Considering the fact that DAB doesn't follow Lambert-Beer criteria while using it to quantify the amount of antigen present during immunohistochemistry, is it still possible to use it for semi-quantitative analysis, especially in comparative studies ?
There are many scientific reasons NOT to use DAB for quantitative measurement of staining, you cite the Beer law but there are others, including the non-linearity of antibody binding. However, DAB staining IS used all the time in a quantitative way, either because people don't understand its limitations, or simply because its the more practical method to use, since other more linear methods of staining (ie. fluorescence) have their own technical problems. My own opinion is that it's OK to use DAB in a semi-quantitative way (ie. with 1+ to 3+ subjective scoring by a trained observer), as long as the stains evaluated have all been done in the same batch, with automated methods. It is accepted by regulatory agencies for some human diagnostic applications (eg HER2 scoring in mammary tumors). It is more dicey from a scientific standpoint to make conclusions using quantitative methods of staining intensity measurement (ie. image analysis), especially when the results indicate only small differences between slides.
- Badges
- Science topic