Clonal Cytopenia of Uncertain Significance or Myelodysplastic Syndrome?

Clinical history

51-year-old man with progressive fatigue and exertional dyspnea.

Complete blood count and differential (reference range)

Bone marrow morphology

Flow cytometry

Mildly decreased side scatter of granulocytes; myeloid progenitors with unremarkable phenotype.

Karyotype

46,XY[20]

Next-generation sequencing

Next-generation sequencing studies showed the following mutations:

Final diagnosis

Variably cellular (40–60 percent) bone marrow showing trilineage hematopoiesis with left-shifted erythropoiesis and granulopoiesis, mild dysmegakaryopoiesis, mild eosinophilia, and mild polytypic plasmacytosis.

Case Discussion

The pathologist’s view
When evaluation of a patient with cytopenias demonstrates ineffective hematopoiesis, the pathologist’s challenge is to differentiate a clonal myeloid malignancy that meets criteria for myelodysplastic syndrome (MDS) from other causes. MDS encompasses a heterogeneous group of clonal hematopoietic stem cell disorders, but morphologic dysplasia in one or more of the myeloid lineages is a near-constant phenotypic manifestation. Therefore, despite major advances in the understanding of the genetic changes associated with MDS, morphologic examination of hematopoietic precursors continues to be a cornerstone in its evaluation. The World Health Organization (WHO) classification (1) continuesto require the presence of quantifiable morphologic dysplasia in >10 percent of any of the myeloid lineages, with or without genetic evidence of clonal disease, to unequivocally establish a diagnosis of MDS. The current WHO classification scheme recognizes certain recurrent chromosomal alterations as presumptive evidence of MDS in the absence of quantifiable morphologic dysplasia; however, they rightfully argue against using somatic gene mutations as diagnostic evidence of MDS in the absence of unequivocal morphologic dysplasia, given the occurrence of these changes in healthy older individuals as age-related clonal hematopoiesis (1,2). Herein lies the conundrum of clonal cytopenia(s) of undetermined significance (CCUS) (3), which often demonstrates only mild dysplasia not meeting the quantitative cutoff for MDS. Any experienced pathologist can attest to the significant interobserver variability of discerning morphologic dysplasia, particularly in cases of low-risk MDS. In this example, subtle morphologic changes – including bone marrow hypercellularity and mild megakaryocytic dysplasia – are not sufficient to warrant a “WHO-sanctioned” MDS diagnosis and myeloid precursors exhibit an essentially normal immunophenotype by flow cytometry, further limiting our ability to establish an unequivocal diagnosis of MDS. Nevertheless, given the presence of multiple gene mutations (some with relatively high allelic frequency), macrocytic anemia, and subtle morphologic changes, it is conceivable to predict that this patient’s disease will evolve into MDS with time. Our task is to alert the oncologist and recommend close observation.

The hematologist’s view

Hematologists are frequently asked to assess patients with persistent cytopenias. Often, despite careful evaluation including marrow aspiration and biopsy, diagnosis remains unclear. Historically, such patients were described as having “idiopathic cytopenia(s) of undetermined significance” (ICUS) (4) – “idiopathic” highlighting unclear etiology, and “undetermined significance” underscoring a variable and unpredictable natural history.

The recognition that most patients with MDS and other myeloid neoplasms have somatic genetic mutations in hematopoietic cells that can be detected with next-generation sequencing (NGS) promised to improve diagnostic testing and offer some clarity in ambiguous cases (5). Detection of SF3B1 mutation, for instance, defines a subtype of MDS with a strong association with ring sideroblast morphology and usually indolent natural history, and can help rule out an exclusively reactive cause for sideroblastic anemia (6).

At the same time, we have learned that clonal hematopoiesis associated with somatic mutations in hematopoietic cells is common with aging, and that the most frequent mutations seen in aging-associated clonal hematopoiesis (e.g., DNMT3A, TET2, ASXL1, TP53, JAK2) are also highly recurrent in myeloid neoplasms (3). Therefore, simply finding an MDS-associated mutation in a patient with cytopenias and indeterminate morphology is not enough to classify the patient as having MDS. When a clonal mutation is present in a patient with cytopenias, it is referred to as “clonal cytopenia(s) of undetermined significance” (CCUS) (7). 

Malcovati and colleagues showed a few years ago that patients with CCUS are much more likely than those with ICUS to progress to MDS or another neoplasm meeting World Health Organization (WHO) diagnostic criteria (8). Patients with multiple mutations and “larger” clones – variant allele frequency (VAF) >10-20 percent in blood-derived DNA – were more likely to progress than those with single mutations and small clones. Similarly, Abelson and colleagues found that a broad red cell distribution width, specific mutations such as splicing factor mutations, and higher VAF indicating greater clonal expansion predicted acute myeloid leukemia (AML) in healthy people with clonal hematopoiesis (9). Effectively, higher-risk forms of CCUS have a natural history comparable to lower-risk MDS and can be thought of as “MDS without dysplasia.” The WHO allows certain clones detected by conventional metaphase karyotyping (e.g., del(5q) or monosomy 7) to support a diagnosis of MDS in the absence of cellular dysmorphology.

This information about the relative risk of disease evolution in people with ICUS versus CCUS can be helpful in counseling patients, but identifying which allelic patterns actually cause cytopenias can be a challenge. The patient described here, who presented with macrocytic anemia without other cytopenias, exhibits some worrisome features. Five clonal mutations were detected, two at VAF >10 percent. SRSF2, ASXL1, and SETBP1 are all enriched in MDS/myeloproliferative neoplasm (MPN) overlap syndromes compared with MDS without MPN features (10,11), so although splenomegaly is not mentioned and there is no leukocytosis or thrombocytosis, his condition may, with time, evolve in that direction.

The patient’s marrow is also not “stone-cold normal.” It is hypercellular for age and there are a few dysplastic cells on the aspirate film. We regularly observe a few atypical or dysplastic cells in the marrow of healthy older people (12) but, at age 51, small hypolobated megakaryocytes are unusual. In some studies, patients with cytopenias whose marrow had some dysplasia, but not enough to meet WHO criteria for MDS, were more likely to progress to incontrovertible MDS than those with normal marrow.

Might some morphologists have called this case MDS rather than CCUS? Inter- and intra-observer reproducibility is not as high as we would like in MDS diagnosis, especially in lower-risk MDS where morphologic changes can be subtle (13,14). In cooperative group trials, historically at least one in five enrolled patients have not had the diagnosis confirmed on central review, either because dysplasia was subtle or the marrow sample provided to the reviewing hematopathologist was inadequate.

I suspect this man with CCUS will eventually develop an overt myeloid neoplasm, either MDS or MDS/MPN. For now, monitoring with serial blood counts and supportive care (possibly with an erythropoiesis-stimulating agent) would be appropriate. But he bears close watching and will likely need disease-modifying therapy and allogeneic hematopoietic cell transplant in the future.

Read on to discover our other myeloid cases...

Case 2: SF3B1-Mutant Chronic Myelomonocytic Leukemia
Case 3: Myelodysplastic Syndrome with Excess Blasts and Fibrosis
Case 4: Elevating the Treatment Standard in Older Patients with AML