ABSTRACT: Pancreatic cancer is the fourth leading cause of cancer death, and it is estimated that over 43,000 people would be diagnosed with and over 36,000 people would die of pancreatic cancer in the United States in 2010. Surgical resection remains the only chance for possible cure, but only 15% to 20% of patients newly diagnosed with pancreatic cancer are considered for surgical resection. Of these, the median five-year survival rate is still less than 20%, with most resections resulting in recurrent disease.[4,5] This suggests that even seemingly resectable pancreatic cancer has microscopic systemic spread before operative intervention occurs. Both adjuvant and neoadjuvant therapies have been studied in an effort to improve survival for patients with resectable pancreatic cancer.
Early trials of adjuvant therapy in pancreatic cancer had multiple limitations including small sample size, population heterogeneity, and inability to distinguish between components of combined modality treatment. The first randomized controlled trial of adjuvant therapy in pancreatic cancer was designed by the Gastrointestinal Tumor Study Group (GITSG), and found that adjuvant 5-FU plus radiation, followed by two years of weekly 5-FU maintenance therapy had better outcomes than surgery alone (Table 1). Of note, a split-course radiation therapy regimen of 40 Gy was used, which is considered to be substandard by modern techniques. The trial was continued as a registry since real-time disclosure led to difficulties in randomizing patients to the surgery-alone arm; a total of 32 patients were registered on the combined modality arm. Later review combining these 32 patients with the 43 patients in the original study showed continued benefit of adjuvant chemoradiation. Although this trial has been criticized for its small size, slow accrual rate, inferior (by modern standards) radiotherapy, and inability to distinguish between the benefit of chemotherapy and that of chemoradiation, it nevertheless served as a significant first step in establishing a place for adjuvant therapy in the treatment of resected pancreatic cancer.
For nearly fifteen years, the GITSG trial provided the only evidence-based rationale for adjuvant therapy in pancreatic cancer—until the European Organization for Research and Treatment of Cancer (EORTC) presented a phase III clinical trial comparing chemoradiation (20 Gy split-course radiation and infusional 5-FU 25 mg/kg) with observation alone in a population of 218 patients with both pancreatic head and periampullary tumors. This trial was criticized for studying only chemoradiation in a heterogeneous population with an inferior (by modern standards) radiotherapy regimen; however it did reveal a nonsignificant trend favoring chemoradiation in the subset of patients with pancreatic head tumors. The European Study Group for Pancreas cancer-1 (ESPAC-1) trial was published in 2004 with the intent of distinguishing between the benefit of chemotherapy and chemoradiation. The ESPAC-1 trial had a complex 2 × 2 factorial design in which 289 patients were randomized to one of four treatment groups: observation alone, chemotherapy, chemoradiation, or chemoradiation followed by chemotherapy. An additional 256 patients were allocated to one of two randomizations: either chemotherapy vs observation or chemoradiation vs observation. The study authors concluded that adjuvant chemotherapy had a beneficial effect on overall survival (median survival 20.1 months vs 15.5 months in those who received chemotherapy vs those who did not, HR for death 0.71, P = .009), but that chemoradiation was actually associated with a decreased survival (median survival 15.9 months vs 17.9 months in those who received chemoradiation vs those who did not, HR for death 1.28, P = .05). It should be noted, however, that the study was not powered to compare the four individual groups. Additionally, the split-course radiation therapy used in this trial is not current practice, so these results cannot be applied to modern techniques that give higher doses of radiation in a more protracted, but continuous, time course.
In 2008, two retrospective, non-randomized, single-institution analyses further suggested a benefit from adjuvant chemoradiation. The Johns Hopkins Hospital evaluated a prospectively collected cohort of patients who underwent either adjuvant chemoradiation with 5-FU, or observation alone. Of this cohort, patients who received adjuvant therapy had improved median, two-year, and five-year survival compared with those treated with surgery alone (21.2 vs 14.4 months; 43.9% vs 31.9%; 20.1% vs 15.4%, respectively; P < .001). The Mayo Clinic also performed a retrospective review of 472 patients who underwent R0 resection of invasive pancreatic cancer from 1975 to 2005. Of the 472 patients, 274 patients received adjuvant external beam radiation using multiple techniques; 98% of these patients also received concurrent 5-FU based chemotherapy. The study found that the patients who received adjuvant chemoradiation had improved median, two-year, and five-year overall survival compared with patients who received surgery alone (25.2 months vs 19.2 months, 50% vs 39%, and 28% vs 17%, respectively; P = .001), with an absolute survival benefit of 11%. As with other retrospective studies, there was inherent bias since only the patients who did well enough postoperatively to receive adjuvant therapy were included in the analysis. However, these data are interesting and should be seen as hypothesis-generating in designing future trials in adjuvant therapy in pancreatic cancer.
As new treatment options in metastatic disease have become available, further trials have been published using modern systemic therapies in the adjuvant setting (see Table 1). In 2007, the Charite Onkologie (CONKO)-001 trial provided data demonstrating a beneficial effect from adjuvant chemotherapy with gemcitabine(Drug information on gemcitabine) (Gemzar) on disease-free survival in pancreatic cancer. The CONKO-001 trial was a rigorously designed phase III randomized controlled clinical trial of 368 patients from 88 academic and community oncology centers in Germany and Austria who underwent complete (R0 or R1) resection of pancreatic cancer. The primary end point was disease-free survival; secondary end points were overall survival, toxicity, and quality of life. Patients were randomized to either adjuvant chemotherapy with six cycles of gemcitabine (1000 mg/m2 for three weeks of a four week cycle for six months) or observation alone. Adjuvant chemotherapy with gemcitabine improved median disease-free survival (13.4 months vs 6.9 months, P < .001). Subgroup analyses revealed that the advantageous effect of gemcitabine was apparent in R0 and R1 resections, large and small primary tumors, and node positive and negative disease. Recurrent disease occurred in 133 of 179 eligible patients (74.3%) in the gemcitabine group and 161 of 175 patients (92%) in the observation group. At the time of initial report, there was also a nonsignificant trend favoring the gemcitabine group over the control group, with median survival of 22.1 months (95% CI 18.4-25.8) vs 20.2 months (95% CI 17-23.4) ,respectively (P = 0.06). Subsequent qualified analysis demonstrated a significant overall survival benefit of gemcitabine over control (median overall survival 24.2 months [95% CI, 18.4-30.0] vs 20.5 months [95% CI 17.0-24.0]; P = .02). This robustly designed trial provided strong evidence favoring adjuvant chemotherapy with gemcitabine.
Soon after CONKO-001 was published, the RTOG (Radiation Therapy Oncology Group)-9704 was published examining the benefits of 5-FU vs gemcitabine in addition to chemoradiation. The primary endpoint was overall survival; disease-free survival and toxicity were secondary endpoints. The study population was comprised of 451 patients who underwent complete gross resection of pancreatic cancer. Patients were randomized to either adjuvant gemcitabine (1000 mg/m2 weekly for three weeks) followed by chemoradiation and three months of gemcitabine or infusional 5-FU (250 mg/m2 per day for three weeks) followed by chemoradiation and three months of 5-FU. Both groups received identical chemoradiation following adjuvant chemotherapy (50.4 Gy with continuous infusion of 250 mg/m2 of 5-FU daily). There was no difference in disease-free survival, and there was a higher number of grade 3 toxicities in the gemcitabine group. This was primarily due to hematologic toxicity, with no difference in febrile neutropenia or infection rates. Of note, the gemcitabine group had more patients with T3/T4 disease (81% vs 70% in the 5-FU group, P = .01). Additionally, gemcitabine was used as salvage chemotherapy in a significant proportion of both groups (82% in the 5-FU group and 62% in the gemcitabine group). On initial analysis, there was a nonsignificant trend favoring gemcitabine over 5-FU in median and three-year overall survival (20.5 months and 31% vs 16.9 months and 22%, HR 0.82 [95% CI, 0.65-1.03]; P = .09). Once differences in nodal status, tumor size, and margin status between the two groups were accounted for on multivariate analysis, gemcitabine was found to have a significant treatment effect on overall survival, with an HR of 0.8 (95% CI, 0.63-1.00; P = .05).
At the 2009 ASCO Annual Meeting, the EORTC group reported results of a phase II study of gemcitabine-based chemoradiation after R1 resection. This study randomized 90 patients to either four cycles of gemcitabine alone or gemcitabine with concurrent radiation. There were no differences in the rate of completion of treatment between the two groups (86.7% and 73.3% of control and experimental arms, respectively), or in grade 4 toxicities (0% and 4.4%), which were the study’s primary endpoints. There was also no difference in median disease-free survival (12 months vs 11 months, P = 0.6) or median overall survival (24 months in both arms).
The ESPAC-3(v2) study was recently published as a derivative of ESPAC-1. This was a multicenter randomized controlled phase III trial comparing adjuvant therapy with a 5-FU-based regimen to gemcitabine in patients with resected pancreatic cancer in 1088 patients from 16 countries with an R0 or R1 resection of pancreatic cancer from 2000-2007. This study was designed to compare the survival benefit of adjuvant fluorouracil(Drug information on fluorouracil) plus folinic acid to that of gemcitabine. After stratification for margin status and country, patients were randomized to receive either “Mayo Clinic regimen” 5-FU/FA (20 mg/m2 of folinic acid iv bolus followed by 425 mg/m2 5-FU, given 1-5d every four weeks) or gemcitabine (1000 mg/m2 iv 1d, 8d, 15d every four weeks) over a six-month period. The study was designed to detect a 10% difference in two-year survival rates with 90% power. However, no difference in median survival between the two groups was found (23 months for the 5-FU/FA group and 23.6 months for the gemcitabine group).
Based on these studies, it appears that patients with resected pancreatic cancer derive benefit from adjuvant chemotherapy, while the benefit of chemoradiation remains unclear. Considering that a high proportion of patients who undergo resection of their pancreatic cancer have microscopic systemic spread, the benefit of improved local control from chemoradiation will likely be limited until better systemic therapies are available. Gemcitabine has been the standard of care for systemic disease and tends to be better tolerated than the “Mayo Clinic regimen” of FU/FA. Consequently, gemcitabine-based therapies are the main focus of several recent and ongoing studies. ECOG 2204 is an intergroup randomized phase II study recently reported at the 2010 ASCO Annual Meeting. This study randomized patients to receive adjuvant therapy with either gemcitabine plus bevacizumab (Avastin) or gemcitabine plus cetuximab(Drug information on cetuximab) (Erbitux), followed by 5.5 weeks of chemoradiation with capecitabine(Drug information on capecitabine) (Xeloda) as well as three additional cycles of the original respective chemotherapy. Although both regimens were well-tolerated, there was no significant difference between the bevacizumab(Drug information on bevacizumab) and cetuximab groups in terms of disease-free survival (22% vs 16%) or overall survival (23% vs 27%) at two years. RTOG 0848 is a phase III trial evaluating gemcitabine alone versus gemcitabine plus erlotinib (Tarceva) as adjuvant chemotherapy. Patients who do not have progression of disease after five months of therapy will undergo a sixth month of their respective chemotherapy with or without the addition of chemoradiation with either 5-FU or capecitabine. The goal of this study is to determine the benefit of erlotinib in addition to gemcitabine, as well as the benefit of consolidation chemoradiation.
A third study, American College of Surgeons Oncology Group (ACOSOG) Z5041, is evaluating the use of gemcitabine and erlotinib in both the neoadjuvant and adjuvant setting. In addition to clarifying the role of erlotinib as an adjunct to gemcitabine, this study aims to determine prognostic markers for response to EGFR inhibitors. Finally, ESPAC-4 is a phase III, multi-center randomized controlled clinical trial comparing combination therapy with gemcitabine and capecitabine with gemcitabine alone as adjuvant chemotherapy, a regimen that has had negative results in phase III trials in the metastatic setting.[17,18] Subject accrual is expected to be completed in 2014.