Malignant brain tumors carry a dismal prognosis despite surgery, chemotherapy, and radiation, hence new therapeutic approaches are needed. New strategies are being investigated using modified viruses (termed “vectors”) to infect cancer cells and deliver genes that serve as blueprints to make therapeutic proteins inside the cancer cells themselves.  

Until recently, most researchers used viral vectors that could deliver therapeutic genes to the initially infected tumor cells but are rendered incapable of infecting any additional cells. However, such ‘non-replicating’ vectors were found to have limited benefit, because not enough tumor cells could be reached.  

A newer approach is to use virus vectors that actively replicate themselves and can spread forth from the initially infected cancer cells within tumors, but not in normal tissues, thereby continuing to infect more cancer cells even as the cells continue to proliferate. These tumor-selectively spreading viruses are used to deliver a “suicide gene,” which converts a non-toxic ‘trigger’ compound (“prodrug”) into a DNA synthesis-blocking chemotherapy drug. Because this virus causes the chemotherapy drug to be generated selectively and directly within the infected tumor itself, there are few adverse side effects.  

The first version of this type of therapeutic virus has shown highly promising results in early-stage clinical trials and is currently being tested in an international Phase 2B/3 trial for recurrent brain cancer. 

We recently discovered that this approach can also activate the immune system to attack tumors. Hence, in this Alliance for Cancer Gene Therapy funded study, we examine whether brain tumors that show a high rate of new mutations, which can be recognized by the immune system, may be correlated with better responses to this treatment.  

We also propose to develop a new tumor-selectively spreading virus vector that delivers a different “suicide gene” which cross-links DNA and thereby generates new mutations for the immune system to attack, and we further propose to combine this mutagenic DNA cross-linking suicide gene therapy with strategies to overcome immune blockade (“checkpoint”) mechanisms within tumors.  

Finally, we will conduct preclinical studies to evaluate the safety of this new virus vector for use in a future clinical trial. If the proposed preclinical studies are successful in validating the safety, efficacy, and mechanism of action for this new immunogenic suicide gene therapy, combined with immune checkpoint inhibition, we anticipate that this approach can be rapidly translated to the clinic in collaboration with a biotech partner that is already testing our previously developed virus vector in clinical trials.  

This Alliance for Cancer Gene Therapy Research Fellow is funded in part by Swim Across America. 


The prevailing view suggests that lung tumors may not be good targets for vaccine therapy. The reason for this belief is the fact that in lung tumor patients we do not find any indication of immune activation including a complete absence of killer cells (cytotoxic T cells) that could recognize and kill the tumors. Lung tumors are therefore classified as non-immunogenic. We suggest that this circumstance is ideal for immunotherapy of non-small cell lung carcinoma by vaccination. 

We have been developing vaccines that can generate killer cells against lung tumors (non-small cell lung carcinoma) by genetic engineering lung tumor cell lines adapted to grow in cell culture. We predict that a vaccine that can generate killer cells in lung tumor patients will be effective and that the patient tumor will be defenseless against killer cell attack because the tumor cells have never before been exposed to killer cells and have not evolved protective mechanisms.  

In a phase I study we found that a lung tumor vaccine, designated B7-vaccine, was able to generate killer cells and seemed to provide clinical benefit to patients. In the study proposed here we have developed a novel vaccine approach for lung tumors based on a genetically modified, secreted form of heat shock protein gp96. The vaccine is comprised of lung tumor cells that secrete hsp-gp96 and thereby provide a very strong stimulus for the immune system to generate killer cells against lung tumors.  

Our initial in vitro data and data in experimental mice indicate that the gp96-vaccine is over one million-fold more efficient in the generation of killer cells than normal proteins in the absence of gp96. These properties of tumor secreted gp96 make it an ideal vaccine. In addition, these properties allow us to study how gp96 induces immunity and how it may be able to restore immunity that is suppressed in the presence of existing tumors. 

In this application we plan to pursue the clinical testing of the gp96-vaccine and to conduct further research into the underlying immune response. Understanding the underlying mechanisms will allow us to develop further reagents to improve vaccine efficiency.