Biomarkers In Brain Tumors With Focus On Glioblastoma

The primary brain tumors, gliomas, are not very common but they are deadly. Each year in Sweden around 500 patients will be diagnosed with a glioma and unfortunately most of them will have the most aggressive type, glioblastoma (GBM). Median survival, even if treated, is poor (14-17 months). Males are diagnosed up to 60% more often than females and they often have a worse prognosis. GBM affects mainly older patients and treatment includes radiochemotherapy with temozolomide (TMZ). The only known predictive biomarker for TMZ treatment is methylation of the O6-methylguanine DNA methyltransferase (MGMT) promoter and patients with methylated MGMT have better outcomes. There is a great need for biomarkers to decipher existing sex differences and others that identify patients that will benefit from radiotherapy (RT) or TMZ despite of unmethylated MGMT. Papers I-III are focused on investigating sex differences in GBM and in Paper IV we examined the methylation-based biomarkers used in diagnostics, on patients from the Nordic trial with exceptionally good and poor survival when treated with TMZ or RT. Loss of the Y chromosome (LOY) in male’s blood cells is associated with aging and, among other diseases, with cancer. We looked at 10 genes located on chromosome Y in 105 males with GBM treated with TMZ concomitant with RT and found that they are often deleted. Detected LOY, as well as deletion of the sex determining region Y (SRY) gene were associated with shorter overall survival. Low SRY gene expression analyzed in an additional cohort of 219 samples from The Cancer Genome Atlas (TCGA) was also associated with a shorter survival. In Paper II we re-analyzed data from three cohorts to compare the frequency of MGMT methylated tumors in males and females and investigated whether sex is an important factor associated with patient’s survival. This was done in a GBM cohort from the randomized, phase 3 Nordic trial, which included patients 60 years or older, treated with standard RT (60Gy) vs. hypofractionated RT vs. TMZ given in up to six 4 weekly cycles; in a population-based cohort, treated with TMZ concomitant with RT and an excerpt of the TCGA cohort of patients treated with different modalities. In all three cohorts there was a higher fraction of MGMT methylated tumors in females and MGMT methylation was predictive of longer survival for those treated with an alkylating agent, such as TMZ. The third study investigated the androgen receptor (AR), located on chromosome X as a potential sex susceptibility factor for GBM. We found that the gene encoding for AR can be amplified or deleted in GBM, and these changes are more common in females. The AR gene expression was enhanced in GBM but did not differ between sexes. At the same time, in a separate analysis for males and females, we found that high AR expression is associated with shorter survival in females and longer survival in males Also, the methylation sites in the AR promoter that correlated with gene expression are sex specific. In Paper IV we included 59 patients from the Nordic trial, equally divided by the treatment arms and MGMT methylation status, with good prognostic factors and with long or short survival. We performed genome-wide methylation analysis and identified differentially methylated sites between those with long and short survival for the TMZ treated, MGMT methylated samples, as well as the 60Gy, MGMT unmethylated and 34Gy MGMT methylated samples. This small pilot study was unable to discern any differentially methylated sites in TMZ treated samples with unmethylated MGMT, associated with long or short survival. By using a methylation-based diagnostic classifier, we were able to detect a misclassified sample that was not a GBM. Lastly, we calculated so called ‘epigenetic age’ of the tumor tissue based on the methylation data and using three different algorithms and found that in all treatment groups short-term survivors tended to have lower epigenetic age, though these results were not significant and need verification in a larger cohort. In summary, this thesis advances our knowledge on the molecular differences between male and female GBM and the association between these alterations and patients’ survival. Our results also suggest that there are potential methylation-based biomarkers apart from the MGMT promoter methylation, that can be used to distinguish between patients with good and poor prognosis, for instance, the epigenetic age.ISBN (PDF) has been added in the PDF-version.