A Pilot Study of Post-Operative Adjuvant Vaccine for Advanced Gastric Cancer

Adjuvant cancer vaccine for gastric cancer

Abstract

Background Previously, we had performed a clinical study using human leukocyte antigen (HLA)-A24-binding peptide vaccines containing a combination of novel cancer-testis antigens and anti-angiogenic peptides derived from DEPDC1, URLC10, FOXM1, KIF20A and VEGFR1 for advanced gastric cancer (AGC) patients who were refractory to chemotherapy. We applied the cocktail vaccine to the combination therapy with S-1 for patients with AGC as a post-operative adjuvant therapy and performed this clinical pilot study.
Methods AGC patients who had curative surgery and were classified as pathologically stage III were enrolled. At each 6-week treatment cycle, patients received weekly subcutaneous administration of the cocktail vaccine with 5 continuous injections and one break for the first 4 cycles and with bi-weekly injections for the following 4 cycles. S-1 (80 mg/m²) was administered orally for 4 weeks with 2-week rest for all 8 cycles. The primary endpoint was the safety of the combination therapy and the secondary was the relative dose intensity for S-1.
Results Fourteen patients were enrolled. Six patients with HLA-A*2402 had received S-1 plus the cocktail vaccine as an adjuvant therapy and the remaining 8 had S-1 monotherapy for eight cycles. Five out of 6 patients subjected to the combination group completed the therapy and one patient discontinued because of Grade 3 injection-site reaction. No adverse events of grade 3 or higher were observed except injection-site reactions shown in 5 out of 6 patients who had vaccine therapy. The mean and median relative dose intensities for S-1 were 75.5% and 88% in the combination group and 67% and 80.5% in S-1.
Conclusion The vaccine therapy combined with S-1 was manageable and safe adjuvant therapy for stage III gastric cancer. Furthermore, the optimal relative dose intensity of S-1 was achieved in combination group, although the injection-site reaction should be considered.