Neoadjuvant Stereotactic Body Radiation Therapy Plus Elective Nodal Irradiation With Concurrent Capecitabine for Patients With Resectable Pancreatic Cancer: Survival Analysis of a Prospective Phase 1 Trial

Abstract

Purpose

Elective nodal irradiation (ENI) in resectable pancreatic cancer remains undefined, although occult nodal disease is common. This study investigated the use of neoadjuvant stereotactic body radiation therapy (SBRT) for primary disease with ENI, with concurrent capecitabine therapy. Safety data for this protocol were previously reported. In this report, we provided an updated survival analysis.

Methods and Materials

This is a prospective, single institution, phase 1A/B dose-escalation trial that enrolled patients with biopsy-proven, resectable, pancreatic adenocarcinoma between 2014 and 2019 (NCT1918644). Patients were enrolled into 1 of the 3 cohorts with escalating dose levels. Neoadjuvant SBRT to the primary tumor was delivered in 5 fractions of 5, 6, or 7 Gy with concomitant capecitabine (1650 mg/m²). All patients received ENI (5 Gy × 5 fractions). Our initial report found no dose-limiting toxicities. Clinicopathologic features were summarized using descriptive statistics. Kaplan-Meier curves were employed for survival analysis.

Results

Of 17 enrolled patients, 16 were evaluable (94.1%). Thirteen (76.5%) proceeded to surgery. The median follow-up was 28.0 (range, 1.7-71.9) months. Four patients (25.0%) received neoadjuvant chemotherapy, and 6 (37.5%) received adjuvant chemotherapy. Pathologic nodal involvement (69.2%) was associated with a higher risk of any relapse (P < .01) and distant metastasis (P = .02). Local failure occurred in 4 (25%) patients, with 2 of 4 of those failures occurring partially within the 25 Gy elective nodal field and 1 of 4 occurring in the 25 Gy elective nodal field and partially within the 35 Gy tumor field. The median overall survival (OS) and disease-free survival (DFS) were 31.1 (range, 2.3-73.6) months and 12.0 (range, 0.4-71.9) months, respectively. Three-year OS and DFS were 50% and 31.3% overall, and 61.5% and 38.5% for the surgical cohort. Patients with positive lymph nodes had worse median OS (23.9 vs 69.3 months; P = .002) and DFS (9.9 vs 58.9 months; P = .002). No further radiation-related toxicities were noted since the prior report.

Conclusions

Neoadjuvant SBRT to the primary tumor with ENI and radiosensitizing chemotherapy is a feasible approach that may improve outcomes in patients with resectable and borderline pancreatic cancer, despite high rates of pathologic nodal involvement. Further investigation of this strategy is warranted in a larger cohort.