Surgical resection remains an important component in the treatment of pancreatic cancer. Unfortunately, only approximately 10% to 20% of patients are eligible for surgical resection at presentation. Many patients with pancreatic cancer, especially cancer of the distal pancreas, present at a late stage and their disease is unresectable secondary to locally advanced or metastatic disease. The surgical approach depends on the location of the tumor: tumors of the pancreatic head require pancreaticoduodenectomy (PD) or pylorus-preserving pancreaticoduodenectomy (PPPD), whereas tumors located in the body or tail require distal pancreatectomy and splenectomy (DP). In hopeful attempts to extend the limits of resection and improve survival, many centers have performed more aggressive resections, including extended lymphadenectomy and vascular resections.
Pancreaticoduodenectomy
The standard of care for resection of cancers of the pancreatic head, neck, and uncinate process is PD ( Fig. 1 ). The development of the PD procedure has been credited in the United States to Allen O. Whipple. In 1935, he described a 2-stage operation; the first operation involves a cholecystogastrostomy followed 3 to 4 weeks later by resection of the stomach, pancreas, and duodenum. Dr. Whipple later described a 1-stage PD in 1941. Experience with pancreatic resections began to grow, but morbidity and mortality rates at that time remained prohibitive.
Since the time of Dr. Whipple, there have been significant improvements in anesthesia, critical care, and surgical technique, which have led to decreased mortality rates and improved survival. In a recent large single-institution study, Winter and colleagues evaluated the outcomes of 1175 patients who underwent PD for ductal adenocarcinoma from 1970 to 2006. Median blood loss was 800 mL and operating time was 380 minutes. Perioperative mortality rate was 2%, morbidity was 38%, length of stay was 9 days, and median survival time was 18 months. They reported median survival in the 1970s to be 8 months, in the 1980s to be 14 months, in the 1990s to be 17 months, and in the 2000s to be 19 months. McPhee and colleagues performed a national database study and also reported a low mortality rate of 5.5% in 2003, which had significantly decreased from 8.2% in 1998. Additional studies have reported morbidity rates of 42% to 47%; the most common complications being delayed gastric emptying, pancreatic fistula, and wound infections. Pancreatic fistulas have traditionally been the Achilles’ heel of PD, being a fairly common and potentially life-threatening complication. Pancreatic fistula rates vary widely in the literature, in part secondary to the lack of a universal definition. The international study group on pancreatic fistula proposed a standard definition in 2005. The group defined a pancreatic fistula as any volume of drain fluid after postoperative day 3, with amylase content greater than 3 times serum. The group further classified the fistulas by grade according to clinical severity. Grade A fistulas require little change in management and are considered “transient fistulas.” Grade B fistulas require a change in management, including parenteral nutrition, antibiotics, or somatostatin analogues. Grade C fistulas require a major change in management, such as additional drainage procedures, and often lead to an extended hospital stay. Hopefully, with the use of a standard definition, accurate comparisons between centers can be made, potentially leading to improvements in the rates or management of pancreatic fistulas.
In their study of 1175 patients, Winter and coworkers reported the significant factors influencing survival to be tumor diameter, resection margin, lymph node status, and histologic grade. Additional studies have confirmed the importance of tumor size, lymph node status, histologic grade, and negative surgical margins. Wagner and colleagues found a significant difference between patients who received R1 resections (median survival, 11.5 months) and R0 resections (median survival, 20.1 months) and found R0 resection to be a significant independent predictor of long-term survival. However, not all studies found resection margin to be a significant factor for survival. This discrepancy may be, in part, because of the lack of a universal consensus on pathologic examination of PD specimens. When margins are examined, the pancreatic resection margin, bile duct margin, and stomach/duodenal margin are evaluated. The difficulty lies in assessing the soft tissue margin that abuts the superior mesenteric vessels. This soft tissue margin is the most common margin involved with the tumor. Some centers circumferentially evaluate an anterior, posterior, and medial margin, whereas other centers examine the medial margin, also known as the retroperitoneal or uncinate margin. Wide variation (16%–85%) in rates of R1 resections has been reported. Esposito and coworkers, when comparing margin results using new, more thorough standardized pathologic reporting, found that rates of R1 resections increased from 14% to 76%. The need for a standardized method of examining PD specimens and evaluating margins is imperative because the rates of local recurrence are high, and these recurrences often are located at the superior mesenteric vessels. Improving the assessment of margins will lead to more accurate assessment of outcomes and surgical and adjuvant treatments. Operative blood loss has also been reported as a factor influencing survival. Kazanjian and colleagues described blood loss of greater than 400 mL to be a predictor of poor survival, and Sohn and colleagues reported an estimated blood loss of less than 750 mL to be a favorable prognostic indicator.
Additionally, multiple studies have supported the regionalization of complex procedures such as the PD. Luft and coworkers initially supported the early regionalization for procedures such as open heart surgery, major vascular surgery, and total hip replacement. In 1999, Birkmeyer and colleagues published a study based on the Medicare claims database showing that hospital volumes for PD were strongly related to perioperative mortality and long-term survival. Subsequent studies looked at the effect of surgeon volume for pancreatic resections and found that higher surgeon procedure volumes correlated with lower mortality rates. These studies suggest that regionalization of pancreatic resections may lead to decreased mortality and improved survival. The PD has become a procedure that can be performed with low mortality and acceptable morbidity rates in large centers and should be performed with the intent of achieving negative margins and minimizing blood loss.
Pancreaticoduodenectomy Versus Pylorus-Preserving Pancreaticoduodenectomy
Dr. Watson first described the PPPD in 1944. However, interest in this procedure was stimulated by Traverso and Longmire after their report of 18 patients who underwent PPPD for chronic pancreatitis and early periampullary cancers, with improved gastrointestinal function. There has since been long debate over the benefits of PPPD versus PD. Proponents of PPPD claim that it allows for fewer cases of dumping syndrome, less blood loss, and shorter operating times. Proponents of PD argue that PPPD is associated with more delayed gastric emptying and question its effectiveness as an oncologic procedure. In a prospective, randomized comparison by Lin and Lin, the same surgeon performed 16 PPPD and 15 PD. Both PPPD and PD were associated with low morbidity and mortality rates with no significant differences between the 2 groups. Patients also had similar operating times and blood loss. Tran and coworkers subsequently published a prospective, randomized, multicenter study out of the Netherlands. In this study, 170 consecutive patients were randomly assigned to undergo either PPPD or PD for suspected pancreatic or periampullary cancer. Both groups were similar with regard to operating times, blood loss, and length of stay. Additionally, rates of delayed gastric emptying and positive resection margins were equivalent. When patients with confirmed pancreatic or periampullary adenocarcinoma were analyzed, patients who were treated with PPPD had a median disease-free survival of 15 months, whereas patients who were treated with PD had a median disease-free survival of 14 months ( P = 0.80). The conclusion of the study was that PD and PPPD were both effective in the treatment of pancreatic cancer. A recent Cochrane review confirmed these findings. The Cochrane review included all randomized, controlled trials, regardless of language. Six studies between March 2006 and January 2011 with a total of 465 patients were included. The study found no differences in morbidity, mortality, and survival for patients receiving PPPD or PD. Despite the theorized pros and cons of preserving the duodenum, the bulk of the literature shows no difference in outcomes after PPPD compared with PD.
Pancreaticoduodenectomy Versus Pylorus-Preserving Pancreaticoduodenectomy
Dr. Watson first described the PPPD in 1944. However, interest in this procedure was stimulated by Traverso and Longmire after their report of 18 patients who underwent PPPD for chronic pancreatitis and early periampullary cancers, with improved gastrointestinal function. There has since been long debate over the benefits of PPPD versus PD. Proponents of PPPD claim that it allows for fewer cases of dumping syndrome, less blood loss, and shorter operating times. Proponents of PD argue that PPPD is associated with more delayed gastric emptying and question its effectiveness as an oncologic procedure. In a prospective, randomized comparison by Lin and Lin, the same surgeon performed 16 PPPD and 15 PD. Both PPPD and PD were associated with low morbidity and mortality rates with no significant differences between the 2 groups. Patients also had similar operating times and blood loss. Tran and coworkers subsequently published a prospective, randomized, multicenter study out of the Netherlands. In this study, 170 consecutive patients were randomly assigned to undergo either PPPD or PD for suspected pancreatic or periampullary cancer. Both groups were similar with regard to operating times, blood loss, and length of stay. Additionally, rates of delayed gastric emptying and positive resection margins were equivalent. When patients with confirmed pancreatic or periampullary adenocarcinoma were analyzed, patients who were treated with PPPD had a median disease-free survival of 15 months, whereas patients who were treated with PD had a median disease-free survival of 14 months ( P = 0.80). The conclusion of the study was that PD and PPPD were both effective in the treatment of pancreatic cancer. A recent Cochrane review confirmed these findings. The Cochrane review included all randomized, controlled trials, regardless of language. Six studies between March 2006 and January 2011 with a total of 465 patients were included. The study found no differences in morbidity, mortality, and survival for patients receiving PPPD or PD. Despite the theorized pros and cons of preserving the duodenum, the bulk of the literature shows no difference in outcomes after PPPD compared with PD.
Distal Pancreatectomy
The standard surgical therapy for pancreatic cancers of the body and tail of the pancreas is distal pancreatectomy and splenectomy. The first distal pancreatic resection was first reported in 1884 by Bilroth. In the years following, pancreatic resection was associated with high morbidity and mortality, and the wisdom of pancreatic resection was questioned. In recent years, however, there has been tremendous improvement in outcomes after DP. Lillemoe and colleagues retrospectively studied 235 patients who underwent DP for a variety of benign and malignant conditions between 1994 and 1997. Mortality rate was 0.9%, operating time was 4.3 hours, and blood loss was 450 mL. Morbidity after DP was 31% and the rate of pancreatic fistula was 5%. Additional studies have reported low mortality rates and morbidity rates ranging from 20% to 50%.
In the study by Lillemoe et al, 18% of the patients who underwent DP had adenocarcinoma of the pancreas diagnosed. Long-term survival rates in this subset of patients were not reported. Sperti and coworkers reported on 24 patients who underwent DP for adenocarcinoma of the body and tail of the pancreas with no adjuvant radiotherapy or chemotherapy. Morbidity reported in the study was 25% and mortality 8%. Five-year survival rate for these patients was 12.5%, similar to rates reported for survival after PD for pancreatic head cancer. The authors recommended an aggressive surgical approach, supporting the resection of adjacent organs if necessary to achieve complete tumor resection.
Extended Resection
In an attempt to extend the limits of resectability for pancreatic tumors, Fortner in the 1970s described extended PD resections, which included total pancreatectomy, extended lymph node resection, and combinations of portal vein, arterial, colon, and gastric resection and reconstruction. These procedures greatly increased resectability rates but resulted in high postoperative morbidity and mortality rates. Since then, there has been reluctance to perform such extensive surgery. However, with improvements in anesthesia, surgical technique, and perioperative care, many centers have been exploring the utility of extended pancreatic resection.
Extended Lymphadenectomy
Regional lymph nodes typically resected with the PD specimen include anterior and posterior pancreaticoduodenal nodes, as shown as the first group of nodes in Fig. 2 . Three prospective, randomized, controlled trials have been published that address the utility of performing PD in conjunction with extended lymphadenectomy. The first study, by Pedrazzoli and coworkers, is a multicenter study that randomly assigned 40 patients to standard PD and 41 patients to PD with extended lymphadenectomy. Extended lymphadenectomy included removal of nodes from the liver hilum and along the aorta from the diaphragm to the inferior mesenteric artery and laterally to both renal hilum. Circumferential clearance of the celiac trunk and superior mesenteric artery was also included. Mean lymph nodes retrieved were significantly higher in the extended group (19.8 vs 13.3), and morbidity and mortality rates were similar. Overall survival did not differ between the 2 groups; however, when examining node-positive patients only, the authors found a longer survival rate in those patients undergoing PD with extended lymphadenectomy.
Related posts:
Modern Management of Pancreatic Disease
Pancreatic Cystic Neoplasms: Diagnosis and Management
Pancreatic Cancer: Medical Management (Novel Chemotherapeutics)
Endoscopic Management of Pancreatic Pseudocysts
Pancreatic Cancer: Radiologic Imaging
The Role of Endoscopic Ultrasonography in the Diagnosis and Management of Pancreatic Cancer
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
