OBJECTIVETo investigate the accuracy and consistency of the detection of BCR-ABL tyrosine kinase domain point mutation among different laboratories.

METHODSEvery one of 6 laboratories prepared 10 cDNA samples from tyrosine kinase inhibitors resistant BCR-ABL (P210 or P190) positive patients'bone marrow or peripheral blood. Each cDNA sample was divided into 6 aliquots and delivered to the laboratories. All 6 laboratories tested BCR-ABL point mutations of 60 samples according to their own protocols. Peking University People's Hospital analyzed the comparison results based on both the reports and sequencing chromatogram from all laboratories.

RESULTSAll laboratories reported the same nucleotide and corresponding amino acid mutations in 37 samples (61.7%）. Of 60 samples, 53 had confirmed mutation types, and a total of 23 types were included; 1 had no mutation; mutation types of 6 samples could not be determined because of the big differences among chromatograms from different laboratories. Low percentages of mutants were significantly related to results inconsistency (P=0.008). Inconsistent result of one sample was caused by the unique chromatogram of the mutant L248V, and one by the non-coverage amplification of PCR product from different laboratories. Amplification was failed in 3 samples. Testing or sequencing mistakes occurred in 7 samples. The differences in the mutant percentages among laboratories were less than 20% in the 80.6% of samples with confirmed results. Low internal control gene copies (ABL<10 000) were significantly related to both failed amplification and big differences among chromatograms from different laboratories (P=0.005 and <0.001, respectively).

CONCLUSIONProblems in the clinical routine detection of BCR-ABL point mutation could be exposed and improvement could be achieved by sample exchange and comparison. Low percentage of mutant is the main reason which causes the discrepancy of BCR-ABL point mutation results among different laboratories.