Combined analyses of clinico-pathological features, genetic alterations and immune infiltration improve understanding of tumor behavior and allow for better prognostication : application to gliomatosis cerebri and uveal melanomas
(2014)
Files
Thesis_NarasimhaiahD.pdf
Open Access - Adobe PDF
- 8.84 MB
Details
- Authors
- Supervisors
- Abstract
- The advances in understanding the genetic basis of cancer have focused largely on cancer cell per se. However, in the last few decades, this interest has extended to include the ‘tumor microenvironment’ (TME). Amongst the spectrum of constituents in the microenvironment, immune component has received much attention. Together, these data have led to identification of novel biomarkers in cancer, from both genetic and immune perspectives. Of course, one should not forget the time honoured clinical and pathological risk factors. An integrated analysis of all these different aspects may lead to improved delineation of prognostic groups and development of cost-effective biomarkers for routine clinical use. With this objective in mind, the present work on gliomatosis cerebri and uveal melanoma was carried out. Gliomatosis cerebri (GC) is an extensively infiltrative glial tumor occurring in adults and children. Its histogenesis had long been controversial and only lately it has been regarded as a glioma subtype. Recently, mutations in isocitrate dehydrogenase genes 1 and 2 (IDH1/2) have been identified as early events in gliomas of mainly adults. In this work, 15 GC (12 adult, 3 pediatric) were tested for IDH1/2 mutations as there was no data available at that point. Only IDH1 mutations were identified in adult GC (40%), but not in pediatric ones. The median overall survival was longer in IDH1-mutant GC than IDH1-wild-type one. These results made the final histogenetic link between GC and diffuse gliomas and emphasized the relevance of IDH mutation testing in GC prognostication (Publication I – Narasimhaiah D, et al. Neuropathology 2012; 32:30-7). Uveal melanomas (UM) are the most common primary, intra-ocular malignant tumors in adults. About 50% of these tumors disseminate, mostly to liver and metastatic disease is a major cause of mortality. Monosomy 3 (M3) is a well-recognized marker of poor prognosis in these tumors. In addition, unlike in most other solid tumors, lymphocytic infiltration is a marker of poor clinical outcome in UM. When this work was started, using single biomarkers, UM were stratified in two groups. The fact that there has been no improvement in their survival over the last three decades indicates a need for better definition of risk-groups. With this aim, we assessed genetic, immune and clinic-pathological features of 89 primary untreated UM. Using these variables, five multivariate Cox models to predict disease-free survival (DFS) were constructed, compared and validated. Amongst them, three performed better (CNA, immune-CNA and combined) and independently delineated presence of three prognostic groups (all, p <0.001). Of these, copy number alteration (CNA) model based on M3, 8q gain, LZTS1 deletion and NBL1 deletion emerged as the strongest one. Another, combining immune and CNA, demonstrated that UM with high degree of immune infiltrate (immune-high) stratified more frequently to high risk-group (73.7%), than to intermediate (21%) and low (5.3%). This suggested a possible reason for the paradoxical association between immune infiltration and poor clinical outcome in UM. Furthermore, the prognostic impact of immune-high varied between risk-groups, with no survival effect in high risk-group, but with a poorer outcome in intermediate risk-group. This indicated that impact of immune on outcome is dependent on associated chromosomal alteration. These findings are relevant for designing clinical trials and application of risk-tailored adjuvant therapies to improve survival. Furthermore, they emphasize the need for innovative strategies to target immune-high UM (Publication II - manuscript in preparation). In the other part of UM work, transcriptomic profile analysis, led to identification of two molecular subgroups based on a gene signature. The latter comprised of immune genes of IFNG/STAT1-IRF1 axis and its presence was correlated with intra-tumoral infiltrate of lymphocytes and macrophages. From prognostic point of view, this signature was associated with poor clinical outcome; which was also validated on two independent datasets. These findings provide an insight into the immune profile of UM; a better understanding of the underlying immunological processes through the elucidated network of genes and pathways and could have implications for designing target-based therapies (Publication III - manuscript in preparation).
- Affiliations
Citations
Narasimhaiah, D. (2014). Combined analyses of clinico-pathological features, genetic alterations and immune infiltration improve understanding of tumor behavior and allow for better prognostication : application to gliomatosis cerebri and uveal melanomas.