Probability Driven Cancer Progression and Evolution in Human Brain

Abstract

We analyzed 31 temporally successive brain tumors collected from ten patients by applying deep whole genomic and whole transcriptomic sequencing. The genomic landscape showed significant variation between the temporal tumors and that between patients. These differences were primarily driven by different cellular progresses. Only a small portion of the genomic mutations existed in all temporal tumors and likely grounded the basis for initiation and progression. TP53, IDH1, and the telomere maintenance genes (ATRX, DAXX, and the TERT) are recurrently seen in almost all secondary GBM and LGG patients, while the primary GBM tumors greater differentiation of mutant genes between patients. These 'early' cancerous genes surprising predicted the overall survival time (R2=0.90). Regardless malignancy, the tumor progression can be summarized by a simple multinomial mathematical equation. This suggests a uniformed probability on mutating genomes in the cancer cells, where the mutating frequency is escalated by cancerous driving genes. Phylogenetic analysis showed that the cancer evolution is highly influenced intrinsically and extrinsically, e.g. the early mutant genes and the therapeutic regimen. Our results are important on future development of cancer research and diagnosis assay for prevention and therapy.