Le mode de vaccination des patients avec un antigène tumoral influence la différenciation des lymphocytes T CD8 anti-vaccinaux

Connerotte, Thierry
(2009)

Files

TConnerotte2009.pdf
  • Open Access
  • Adobe PDF
  • 8.56 MB

Details

Authors
  • Connerotte, ThierryUCLouvain
    author
Supervisors
Coulie, Pierre
Abstract
(en) The identification in 1991 of antigens recognized on human tumor cells by cytolytic T lymphocytes (CTL) opened new prospects for cancer immunotherapy. Several small scale clinical trials of vaccination with MAGE antigens have been conducted in metastatic melanoma patients with measurable disease. Tumor regressions have been observed in 15-20% of the patients, and a third of these regressions corresponded to 'partial responses' or better. These clinical results are similar whatever the vaccine modality (peptide, protein, recombinant virus, or dendritic cells). Vaccination with peptide MAGE-A3(168-176) under different modalities induced low level anti-vaccine CTL responses in the blood, reached between 10-5 and 10-3 of the CD8+ T cells. These responses were detectable after in vitro restimulation in limiting dilution condition, followed by labeling the responder cells with the HLA-A1/MAGE-A3(168-176) tetramer. The anti-vaccine CTL responses were correlated with the tumor regressions, suggesting a causative link. It is however surprising that so few anti-vaccine CTL could by themselves eliminate one or several metastases, leading in some cases to complete remission. Several hypotheses could explain this paradox, and this work consisted in exploring two of them. The low level anti-vaccine CTL responses could be associated to other anti-vaccine CTL responses, possibly of high magnitude, but not detectable by our methodology. Indeed, labeling the microcultures with the HLA-A1/MAGE-A3(168-176) tetramer, we were not able to detect T lymphocytes recognizing peptide MAGE-A3(168-176) presented by other HLA molecules than HLA-A1. I used two approaches to reasonably exclude this possibility: a screening of the microcultures with a cytokine secretion assay instead of a tetramer, and an in vitro assessment of the binding of peptide MAGE-A3(168-176) to other recombinant class I HLA molecules. I obtained negative results, strengthening the hypothesis that the low level anti-MAGE-A3(168-176)/HLA-A1 CTL responses detected in some patients played a crucial role in the tumor regressions. Anti-vaccine CTL could have functional properties that could explain their anti-tumoral effect, even in low numbers. As the low blood frequencies of anti-vaccine T cells prevented robust ex vivo functional analyses, I analyzed a representative set of 15 anti-MAGE-A3(168-176)/HLA-A1 CTL clones derived from 8 melanoma patients who displayed tumor regression. Using gene expression profiling, I observed that the expression levels of about 20 genes distinguished the anti-MAGE-A3(168-176)/HLA-A1 CTL clones derived from patients vaccinated with peptide alone or with a recombinant poxvirus containing MAGE minigenes, and the anti-MAGE-A3(168-176)/HLA-A1 CTL clones derived from patients vaccinated with peptide-pulsed dendritic cells. These results indicated that the vaccination modality with a MAGE tumor-specific antigen influences the differentiation of anti-vaccine CTL. This might impact on their capacity to trigger tumor regression. In addition they suggest that it might be important to carry out the immunological monitoring of vaccinated cancer patients with methods that do not only evaluate quantitative aspects of the response, but that can also compare the functional properties of the detected anti-vaccine T lymphocytes.
Affiliations

Citations

Connerotte, T. (2009). Le mode de vaccination des patients avec un antigène tumoral influence la différenciation des lymphocytes T CD8 anti-vaccinaux. https://hdl.handle.net/2078.5/132006