Overview of surgery in the treatment of exocrine pancreatic cancer and prognosis

INTRODUCTION — Approximately 64,000 patients are diagnosed with exocrine pancreatic cancer annually in the United States [1]. Global and country-specific incidence, and mortality rates, are available from the World Health Organization Globocan database. (See "Epidemiology and nonfamilial risk factors for exocrine pancreatic cancer".)

For patients with exocrine pancreatic cancer, a majority (85 percent) present with adenocarcinomas arising from the ductal epithelium. Surgical resection is the only potentially curative treatment.

An overview of the surgical management of cancers of the exocrine pancreas will be reviewed here. The clinical manifestations, diagnosis, and surgical and nonsurgical therapy are discussed separately. (See "Clinical manifestations, diagnosis, and staging of exocrine pancreatic cancer" and "Surgical resection of lesions of the head of the pancreas" and "Surgical resection of lesions of the body and tail of the pancreas" and "Treatment for potentially resectable exocrine pancreatic cancer".)

CANDIDATES FOR RESECTION — Surgical resection is the only potentially curative treatment. Unfortunately, because of the late presentation of the disease, only 15 to 20 percent of patients are candidates for pancreatectomy. The prognosis of pancreatic cancer is poor even in those with potentially resectable disease, and despite progress in surgical techniques and adjuvant therapy, improvement in outcomes over time has lagged behind that of other common cancers. (See 'Outcomes of pancreaticoduodenectomy' below and "Treatment for potentially resectable exocrine pancreatic cancer".)

Notably, in a study of the National Cancer Database, among those who were candidates for resection of pancreatic cancer, 38 percent were not offered surgery [2]. Underutilization of surgery for early-stage pancreatic cancer may be related to socioeconomic factors as well as physician pessimism regarding the prognosis of the disease.

Preoperative imaging evaluation determines candidacy for resection. Radiographic staging and the role of endoscopic ultrasound (EUS) are discussed in detail elsewhere. (See "Clinical manifestations, diagnosis, and staging of exocrine pancreatic cancer", section on 'Imaging studies' and "Endoscopic ultrasound in the staging of exocrine pancreatic cancer".)

Disease that is limited to the pancreas is most likely to be cured by resection, although approximately 30 percent of individuals who undergo a complete (R0) resection of a pancreatic cancer with limited nodal involvement may also be long-term survivors (figure 1) [3]. Tumors with limited involvement of the major peripancreatic vessels such as the superior mesenteric vein, portal vein, superior mesenteric artery, or hepatic artery may be technically resectable. However, the impact of more aggressive resections (particularly arterial resection) on long-term prognosis is controversial. (See 'Vascular resection' below and "Surgical resection of lesions of the head of the pancreas", section on 'Vascular evaluation'.)

Some cases are considered "borderline" resectable, although the definition is variable. Some reserve the term "borderline resectable" for cases where there is focal (less than one-half of the circumference) tumor abutment of the visceral arteries or short-segment occlusion of the superior mesenteric vein (SMV) or SMV/portal vein confluence. Others suggest that venous narrowing without occlusion should be included in the definition of borderline resectable disease. Encasement (more than one-half of the vein circumference) or occlusion of the SMV or the SMV-portal vein confluence was previously considered unresectable. However, many centers have demonstrated the feasibility of SMV reconstruction, and provided venous reconstruction is possible, this is now considered by many to represent borderline resectable disease. Although vein resection adds a measure of complexity to pancreaticoduodenectomy, an experienced surgical team in a high-volume center can safely perform such a resection when it is required. Several groups have issued guidelines to define borderline resectable disease based upon imaging studies. This subject is discussed in detail elsewhere. (See "Clinical manifestations, diagnosis, and staging of exocrine pancreatic cancer", section on 'Definitions of unresectable and borderline resectable disease' and 'Vascular resection' below.)

Patients who have borderline resectable or locally advanced unresectable pancreatic cancer, but no metastatic disease, are potential candidates for downstaging with neoadjuvant therapy and should be referred for medical oncology and radiation therapy consultation. There have been reports that when surgery was performed after neoadjuvant chemotherapy in such patients, the rates of lymph node positivity, perineural invasion, and positive margins were lower compared with when surgery was performed without neoadjuvant chemotherapy, and that long-term survival may be similar or even better than after upfront resection [4,5]. (See 'Role of neoadjuvant chemotherapy' below and "Initial chemotherapy and radiation for nonmetastatic, locally advanced, unresectable and borderline resectable, exocrine pancreatic cancer".)

Contraindications — Absolute contraindications to resection include the presence of metastases in the liver, peritoneum, omentum, or any extra-abdominal site. Other indications of unresectability include encasement (more than one-half of the vessel circumference) or occlusion/thrombus of the superior mesenteric artery; unreconstructable SMV or SMV-portal vein confluence occlusion; or direct involvement of the inferior vena cava, aorta, or celiac axis, as defined by the absence of a fat plane between the low-density tumor and these structures on computed tomography (CT) scan or EUS. These patients should not go directly to surgery but rather receive neoadjuvant chemotherapy and radiotherapy. After neoadjuvant treatment, radiological changes of encasement, particularly around the arteries, do not necessarily reflect involvement by tumor [4]. (See "Clinical manifestations, diagnosis, and staging of exocrine pancreatic cancer", section on 'Definitions of unresectable and borderline resectable disease' and "Initial chemotherapy and radiation for nonmetastatic, locally advanced, unresectable and borderline resectable, exocrine pancreatic cancer" and 'Staging laparoscopy' below and "Clinical manifestations, diagnosis, and staging of exocrine pancreatic cancer", section on 'Importance of peritoneal cytology'.)

The presence of ascites is not necessarily a contraindication to attempted resection unless peritoneal cytology is positive. However, the management of patients with isolated positive peritoneal washings at the time of staging laparoscopy is controversial. The presence of these cells is associated with a worse prognosis in patients with otherwise resectable disease [6], and the American Joint Committee on Cancer (AJCC) tumor, node, metastasis (TNM) staging system considers positive peritoneal washings to represent distant metastatic disease (table 1). In general, most patients who have cytologically positive washings have other findings that suggest advanced disease and unresectability such as extensive ascites and/or the presence of metastases in the liver, pelvis, or omentum. However, if these are absent, most pancreatic surgeons would not rely solely upon the results of peritoneal washings obtained at the time of laparoscopy to guide decision making regarding resectability. (See "Clinical manifestations, diagnosis, and staging of exocrine pancreatic cancer", section on 'Importance of peritoneal cytology'.)

PREOPERATIVE CONSIDERATIONS

Staging laparoscopy — Currently available imaging techniques are highly accurate at predicting unresectable disease, but they fall short in predicting resectable disease, mainly because of limited sensitivity for small-volume metastatic disease. Radiographically occult metastases (<1 cm in diameter) on the surface of the liver or peritoneum, which are rarely visible by computed tomography (CT), magnetic resonance imaging (MRI), or transabdominal ultrasound, may be visualized laparoscopically [7]. This also includes patients with a tumor of the body or tail of the pancreas who appear to have potentially resectable disease by CT scan (one-half of whom will have occult peritoneal metastases), large (>3 cm) primary tumors, any patient for whom high-quality imaging is in any way suggestive of occult metastatic disease, and those with a CA 19-9 level >100 units/mL in absence of jaundice.

A selective approach to staging laparoscopy maximizes yield by limiting the procedure to those with the highest likelihood of occult metastatic disease. (See "Clinical manifestations, diagnosis, and staging of exocrine pancreatic cancer", section on 'Staging laparoscopy' and "Surgical resection of lesions of the head of the pancreas", section on 'Staging laparoscopy' and "Surgical resection of lesions of the body and tail of the pancreas", section on 'Staging laparoscopy'.)

Role of preoperative biliary drainage — Patients with pancreatic cancer who are jaundiced at presentation are at risk for perioperative complications. Some surgeons advocate an endoscopically placed stent prior to surgery while others reserve decompression for patients with a bilirubin of >20 mg/dL, with signs of cholangitis, or in those for whom surgery will be delayed for longer than two weeks. In practice, jaundiced patients may have already undergone biliary stenting before resectability or the time frame for resection has been determined, given that they are frequently initially seen by a gastroenterologist. (See "Supportive care for locally advanced or metastatic exocrine pancreatic cancer", section on 'Stents' and "Endoscopic stenting for malignant biliary obstruction" and "Surgical resection of lesions of the head of the pancreas", section on 'Preoperative biliary drainage'.)

Role of neoadjuvant chemotherapy — The low rate of resectability, the high rate of lymphatic disease at presentation, the poor long-term outcomes following pancreaticoduodenectomy with adjuvant therapy, and the fact that prolonged recovery prevents the delivery of postoperative adjuvant chemotherapy in approximately one-fourth of patients [8] have led to the investigation of neoadjuvant therapy in patients with potentially resectable pancreatic exocrine cancer. This approach has become standard for individuals with borderline resectable pancreatic cancer. (See "Initial chemotherapy and radiation for nonmetastatic, locally advanced, unresectable and borderline resectable, exocrine pancreatic cancer".)

While neoadjuvant therapy can be safely delivered to patients with localized, potentially resectable pancreatic cancer without adversely influencing perioperative morbidity or mortality, no study has clearly demonstrated improved resectability or survival compared with patients treated with surgery alone, and it remains unclear whether this approach provides benefit compared with modern adjuvant (postoperative) therapy. Furthermore, whether the neoadjuvant approach should include radiation therapy, which chemotherapy combination is optimal in the neoadjuvant setting and the optimal duration of neoadjuvant therapy remain unanswered questions. Randomized trials are currently underway.

At many institutions, neoadjuvant therapy is becoming an accepted option for patients with potentially resectable pancreatic cancer, although this is an evolving area. Eligible patients should be encouraged to enroll in trials testing novel strategies. This subject is discussed in detail elsewhere. (See "Treatment for potentially resectable exocrine pancreatic cancer", section on 'Neoadjuvant therapy'.)

TUMORS IN THE HEAD OF THE PANCREAS

Pancreaticoduodenectomy — The conventional operation for pancreatic cancer of the head or uncinate process is pancreaticoduodenectomy. Conventional pancreaticoduodenectomy (ie, Whipple procedure) involves removal of the pancreatic head, duodenum, first 15 cm of the jejunum, common bile duct, and gallbladder, and a partial gastrectomy (figure 2). Modifications of the conventional pancreaticoduodenectomy procedure have been developed in an attempt to improve outcomes or minimize the morbidity associated with the operation. These include:

●Pylorus-preserving pancreaticoduodenectomy – Pylorus-preserving pancreaticoduodenectomy preserves the gastric antrum, pylorus, and proximal 3 to 6 cm of the duodenum, which is anastomosed to the jejunum to restore gastrointestinal continuity (figure 3). The procedure may decrease the incidence of postoperative dumping, marginal ulceration, and bile reflux gastritis that can occur in some patients undergoing partial gastrectomy. The available data suggest that, for suitable cases, perioperative morbidity and mortality and long-term survival are not adversely affected using a pylorus-preserving technique. (See "Pylorus-preserving pancreaticoduodenectomy" and "Surgical resection of lesions of the head of the pancreas".)

●Subtotal stomach-preserving pancreaticoduodenectomy – Subtotal stomach-preserving pancreaticoduodenectomy aims to preserve as much stomach as possible while minimizing problems related to delayed gastric emptying that are associated with preserving the pyloric ring in the face of a loss of vagal innervation [9,10]. (See "Surgical resection of lesions of the head of the pancreas", section on 'Pylorus-preserving pancreaticoduodenectomy'.)

Minimally invasive (laparoscopic, robotic-assisted) pancreaticoduodenectomy is technically feasible. However, even with the available technology, minimally invasive pancreaticoduodenectomy is a complex operation that is only suited for selected patients. Robotic-assisted pancreaticoduodenectomy has not reduced rates of perioperative morbidity (eg, pancreatic fistula) or mortality [11]. (See "Surgical resection of lesions of the head of the pancreas", section on 'Open versus minimally invasive'.)

Vascular resection — Vascular resection and reconstruction at the time of pancreaticoduodenectomy is controversial because of the added complexity of the operative procedure and the poor quality of the published data that examine whether vascular resection represents a poor prognostic factor for survival duration (mainly due to the lack of information as to completeness of resection). (See 'Candidates for resection' above.)

Because of these issues and the inability to consistently determine the presence or absence of tumor adherence or invasion preoperatively, practice guidelines have suggested making decisions about vascular resection at the time of surgery after the pancreatic neck has been divided [12]. Although this is a controversial area, our position is that pancreaticoduodenectomy with portal vein (PV) or superior mesenteric vein (SMV) resection and reconstruction should be considered a standard approach for pancreatic adenocarcinomas that focally involve the PV or SMV, provided that adequate inflow and outflow veins are present, the tumor does not involve the superior mesenteric artery or hepatic artery [13], and an R0/R1 resection can be accomplished [14]. By contrast, we do not advocate arterial resection and reconstruction. These resections (mostly the superior mesenteric and hepatic arteries) are performed infrequently, and few data are available to support the practice [15-17]. Morbidity and mortality increase markedly when arterial resection and reconstruction is performed [18,19].

The available data suggest that patients with tumors involving the PV appear to be no more likely to have positive lymph nodes or poor prognostic histologic parameters (eg, aneuploidy) than those without portal vein involvement, suggesting that vein invasion is a function of tumor location rather than an indicator of aggressive tumor biology [20-22]. Whether perioperative mortality rates are different is unclear:

●Single-center reports that largely reflect highly selected patients treated at high-volume institutions report similar morbidity and perioperative mortality rates and comparable survival durations of patients who undergo venous resection and reconstruction versus those who do not, as long as a macroscopically complete (R0 or R1) resection has been accomplished [23-25]. A systematic review of 12 such reports concluded that PV/SMV resection combined with pancreatectomy is a safe and feasible procedure that increases the number of patients who undergo curative resection and therefore provides important survival benefits to selected groups of patients [26].

●However, in contrast to these data from single-center analyses, analysis of data on 3582 patients undergoing pancreatectomy for pancreatic cancer and derived from the American College of Surgeons National Surgical Quality Improvement Program database suggests significantly higher rates of postoperative morbidity (39.9 versus 33.3 percent) and mortality (5.7 versus 2.9 percent) in patients who undergo pancreaticoduodenectomy with, as compared to without, vascular resection [27]. This database includes patients treated at institutions with different levels of expertise, and it is likely that in many, if not most, cases the vascular resection was unplanned. These two factors could contribute to the higher perioperative morbidity and mortality rates.

Regional pancreatectomy — Regional pancreatectomy involves resection and reconstruction of the SMV-PV confluence and extensive en bloc regional lymph node dissection. However, the morbidity of a regional pancreatectomy is higher than that associated with conventional pancreaticoduodenectomy, and there is no apparent improvement in either local control or survival when a regional pancreatectomy is performed [28-30].

Extent of lymphadenectomy — Standard lymphadenectomy for pancreatoduodenectomy should strive to resect lymph node stations 5, 6, 8a, 12b1, 12b2, 12c, 13a, 13b, 14a, 14b, 17a, and 17b (figure 4) [31]. Some groups (mainly in Japan) routinely perform extensive lymph node dissection (extended lymphadenectomy including all 8, 9, all 12, all 14, 16a2, and 16b1 lymph nodes) in conjunction with pancreaticoduodenectomy since periampullary malignancies frequently metastasize to lymph nodes that are beyond the confines of the conventional pancreaticoduodenectomy [32,33]. However, the evidence to support a benefit from extended lymphadenectomy is conflicting, with some trials suggesting benefit limited to the setting of positive nodes [34], others suggesting no survival benefit in any subgroup [35-37], and two suggesting worse quality of life following extended lymphadenectomy [35,36].

A systematic review of standard versus extended lymphadenectomy during pancreaticoduodenectomy included 938 patients from nine studies (including four randomized trials [34-36,38]) [39]. Overall survival was no different between the groups at one, three, or five years. Outcomes were not stratified according to nodal status. The risk for some complications was significantly increased in the extended lymphadenectomy group (lymphatic fistula odds ratio [OR] 6.1, 95% CI 1.0-35.3; delayed gastric emptying OR 2.0, 95% CI 1.2-3.5; OR bile leak 2.6, 95% CI 1.0-6.7; OR pancreatic leak 1.7, 95% CI 1.0-2.9). The analysis was limited by differences in the ranges of lymphadenectomy from several studies, differences in the proportion of patients with different adjuvant therapies, and differences in the diagnostic criteria used for complications and overall mortality. Nevertheless, extended lymphadenectomy does not appear to convey any survival benefit and may be associated with increased perioperative morbidity and compromised quality of life, particularly in the early postoperative period. Pancreaticoduodenectomy with standard lymphadenectomy is the operation of choice.

Outcomes of pancreaticoduodenectomy — The prognosis for resection of adenocarcinoma of the head of the pancreas remains poor even with pancreaticoduodenectomy with surgically negative margins. Large series show five-year survival rates of only 10 to 25 percent and median survival between 10 and 20 months [40-46]. These series, however, include patients who did not undergo adjuvant treatment or received single-agent chemotherapy. Newer multiagent adjuvant chemotherapy regimens significantly increase both median and five-year survival after surgical resection. (See "Treatment for potentially resectable exocrine pancreatic cancer", section on 'Modern combination regimens (FOLFIRINOX and gem-nabpaclitaxel)'.)

Although previously associated with high perioperative morbidity and mortality rates, perioperative mortality of pancreaticoduodenectomy in most modern series is less than 3 percent [40,47-52]. This relatively low perioperative mortality rate represents a decline from over 15 percent in the 1970s.

One of the most important reasons for this improvement is the greater experience of a more limited number of surgeons who perform the procedure on a regular basis in high-volume hepatobiliary centers [53-56]. The relationship between surgeon and hospital volume, and postoperative mortality and survival after pancreatic cancer surgery, was addressed in a meta-analysis of 14 studies [57]. There was a significant association between hospital volume and postoperative mortality (OR 0.32, 95% CI 0.16-0.64) as well as overall survival (hazard ratio for death 0.79, 95% CI 0.70-0.89).

Given the increase in the proportion of patients undergoing surgery at higher-volume teaching hospitals, diminished surgical mortality, and greater use of adjuvant chemotherapy and chemoradiotherapy in more recent years [58], it was hoped that outcomes from resection would improve over time. However, there are conflicting reports in the literature regarding progress in the prognosis of resected pancreatic cancer:

●In one report of 396 Medicare patients residing in 1 of 11 Surveillance, Epidemiology, and End Results (SEER) reporting registries who underwent resection of pancreatic cancer with curative intent between 1991 and 1996, the three-year survival rate was 34 percent, a rate that compares favorably with published reports from the earlier literature [59]. In multivariate analysis, one of the strongest predictors for survival was the use of adjuvant chemoradiotherapy; both median survival and three-year survival rates were significantly higher among those who received it compared with those who did not (29 versus 12.5 months, and 45 versus 30 percent, respectively). (See "Treatment for potentially resectable exocrine pancreatic cancer", section on 'Chemoradiotherapy'.)

Similarly, a comparison of outcomes following pancreatectomy over two different time periods (1991 to 2000 and 2001 to 2010) in a single high-volume academic institution revealed significantly lower rates of perioperative mortality (1.6 versus 6.76 percent), a greater number of lymph nodes resected (17 versus 7), higher rates of adjuvant therapy use (63 versus 30 percent), and significantly longer survival among patients treated during the later time period (median 27 versus 16 months, five-year survival 27 versus 15 percent) [60].

Improvements in outcomes over time were also noted in a Korean series comparing outcomes for 746 patients treated between 2000 and 2009 with those of 1283 patients treated between 2010 and 2016 [61]. Five-year survival rates were 27.6 versus 22.3 percent in the two groups.

●In contrast to these data, a lack of progress in long-term survival over time was noted in a large series of 1147 pancreatic resections performed over three decades at the Memorial Sloan Kettering Cancer Center [62]. Although patients treated between 2000 and 2009 experienced lower rates of operative mortality and had greater one-year survival, long-term survival rates were nearly identical for patients treated in the 1980s, 1990s, and 2000s (median 23.2, 25.6, and 24.5 months, respectively). The corresponding five-year survival rates were 17, 20, and 8 percent. These data underscore the need for earlier detection and more effective adjuvant therapies.

TUMORS IN THE BODY OR TAIL — Because ductal adenocarcinomas involving the body or tail of the pancreas usually do not cause obstruction of the intrapancreatic portion of the common bile duct, early diagnosis is rare; the majority have locally advanced or metastatic disease at the time of presentation. In the rare patient who appears to have potentially resectable disease by computed tomography (CT) scan, laparoscopic exploration should precede attempted resection, since a significant proportion will have occult peritoneal metastases. (See "Clinical manifestations, diagnosis, and staging of exocrine pancreatic cancer", section on 'Staging laparoscopy'.)

Surgical resection of cancers located in the body or tail of the pancreas consists of a distal subtotal pancreatectomy, usually combined with splenectomy. Many of these procedures can be performed laparoscopically or robotically, which has not affected the ability to achieve a complete resection. Thus far, oncologic outcomes have not been inferior. Clinical factors that may influence the choice between an open or minimally invasive approaches are reviewed separately. (See "Surgical resection of lesions of the body and tail of the pancreas", section on 'Open surgical versus minimally invasive distal pancreatectomy'.)

Outcomes of distal pancreas resection — The scant data available regarding the outcome of surgical resection for tumors of the body and tail of the pancreas suggest a worse prognosis compared with those with cancers involving the head of the pancreas [63,64], although this is not a universal finding [65]. In one study, for example, only 13 of 105 patients (12 percent) with cancer of the body or tail of the pancreas had resectable tumors, and median survival was only 13 months after surgery, with only five patients remaining alive at two years [63].

TUMORS INVOLVING THE ENTIRE GLAND — Total pancreatectomy is sometimes required in order to achieve a microscopically negative resection margin [66-68]. However, the metabolic consequences of total pancreatectomy, which include permanent exocrine insufficiency and brittle diabetes, have a detrimental impact on quality of life and long-term survival [69-72]. Contemporary data on total pancreatectomy are lacking, and single-institution series are limited by small numbers [68,73,74].

Total pancreatectomy was first advocated to remove more tissue potentially involved with the malignancy and to avoid a pancreaticojejunal anastomosis, the source of considerable morbidity and mortality. However, single-institution studies suggested higher operative morbidity and mortality compared with pancreaticoduodenectomy [69,72,75]. Over time, perioperative (30 day) mortality has progressively declined; however, as with pancreaticoduodenectomy, morbidity remains high [74]. A review of the American College of Surgeons-National Surgical Quality Improvement Program (ACS-NSQIP) database (2005 to 2009) found a mortality rate of 5.4 percent among 166 patients who underwent total pancreatectomy [52]. Perioperative mortality for pancreaticoduodenectomy (n = 4317) and distal pancreatectomy (n = 2364) was 2.9 and 1.7 percent, respectively. The incidence of complications was similar between the procedures at 31.9, 34.7, and 27.8 percent for total pancreatectomy, pancreaticoduodenectomy, and distal pancreatectomy, respectively.

Long-term oncologic outcomes are not necessarily better following total as compared with partial pancreatectomy [74,76]. This was illustrated in a retrospective cohort study of 4021 patients who had a pancreatectomy with curative intent for adenocarcinoma and were reported to the National Cancer Institute Surveillance, Epidemiology, and End Results (NCI SEER) database between 1998 and 2004 [76]. The following findings were noted:

●Among the 376 patients who had a total pancreatectomy, perioperative mortality rates at one month (8.6 versus 6.3 percent) and three months (13.8 versus 10.8 percent) were not significantly higher than in patients who underwent partial pancreatectomy. There were no significant differences when the analysis was conducted separately according to primary tumor site (head, body/tail, unspecified location).

●Kaplan-Meier survival estimates for all three tumor locations demonstrated similar long-term survival after total as compared with partial pancreatectomy (at three years, 20 versus 15 percent for head, 22 versus 24 percent for body/tail, and 20 versus 25 percent for unspecified location, respectively).

These data support the use of total pancreatectomy where oncologically appropriate (ie, under conditions where it would yield a tumor-free resection margin while a conventional Whipple procedure would not). This typically occurs when a tumor of the pancreatic head extends into the body or tail of the pancreas. Total pancreatectomy also plays a role in the treatment of some main-duct intraductal papillary mucinous neoplasms (IPMNs), which may involve the entire length of the pancreatic duct, either continuously or in a multifocal fashion. When both types of pancreatic resection would provide a tumor-free margin, total pancreatectomy provides no additional benefit. (See "Intraductal papillary mucinous neoplasm of the pancreas (IPMN): Evaluation and management".)

PROGNOSIS AND PROGNOSTIC FACTORS — As noted above, even in the setting of completely resected, node-negative pancreatic cancer, the majority of patients die of their disease. The most important prognostic factor for completely resected patients is nodal status. Five-year survival after pancreaticoduodenectomy is only approximately 10 percent for node-positive disease (even if only one node is positive [77]), while it is approximately 30 percent for node-negative disease [3]. Importantly, two-thirds of patients with newly diagnosed pancreatic cancer would be expected to have positive lymph nodes, which carry a worse prognosis. This fact has been used as a justification for neoadjuvant chemotherapy for all patients with either resectable or borderline-resectable disease. (See "Treatment for potentially resectable exocrine pancreatic cancer", section on 'Neoadjuvant therapy' and "Initial chemotherapy and radiation for nonmetastatic, locally advanced, unresectable and borderline resectable, exocrine pancreatic cancer".)

Tumor stage is the most important prognostic factor. The influence of tumor stage on survival according to the newest eighth edition tumor, node, metastasis (TNM) staging classification (table 1) can be illustrated by a series of 8960 patients undergoing treatment for potentially resectable pancreatic adenocarcinoma and reported to the Surveillance, Epidemiology, and End Results (SEER) database between 2004 and 2013 (figure 5) [78].

Although five-year overall survival rates remain poor overall, survival estimates for individual patients are dynamic, and they may change over time, based upon the time already survived (a concept referred to as "conditional survival"). This was illustrated in a retrospective analysis of 1822 patients undergoing curative-intent surgery for pancreatic cancer at Johns Hopkins between 1970 and 2008 [79]. The two-year conditional survival at three years (ie, the probability of surviving to postoperative year 5 given that the patient had already survived three years) was 66 percent, versus a five-year actuarial survival calculated from the time of surgery of 18 percent. Patients with high lymph node ratios (the number of metastatic divided by the total number of resected nodes) or positive margins saw the greatest increases in two-year conditional survival as more time elapsed since treatment.

Only limited prognostic information exists for 5 to 15 year survivors. In an analysis of data from the SEER database of the National Cancer Institute, prognosis continued to improve with accrued survival beyond five years (table 2) [80]. However, deaths due to pancreatic cancer occurred as late as 20 years after diagnosis. The hazard of overall and pancreatic cancer-related death was highest at year 3 after diagnosis, declined until year 13, and then remained less than 3 percent per year. Between years 9 and 13, there was still an approximately 10 percent per-year risk of pancreatic cancer-specific death. The hazard of death due to other causes did not exceed that of pancreatic cancer-related death until 8.75 years postdiagnosis.

In addition to stage, other factors that influence prognosis after resection are the status of the surgical margins (involved or uninvolved), tumor differentiation and the presence or absence of lymphatic invasion within the tumor, both preoperative and postoperative serum CA 19-9 levels, and cigarette smoking [24,40-43,81-89]. (See "Clinical manifestations, diagnosis, and staging of exocrine pancreatic cancer".)

However, long-term survival is possible, even in the presence of positive nodes or involved margins. As an example, in the Charité Onkologie (CONKO)-001 trial examining the benefit of adjuvant gemcitabine, 15 percent of patients survived five years or more [90]. Among the 53 long-term survivors, 29 (54 percent) had node-positive disease, while 7 (13 percent) had undergone R1 (microscopically positive margins) resection.

The number of positive nodes is of prognostic significance. The eighth edition TNM staging system (2017 (table 1)) subdivides node-positive disease according to the number of positive nodes to provide better prognostic stratification (figure 1) [3]. (See "Clinical manifestations, diagnosis, and staging of exocrine pancreatic cancer", section on 'Staging system and the staging workup'.)

Finally, as has been seen in colon and gastric cancer, the total number of nodes examined also impacts prognosis [91,92]. At least some data support the view that pathologic examination of at least 15 nodes in the pancreatectomy specimen is necessary to accurately stage a node-negative adenocarcinoma [92].

A postresection nomogram has been developed [93] and validated [91] to predict the probability that a patient will die of pancreatic cancer within three years of surgery. In addition to the T and N status (according to the seventh edition [2010] criteria), this nomogram incorporates clinical (age, sex, presence of back pain or weight loss, tumor location), pathologic (histologic differentiation, tumor size, margin status, number of positive nodes), and surgical (type of resection) variables. Other nomograms are available that incorporate adjuvant treatment [94], but few have integrated the eighth edition (2017) American Joint Committee on Cancer (AJCC) staging criteria [95]. (See "Clinical manifestations, diagnosis, and staging of exocrine pancreatic cancer", section on 'Staging system and the staging workup'.)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Pancreatic cancer".)

INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●Basics topics (see "Patient education: Pancreatic cancer (The Basics)")

●Beyond the Basics topics (see "Patient education: Pancreatic cancer (Beyond the Basics)")

SUMMARY AND RECOMMENDATIONS

●General principles

•Surgical resection is the only potentially curative treatment for pancreatic adenocarcinoma. Unfortunately, because of the late presentation, only 15 to 20 percent of patients are candidates for pancreatectomy. Although uncommon, some patients with initially unresectable or borderline resectable disease may be downstaged to resectable disease with neoadjuvant treatment. (See 'Introduction' above.)

•Absolute contraindications to resection include (see 'Contraindications' above and 'Vascular resection' above and "Clinical manifestations, diagnosis, and staging of exocrine pancreatic cancer", section on 'Definitions of unresectable and borderline resectable disease'):

-Metastases in the liver, peritoneum, omentum, or any extra-abdominal site.

-Encasement (more than one-half of the vessel circumference) or occlusion/thrombus of the superior mesenteric artery; unreconstructable superior mesenteric vein (SMV) or SMV-portal vein confluence occlusion; or direct involvement of the inferior vena cava, aorta, celiac axis, or hepatic artery, as defined by the absence of a fat plane between the low-density tumor and these structures on computed tomography (CT) scan or endoscopic ultrasound (EUS). After neoadjuvant chemotherapy, radiological changes of encasement, particularly around the arteries, do not necessarily reflect involvement by tumor.

•Some cases are considered "borderline" resectable, although the definition is variable. Some reserve the term "borderline resectable" for cases where there is focal (less than one-half of the circumference) tumor abutment of the visceral arteries or short-segment occlusion of the SMV or SMV/portal vein confluence. Others suggest that venous narrowing without occlusion be included in the definition. (See 'Candidates for resection' above and "Clinical manifestations, diagnosis, and staging of exocrine pancreatic cancer", section on 'Definitions of unresectable and borderline resectable disease'.)

•Patients who have borderline resectable or locally advanced unresectable pancreatic cancer, but no metastatic disease, are potential candidates for downstaging with neoadjuvant therapy and should be referred for medical oncology and radiation therapy consultation. (See 'Candidates for resection' above and "Initial chemotherapy and radiation for nonmetastatic, locally advanced, unresectable and borderline resectable, exocrine pancreatic cancer".)

At many institutions, neoadjuvant therapy is increasingly considered an acceptable option for patients with potentially resectable pancreatic cancer. Eligible patients should be encouraged to enroll in clinical trials testing novel strategies. (See 'Role of neoadjuvant chemotherapy' above.)

●Tumors in the head/uncinate process of the pancreas

•For potentially resectable pancreatic cancers within the head or uncinate process, the standard operation is pancreaticoduodenectomy. (See 'Pancreaticoduodenectomy' above.)

•For patients undergoing pancreaticoduodenectomy, there is no evidence that either a conventional or pylorus-preserving Whipple operation is superior, and the preferences of the surgeon and patient should prevail. Although some studies suggest modestly higher rates of delayed gastric emptying with a pylorus-preserving approach, the impact on gastrointestinal function remains an open question. (See 'Pancreaticoduodenectomy' above.)

•There is no added benefit to be gained by regional pancreatectomy or extended lymphadenectomy, and we recommend that these procedures not be performed (Grade 1B). (See 'Regional pancreatectomy' above and 'Extent of lymphadenectomy' above.)

•Although controversial, vein resection and reconstruction is a standard approach for pancreatic adenocarcinomas focally involving the portal vein or SMV, providing that adequate inflow and outflow veins are present, the tumor does not involve the superior mesenteric artery (SMA) or hepatic artery, and a macroscopically complete (R0/R1) resection can be accomplished. (See 'Vascular resection' above.)

•For patients presenting with obstructive jaundice, uncertainty as to the benefit of preoperative drainage has led to differing approaches. We suggest biliary decompression for selected patients in whom surgery will be delayed for longer than two weeks (Grade 2C), and we recommend it in the presence of cholangitis (Grade 1B). (See 'Role of preoperative biliary drainage' above and "Surgical resection of lesions of the head of the pancreas", section on 'Preoperative biliary drainage'.)

●Tumors of the tail/body of the pancreas

•Surgical resection of cancers located in the body or tail of the pancreas consists of a distal pancreatectomy, usually combined with splenectomy. (See 'Tumors in the body or tail' above.)

•We suggest laparoscopic exploration prior to attempted resection since a significant proportion will have occult peritoneal metastases (Grade 2B). (See "Surgical resection of lesions of the body and tail of the pancreas", section on 'Staging laparoscopy'.)

●Tumors involving the entire gland

•Total pancreatectomy is sometimes needed for tumors involving the entire gland to achieve a microscopically negative resection margin. (See 'Tumors involving the entire gland' above.)

•However, the metabolic consequences of total pancreatectomy, which include permanent exocrine insufficiency and diabetes, have a detrimental impact on quality of life and long-term survival.

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Michael Steer, MD, who contributed to an earlier version of this topic review.