Pancreatic Adenocarcinoma: Long-Term Outcomes of Adjuvant Therapy in the ESPAC4 Phase III Trial

Abstract

Purpose: The ESPAC4 trial showed that adjuvant chemotherapy with gemcitabine plus capecitabine (GemCap) produced longer overall survival (OS) than gemcitabine monotherapy. Subsequently, the PRODIGE24-CCTG PA.6 trial showed even longer survival for modified fluorouracil, folinic acid, irinotecan, and oxaliplatin (mFOLFIRINOX) than gemcitabine but had more restrictive eligibility criteria. Our aim was to analyze the ESPAC4 survival on long-term follow-up.

Methods: The OS of 732 ESPAC4 patients comparing 367 randomly assigned to gemcitabine and 365 to GemCap was previously reported after a median follow-up time of 43.2 months (95% CI, 39.7 to 45.5) and 458 deaths. Analysis was now carried out after a median follow-up of 104 months (101-108) and 566 deaths.

Results: The median OS was 29.5 months (27.5-32.1) for all patients, 28.4 months (25.2-32.0) in the gemcitabine group and 31.6 months (26.5-38.0) in the GemCap group (hazard ratio [HR], 0.83 [0.71 to 0.98]; P = .031). R0 patients given gemcitabine had a median survival of 32.2 months (27.9-41.6) compared with 49.9 months (39.0-82.3) for those given GemCap (HR, 0.63 [0.47 to 0.84]; P = .002). Lymph node-negative patients had significantly higher 5 year OS rates on GemCap (59% [49%-71%]) than gemcitabine (53% [42%-66%]; HR, 0.63 [0.41 to 0.98]; P = .04) but not those with positive lymph nodes (P = .225). The OS advantage for GemCap was retained in the PRODIGE24 subgroup of 193 (26.4%) ESPAC4 patients not eligible for PRODIGE24 with a median survival of 20.7 (16.2-27.3) months in patients allocated to gemcitabine compared with 25.9 (22.3-30.2) months for ineligible patients allocated to GemCap (HR, 0.71 [95% CI, 0.52 to 0.98]; χ²_log-rank-1df = 4.31; P = .038).

Conclusion: GemCap is a standard option for patients not eligible for mFOLFIRINOX. Exploratory evidence suggests that GemCap may be particularly efficacious in R0 patients and also in lymph node-negative patients.

Trial registration: ClinicalTrials.gov ISRCTN96397434.

FIG 1.

CONSORT diagram. GemCap, gemcitabine plus capecitabine.

FIG 2.

Extended follow-up analysis for OS for patients in ESPAC4. (A) OS for patients randomly assigned to Gem versus GemCap. (B) OS in patients with R0 versus R1 resection margin status randomly assigned to Gem versus GemCap. Gem, gemcitabine; GemCap, gemcitabine plus capecitabine; HR, hazard ratio; OS, overall survival.

FIG 3.

Analysis for OS for patients in ESPAC4 analyzed according to PRODIGE24 eligibility criteria. (A) OS for all patients in ESPAC4 analyzed according to whether eligible (yes) or not eligible (no) to PRODIGE24 criteria. (B) Analysis for OS for patients in ESPAC4 analyzed according to whether they were randomly assigned to Gem versus GemCap whether they eligible or not eligible to PRODIGE24 criteria. Gem, gemcitabine; GemCap, gemcitabine plus capecitabine; HR, hazard ratio; OS, overall survival.

FIG 4.

Forest plot summarizing the effects of prognostic factors and PRODIGE24 eligibility on survival by treatment. CA 19-9, carbohydrate antigen 19-9; Gem, gemcitabine; GemCap, gemcitabine plus capecitabine; HR, hazard ratio; NA, not available.