Estimating the Prevalence of Possible Undiagnosed Chronic Myelomonocytic Leukemia in the General Population: An Nhanes Analysis

Background:

Chronic myelomonocytic leukemia (CMML) is a chronic hematologic malignancy characterized by peripheral blood monocytosis and cytopenias, constitutional symptoms, and splenomegaly. It has been reported to have a prevalence of 1-2 cases per 100,000, however, given its chronic nature and often asymptomatic presentation, particularly in low-risk disease, CMML is likely underdiagnosed in the general population. To provide a rough estimation of possible undiagnosed CMML, we examined the National Health and Nutritional Examination Survey (NHANES) database.

Methods:

NHANES is a cross-sectional, nationally representative, population-based survey administered by the Centers for Disease Control and Prevention that utilizes a combination of interviews, physical examinations, and laboratory testing. Information on complete blood count, demographic data, and medical conditions were exported from the NHANES database from 1999-2020. Our definition for possible undiagnosed CMML included an absolute monocyte count (AMC) of greater than 1000/uL, >10% monocytes of the white blood cell (WBC) differential, and, to align with the International Consensus Classification (ICC) 2022 diagnostic criteria, patients were also required to have either anemia with hemoglobin (Hb) <13g/dL for males and <12g/dL for females, neutropenia with absolute neutrophil count <1.8 x 10⁹/L, or thrombocytopenia with platelet count <150 x 10⁹/L. Given the predilection for macrocytic anemia in CMML, patients qualifying based on anemia were excluded if MCV was less than 100 or if they had evidence of B12 or folate deficiency. Patients were also excluded from the possible undiagnosed CMML group if they reported any history of cancer or any infection in the past 1 month based on the patient questionnaires. Additionally, although rheumatologic disease data was not easily accessible from the NHANES data, we excluded patients with a C-reactive protein (CRP) or high sensitivity CRP of > 1 mg/dL to account for a while range of non-oncologic hyperinflammatory conditions. Patients with missing data necessary to determine possible undiagnosed CMML were also excluded. All analyses accounted for the complex sampling design of NHANES by incorporating the NHANES provided design variables and reported statistics represent weighted values.

Results:

Of the 149240 participants available in our analysis, 31125 were included who all had available laboratory data, without recent illness or cancer, and had CRP which was available and <1 mg/dL. The estimated prevalence of possible undiagnosed CMML was 44.8 per 100,000 patients. The prevalence of possible undiagnosed CMML in the age cohorts 18-30, 30-40, 40-50, and 50-60, were 4.3, 8.9, 23.5, and 21.3 per 100,000, respectively. The prevalence increased significantly after age 60 with 28.9 patients per 100,000 for ages 60-70, 40.8 per 100,000 for ages 70-80, and 885.3 per 100,000 for age > 80.

The median age of the possible undiagnosed CMML group was 70.5 years, with 53% male and 81% white. This was significantly older than patients without possible undiagnosed CMML (45.9 years). The mean WBC, Hb, and platelet count in the possible undiagnosed CMML group was 8.0 x 10⁹/L, 13.4 g/dL, and 138.5 x 10⁹/L respectively compared to 7.2 x 10⁹/L, 14.4 g/dL, and 253.9 x10⁹/L in the participants who did not meet possible CMML criteria.

Conclusions:

This study is limited by the inability to extract the full information needed for diagnostic criteria of CMML, including bone marrow biopsy, mutational and cytogenetic data, as well as follow up monocyte counts to ensure persistent monocytosis. Despite these limitations, this NHANES analysis suggests that the prevalence of possible undiagnosed CMML may be higher than commonly appreciated, with an estimated prevalence of 44.8 per 100,000 patients. Accordingly, undiagnosed CMML is more prevalent with older age and in white males. While it is important to acknowledge significant limitations, these data are the first to our knowledge that attempt to estimate the burden of possible CMML underdiagnoses, underscoring the need to expand awareness of the optimal evaluation of monocytosis.