AUTHORS
Alisha Monnette Kimble, PhD, MPH, Robert Reid, MD2, Wei Dai3, Christina Lewis4, Elizabeth Pulte5, Martha Donoghue6, Arup Sinha7, Fatima Rizvi8
AFFILIATIONS
1 Ontada, Boston, Massachusetts, United States
2 Ontada, Boston, Massachusetts, United States
3 Ontada, Boston, Massachusetts, United States
4 Ontada, Boston, Massachusetts, United States
5 U.S. Food and Drug Administration, Silver Spring, Maryland, United States
6 U.S. Food and Drug Administration, Silver Spring, Maryland, United States
7 U.S. Food and Drug Administration, Silver Spring, Maryland, United States
8 U.S. Food and Drug Administration, Silver Spring, Maryland, United States
ABSTRACT
Background: Chronic myelomonocytic leukemia (CMML) is a rare myeloid neoplasm with myelodysplastic (MD) and myeloproliferative (MP) features, no CMML-specific approved therapy, and outcomes historically defined by academic referral cohorts. Real-world data from community oncology practice, where most U.S. patients are treated, remain limited. Using structured electronic health record (EHR) data, this study characterized the clinical characteristics, treatment patterns, and overall survival (OS) of CMML in a large community network.
Methods: In this retrospective cohort study, adults (≥18 years) diagnosed with CMML in the U.S. Oncology Network from 2016 to 2024 (follow-up through 2025) using structured iKnowMed EHR data. Eligible patients were required to have at least one documented vital visit through 2025. Demographics, clinical characteristics, baseline labs, first-line (LOT1) treatment, and OS were assessed and stratified by WHO CMML subtype (MD-CMML, white blood cell [WBC] <13 ×10⁹/L; MP-CMML, WBC ≥13 ×10⁹/L). OS from initial diagnosis and from LOT1 initiation was estimated by Kaplan-Meier methods and summarized descriptively by subtype.
Results: Among 1,909 patients (median age 76 years [inter-quartile range (IQR) 69–82]; 63% male; 73% White), MD-CMML comprised 54% (n=1,039) and MP-CMML 39% (n=739), with subtype undocumented in 7% (n=131). Median follow-up was 17.3 months (IQR 7–33) and differed by subtype (MD-CMML 19.6 vs MP-CMML 16.0 vs undocumented 8.0 months). Baseline anemia (hemoglobin <12 g/dL, 56%), thrombocytopenia (platelets <150 ×10³/µL, 56%), and monocytosis (73%) were common, and MP-CMML had higher median WBC than MD-CMML (22.0 vs 6.4 ×10⁹/L).
With 1,056 deaths (55%), median OS from diagnosis was 29.1 months (95% CI 27.5–31.9); estimated OS was 75% at 12 months. OS differed by subtype: median OS was 37.6 months (95% CI 33.4–42.1) for MD-CMML, 23.2 (20.9–26.6) for MP-CMML, and 11.8 (8.9–15.4) for undocumented subtype. These medians fell within the reported range for CMML (~20–40 months), and were consistent with SEER population-based estimates.
Only 45% (n=865) of patients received systemic LOT1 therapy (MP-CMML 56% vs MD-CMML 40%), initiated a median of 1.1 months (IQR 0.4–5.8) from diagnosis. Consistent with guideline-recommended (NCCN) options, hypomethylating agents (HMAs) predominated (74%), most commonly azacitidine (46%), followed by hydroxyurea (23%, more common in MP-CMML) and decitabine (14%); cytotoxic and targeted agents comprised 23% and 1.9%, respectively. Among treated patients, median OS from LOT1 initiation was 19.0 months (95% CI 16.8–21.8) and again favored MD-CMML (21.2 vs MP-CMML 18.7 vs undocumented 11.0 months). Only 6 patients (0.3%), of the full 1,909, enrolled in a clinical trial during follow-up.
Conclusions: In this large community-based CMML cohort, demographically concordant with population-based registries, survival was poor: median OS was under 2.5 years, confirming that CMML's poor prognosis persists in community practice. WHO subtype was a strong, readily available prognostic discriminator: MP-CMML was inferior to MD-CMML both across the full cohort (median OS 23.2 vs 37.6 months) and among LOT1-treated patients (18.7 vs 21.2 months). Undocumented subtype, marked by the shortest follow-up (8.0 months) and worst survival (11.8 months)—likely reflecting early mortality and incomplete diagnostic workup—identifies incomplete classification as an actionable data-quality gap with prognostic consequences. Treatment aligned with guideline-endorsed options (HMAs and hydroxyurea), yet fewer than half of patients received systemic therapy and survival after initiation remained short (19.0 months), mirroring the limited survival benefit demonstrated for available agents in CMML. As an exploratory, hypothesis-generating structured-data analysis, this study lacked molecular/cytogenetic risk stratification, did not capture reasons for treatment initiation or deferral, and may be subject to misclassification and, for LOT1 initiation survival, immortal time bias. Given the short median follow-up (17 months), longer-term survival estimates (e.g., 5-year) are imprecise and should be considered preliminary. Nevertheless, these findings confirm a substantial unmet need for effective CMML-directed therapies and consistent subtype classification, and provide real-world data from a U.S. community oncology setting to help contextualize outcomes observed in clinical trials of emerging treatments.
DISCLOSURES
Alisha Monnette Kimble, PhD, MPH: McKesson - Current equity holder in publicly-traded company; McKesson - Current holder of stock options in a privately-held company; McKesson - Stock (publicly traded); McKesson - Current Employment