As an initial randomized trial, the Gastrointestinal Tumor Study Group (GITSG) showed a survival benefit with the addition of chemoradiation to surgical resection on interim analysis. Forty-three patients with R0 resection were randomized to chemoradiotherapy versus observation. Twenty-two patients randomized to no adjuvant treatment and 21 to combined therapy were analyzed. Radiotherapy was delivered to 40 Gy in 20 fractions via split course, with a 2-week break after 20 Gy. Bolus 5-fluorouracil (FU) (500 mg/m²) was administered during radiotherapy and then once weekly for 2 years after radiotherapy. Median survival for chemoradiation group with 20 months was significantly longer than that observed for the control group with 11 months. Two-year and 5-year overall survival (OS) were 42% and 14% versus 15% and 4%, respectively [9]. After survival benefit on interim analysis, the results of an additional 30 patients treated with identical chemoradiation regimen showed a similar survival, with median survival of 18 months and a 2-year survival of 46% [10]. Because of this significant improvement in survival, the adjuvant chemoradiotherapy became a standard treatment, particularly in North America. However, the GITSG trial has been criticized for several limitations. The trial was terminated prematurely because of an unacceptably low rate of accrual combined with the observation of increasingly large survival differences between the study groups. The trial accrued only 43 patients over 8 years. Since this trial used the split-course technique with low radiation dose (40 Gy) and used two-dimensional therapy with AP-PA (anteroposterior-posteroanterior) fields, which encompassed the entire pancreas or pancreatic bed and the celiac, pancreaticosplenic, peripancreatic, and retroperitoneal regional lymph nodes, the results may not be applicable to modern radiotherapy practice. Additionally, this trial evaluated the use of adjuvant chemoradiotherapy, but did not address the effect of adding radiation to chemotherapy.

On the other hand, the addition of chemoradiation to surgical resection did not show an overall survival benefit in the European Organization of Research and Treatment of Cancer (EORTC) study. Two hundred eighteen patients with pancreatic head or periampullary cancer were randomized to chemoradiotherapy or observation after surgical resection. The radiation dose of 40 Gy was delivered with split course as in the GITSG trial. Unlike the GITSG trial, concurrent continuous infusional 5-FU (25 mg/kg) with no maintenance chemotherapy was administered. A trend toward improved survival was identified in an analysis including only pancreatic head cancer patients (p = 0.099). The median survival for 81 patients with pancreatic head cancer was 17.1 months in the chemoradiation group versus 12.6 months in the observation group, but this study confirmed the absence of a statistical significant advantage for adjuvant chemoradiotherapy. The median survival was 19.0 months for the observation group and 24.5 months for the chemoradiation group (p = 0.208), and the 2-year survival were 41% and 51%, respectively [2]. With long-term follow-up of median 11.7 years, EORTC trial further confirmed no statistical advantage for adjuvant chemoradiotherapy over observation [11]. However, there are several contributing factors for a lack of survival benefit of adjuvant chemoradiotherapy in this trial. One potential explanation is its heterogeneity of patient population. This trial included patients with both pancreatic and periampullary cancer; it is known that periampullary cancer has a significantly better prognosis compared with pancreatic cancer. In an analysis including only patients with pancreatic cancer, improved survival of adjuvant chemoradiotherapy was shown in this study. And more than 20% of patients in the chemoradiation group did not apply the planned protocol due to postoperative complications or patient refusal. Also, this trial allowed 25% of patients with positive surgical margins and 47% of patients with node positive which carries a worse prognosis. With respect to the radiotherapy, it employed suboptimal therapy with a low dose delivered in a split course similar to the GITSG study. In addition, this study omitted a maintained chemotherapy and included small sample size. The result for discordant survival benefit observed in the EORTC trial as opposed to the GITSG trial was considered as the absence of maintenance of chemotherapy rather than administration of radiotherapy. Therefore, the adjuvant chemotherapy is considered as the standard treatment for patients with resected pancreatic cancer in Europe [12, 13].

The European Study Group for Pancreatic Cancer-1 (ESPAC-1) was a randomized controlled trial, which evaluated the roles of chemoradiotherapy and chemotherapy consisting of four arms. After resection, 285 patients enrolled (1) observation (n = 69), (2) adjuvant chemotherapy alone (n = 74), (3) adjuvant radiation with concurrent chemotherapy (n = 70), or (4) adjuvant radiation with concurrent chemotherapy followed by maintenance chemotherapy (n = 72). In addition, clinicians could choose to randomize the patients either (1) observation versus chemoradiotherapy (n = 68), consisting of 20 Gy over 2 weeks with 5-FU (500 mg/m²) and then repeated after 2-week break, or (2) observation versus chemotherapy (n = 188). The data from the treatment groups of the two-by-two factorial design plus two optioning trials were pooled for analysis. Chemoradiation regimen was similar to those of the GITSG and EORTC trials, but the total radiation dose was either 40 Gy or 60 Gy. Positive margins were allowed, and this trial included 18% of patients with positive margins. Overall results showed no benefit for adjuvant chemoradiotherapy. Median survival was 15.5 months in 175 patients with chemoradiotherapy versus 16.1 months in 178 patients without chemoradiotherapy. There was evidence of a survival benefit for adjuvant chemotherapy [12]. In update of the interim results reporting only 289 patients who underwent randomization using a two-by-two factorial design, median survival of observation, chemotherapy alone, chemoradiation, and chemoradiation followed by chemotherapy was 16.9 months, 21.6 months, 13.9 months, and 19.9 months, respectively. On analysis performed grouping patients who received chemotherapy versus patients who received no chemotherapy, the 5-year survival rate was 21% among patients who received chemotherapy and 8% among patients who did not receive chemotherapy (P = 0.009). On another analysis performed grouping patients who received radiotherapy versus patients who received no radiotherapy, patients who received radiotherapy had a survival detriment compared to those who did not (p = 0.05). The authors concluded that adjuvant chemotherapy had a beneficial effect on survival, but adjuvant chemoradiotherapy had a deleterious effect on survival based on these results [14]. However, the ESPAC-1 trial has been strongly criticized for several problems. Regarding that physicians could choose the randomization arm, the trial design has the potential for selection bias in the enrollment process [15]. Similar to the GITSG and EORTC study, this trial employed suboptimal radiotherapy including outdated radiotherapy regimen using split-course and low radiation dose. Also, there was absence of quality assurance of radiotherapy plans, and radiotherapy field size and technique were not specified. In addition, this trial included patients with uncontrolled and previous therapy substantially as well as a high proportion of noncompliance to the treatment regimens. Only 62% of patients received full course of chemoradiation treatment, and 42% of patients in the chemotherapy arms completed the scheduled regimen, which questions the validity of any analysis and therefore its conclusions. All of these factors could have adversely impacted the outcomes against the chemoradiotherapy arm.

The development of gemcitabine may be considered a major advance in the treatment of pancreatic cancer. Based on the data showing potential benefit to adjuvant chemotherapy, the Charite Onkologie (CONKO)-1 trial was initiated, which evaluated surgery alone versus surgery plus six cycles of gemcitabine-based chemotherapy (1,000 mg/m²). Three hundred sixty-eight patients were enrolled, and patients treated with gemcitabine achieved a statistically significantly lower disease-free survival (DFS) than those observed after surgical resection (13.4 vs. 6.9 months) [16]. To address the role of radiotherapy, the results of the gemcitabine arm of the CONKO-1 trial have been compared to the gemcitabine arm of Radiation Therapy Oncology Group (RTOG) 97-04 including radiotherapy. Given differences in the two trials, such comparisons are statistically invalid, which cannot draw conclusions regarding the benefit of RT in addition to chemotherapy [8]. The most difference between two studies is that the CONKO-1 trial included only patients with carbohydrate antigen (CA) 19-9 serum values of less than 2.5 times normal, whereas RTOG 97-04 did not define an upper limit. When 385 patients were stratified based on CA 19-9 levels (<180 IU/mL vs. ≥180 IU/mL, ≤90 IU/mL vs. >90 IU/mL), there was a significant survival difference favoring patients with CA 19-9 levels of <180 IU/mL [17]. In an analysis of 200 patients with CA 19-9 levels of ≤90 IU/mL, median survival was similar to that observed in the gemcitabine arm of the CONKO-1 trial. Despite the use of radiotherapy and including a higher percentage of patients with positive margins in the RTOG 97-04 trial compared to CONKO-1 trial, local recurrence rates were similar in the gemcitabine arm of both trials.

To determine whether gemcitabine is superior to 5-FU in terms of overall survival as adjuvant treatment, ESPAC-3, phase III randomized controlled trial enrolled 1,088 patients between 2000 and 2007 and underwent at least 2 years of follow-up. Patients received either six cycles of 5-FU (425 mg/m²) plus folinic acid (20 mg/m²) (n = 551) or gemcitabine (1,000 mg/m²) (n = 537). After a median 34.2 months of follow-up, there were no significant differences in either progression-free survival (PFS) or global quality-of-life scores between the treatment groups. However, 14% of patients receiving 5-FU had 97 treatment-related serious adverse events, compared with 7.5% of patients receiving gemcitabine, who had 52 events (P < 0.001). Given its favorable toxicity profile, gemcitabine is considered the standard adjuvant treatment in many parts of Europe [18].

Unlike Europe, the focus of future adjuvant therapy for resectable pancreatic cancer has been chemoradiation in the United States. The RTOG 97-04 evaluated the efficacy of gemcitabine in the adjuvant setting compared to 5-FU, with both regimens followed by chemoradiotherapy. Chemoradiation was provided with 50.4 Gy and continuous 5-FU. Univariate analysis showed no difference in OS between two groups. On analysis of pancreatic head tumor patients (n = 388), a median survival and 5-year OS were 20.5 months and 22% with gemcitabine versus 17.1 months and 18% with 5-FU. Also, patients on the gemcitabine arm with pancreatic head tumors showed a trend toward improved OS on multivariate analysis (P = 0.08). The distant relapse rate was still remained higher over 70% of patients although the local recurrence was half of that reported in previous trials [19, 20].

Currently, EORTC/RTOG 0848 phase III trial evaluates the impact of the small-molecule epidermal growth factor receptor (EGFR), erlotinib, and chemoradiation on OS after completion of a full course of gemcitabine. Patients with resected pancreatic head tumor are randomized to receive treatment with either gemcitabine alone or gemcitabine combined with erlotinib for five cycles. If no progression is seen following the completion of systemic therapy, patients are further randomized either to receive an additional cycle of the previously administered chemotherapy and no further treatment or to receive 50.4 Gy radiation with concurrent capecitabine or 5-FU. This trial was designed to address the issue of high rate of distant metastasis as well as to further define the role of chemoradiotherapy in adjuvant setting.

Two nonrandomized trials from Johns Hopkins Hospital and Mayo Clinic also suggested a survival benefit with adjuvant chemoradiotherapy in pancreatic cancer. In a prospective review from Johns Hopkins Hospital, 616 patients receiving 5-FU-based chemoradiotherapy after resection experienced an improved median survival with 21.2 months versus 14.4 months (P < 0.001). Both 2-year (43.9% vs. 31.9%) and 5-year (20.1% vs. 15.4%) survival were better compared with no adjuvant therapy [21]. Similarly, the Mayo Clinic experience reported the outcomes of 472 patients who underwent complete surgical resection with negative margins between 1975 and 2005. For the 466 surviving patients, median OS after adjuvant chemoradiotherapy was 25.2 versus 19.2 months after no adjuvant therapy (P = 0.001). A 2-year OS was 50% versus 39%, and a 5-year OS was 28% versus 17%. Despite patients receiving adjuvant therapy had more adverse prognostic factors such as higher frequency of positive lymph nodes and histologic grade than those not receiving adjuvant therapy (P = 0.001), adjuvant chemoradiotherapy improved median, 2-year, and 5-year survival significantly compared to surgery alone [22]. Both studies applied 50.4 Gy in 28 fractions of radiation dose. A subsequent pooled analysis of approximately 1,300 patients from both institutions showed that OS was longer for those who received chemoradiotherapy compared to surgery alone. Median survival was 21.1 versus 15.5 months, and 2- and 5-year OS were 44.7% versus 34.6% and 22.3% versus 16.1%, respectively (P < 0.001) [23].

Unlike randomized trials, reports of single-institution experiences have provided evidence to the benefit of adjuvant therapy for resected pancreatic cancer. Given that 11–26% of patients experience distant progression during radiotherapy [24, 25], multimodality therapy seems necessary in adjuvant setting; therefore, a reasonable consideration is to begin with adjuvant chemotherapy, followed by radiotherapy in patients who do not progress.

The use of neoadjuvant treatment in pancreatic cancer offers several theoretical advantages compared to adjuvant treatment: (1) Pancreatic cancer overall has the poor prognosis because it has a low rate of resectability. Neoadjuvant therapy may downstage the local, borderline, and unresectable disease, which potentially facilitates resectability with clear margins (R0 resection) and decreases lymphatic spread [26]. (2) Pancreatic cancer is more likely a systemic disease with high incidence of recurrence, and 80–85% of patients experienced recurrence even after undergoing curative resection [13, 14]. To start, early chemoradiation therapy may reduce the incidence of distal metastasis and contribute to improved survival. (3) A proportion of patients will develop distant metastatic disease during neoadjuvant therapy, in whom a major unnecessary surgical procedure can be avoided [8]. (4) In naïve tumor bed, radiotherapy can be more effective due to rich oxygenated tissue compared to postoperative status. In addition, by avoiding bowel displacement due to surgery, radiotherapy can be well tolerated without higher gastrointestinal toxicity [27, 28]. However, there have been no large randomized trials of neoadjuvant therapy in resectable pancreatic cancer, and several institutions have used this strategy to improve the survival rates of patients with pancreatic cancer (Table 38.2).

Surgical resection offers the only potentially curative treatment in pancreatic cancer, and subset of patients with borderline resectable disease who do not develop progressive disease will benefit from surgery after neoadjuvant approach [43]. On the other hand, pancreatic tumors are usually considered unresectable/locally advanced if it has the following features: (1) involvement of nodes outside resection field, (2) encasement of more than half circumference of the superior mesenteric artery, (3) abutting or encasement of more than half circumference of celiac axis, (4) superior mesenteric vein or portal vein occlusion of without suitable vessel for reconstruction, and (5) invasion or encasement of the aorta. These patients have poor prognosis with median survival that ranges from 8 to 12 months [44]. Treatment options in patients with locally advanced/unresectable cancer are chemotherapy alone, chemotherapy and radiation including intensity-modulated radiotherapy (IMRT), and stereotactic radiation therapy (SBRT) which can also give chemotherapy followed by radiotherapy. With conflicting results, there is little consensus as to the appropriate management of locally advanced patients (Table 38.4). In addition, the role of radiotherapy in unresectable, locoregionally advanced pancreatic cancer remains unclear. The addition of radiation may slow the local progression and offer palliation of symptoms such as pain, biliary, or bowel obstruction. On the other hand, the likelihood of micrometastatic distant disease is high, so that locally advanced cancer is quite often treated with chemotherapy, which improves quality of life and survival when compared with supported care. Also, when chemotherapy and radiotherapy are combined, long-term survival has been reported. However, several issues remain to be defined about the optimal treatment of locally advanced pancreatic cancer: (1) defining the optimal systemic regimen which administers with or without radiation, (2) determining whether radiation should be added to chemotherapy, and (3) determining when radiation and how radiation should be delivered.