Evaluation of various methods to quantify endothelial cells attached to vascular prostheses: comparison with a new "gold standard" FACS method.

For in vitro evaluation of functional properties of endothelial cells seeded on synthetic vascular prostheses accurate and reproducible quantification of cells is mandatory. Comparison of these properties with those resulting from other studies requires correlation of the functional parameters to reliably counted cell numbers. The accuracy of methods of quantification currently being used is unknown due to the lack of a "gold standard" method to which these methods can be compared. To determine the accuracy and reproducibility of four widely used methods, we have developed a "gold standard" model, using a flow cytometer (FACS). Endothelial cells, attached to collagen-coated Dacron vascular prostheses, were counted by four conventional methods and a new method of quantification after the attached number of cells had been determined with 99% accuracy by FACS. Subsequently, ratios were computed by dividing the cell numbers determined by the methods under investigation by those determined by FACS (x100%). The four conventional methods investigated were (1) removal and subsequent counting of cells from substrata by trypsin (T), (2) digestion of cells by citric acid and counting of crystal violet-stained cell nuclei (CV), (3) light microscopy after hematoxylin staining (LM), and (4) scanning electron microscopy (SEM). The new method consists of the measurement of cell fluorescence after labeling with fluorescein-diacetate (FDA). T and CV had average accuracy ratios of 127 +/- 58% and 96 +/- 48%, respectively (+/- standard deviation). The ratios for LM and SEM were 116 +/- 101% and 44 +/- 10% (respectively). FDA had a ratio of 99 +/- 7%. Reproducibility of cell quantification by T and CV was significantly less than that of quantification by LM, SEM, and FDA, as expressed by data on inter- and intraobserver agreement. Our results indicate that the investigated conventional methods of quantification failed to meet criteria of both high accuracy and reproducibility. Light microscopy and scanning microscopy methods were inaccurate but yielded reproducible countings. We conclude that the FACS method can serve as a "gold standard" to compare the accuracy and reproducibility of cell quantification methods. Moreover, the FDA method results in both accurate and reproducible quantification of endothelial cells attached to vascular prosthetic material.