Prognostic factors in soft tissue sarcoma. Tissue microarray for immunostaining, the importance of whole-tumor sections and time-dependence

University dissertation from Oncology, Lund

Abstract: In adult soft tissue sarcoma (STS) of the extremities and trunk wall, improved prognostic factors are needed to identify patients at high-risk for metastasis. Various factors are included in the many prognostic systems currently in use and the prognostic value of immunohistochemical (IHC) expression of biological markers is unclear. The tissue-preserving, high throughput tissue microarray (TMA) technique for analysis of immunohistochemical expression of biological markers was validated for Ki-67, and was found to yield results comparable to conventional staining methods. TMA was used to study the IHC expression of multiple markers (Ki-67, p53, cyclin A, bcl-2, ß-catenin, CD44, and Pgp) in 218 malignant fi brous histiocytomas (MFH) and in 140 mixed STS. In the MFH series, tumor size and Ki-67, as the only IHC marker, provided independent prognostic information. In the mixed STS series whole-tumor sections were used and TMA was performed in the peripheral tumor growth zone. Whole-tumor sections facilitated assessment of the strong independent prognostic factors for metastasis vascular invasion, hazard ratio (HR) 3.5, tumor necrosis (HR 2.8), and tumor growth pattern (HR 3.2), and the latter also correlated with local recurrence (LR). In comparison, histological malignancy grade, tumor size, and depth were not of independent prognostic value. When TMA was performed from the peripheral tumor growth zone, the IHC expression of Ki-67 (HR 1.9), ß-catenin (HR 2.7), CD44 (HR 2.1) and Pgp (HR 2.4) were independent prognostic factors. Finally, prognostic factors were found to be time-dependent, and most factors lost their prognostic value after 2 years, whereas LR was a strong prognostic factor for metastasis whenever it occurred.

  This dissertation MIGHT be available in PDF-format. Check this page to see if it is available for download.