Abstract
Background:
Safety of pancreaticoduodenectomy has improved significantly in the past 3 decades. Current inpatient and 30-d mortality rates are low. However, incidence and causes of 90-d and 1-y mortality are poorly defined and largely unexplored.
Methods:
All patients who had pancreaticoduodenectomy between 2007 and 2016 were included in this single institution, retrospective cohort study. Distributions of pancreaticoduodenectomy-specific morbidity and cause-specific mortality were compared between early (within 90 d) and late (91–365 d) postoperative recovery periods.
Results:
A total of 551 pancreaticoduodenectomies were performed during the study period. Of these, 6 (1.1%), 20 (3.6%), and 91 (16.5%) patients died within 30, 90, and 365 d after pancreaticoduodenectomy, respectively. Causes of early and late mortality varied significantly (all P ≤ 0.032). The most common cause of death within 90 d was due to multisystem organ failure from sepsis or aspiration in 9 (45%) patients, followed by post-pancreatectomy hemorrhage in 5 (25%) patients, and cardiopulmonary arrest from myocardial infarction or pulmonary embolus in 3 (15%) patients. In contrast, recurrent cancer was the most common cause of death in 46 (65%) patients during the late postoperative period between 91 and 365 d. Mortality from failure to thrive and debility was similar between early and late postoperative periods (15% versus 19.7%, P = 0.76).
Conclusions:
Most quality improvement initiatives in patients selected for pancreaticoduodenectomy have focused on reduction of technical complications and improvement of early postoperative mortality. Further reduction in postoperative mortality after pancreaticoduodenectomy can be achieved by improving patient selection, mitigating postoperative malnutrition, and optimizing preoperative cancer staging and management strategies.
Keywords: Pancreaticoduodenectomy, Whipple, Mortality, Complications, Outcomes, Quality improvement
Introduction
Mortality after pancreaticoduodenectomy has decreased considerably from as high as 25%−30% in the 1970–1980s to less than 2%−4% in the modern era.1–4 The decline in short-term, perioperative mortality is observed across different indications for resection and variations of surgical technique.3,5,6 Improvements in operative and anesthetic techniques, regionalization to high-volume centers, implementation of standardized recovery pathways, and better understanding and management of common complications have all contributed to the dramatic decline in postoperative mortality.7–10
While the decrease in postoperative mortality is encouraging, the definition of mortality used in many studies is limited to either in-hospital or within 30-d from the index operation.3,4 Few studies explore surgery related mortality beyond immediate 30-day postoperative period; but in general, these studies report higher rates of patient death than 30-d or inpatient mortality.11–13
The primary objective of this study was to compare differences between early (defined as within 90 d of operation) and late (defined as between 91–365 d after operation) patient-specific mortality after pancreaticoduodenectomy. The secondary objective was to identify potentially modifiable factors that result in patient death.
Materials and methods
Patient population
This single institution, retrospective study included all adult patients (≥18 y of age) who underwent standard or pylorus-preserving pancreaticoduodenectomy for benign or malignant pathology between 2007 and 2016 at the University of Virginia Health System. The study was reviewed and approved by the University of Virginia Institutional Review Board for Health Sciences Research (protocol #19754). Abstracted demographic information included age, sex, and race/ethnicity. Pathologic diagnosis was defined based on postresection pathology and categorized as pancreatic adenocarcinoma (PDAC), other malignancies (e.g., duodenal cancer, periampullary cancer, neuroendocrine carcinoma, and others), premalignant neoplasms (e.g., intraductal papillary mucinous neoplasms without malignant transformation to carcinoma, premalignant duodenal adenomas, and others), and benign disease (e.g., chronic pancreatitis, serous cystadenoma, and others).
Outcome definitions
The primary outcome was patient-specific mortality categorized as early if it occurred between 0 and 90 d after pancreaticoduodenectomy and late if it occurred between 91 and 365 d after pancreaticoduodenectomy. Causes of mortality were categorized as (1) multisystem organ failure (MSOF) secondary to sepsis, aspiration, or hemorrhage; (2) cardiac arrest secondary to myocardial infarction or pulmonary embolus; (3) debility with failure to thrive; and (4) cancer recurrence. Secondary outcomes included causes of early and late morbidity, including technical complications such as pancreatic fistula, delayed gastric emptying (DGE), bile leak, gastrointestinal anastomotic leak, hemorrhage, and reoperations. Pancreatic fistula and DGE were categorized based on standardized International Study Group of Pancreatic Surgery definitions.14,15 Bile leak was defined using the standardized guidelines from the International Study Group of Liver Surgery.9 Reoperations related to pancreaticoduodenectomy were abstracted from clinical records. Indications for reoperation were categorized as hemorrhage, gastrointestinal complications (i.e., ischemia, anastomotic leak, perforation, obstruction), and abdominal wall incisional complications (i.e., dehiscence, evisceration, hernia).
Data analysis
Categorical variables were reported as percentages and compared using either chi-squared or Fisher’s exact tests, as appropriate. Continuous data were summarized as median with interquartile range and compared with Wilcoxon rank-sum test. Statistical analyses were performed using Stata software, version 14.2 (StataCorp LP, College Station, TX), and P values less than 0.05 were considered significant.
Results
Patient demographics and clinical factors
A total of 551 patients, median age 65 y (interquartile range 57–73), underwent pancreaticoduodenectomy during the study period (Table 1). The majority of patients were men (n = 300, 54.4%) and white (n = 483, 87.7%). The most common indication for pancreaticoduodenectomy was pancreatic adenocarcinoma (n = 226, 41.0%), followed by other malignancy (n = 182, 33.0%). Proportions of postoperative complications are summarized in Table 1.
Table 1 –
Patient demographics and clinical factors of total study population.
| Variable | n = 551 |
|---|---|
| Age, median (IQR) | 65 (57–73) |
| Sex, n (%) | |
| Male | 300 (54.4) |
| Female | 251 (45.6) |
| Race, n (%) | |
| White | 483 (87.7) |
| Black | 54 (9.8) |
| Asian | 5 (0.9) |
| Unknown | 9 (1.6) |
| Diagnosis, n (%) | |
| Benign | 81 (14.7) |
| Premalignant | 62 (11.3) |
| PDAC | 226 (41.0) |
| Other malignancy | 182 (33.0) |
| Complications, n (%) | |
| Pancreatic fistula | 98 (17.8) |
| BL | 22 (4.0) |
| Grade B | 57 (10.3) |
| Grade C | 19 (3.4) |
| Delayed gastric emptying | 103 (18.7) |
| Grade A | 39 (7.1) |
| Grade B | 22 (4.0) |
| Grade C | 42 (7.6) |
| Bile leak | 15 (2.7) |
| Grade A | 2 (0.4) |
| Grade B | 8 (1.5) |
| Grade C | 5 (0.9) |
| Gastrointestinal anastomotic leak | 5 (0.9) |
| Reoperations, n (%) | 46 (8.3) |
| Hemorrhage* | 6 (1.1) |
| Gastrointestinal complications† | 27 (4.9) |
| Incisional complications | 16 (2.9) |
Excludes 12 patients in the early period and 1 patient in the late period who had endovascular management only.
Gastrointestinal complications required multiple reoperations in 3 patients.
Early and late postoperative mortality
Of the 551 patients, 91 (16.5%) died within 1 y after pancreaticoduodenectomy (Table 1). Six patients (1.1%) died within 30 d of the index operation. Proportions of early and late mortality were 20 patients (3.6%) in the early (0–90 d postoperative period) and 71 patients (12.9%) in the late (91–365 d) postoperative period. Comparisons between the two groups are summarized in Table 2.
Table 2 –
Demographics and clinical factors of patients who died within 1 y.
| Early (0–90 d) n = 20 |
Late (91–365 d) n = 71 |
P value | |
|---|---|---|---|
| Age, median (IQR) | 71 (63–75) | 65 (57–77) | 0.31 |
| Sex, n (%) | 0.23 | ||
| Male | 14 (70) | 39 (54.9) | |
| Female | 6 (30) | 32 (45.1) | |
| Race, n (%) | 0.75 | ||
| White | 19 (95) | 63 (88.7) | |
| Black | 1 (5) | 7 (9.9) | |
| Unknown | 0 | 1 (1.4) | |
| Diagnosis, n (%) | 0.20 | ||
| Benign | 3 (15) | 3 (4.2) | |
| Premalignant | 2 (10) | 4 (5.6) | |
| PDAC | 9 (45) | 44 (62) | |
| Other malignancy | 6 (30) | 20 (28.2) |
One-year mortality was lower in patients with benign or premalignant diagnosis compared to patients with PDAC or other malignancies (8.4% versus 19.4%, P < 0.001). One-year mortality was highest in patients with PDAC (23.5%). Of the 226 patients with PDAC, 174 (77%) received neoadjuvant therapy and 180 (79.6%) received adjuvant therapy. Adjuvant therapy was initiated in 2 of 9 (22.2%) PDAC patients who died in the early period and in 31 of 44 (70.5%) PDAC patients who died in the late period.
Causes of death are summarized in Table 3. MSOF and cardiac arrest were the most common causes of early mortality and were significantly more common in the early postoperative period (all P ≤ 0.032). MSOF resulting from sepsis (n = 6, 30%), aspiration (n = 3, 15%), or hemorrhage (n = 5, 25%) were the most common cause of death within 90 d of pancreaticoduodenectomy. In contrast, the two most common causes of death between 91 and 365 d after operation were cancer recurrence (n = 46, 64.8%) and chronic and irreversible debility (n = 14, 19.7%). When stratified by diagnosis, the most common causes of death within 1 y in patients with benign or premalignant diseases were MSOF from sepsis (n = 5, 41.7%) and debility (n = 5, 41.7%). In contrast, the most common causes of death in patients with PDAC and other malignancies were cancer recurrence (n = 46, 58.2%) and debility (n = 12, 15.2%).
Table 3 –
Causes of 1-y mortality.
| Early (0–90 d) n = 20 |
Late (91–365 d) n = 71* |
P value | |
|---|---|---|---|
| MSOF from sepsis | 6 (30) | 4 (5.6) | 0.007 |
| MSOF from aspiration | 3 (15) | 1 (1.4) | 0.032 |
| MSOF from hemorrhage | 5 (25) | 0 | <0.001 |
| Cardiac arrest due to MI or PE | 3 (15) | 0 | 0.009 |
| New primary cancer | 0 | 2 (2.8) | 1.00 |
| Cancer recurrence | 0 | 46 (64.8) | <0.001 |
| Debility/FTT | 3 (15) | 14 (19.7) | 0.76 |
Data reported as n (%).
FTT = failure to thrive; MSOF = multisystem organ failure; MI = myocardial infarction; PE = pulmonary embolism.
The cause of death could not be determined in 4 patients who died in the late period.
Early and late postoperative morbidity
Pancreatic fistula was present in 98 (17.8%) patients. Of these, 22 (4%) were biochemical leaks of no clinical significance. Remaining 76 (13.7%) patients had clinically significant pancreatic fistulae: 57 (10.3%) grade B and 19 (3.4%) grade C (Table 1). DGE occurred in 103 (18.7%) patients: 39 (7.1%) grade A, 22 (4%) grade B, and 42 (7.6%) grade C. Bile leaks and gastrointestinal anastomotic leaks occurred in 15 (2.7%) and 5 (0.9%) patients, respectively. All cases of pancreatic fistula, DGE, bile leak, and gastrointestinal anastomotic leak occurred in the early postoperative period with the exception of a single patient who had a new diagnosis of biliary leak >90 d after pancreaticoduodenectomy. Reoperations were performed in 28 (5%) patients in the early postoperative period and 18 (3.3%) patients in the late postoperative period. Indication for reoperation in the early period included hemorrhage, anastomotic leak, bowel necrosis, and wound dehiscence; indications for reoperations in the late period included small bowel obstructions and abdominal wall hernias. Categorized indications for reoperations are summarized in Table 1. A total of 13 patients required endovascular management of bleeding complications; one of these patients required endovascular management after the initial 90 d.
Discussion
In this single institution study, 30-d mortality after pancreaticoduodenectomy is 1.1%. Early mortality within 90 d of resection is 3.6% and increases to 16.5% by the end of first year after pancreaticoduodenectomy. As we hypothesized, cause of death between the early and late post-pancreaticoduodenectomy patients differs significantly. While deaths in the early period are predominantly due to multisystem organ failure from postoperative complications, the majority of deaths in the late period are due to prolonged postoperative malnutrition and debility or recurrence of cancer.
Acute postoperative technical complications such as pancreatic fistulae, bile leaks, DGE, and hemorrhage reported in this study are comparable with published results.16–20 Reoperations were required in 46 patients, of which 28 were in the early period and 18 were in the late period. Not surprisingly, the indications for reoperations between the early and late postoperative periods differ. While reoperations in the early period were performed to manage acute abdominal wall, gastrointestinal, or vascular complications, reoperations in the late period were performed to address bowel obstruction or hernia. Currently published estimates of reoperations, which are mostly limited to the early postoperative period, range from 5% to 9% and include indications similar to those for the early reoperations presented in this study.21,22
Variations in causes of death between the early and late time periods suggest that different strategies are necessary to reduce long-term mortality beyond the initial recovery period. As technical complications resulting in MSOF from sepsis, aspiration, or hemorrhage were the most common causes of death in the early period, meticulous operative technique and standardized postoperative care continue to be of importance in mitigating early mortality. Because accumulation of complications can lead to failure to rescue,23 avoidance as well as prompt identification and management of postoperative complications will continue to remain a key strategy to further reduce early postoperative mortality after pancreaticoduodenectomy.
Although mortality in the early period can be reduced by further improvements in operative technique and patient management, patient selection is a critical component in decreasing late mortality after pancreaticoduodenectomy. Most patients with cancer recurrence after pancreaticoduodenectomy had pancreatic ductal adenocarcinoma that recurred with metastatic disease. Up to 75% of patients who complete curative resection for pancreatic cancer develop and die from distant metastases.24 Improvement in 1-y survival after pancreaticoduodenectomy for cancer is dependent on advances in preoperative detection and staging, as well as expansion of successful neoadjuvant treatment strategies for patients with pancreatobiliary malignancies.
Patient selection as it relates to preoperative functional performance is also paramount. Debility and failure to thrive was the second most common cause of patient mortality in the late postoperative period after pancreaticoduodenectomy. Preoperative rehabilitation programs aimed at assessment and treatment of frailty before resection have been shown to reduce complications and mortality after select gastrointestinal operations.25–27 The effectiveness of prehabilitation in patients who need pancreaticoduodenectomy requires further study.
Retrospective single institution design and data collection are limitations of this study. However, granular data analysis of patient-specific cause of death and complications provide an important contribution to understanding of mortality after pancreaticoduodenectomy.
Conclusions
Improvements in operative technique and postoperative management have significantly reduced early mortality after pancreaticoduodenectomy. Judicious patient selection, timely management of malnutrition, and advances in cancer staging and treatment will help reduce long-term mortality after pancreaticoduodenectomy.
Acknowledgment
This study was supported in part by funding provided by the Institutional National Research Service Award T32 CA 163177 from the National Cancer Institute to A.N.M. and Loan Repayment Program Award L30 CA220861 from the National Cancer Institute to V.M.Z.
Footnotes
Presented in part at the 13th Annual Academic Surgical Congress (ASC) January 30-February 1, 2018, in Jacksonville, Florida.
Disclosure
None of the authors have any personal, financial, or other conflicts of interest to disclose.
REFERENCES
- 1.Crist DW, Sitzmann JV, Cameron JL. Improved hospital morbidity, mortality, and survival after the Whipple procedure. Ann Surg. 1987;206:358–365. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Grace PA, Pitt HA, Tompkins RK, DenBesten L, Longmire WP. Decreased morbidity and mortality after pancreatoduodenectomy. Am J Surg. 1986;151:141–149. [DOI] [PubMed] [Google Scholar]
- 3.Nassour I, Wang SC, Christie A, et al. Minimally invasive versus open pancreaticoduodenectomy: a propensity-matched study from a national cohort of patients. Ann Surg. 2018;268:151–157. [DOI] [PubMed] [Google Scholar]
- 4.Gleeson EM, Shaikh MF, Shewokis PA, et al. WHipple-ABACUS, a simple, validated risk score for 30-day mortality after pancreaticoduodenectomy developed using the ACS-NSQIP database. Surgery. 2016;160:1279–1287. [DOI] [PubMed] [Google Scholar]
- 5.Hüttner FJ, Fitzmaurice C, Schwarzer G, et al. Pylorus-preserving pancreaticoduodenectomy (pp Whipple) versus pancreaticoduodenectomy (classic Whipple) for surgical treatment of periampullary and pancreatic carcinoma. Cochrane Database Syst Rev. 2016;2:CD006053. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Newhook TE, LaPar DJ, Lindberg JM, Bauer TW, Adams RB, Zaydfudim VM. Morbidity and mortality of pancreaticoduodenectomy for benign and premalignant pancreatic neoplasms. J Gastrointest Surg. 2015;19:1072–1077. [DOI] [PubMed] [Google Scholar]
- 7.McPhee JT, Hill JS, Whalen GF, et al. Perioperative mortality for pancreatectomy: a national perspective. Ann Surg. 2007;246:246–253. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Kennedy EP, Rosato EL, Sauter PK, et al. Initiation of a critical pathway for pancreaticoduodenectomy at an academic institutionethe first step in multidisciplinary team building. J Am Coll Surg. 2007;204:917–923 [discussion 923–924]. [DOI] [PubMed] [Google Scholar]
- 9.Koch M, Garden OJ, Padbury R, et al. Bile leakage after hepatobiliary and pancreatic surgery: a definition and grading of severity by the International study group of liver surgery. Surgery. 2011;149:680–688. [DOI] [PubMed] [Google Scholar]
- 10.Lee GC, Fong ZV, Ferrone CR, et al. High performing Whipple patients: factors associated with short length of stay after open pancreaticoduodenectomy. J Gastrointest Surg. 2014;18:1760–1769. [DOI] [PubMed] [Google Scholar]
- 11.House MG, Fong Y, Arnaoutakis DJ, et al. Preoperative predictors for complications after pancreaticoduodenectomy: impact of BMI and body fat distribution. J Gastrointest Surg. 2008;12:270–278. [DOI] [PubMed] [Google Scholar]
- 12.Venkat R, Puhan MA, Schulick RD, et al. Predicting the risk of perioperative mortality in patients undergoing pancreaticoduodenectomy: a novel scoring system. Arch Surg. 2011;146:1277–1284. [DOI] [PubMed] [Google Scholar]
- 13.Swanson RS, Pezzi CM, Mallin K, Loomis AM, Winchester DP. The 90-day mortality after pancreatectomy for cancer is double the 30-day mortality: more than 20,000 resections from the national cancer data base. Ann Surg Oncol. 2014;21:4059–4067. [DOI] [PubMed] [Google Scholar]
- 14.Wente MN, Bassi C, Dervenis C, et al. Delayed gastric emptying (DGE) after pancreatic surgery: a suggested definition by the International Study Group of Pancreatic Surgery (ISGPS). Surgery. 2007;142:761–768. [DOI] [PubMed] [Google Scholar]
- 15.Bassi C, Marchegiani G, Dervenis C, et al. The 2016 update of the International Study Group (ISGPS) definition and grading of postoperative pancreatic fistula: 11 years after. Surgery. 2017;161:584–591. [DOI] [PubMed] [Google Scholar]
- 16.Burkhart RA, Relles D, Pineda DM, et al. Defining treatment and outcomes of hepaticojejunostomy failure following pancreaticoduodenectomy. J Gastrointest Surg. 2013;17:451–460. [DOI] [PubMed] [Google Scholar]
- 17.Harnoss JC, Ulrich AB, Harnoss JM, Diener MK, Büchler MW, Welsch T. Use and results of consensus definitions in pancreatic surgery: a systematic review. Surgery. 2014;155:47–57. [DOI] [PubMed] [Google Scholar]
- 18.Park JS, Hwang HK, Kim JK, et al. Clinical validation and risk factors for delayed gastric emptying based on the International Study Group of Pancreatic Surgery (ISGPS) classification. Surgery. 2009;146:882–887. [DOI] [PubMed] [Google Scholar]
- 19.Welsch T, Borm M, Degrate L, Hinz U, Büchler MW, Wente MN. Evaluation of the International Study Group of Pancreatic Surgery definition of delayed gastric emptying after pancreatoduodenectomy in a high-volume centre. Br J Surg. 2010;97:1043–1050. [DOI] [PubMed] [Google Scholar]
- 20.Malleo G, Crippa S, Butturini G, et al. Delayed gastric emptying after pylorus-preserving pancreaticoduodenectomy: validation of International Study Group of Pancreatic Surgery classification and analysis of risk factors. HPB (Oxford). 2010;12:610–618. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.McMillan MT, Allegrini V, Asbun HJ, et al. Incorporation of procedure-specific risk into the ACS-NSQIP surgical risk calculator improves the prediction of morbidity and mortality after pancreatoduodenectomy. Ann Surg. 2017;265:978–986. [DOI] [PubMed] [Google Scholar]
- 22.Lyu HG, Sharma G, Brovman E, et al. Risk factors of reoperation after pancreatic resection. Dig Dis Sci. 2017;62:1666–1675. [DOI] [PubMed] [Google Scholar]
- 23.Varley PR, Geller DA, Tsung A. Factors influencing failure to rescue after pancreaticoduodenectomy: a national surgical quality improvement project perspective. J Surg Res. 2017;214:131–139. [DOI] [PubMed] [Google Scholar]
- 24.Hishinuma S, Ogata Y, Tomikawa M, Ozawa I, Hirabayashi K, Igarashi S. Patterns of recurrence after curative resection of pancreatic cancer, based on autopsy findings. J Gastrointest Surg. 2006;10:511–518. [DOI] [PubMed] [Google Scholar]
- 25.Barberan-Garcia A, Ubré M, Roca J, et al. Personalised prehabilitation in high-risk patients undergoing elective major abdominal surgery: a randomized blinded controlled trial. Ann Surg. 2018;267:50–56. [DOI] [PubMed] [Google Scholar]
- 26.Mazzola M, Bertoglio C, Boniardi M, et al. Frailty in major oncologic surgery of upper gastrointestinal tract: how to improve postoperative outcomes. Eur J Surg Oncol. 2017;43:1566–1571. [DOI] [PubMed] [Google Scholar]
- 27.Bolshinsky V, Li MH-G, Ismail H, Burbury K, Riedel B, Heriot A. Multimodal prehabilitation programs as a bundle of care in gastrointestinal cancer surgery: a systematic review. Dis Colon Rectum. 2018;61:124–138. [DOI] [PubMed] [Google Scholar]