P496: phenotypic subtypes of leukemic transformation in chronic myelomonocytic leukemia

Guillermo Montalban‐Bravo,Rashmi Kanagal‐Shamanna,Ziyi Li,Danielle Hammond,Kelly S. Chien,Juan José Rodríguez‐Sevilla,Koji Sasaki,Elias Jabbour,Courtney D. DiNardo,Koichi Takahashi,Nicholas J. Short,Ghayas C. Issa,Tapan M. Kadia,Naveen Pemmaraju,Farhad Ravandi,Naval Daver,Gautam Borthakur,Sanam Loghavi,Sherry Pierce,Claudio Rodríguez Ramos,Guillermo Garcia‐Manero
DOI: https://doi.org/10.1097/01.hs9.0000968892.25217.59
2023-01-01
HemaSphere
Abstract:Background: Chronic myelomonocytic leukemia (CMML) is a hematologic disorder with high risk of transformation to acute myeloid leukemia (AML). The clinical outcomes of patients with CMML whose disease progresses to AML remains dismal. Identification of disease subtypes of AML after CMML is necessary to inform development of future therapies. Aims: We sought to analyze the distinct patterns of CMML transformation and dissect their genomic determinants, clonal architecture, and clinical outcomes with distinct therapeutic modalities to identify disease subsets with distinct clinicopathological features. Methods: We evaluated all consecutive patients (pts) with AML evolving from antecedent CMML treated at our institution from 2001 to 2022. Based on blast morphology and immunophenotype cases were classified as monocytic (≥20% monoblasts and/or promonocytes or ≥20% CD14+/CD64+ blasts with otherwise non-monocytic morphology), myeloid (≥20% blasts not meeting criteria for monocytic) or erythroid (≥80% immature erythroid precursors positive for CD71, E-cadherin and/or CD235a). The Kaplan-Meier method was used to estimate the median overall (OS) and event-free (EFS) survival. Cox proportional hazards regression was used to study association of clinical or treatment variables with survival. Results: A total of 189 pts were identified: 101 (53%) with monocytic (Mo-AML), 85 (45%) with myeloid (My-AML) and 3 (2%) with erythroid leukemia (Ery-AML). Median age was 71 years (range 28-90) with male predominance (66%). Compared to My-AML, pts with Mo-AML had higher median white blood cell count (23.5 vs 6.0 x109/L, p<0.001), peripheral blood (31% vs 12%, p<0.001) and bone marrow monocyte % (12% vs 6%, p<0.001). Mo-AML was associated with higher frequency of CD56 (51% vs 21%, p<0.001), and lower CD34 (30% vs 89%, p<0.001), CD38 (83% vs 97%, p=0.006) and CD117 (35% vs 87%, p<0.001) blast positivity. Pts with Mo-AML were more likely to have CBL (37% vs 7%, p=0.005), SRSF2 (57% vs 27%, p=0.018) and RAS pathway (BRAF, CBL, KRAS, NRAS, NF1, PTPN11) mutations (85% vs 57%, p=0.006) while CEBPA mutations were more common among My-AML (23% vs 3%, p=0.023) (Fig 1A-B). Median VAF of TET2 mutations was higher among Mo-AML (45% vs 39%, p=0.033). Pts with Mo-AML were defined by concurrent TET2, SRSF2 and RAS pathway mutations (43% vs 13%, p=0.01), compared to My-AML, and more frequently had RAS pathway mutations as dominant clones (56% vs 27%, p=0.013). Pts with Mo-AML were more likely to evolve from SRSF2 (58% vs 27%, p=0.025) and TET2 (65% vs 40%, p=0.038) mutant CMML, particularly with RAS pathway or ASXL1 mutations (Fig 1C). Cytogenetic evolution was observed in 42/107 (39%) pts, with new previously undetectable somatic mutations or expansion of pre-existing clones being observed in 28/42 (82%) and 14/42 (41%) pts, respectively. Pts with Mo-AML acquired higher number of balanced/unbalanced translocations and clonal evolution of RAS signaling mutations (81% vs 33%, p=0.005) (Fig 1D). With a median follow up of 49.5 months, the median OS and EFS was 5.9 and 2.6 months, respectively. Trends to improved EFS with venetoclax and cladribine/clofarabine containing regimens were observed in My-AML and Mo-AML, respectively (Fig 1E-F). Summary/Conclusion: Our data support that disease transformation in CMML can occur through three distinct phenotypic trajectories characterized by specific genomic profiles and clonal evolution. This sets the basis for potential future studies to develop phenotype-specific targeted novel therapeutics.Keywords: Genomics, CMML, Acute monoblastic leukemia, Acute myeloid leukemia
What problem does this paper attempt to address?