Dr. James Boyett

Influence of Central Review on Outcome Associations in Childhood Malignant Gliomas: Results from the CCG-945 Experience

My major contribution to this publication was to insist that the most important inference to be drawn from my analysis is that eligibility for clinical trials should not be determined by retrospective histological reviews by recognized “expert” neuro-pathologists whose decisions, unlike decisions of the institutional pathologist, are only academic and in no way affects treatment of the patients. This work points out clearly that “experts” differ and should that be a surprise to anyone? After all, “experts” are so recognized because they do differ from the norm and have convinced others that their difference makes a difference!

The attached PowerPoint presentation demonstrates that the survival distribution of childhood malignant high-grade gliomas is more dependent on the neuro-pathologist than the patient’s treatment; thus histological diagnoses of malignant gliomas perhaps should be subscripted by the initials of the pathologist making the call; e.g., GBMJMB, AARAF, etc.

These data strongly argue objective criteria based on molecular or genetic data are needed to classify malignant gliomas. This is an emphasis of the Brain Tumor and Neurobiology program in the St. Jude Children’s Research Hospital Cancer Center.

This article appears in The Neuro-Oncology 2003.  Other co-first authors of the paper include Ian Pollack (Children’s Hospital of Pittsburgh), Allan Yates (Ohio State University), Peter Burger (Johns Hopkins University School of Medicine), Floyd Gilles (University of Southern California), Richard Davis (University of California, San Francisco), and Jonathan Finlay (New York University Medical Center).