NEW YORK (GenomeWeb) – Researchers from several Boston-area academic and research institutions have developed a rapid multiplexed PCR-based genotyping tool to diagnose patients with a mutated brain tumor cell line during surgery, which they believe could be used in tandem with a therapeutic tool to eradicate the glioma.
Stage II diffuse gliomas are normally challenging to identify using intraoperative histologic methods because of their location, infiltrative growth, low cellularity, and small stereotactic biopsy size. Diagnosis often requires the patient to undergo multiple neurosurgical procedures to collect tumor tissue samples.
In a study published earlier this week in the Proceedings of National Academy of Sciences, the researchers described a simultaneous precision intraoperative treatment method that integrates rapid genotyping of IDH mutations — which are typically associated with lower-grade gliomas — with a sustained release microparticle drug delivery system targeting IDH-mutated cells.
Study author and Harvard Medical School assistant professor Giovanni Traverso explained that his team sought a method to quickly treat glioma patients with cancer-inhibiting drugs during neurosurgery the operating room.
In order to guide the correct metabolic local therapy for gliomas, however, the researchers required a workflow that "would identify the genotype fast enough to fit within the four-hour timeframe during the surgery and facilitate the selection of the therapeutic agent," Traverso said.
The researchers' multiplexed qPCR solution was previously described in an August 2015 study in JAMA Oncology, and used to detect somatic single-nucleotide variants in TERT promoter and IDH1 genes to diagnose low-cellularity tumors like glioma.
Daniel Cahill, a neurology associate professor at Massachusetts General Hospital and author on both papers, noted that brain tumors are particularly challenging to diagnose. While other cancers grow as a mass of tumor cells, gliomas are "infiltrative tumors" and are often mixed with normal blood cells.
"Normally, this is an issue for PCR techniques, since most of the alleles you're going to be amplifying from the tumor mass are actually normal alleles from normal cells," Cahill said.
Cahill's team needed to figure out a way to suppress normal alleles, while only allowing amplification events to occur if mutant alleles were present.
In the 2015 study, the team incorporated both peptide nucleic acid oligonucleotides that block the amplification of wild-type alleles, as well as locked nucleic acids to increase specific binding to the mutant allele acids in the detection probes. The team then applied the assay to tissue samples from 190 patients with diffuse gliomas, including frozen and fixed samples collected intraoperatively.
By blocking the wild type allele, the team's assay successfully identified the key specific mutation in the qPCR product for grade II and grade III gliomas, with a 96 percent clinical sensitivity and a 100 percent clinical specificity.
Cahill's team also identified glioma-defining mutations in a series of live cases within an hour, demonstrating its potential use as a diagnostic for gliomas during a neurosurgery timeframe.
"[We] have found that a high frequency of cases can be classified correctly by scoring a handful of these point mutations," Cahill said. "We suggested an opportunity, where we could, through an optimized PCR technique, prove a clinically useful categorizing technology in a rapid way."
In the past three years, Cahill and his team have halved the assay time to 30 minutes or less, further increasing its utility. In addition, Cahill's group added three more alleles — BRAF, IDH2, H3F3A— that categorize the 1 percent to 2 percent of glioma cases undetected by the initial set of biomarkers.
In the study published this week, Cahill and his researchers first characterized the patterns of IDH mutant glioma progression by analyzing a cohort of 130 patients who underwent resection. Afterwards, they integrated the genotyping tool with the drug delivery system on glioma tissue samples resected from neurosurgery operations to develop a prototype for the proposed combined surgical model.
By suppressing wild-type alleles, the team detected mutations in IDH1, TERT promotor mutations, H3F3A, and BRAF within 27 minutes.
Cahill's team then validated the genetic diagnostic tool on extracted human samples, testing the assay on 87 clinically annotated brain tumors. They found that the presence of at least one or more mutations captured 75 brain tumor samples, noting in their paper that the high rate of "positive assignment allows for the confidence to distinguish tumor versus nonneoplastic pathologies."
Cahill explained that the strategy allows immediate intraoperative genotyping and a genotype-specific treatment in surgical scenarios. During a surgery, the researcher can take a small sample and immediately analyze it using any standard real-time PCR platform. The qPCR assay runs while the operation occurs, and the researchers can discover information about the tissue before the surgery finishes.
"This is a very controversial idea … as many [surgeons] would do the [operation] now and look at the genome later," Cahill said. "But now …. it's helpful to know the [genetic] information, especially to pair it with targeted therapy given during the operation."
Cahill noted that the team's main challenge in developing the integrated genotype and targeted therapy tool is to explain to surgeons the demonstrated need or advantage over the standard of care.
"Our key problem is that we have to show that we have 100 percent concordance with the standard procedure," Cahill said. "Researchers and clinicians who are using the existing standards have to understand that it's important and [we are] beholden … to show our technology can help them. "
Cahill believes that by optimizing the diagnostic workflow, the process would improve in the future and offer diagnostic opportunities for clinicians.
Cahill also noted that the workflow could be used in surgical oncology for other tumors characterized by targetable molecular alterations. While the TERT promoter and BRAF mutations are rare in brain tumors, Cahill noted that the SNVs are very common in melanoma.
While acknowledging that the clinical situations are "completely different in the two cancers," Cahill highlighted that having rapid genotyping tools in different clinical spaces would be very useful. However, he noted that a melanoma expert, for instance, would need to identify the utility of the information in a specific clinical scenario.
Moving forward, Traverso said that the team is applying for more grant funding as it pushes the tool for clinical use. Meanwhile, Cahill said that the team has filed patents for both the genotyping tool and sustained release microparticle drug delivery system. The researchers are also in talks with undisclosed commercial partners to further develop the assay.
"Ideally, we'd like to begin human clinical testing in two to three years, if not sooner," Traverso said. "We want to have this technology available in the operating room, as we see tremendous opportunity in how it can help patients.