Abstract LB-218: Immune response assessment in a phase II trial of AE37 HER2 peptide vaccine

Abstract

Abstract Introduction: AE37 is the Ii-Key hybrid of the HER2 derived peptide AE36 (776-790). A phase I trial administering AE37 with the immunoadjuvant GMCSF demonstrated the vaccine to be safe and capable of stimulating CD4+helper T cells with HER2 specific anti-tumor activity. Here we present an analysis of our immunologic testing of our prospective, randomized, single-blinded, phase II trial of the AE37+GMCSF vs. GMCSF alone for the prevention of breast cancer recurrence. Methods: After completion of indicated standard therapy; disease-free, node positive or high risk node negative breast cancer patients were randomized to receive either AE37+GMCSF (VG) or GMCSF (CG) in 6 monthly intradermal inoculations. Patients were enrolled with any level of HER2 expression (IHC 1+ 2+ or 3+). Specific immunologic responses to both AE36 and AE37 were evaluated. In vitro responses were measured using [3H]-thymidine incorporation assay and in vivo responses using delayed-type hypersensitivity (DTH) reactions. T regulatory cells (Tregs) were measured throughout vaccination series. Data was analyzed using Pearson chi-squared tests. Results: To date 217 patients have enrolled (VG=109, CG=108). HER2 over-expression was present in 54 (49.5%) VG and 51 (47.2%) CG, p=0.783. VG had 59 (51.4%) ER positive and CG had 58 (53.7%, p=0.985). The in vitro proliferation responses with a stimulation index (SI) ≥2 were classified as high responders VG 36 (33.0%) vs CG 8 (7.4%, p<0.001), SI 1.5-2 were classified as low responders VG 19 (17.4%) vs CG 16 (14.8%, p=0.600) or SI <1.5 as non responders VG 54(49.5%) vs CG 84(77.8%, p<0.001). The in vivo DTH reactions measured in 149 (VG n=86, CG n=63) were stratified as responders with ≥5mm (VG 74(86.0%) vs CG 17(27.0%) p<0.0001) or non responders with <5mm (VG 12 (14.0%) vs CG 46(73.0%) p<0.001). Tregs responses were measured in 107 patients (VG n=56, CG=51) and categorized according to percent change from pre-vaccination baseline as an increase >110% (5(8.9%) VG vs 10 (19.6%) CG, p=0.112), no change 90-110% (10 (17.9%) VG vs 13 (25.5%) CG, p=0.337), or decrease <90% (41 (73.2%) VG vs 28 (54.9%), p=0.048). Conclusion: VG patients had significant immunologic responses compared to CG. VG had a statistical decrease in Tregs compared to CG. Monitoring immunologic tests and Tregs throughout the vaccination process may stratify patients into responders and non responders and thus assist in identifying patients that will have recurrence. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr LB-218. doi:1538-7445.AM2012-LB-218