Refined diagnosis and prognosis in soft tissue sarcoma - genetic profiles, biomarkers and prognostic models

Abstract: This work aimed at evaluate diagnosis and prognosis in soft tissue sarcoma (STS): In study I, 32K bacterial artificial chromosome (BAC) arrays and gene expression profiling were applied to 18 leiomyosarcomas and 31 undifferentiated pleomorphic sarcomas (UPS), with the aim of identifying molecular subtype signatures. Both the gains/losses profiles and the gene expression signatures revealed striking similarities between the two tumour types. Leiomyosaromas and UPS were indistinguishable using unsupervised hierarchical cluster analysis and significance analysis for microarrays, which suggests a shared lineage. In study II, whole-tumour sections from 239 STS were reviewed for size, vascular invasion, necrosis, and peripheral growth pattern. All factors provided independent prognostic information with hazard ratios (HRs) of 2.2-2.6 for development of metastases in multivariate analysis. When combined into a prognostic model, referred to as SING (Size, Invasion, Necrosis, Growth), high-risk tumours were identified with a sensitivity of 74% and a specificity of 85%. SING compared favourably with other currently used prognostic systems. In study III, the prognostic value and clinical applicability of five proliferation markers were assessed: Ki-67, Top2a, p21, p27Kip1 and S-phase fraction in a mixed series of 196 STS of the extremities and the trunk wall, encompassing MFH/UPS, leiomyosarcomas and liposarcomas. High S-phase fraction and high expression of Ki- 67 and Top2a significantly correlated to risk for metastasis with HRs of 1.9-4.4. Classification and regression tree analysis showed that Ki-67, Top2a and S-phase identified different prognostic subgroups. This explorative analysis suggests that proliferation markers could have a role in STS prognostication, in particular when few other factors can be evaluated, as in the preoperative setting. In study IV, ezrin expression was evaluated by immunohistochemistry on tissue microarrays from a mixed series of 256 STS. Positive ezrin expression predicted development of metastasis (HR=1.8) and local recurrence (HR=1.8) and was strongly correlated to necrosis and growth pattern. Ezrin represents therefore a potential marker for identification of high-risk sarcoma patients.