BackgroundCancer vaccines aim at eliciting not only an immune response against specific tumor antigens, but also at enhancing a preexisting immunity against the tumor. In this context, we recently reported on the levels of preexisting immunity in prostate cancer patients vaccinated with the HER-2 hybrid peptide (AE37), during a phase I clinical trial. The purpose of the current study was to correlate between preexisting immunity to the native HER-2 peptide, AE36, and expression of HLA-A2 and -A24 molecules with the clinical outcome. Additionally, we investigated the ability of the AE37 vaccine to induce an antitumor immune response against other tumor associated antigens, not integrated in the vaccine formulation, with respect to the clinical response.MethodsWe analyzed prostate cancer patients who were vaccinated with the AE37 vaccine [Ii-Key-HER-2/neu
(776–790) hybrid peptide vaccine (AE37), which is a MHC class II long peptide vaccine encompassing MHC class I epitopes, during a phase I clinical trial. Preexisting immunity to the native HER-2/neu
(776–790) (AE36) peptide was assessed by IFNγ response or dermal reaction at the inoculation site. Antigen specificity against other tumor antigens was defined using multimer analysis. Progression free survival (PFS) was considered as the patients’ clinical outcome. Two-tailed Wilcoxon signed rank test at 95 % confidence interval was used for statistical evaluation at different time points and Kaplan–Meier curves with log-rank (Mantel-Cox) test were used for the evaluation of PFS.ResultsPreexisting immunity to AE36, irrespectively of HLA expression, was correlated with longer PFS. Specific CD8+ T cell immunity against E75 and PSA146–151 (HLA-A2 restricted), as well as, PSA153–161 (HLA-A24 restricted) was detected at relatively high frequencies which were further enhanced during vaccinations. Specific immunity against PSA153–161 correlated with longer PFS in HLA-A24+ patients. However, HLA-A2+ patients with high preexisting or vaccine-induced immunity to E75, showed a trend for shorter PFS.ConclusionsOur data cast doubt on whether preexisting immunity or epitope spreading specific for HLA-class I-restricted peptides can actually predict a favorable clinical outcome. They also impose that preexisting immunity to long vaccine peptides, encompassing both HLA class II and I epitopes should be considered as an important prerequisite for the improvement of future immunotherapeutic protocols.Protocol ID Code: Generex-06-07National Organization for Medicines (EOF) ID Code: IS-107-01-06NEC Study Code: EED107/1/06EudraCT Number: 2006-003299-37Date of registration: 07/06/2006Date of enrolment of the first participant to the trial: Nov 1st, 2007Electronic supplementary materialThe online version of this article (doi:10.1186/s40425-016-0183-4) contains supplementary material, which is available to authorized users.