1006. Analysis of Resistance to Viral Oncolysis in Primary Ovarian Cancer Cell Cultures

Abstract

We have generated a new type of capsid-modified, conditionally replicating, oncolytic adenovirus (Ad5/35.E1A/TRAIL). Tumor targeting of this vector is achieved via binding of the Ad serotype 35 fiber to CD46, which is upregulated, several orders of magnitude, on most tumor cells in situ. Furthermore, this vector employs a new concept for tumor-specific gene expression that is based on homologous recombination between inverted repeats (IR) in Ad genomes (Ad.IR). Ad5/35.IR-E1A/TRAIL was highly efficient in elimination of liver metastases derived from various human tumor cell lines in mice and did not cause unspecific toxicity in mice or baboons. So far, Ad5/35.IR-E1A/TRAIL has been tested in vitro and in vivo on established tumor cell lines, including ovarian cancer cell lines. Established tumor cell lines, however, often do not reflect the phenotypic and genetic heterogeneity of tumors in situ and appear not to be adequate models to assess the clinical performance of oncolytic adenoviruses. We therefore tested our oncolytic vector on cultures of primary ovarian cancer cells. Cultures were established from 25 biopsies and 6 ascites fluids from ovarian cancer patients. Tumor cells were either expanded in culture or as subcutaneous xenografts in immuno-deficient mice. Tumor cells within a given culture showed great heterogeneity in Ad infectibility. We therefore obtained clones from single tumor cells after limiting dilution and analyzed them for susceptibility to lysis by Ad5/35.IR-E1A/TRAIL. So far we have studied 16 clones derived from one ovarian biopsy culture. In 14 clones, 100% of tumor cells were lysed by the oncolytic vector. Resistant tumor cells can be efficiently infected with Ad5/35-GFP and supported viral replication. This indicates that resistance to Ad5/35.IR-E1A/TRAIL is due to a block in execution of TRAIL-mediated apoptosis. To corroborate heterogeneity of tumor cells in susceptibility to our oncolytic vector, over the next two months we will perform transduction with clones derived from biopsies from at least four additional ovarian cancer patients. The fact that clones derived from the same tumor cultures display different susceptibility also gives us a means to study and identify genes or pathways that are involved in mediating resistance. So far, using DNA arrays, we have compared expression profiles of the original biopsy with those of the corresponding tumor cell culture at different passage numbers. We also included SKOV-3 cells, an established ovarian cancer cell line, in this study. These studies indicated that the expression profile of primary ovarian cancer cultures were closer to that of SKOV-3 than to the original biopsy. We plan to repeat this study and use laser capture to obtain tumor cells from biopsy sections. We have also started array analyses with RNA isolated from resistant and susceptible clone and the results of these studies will be presented.