I'm pleased to welcome our next speaker, Dr. Ganesh Shankar, winner of the Mahaly Clinical Research Award for his talk entitled Genotype-Based Local Targeted Therapy for Glioma. All of our award winners previously received their awards prior to the presentations. Thank you, Dr. Jones. And I'd like to thank the scientific committee for inviting me to speak here on our recent work on Genotype-Based Local Targeted Therapy for Glioma. And also a big thank you for the honor of the Mahaly Award, which I share on behalf of our team. These are the disclosures for this work. I'd like to motivate our work by presenting a clinical scenario that's familiar to us, especially in light of Dr. Rosenblum's talk this afternoon. The 16-year-old who's presenting with headaches, who's found to have a non-enhancing lesion in the left mesial temporal lobe, 2006. This was biopsied. And incidentally, it was an inconclusive biopsy. She was subsequently followed with serial annual imaging with minimal growth. And then she represented in 2014 with local progression at the site of her initial disease. Since 2009, we've appreciated the central feature of IDH, mutations in IDH genes of one and two, as being a hallmark signature of diffuse astrocytomas. And indeed, in this patient, when we went back to look at her initial tumor, as well as the recurrence, there was a densely nodular positivity for IDH1R132H by immunohistochemistry. When we looked at our own series of patients with IDH mutated gliomas who were treated at Mass General, as well as at Dresden in Germany, we found that similar to this patient, there was a vast majority of patients of 82% who presented with recurrent disease within two centimeters of their initial lesion. To address this pattern of local failure in the natural history of diffuse astrocytomas, we sought to expand on the current surgical strategy of safe maximal resection by introducing the application of local targeted therapy that's predicated on knowing the tumor's genotype at the time of surgery. This required the development of these two technical hurdles. And first, I'll focus on our development of a rapid protocol for returning glioma-specific mutations within an intraoperative timeframe. Over the last 10 years, there have been many broad-based genomic characterization studies of gliomas, which have revealed recurrent clonal hotspot mutations in IDH, the TERT promoter, H3F3, and BRAF. And these mutations correlate well with the grade, the histology, the prognosis, as well as the patient's age. In appreciating this accident of nature, we initially developed a protocol for genotyping IDH1 and the TERT promoter within 65 minutes. And when we published this paper, the two pieces of feedback that we received is, one, we need to make it faster, and the second is to include H3F3 and BRAF in this panel. The basis of this qPCR approach, the sensitivity of this qPCR approach really lies in the inclusion of peptide nucleic acid oligonucleotides designed to be complementary to the wild-type sequence to clamp their amplification, allowing for permissive amplification of mutant alleles with detection using a TaqMan-based system. This approach has allowed us to detect tumor mutant alleles to a purity of 0.1%, or 1 in 1,000 cells being tumor. And through reengineering of this assay, we have now reduced the tissue-to-result time to 30 minutes. In addition, we developed this as a multiplexed qPCR assay so that we would have GAP-BEH as an internal control to ensure validity of the reaction. And this is important for developing this as a clinical test, and together with our clinical molecular pathology lab at Mass General, we've now begun rolling this out as a CLIA-certified molecular test. We validated this panel, which was, we validated this panel on 87 clinically-annotated brain tumor specimens, and now our panel includes BRAP as well as H3F3. And greater than 90% of glioma specimens could be characterized by one of these mutations. So now I'll turn to the second part of the workflow, which is the development of sustained release formulations that can be applied at the surgical margin. We, for this, this part of this work, we developed our strategy against IDH-mutated gliomas, given that they're well-characterized and represent a prototypic tumor for this approach. What we know from early clinical trials is that direct inhibition of mutant IDH enzyme has not yet yielded promising clinical results in the management of recurrent low-grade gliomas. However, Dan Cahill's lab at Mass General recently identified a sensitivity of IDH mutant cells to the inhibition of NAMPT, which is an enzyme required in the, as a salvage, uses a salvage pathway for IDH mutant cells in the generation of NAD. GMX1778 is a brain-permeable NAMPT inhibitor and can be dosed orally, and we know therapeutic concentrations of this compound in the brain following a single dose of this compound. However, after five days of this compound orally dosed, we noted significant weight loss as well as anemia and uremia, and we and others have also noted retinal toxicity as well as hepatotoxicity associated with systemic administration of this compound. And this has really precluded the translation of this compound into clinical use for oncology. Acknowledging that systemic administration of NAMPT inhibitors may not be possible given the toxicity profile, we then asked whether we could develop a sustained release formulation that could be applied locally. And to achieve this, we collaborated with Gio Traverso and Bob Langer over at MIT to develop a biodegradable copolymer formulation of NAMPT inhibitors in a PLGA backbone. This requires the optimization of multiple formulation parameters to ultimately achieve maximal drug loading, but also minimal drug crystals, as shown in this electron microscopy image. The median size distribution of these NAMPT inhibitor microparticles is on the order of three to four microns, with release kinetics that allow for 60 to 70% of the drug that's loaded to be released over one to two weeks. We then tested the in vivo efficacy of these in cell culture. And we noted a selective vulnerability of IDH mutated cells, including the MGG152, a patient-derived cell line from a diffuse astrocytoma that was in a time-dependent fashion, but not against a glioblastoma cell line such as U87. This was secondary to an off, sorry, an on-target effect through NAD depletion. And when we surveyed multiple cell lines, what we found is a universal sensitivity amongst IDH mutated cells, but not against glioblastoma cells, which are IDH wild type, but toward promoter mutated, or in DIPG cell lines, which are characterized by H3F3. And this underscores the fact that to reliably and confidently use a form of local target therapy such as this, we would need to know the molecular phenotype of the tumor at the time of the surgery. So then we tested this in an orthotopic model of glioblastoma for its in vivo efficacy using the U87 cell line as well as in diffuse astrocytomas using an MGG152, which is IDHR132H mutated. These cells were engineered to express luciferase so that we can monitor tumor burden in vivo. And what we found with serotactic implantation of microparticles in the U87 orthotopic model is that there was no change in tumor growth, nor in the overall survival of these animals. However, when these microparticles were implanted in the MGG152 established tumor, we noted a greater than hundredfold decrease in tumor volume growth, as well as an increase in overall survival between 58 to 79 days. So we proposed this as one potential avenue for the future workflow in neurosurgical oncology. Currently, we are working on developing a strategy like this to target H3F3 using EZH2 or HDAC inhibitors and BRAF using BRAF and MEK inhibitors. And lastly, I'd just like to acknowledge this largely collaborative effort, which made this work possible. I'd like to highlight Dan Cahill's work and his lab who have pioneered work on IDH biology, as well as my mentor Dr. Carter, as well as our collaborators at MIT, Dr. Traverso and Bob Lang. Thank you.

Genotype-Based Local Targeted Therapy for Glioma

Mahaly Clinical Research Award

diffuse astrocytomas

genotyping glioma-specific mutations

sustained release formulation of NAMPT inhibitors