Abstract
The French, American and British cooperative group (FAB) proposed in 1976 a classification of acute leukemia based on morphological and cytochemical criteria. In order to test the applicability and clinical usefulness of the FAB classification, an evaluation was performed on 104 cases of acute leukemias and related disorders treated in our institute. We followed the criteria of the original proposal with a clearer definition of promyelocytes as immature cells with 20 or more azurophilic granules. Those with less than 20 granules were considered to be blasts.
The number of cases of each category as determined by consensus of the 3 observers was 16 L1, 9 L2, 0 L3, 20 M1, 15 M2, 8 M3, 11 M4, 7 M5, 2 M6 and 16 myelodysplastic syndrome (MDS). The rate of complete coincidence of the classification among the 3 observers was 81% in ranged from 56% in L2 to 100% in M3. Eight cases of “ALL” of the FAB classification had been classified as peroxidase negative AML in the former conventional classification, and 4 cases of MDS as erythroleukemia. “ALL” might include in its category peroxidase negative AML and other undefinable forms, and the distinction between L1 and L2 was not always clear even after scoring system was applied. We found that all M2 cases had more than 10% promyelocytes, or more than 25% granulocytes beyond the promyelocytic stage. In the classification of M4 and M5, non-specific esterase staining with α-naphthyl acetate used as substrate was very useful.
The FAB classification was found to be more objective in its criteria than previously employed conventional one, and could be used as a method of standard classification among different institutes. However, it should need further refinement supplemented with electronmicroscopic and immunologic examinations.