Amar Gajjar, MD

My work has focused on gaining a better molecular understanding of medulloblastoma and using this information to improve our risk-stratification models. Our findings also inform the next generation of medulloblastoma treatment protocols.

Tumors of the central nervous system make up approximately one-third of all childhood cancers. Of these, medulloblastoma is the most common malignant brain tumor diagnosed in children. Previously, medulloblastoma was thought to be a single disease entity, and all patients received a similar treatment strategy. We now know that this disease is genetically multifaceted, consisting of four molecularly distinct groups: WNT, SHH, Group 3 and Group 4. 

We previously relied on clinical characteristics to classify disease as high- or low-risk as well as inform the course of treatment. We are now applying molecular techniques–such as DNA methylation profiling and next-generation sequencing–to medulloblastoma tumors, and we have shown these tumors to be molecularly diverse.

Four groups of medulloblastoma have been established, each with a unique molecular profile. Variations have been noted, for example, in each group’s expression of mutations affecting oncogenic cell signaling and in each group’s pattern of DNA methylation. Ultimately, these molecular differences affect prognosis.  

We are now taking these molecular idiosyncrasies into consideration as we adjust our treatment strategies for patients. Our goal is to lessen therapeutic intervention for children with less aggressive disease and introduce novel therapies to those exhibiting molecularly and clinically aggressive disease. This tailored approach will also help us to lower treatment-related side effects for some of our patients. 

In addition, we are actively engaging in intervention studies to assess and subsequently minimize treatment-related side effects such as cognitive deficits and growth inhibition.

While frontline multimodal therapy cures approximately 70% of patients, those with progressive disease stand a high chance of succumbing to the cancer. Therefore, it is imperative that we monitor measurable residual disease (MRD) in conjunction with our treatment strategies. The inability to detect MRD suggests a positive response to treatment while those with persistent MRD-positive tests have a significantly higher risk of progression.

We are currently monitoring MRD via cerebrospinal fluid (CSF)-derived cell-free DNA. Medulloblastoma genomes are characterized by tumor-associated chromosomal copy-number variations (CNVs), and we have successfully shown that we can use the presence or absence of CNVs in the spinal fluid to determine MRD. This finding has opened the door for using CSF as a clinically relevant biomarker for medulloblastoma and possibly other cancers.