Treatment for potentially resectable exocrine pancreatic cancer

INTRODUCTION — For patients with exocrine pancreatic cancer, a majority (85 percent) present with adenocarcinomas arising from the ductal epithelium. (See "Pathology of exocrine pancreatic neoplasms".)

Surgical resection offers the only chance of cure for exocrine pancreatic cancer, but only 15 to 20 percent of cases are potentially resectable at presentation. Local unresectability is usually (but not always) due to vascular invasion. (See "Overview of surgery in the treatment of exocrine pancreatic cancer and prognosis", section on 'Candidates for resection'.)

Unfortunately, prognosis is poor, even for those undergoing microscopically complete (R0) resection. Given the high rates of both systemic (>80 percent) and local (>20 percent) recurrence after surgery alone, systemic chemotherapy, radiation therapy, and combined approaches (chemoradiotherapy) have been used both prior to and following surgical resection in an effort to improve cure rates.

Management of potentially resectable pancreatic cancer will be discussed here. Neoadjuvant strategies for locally advanced, unresectable or "borderline" resectable exocrine pancreatic cancer; surgical management of localized exocrine pancreatic cancer; and chemotherapy for advanced disease are discussed separately. (See "Overview of surgery in the treatment of exocrine pancreatic cancer and prognosis" and "Ampullary carcinoma: Epidemiology, clinical manifestations, diagnosis and staging" and "Initial chemotherapy and radiation for nonmetastatic, locally advanced, unresectable and borderline resectable, exocrine pancreatic cancer" and "Initial systemic chemotherapy for metastatic exocrine pancreatic cancer".)

APPROACH TO THE PATIENT

Assessing resectability

●Radiographic staging – Multiphasic computed tomography scans of the chest, abdomen, and pelvis are indicated to assess disease extent [1]. Other staging studies should be performed only as dictated by symptoms. (See "Clinical manifestations, diagnosis, and staging of exocrine pancreatic cancer", section on 'Abdominal CT' and "Clinical manifestations, diagnosis, and staging of exocrine pancreatic cancer", section on 'Carbohydrate antigen 19-9'.)

Pancreatic cancers can be categorized on a continuum from potentially resectable to unresectable according to the involvement of adjacent structures and the presence of distant metastases (figure 1) [2]. (See "Clinical manifestations, diagnosis, and staging of exocrine pancreatic cancer", section on 'Imaging studies' and "Overview of surgery in the treatment of exocrine pancreatic cancer and prognosis", section on 'Candidates for resection'.)

Three general categories are recognized based on radiographic stating that influence the treatment approach:

●Potentially resectable – A commonly accepted definition of potentially resectable pancreatic cancer is a tumor that is limited to the pancreas with no arterial contact (superior mesenteric artery [SMA], celiac axis, common hepatic artery); some also include superior mesenteric vein (SMV) involvement ≤90 degrees in the definition [3]. Tumors with very limited involvement of the mesenteric vasculature may be technically resectable upfront, but the impact of more aggressive resections (particularly arterial resection) on long-term outcomes is controversial.

●Borderline resectable – Although definitions are variable, the National Comprehensive Cancer Network (NCCN) formally defines the category of "borderline resectable" pancreatic cancer as follows [4]:

•For tumors of the head or uncinate process, solid tumor contact with any of the following:

-The SMV or portal vein >180 degrees with contour irregularity or thrombosis of the vein, but with suitable vessel proximal and distal to the site of involvement, allowing for safe and complete resection and vein reconstruction.

-The inferior vena cava.

-The common hepatic artery without extension to the celiac axis or hepatic artery bifurcation.

-The SMA ≤180 degrees.

-Variable anatomy (eg, accessory right hepatic artery, replaced right hepatic artery, replaced common hepatic artery, and the origin of replaced or accessory artery), and the presence and degree of tumor contact should be noted if present, as it may affect surgical planning.

•For tumors of the body/tail, solid tumor contact with:

-The celiac axis of ≤180 degrees.

-It should be noted that some institutions will still attempt resection on patients who have CT evidence of >180 degree celiac axis involvement on the basis that some studies have shown that this 'involvement' represents scar tissue and not cancer in some patients.

While a subset of these patients will prove to be resectable with initial surgery, the likelihood of an incomplete resection is high. A preferred approach is attempted downstaging with neoadjuvant therapy (algorithm 1). (See "Overview of surgery in the treatment of exocrine pancreatic cancer and prognosis", section on 'Vascular resection' and "Initial chemotherapy and radiation for nonmetastatic, locally advanced, unresectable and borderline resectable, exocrine pancreatic cancer", section on 'Borderline resectable disease'.)

●Unresectable – Absolute contraindications to resection include the presence of metastases in the liver, peritoneum, omentum, or any extra-abdominal site. Other indications of local unresectability include encasement (more than one-half of the vessel circumference) or occlusion/thrombus of the SMA, 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 endoscopic ultrasound. Management of these locally advanced but unresectable patients usually entails initial chemotherapy and not surgical exploration (algorithm 1). (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 "Initial systemic chemotherapy for metastatic exocrine pancreatic cancer".)

●Contribution of serum tumor markers – Preoperative serum carbohydrate antigen 19-9 (CA 19-9) levels should not be used to direct the initial treatment strategy. Elevated levels of CA 19-9 can help to predict the presence of radiographically occult metastatic disease, the likelihood of a microscopically complete (R0) resection, and long-term outcomes in patients with potentially resectable pancreatic cancer [5]. However, while high levels of CA 19-9 may help surgeons to better select patients who need staging laparoscopy, the use of CA 19-9 alone as an indicator of operability is not advised [6]. Furthermore, although a year 2016 clinical practice guideline on management of potentially curable pancreatic cancer from the American Society of Clinical Oncology (ASCO) and guidelines from the NCCN suggest that chemotherapy could be used before surgery for patients who had potentially anatomically resectable but high-risk tumors (as judged by elevated preoperative levels of CA 19-9, among other factors), there was no recommendation for a specific cutoff value of CA 19-9 to select patients for neoadjuvant therapy [4,6]. (See "Clinical manifestations, diagnosis, and staging of exocrine pancreatic cancer", section on 'Carbohydrate antigen 19-9'.)

Multidisciplinary assessment and symptom burden — Multidisciplinary collaboration to formulate treatment plans is preferred. Baseline performance status and a comorbidity profile should be evaluated. Goals of care, patient preferences, psychologic status, support systems, and symptoms should guide decisions for treatment [7]. (See "Discussing goals of care" and "Supportive care for locally advanced or metastatic exocrine pancreatic cancer".)

All patients should have a full assessment of symptom burden, psychologic status, and social supports as early as possible. In some cases, this will indicate a need for formal palliative care consultation and services [7]. Early initiation of palliative care services improves clinical and quality of care outcomes, and may prolong survival. (See "Benefits, services, and models of subspecialty palliative care", section on 'Rationale for palliative care'.)

Patients with pancreatic cancer who are jaundiced are at risk for associated coagulopathy, malabsorption, and malnutrition. It was hoped that relief of biliary obstruction preoperatively would correct these defects and decrease postoperative morbidity and mortality rates. However, results from randomized trials have not been consistent. Uncertainty as to the benefit of preoperative biliary drainage has led to differing approaches. Some surgeons routinely decompress jaundiced patients with an endoscopically placed stent prior to surgery. However, most reserve biliary decompression for patients whose serum bilirubin concentration exceeds 20 mg/dL, for those in whom surgery will be delayed for longer than two weeks (particularly if neoadjuvant therapy is chosen), and in those with debilitating pruritus or cholangitis. (See "Overview of surgery in the treatment of exocrine pancreatic cancer and prognosis", section on 'Role of preoperative biliary drainage'.)

Initial resection versus neoadjuvant therapy — Treatment plans for all patients with nonmetastatic pancreatic cancer should be guided by radiographic disease extent, performance status, comorbidities, goals of care, patient preferences, psychosocial issues, and symptom burden. Management of potentially resectable disease is in evolution. We consider neoadjuvant therapy to be a reasonable alternative to upfront surgery followed by adjuvant therapy in patients who appear to have potentially resectable tumors, as long as performance status and comorbidity are sufficient to tolerate treatment. Surgery could then be reconsidered at a later time if appropriate. Further studies are necessary to clarify the optimal approach, and we encourage patients to enroll in clinical trials comparing neoadjuvant/perioperative therapy with upfront resection followed by adjuvant therapy.

Given the poor outcomes with upfront surgery, many institutions have embraced neoadjuvant combination chemotherapy for initial therapy of borderline resectable pancreatic cancer, followed by re-evaluation for surgical exploration, and this is now a preferred approach. (See "Initial chemotherapy and radiation for nonmetastatic, locally advanced, unresectable and borderline resectable, exocrine pancreatic cancer", section on 'Benefit of neoadjuvant therapy'.)

However, many centers, including those of several of the authors and editors associated with this review, have taken the view that nearly all patients who appear to have "potentially resectable" pancreatic cancer should in fact be considered to have "borderline resectable" disease given the inaccuracy of imaging and the very high rates of a positive resection margins with initial surgery, a finding that is associated with poor overall survival (OS). Although only limited data are available in patients with potentially resectable disease, accumulating data from prospective randomized trials and a meta-analysis support benefit for neoadjuvant therapy in combined populations of patients with potentially resectable or borderline resectable pancreatic cancer. (See 'Rationale and benefit over initial resection' below.)

Although guidelines from ASCO and NCCN suggest that neoadjuvant therapy could be offered to certain "high-risk" patients with initially resectable disease, including those with a high initial CA 19-9 level [4,6], in our view, preoperative serum CA 19-9 levels should not be used to direct the initial treatment strategy. (See 'Assessing resectability' above.)

Our general approach to patients with nonmetastatic pancreatic cancer is outlined in the algorithm (algorithm 1), and the approach to patients with potentially resectable tumors is described in detail in the sections below.

UPFRONT PANCREATECTOMY — Primary surgical resection is an appropriate option for patients with no distant metastatic disease, a performance status and comorbidity profile that are appropriate for major abdominal surgery, and no radiographic interface between the primary tumor and the celiac axis, superior mesenteric artery, or common hepatic artery. For those patients who have radiographic interface between the tumor and the mesenteric vasculature as well as those who have potentially resectable tumors, and a performance status or comorbidity profile that is not appropriate for a major abdominal operation but is potentially reversible, we suggest initial chemotherapy rather than upfront surgery. Surgery could then be reconsidered at a later time if appropriate. (See 'Neoadjuvant therapy' below.)

An overview of surgical treatment for pancreatic cancer is provided elsewhere. (See "Overview of surgery in the treatment of exocrine pancreatic cancer and prognosis".)

Adjuvant therapy

Indications

Ductal adenocarcinoma — We recommend adjuvant chemotherapy for all patients with resected pancreatic ductal adenocarcinoma who did not receive neoadjuvant therapy, including those with resected T1N0 disease (table 1). This recommendation is consistent with guidelines from American Society of Clinical Oncology (ASCO) [6,8], the National Comprehensive Cancer Network (NCCN) [4], and the European Society for Medical Oncology (ESMO) [9].

Efficacy of different regimens, and benefit for specific subgroups, including those with stage I disease, and older individuals, are discussed in detail below. (See 'Chemotherapy' below.)

Acinar cell histology — Indications for adjuvant therapy in patients with acinar cell carcinoma (ACC) are undefined. Despite the better overall survival (OS) compared with ductal adenocarcinoma, we treat these patients the same as those with pancreatic ductal adenocarcinoma, offering adjuvant chemotherapy to most patients with resected disease, with the possible exception of those with well-differentiated T1 tumors and negative resection margins, as they have a relatively favorable prognosis. (See 'Choice of therapy' below.)

We would also consider these patients to be appropriate candidates for neoadjuvant therapy, just as we do for those with pancreatic ductal adenocarcinoma. (See 'Neoadjuvant therapy' below.)

ACCs are rare malignant neoplasms that constitute approximately 1 percent of all malignant nonendocrine pancreatic tumors. Histologically, there is clear evidence of acinar cell differentiation, which can be identified immunohistochemically by staining for trypsin, chymotrypsin, elastase, or lipase; or can be ultrastructurally identified by the presence of zymogen granules, although variant histologic patterns reflect different degrees of differentiation (picture 1). (See "Pathology of exocrine pancreatic neoplasms", section on 'Acinar cell carcinoma'.)

Some ACCs give rise to a clinical syndrome related to lipase hypersecretion with distant manifestations: subcutaneous fat necrosis (termed "pancreatic panniculitis"), eosinophilia, and polyarthralgia (the so-called "Schmidt's triad"). This syndrome is associated with poor prognosis [10,11]. (See "Cutaneous manifestations of internal malignancy".)

The available evidence suggests high recurrence rates, even after complete surgical resection [12-14]. However, the overall prognosis for patients with ACC is better than that for patients with ductal adenocarcinoma [12,15-17]. Patients with ACC tend to present at a younger age than do those with pancreatic ductal adenocarcinoma; they are more likely to have localized disease (15 versus 8 percent) and are more likely to undergo resection (39 versus 11 percent) [16]. Five-year survival rates among patients with resected disease range from 44 to 72 percent [16-19]; in one large series, the stage-stratified five-year survival rates for ACC versus pancreatic ductal adenocarcinoma were as follows [17]:

●Stage I – 52 versus 28 percent

●Stage II – 40 versus 10 percent

●Stage III – 23 versus 7 percent

Timing and duration — For all patients who undergo initial resection without neoadjuvant therapy, including those with pathologic T1N0 disease (table 1), we recommend six months of adjuvant chemotherapy. Adjuvant treatment should be started within eight weeks of surgery, if possible, but it can be started beyond eight weeks if treatment is delayed by postoperative complications. Prior to beginning adjuvant therapy, all patients should undergo formal restaging with CT scans and a serum level of the tumor marker carbohydrate antigen 19-9 (CA 19-9). We do not use postoperative CA 19-9 levels to determine whether or not to give postoperative adjuvant therapy.

The optimal timing and duration of adjuvant therapy are not established. A focused guideline update on potentially curable pancreatic cancer from ASCO recommends adjuvant systemic chemotherapy for six months starting within eight weeks of surgery, assuming adequate recovery from surgery [8]. Modern adjuvant chemotherapy trials have utilized six months as the standard duration of postresection chemotherapy, and they permitted enrollment up to 12 weeks postoperatively [20]. However, there are no randomized trials addressing the impact of delayed initiation of adjuvant therapy on outcomes or addressing the effect of a longer duration of therapy.

Some information on timing and duration of adjuvant chemotherapy is available from an analysis of the 985 patients who received adjuvant gemcitabine or FU after resection of pancreatic cancer in the ESPAC-3 trial [21] (see 'Other gemcitabine and FU-based approaches' below):

●OS favored patients who completed the full six months of therapy over those who did not (median survival 28 versus 15 months, hazard ratio [HR] for death 0.51, 95% CI 0.44-0.60).

●Time to start of chemotherapy (within eight weeks of surgery versus later) was an important survival factor only for the subgroup of patients who did not complete all six months of therapy (and in this group, survival inexplicably favored later initiation of therapy). There seemed to be no difference in outcomes if chemotherapy was delayed for up to 12 weeks.

While these data support the view that delaying the initiation of adjuvant therapy to allow full recovery from surgery does not compromise the benefit of adjuvant therapy, the subset analysis must be viewed cautiously. There are several confounding reasons as to why patients who receive less than six months of adjuvant therapy might do worse.

Results from retrospective analyses are conflicting on the issue of timing:

●An analysis of approximately 10,000 patients with resected stage I to III pancreatic cancer derived from the National Cancer Database (NCDB) found a similar proportional benefit for adjuvant chemotherapy initiated within 12 weeks or beyond 12 weeks postresection, compared with surgery alone [22].

●On the other hand, in a separate analysis derived from the NCDB, patients with stage I or II pancreatic cancer who commenced adjuvant therapy between 28 and 59 days after primary surgical resection had better survival outcomes than did those initiating adjuvant therapy before 28 or beyond 59 days [23]. However, even patients who initiated adjuvant systemic therapy more than 12 weeks postresection had improved OS relative to those undergoing surgery alone.

Persistent postoperative elevations of serum tumor marker CA 19-9 are associated with a poor long-term prognosis. However, CA 19-9 levels are prognostic and not necessarily predictive of benefit from adjuvant therapy.

Some suggest withholding adjuvant therapy from such patients or treating them as if they have advanced metastatic disease [24], and some adjuvant therapy protocols (such as Radiation Therapy Oncology Group [RTOG] 0848) only allow enrollment if the post-treatment CA 19-9 level is ≤180 units/mL. However, we (and others [25]) suggest not using postoperative CA 19-9 levels to determine whether or not to give postoperative adjuvant therapy outside of the context of a clinical trial. (See "Clinical manifestations, diagnosis, and staging of exocrine pancreatic cancer", section on 'Carbohydrate antigen 19-9' and "Overview of surgery in the treatment of exocrine pancreatic cancer and prognosis", section on 'Prognosis and prognostic factors'.)

Choice of therapy

Chemotherapy — If protocol therapy is not available or declined, for patients with an excellent performance status who are able to tolerate it, we suggest modified FOLFIRINOX (oxaliplatin plus irinotecan with leucovorin [LV] and short-term infusional fluorouracil [FU]) rather than gemcitabine alone or gemcitabine plus capecitabine. For less fit patients, gemcitabine plus capecitabine or gemcitabine plus nabpaclitaxel are options. We reserve gemcitabine alone for patients with a borderline performance status or a comorbidity profile that precludes multiagent therapy. In Japan, oral therapy with S-1, where available, is a preferred alternative to gemcitabine monotherapy, given the results of a randomized trial demonstrating therapeutic noninferiority and better tolerability for S-1 [26]. (See 'Other gemcitabine and FU-based approaches' below.)

Early trials demonstrating benefit of adjuvant therapy

●ESPAC-1 – An ambitious trial sponsored by European investigators, the ESPAC-1 trial, initially set out to randomize patients to a two-by-two factorial design in which the relative benefits of adjuvant chemotherapy, chemoradiotherapy (CRT), or CRT followed by chemotherapy would be compared with observation alone. However, fear of poor accrual led the investigators to permit the clinician to choose from this or two other randomization schemes (described below).

The pooled analysis included 541 patients (including those with positive margins) who were randomly assigned to postoperative treatment following resection of exocrine pancreatic cancer in the following three parallel studies [27]:

•CRT versus no CRT – Sixty-eight patients enrolled; CRT consisted of 20 Gy of external beam radiation therapy (RT) plus three days of concomitant FU, repeated after a planned break of two weeks.

•Adjuvant chemotherapy versus no chemotherapy – One hundred and eighty-eight patients enrolled; adjuvant chemotherapy consisted of bolus LV-modulated FU (LV 20 mg/m², FU 425 mg/m²) administered daily for five days every 28 days for six months.

•A two-by-two factorial design trial with four groups – CRT (n = 73), chemotherapy (n = 75), both (n = 72), or observation (n = 69) [28]; the CRT and chemotherapy treatments were as described above.

In the latest intent-to-treat analysis, both the two-year (40 versus 30 percent) and five-year (21 versus 8 percent) survival rates were significantly greater among patients randomized to postoperative chemotherapy alone compared with those who did not receive it, despite the fact that 33 percent of those assigned to adjuvant chemotherapy did not complete all six courses and 17 percent received no chemotherapy at all [28]. By contrast, there was no significant benefit for CRT in the two groups that received it, and in fact, the data suggested a trend toward worse survival for this group. These results are discussed below. (See 'Chemoradiotherapy' below.)

●CONKO-001 – A survival benefit from adjuvant gemcitabine monotherapy was shown in the multinational European CONKO-001 trial, in which 368 patients with a grossly complete (microscopically complete [R0] or macroscopically complete [R1]) surgical resection and a preoperative CA 19-9 level <2.5 times the upper limit of normal were randomly assigned to gemcitabine (1000 mg/m² on days 1, 8, and 15 every four weeks for six months) or no treatment after surgery [29]. In the most recent update, there was a modest but significant improvement in OS that favored gemcitabine and persisted long term (five-year OS 21 versus 10 percent; 10-year OS 12.2 versus 7.7 percent) [30]. These results established gemcitabine monotherapy as a standard adjuvant treatment for resected pancreatic cancer.

Modern combination regimens (FOLFIRINOX and gem-nabpaclitaxel)

●Modified FOLFIRINOX – Better outcomes may be achieved using multiagent chemotherapy combinations, such as FOLFIRINOX (oxaliplatin plus irinotecan with LV and short-term infusional FU), which have been shown to be more effective than gemcitabine alone in the setting of metastatic disease. (See "Initial systemic chemotherapy for metastatic exocrine pancreatic cancer".)

The benefits of modified FOLFIRINOX over gemcitabine alone were shown in the multicenter PRODIGE-24 trial, which randomly assigned 493 patients with histologically proven pancreatic ductal adenocarcinomas 21 to 84 days after R0 or R1 resection, a performance status of 0 or 1, adequate hematologic and renal function, and no cardiac ischemia to six months of gemcitabine alone (28-day cycles of gemcitabine 1000 mg/m² on days 1, 8, and 15) or modified FOLFIRINOX (table 2) [31,32]. At median follow-up of 70 months, compared with gemcitabine alone, modified FOLFIRINOX improved disease-free survival (DFS; median 21 versus 13 months, five-year DFS 26 versus 19 percent, HR 0.66, 95% CI 0.54-0.82) and OS (median 54 versus 36 months, five-year OS 43 versus 31 percent, HR 0.68, 95% CI 0.54-0.85) [32]. Grade 3 to 4 adverse events that were more frequent with modified FOLFIRINOX included diarrhea (19 versus 4 percent), sensory neuropathy (9 versus 0 percent), fatigue (11 versus 5 percent), and vomiting (5 versus 1 percent).

There are no trials directly comparing adjuvant FOLFIRINOX with gemcitabine plus capecitabine.

Based on these results, off protocol, for patients with an excellent performance status (ECOG 0) (table 3) who are able to tolerate it, we suggest modified FOLFIRINOX rather than gemcitabine alone for adjuvant therapy, and we also prefer this regimen over gemcitabine plus capecitabine or gemcitabine plus nabpaclitaxel in patients with a good performance status (ECOG 0 to 1 (table 3)).

●Gemcitabine plus nabpaclitaxel – Gemcitabine plus nanoparticle albumin-bound paclitaxel (nabpaclitaxel) is a highly active regimen in metastatic pancreatic cancer, leading to interest in its use as adjuvant therapy in resected pancreatic cancer [33-35]. (See "Initial systemic chemotherapy for metastatic exocrine pancreatic cancer", section on 'Gemcitabine plus nabpaclitaxel'.)

In an international, open-label phase III trial (APACT), 866 patients with R0 or R1 resected pancreatic cancer were randomly assigned to six months of gemcitabine (1000 mg/m² on days 1, 8, and 15 of every 28-day cycle) alone or in combination with nabpaclitaxel (125 mg/m² on days 1, 8, and 15 of every 28-day cycle) [35]. At median follow-up of 38 months, compared with single-agent gemcitabine, the combination of gemcitabine and nabpaclitaxel failed to improve DFS as assessed by independent review (median 19.4 versus 18.8 months, HR 0.88, 95% CI 0.73-1.063) but did improve OS in extended follow-up (median OS 42 versus 38 months, five-year OS 38 versus 31 percent, HR 0.80, 95% CI 0.68-0.95).

For patients who are ineligible for FOLFIRINOX, acceptable options for adjuvant therapy include either gemcitabine plus nabpaclitaxel or gemcitabine plus capecitabine. (See 'Other gemcitabine and FU-based approaches' below.)

Other gemcitabine and FU-based approaches

●ESPAC-3 – The multicenter ESPAC-3 trial randomly assigned 1088 patients with resected exocrine pancreatic cancer to six months of postoperative adjuvant treatment with either weekly gemcitabine or monthly LV-modulated FU [36]. At a median follow-up of 34 months, median survival was similar (23.6 versus 23 months with gemcitabine and LV/FU therapy, respectively). However, the patients assigned to LV/FU had more grade 3 to 4 treatment-related toxicity, and more treatment-related hospitalizations. Otherwise, progression-free survival and global quality of life scores were similar between the two groups.

●RTOG-9704 – A slightly different question, the relative value of adjuvant gemcitabine monotherapy both before and after concomitant FU-based CRT versus all-FU therapy in which FU was given before, during, and after RT, was addressed in the RTOG 9704 trial [37]. Patients who had undergone gross total resections for T1-4, N0-1 (according to the 2010 tumor, node, metastasis [TNM] classification (table 4)) exocrine pancreatic cancer and who were taking in at least 1500 calories daily postoperatively were randomly assigned to one of the following two treatment arms:

•FU arm – Three weeks of continuous-infusion FU (250 mg/m² daily) followed by CRT (50.4 Gy in 1.8 Gy daily fractions for 5.5 weeks with concurrent infusional FU 250 mg/m² daily) and, starting three to five weeks later, two four-week courses of continuous-infusion FU (250 mg/m² daily with a two-week rest in between courses)

•Gemcitabine arm – Three weekly doses of gemcitabine alone (1000 mg/m² per week) followed by the same CRT protocol as for the conventional CRT arm and, starting three to five weeks later, three months of single-agent gemcitabine (1000 mg/m² weekly for three of every four weeks)

In the latest update, there were no significant differences in five-year OS or DFS between the two groups [38]. The study was sufficiently powered to separately analyze results according to the primary site (head versus body/tail). Among patients with pancreatic head tumors (n = 388), there was a nonstatistically significant trend toward better median (20.5 versus 17.1 months) and five-year survival (22 versus 18 percent, ) with gemcitabine-based adjuvant therapy, although neither difference was statistically significant. There were no differences between the two treatments in patients with body/tail tumors (n = 63). Compared with the all-FU treatment, the gemcitabine group had similar rates of nonhematologic grade 3 or 4 toxicity and febrile neutropenia, despite significantly more grade 4 hematologic toxicity (14 versus 1 percent) [37].

●Gemcitabine versus S-1 – S-1 is an oral fluoropyrimidine that includes three different agents: ftorafur (tegafur), gimeracil (5-chloro-2,4 dihydropyridine; a potent inhibitor of dihydropyrimidine dehydrogenase), and oteracil (potassium oxonate; which inhibits phosphorylation of intestinal FU, thought to be responsible for treatment-related diarrhea). It is approved in Japan for adjuvant treatment of gastric cancer and in Europe for treatment of advanced gastric cancer; it is not available in the United States.

S-1 (40 to 60 mg twice daily for four weeks and repeated every six weeks for four courses) was directly compared with gemcitabine (1000 mg/m² on days 1, 8, and 15 and repeated every four weeks for six courses) in a multicenter trial of 385 Japanese patients with completely resected stage I, II, or III pancreatic cancer [39]. S-1 was found to be noninferior to gemcitabine, and patients treated with S-1 actually had a lower mortality rate (HR for death 0.57, 95% CI 0.44-0.72; five-year survival 44 versus 24 percent). Although both treatments were associated with similarly low rates of grade 3 or 4 anorexia, thrombocytopenia, and anemia, gemcitabine was associated with more leukopenia (39 versus 9 percent) and transaminitis (5 versus 1 percent). Whether these results can be extrapolated to other populations is unclear.

●Gemcitabine plus capecitabine – The ESPAC-4 trial randomly assigned 730 patients with R0 or R1 resected pancreatic adenocarcinoma to six months of gemcitabine alone (1000 mg/m² on days 1, 8, and 15 of each 28-day cycle) or to the same dose of gemcitabine plus capecitabine (1660 mg/m² per day on days 1 through 21 of each 28-day cycle) (table 5) [20]. The following outcomes were reported:

•The majority of patients in both arms had R1 resection margins (60 percent) and positive lymph nodes (80 percent).

•Of the grade 3 or 4 adverse effects, diarrhea (5 versus 2 percent), hand-foot syndrome (7 versus 0 percent), and neutropenia (38 versus 24 percent) were significantly more common with combined therapy, although there were no significant differences in the rates of treatment-related serious adverse events.

•In the latest analysis, at a median follow-up of 43 months, combination therapy was associated with a significant survival benefit (median OS 30.2 versus 27.9 months, HR for death 0.81, 95% CI 0.68-0.98, and five-year OS was 28 versus 17 percent [40]).

Chemotherapy for BRCA and PALB2 mutation carriers — The optimal chemotherapy regimen to use for BRCA1/BRCA2 or partner and localizer of BRCA2 (PALB2) mutation carriers with pancreatic adenocarcinoma is not established. Until further information becomes available, we suggest using the same adjuvant chemotherapy regimens as are used for nonmutation carriers.

Hereditary breast and ovarian cancers are characterized by the presence of germline mutations in one of two cancer susceptibility genes, BRCA1 and BRCA2. The risk of pancreatic cancer is elevated in those with BRCA2 mutations, and it may also be elevated in BRCA1 mutation carriers, although the risk is not as well established as it is with BRCA2 gene mutations. In addition, the PALB2 gene is a breast cancer susceptibility gene that encodes a BRCA2-interacting protein; patients with inherited germline PALB2 mutations are also at a higher risk for pancreatic cancer. (See "Cancer risks and management of BRCA1/2 carriers without cancer", section on 'Pancreas' and "Familial risk factors for pancreatic cancer and screening of high-risk patients", section on 'Hereditary breast cancer: BRCA and PALB2' and "Overview of hereditary breast and ovarian cancer syndromes", section on 'PALB2'.)

Cells that lack functioning BRCA1 or BRCA2 have a deficiency in the repair of deoxyribonucleic acid (DNA) double-strand breaks. There is accumulating evidence of increased sensitivity to platinums and poly (ADP-ribose) polymerase inhibitors in BRCA associated breast and ovarian cancers. (See "Medical treatment for relapsed epithelial ovarian, fallopian tube, or peritoneal cancer: Platinum-sensitive disease", section on 'PARP inhibition in BRCA carriers' and "ER/PR negative, HER2-negative (triple-negative) breast cancer", section on 'Germline BRCA mutation'.)

Whether these findings apply to adjuvant treatment of pancreatic cancers that arise in the setting of a BRCA or PALB2 mutation is unclear. At least some retrospective data support the view that exposure to a platinum-based perioperative chemotherapy regimen (cisplatin, oxaliplatin, or carboplatin) might confer a survival benefit for patients with resected pancreatic ductal adenocarcinoma and a pathogenic BRCA or PALB2 mutation over the use of non-platinum-containing chemotherapy [41,42]. However, these results are, at best, hypothesis generating, and the optimal chemotherapy combination to use in these patients is not yet established.

Updated consensus-based guidelines from the NCCN suggest consideration of a platinum-based regimen for advanced-stage pancreatic cancer that arises in the setting of a BRCA mutation and in the neoadjuvant setting [4], but there are no guidelines for the adjuvant setting. Nevertheless, FOLFIRINOX is the reference standard for adjuvant therapy and represents a reasonable choice for BRCA mutant patients who are able to tolerate it.

Stage 1 disease — Because of the relatively poor five-year OS rates for stage I pancreatic ductal adenocarcinoma (figure 2) [43] adjuvant chemotherapy is widely considered to represent a standard of care in this subgroup [4,6,8,9], although benefit for resected, subcentimeter, stage IA pancreatic cancer has been questioned [44].

The evidence to support benefit from adjuvant chemotherapy in patients with stage I pancreatic cancer is less robust than for stage II or III disease:

●Although the majority of the randomized trials of adjuvant chemotherapy included those with stage I disease, few of these patients were included in the pivotal randomized adjuvant therapy trials with a surgery alone control group (eg, the CONKO-001 and JSAP-02 trials enrolled a total of 14 and 12 patients with stage I disease, representing 4 and 10 percent of the total populations, respectively [29,45]) [46]. Unplanned subgroup analysis in the CONKO-001 trial suggested a comparable magnitude of benefit for T3/4 tumors (HR 0.78, 95% CI 0.61-0.99) as with T1/2 tumors (previously staged as stage IA and IB, respectively in prior staging systems), but the confidence intervals were wider, and the result was no longer statistically significantly overlapped (HR for death 0.58, 95% CI 0.30-1.10) [29]. The test for interaction was not reported. Thus, it cannot be concluded that adjuvant chemotherapy did not benefit those with stage I disease.

●In a retrospective study that included 3909 patients with stage I pancreatic cancer treated between 2006 and 2012, both adjuvant chemotherapy (received by 45.5 percent) and CRT (received by 20 percent) were associated with a survival benefit for those with T1N0 disease in multivariate analysis [47]. However, patients who received adjuvant therapy were also likely to be younger and have tumors in the pancreatic head rather than in the body/tail, both differences that could have biased the results in favor of better outcomes in this group.

Older individuals — Similarly, the evidence to support benefit from adjuvant chemotherapy in older adults is limited. While some of the trials described above describe the proportion of patients older than 65 (ie, 62 percent in the CONKO-001 phase III trial [29]), the representation of patients 80 years of age and older is not well described. However, support for the benefit of chemotherapy in these patients is provided by a retrospective analysis of data from the NCDB. Between 2004 and 2016, there was a 19 percent increase in the use of adjuvant chemotherapy for individuals 80 years of age and older, but it was still administered to fewer than 50 percent of individuals undergoing pancreaticoduodenectomy for stage I to III pancreatic ductal adenocarcinoma [48]. In a multivariate analysis, receipt of adjuvant chemotherapy (HR 0.72, 95% CI 0.65-0.79), female sex, and surgery in more recent time period (2011 or beyond) were associated with a decreased hazard of death.

Chemotherapy dosing in obese patients — For cancer patients with a large body surface area, chemotherapy drug doses are often reduced because of concern for excess toxicity. However, there is no evidence that fully dosed obese patients experience greater toxicity from chemotherapy for pancreatic cancer. Guidelines from ASCO recommend that full weight-based cytotoxic chemotherapy doses be used to treat obese patients with cancer, particularly when the goal of treatment is cure [49]. (See "Dosing of anticancer agents in adults", section on 'Dosing for overweight/obese patients'.)

Chemoradiotherapy — For most patients who received either gemcitabine alone or gemcitabine plus capecitabine for adjuvant therapy, we also suggest the addition of CRT to chemotherapy. The tolerability and benefit of adjuvant CRT in patients who undergo adjuvant FOLFIRINOX chemotherapy are unclear. Until further information is available, we do not offer adjuvant CRT to all individuals who have had a six-month adjuvant course of FOLFIRINOX, but instead, we reserve this approach for selected patients who desire aggressive therapy and retain an excellent performance status after surgery and chemotherapy.

Our approach is consistent with guidelines from ASCO, which suggest adding postoperative CRT to six months of adjuvant gemcitabine-based chemotherapy for patients with node-positive or margin-positive disease [6], and the American Society for Radiation Oncology, which provide a strong recommendation for CRT following four to six months of systemic chemotherapy after resection [50]. Updated 2019 guidelines from the NCCN suggest chemotherapy alone (gemcitabine plus capecitabine, FOLFIRINOX) as the preferred regimen for adjuvant therapy, but they include induction chemotherapy followed by CRT as an acceptable option for adjuvant therapy [4]. ESMO suggests chemotherapy alone in this setting [9].

Benefit, international practice, and rationale — The benefit of adjuvant RT in the adjuvant setting, especially in patients who receive adjuvant FOLFIRINOX chemotherapy, is controversial, and the approach differs globally (see 'Does chemoradiotherapy add benefit to chemotherapy?' below):

●Most European clinicians advocate chemotherapy alone, emphasizing the survival benefit of chemotherapy in the German CONKO-001 trial, the lack of a significant survival benefit with CRT in the EORTC trial, and the detrimental impact of CRT on survival seen in the ESPAC-1 trial. The most recent (2015) guidelines for treatment of pancreatic adenocarcinoma suggest that CRT in the adjuvant setting should only be undertaken within the context of a randomized controlled trial [9]. (See 'Fluorouracil-based approaches' below.)

●Similarly, the Japanese approach also includes chemotherapy alone [26].

●On the other hand, the American approach, which more often includes CRT in addition to adjuvant chemotherapy, emphasizes the following points (see 'Fluorouracil-based approaches' below):

•The high risk of local failure after resection of pancreatic cancer and the potential for benefit from CRT.

•The high rate of positive retroperitoneal margins and the adverse impact of this finding on survival.

•The survival benefit from CRT in the Gastrointestinal Tumor Study Group (GITSG) study. (See 'Fluorouracil-based approaches' below.)

•The trend toward improved survival seen with CRT in the underpowered EORTC study.

•The serious design flaws of the ESPAC-1 trial and the inherent difficulty in drawing definitive conclusions from this study.

The rationale for adding RT to the postoperative adjuvant strategy is that failure patterns often include local recurrence in patients treated with resection alone or adjuvant chemotherapy, and local RT may improve local control [51-53]. In the Fédération Francophone de Cancérologie Digestive (FFCD)/GERCOR trial, the addition of CRT to postoperative gemcitabine reduced the rate of local recurrence alone at first progression [54]. This trial is discussed in more detail below. (See 'Does chemoradiotherapy add benefit to chemotherapy?' below.)

However, not all studies have demonstrated improvements in local control with the use of postoperative combined modality therapy [28,55]. Furthermore, randomized trials and meta-analyses [56] have failed to confirm any survival benefit from the addition of RT. As a result, the benefit of RT has become controversial, even within the United States. (See 'Does chemoradiotherapy add benefit to chemotherapy?' below.)

Fluorouracil-based approaches

●Early trials

•In a study conducted in the late 1970s and early 1980s by GITSG patients with resected pancreatic cancer were randomly assigned to either observation or external beam RT (40 Gy) plus concurrent bolus FU (500 mg/m² per day on the first three and last three days of RT) followed by maintenance chemotherapy (FU 500 mg/m² per day for three days monthly) for two years or until disease progression [57].

Despite the relatively low RT dose, the small number of patients (43) and the fact that 25 percent of the patients on the treatment arm did not begin postoperative treatment until more than 10 weeks following resection, patients receiving postoperative CRT had significantly longer median OS (20 versus 11 months) and a doubling of the two-year survival rate (20 versus 10 percent). Following closure of the study, an additional 32 patients were registered on the combined modality arm, and a subsequent report that included these and the original 43 patients confirmed the initial survival benefit [58].

•In an effort to reproduce these findings, an EORTC study randomly assigned 114 patients with resected pancreatic cancer to postoperative concurrent FU (25 mg/kg per day by continuous infusion) plus external beam RT (40 Gy in split courses) or to observation [55]. In contrast to the GITSG findings, there was only a nonstatistically significant trend toward improved two-year survival for CRT (26 versus 34 percent), and no reduction in locoregional recurrence.

However, this trial was also flawed: RT was delivered in a split-course manner (potentially allowing for tumor repopulation between courses), the dose was suboptimal, and there was no prospective assessment of the completeness of surgical margins. Furthermore, 20 percent of patients randomized to treatment never received it.

●ESPAC-1 – As noted above, the ambitious ESPAC-1 trial initially set out to randomize patients to a two-by-two factorial design in which the relative benefits of adjuvant chemotherapy, CRT, or CRT followed by chemotherapy would be compared with observation alone. However, fear of poor accrual led the investigators to permit the clinician to choose from this or two other randomization schemes. (See 'Early trials demonstrating benefit of adjuvant therapy' above.)

As discussed above, the final results were presented in two separate publications:

•In an initial pooled analysis of the three parallel randomized trials [27], there was no survival difference when the 175 patients who received postoperative CRT were compared with the 178 who did not receive it (median OS 15.5 versus 16.1 months, respectively).

•However, in a subsequent intent-to-treat analysis of the 289 patients randomized in the four-arm study, there was a trend toward worse survival for the group receiving CRT (two- and five-year survival rates 29 versus 41 and 10 versus 20 percent for the CRT and no CRT groups, respectively) [28].

Many European clinicians cite these data as a main reason for not recommending concomitant CRT after resection of pancreatic cancer. In contrast others, especially American clinicians, consider the study's flaws (including the lack of radiation quality control and because it is the only study showing worse outcomes with CRT, whereas all others show either equivalence or an advantage to RT) to preclude any ability to draw firm conclusions regarding the benefit of CRT.

Gemcitabine-based approaches — Preliminary data support the tolerability and favorable short-term outcomes of regimens that use gemcitabine as a radiation sensitizer [54,59-61]; but no trials have compared this approach with CRT using FU as the radiation sensitizer, at least in the postoperative setting. Thus, we consider FU-based CRT to represent a standard approach. A randomized phase II trial comparing gemcitabine-based CRT with gemcitabine alone is discussed below. (See 'Does chemoradiotherapy add benefit to chemotherapy?' below.)

Oral fluoropyrimidines as a substitute for infusional FU — The substitution of capecitabine (or where available, S-1) for infusional FU is reasonable in patients for whom ambulatory infusional FU therapy using a pump is not feasible.

Accumulating data from uncontrolled trials support the view that oral capecitabine can safely replace infusional FU as a radiation sensitizer in patients treated for locally advanced pancreatic cancer, although there are no data from the adjuvant setting. However, extrapolating from these and other data from phase III studies comparing capecitabine with infusional FU during neoadjuvant CRT for rectal cancer, many investigators feel that substituting capecitabine for infusional FU as a radiation sensitizer is reasonable for other gastrointestinal malignancies and that the question is not worthy of phase III studies in each tumor type. We agree with this point of view. (See "Neoadjuvant therapy for rectal adenocarcinoma", section on 'Fluoropyrimidines' and "Initial chemotherapy and radiation for nonmetastatic, locally advanced, unresectable and borderline resectable, exocrine pancreatic cancer", section on 'Radiation sensitizer during RT'.)

Does chemoradiotherapy add benefit to chemotherapy?

●Randomized trials – Few trials have directly compared chemotherapy with and without CRT as an adjuvant strategy. Besides the ESPAC-1 trial (discussed above), the only other trial to compare the relative benefits of adding CRT to systemic therapy versus systemic therapy alone was the EORTC 40013/FFCD 9203/GERCOR phase II study [54]. Ninety patients with resected pancreatic cancer (70 percent node positive, 97 percent R0 resected) were randomly assigned to gemcitabine-based CRT (two cycles of weekly gemcitabine alone [1000 mg/m² weekly, three weeks on, one week off] followed by RT [50.4 Gy in 28 daily 1.8 Gy fractions] with concurrent gemcitabine [300 mg/m² once weekly, four hours before RT, for five to six weeks]) or a control group. Initially, the control group was observation alone (n = 4), but the protocol was amended, and the remainder of the control group (n = 41) received four cycles of gemcitabine alone (1000 mg/m² for three consecutive weeks followed by a one-week rest). Treatment started within eight weeks of surgery.

In contrast to the results of the ESPAC-1 analysis, the addition of CRT to postoperative chemotherapy was not deleterious. Relative to postoperative chemotherapy alone, CRT plus postoperative chemotherapy reduced the rate of local recurrence alone at first progression (11 versus 24 percent), with similar DFS (median 12 versus 11 months) and OS (median 24 months each). By contrast, the rates of simultaneous local and distant progression for CRT plus chemotherapy versus chemotherapy alone were 20 and 13 percent respectively, with similar rates of distant progression in both treatment arms (40 versus 42 percent). (See 'Fluorouracil-based approaches' above.)

●Meta-analysis — The benefits of CRT with and without chemotherapy were addressed in a 2013 Bayesian network meta-analysis of nine randomized trials comparing six different adjuvant strategies (observation alone, FU alone, gemcitabine alone, CRT alone, CRT followed by FU, and CRT followed by gemcitabine) [56]. The following results were noted:

•A statistically significant survival benefit for adjuvant CRT could not be shown. Compared with observation alone, the HRs for death were 0.91 (95% CI 0.55-1.46) for CRT alone, 0.54 (95% CI 0.15-1.80) for CRT plus FU, and 0.44 (95% CI 0.10-1.81) for CRT plus gemcitabine. However, the very wide confidence intervals reflect a lack of precision for these estimates.

•When CRT plus chemotherapy (FU or gemcitabine) was compared with chemotherapy alone (FU or gemcitabine), the point estimates of the HRs for survival all favored the CRT arms; however, the confidence intervals were all wide, reflecting a lack of precision. As an example, in the comparison of CRT plus gemcitabine versus gemcitabine alone, the HR for survival was 0.65 (95% CI 0.14-2.70). These results include the possibility of an 86 percent reduction and a 2.7-fold increase in the risk of death with the addition of CRT. It is difficult to draw any meaningful conclusions from these data.

Additional information on the benefits of chemoradiotherapy in conjunction with adjuvant chemotherapy should be forthcoming from RTOG trial 0848 although the chemotherapy regimen is not FOLFIRINOX (gemcitabine with or without erlotinib followed by the same chemotherapy regimen with or without RT and capecitabine or fluorouracil after potentially curative resection).

NEOADJUVANT THERAPY — We consider neoadjuvant therapy to be a reasonable alternative to upfront surgery followed by adjuvant therapy in patients who appear to have potentially resectable tumors, especially if there is any radiographic interface between the tumor and the mesenteric vasculature, as long as performance status and comorbidity are sufficient to tolerate treatment. Before recommending this approach over upfront surgery, we await additional data to clarify which patients might benefit the most from preoperative therapy as opposed to those for whom it is best to undertake upfront resection followed by adjuvant therapy.

Rationale and benefit over initial resection — As noted above, the rationale for neoadjuvant chemotherapy is as follows (see 'Approach to the patient' above):

●Improved selection of patients for whom resection will not offer a survival benefit (ie, those who rapidly progress to metastatic disease during preoperative therapy).

●Higher rates of margin-negative resection, which is the major goal of surgery.

●Early treatment of micrometastases, which are presumed to be present even without evidence of distant metastases on initial imaging, and responsible for poor long-term outcomes from initial surgical treatment of apparently localized disease.

●Greater chance of receiving effective perioperative chemotherapy, given that prolonged recovery from upfront resection prevents the delivery of postoperative adjuvant chemotherapy in approximately one-fourth of patients [62]. (See 'Chemotherapy' above.)

Neoadjuvant therapy for potentially resectable pancreatic cancer is in evolution. Several case series [63], retrospective analyses [64,65], phase II studies [66,67], an early meta-analyses [63], and a pooled analysis of three randomized trials (all prematurely discontinued and together totaling only 130 patients) [68], all suggest potential benefits over initial resection, but selection bias and patient heterogeneity (most studies included patients with initially resectable, borderline resectable and locally advanced unresectable tumors) preclude definitive conclusions. Three small randomized phase II trials limited to patients with potentially resectable tumors failed to demonstrate a meaningful survival benefit using chemotherapy with or without chemoradiotherapy (CRT) [69-71]. However, most of these studies utilized older chemotherapy regimens (gemcitabine alone, gemcitabine plus cisplatin), mostly with FU-based CRT.

More recently, accumulating data from an updated analysis of the phase III PREOPANC trial, an early report of JSAP05 trial, and a newer meta-analysis provide additional support for the benefit of neoadjuvant therapy in combined populations of patients with both potentially resectable and borderline resectable pancreatic cancer:

●PREOPANC trial – The phase III PREOPANC trial randomly assigned 248 patients with potentially resectable or borderline resectable (approximately 50 percent each) pancreatic cancer to upfront surgery followed by six months of adjuvant gemcitabine or to gemcitabine-based CRT (two weekly doses of gemcitabine followed by one week of rest, gemcitabine-based CRT [36 Gy in 15 fractions with gemcitabine 1000 mg/m² on days 1, 8, and 15], and then two weekly doses of gemcitabine alone) followed by surgery and four months of gemcitabine alone [72]. Resectable disease was strictly defined as no contact with the superior mesenteric artery, celiac trunk, or common hepatic artery, and ≤90 degrees contact with the superior mesenteric vein.

The R0 resection rate was significantly higher with preoperative therapy (71 versus 40 percent), and this group also had significantly better disease-free survival and locoregional failure-free interval as well as significantly lower rates of pathologic lymph nodes, perineural invasion, and venous invasion. In the latest analysis at a median follow-up of 59 months, the difference in median survival between the neoadjuvant and adjuvant therapy groups was relatively small but statistically significant (median 15.7 versus 14.3 months, hazard ratio [HR] for death 0.73, 95% CI 0.56-0.96) [3]. The five-year overall survival (OS) rate with neoadjuvant therapy was 20.5 percent compared with 6.5 percent in the control group. In a preplanned subgroup analysis, the benefits of neoadjuvant therapy were similar for resectable (HR 0.79, 95% CI 0.54-1.16) and borderline resectable disease (HR 0.67, 95% CI 0.45-0.99).

●JSAP-05 trial – Additional support for neoadjuvant chemotherapy is provided by the multicenter Japanese Prep-02/JSAP-05 trial, which randomly assigned 364 patients with potentially resectable pancreatic cancer to two preoperative courses of gemcitabine (1000 mg/m² on day 1 and 8) plus S-1 (40 mg/m² twice daily on days 1 to 14) versus upfront surgery; patients in both groups who were completely resected and fully recovered by 10 weeks postoperation also received six months of therapy with S-1 alone [73]. In a preliminary report, despite no improvement in overall and R0 resection rates, patients receiving neoadjuvant chemotherapy had significantly longer median OS (36.7 versus 26.6 months).

●Meta-analysis – A meta-analysis included seven trials (totaling 938 patients) in which the comparator arm was any form of neoadjuvant therapy (including gemcitabine and gemcitabine/cisplatin-based CRT) compared with upfront surgery followed by adjuvant gemcitabine (none of the trials used adjuvant FOLFIRINOX) for potentially or borderline resectable pancreatic cancer [74]. Overall neoadjuvant therapy improved OS compared with upfront surgery (HR for death 0.66, 95% CI 0.52-0.85), representing an increase in median survival from 19 to 29 months. In an unplanned subgroup analysis, the OS benefit was limited to those patients with borderline resectable tumors (HR for death 0.61, 95% CI 0.44-0.85), and not those with potentially resectable disease upfront. However, a test for interaction was not provided.

Guidelines from expert groups — Given the evolving role of neoadjuvant therapy in potentially resectable pancreatic cancer, it is not surprising that guidelines from expert groups are conflicting:

●Year 2016 guidelines on management of potentially curable pancreatic cancer from the American Society of Clinical Oncology (ASCO) [6] suggest not administering neoadjuvant chemotherapy or CRT to most patients with potentially resectable tumors in the absence of radiographic interface with the mesenteric vessels, unless in the context of a clinical trial. However, they also state that chemotherapy could be used before surgery for patients who had potentially anatomically resectable but high-risk tumors (as judged by elevated preoperative levels of carbohydrate antigen 19-9 [CA 19-9]).

●Updated 2021 guidelines from the National Comprehensive Cancer Network also suggest that this approach could be considered in high-risk potentially resectable tumors (ie, those with concerning imaging findings, very elevated CA 19-9 levels, large primary tumors, or large regional lymph nodes, and in patients with excessive weight loss and extreme pain), however, there are no specific guidelines as to the threshold for any of these factors to define a patients as having "high-risk" disease.

●The European Society for Medical Oncology 2015 guidelines do not address the benefit of neoadjuvant therapy for potentially resectable pancreatic cancer [9].

Important pretreatment considerations — Management of nonmetastatic pancreatic cancer is in evolution, but increasingly, initial chemotherapy using combination regimens is favored, despite the paucity of high-quality trials demonstrating benefit in patients with potentially resectable tumors. (See 'Neoadjuvant therapy' above.)

Important pretreatment considerations for individuals who will be treated initially with chemotherapy include the following:

●A tissue diagnosis should be established before initiation of therapy. This differs from a surgery-first strategy, in which preoperative tissue diagnosis may not be needed. (See "Clinical manifestations, diagnosis, and staging of exocrine pancreatic cancer", section on 'Diagnostic algorithm and need for preoperative biopsy'.)

●For patients who present with obstructive jaundice, delivery of neoadjuvant therapy necessitates durable biliary decompression for as many as six months. This can usually be accomplished through placement of a biliary stent. This subject, which includes a discussion of different types of stents, is provided elsewhere. (See "Surgical resection of lesions of the head of the pancreas", section on 'Preoperative biliary drainage' and "Endoscopic stenting for malignant biliary obstruction" and "Supportive care for locally advanced or metastatic exocrine pancreatic cancer", section on 'Jaundice'.)

Choice of neoadjuvant strategy — If it is chosen, the best regimen for neoadjuvant therapy is not currently defined; there are few comparator trials. For most patients who are able to tolerate it, we suggest therapy with the multiagent modified FOLFIRINOX (oxaliplatin plus irinotecan with leucovorin and short-term infusional fluorouracil [FU]) regimen, followed by CRT in the absence of distant metastases, rather than a gemcitabine-based regimen. (See 'FOLFIRINOX plus chemoradiotherapy' below.)

Most of the reported studies of neoadjuvant therapy have utilized CRT alone or chemotherapy plus CRT Experience is more limited with neoadjuvant chemotherapy alone. (See 'Induction chemotherapy alone' below.)

Chemotherapy plus chemoradiotherapy — Although early reports of preoperative radiation therapy (RT) with or without concurrent FU demonstrated that this approach did not worsen perioperative morbidity and mortality, there was no obvious improvement in either resectability or OS [70,75-79]. A possible limitation of these initial studies is that most used single-agent bolus FU. Subsequent studies have focused on improving the treatment regimen by optimizing chemotherapy. Two specific approaches are FOLFIRINOX plus FU-based CRT, and gemcitabine-based CRT.

FOLFIRINOX plus chemoradiotherapy — There are no dedicated trials examining the benefit of neoadjuvant FOLFIRINOX (table 2) plus CRT that are limited to individuals with potentially resectable tumors. The following data are available to inform this issue:

●In the setting of borderline resectable pancreatic cancer, neoadjuvant therapy using FOLFIRINOX in conjunction with CRT results in a high rate of R0 resection and prolonged median progression-free survival and OS compared with historical controls [80]. (See "Initial chemotherapy and radiation for nonmetastatic, locally advanced, unresectable and borderline resectable, exocrine pancreatic cancer", section on 'Benefit of neoadjuvant therapy'.)

●Additional information is available from a retrospective analysis of data on 1835 consecutive patients presenting with localized pancreatic cancer (52 percent locally advanced, 29 percent borderline resectable, 19 percent potentially resectable) who received at least one cycle of modified FOLFIRINOX chemotherapy as initial treatment in one of five referral centers in the United States and Netherlands between 2012 and 2019 [81]. The median number of chemotherapy cycles was 6 (interquartile interval 4 to 8), and subsequent treatment included radiotherapy in 49 percent, and resection in 38 percent. The margin-negative (R0) resection rate for those with borderline or potentially resectable tumors was 63 and 79 percent, respectively, and the corresponding median OS was 23.2 and 31.2 months, respectively, and the five-year OS rates were 18.4 and 33.7 percent, respectively.

The lack of a control group in studies such as these limits the assessment of whether the results are better than those that could be achieved using upfront resection followed by adjuvant FOLFIRINOX. (See 'Modern combination regimens (FOLFIRINOX and gem-nabpaclitaxel)' above.)

However, rates of R0 resection have historically been much lower with initial surgery for borderline resectable disease than are reported in these studies [82]. Furthermore, as noted above, some centers, including ours, have taken the view that nearly all patients who appear to have "potentially resectable" pancreatic cancer should be considered to have "borderline resectable" disease given the inaccuracy of imaging and the very high rates of a positive margin with initial surgery, which is associated with poor OS. As a result, we consider that these results are applicable to patients with potentially resectable disease. (See 'Initial resection versus neoadjuvant therapy' above.)

Gemcitabine-based chemoradiotherapy — While there are no trials directly comparing FOLFIRINOX plus fluoropyrimidine-based CRT versus gemcitabine-based CRT, for most patients who are able to tolerate it, we prefer FOLFIRINOX plus CRT.

Gemcitabine-based CRT may provide an enhanced local effect, although with the potential for more toxicity than FU-based regimens [83-86]. Preoperative gemcitabine-based CRT was evaluated in a phase III trial (PREOPANC), which is discussed separately [72]. (See 'Rationale and benefit over initial resection' above.)

Induction chemotherapy alone — Another option for neoadjuvant therapy is chemotherapy alone. Several randomized trials have addressed the benefit of neoadjuvant chemotherapy without CRT in potentially resectable pancreatic cancer:

●In a randomized phase II trial (NEONAX), 127 patients with resectable pancreatic cancer were treated with six cycles of gemcitabine plus nabpaclitaxel either in the perioperative (two preoperative [neoadjuvant] and four postoperative [adjuvant] cycles) or postoperative setting [33]. Among the modified intention to treat population (patients with an R0/R1 resection who started either neoadjuvant or adjuvant chemotherapy), disease-free survival (DFS) was similar for both preoperative and postoperative therapy (18-month DFS 33 versus 41 percent), but neither arm met the prespecified primary endpoint (18-month DFS of 55 percent). In the entire study population, for perioperative versus postoperative chemotherapy, median DFS was 12 versus 6 months and median OS was 26 versus 17 months, but these differences were not statistically significant

Perioperative chemotherapy did demonstrate other clinical benefits. For example, the R0 resection rate was higher for perioperative versus postoperative chemotherapy (88 versus 67 percent). Patients who received perioperative therapy were also more likely to be exposed to chemotherapy; a majority of these patients completed neoadjuvant treatment (90 percent) and approximately 42 percent also completed postoperative therapy, whereas less than half of the patients assigned to postoperative therapy alone started treatment (42 percent).

●A randomized phase II trial assigned 93 patients (88 eligible for analysis) to receive surgery followed by six months of single-agent adjuvant gemcitabine (group A); surgery followed by six months of adjuvant cisplatin, epirubicin, gemcitabine, and capecitabine (PEGX; group B); or three cycles of PEGX before and after surgery (group C) [67]. At a median follow-up of 55 months, the event-free (defined as freedom from progression, relapse, new tumor occurrence, distant metastasis, or death) survival in group A was 23 percent (95% CI 7-39 percent); in group B, it was 50 percent (95% CI 32-68 percent); and in group C, it was 66 percent (95% CI 49-83 percent). The corresponding values for three-year survival were 35, 43, and 55 percent, respectively. The R0 resection rates were 27 and 37 percent in both the adjuvant therapy groups, while it was 63 percent in group C.

The trial was initially conceived as a combined phase II/III trial, but it was decided not to continue with the phase III component for two reasons: because the standard of care for adjuvant therapy had changed and because of the existence of newer chemotherapy regimens for which the evidence supporting benefit, at least in the setting of metastatic disease, was stronger than that for the PEGX regimen.

●The multicenter Japanese Prep-02/JSAP-05 trial randomly assigned 364 patients with potentially resectable pancreatic cancer to two preoperative courses of gemcitabine plus S-1 versus upfront surgery; patients in both groups who were completely resected and fully recovered by 10 weeks postoperation also received six months of therapy with S-1 alone [73]. In a preliminary report, despite no improvement in overall and R0 resection rates, patients receiving neoadjuvant chemotherapy had significantly longer median OS (36.7 versus 26.6 months).

Is there an optimal regimen? — SWOG S1505 directly compared perioperative (12 weeks preoperative and 12 weeks postoperative) treatment with FOLFIRINOX versus gemcitabine plus nabpaclitaxel (gem/nab) in 102 patients with potentially resectable pancreatic cancer [87]. Resectability rates were high with both regimens (73 versus 70 percent), although more patients receiving gem/nab had a complete or major pathologic response (42 versus 25 percent). Two-year OS, the primary endpoint, was 47 percent with FOLFIRINOX, and 48 percent with gem/nab, and neither arm met the prespecified endpoint (58 percent two-year survival). Interpretation of these results is limited by the high rate of ineligibility after central radiologic review (44 of the 147 enrolled patients). This trial raises the obvious need for biomarkers to determine which regimen is best for patients with pancreatic cancer as the two regimens have different mechanisms of action.

Additional information is expected from the Dutch PREOPANC-2 trial comparing neoadjuvant FOLFIRINOX to gemcitabine-based CRT for borderline and potentially resectable pancreatic cancer [88]; accrual was completed in January 2021, and the data are maturing.

Role of adjuvant therapy after neoadjuvant therapy — For patients with pancreatic cancer who received preoperative (neoadjuvant) therapy, we suggest additional postoperative (adjuvant) systemic therapy so that patients receive a total of six months of systemic chemotherapy, including the preoperative regimen. This recommendation is based on an extrapolation of data from adjuvant therapy trials and is generally consistent with ASCO guidelines. (See 'Chemotherapy' above.)

If protocol therapy is not available or declined, for patients with an excellent performance status who are able to tolerate it, we suggest modified FOLFIRINOX (oxaliplatin plus irinotecan with leucovorin and short-term infusional FU) rather than gemcitabine-based regimens. For less fit patients, options include gemcitabine plus capecitabine, gemcitabine plus nabpaclitaxel, gemcitabine alone, or, where available, S-1 alone [89].

For patients who receive preoperative (neoadjuvant) therapy, there are limited data from randomized trials to guide the administration of postoperative (adjuvant) therapy. These data are discussed separately. (See 'Induction chemotherapy alone' above and 'Adjuvant therapy' above.)

While observational studies are conflicting, most suggest some overall survival benefit with adjuvant therapy in patients treated with neoadjuvant therapy followed by surgical resection [64,90,91]. However, the degree of benefit varies based on tumor characteristics such as tumor (T) stage, tumor differentiation, nodal involvement, and margin status [91].

POST-TREATMENT SURVEILLANCE — Patients who have completed treatment for potentially resectable pancreatic cancer should be monitored for recovery of treatment-related toxicities and disease recurrence [6].

There is no evidence to guide the post-treatment surveillance strategy in patients with pancreatic cancer, and clinical practice is variable, particularly with regard to CT scanning [92,93].

We suggest a periodic history and physical examination every three to six months for at least two years. If carbohydrate antigen 19-9 (CA 19-9) levels were initially elevated, we follow them every three to six months for at least two years and, if elevated, perform a CT scan. An elevated post-treatment CA 19-9 increases the risk for recurrence, with the highest risks in those with persistent elevations from diagnosis through follow-up [94,95]. An important point is that mild elevations in CA 19-9 can occur with biliary tract dysfunction, which often occurs in patients after a pancreaticoduodenectomy. Thus, noncancerous causes can lead to abnormal results in these patients.

Given that there is no known curative treatment for patients with recurrent pancreatic adenocarcinoma, many clinicians do not routinely obtain CT scans in the follow-up of patients following resection of a pancreatic cancer unless indicated by symptoms or a rising CA 19-9 level. Patients with metastatic pancreatic cancer inevitably develop symptoms and will present for treatment discussions. Additionally, it is not clear that early initiation of therapy in asymptomatic individuals is associated with a survival benefit, thereby calling into question any surveillance technique. This approach is consistent with guidelines from the American Society of Clinical Oncology (ASCO) [6]. However, others disagree with this recommendation and follow the consensus-based guidelines of the National Comprehensive Cancer Network (NCCN) and the European Society for Medical Oncology (ESMO), which recommend periodic radiographic surveillance. One advantage to CT follow-up is that patients will be detected prior to deterioration of performance status or laboratory abnormalities that might preclude clinical trial enrollment. (See 'Guidelines from expert groups' below.)

The majority of recurrences after potentially curative treatment of exocrine pancreatic cancer occur within two years, and they can be locoregional or at distant sites, most often the liver, lung, and peritoneal cavity [96]. In one autopsy series of patients with known pancreatic cancer, approximately 30 percent died with locally destructive disease without evidence of metastasis, while 70 percent died with widespread metastatic disease [51].

The primary goal of surveillance after curative treatment for any cancer is to detect local or distant recurrence when available interventions can prolong survival. However, for pancreatic cancer, the vast majority of recurrences are not amenable to potentially curative therapy, and the evidence that early identification of recurrent or metastatic disease in asymptomatic patients improves long-term survival is limited:

●In a study of 216 patients with pancreatic cancer who developed a postoperative recurrence during the course of a surveillance regimen that consisted of physical examination, CA 19-9 assay, chest radiograph, and abdominal CT every three to four months for the first two years after surgery, then every six months until five years, and then annually, the surveillance regimen was able to detect an asymptomatic recurrence in 55 percent of patients [97]. Although the median time to recurrence was not different between patients with a symptomatic recurrence and patients with an asymptomatic recurrence, median survival was significantly less in symptomatic patients (5.1 versus 13 months). Asymptomatic patients were much more likely to receive treatment after a recurrence was identified.

●On the other hand, analysis of a large national database demonstrated no significant survival benefit for annual radiographic surveillance among patients with curatively resected pancreatic cancer [98].

Despite the lack of a consistent survival benefit, secondary benefits for detection of metastatic disease at an asymptomatic stage include early introduction of palliative chemotherapy or radiation therapy to slow disease progression and, for patients who can tolerate an aggressive regimen such as FOLFIRINOX (oxaliplatin plus irinotecan with leucovorin and short-term infusional fluorouracil), improve survival.

Guidelines from expert groups — Recommendations for post-treatment surveillance are available from several expert groups, and not surprisingly, they differ:

●ASCO guidelines for management of potentially resectable pancreatic cancer suggest visits at three- to six-month intervals to monitor for recovery of treatment-related toxicities and recurrence, with tapering of visits after two years [6]. If elevated preoperatively, serum CA 19-9 levels can be measured every three to six months as well. The benefit of surveillance imaging is less clear, and at best, it seems to result in greater detection of asymptomatic recurrence. The timing of imaging should be determined after discussion with the individual patient, taking into account their preferences, level of emotional stress, and concern regarding financial burden.

●Consensus-based guidelines from the NCCN recommend a history and physical examination for symptom assessment, and CA 19-9 determinations and follow-up CT scans every three to six months for two years, then every 6 to 12 months, with a low level of evidence but uniform consensus [4].

●Consensus-based guidelines from ESMO state that there is no evidence that regular follow-up after initial therapy with curative intent has any impact on outcome [9]. They suggest that follow-up visits concentrate on symptoms, nutrition, and psychosocial support.

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 topic (see "Patient education: Pancreatic cancer (The Basics)")

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

SUMMARY AND RECOMMENDATIONS

●Who has potentially resectable disease? – Patients with potentially resectable exocrine pancreatic cancers are those with no distant metastatic disease, a performance status and comorbidity profile that are appropriate for major abdominal surgery, and no radiographic interface between the primary tumor and the mesenteric vasculature as assessed by cross sectional imaging. (See 'Assessing resectability' above.)

●Upfront pancreatectomy versus neoadjuvant therapy – For patients with potentially resectable exocrine pancreatic cancer, we consider either treatment strategy to be appropriate (algorithm 1); authors and editors of this topic use different approaches. Their rationale for preferring one or the other approach is discussed above. (See 'Initial resection versus neoadjuvant therapy' above.)

●Patients selected for upfront pancreatectomy – The surgical treatment of exocrine pancreatic cancer including preoperative management is discussed separately. (See "Overview of surgery in the treatment of exocrine pancreatic cancer and prognosis".)

•Adjuvant chemotherapy – For all patients who undergo upfront pancreatectomy, including those with pathologic T1N0 disease (table 1), we recommend six months of adjuvant chemotherapy (Grade 1B). (See 'Indications' above.)

Adjuvant treatment should be started within eight weeks of surgery, if possible, and preceded by restaging with CT scans and a serum level of the tumor marker carbohydrate antigen 19-9 (CA 19-9). (See 'Timing and duration' above.)

Off protocol, we suggest the following approach (see 'Choice of therapy' above):

-For patients with an excellent performance status, we suggest modified FOLFIRINOX (oxaliplatin plus irinotecan with leucovorin and short-term infusional fluorouracil (table 2)) rather than a gemcitabine-based regimen (Grade 2C).

-For less fit patients, gemcitabine plus capecitabine (table 5) or gemcitabine plus nabpaclitaxel (table 6) are reasonable alternatives. We reserve therapy with gemcitabine alone (table 7) for patients with a borderline performance status or a comorbidity profile that precludes multiagent therapy. S-1, where available, is another alternative that appears more effective than gemcitabine monotherapy.

•Adjuvant chemoradiotherapy – We do not offer adjuvant chemoradiotherapy (CRT) to individuals who have had a six-month adjuvant course of FOLFIRINOX, but reserve this for selected patients who desire highly aggressive therapy and retain an excellent performance status after surgery and chemotherapy. (See 'Chemoradiotherapy' above.)

For most patients who have received a gemcitabine-based adjuvant chemotherapy regimen, we suggest the addition of concurrent CRT to chemotherapy (Grade 2C).

●Patients selected for neoadjuvant therapy – If neoadjuvant therapy is planned, a tissue diagnosis is required before initiation of therapy. (See 'Important pretreatment considerations' above.)

•Neoadjuvant regimen – For most patients who are able to tolerate it, we suggest the multiagent modified FOLFIRINOX regimen (table 2) followed by CRT (Grade 2C). Other options include gemcitabine-based CRT and chemotherapy without radiation. Surgery could then be reconsidered at a later time if appropriate. (See 'Choice of neoadjuvant strategy' above.)

•Adjuvant therapy – After resection, we suggest a total of six months of systemic therapy, to include the preoperative regimen (Grade 2C). For patients with an excellent performance status, we suggest modified FOLFIRINOX (table 2). For less fit patients, alternatives include gemcitabine plus capecitabine (table 5), gemcitabine plus nabpaclitaxel (table 6), gemcitabine alone or, where available, S-1 alone. (See 'Role of adjuvant therapy after neoadjuvant therapy' above.)

●Post-treatment surveillance – We perform a history and physical examination and check CA 19-9 level (if initially elevated) every six months for two years, then annually. We do not routinely obtain follow-up CT scans after pancreaticoduodenectomy, unless indicated by symptoms or a rising CA 19-9 level. Others disagree, and follow the consensus-based guidelines of the National Comprehensive Cancer Network, which suggest follow-up CT scans every three to six months for at least two years, then every 6 to 12 months. (See 'Post-treatment surveillance' above.)

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges David P Ryan, MD, who contributed to earlier versions of this topic review.