Optimal Pancreatic Surgery: Are We Making Progress in North America?

Joal D. Beane, MD; Jeffrey D. Borrebach, MS; Amer H. Zureikat, MD; E. Molly Kilbane, RN; Vanessa M. Thompson, PhD; Henry A. Pitt, MD

Disclosures

Annals of Surgery. 2021;274(4):e355-e363. 

In This Article

Results

Patient Population and Trends in Preoperative Processes

Between 2013 and 2017, 16,222 patients who underwent PD and 7946 who underwent DP met the inclusion/exclusion criteria. Demographics for these patients are shown in Table 1. The median age for patients undergoing PD was 64.7 years, and 53.2% were male. The median age for patients undergoing DP was 61.0 years, and 56.5% were female. The majority of PD (77.8%) and DP (75.8%) patients were White. The average BMI for PD and DP was 27.4 and 28.9, respectively. DP patients were more likely to have diabetes (P < 0.001) and disseminated cancer (P < 0.01), whereas PD patients more commonly were smokers (P < 0.01) and lost weight (P < 0.001). PD patients also were more likely to have an ASA score of ≥3 (P < 0.001).

On unadjusted analysis, the percentage of patients who underwent PD for pancreatic cancer over the study period was unchanged, but a greater percentage of patients had a biliary stent (P < 0.05) and received neoadjuvant therapy (P < 0.001; Table 2, Figure 1A). However, the percentage of patients who received preoperative radiation therapy remained low (7.8%) and was unchanged over time (Table 2, Figure 1A). Similar trends were observed with respect to the percentage of patients with pancreatic cancer, receipt of neoadjuvant chemotherapy, and treatment with preoperative radiation therapy in patients who underwent DP (Table 3, Figure 1B).

Figure 1.

(A) Preoperative factors in patients undergoing pancreatoduodenectomy. (B) Operative approach in patients undergoing pancreatoduodenectomy. (C) Preoperative factors in patients undergoing distal pancreatectomy. (D) Operative approach in patients undergoing distal pancreatectomy.

Trends in Perioperative Processes and Operative Management

During the study period, 8.5% (n = 1368) of PD patients underwent a minimally invasive approach (MIPD). Of all PD patients, 3.7% (n = 597) were performed using the robotic platform (RPD) and 4.4% (n = 703) were performed using standard laparoscopy (LPD). While the overall percentage of patients who underwent MIPD did not change over the study period, the percentage of patients who underwent RPD increased from 2.5% in 2013 to 4.2% in 2017 (P < 0.001), and LPD decreased from 5.8% in 2013 to 4.3% in 2017 (P = 0.018; Table 2, Figure 1C). In contrast, the percentage of DP patients who underwent a minimally invasive approach (MIDP) was 52.2% (n = 4136) and increased over time (P = 0.004). Of all patients, the percentage undergoing standard laparoscopic DP (LDP) was 39.9% (n = 3163) and did not change over time, but the percentage of patients who underwent DP using a robotic approach (RDP) was 11.9% (n = 941) and increased over time (P < 0.001; Table 3, Figure 1D).

With regards to the technical details of PD, the percentage of patients who had a pylorus preserving PD reduced over time (P < 0.001; Table 2). The percentage of patients who had a duct-to-mucosal pancreaticojejunal anastomosis was 85% (n = 11,139) and use of this technique increased over time (P = 0.018; Table 2). The percentage of patients who had a pancreaticojejunal invagination (8.7%, n = 1145) decreased over time (P < 0.001; Table 2). The percentage of patients who underwent a concomitant major vascular resection (venous and/or arterial) was 17.4% (n = 2320) in the PD cohort and did not change over time (P = 0.194; Table 2). In the subset of patients with pancreatic cancer, the percentage of patients who underwent a concomitant major vascular resection was 23.9% (n = 1760) and also did not change over time (P = 0.355).

Trends in Operative Time and Perioperative Transfusions

The mean operative time in patients undergoing PD and DP were 372 and 230 minutes, respectively. Operative time for patients undergoing PD decreased over time (P < 0.05; Table 2, Figure 2A), but not for patients undergoing DP (Table 3). The mean percentage of patients receiving perioperative transfusions was 17.7% for PD patients (Table 2) and 10.9% for DP patients (Table 3). Transfusions for patients undergoing PD decreased over time (P < 0.001; Table 2, Figure 2A), but not for patients undergoing DP (Table 3).

Figure 2.

(A) Operative time and transfusions in pancreatoduodenectomy patients. (B) Mortality and overall morbidity in pancreatoduodenectomy patients. (C) Sepsis and percutaneous drainage in pancreatoduodenectomy. (D) Length of stay (LOS) and readmissions in pancreatoduodenectomy patients.

Trends in Drain Placement and Management

Intraoperatively placed drains were used in 88.3% (n = 11,898) and 85.9% (n = 5712) of patients who underwent PD and DP, respectively. No significant change was observed over time in the use of surgical drains for either operation. However, an increasing percentage of patients who underwent PD and DP had a POD 1 drain fluid amylase (DFA-1) analyzed as part of their management (both P < 0.001). In addition, both cohorts had a greater percentage of patients who had their surgically placed drains removed by POD 3 (both P < 0.001). However, the percentage of patients with a DFA-1 less than 5000 IU/L who had their drains removed by POD 3 was only 26.9% overall for patients who had PD and 30.1% for those who had DP; neither of these processes changed significantly over time.

Trends in Postoperative Outcomes

Unadjusted postoperative outcomes for patients who underwent PD are shown in Table 2. Mortality after PD was 1.7% and overall morbidity was 47.0% during the study period, and both improved over time (P < 0.05; Table 2, Figure 2B). On risk-adjusted analysis, a statistically significant improvement in overall morbidity was observed over time [odds ratio (OR_ 0.963, P = 0.018]. The incidence of CR-POPF and delayed gastric emptying were 13.9% and 16.4%, and both were unchanged over time (Table 2). A reduction was observed in the percentage of patients undergoing PD who developed postoperative sepsis/septic shock (P = 0.003; Table 2, Figure 2C), and required percutaneous drainage (P = 0.005; Table 2, Figure 2C). The reduction in each was confirmed on risk-adjusted analyses (Table 4). A statistically significant reduction was observed in PD patients in the mean postoperative LOS (P = 0.002; Table 2, Figure 2D). On risk-adjusted analysis, an increase in the percentage of patients with an established LOS ≤12 days also was observed (P = 0.009; Table 4).

Postoperative outcomes for patients undergoing DP are shown in Table 3. Mortality and overall morbidity for patients undergoing DP were 0.6% and 39.7%, respectively, and remained relatively constant over the study period (Table 3). The incidence of CR-POPF and delayed gastric emptying were 11.5% and 4.0%, respectively, and also were not significantly different over time (Table 3). However, the use of percutaneous drainage after surgery decreased over time on both unadjusted (P = 0.038; Table 3) and risk-adjusted analyses (OR 0.927, P = 0.037; Table 4). The percentage of DP patients with acceptable LOS ≤7 days increased over time on both unadjusted (P = 0.007; Table 3) and risk-adjusted analyses (OR 0.936, P = 0.008; Table 4).

Trends in Optimal Pancreatic Surgery

The percentage of patients who achieved optimal pancreatic surgery after PD was 55.4% (Table 2). Optimal pancreatic surgery for PD patients increased over time on both unadjusted (P = 0.002; Table 2, Figure 3) and risk-adjusted analyses (OR 1.058, P < 0.001; Table 4). The percentage of patients who achieved optimal pancreatic surgery after DP was 55.8% (Table 3). Optimal pancreatic surgery for DP patients also increased over time on both unadjusted (P < 0.001; Table 3, Figure 3) and risk-adjusted analyses (OR 1.085, P < 0.001; Table 4).

Figure 3.

Optimal pancreatic surgery in pancreatoduodenectomy (Whipple) and distal pancreatectomy patients.

processing....