Erill N1, Colomer A1, Calvo M3, Vidal A2, Román R1, Verdú M2, Cordon-Cardo C4, Puig X1,2.
1BIOPAT, 2HISTOPAT Laboratoris, 3Statistics Dept. of the Universitat de Barcelona, Spain; and 4Division of Molecular Pathology, Memorial Sloan-Kettering Cancer Center, New York.
Chromosome 18q allelic loss (c18q LOH) seems to have prognostic significance in stage II colorectal carcinoma. In our laboratory c18q LOH was prospectively analyzed in a series of primary colorectal tumors using five markers (D18S55/58/61/64 and 69). Results of 207 PCR-based assays previously analyzed using commercial polyacrylamide pre-cast gels were re-evaluated by two independent observers to determine the degree of concordance of visual valuation. A subset of 61 tumors was retested for the same markers and analyzed by capillary electrophoresis, and results were compared with those obtained before. Interobserver discordance affecting LOH status valuation was 9.18% (CI 5.25-13.11). When visual results were compared with those obtained from electropherograms, discrepancies (worse observer) were detected in 16.39% of determinations (CI 7.10-25.68). Analysis of D18S58 and D18S64 was sufficient to predict c18q status in 95% of cases, and the addition of D18S61 raised it to 100%. Our conclusion based on discrepancies analysis points to the automated assay of D18S58/61 and 64 as the option for c18q LOH determination.