Praveen Raju   Assistant Professor of Pediatrics

  • +1 212 746 3278

Laboratory for Childhood Brain Tumor Research

Medulloblastoma is the most common malignant brain tumor in children and one of the leading causes of death in this age group. Over one-third of children with medulloblastoma die within 5 years of diagnosis and the vast majority of survivors have significant neurological deficits due to the toxicity of combined surgery, radiation, and chemotherapy. To improve survival and minimize side effects of current therapy, a more detailed understanding of the biology of medulloblastoma will better characterize the disease and will aid in prognosis and in identification of novel treatments.

Cancer, in general, is a disease that is believed to originate sporadically from one or a few cells within a tissue. After a cell acquires genetic mutations over time, the cell then has a growth advantage over normal surrounding cells in the tissue, and has the potential to become cancerous. Current mouse models of cancer, including those for medulloblastoma, have been generated but have inherent drawbacks and do not accurately reflect the sporadic nature of tumor formation. As a postdoctoral research fellow at MSKCC, I helped to develop a novel genetic technique in mice where we can mutate genes implicated in cancer within a small number of cells in a tissue and simultaneously mark the mutant cancer cell so that it can be unequivocally identified from its normal neighboring cells. This general approach for modeling sporadic cancer will allow the study of mutant tumor cells within the context of a normal surrounding environment during different tumor stages and should more accurately model sporadic tumor formation, which occurs in the majority of human patients.

The ability to identify mutant cancer cells from normal surrounding tissue has tremendous advantages and implications for therapy. Characterizing marked cancer cells at the molecular level compared to normal surrounding cells may provide insight into identifying novel and more effective therapies targeted to cancer cells and not surrounding noncancerous cells. This approach will protect normal surrounding tissue from the toxicity of current treatments that do not have this selectivity.

Furthermore, relapse after treatment of cancers, including medulloblastoma, is also a major issue in the clinical setting. Even if tumor size is reduced significantly after surgery, standard chemotherapy, and/or radiation therapy, cancer cells can often evade these therapies by acquiring new genetic mutations that allow a survival advantage in this setting. The ability to identify marked tumor cells that are resistant to treatment from surrounding cells using our genetic approach in preclinical drug testing will allow us to characterize these new genetic changes and would provide additional targets for developing novel therapies. In addition, as cancer stem cells are believed to be relatively resistant to standard chemotherapy and radiation therapy, the identification of treatment resistant cancer cells in our mouse model will shed light on cancer stem cell biology.

In addition to providing more physiologically and genetically relevant preclinical mouse models for medulloblastoma subgroups, our novel mosaic approach will be easily applicable to other pediatric and adult cancers, including neural and non-neural tumors. My group is also actively working on improved animal models for other pediatric brain tumors including atypical teratoid rhabdoid tumors (ATRT), diffuse intrinsic pontine gliomas (DIPG), and subependymal giant cell astrocytomas (SEGA) that are commonly associated with tuberous sclerosis in order to identify improved therapies that are sorely needed in patients with these devastating diseases.



Sort by

Selected publications


Sort by

Grants awarded



Primary Email