsburgh Medical Center; WMUH, Wakayama Medical University Hospital

Fig. 2

Discrimination curves for 90-day mortality prediction. Receiver operator curves (ROC) for the 90-day mortality prediction model. The area under the curve (AUC) was 0.79 (95% confidence interval [CI], 0.65–0.93) in the design cohort (n = 71) and 0.74 (95% CI, 0.56–0.92) in the validation cohort (n = 120). The difference is not significant (P = 0.642)