6th ESACP Congress, Heidelberg, April 7-11, 1999

A141
PLOIDYVAR - INVESTIGATING VARIABILITY IN PLOIDY PARAMETERS
Tucker J *, Haroske G +, Giroud F ++, Sowter C **

* Pathology Dept., Edinburgh University, ** Dept. Pathology, St. Bartholomew's Hospital, London, UK , + Inst. Pathology, University of Technology, Dresden, Germany, ++ Inst. Albert Bonniot, Université Joseph Fourier, Grenoble, France

Measurement of total nuclear DNA content ("ploidy") is a widely-used quantitative pathology procedure. However one of the major problems in the field has been the wide variation in results reported in the literature. A multi- laboratory experiment was therefore carried out to study the effects of a) cell selection, and b) sample size on ploidy parameters. Seven Feulgen-stained breast imprint specimens were circulated between three different laboratories whose personnel are experienced in ploidy measurement. From each Laboratory, two sets of ploidy measurements were obtained for each specimen: a) from their own selection of cells, made according to their usual measurement protocols, and b) from a "central" selection of cells, which are relocated using the AxioHOME computerised microscope or other similar equipment. All results were processed using the EUROQUANT remote ploidy analysis system via the world wide web. The whole experiment was managed by means of a web site which distributed aims, experimental protocols, current status, data and results. Although broadly similar histogram patterns were obtained by all laboratories for all specimens, numerical results for two major parameters (DNA Index and S-Phase fraction) showed substantial differences. Also very large variations were occasionally found. The major source of variation was found to be the selection of internal reference cells; the selection of tumour cells had a relatively minor effect. DNA Index figures usually stabilise after relatively small samples, but large sample sizes are required for accurate S-phase Fraction measurements.