The flow cytometric pattern of CD34, CD15 and CD13 expression in acute myeloblastic leukemia is highly characteristic of the presence of PML-RARalpha gene rearrangements

Abstract

BACKGROUND AND OBJECTIVE: Rapid identification of AML patients carrying the t(15;17) translocation for treatment decision-making is currently made on the basis of morphologic screening. However, the existence of both false positives and negatives highlights the need for more objective methods of screening AML cases and further molecular confirmation of the t(15;17) translocation. DESIGN AND METHODS: In the present study we analyzed a total of 111 AML cases in order to investigate whether immunophenotyping based on the assessment of multiple-stainings analyzed at flow cytometry could improve the sensitivity and specificity of morphologic identification of acute promyelocytic leukemia (APL) carrying the t(15;17) translocation. FISH analysis was used as a complementary technique for cases in which morphology and molecular biology yielded discrepant results. RESULTS: Concordant results between morphology and RT-PCR were found in 102/111 (91.8%) cases: 34 patients had M3/PML-RARalpha+ and 68 non-M3/PML-RARalpha- disease. Nine cases showed discrepants results. Multivariate analysis showed that the best combination of immunologic markers for discriminating between M3/PML-RARalpha+ and non-M3/PML-RARalpha- cases was that of the presence of heterogeneous expression of CD13, the existence of a single major blast cell population, and a characteristic CD34/CD15 phenotypic pattern (p<0.02). A score system based on these parameters was designed, and the 34 M3/PML-RARalpha+ cases showed a score of 3 (presence of the 3 phenotypic characteristics). In contrast, only 1 out of the 68 (1.3%) non-M3/PML-RARalpha- cases had this score, most o these latter cases (53/68, 78%) scoring either 0 or 1. Therefore, among these cases, immunophenotyping showed a sensitivity of 100% and a specificity of 99% for predicting PML/RARalpha gene rearrangements. Of the 9 cases in which morphology and molecular biology results were discrepant, four cases displayed M3 morphology without PML/RARalpha rearrangements by RT-PCR. In only one of these 4 cases did the immunophenotype score 3, this being the only FISH positive case. From the remaining five discrepant cases (non-M3 morphology while positive for PML/RARalpha) two cases had a phenotypic score of 3 and were FISH positive while the other three were negative by FISH. Upon repeating RT-PCR studies, two of these latter three cases became negative. INTERPRETATION AND CONCLUSIONS: Our results show that immunophenotyping may be of great value for quick screening of APL with PML/RARalpha rearrangements.