The prognostic value of clinical factors and cancer stem cell-related markers in gliomas.Dan Med J 2014; 61(10):B4944DM
Gliomas are the most frequent brain tumours among adults, and it is estimated that gliomas constitute half of the about 1500 new brain tumours diagnosed in Denmark every year. Existing treatment strategies include neurosurgery, radiation, and chemotherapy. Therapy selection is based on experiences from clinical trials, with the risk that the results obtained are restricted to highly selected patients only. Moreover, these studies provided only little knowledge of the clinical behaviour of the tumours. For some time, it has been believed that somatic stem cells are responsible for self-renewal, proliferation, and differentiation during development of different (normal) tissues. The same characteristics were identified in cancer cells, and recently a major part of the glioma research has focused on the cancer stem cell (CSC) hypothesis, suggesting that only CSCs posses the ability of initiating new tumours. Moreover, CSCs have been suggested as the cause of resistance towards radiotherapy and chemotherapy. In gliomas, CSCs were originally identified by means of the expression of CD133, but other proteins have subsequently been suggested as CSC related. To improve patients' survival, further knowledge about the biological but also about the clinical presentation of gliomas and of glioma patients in an entire population was needed. Identification of patients who would benefit from standard treatment as well as identification of patients who need more aggressive treatment at the time of diagnosis is essential. Equally important is the identification of patients who will not benefit from current standard treatment. Moreover, as common exclusion criteria in clinical trials are age, performance status, and a histologically verified diagnosis, knowledge regarding clinical characteristics in the total population was highly needed. In manuscripts 1 and 2, sampling from national registries was performed and clinical data were collected in order to indentify a clinical prognostic profile for patients with WHO grade I-II tumours (LGG) and WHO grade III-IV tumours (HGG). By using a population-based setup, we identified 433 patients who were diagnosed with a primary glioma in the period 1 January 2005 to 31 December 2009, and of these 76 patients were clinically diagnosed and 357 had a histologically verified diagnosis. We found that younger age, a non-astrocytic histology, having performance status (PS) 0-1, and the absence of neurological deficits were associated with a better prognosis in patients with LGGs. In patients with HGGs younger age, having PS 0-1, absence of neurological deficits, having a tumour that does not cross the midline, and receiving curatively intended post-surgical treatment were associated with a superior prognosis. Older patients also benefitted from curatively intended treatment, although their survival was inferior as compared to younger patients receiving similar treatment. In addition, the prognostic value of having somatic mutation affecting the protein isocitrate dehydrogenase 1 (IDH1) was evaluated. Presence of a mutated IDH1 was associated with a better prognosis in patients with WHO grade II and III tumours, whereas no prognostic potential was identified in the group of GBMs. In manuscript 3, the independent prognostic value of the RNA-binding protein Musashi-1 was evaluated using fluorescence-based automated quantitative image acquisition. The prognostic significance was subsequently investigated in relation to the observed clinical prognostic variables. We found that Musashi-1 was not prognostic in WHO grade II tumours, but in WHO grade III high levels of Musashi-1 were associated with poor survival, although the conclusion is based on very few patients. The opposite effect was identified in a sub-group of postsurgical treated GBM patients expressing high levels of Musashi-1 and a superior prognosis. It may be speculated that Musashi-1 status has a predictive value to the effect of chemo radiotherapy in GBM patients, but the study was not designed to explore a potential predictive potential, and this should be investigated in further material. In manuscript 4, a double staining of CD133 and nestin was performed. The use of fluorescence made it possible to identify expression of CD133 and nestin in the same cell, which has never been done before. However, neither co-localisation nor expression of CD133 or nestin was associated with survival.
Clinical variables associated with better survival were identified for patients with both LGGs and HGGs. All variables are already used in clinical decision making, and they can be used in prognostic counselling of the patients and to guide clinicians regarding the potential benefit from standard treatment in specific patients. Musashi-1 was a predictor of poor survival in WHO grade III tumours, but in patients with GBMs, high levels of Musashi-1 were associated with improved survival. No prognostic value was identified regarding CD133, nestin, or co-localisation of these markers in multivariate analysis adjusted for clinical variables. None of the investigated CSC markers can be used in a clinical setting at the present time. Quantitative automated image acquisition and processing was demonstrated to be a feasible, robust, and reproducible method that will be used in future projects investigating other potential prognostic factors.