EVALUATION OF 6 HISTOLOGICAL FIXATIVES USING IMAGE-ANALYSIS TO MEASURE REACTION-PRODUCT CONCENTRATION

The purpose of this study was two-fold: to determine the effectiveness of manual image analysis (IA) and computer enhanced IA as a quantitat ive tool to measure reaction products from immunohistological reaction s and to determine the effects of different fixatives on the conservat ion of epitopes when immunostaining for prostate specific antigen. Ide ntical slides were evaluated with both computer enhancer and manual IA . Computer enhanced image analysis showed an overall sensitivity of 84 %, was a turn key system, and provided more control for measuring para meters than did manual IA. Manual image analysis was inexpensive, sens itive (77%), and practical to adapt to current lab methodologies. Both IA techniques were more sensitive to small but statistically signific ant differences than either visual or photographic techniques. Image a nalysis should be incorporated into laboratory protocols to support ex isting subjective measurements currently in use. The effect on the con servation of epitopes was evaluated with 6 common histological fixativ es (B5, Bouin, Carnoy, Stat Fix, HistoChoice, and 10% neutral buffered formalin [NBF]) used to fix prostate tissue collected from surgery. F ixation times for tissues were limited to 4 hr except 10% NBF, which w as fixed for 4 hr and from 24-295 hr. Bouin and B5 conserved prostate specific antigen epitope most efficiently as demonstrated photographic ally and by image analysis. A selected patient trial was repeated to d etermine if counterstaining or antigen retrieval would affect the rank ing of the fixatives. Antigen retrieval was effective in recovering an tigenic sites from HistoChoice and Carnoy fixed tissues, but neither a ntigen retrieval or counterstaining altered the final ranking of fixat ives.