Influence of age and subtype in outcome of operable liposarcoma.

Liposarcoma (LPS) is rare tumor deriving from adipocytes. LPS is classified into histological subtypes: well-differentiated (WDLPS), dedifferentiated (DDLPS), myxoid (MLPS) and pleomorphic (PLPS). A tailored approach taking into account the specificity of disease subtype and age at presentation could be helpful in delineating therapeutic management of liposarcoma. In this paper, we report a retrospective series of a single-institution cohort of patients with LPS, undergoing surgery and radiotherapy and/or chemotherapy. The aim of this study is to evaluate whether clinical characteristics, tumor- and treatment-related features affect clinical outcome in patients treated with curative intent for non-metastatic liposarcoma. Data of patients with locally advanced, non-metastatic liposarcoma treated between 1990 and 2015 were retrospectively reviewed. Data about patient, tumor and treatment features were collected. Two patients subgroups were identified according to age (cutoff: age < 65 years or > 65 years). Statistical analysis was performed to assess correlation between the above-cited variables and local recurrence-free survival (DFS-LR), distant metastasis-free survival, overall survival (OS) and disease-specific survival (DSS); moreover, differences in clinical outcome between the two age groups were identified. Data of 186 patients were collected. At diagnosis, 27.4% of patients were 65 years or older. At a median follow-up of 8.6 years (range 0.1-27.3 years), Kaplan-Meier (KM) survival analysis showed that LR, DM, OS and DSS were 75.5%, 76.6%, 48.1% and 72.1%, respectively. KM analysis showed that age > 65, DDLPS and lower limb localization were related to LR (p = 0.001, p = 0.0001 and p = 0.0001, respectively). Association between LR, age and DDLPS persisted both at univariate (p = 0.003 and p = 0.0001, respectively) and multivariate Cox regression (CR) analysis (p = 0.024 and p = 0.002). Age, tumor depth and grading influenced distant recurrence, both at KM (p = 0.023, p = 0.026 and p = 0.016) and univariate CR (p = 0.026, p = 0.042 and p = 0.012). Age and grading were confirmed at multivariate analysis (p = 0.009 and p = 0.017). Patients with WDLPS and wide excision had significantly better OS (p = 0.001 and p = 0.03, respectively), while histological G3 and age > 65 were related with worse OS (p = 0.008 and p = 0.0001, respectively). Age, DDLPS and grade were related to OS at univariate (p = 0.0001, p = 0.0001 and p = 0.03, respectively) and multivariate CR analysis (p = 0.031, p = 0.0001 and p = 0.001, respectively). However, analyzing the specific causes of death, female died less often for tumor-related causes, with a DSS of 91.0% compared to 57.4% of male counterpart (p = 0.005). At Kaplan-Meier analysis, postoperative radiotherapy resulted in a statistically significant better disease-specific survival than postoperative radiotherapy (82.9% vs. 46.2%, p = 0.045). High grade correlated with poorer disease-specific survival (59.3%) than intermediate and low grade (73.4% and 91.6%, respectively) (p = 0.008). Association between DSS, sex and grade persisted both at univariate (p = 0.008 and p = 0.022, respectively) and multivariate Cox regression (CR) analysis (p = 0.014 and p = 0.038). Histotype-driven schedules of treatment should be developed to take into account biological heterogeneity of this disease. Further studies are needed to stratify patients subgroup and develop tailored treatment strategies (i.e., altered fractionations and different chemotherapy regimens in aggressive subtypes), in particular more prospective trials are needed to develop treatment guidelines in elderly STS, taking into account the frailty and the peculiarity of this subgroup.