Stage III Pancreatic Cancer
Table 5. Randomized Studies in Stage III Pancreatic Cancer: Median Survival continued...
As it became clear that radiation therapy alone was an inadequate treatment, investigators evaluated combined modality approaches versus chemotherapy alone. Investigators from the FFCD-SFRO randomly assigned 119 patients to induction chemoradiation therapy (60 Gy in 2 Gy fractions with 300 mg/m2 /day of continuous infusion 5-FU on days 1 through 5 for 6 weeks and 20 mg/m2 /day of cisplatin on days 1 through 5 during weeks 1 and 5) or induction gemcitabine (1,000 mg/m2 weekly for 7 weeks).[Level of evidence: 1iiA] Maintenance gemcitabine was administered to both groups until stopped by disease progression or treatment discontinuation as a result of toxicity. Median survival was superior in the gemcitabine arm (13 vs. 8.6 months, P = .03).
Nonhematological grade 3 to 4 toxicities (primarily gastrointestinal) were significantly more common in the chemoradiation arm (44% vs. 18%, P = .004), and fewer patients completed at least 75% of induction therapy (42% vs. 73%). Nonetheless, the survival benefit persisted in a per-protocol analysis of patients receiving at least 75% of planned therapy. Notably, the dose intensity of maintenance gemcitabine was significantly less in the chemoradiation arm because of a greater incidence of grade 3 to 4 hematological toxicities (71% vs. 27%, P = .0001). As a result of this study, induction chemoradiation has fallen out of favor.
The results of the FFCD study stand in contrast to the results of a study from ECOG where investigators randomly assigned 74 patients to either gemcitabine alone or gemcitabine with radiation followed by gemcitabine. Of note, the study was closed early as the result of poor accrual. The primary endpoint was survival, which was 9.2 months (95% CI, 7.9–11.4 months) and 11.1 months (95% CI, 7.6–15.5 months) for chemotherapy and combined modality therapy, respectively (one-sided P = .017 by stratiﬁed log-rank test). Grade 4 and 5 toxicity was greater in the chemoradiation arm than in the chemotherapy arm (41% vs. 9%).
Given the increased toxicity of chemoradiation and the early development of metastatic disease in a large percentage of patients with stage III pancreatic cancer, investigators are pursuing a strategy of selecting patients with localized disease for chemoradiation. With this strategy, the selected patients have an absence of progressive disease locally or systemically after several months of chemotherapy. A retrospective analysis of 181 patients enrolled in prospective phase II and III GERCOR studies revealed that 29% had metastatic disease after three months of gemcitabine-based chemotherapy. For the remaining 71%, median OS was significantly longer among patients treated with chemoradiation compared to additional chemotherapy (15.0 months vs. 11.7 months, P = .0009).[Level of evidence: 3iiiA] Taken together, the FFCD and GERCOR studies provide support for gemcitabine-based chemotherapy for at least 3 months, followed by chemoradiation in the absence of metastatic disease. This approach has yet to be validated in a prospective phase III trial.