Impact of New Molecular Understandings on Brain Tumor Clinical Trial Design Performance

Dr. Roger J. Packer, M.D.
ISPNO 2016 ISPNO 2016

The Childhood Brain Tumor Foundation was pleased to be a Silver Sponsor for the comprehensive 17th International Symposium on Pediatric Neuro-Oncology. Our Neurotransmitter Fall-Winter edition will cover other topics from the other conference.

This summary covers a topic from Family Day and Educational Day, held June 11, 2016, related to molecular studies and clinical trials. The presentations were given by Dr. Roger J. Packer, Senior Vice-President, Neuroscience and Behavioral Medicine, Children’s National Health System.

The new biologic understandings of childhood brain tumors have led to the development of an exciting array of potentially more effective therapies. However, to clinically best incorporate these therapies into patient care and evaluate the potential benefits of such therapies, clinical trial designs for patients must change. This includes trials done by national organizations such as Children’s Oncology Group (COG) and the Pediatric Brain Tumor Consortium (PBTC), as well as more limited institutional studies and clinical trials supported by industry.

In the past, clinical trials have essentially been broken into three major subtypes, as described below. The subtypes remain, but trials need to get smarter and shorter to answer important questions in the care of children with brain tumors in a world that is rapidly changing.

General Principles of Clinical Trial Design 

Phase I clinical trials are primarily designed to determine the maximum tolerated dose of a drug. For molecular targeted therapies, where toxicities may be less severe than standard chemotherapeutics, the maximum tolerated dose may not be reached and instead the trial is modified to determine the optimal dose to saturate the molecular target that can be safely given. Although phase I clinical trials are less likely than other stage trials to show clinical benefit for patients, they may result in tumor control, especially if the agent used is a molecular agent and the patient harbors a tumor which has a genetic aberration that is targeted by that molecular agent. Phase I trials are done in groups of three patients to assess toxicity at each drug dose, although another form of these trials is to use a rolling six design. A limitation of these trials is often availability, as the trials must be temporarily halted until each level of testing is completed.

Phase II clinical trials are performed to determine the efficacy of the agent chosen. Trials with molecularly-targeted agent are more effective when the patients chosen for the trial are pre-screened for having the molecular abnormality thought to be targeted by the drug being utilized. Biomarkers are often built into phase II clinical trials. For these trials, improved outcome measures are needed, such as assessment of functional outcomes, rather than just relying on radiographic

Phase III trials are the goal standard and are randomized trials where the novel therapy is assessed against the best available therapy. Phase III trials are usually of longer duration, possibly for 5-10 years. Just as in phase II trials, better outcome measures are needed, especially functional outcome measures.

All of these trials need to be modified somewhat to better assess efficacy in the molecular era. Because molecular targeted therapy will only work in specific subsets of patients with tumors having the appropriate biologic makeup, such trials are highly dependent on selecting the right study population, the so called “enriched study population”. Most trials are designed only to assess radiographic outcomes. Functional outcomes are often not included and functional outcomes can be more difficult to assess. Also there is reluctance by sponsoring agencies to fund the assessment of the needed functional outcomes, as they may significantly increase cost of the overall study.

To try to better conceptualize these issues, Dr. Packer addressed three tumor types and the impact biology is having on how trials are developed for those tumor types.

Challenges for Medulloblastoma

Medulloblastoma is the most common pediatric malignant brain tumor. One of the primary challenges for medulloblastoma is that medulloblastoma is not one tumor, but it is really multiple different tumors (at least biologically) lumped under the same name. Survival rates have been shown for patients with medulloblastoma to be as high as 90% at five years, the majority of 5-year survivors being cured of their disease. Although survival has improved, patients may suffer long-term sequelae from the tumor or its treatment.

It is no longer rational to put all patients with medulloblastoma on a trial treating patients in the same way, as subgroups of patients with medulloblastoma have varying genetically-driven prognoses. For example, the WNT subgroup, most commonly found in older children, has the best survival and trials need to be designed that actually reduce therapy for those patients. A second major subset of patients with medulloblastoma, those with the Sonic-Hedgehog signature, are treated differently dependent on whether they have what is known as an “upstream” or “downstream” mutation in the Sonic-Hedgehog pathway. Therapy has to be tailored to those groups differently. As example, infants with one variety of Sonic-Hedgehog driven medulloblastomas are probably best treated with chemotherapy alone.

There still needs to be a consensus reached in how to best identify these molecular subtypes in those with medulloblastoma. Molecular testing is not available at many sites and it may take time to obtain molecular results, risking delays in treatment. Also, studies in patients with medulloblastoma, especially phase III studies, have to be shorter in duration, as the science moves quickly.

Challenges for Low-Grade Glioma (LGG)

The understanding of LGGs is in rapid transition. Many questions exist: including whether BRAF-fusion pi tumors are the only true type of pilocytic astrocytoma and whether non-BRAF fusion tumors, even though, at times, having histological features consistent with pilocytic tumors, should be considered a different tumor type. For LGGs it is unclear what is the best outcome measure is. Should response be based on neuroradiographic features alone or should it be based on functional improvement? It is known that in some children with LGGs may have radiographic improvement, but functional deterioration, especially in vision. Molecular targeted therapies are available to treat low grade tumors and their use needs to be once again studied in molecularly enriched populations.

Challenges for Diffuse Intrinsic Pontine Gliomas (DIPGs)

Over the past 50 years, despite multiple different attempts at altering therapy, no improvement has been seen in the outcome of children with DIPGs. Until recently the biology of these tumors has been poorly understood, primarily because tissue was not available for analysis. Newer protocols are requiring biopsy, in attempts to couple the new molecular understandings of these tumors to novel therapies.

In addition, DIPGs are just one type of diffuse infiltrating lesions that occurs in the brain. It is likely that infiltrating glial tumors, which involve the thalamus or the entire brain (gliomatosis cerebri), are biologically similar lesions and should be treated as a DIPG, if they have the same molecular signature.

There is an overall consensus that tissue removed by stereotactic biopsy is needed, not only for research purposes, but also for management. Autopsy tissue is also another very important type of tissue resource and this requires difficult discussions with the family, usually prior to the child’s death. Molecular agents are nearly ready for treatment of childhood DIPGs and these agents must be rapidly incorporated into clinical trials. 

To do all of these trials efficaciously, partnerships and trust between the physicians, the patients and their families are mandatory. Also there has to be a close working relationship with government agencies to allow these trials to move ahead more quickly and with industry to provide the needed drugs. The future of childhood brain tumor therapy is based on more exact molecular profiling of these tumors and using this information to determine what type of treatment is best for the child.

The Childhood Brain Tumor Foundation thanks Dr. Roger J. Packer for reviewing and including additional detail to the contents of this summary.

Thanks to Our Sponsors