Skip to main content
Premium Trial:

Request an Annual Quote

Study Supports MDxHealth's MGMT Methylation Test as Strong Brain Cancer Prognostic


By Molika Ashford

A multi-center study that used MDxHealth's MGMT methylation test to characterize brain cancer progression determined that the diagnostic was able to gauge which patients would have a survival advantage when treated with the oral alkylating agent temozolomide.

In the study, published this month online in the Journal of Clinical Oncology, researchers led by Jaime Gállego Pérez-Larraya of France's Pitié-Salpêtrière Hospital evaluated the efficacy and safety of temozolomide alone in a cohort of patients aged 70 years or older with newly diagnosed glioblastoma. The study found temozolomide increased survival compared with supportive care alone, but especially so in patients with methylated MGMT.

In 31 tumors evaluated for MGMT promoter methylation, the study authors reported, methylated status was linked with longer progression-free survival of 26 weeks over 11 weeks for non-methylated status, and a longer overall survival of 31 weeks over 19 weeks for the non-methylated group.

Patients' MGMT methylation status was established in the study with MDxHealth's test, called PredictMDx for Brain. The chemotherapeutic temozolomide is marketed by Merck under the brand name Temodar.

According to MDxHealth VP of R&D Wim van Criekinge, the results from the study confirm data presented earlier this year at the American Society for Clinical Oncology annual meeting from a Phase III study of temozolomide, and represents a growing body of research validating the test's prognostic power and "importance for clinical decision-making."

The Phase III study, called RTOG 0525, was a randomized trial investigating how well patients responded when they were given two different doses of temozolomide in conjunction with radiation therapy. While the study did not demonstrate improved efficacy for double dose temozolomide for newly diagnosed glioblastoma patients regardless of methylation status, it confirmed the prognostic significance of MGMT methylation in brain cancer patients.

"Additionally, it demonstrated the feasibility of tumor tissue collection, molecular stratification and collection of patient outcomes in a large trans-Atlantic intergroup trial and established this as a viable clinical trial paradigm," the authors, comprised of researchers from Genentech, Merck, and MDxHealth, conclude in the ASCO abstract.

MDxHealth's PCR-based test predicts which patients will respond to alkylating chemotherapy drugs based on the methylation status of MGMT. When the MGMT gene is methylated, it stops producing a critical DNA repair enzyme, allowing alkylating chemotherapy drugs such as temozolomide to be more effective in fighting cancer.

The positive clinical data comes as the company is seeking US regulatory approval of PredictMDx as a companion diagnostic that assesses best responders to cilengitide, another Merck drug being investigated in a Phase III trial involving MGMT promoter methylated glioblastoma patients identified by MDxHealth's test (PGx Reporter 10/20/2010).

"These types of studies are just reiterating the point that MGMT [promoter methylation] is something you need to know if you are looking at [glioma]," van Criekinge told PGx Reporter this week.

"One of the difficult things… [is] it's really hard to find patient arms without treatment [to allow predictive claims]. That's why we are now more or less stuck in confirming study after study, and I think there are at least 25 or 30 studies [now being conducted] confirming MGMT methylation as a prognostic factor."

According to van Criekinge, one of the issues resolved by the RTOG 0525 study was whether a doubled dose of temozolomide would benefit patients who didn't test positive for methylation of MGMT, and thus shouldn't have an increased sensitivity to the alkylating treatment treatment.

"The idea was, they don't have methylation of MGMT, they are repairing damage, so let's give more agent, let's induce more damage so the enzyme that is there cannot follow and we might induce damage in the tumor and get the response we require," he said. But RTOG 0525 "showed that the double dose regimen didn't make a difference in survival for the patients, so that was [disappointing] from a therapeutic point of view."

Nevertheless, the data suggested that MGMT was an independent prognostic factor in glioblastoma patients, regardless of the drug dose, van Criekinge added.

MGMT methylation is being evaluated as a prognostic in up to 30 other ongoing trials, van Criekinge said, including the Merck Phase III study, CENTRIC.

Van Criekinge believes that the company's proprietary methylation-specific PCR method is the most appropriate technology platform to gauge best responders to alkylating chemotherapies. "Because of PCR we have a very sensitive readout, and that's really needed here … typically the tumor has a minority of molecules that are methylated, so its not really a playing field for sequencing methods, its really [suited] for PCR," he said.

This puts the company in a good position, he said, to maintain a large share of the market for MGMT methylation testing. While other people can do it, he said, and other groups are probably doing test development in earlier phases, [methylation-specific PCR is ] probably the easiest, most robust way of doing it."

In addition to pursuing partnerships with drug companies through the trials van Criekinge mentioned, the company also has plans to try to move PredictMDx beyond brain cancer, into other cancer types.

Because glioma represents a relatively small population of patients, van Criekinge said it is maybe a less obvious choice for developing a full companion diagnostic kit. "In this particular case we can offer it as a service test, where it is quicker for us because we control more and can guarantee a turnaround time of four to five days, which is needed in these trials; especially if it’s a rule-in for a treatment, we can't delay that result too long," he said.

"But in other treatments, if we are talking about breast or ovarian or lung, like with Pfizer or [GlaxoSmithKline], those are bigger markets, and there indeed we are looking for opportunities to collaborate with people who have more capacity to put the final test onto the market."

According to van Criekinge, other cancer types, specifically colon cancer, where around "30 percent of all primary cancers are methylated for MGMT, coinciding with those that have KRAS-activating mutation," are areas where the company thinks there is an opportunity for developing MGMT testing.

Beyond MGMT, van Criekinge said MDxHealth also hopes to explore the role of methylation in other repair genes, taking what it has learned for MGMT and trying to reproduce it elsewhere in the genome.

"It took us a while to move from the candidate gene to the genome-wide approach [with MGMT], so now we are really asking ourselves how many other repair enzymes are in the genome which have an epigenetic input and could be predictive for platinum-based therapies, alkylating agents, and also novel therapies like PARP inhibitors" for treating breast and ovarian cancer, he said.

MDxHealth recently announced an agreement with Newcastle University, Cancer Research Technology Limited, and Pfizer to identify and develop a predictive biomarker based on methylation patterns of repair genes for response to the groups' drug candidate for PARP inhibition, PF-01367338. (PGx, 02/02/2011)

Overall, van Criekinge said he views the possibilities for methylation-based prognostics to be promising.

"There has been this historical debate between people saying a tumor is driven by mutations … and what is turning out to be a theme now is that if you have no repair present, if the repair machinery is broken down, then you accumulate mutation, and the accumulation might drive the tumor," he noted.

"But what is popping up in the literature now is that the reason the repair is not active anymore is methylation … People are seeing the methylator phenotype preceding the mutated phenotype… and it's becoming more clear that methylation should be considered."

Have topics you'd like to see covered in Pharmacogenomics Reporter? Contact the editor at mashford [at] genomeweb [.] com.