Assessing the added value of bone marrow morphologic evaluation beyond flow cytometry for disease detection in the setting of acute myeloid leukemia after chemotherapy

Abstract

Abstract Complete response after acute myeloid leukemia (AML) therapy historically requires a decrease in bone marrow (BM) blasts to <5% by morphologic assessment. However, accumulating data indicate that measurable residual disease by more sensitive multiparameter flow cytometry (MFC) is a better predictor of outcome than traditional morphologic evaluation. In most cases, morphology alone cannot differentiate between regenerative and neoplastic blasts, raising the question of its added value to more specific MFC. The objective of this study was to assess the value morphologic evaluation adds to MFC for the detection of residual AML in the post-chemotherapy (induction or consolidation) setting. A search of our pathology database identified 47 consecutive nontransplant BM evaluations (41 total patients) in the post-chemotherapy setting between January and March 2020. The morphology portion for each case was reviewed by three independent hematopathologists for evaluation of adequacy (>200 cells), blast count, trilineage hematopoiesis, dysplastic features, focal/patchy disease, and any non-hematopoietic or alternative diagnoses that would not have been discovered by the associated MFC study. The MFC data for each case was separately reviewed by two hematopathologists. Six of the 47 cases showed >5% blasts by morphology (range 8%-70%); in all six cases, >5% abnormal blasts were also identified by MFC (range 23%-78%). Of cases with <5% blasts by morphology (defined as negative by morphology), MFC detected >5% abnormal blasts in four cases and <5% abnormal blasts in five cases (range 0.007-14%). One of these additional nine cases by MFC was identified to have focally clustered blasts by morphology (7% blasts by MFC, <5% overall by morphology). Two of the nine additional cases by MFC were deemed to have inadequate aspirates by morphology. In total, four of the 47 morphology cases had inadequate aspirate smears (9%). Of these four cases, three had adequate particle preparations. Eleven cases overall had corresponding core biopsies submitted, of which five were inadequate (45%). Morphologic examination identified seven cases with dysplastic features; abnormal blasts were identified in four of these cases by MFC (all >5% blasts). In the remaining three cases with dysplasia, ancillary molecular studies supported molecular evidence of residual disease in one case (NPM1 mutation and FLT3-ITD positivity by PCR). No unexpected carcinomas or other findings that would not be detected by the associated MFC study were identified. In conclusion, in our study, morphologic evaluation did not definitively identify any additional cases of residual/recurrent AML that were not identified by MFC. Morphologic evaluation detected dysplasia without abnormal blasts in three cases; however, the significance was unclear without supportive molecular studies. Reducing the number of cases for morphologic evaluation could save resources and improve turnaround time if MFC can provide the relevant details to assess response.