Kinomics platform using GBM tissue identifies BTK as being associated with higher patient survival.

Better understanding of GBM signalling networks in-vivo would help develop more physiologically relevant ex vivo models to support therapeutic discovery. A "functional proteomics" screen was undertaken to measure the specific activity of a set of protein kinases in a two-step cell-free biochemical assay to define dominant kinase activities to identify potentially novel drug targets that may have been overlooked in studies interrogating GBM-derived cell lines. A dominant kinase activity derived from the tumour tissue, but not patient-derived GBM stem-like cell lines, was Bruton tyrosine kinase (BTK). We demonstrate that BTK is expressed in more than one cell type within GBM tissue; SOX2-positive cells, CD163-positive cells, CD68-positive cells, and an unidentified cell population which is SOX2-negative CD163-negative and/or CD68-negative. The data provide a strategy to better mimic GBM tissue ex vivo by reconstituting more physiologically heterogeneous cell co-culture models including BTK-positive/negative cancer and immune cells. These data also have implications for the design and/or interpretation of emerging clinical trials using BTK inhibitors because BTK expression within GBM tissue was linked to longer patient survival.

MYD88 L265P mutation in primary central nervous system lymphoma is associated with better survival: A single-center experience.

BACKGROUND: The myeloid differentiation primary response gene (MYD88) mutation in primary central nervous system lymphomas (PCNSL) may be associated with unfavorable prognosis; however, current evidence remains limited. We aimed to characterize PCNSLs by integration of clinicopathological, molecular, treatment, and survival data. METHODS: We retrospectively identified and validated 57 consecutive patients with PCNSLs according to the 2017 WHO classification of lymphoid neoplasms over 13 years. Formalin-fixed paraffin-embedded tumor samples underwent polymerase chain reaction assay to detect MYD88 mutation. We used Cox regression for survival analysis, including age, treatment, and MYD88 as covariates. We searched the literature for studies reporting demographics, treatment, MYD88, and survival of PCNSL patients and incorporated individual patient data into our analyses. RESULTS: The median age was 66 years and 56% were women. All 57 patients had PCNSL of non-germinal center cell subtype and the majority (81%) received either single or combined therapies. There were 46 deaths observed over the median follow-up of 10 months. MYD88 mutation status was available in 41 patients of which 36 (88%) were mutated. There was an association between MYD88 mutation and better survival in the multivariable model (hazard ratio [HR] 0.277; 95% confidence interval [CI]: 0.09-0.83; P = .023) but not in a univariable model. After incorporating additional 18 patients from the literature, this association was reproducible (HR 0.245; 95% CI: 0.09-0.64; P = .004). CONCLUSIONS: Adjusting for confounders, MYD88-mutant PCNSL appears to show improved survival. While further validation is warranted, detection of MYD88 mutation will aid the identification of patients who may benefit from novel targeted therapies.

Clinical importance of molecular markers of adult diffuse glioma.

In 2016, the WHO incorporated molecular markers, in addition to histology, into the diagnostic classification of central nervous system (CNS) tumours. This improves diagnostic accuracy and prognostication: oligo-astrocytoma no longer exists as a clinical entity; isocitrate dehydrogenase (IDH) mutant and 1p/19q co-deleted oligodendroglioma is a smaller category with better prognosis; IDH wild-type 'low-grade' glioma has a much poorer prognosis; and glioblastoma is divided into IDH mutant (with an better prognosis than pre-2016 glioblastoma) and IDH wild type (with a poorer prognosis). Previous advice based on phenotype alone will change with respect to median survival, best management plan and response to treatment. There are implications for routine neuropathology reporting and future trial design. Cases that are difficult to classify may need more advanced molecular genetic classification through DNA methylation-based classification of CNS tumours (Heidelberg Classifier). We discuss the practical implications.