Identifying Low-Risk Medulloblastoma to De-escalate Therapy

Liam Davenport

November 08, 2018

Some children with the most common form of malignant brain cancer could receive lower-dose radiotherapy and chemotherapy, depending on the chromosome signature of their tumor. Such an approach would improve patients' quality of life and could potentially increase cure rates, say British researchers.

The comments come after a new analysis of clinical trial data, conducted by Steven C. Clifford, PhD, from the Northern Institute for Cancer Research at Newcastle University, Newcastle upon Tyne, United Kingdom, and colleagues.

The team analyzed tumor samples from more than 130 medulloblastoma patients from a previous clinical trial.

They found that among two subgroups of patients whose conditions had traditionally been associated with heterogeneous outcomes, a combination of chromosomal alterations could identify a subgroup at favorable risk, for whom the 5-year event-free survival rate was 100%, and a subgroup at higher risk, whose event-free survival was 60%.

"Our findings provide a new blueprint for the personalization of treatment in medulloblastoma so that all children are not given the same intensity of therapy," Clifford said in a press statement.

"This study shows that low-risk patients may receive kinder treatments aimed at reducing toxicity and side effects, while targeting more intensive treatments to the high-risk patients who need it most," he added.

He hopes that this will increase the cure rate for medulloblastoma, "but critically, for those children who survive, we want to make sure their quality of life is good after treatment," he said.

Coauthor Edward C. Schwalbe, PhD, senior lecturer in bioinformatics and biostatistics at Northumbria University, Newcastle upon Tyne, underlined that the chromosome signature is "easily testable" and "predicts excellent outcomes.

"Importantly, we also show that this signature works as expected in an independent set of patients," he said.

Echoing Clifford's comments, Schwalbe noted: "Looking forward, we hope that children whose tumors have this signature could be treated less aggressively, reducing the life-long side effects of the grueling treatments, while maintaining a cure."

The study was published online November 1 in the Lancet Oncology.

Speaking to Medscape Medical News, Clifford said that that although tests based on the risk profiles they identified could be implemented immediately, the results of their study would need to be validated before the tests are introduced to the clinic.

He explained that all the techniques used in the study are already used routinely in his institution to characterize medulloblastomas. "Technically," he said, "we're set up, we can absolutely do it, and we can turn those results around in real time."

He continued: "We've made these really exciting findings, we showed in the study that...we could reproduce those findings in an independent cohort, but we need to take them forward into clinical trials and into a clinical setting.

"We would really like those results replicated by an independent group, perhaps one of the US trials groups, and then, once we have that confidence, we'd then look set to up our next clinical trial incorporating that information," he said.

Several Distinct Profiles

Medullobastoma comprises several different diseases that have distinct genetic and biological profiles.

The classification of medulloblastomas into separate groups is an ongoing process. Currently, medulloblastomas are classified into four broad categories:

  • WNT, the least common, which has the most favorable risk profile;

  • SSH, the highest-risk type;

  • Group 3, non-WNT/non-SSH, typically characterized by MYC amplification; and

  • Group 4, non-WNT/non-SSH, without MYC amplification.

For the two latter groups, which have similar biological profiles, outcomes following treatment are heterogeneous, the researchers note.

Previous studies have shown that "standard-risk" medulloblastoma, which accounts for up to 60% of cases, is characterized by the absence of high-risk features and has a 5-year event-free survival of 75% to 85%.

However, it has not been possible to accurately and reproducibly know which non-WNT patients have a favorable prognosis. As a result, all patients currently undergo surgery followed by radiotherapy and chemotherapy.

The researchers therefore examined data from the prospective SIOP PNET 4 trial, which recruited 338 patients aged 4 to 21 years with medulloblastoma from 120 centers in seven European countries between 2001 and 2006.

For the trial, the patients were randomly assigned to receive standard or hyperfractionated radiotherapy, followed by eight cycles of chemotherapy. There was no difference in event-free survival between the groups.

Retrospective Analysis of Tumor Samples

As a part of the current study, formalin-fixed, paraffin-embedded tumor samples were collected. This allowed the researchers to retrospectively analyze samples from 136 trial participants. The average age of the patients at diagnosis was 9.0 years, and 62% were boys.

Owing to the scarcity and low quality of available genomic material in the tumor samples, the team determined the methylation subgroup using a mass spectrometry–minimal methylation classifier assay and a molecular inversion probe array to detect genome-wide copy number aberrations.

Clifford explained that the material was not "really up to the standards of being able to do current contemporary genomics on it, so we had to develop the methods that were necessary to extract the information that we needed from those samples.

"Of course, things have changed now," he said. "We now collect both more and better-quality material routinely, but this was really about getting the very best information we could out of this really well-described, well-annotated clinical trial."

The team reports that 28 of the patients (21%) had WNT disease, 17 (13%) had SSH disease, and 91 (67%) had disease of non-WNT/non-SSH groups 3 and 4.

After a median follow-up of 6.7 years, overall, there were no significant differences in 5-year event-free survival between WNT, SSH, group 3, and group 4 patients.

There was, however, a significant improvement in survival of patients with tumors carrying ≥1 whole chromosomal aberrations compared to those without such aberrations; the 5-year survival rate was 87.1% vs 60.0% (hazard ratio [HR], 4.05; P = .00077).

Among WNT patients, 5-year survival was 100% among those aged <16 years at diagnosis. Three relapses occurred, all in patients aged 16 to 21 years. Among SHH patients, four relapses occurred, all in those with a TP53 mutation or chromosome 17p loss. Both of these findings confirmed findings from previous studies.

In non-WNT/SSH group 3 and 4 tumors, a novel whole chromosomal aberration signature comprising at least two of chromosome 7 gain, chromosome 8 loss, and chromosome 11 loss was found to be associated with a favorable prognosis.

The signature was identified in 38 patients (42%). For these patients, the 5-year event-free survival was 100%, vs 68% for higher-risk patients whose disease did not have those features (P = .00014).

Tested in Validation Cohort

The team then tested the signature in an independent, demographically matched validation cohort of 70 medulloblastoma patients with non-WNT/non-SSH standard-risk disease who were treated between 1990 and 2014.

The 5-year event-free survival was 94.7% in the favorable-risk group, vs 58.6% in the higher-risk group (HR, 7.41; P = .029).

Overall, the researchers were able to develop subgroup-specific risk models that allowed 69 of the 134 patients (51%) for whom complete data were available to be assigned to a favorable-risk group.

Ongoing Clinical Trial

Currently, the team is participating in a clinical trial in which approximately 10% of patients with a favorable-risk profile are being treated with reduced radiotherapy and some chemotherapy.

Clifford said that the current findings will potentially allow them "to increase that group from 10% to more like 40%, and that really is the next big step.

"I would have thought in the next probably 2 to 5 years, we would be in a position to have a next trial or study planned, based on the information we've got, again provided it validates," he said.

He emphasized that de-escalating therapy must, for now, take place in the context of a clinical trial, because this would ensure that it would be "highly controlled.

"It's got to have appropriate stopping rules as well, such that, if those reductions do happen to cause increased relapses or anything like that, then we need to be able to stop that and stop quickly," he said.

This Study "Sets the Foundation"

In an accompanying comment, Elisabetta Ferretti, PhD, and Agnese Po, PhD, from Sapienza University of Rome, Italy, say that the risk model identified in the study is "more solid" in comparison with previous models and increases the number of patients who can be assigned to therapy de-escalation.

They point out that other molecular markers, such as noncoding RNAs, should also be considered, because they "have been shown to be important regulators of medulloblastoma biology, and could represent valuable biomarkers for risk stratification and targeted therapy."

They conclude that although the results require validation in a multicenter study, "they set the foundation that is needed to improve genomic patient characterisation for more accurate risk stratification in routine clinical settings."

The study was funded by Cancer Research UK, the Swedish Childhood Cancer Foundation, the French Ministry of Health/French National Cancer Institute, and the German Children's Cancer Foundation. Dr Clifford and Dr Schwalbe have disclosed no relevant financial relationships. Coauthor François Doz, MD, has received personal fees from Bristol-Myers Squibb, Tesaro Oncology, Servier, and Celgen. Dr Ferretti and Dr Po have disclosed no relevant financial relationships.

Lancet Oncol. Published online November 1, 2018. Full text, Comment

For more from Medscape Oncology, follow us on Twitter: @MedscapeOnc


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: