WASHINGTON, D.C.—Tumors with mutations in the proteins isocitrate dehydrogenase-1 or -2 (IDH1/2) exhibited features similar to BRCA-mutant tumors and were more likely to respond better to PARP inhibitors than to IDH inhibitors, according to preclinical data presented at the AACR Annual Meeting 2017, April 1-5 (Abstract LB-290).
"IDH is an enzyme that plays a role in the citric acid cycle, the energy-producing unit of the cells, and IDH mutations are commonly found in brain tumors and some leukemias," said Ranjit Bindra, MD, PhD, Assistant Professor of Therapeutic Radiology and Experimental Pathology at Yale School of Medicine and Yale Cancer Center in New Haven, Conn. "We found that the oncometabolites produced by IDH mutations induce a state in the cell where DNA repair is profoundly inhibited. This essentially makes them quite similar to breast and ovarian cancers which harbor mutations in key DNA repair genes such as BRCA1 and BRCA2.
"Our data strongly suggest that exploiting this DNA repair deficiency in IDH-mutant tumors, rather than inhibiting the function of the mutant IDH proteins, likely will be a better strategy for treating brain tumors with these specific mutations, a devastating disease with an urgent need for better therapies," he added.
IDH mutations are neomorphic, meaning that they change the normal function of the encoded protein, Bindra explained. Instead of making molecules that produce energy in the cell, the mutant IDH proteins start making an oncometabolite, 2-hydroxyglutarate (2HG), which is thought to be important for tumorigenesis.
Based on this knowledge, therapeutics that block the mutant IDH1 and IDH2 proteins and decrease the levels of 2-HG are being developed and tested in clinical trials. However, it is not entirely clear whether these therapeutics have been effective in shrinking solid tumors, according to Bindra.
"The remarkable thing is that we are proposing a therapy which does exactly the opposite of what IDH inhibitors are designed to do," Bindra said. Laboratory research led by his team showed that by suppressing DNA repair, the IDH mutations essentially create an Achilles' heel in the tumor cell, and that this weakness can be exploited by treating these tumors with PARP inhibitors, instead of IDH inhibitors. "In fact, small molecule inhibitors of mutant IDH actually reverse PARP inhibitor sensitivity."
A PARP inhibitor is a DNA repair inhibitor, and when applied to a tumor that has a DNA repair defect, the effects of the defect and inhibitor are synergistic, as a result of a process known as "synthetic lethality," leading to tumor cell death. Therapeutics developed based on this approach have been effective in treating hereditary breast and ovarian cancers, and the FDA has approved two PARP inhibitors, olaparib and rucaparib, to treat certain BRCA-mutant ovarian cancers.
Bindra, Peter Glazer, MD, PhD, and colleagues conducted a series of laboratory experiments to show that IDH-mutant patient-derived glioma cells, bone-marrow cultures of IDH-mutant acute myeloid leukemia, and mice bearing IDH-mutant human tumor cells were all susceptible to PARP inhibition. Results from these studies were published recently in Science Translational Medicine.
Since then, the investigators have conducted a series of experiments in genetically matched IDH-wild type (normal form of the protein) and IDH-mutant cancer cell line models to directly compare the efficacy of IDH inhibitors and PARP inhibitors in inhibiting IDH-mutant tumor cell viability. "When you carefully control the conditions and simply ask the question of which drugs best target solid tumor cell lines with IDH mutations, we do not see a significant effect of mutant IDH inhibitors on viability, however, we do see profound differences with PARP inhibitors," said Bindra.
They also found that in addition to R-2HG, the other form of 2HG, S-2HG, also induced a similar DNA repair deficiency, which made the brain tumor cell lines doubly sensitive to PARP inhibitors. S-2HG is produced in response to tumor hypoxia, and also at baseline levels in subsets of renal cell cancers and pediatric gliomas.
"We believe that exploiting oncometabolite-induced defects in IDH-mutant cancers may be the best path forward," said Bindra. "We will be presenting for the first time the schema for a multicenter, biomarker-driven NCI-sponsored phase II trial across approximately 35 centers in the United States, which will directly translate our recent discovery to patients," Bindra said.
Limitations of the study include that their hypothesis has not been tested specifically in patient-derived, intracranial glioma xenograft animal models, Bindra said.