Oligoblastic (<20%) Myeloid Neoplasms with KMT2A (MLL) Rearrangement Show Significant Overlap with Acute Myeloid Leukemia (AML) and Should be Regarded As AML

Introduction: The current World Health Organization (WHO) classification considers acute myeloid leukemia (AML) with PML-RARA, RUNX1-RUNX1T1, and CBFB-MYH11 to be AML regardless of blast count. Myeloid neoplasms with KMT2A rearrangement present with high blast percentages and are classified as AML in the vast majority of cases; however, rare myeloid neoplasms with KMT2A rearrangement may present with <20% blasts and these are currently considered to be myelodysplastic syndromes (MDS) or chronic myelomonocytic leukemia (CMML). We questioned whether these oligoblastic KMT2A rearranged neoplasms shared similar biology with AML, and if this genetic finding should be considered to be AML-defining.

Methods: We searched the archives of three large institutions in the United States for untreated adult with oligoblastic (<20%) myeloid neoplasms with KMT2A rearrangement and untreated adult AML patients with KMT2A rearrangement and compared their clinical and laboratory characteristics. KMT2A rearrangement was detected by chromosomal analysis and/or fluorescence in-situ hybridization (FISH) using KMT2A breakapart probe. KMT2A rearrangement partners were assessed based on a conventional karyotype results and/or RNA based fusion assay.

Results: Between 2010 and 2021 we identified 22 patients with MDS or CMML and KMT2A rearrangement and 88 patients with AML and KMT2A rearrangement. The clinical and laboratory information are summarized below. MDS/CMML with KMT2A-rearrangement showed a similar age and sex distribution to AML, but had more patients with preceding history of chemotherapy for unrelated malignancies (68% vs 37%, p=0.012). Aside from a low blast count in PB and BM, patients with MDS/CMML had a lower WBC and were less anemic and thrombocytopenic. MDS/CMML patients also showed a lower percentage of BM cells with KMT2A rearrangement by FISH (58% vs. 92%, p<0.0001), and significantly less likely to have MLLT3 as the partner gene (23% vs. 58%, p=0.0005). With a similar mutation frequency among patients tested (42% versus 34%, p=0.22), MDS/CMML patients had significantly fewer mutations involving the RAS pathway (KRAS, NRAS, FLT3, PTPN11) than AML patients (p=0.0035).

Seventeen of 22 MDS/CMML patients (77%) progressed to AML with a median time of progression being 4 months (range, 1-27 months). Of the 5 patients who did not progress to AML, 3 received allogeneic stem cell transplantation (SCT) and remain in complete remission, while 2 remain AML-free 3 and 16 months after the initial diagnosis.

With a median follow-up of 12 months (range, <1-117), 15 MDS/CMML patients and 58 AML patients died. There was no difference in use of SCT to treat MDS/CMML versus AML (p=0.47). MDS/CMML and AML patients displayed similar overall survival (OS) (p=0.9059, Figure 1). There was also no statistical difference in OS between MDS/CMML and AML when therapy-related (p=0.569) and de novo (p=0.962) neoplasms were analyzed separately. On multivariable analysis, therapy-related disease (HR 1.774 [1.058-2.975], p=0.03), SCT (HR 0.166 [0.000-0.291], p<0.0001), and platelet count (HR 0.992 [0.000-0.998], p=0.013) were independently associated with shorter OS. Neither AML versus MDS/CMML nor blast percentage (whether dichotomized as <5% or >=5% or as a continuous variable) were associated with shorter OS.

Conclusion: Oligoblastic myeloid neoplasms with KMT2A rearrangement tend to progress rapidly to overt AML and demonstrate similar aggressive behavior to AML patients with KMT2A rearrangement, irrespective of therapy relatedness. Patients with KMT2A rearrangement and <20% blasts are currently classified as MDS or CMML, and may be inadequately treated. Our findings suggest that all myeloid neoplasms with KMT2A (MLL) rearrangement should be classified as AML, irrespective of the blast count.

Figure 1

View largeDownload slide

Figure 1

View largeDownload slide

Close modal