Faculty Bibliography

BACKGROUND:Two types of pancreatic duct stents are used to improve postoperative outcomes of pancreatic anastomosis. The aim of this meta-analysis was to evaluate and compare the postoperative outcomes of patients with internal or external stenting during pancreaticoduodenectomy (PD). METHODS:We searched PubMed, EMBASE, the Cochrane Library and Web of Science databases until the end of December, 2014. Studies comparing outcomes of external vs. internal stent placement in PD were eligible for inclusion. Included literature was extracted and assessed by two independent reviewers. RESULTS:Seven articles were identified for inclusion: three randomized controlled trials (RCTs) and four observational clinical studies (OCS). The meta-analyses revealed that use of external stents had advantage on reducing the incidences of pancreatic fistula (PF) in total [odds ratio (OR) =0.69; 95% confidence interval (CI), 0.48-0.99; P=0.04], PF in soft pancreas (OR =0.30; 95% CI, 0.16-0.56; P=0.0002) and delayed gastric emptying (DGE) (OR =0.58; 95% CI, 0.38-0.89; P=0.01) compared with internal stents. There were no significant differences in other postoperative outcomes between two stenting methods, including postoperative morbidity (OR =0.93; 95% CI, 0.39-2.23; P=0.88), overall mortality (OR =0.70; 95% CI, 0.22-2.25; P=0.55), and intra-abdominal collections (OR =0.67; 95% CI, 0.26-1.71; P=0.40). CONCLUSIONS:Based upon this meta-analysis, the use of external pancreatic stents might have potential benefit in reducing the incidence of PF and DGE. Due to the limited number of original studies, more RCTs are needed to further support our result and clarify the issue.

The majority of patients with curative resection of pancreatic ductal adenocarcinoma recur within 5 years of resection. However, the prognosis associated with different patterns of recurrence has not been well studied. A retrospective review of patients who underwent curative surgical resection of pancreatic cancer was performed. Of the 209 patients, 174 patients developed recurrent disease. Of these 174, 28(16.1%) had recurrent disease limited to lung metastases, 20(11.5%) had recurrence in the lung plus one or more other sites excluding the liver, 73(42.0%) had liver metastasis alone or liver metastasis with any other site except lung, 28(16.1%) local recurrence only, and 25(14.3%) peritoneal recurrence alone or together with local recurrence. Patients with recurrence limited to lung had a 8.5 months(Mo) median survival from recurrence to death, which was significantly better than the survival associated with recurrence in the liver(5.1Mo), in the peritoneum(2.3Mo) or locally(5.1Mo) in multivariable analyses. Among all groups, the time from surgery to the diagnosis of recurrence in patients who recurred in only in the lung was also the longest. However, 75% of patients were found to have indeterminate lung nodules on their surveillance CT scans prior to the diagnosis of recurrence in lung. This delayed diagnosis of lung recurrence may have a negative impact on survival after recurrence. In conclusion, pancreatic cancer with lung recurrence has a significantly better prognosis than recurrence in other sites. Further studies are needed to investigate how different diagnostic and treatment modalities affect the survival of this unique subpopulation of pancreatic cancer patients.

BACKGROUND:The purpose of this study was to investigate the prognostic significance of early (30-day) hospital readmission (EHR) on mortality after pancreatectomy. METHODS:Using a prospectively collected institutional database linked with a statewide dataset, we evaluated the association between EHR and overall mortality in all patients undergoing pancreatectomy at our tertiary institution (2005 to 2010). RESULTS:Of 595 pancreatectomy patients, EHR occurred in 21.5%. Overall mortality was 29.4% (median follow-up 22.7 months). Patients with EHR had decreased survival compared with those who were not readmitted (P = .011). On multivariate analysis adjusting for baseline group differences, EHR for gastrointestinal-related complications was a significant independent predictor of mortality (hazard ratio 2.30, P = .001). CONCLUSIONS:In addition to known risk factors, 30-day readmission for gastrointestinal-related complications following pancreatectomy independently predicts increased mortality. Additional studies are necessary to identify surgical, medical, and social factors contributing to EHR, as well as interventions aimed at decreasing postpancreatectomy morbidity and mortality.

Our neoadjuvant clinical trial of a GM-CSF secreting allogeneic pancreas tumor vaccine (GVAX) revealed the development of tertiary lymphoid aggregates (TLAs) within the pancreatic ductal adenocarcinoma (PDA) tumor microenvironment 2 weeks after GVAX treatment. Microarray studies revealed that multiple components of the TGF-β pathway were suppressed in TLAs from patients who survived greater than 3 years and who demonstrated vaccine-enhanced mesothelin-specific T cell responses. We tested the hypothesis that combining GVAX with TGF-β inhibitors will improve the anti-tumor immune response of vaccine therapy. In a metastatic murine model of pancreatic cancer, combination therapy with GVAX vaccine and a TGF-β blocking antibody improved the cure rate of PDA-bearing mice. TGF-β blockade in combination with GVAX significantly increased the infiltration of effector CD8+ T lymphocytes, specifically anti-tumor-specific IFN-g producing CD8+ T cells, when compared to monotherapy controls (all p < 0.05). TGF-β blockade alone did not deplete T regulatory cells (Tregs), but when give in combination with GVAX, GVAX induced intratumoral Tregs were depleted. Therefore, our PDA preclinical model demonstrates a survival advantage in mice treated with an anti-TGF-β antibody combined with GVAX therapy and provides strong rational for testing this combinational therapy in clinical trials.

Pancreatic resection has become a feasible treatment of pancreatic neoplasms, and with improvements in surgical techniques and perioperative management, mortality associated with pancreatic surgery has decreased considerably. Despite this improvement, a high rate of complications is still associated with these procedures. Among these complications, delayed gastric emptying (DGE) and postoperative pancreatic fistula (POPF) have a substantial impact on patient outcomes and burden our healthcare system. Technical modifications and postoperative approaches have been proposed to reduce rates of both POPF and DGE in patients undergoing pancreatectomy; however, to date, their rates have remained unchanged. In the present study, we summarize the findings of the most significant studies that have investigated these complications. In particular, several studies focused on technical modifications including extent of dissection, stent placement, nature of anastomosis, type of reconstruction, and application of biological or non-biological agents to site of anastomosis. Moreover, postoperatively, drain placement, duration of drain usage, postoperative feeding, and use of pharmacological agents were studied to reduce rates of POPF and DGE. In this review, we summarize the most relevant literature on this fundamental aspect of pancreatic surgery. Despite studies identifying the potential benefit of technical modifications and postoperative approaches, these findings remain controversial and suggest need for further extensive investigation. Most importantly, we recommend that all surgeons performing these procedures base their practice on the most updated and highest available level of evidence.

OBJECTIVE:The study aim was to quantify the burden of complications of pancreatoduodenectomy (PD). BACKGROUND:The Postoperative Morbidity Index (PMI) is a quantitative measure of the average burden of complications of a procedure. It is based on highly validated systems--ACS-NSQIP and the Modified Accordion Severity Grading System. METHODS:Nine centers contributed ACS-NSQIP complication data for 1589 patients undergoing PD from 2005 to 2011. Each complication was assigned a severity weight ranging from 0.11 for the least severe complication to 1.00 for postoperative death, and PMI was derived. Contribution to total burden by each complication grade was used to generate a severity profile ("spectrogram") for PD. Associations with PMI were determined by regression analysis. RESULTS:ACS-NSQIP complications occurred in 528 cases (33.2%). The non-risk-adjusted PMI was 0.115 (SD = 0.023) for all centers and 0.113 (SD = 0.005) for the 7 centers that contributed at least 100 cases. Grade 2 complications were predominant in frequency, and the most common complication was postoperative bleeding/transfusion. Frequency and burden of complications differed markedly. For instance, severe complications (grades 4/5/6) accounted for only about 20% of complications but for more than 40% of the burden of complications. Organ space infection had the highest burden of any complication. The average burden in cases in which a complication actually occurred was 0.346. CONCLUSIONS:This study develops a quantitative non-risk-adjusted benchmark for postoperative morbidity of PD. The method quantifies the burden of types and grades of postoperative complications and should prove useful in identifying areas that require quality improvement.