^{What immune parameters correlate with effective anti-tumor immunotherapy? Lessons learned using Listeria monocytogenes as a live vector for HPV associated tumors.}

^{Yvonne Paterson, Ph.D.}

^{University of Pennsylvania, 323 Johnson Pavilion, 36th St. and Hamilton Walk, Philadelphia, PA 19104-6076}

^{We have been developing Listeria monocytogenes as a vaccine vector for viral and tumor antigens for a number of years. We have now shown, in various mouse tumor models, that tumor associated antigens can be delivered to the immune system by listerial vectors with the subsequent regression of established tumors. Curiously, we have discovered that effective anti-tumor immune responses are exquisitely dependent on the form of the antigen and the method of transformation of the L. monocytogenes vector. Here I will discuss immune parameters that appear to correlate with anti-tumor efficacy and those that are counter-productive in a mouse model of HPV-16 associated cervical cancer. HPV-16 immortalizes cells using two oncogenes E6 and E7. These antigens have been an intense focus of cancer immunotherapies using a variety of vaccine vectors. Because of the intra-cellular localization of these antigens, these therapies are mostly directed at inducing cellular immune responses of the type that is the forte of L. monocytogenes. We have engineered a number of expression systems for HPV-16 E7 in L. monocytogenes. These vaccines show different levels of efficacy in curing mice of established HPV-16 immortalized tumors. In one very effective vaccine the bacterium secretes a fusion protein composed of a truncated, non-hemolytic LLO, joined at the C-terminus to E7. In a second, relatively ineffective recombinant, Lm-E7, the E7 protein is expressed from the chromosome and is secreted using only the signal sequence of hemolysin. Using these two E7 expressing vaccine strains of L. monocytogenes, we have intensely examined the type of cell mediated immunity they induce and find significant differences in the CD4+ and CD8+ T cell subsets that are made in peripheral lymphoid tissue and the ability of these T cells to penetrate the tumor mass. These differences correlate well with the effect that these vaccines have on professional antigen presenting dendritic cells both in vitro and in vivo. It is well known that there are no good surrogate markers for clinical responses to cancer immunotherapy in both human clinical trials and in mouse models of cancer. We believe that we have identified a number of positive and negative immune markers that may provide a guide for what is needed for good cancer immunotherapy.}