442. Novel cox-2 Driven Infectivity Enhanced CRAds for Treatment of Ovarian Cancer

Abstract

CRAds are a promising strategy for adenoviral gene therapy of cancer. Although preclinical and the first clinical data with first generation CRAds show good results, more specificity of CRAd replication and an infectivity enhancement for target cells would be desirable. To this end we constructed a novel generation of CRAds specific for ovarian cancer treatment, containing the cox-2 promoter driving different mutations of E1A, combined with an adenovirus type 3 fiber-knob mutation on otherwise type 5 viruses. Two length variants of the cox-2 promoter were used, a shorter clone being more active in and a longer clone being more specific for ovarian cancer cells. The aim of this strategy is to investigate whether a more specific or a more active promoter for CRAds is more advantageous. E1A was used as wild type, as delta24-mutation and as delta2-delta24-mutation, testing the concept of specificity enhancement for cancer cells due to delta-mutations. These mutations inactivate Rb binding and thus allow replication only in cell deficient in the p16-Rb pathway, such as most human cancer cells, including ovarian cancers. Since the adenovirus type 3 receptor seems to be highly overexpressed in ovarian cancer cells resulting in a one to two log infectivity enhancement for these cells, we selected the adenovirus 5/3 chimera approach for enhanced transduction of tumor cells. Six different viruses were constructed, amplified and their constructs verified with PCR and restriction enzymes. The replication of all six viruses was one to two logs improved in cox-2 positive cells and one to two logs reduced in cox-2 negative cells compared to wild type adenovirus. In ovarian cancer cells the replication was improved up to three logs as result of the combination of specific replication in and infectivity enhancement for these cells. In addition, the liver-off phenotype of these viruses was tested with primary normal liver cells. Whereas the wild type virus replicated in and lysed the liver cells, our newly constructed CRAds replicated at least two logs lower than wild type. Further more, the cox-2 driven delta24-CRAds showed stronger oncolysis in ovarian tumor cells than an otherwise identical delta24-CRAd without cox2 promoter. Thus these viruses seem to replicate specifically and amplified in ovarian cancer cells while preserving cox-2 negative and especially liver cells. Therefore these viruses could improve the safety and efficacy profile of adenoviral gene therapy strategies for ovarian cancer.