Should flow cytometry be required for CMML diagnosis? Position of the ELN iMDS Flow Working Group

The most recent revision of the World Health Organization (WHO) classification, ¹ as well as the International Consensus Classification of Myeloid Neoplasms (ICC), ² has lowered the historical threshold for monocytosis required for the diagnosis of chronic myelomonocytic leukemia (CMML) from 1 × 10⁹/L to 0.5 × 10⁹/L. This reconsideration followed the 2017 description of a new entity, termed oligomonocytic CMML (OM‐CMML), defined by a peripheral blood monocyte percentage above 10% and an absolute monocyte count (AMC) between 0.5 and 1 × 10⁹/L, and associated progression to overt CMML in 38%–47% of cases. ³ , ⁴ In parallel, a relative accumulation of classical monocytes (cMo) exceeding 94% of total circulating monocytes, identified more than a decade ago as a phenotypic hallmark of CMML, has since been incorporated into the updated WHO classification. ⁵ , ⁶ , ⁷ , ⁸ , ⁹ , ¹⁰ , ¹¹ Briefly, the three main circulating monocyte subsets are defined by their differential expression of CD14 and CD16 using flow cytometry (Figure 1): cMo, characterized by a CD14⁺⁺CD16⁻ phenotype, represent the majority of circulating monocytes (approximately 85%); they are distinguished from intermediate monocytes (iMo, 5%), which express CD16, thereby exhibiting a CD14⁺⁺CD16⁺ phenotype, and from non‐classical monocytes (ncMo, 10%), characterized by low or absent CD14 expression and CD16 expression (CD14^low/negative CD16⁺). In the specific setting of AMC between 0.5 and 1 × 10⁹/L, only dysplasia and acquired clonal cytogenetic or molecular abnormality are required for CMML diagnosis according to the WHO classification. ¹ In this situation, abnormal partitioning of circulating monocyte subsets is not required for CMML diagnosis. However, as we will outline in this article, there is now evidence showing that identification of patients at risk of developing overt CMML may significantly improve if immunophenotypic criteria are included also in the context of AMC between 0.5 and 1 × 10⁹/L; and we therefore recommend that future WHO classification incorporate phenotypic features also for this group of patients. This article represents an expert consensus statement rather than a formal evidence‐based guideline. Accordingly, no prespecified Delphi process or systematic review methodology was used. Instead, consensus was reached through structured expert discussion, review of the relevant literature, and iterative revision of the manuscript until all authors agreed on the final recommendations.

Figure 1.

Example of three immunophenotypic profiles showing the distribution of the three main circulating monocyte subpopulations (classical monocytes [cMo], intermediate monocytes [iMo], non‐classical monocytes [ncMo], and left CD14/CD16 histograms) and the slan⁺ncMo subpopulation (right CD16/slan panels) obtained. (A) In a subject with reactive monocytosis (cMo: 75.2%; slan⁺ncMo: 2.4%; analysis performed on a BD FACSCanto™ II cytometer, BD Biosciences). (B) In a patient with chronic myelomonocytic leukemia (CMML) (cMo: 98.3%; slan⁺ncMo: 0.4%; analysis performed on a BD FACSCanto™ II cytometer, BD Biosciences). (C) In a patient with CMML associated with an inflammatory state (“bulbous” profile with cMo: 90.2%; slan⁺ncMo: 0.5%; analysis performed on a Navios™ cytometer, Beckman Coulter). The percentage of cMo and slan⁺ncMo is highlighted when values are pathological.

THE HISTORICAL THRESHOLD OF 1G/L WAS LIKELY OVERESTIMATED

Defined in 1976 by the French–American–British Cooperative Group (FAB), the AMC threshold of 1 × 10⁹/L used to define monocytosis and underpin CMML diagnosis for nearly five decades was likely overestimated. In a cohort of 192 healthy adults (median age 51 years; range 19–99 years), the median AMC was 0.49 × 10⁹/L (interquartile range: 0.41–0.59 × 10⁹/L). ⁵ Age‐adjusted mathematical modeling demonstrated a gradual increase in AMC with advancing age, reaching approximately 0.6 × 10⁹/L at 100 years. These findings suggest that the historical FAB threshold for CMML diagnosis may have been set too high. In the revised proposal, monocytosis is defined using both absolute (>0.5 G/L) and relative (>10% of total white blood cell count) criteria. This combined approach provides a more robust framework for interpretation, thereby mitigating the potential risk associated with lowering the absolute threshold, as the proportion of monocytes within the overall leukocyte population is assessed concurrently.

ONE‐THIRD OF MDS PATIENTS EXHIBITED A CMML‐LIKE PHENOTYPIC SIGNATURE, IRRESPECTIVE OF THE AMC

During the same period in 2017, two independent studies reported that 28%–35% of myelodysplastic neoplasm (MDS) patients display a CMML‐like phenotype with a relative increase in cMo, regardless of their AMC. ⁶ , ¹² Half of these patients were likely to progress to overt CMML within 1 year. ¹² , ¹³ These findings were confirmed in a third study, in which CMML‐like MDS patients exhibited granulocyte‐macrophage colony‐stimulating factor (GM‐CSF) hypersensitivity similar to that observed in CMML patients at low GM‐CSF concentrations. ¹³ The number of mutations per sample in CMML‐like MDS patients (3.5 ± 0.2) was significantly higher than in classical MDS patients (1.7 ± 0.3; P ≤ 0.0001) and similar to that observed in CMML patients (3.5 ± 0.3). TET2 was the most frequently mutated gene across the three groups, followed by SRSF2 and ASXL1. ¹³ Notably, these three genes are also the most commonly mutated in CMML (TET2: 29%–61%; SRSF2: 29%–52%; and ASXL1: 32%–44%). ¹ Moreover, multiple TET2 mutations were significantly more frequent in CMML and CMML‐like MDS compared to classical MDS. MDS patients harboring multiple TET2 mutations more frequently evolved to genuine CMML according to the 2017 WHO classification. Similarly, OM‐CMML patients were found to carry TET2 multiple mutations more frequently and exhibit relative accumulation of cMo. ¹⁴ , ¹⁵

ARE OM‐CMML AND CMML‐LIKE MDS PRE‐CMML MDS ENTITIES OR GENUINE CMML?

A recent comprehensive study comparing 356 newly classified CMML cases (representing 17% of patients formerly considered as having OM‐CMML) with 3311 established CMML cases supports the notion that OM‐CMML represents an independent entity, characterized by different cytogenetic, molecular, immunophenotypic, and transcriptional landscapes. ¹⁶ The abnormal partitioning of circulating monocyte subsets was not included as a classification criterion in that analysis. Discrepancies with previous studies were observed: newly classified CMML cases showed higher frequencies of mutations in DNMT3A and SF3B1, whereas mutations in ASXL1, RUNX1, SRSF2, and TET2 were less frequent compared with established CMML cases. Notably, TET2 mutations were significantly less common in newly classified CMML cases (30% vs. 47.4% in established CMML; P < 0.001), in contrast to findings reported in previous studies. ¹⁴ , ¹⁵

However, comparable overall survival outcomes between OM‐CMML and myelodysplastic CMML were reported in the Munich Leukemia Laboratory (MLL) comprehensive study. ¹⁶ In another recent study including 104 patients with long‐term follow‐up, OM‐CMML was, conversely, associated with a better prognosis, including a significantly superior 5‐year overall survival and a lower 5‐year cumulative incidence of progression to AML compared with CMML. ⁴

A remaining question was whether these two entities, OM‐CMML, defined by quantitative criteria, and CMML‐like MDS, defined by phenotypic criteria, overlap. Applying both criteria to 52 MDS patients showed that although 34/52 fulfilled the OM‐CMML criterion and 31/52 met those for CMML‐like MDS, only 14 patients achieved both. ¹⁷ The positive predictive values (PPVs) for OM‐CMML and CMML‐like MDS defining criteria were found at 44.1% and 45.2%, respectively, similar to those described in the original studies. ³ , ¹² The combination of both quantitative and phenotypic criteria increased PPV to 71.4%. Therefore, combining these criteria may significantly improve the early identification of patients at risk of developing overt CMML.

FLOW CYTOMETRY SHOULD BE INCORPORATED INTO THE DEFINITION OF PRE‐CMML SUBTYPE

The ELN iMDS Flow Working Group has validated the monocyte partitioning for CMML diagnosis in a multicenter study. ¹⁸ The updated WHO classification states that an increase in cMo above 94% supports the diagnosis of CMML in the absence of known active autoimmune diseases and/or systemic inflammatory syndromes.^{1,Footnote#eTable6} The dedicated chapter to CMML in the published Blue Book highlights the potential interest of the slan marker in this context. In such cases, patients exhibit a distinctive phenotypic “bulbous” profile, driven by increased iMo, resulting in a relative decrease in cMo below the 94% diagnostic threshold (Figure 1). Indeed, in a cohort of 55 CMML patients, including seven with inflammatory features and a bulbous profile of cMo in CD14/CD6 expression pattern, the proportion of slan⁺ncMo was markedly reduced (0.4% ± 0.5%) compared with healthy controls (5.7 ± 3.3%) and patients with reactive monocytosis (4.3 ± 3.0%). Even CMML patients with cMo < 94% and an associated inflammatory state maintained very low slan⁺ncMo levels (0.5% ± 0.4%). Using a 1.7% slan⁺ncMo cutoff increased the sensitivity of flow cytometry diagnosis to 100% (vs. 86%) while maintaining 94% specificity. ¹⁹ In addition, an ongoing multicenter study within our group is evaluating the diagnostic value of the slan marker for CMML, particularly in the context of an inflammatory condition.

The ELN iMDS Flow Working Group recommends that future WHO classification incorporate previously proposed defining criteria (Figure 2), including both quantitative and phenotypic features with slan marker assessment in case of inflammatory context, to more robustly delineate this pre‐CMML condition, which may be distinguished from overt CMML.

Figure 2.

Revised World Health Organization (WHO) criteria for chronic myelomonocytic leukemia (CMML) diagnosis emphasizing the role of flow cytometry. The 2022 WHO criteria, including both essential and desirable features, are presented in a simplified format. In case of monocytosis with an absolute monocyte count (AMC) above 1 G/L, all essential criteria must be fulfilled, along with at least one desirable criterion, in accordance with the 2022 WHO classification. We proposed here that, in cases of absolute monocyte count (AMC) between 0.5 and 1 G/L, all three desirable criteria including relative accumulation of classical monocytes ≥ 94%, should be required to establish the diagnosis. BM, bone marrow; CML, chronic myeloid leukemia; MPN, myeloproliferative neoplasms; PB, peripheral blood.

AUTHOR CONTRIBUTIONS

Orianne Wagner‐Ballon: Conceptualization; writing—review and editing; writing—original draft. Leonie Saft: Writing—review and editing. Ulrika Johansson: Writing—review and editing. Wolfgang Kern: Writing—review and editing; conceptualization. Arjan A. van De Loosdrecht: Conceptualization; writing—review and editing.

CONFLICT OF INTEREST STATEMENT

O.W.‐B. has a patent relevant to this work. L.S. has no conflict to disclose. U.J. has no conflict to disclose. W.K. is declaring part‐ownership of MLL Munich Leukemia Laboratory. A.A.v.D.L. has no conflict to disclose.

FUNDING

This research received no funding.

DATA AVAILABILITY STATEMENT

Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.