What is the Incidence of Malignancy in Resected Intraductal Papillary Mucinous Neoplasms? An Analysis of Over 100 US Institutions in a Single Year

BACKGROUND

A subset of intraductal papillary mucinous neoplasms (IPMNs) will progress to invasive adenocarcinoma, however identifying invasive from non-invasive disease preoperatively remains challenging. The rate of malignancy in resected IPMNs in the US remains unclear.

OBJECTIVE

We aimed to determine the rate of malignancy and factors associated with high-risk pathology in resected IPMNs.

METHODS

The most recent annual cohort of patients undergoing pancreatectomy included in the American College of Surgeons National Surgical Quality Improvement Program were assessed, and contributions of demographics, preoperative laboratory values, and outcome data to level of IPMN dysplasia were analyzed. The main outcomes were incidence of invasive carcinoma or high-grade dysplasia.

RESULTS

Of 5025 pancreatectomies in 1 year, 478 patients underwent pancreatectomy for IPMN. Invasive carcinoma/high-grade dysplasia was identified in 23% of resected lesions, and there was no difference in patient characteristics or type of resection performed in patients with invasive versus non-invasive pathology. Patients with invasive IPMNs presented significantly more often with high liver function tests, >10% weight loss, clinical jaundice and stent placement, and were more likely to undergo an open operation (p = 0.03). There were no differences in perioperative outcomes. Adjusted logistic regression identified an association between invasive disease and non-soft pancreatic gland texture (odds ratio 0.19, 95% confidence interval 0.05-0.68, p < 0.01).

CONCLUSIONS

Approximately 10% of all pancreatectomies in the US are for IPMNs. In these patients, treated after the revised international consensus guidelines, only 23% of IPMNs contained invasive or high-grade histology. Resections carried similar morbidity regardless of pathology. Improved biomarkers are needed to aid in surgical selection.

Perioperative Broad-spectrum Antibiotics are Associated With Decreased Surgical Site Infections Compared to 1st-3rd Generation Cephalosporins After Open Pancreaticoduodenectomy in Patients With Jaundice or a Biliary Stent

OBJECTIVE

Compare the effectiveness of 1st-3rd generation cephalosporins (1st-3rdCE) to broad-spectrum antibiotics in decreasing surgical site infections (SSI) after pancreatectomy.

SUMMARY OF BACKGROUND DATA

SSI is one of the most common complications after pancreatic surgery. Various antibiotic regimens are utilized nationwide with no clear guidelines for pancreatectomy. As we await results of a recently initiated prospective trial, this study retrospectively evaluates over 15,000 patients using the same administrative data abstraction tools as in the trial.

METHODS

All relevant clinical variables were collected from the 2016-2018 targeted-pancreatectomy database from the American College of Surgeon National Surgical Quality Improvement Program. Preoperative antibiotics were initially collected as first-generation cephalosporin, second or third-generation cephalosporin, and broad-spectrum antibiotics (Broad-abx).

RESULTS

Of the 15,182 patients who completed a pancreatic surgery between 2016 and 2018, 6114 (40%) received a first-generation cephalosporin, 4097 (27%) received a second or third-generation cephalosporin, and 4971 (33%) received Broad-abx. On multivariate analysis, Broad-abx was associated with a decrease in all-type SSI compared to 1st-3rdCE (odds ratio = 0.73-0.77, P < 0.001) after open pancreaticoduodenectomy (PD). There was no difference in SSI between antibiotic-types after distal pancreatectomy. Subgroup multivariate analysis of open PD revealed decrease in all-type SSI with Broad-abx amongst patients with jaundice and/or biliary stent only, regardless of wound protector use (odds ratio = 0.69-0.70, P < 0.001). Propensity score matching of open PD patients with jaundice and/or biliary stent confirmed a decrease in all-type SSI (19% vs 24%, P = 0.001), and organ-space SSI (12% vs 16%, P < 0.001).

CONCLUSION

Broad-abx are associated with decreased SSI after open PD and may be preferred specifically for patients with preoperative biliary stent and/ or jaundice.

Should Drains Suck? A Propensity Score Analysis of Closed-Suction Versus Closed-Gravity Drainage After Pancreatectomy

BACKGROUND

Post-operative pancreatic fistula (POPF) remains one of the most common complications after pancreatic surgery. We previously reported that the majority of US surgeons leave drains after pancreatectomy. However, there remains controversy and surgeon bias on the use of gravity compared with suction drainage with limited data on patient outcomes to guide management.

METHODS

Demographics, comorbidities, perioperative, and outcome data were captured from the most recent ACS National Surgical Quality Improvement Program (NSQIP)-targeted pancreatectomy databases. This is a retrospective cohort analysis comparing closed-suction to closed-gravity drains with multivariate analysis and propensity score matching (PSM).

RESULTS

Of 9232 patients that underwent a pancreatectomy with closed drain placement, 1345 (15%) were to gravity and 7887 (85%) were to suction. On multivariate and PSM, stratified by surgery-type, there was no difference in biochemical leak (Whipple, 4 vs. 4%; distal, 8 vs. 6%) or clinically relevant (CR)-POPF (Whipple, 13 vs. 15%; distal, 12 vs. 15%). On multivariate analysis, there was an increase in organ-space surgical site infections with suction drains for patients undergoing Whipple procedure (12 vs. 16%, p = 0.004), which did not persist on PSM (p = 0.088). Finally, there were no significant differences in amylase level, time to drain removal, or superficial surgical site infections for patients undergoing either procedure based on drain type.

CONCLUSION

The majority of drains utilized after pancreatectomy in the USA are placed to suction, though a significant proportion are kept to gravity. Neither type of drain is associated with increased CR-POPF or other post-operative outcomes compared with the other; therefore, both types remain reasonable options if drains are to be placed.

Do Drains Contribute to Pancreatic Fistulae? Analysis of over 5000 Pancreatectomy Patients

INTRODUCTION

Conflicting evidence exists from randomized controlled trials supporting both increased complications/fistulae and improved outcomes with drain placement after pancreatectomy. The objective was to determine drain practice patterns in the USA, and to identify if drain placement was associated with fistula formation.

METHODS

Demographic, perioperative, and patient outcome data were captured from the most recent annual NSQIP pancreatic demonstration project database, including components of the fistula risk score. Significant variables in univariate analysis were entered into adjusted logistic regression models.

RESULTS

Of 5013 pancreatectomy patients, 4343 (87%) underwent drain placement and 18% of patients experienced a pancreatic fistula. When controlled for other factors, drain placement was associated with ducts < 3 mm, soft glands, and blood transfusion within 72 h of surgery. Age, obesity, neoadjuvant radiation, preoperative INR level, and malignant histology lost significance in the adjusted model. Drained patients experienced higher readmission rates (17 vs. 14%; p < 0.05) and increased (20 vs. 8%; p < 0.01) pancreatic fistulae. Fistula was associated with obesity, no neoadjuvant chemotherapy, drain placement, < 3 mm duct diameter, soft gland, and longer operative times. Drain placement remained independently associated with fistula after both distal pancreatectomy (OR = 2.84 (1.70, 4.75); p < 0.01) and pancreatoduodenectomy (OR = 2.29 (1.28, 4.11); p < 0.01).

CONCLUSIONS

Despite randomized controlled clinical trial data supporting no drain placement, drains are currently placed in the vast majority (87%) of pancreatectomy patients from > 100 institutions in the USA, particularly those with soft glands, small ducts, and perioperative blood transfusions. When these factors are controlled for, drain placement remains independently associated with fistulae after both distal and proximal pancreatectomy.

Epidural Analgesia Is Associated with Prolonged Length of Stay After Open HPB Surgery in Over 27,000 Patients

BACKGROUND

The impact of epidural analgesia (EA) on postoperative morbidity and length of stay (LOS) after HPB surgery remains to be determined. These specific outcomes have been highlighted by the implementation of multiple enhanced recovery pathways (ERAS). The authors hypothesized that EA in the current environment may be associated with LOS and other outcomes.

METHODS

The American College of Surgeons (ACS) National Surgical Quality Improvement Program (NSQIP) databases from 2014 to 2017 for patients undergoing open hepatopancreaticobiliary (HPB) surgery were included in a retrospective cohort analysis with propensity score matching (PSM) comparing EA with control.

RESULTS

Twenty-seven thousand two hundred eighteen patients underwent open HPB surgery, of which 6048 (22%) received EA. There was an increase use of EA over time (from 19.3 to 25.5%, p = 0.001). On PSM, EA was associated with more than half of a day increase in LOS for both pancreatic (p < 0.001) and hepatic surgery (p < 0.001). Furthermore, for pancreatic surgery, there was an increase in urinary tract infection (2.5% vs. 3.3%, p = 0.018), time to drain removal (7.8 vs. 8.7 days, p < 0.001), and discharge to rehabilitation (2.9% vs. 4.3%, p = 0.029). For hepatic surgery, there was an increase in blood transfusion requirements (17% vs. 20%, p = 0.019). There were no differences in overall morbidity and mortality.

CONCLUSION

In this cohort of over 27,000 patients with granular surgical details, there was a significant increase in LOS associated with EA after HPB surgery, along with increased procedure-specific UTI and blood transfusion. With the ever-increasing need for standardized and efficient patient care pathways that reduce LOS, alternative analgesic adjuncts may be considered to optimize patient outcomes.

A propensity score analysis of over 12,000 pancreaticojejunal anastomoses after pancreaticoduodenectomy: does technique impact the clinically relevant fistula rate?

BACKGROUND

Clinically relevant postoperative pancreatic fistula (CR-POPF) remains a major cause of morbidity in patients undergoing pancreatic surgery. Controversy exists as to whether there is any difference in CR-POPF with a Duct-to-Mucosa (DTM) versus an Invagination (IG) pancreaticojejunostomy (PJ).

METHODS

Demographic, perioperative, intraoperative, and postoperative data were captured from the American College of Surgeons (ACS) National Surgical Quality Improvement Program (NSQIP) 2014-2017 databases. Potential confounders were included in a logistic regression and a propensity score model. The primary outcome was CR-POPF.

RESULTS

A total of 12,361 pancreaticojejunal anastomoses were performed with 11,168 patients undergoing DTM (90%) and 1193 undergoing IG (10%) after pancreaticoduodenectomy. Amongst all patients, there was no significant difference in CR-POPF between DTM and IG on multivariate (OR = 0.95, p = 0.64) or propensity score analysis (OR = 0.99, p = 0.93). After stratification by pancreatic gland texture and duct size, there was a decrease in CR-POPF with DTM amongst patients with duct size greater than 6 mm on multivariate analysis (OR = 0.35, p = 0.009) and propensity score analysis (OR = 0.40, p = 0.018). There were no significant differences in any other strata.

CONCLUSION

DTM or IG technique are not associated with CR-POPF for patients with average size pancreatic ducts; however, DTM is preferable in patients with large pancreatic duct diameter (>6 mm).

Drain Placement After Pancreatic Resection: Friend or Foe For Surgical Site Infections?

BACKGROUND

Despite multiple studies and randomized trials, there remains controversy over whether drains should be placed, and if so for how long, after pancreas resection. The aim was to determine if post-pancreatectomy drain placement and timing of drain removal were associated with differences in infectious outcomes and, if so, which specific procedures and infectious sites were most at risk.

METHODS

The ACS-NSQIP targeted pancreatectomy database was utilized to identify patients who underwent pancreatectomies between 2015 and 2020 with postoperative drain placement for retrospective cohort analysis. A propensity score matching analyses was conducted to determine associations between drain placement and surgical site infections (SSI).

RESULTS

Of 39,057 pancreatic resections, 66.4% were proximal pancreatectomies, and 33.6% were distal pancreatectomies. After propensity score matching, drain placement was not associated with significantly lower rates of superficial SSI (7% vs 9%, p = 0.755) or organ/space SSI (17% vs 16%, p = 0.647) after proximal pancreatectomy. After distal pancreatectomy, drain placement was associated with higher rates of organ/space SSI (12% vs 9%, p = 0.010). Drain removal on or after postoperative day 3 was significantly associated with higher rates of SSI in both proximal and distal pancreatectomy.

CONCLUSIONS

Drain placement is associated with an increased rate of organ/space SSI after distal pancreatectomy and not after pancreaticoduodenectomy. When drains are utilized, early removal is associated with a reduction of SSI after all types of pancreatectomy. In surgical units where post-pancreatectomy SSI is a concern, selective drain placement for high-risk glands or after distal pancreatectomy, combined with early drain removal, may be considered.