Erlotinib and Proton Therapy for Locally Advanced Pancreatic Cancer: A Phase II Trial

Purpose/Objective(s): Epidermal growth factor receptor (EGFR) is overexpressed in pancreatic cancer. EGFR expression plays a potentially important role in modulation of tumor sensitivity to either chemotherapy or radiotherapy. Erlotinib is a receptor tyrosine kinase inhibitor with specificity for EGFR/HER1. A phase II trial was conducted to explore the efficacy of a regimen utilizing erlotinib and proton therapy.

Materials/Methods: Patients with unresectable or borderline resectable non-metastatic adenocarcinoma of the pancreas, as defined by 2012 NCCN guidelines, were included. Patients received neoadjuvant gemcitabine 1000 mg/m² IV on days 1, 8, 15, 22, 29, 36, and 43 and erlotinib 100 mg by mouth every day for 1–43 days (GE). If there was no evidence of metastatic disease after GE, then patients preceded with proton therapy to 50.4 Gy in 28 fractions with concurrent capecitabine 825 mg/m² twice per day (PCT). This was followed with maintenance oxaliplatin 130 mg/m² on day 1 and capecitabine 1000 mg/m² twice per day on days 2 to 15 (CapOx) for 4 cycles. The primary study objective was 1-year overall survival (OS). The benchmark was 43% 1-year survival as demonstrated in RTOG/NRG 98–12. The Kaplan-Meier method was used to estimate the one-year OS and the median OS and progression-free survival (PFS).

Results: The study enrolled 9 patients (5 male) ages 47–81 years old (median 62) between January 2013 and March 2016, when the trial was closed due to poor accrual. The 1-year OS rate was 55.6% (95% CI, 31% to 99%). The median OS was 14.1 months (95% CI, 11.4-NE) and the median PFS was 10.8 months (95% CI, 7.44-NE). A majority of patients completed PCT and GE, but only 33.3% completed the four cycles of CapOx. A third of patients experienced grade 3 toxicities, which were all hepatic along with one patient who also had grade 3 diarrhea. There were no grade 4 or 5 toxicities. Four patients were enrolled with borderline resectable disease (NCCN), three of which were eligible for pancreaticoduodenectomy after GE and PCT treatment. One of two patients who underwent resection had a negative margin.

Conclusion: This regimen for locally advanced pancreatic cancer exceeded the pre-specified benchmark and was safe and well tolerated. Additional investigations utilizing more current systemic treatment regimens with proton therapy are warranted.