COMPARISON OF VISUALLY CONTROLLED AND AUTOMATIC HISTOMORPHOMETRIC EVALUATION OF SOFT-TISSUE

For histomorphometric evaluation of human soft tissue a visually contr olled method for determination and counting of Giemsa stained histolog ical structures and an automatic image analysis after monoclonal stain ing were applied in order to compare the two evaluation systems. Human tissue adjacent to commercially pure titanium fracture fixation plate s was obtained during routine removal of the plates after 1.5 years of implantation. Tissue samples of 20 patients were examined with both t echniques. For monoclonal staining the following antibodies were used: HAM 56 (macrophages), propyl-4-hydroxylase (fibroblasts), CD 20 (B-ly mphocytes), CD 2 (T-lymphocytes), CD 25 (activated T-lymphocytes), CD 4 (T-helper/inducer cells), CD 8 (T-suppressor/cytotoxic cells). For t he visually controlled evaluation, we observed connective tissue cells (fibrocytes, fibroblasts) 709 +/- 19/mm(2) (mean +/- SEM per mm(2)), macrophages 4.0 +/- 0.9/mm(2) and lymphocytes 48.9 +/- 15/mm(2) With i mage analysis we found fibroblasts 223 +/- 23/mm(2), macrophages (HAM 56) 112 +/- 16/mm(2), T-lymphocytes (CD2) 100 +/- 10/mm(2), activated T-lymphocytes (CD 25) 49 +/- 6/mm(2). No B-lymphocytes were observed i n the tissue samples. The results for macrophages (p = 0.0001) and lym phocytes (p = 0.0001) are statistically significantly higher using the image analysis. The antibody staining method in combination with an i mage analysis is to be recommended because of the time saved and the m ore precise identification of cells. Whenever no specific antibodies a re available for the structure to be analysed, as in many animal studi es, the visual method should be applied.