Neha Vapiwala, MD
Abramson Cancer Center of the University of Pennsylvania
Last Modified: November 6, 2005
AGT is a human protein that modifies DNA in both normal and tumor cells in order to protect it from damage by certain chemotherapy drugs. In both petri dishes and animal studies, AGT has been observed to prevent the death of cancer cells that are in the presence of cancer cell-killing chemotherapy. Glioblastoma multiforme (GBM) cells are among the ones studied, and researchers have documented the inverse relationship between AGT level and temozolamide activity (Friedman et al. 1998, Hegi et al. 2005). In fact, the survival benefit that was recently shown by Stupp et al. using temozolamide (and radiation therapy) in newly diagnosed GBM is only valid in those patients (pts) who are "methylators", and thus are AGT-negative. In fact, a Duke University phase II trial found a markedly higher number of temozolamide responders (15 pts vs. only 1 pt) when the percent of of AGT-positive cells is <20% compared to >20%. This leaves much need for therapeutic improvement in pts with AGT-positive tumors.
So how does one overcome these limits? One option would be to come up with a way to decrease AGT levels. BG, or O6-benzylguanine, is a DNA base (guanine) analog that binds AGT and thereby effectively reduces its tumor-protective activity. Phase I safety trials of O6-benzylguanine in pts undergoing surgery for malignant glioma, as well as in combination with reduced doses of temozolamide, suggest an acceptable safety profile and efficacy of O6-benzylguanine in reducing tumor AGT levels. A phase II study of patients resistant to temozolamide shows that BG can restore temozolamide sensitivity in 20% of the refractory patients with anaplastic astrocytoma.
Another approach to address AGT-positive tumors would be to develop alternatives to temozolamide. One alternative under study is to use human interleukin 13 fused to psuedomonas bacterial exotoxin, creating the IL13-pseudomonas toxin conjugate. Cancer cells have a binding site for IL13,. The toxin conjugate is bound to the tumor cell and internalized, after which it causes death of the cell. Normal cells lack this receptor, and thus are not negatively affected. A phase III study is currently randomizing patients to resection + tumor bed gliadel chemotherapy wafers vs. resection and postoperative IL13-pseudomonas toxin delivery into the resection bed via catheters. A phase I study is looking at IL13-pseudomonas toxin therapy in conjunction with radiation therapy +/- concurrent and adjuvant temozolamide. The results of these and other studies designed to advance the treatment of malignant gliomas are eagerly awaited and badly needed.