Comparative analysis of prognostic indicators for sarcomas of the soft parts and the viscerae

Purpose of this study was to determine which parameters may be best applied to determine the prognosis of soft tissue and visceral sarcomas, the two groups being regarded as biologically different. In a cohort of 184 soft tissue tumors (STT) and 53 gastrointestinal stromal tumors (GIST), the following factors were examined for their diagnostic and prognostic relevance: patient age, sex, tumor location, histological type, tumor size, tumor grade, DNA ploidy status, mitotic count, and immunohistochemical proliferation index. Tumors were graded according to the FNCLCC system, and antibody Ki-S11 (Ki-67) served as a proliferation marker. Median clinical follow-up time was 48 months. In STT, morphological criteria allowed a ready discrimination between benign and malignant lesions, which was only warranted by histopathological grading in GIST. 178 of all 236 tumors were thus classified as malignant. Whilst most parameters yielded significant results in the univariate analysis, age, sex, and histological type were irrelevant. A proliferation index > 20% predicted a poor outcome in soft tissue sarcomas, in contrast to a threshold of 10% for GIST. In both groups, the Cox multivariate analysis selected the proliferation index as the sole independent predictor of overall survival, whereas it was superseded by the tumor grade with respect to metastatic spread. In conclusion, soft tissue and visceral sarcomas appear to behave basically in a similar manner. Both tumor grade and immunohistochemical proliferation index are of major prognostic value. Concerning the growth fraction, however, different cut-off points should be selected for sarcomas of the soft tissues and those of the digestive tract.