Proteogenomics of 112 young glioma tumors finds a biological break near age 26

Normal brain development changes with age and differs between males and females, and those changes sit in the same tissue as the tumor. Telling the two apart was the design problem. "Because age and sex shape normal brain development, we wanted to separate features of the cancer from those associated with a patient's developmental stage," said Pei Wang, co-senior author of the paper in Cell Reports Medicine ^[21][15]. Wang is a professor of genetics and genomic sciences at the Icahn School of Medicine at Mount Sinai ^[13].

The work was done by the National Cancer Institute's Clinical Proteomic Tumor Analysis Consortium with the Children's Brain Tumor Network, which collected the samples, and the Philadelphia Coalition for a Cure ^[16][4].

A cohort of 112 patients spread across roughly 40 years of age averages fewer than three patients per year of age ^[3][23]. Modelling profiles as a continuous function of age uses every sample; a hard bin drawn at 26 would be estimated from thin cells on either side ^[7]. The adult end is anchored by 99 glioblastoma tumors, along with clinical and genetic data from a reference group of more than 5,000 people with high-grade glioma. That brings the count of tumors contributing molecular data to 211 ^[5][22].

Computational analysis plus laboratory work flagged four kinases: CDK8, ATM, ATR and LCK ^[8]. In tumor-derived cell lines, knocking out or drugging one of them slowed growth when that kinase was most active ^[9]. CDK8 drew the most attention because it is less well characterized in high-grade glioma than ATM and ATR ^[10]. The argument for it runs through two links: the analyses indicate CDK8 suppresses oxidative phosphorylation, and the study associates oxidative phosphorylation with more favorable survival ^[11]. The first link was tested in cells. The second is an association across patients, which is not evidence that raising OXPHOS extends a life. "Causal network analysis and cell line studies provide a rationale for personalized therapies targeting candidate kinases, such as CDK8," the investigators wrote ^[20].

For trial design, what is on the record is that two AYA groups differ in molecular profile and in survival, and that the researchers suggest using the results to guide trial design and new therapies ^[6][19]. Whether stratifying by developmental stage changes what a trial concludes is a question a trial has to answer. In a disease with a five-year survival rate below 10%, few groups get to run the comparison twice ^[17]. Adolescents and young adults have been understudied in part because their tumors straddle pediatric and adult disease ^[18]. On sex, the authors wrote that "differences between male and female patients, which influence both brain development and cancer biology, represent another underexplored dimension" ^[24]. The paper does not report effect sizes.