CCUS with dysplasia: A clinically relevant precursor to MDS

Clonal hematopoiesis of indeterminate potential (CHIP) evolves into hematologic cancers at a much slower rate compared to clonal cytopenia of undetermined significance (CCUS). CCUS therefore appears to be a more immediate precursor to myelodysplastic syndromes (MDS) than CHIP. However, there is limited evidence on the indication of CCUS progression and its molecular mechanism. The MDS Hub has previously published two articles on this theme: Immune dysregulation and genetic mutations associated with MDS pathogenesis and Genetic pathogenesis of MDS: Translating molecular advances into better diagnosis and prognosis of CCUS and MDS, both of which focused on gene mutations and their association with CCUS.

Jojosky and colleagues recently published a study in the European Journal of Haematology, examining the clinical, histologic, and molecular features of patients with CCUS to provide an improved understanding of the factors that influence CCUS progression.¹ Here, the MDS Hub summarizes the key findings.

Study design

An observational cohort study was conducted, including patients who underwent bone marrow (BM) biopsies for unexplained cytopenias who were categorized into CCUS–sub-diagnostic dysplasia (CCUS‑D; n = 25), CCUS-no dysplasia (CCUS-ND; n = 24), and MDS groups (n = 49). Patients were classified as CCUS-D and CCUS-ND based on the criteria shown in Table 1. Myeloid next-generation sequencing and flow cytometry on either blood samples and/or BM aspirates were used.

Table 1. Classification criteria*

CCUS-D, clonal cytopenia of undetermined significance sub-diagnostic dysplasia; CCUS-ND clonal cytopenia of undetermined significance with no dysplasia. *Data adapted from Jajosky et al.1
CCUS-D	CCUS-ND
2–10% erythroid precursors	≤2% atypical erythroid cells
>50% erythroid cells with megalo-blastoid changes and/or basophilic stippling	≤5% atypical granulocyte cells
5–10% hypogranular and/or hypolobated granulocytes	≤10% atypical megakaryocytic cells
≥10% megakaryocytes (small, hypolobated, or other nuclear abnormalities)

Results

Blood cell indices and BM composition

• More than half the patients with CCUS (25 out of 49) were classified as CCUS-D and the remaining patients (24 out of 49) were classified as CCUS-ND. There were no significant differences in age, gender, or history of smoking or alcohol abuse between the groups (as shown in Table 2).
• No significant differences were observed in the frequency of red blood cell macrocytosis or the degree of neutropenia, anemia, or thrombocytopenia, indicating no correlation between blood cell indices and low-level BM dysplasia in the CCUS groups. However, BM hypercellularity and erythroid predominance in CCUS was significantly different to the MDS group.

Table 2. Clinical characteristics and BM features of the CCUS-ND, CCUS-D, and MDS groups*

ANC, absolute neutrophil count; BM, bone marrow; CCUS-D, clonal cytopenia of undetermined significance sub-diagnostic dysplasia; CCUS-ND, clonal cytopenia of undetermined significance with no dysplasia; MDS, myelodysplastic syndromes; M:E, myeloid:erythroid. *Data adapted from Jajosky et al.1 †Bold font indicates statistically significant p values. ‡Hypercellularity was defined as bone marrow cellularity of at least 10% above normal.
Characteristic (Data presented as % unless stated otherwise)	CCUS-ND n = 24	CCUS-D n = 25	MDS n = 49	p value†
				CCUS-ND vs CCUS-D	CCUS-ND vs MDS	CCUS-D vs MDS
Average age, years (range)	74.4 (53–87)	76.4 (64–93)	72.9 (55–87)	0.43	0.54	0.11
Sex, males	54	72	74	0.20	0.12	0.90
History of smoking	63	60	60	0.86	0.87	0.97
History of alcohol abuse	8	8	13	0.97	0.58	0.54
ANC <1.8 × 109/L	13	32	43	0.10	0.003	0.37
Platelets <100,000 µL	46	40	51	0.69	0.68	0.38
Hemoglobin <10 g/dL	42	44	71	0.87	0.019	0.028
Hypercellular BM‡	46	64	92	0.21	<0.001	0.013
BM M:E ratio ≤1	13	12	35	0.96	0.023	0.020
Flow cytometry abnormalities	33	60	86	0.06	<0.0001	0.027

Molecular abnormalities

• The molecular abnormalities were more comparable to MDS in CCUS-D patients compared to CCUS-ND patients. There were increased non-TET2/DNMT3A/ASXL1 variants (p = 0.004) and co-mutations of TET2, DNMT3A, or ASXL along with other genes (p = 0.024) in CCUS-D patients (as seen in Table 3).
• The mutational burden was also similar in CCUS-D and MDS patients. Larger clone size and >2 pathogenic variants (p = 0.011) occurred more often in CCUS-D patients.  

Table 3. Comparison of molecular features of the CCUS-D, CCUD-ND, and MDS groups*

CCUS-D, clonal cytopenia of undetermined significance sub-diagnostic dysplasia; CCUS-ND, clonal cytopenia of undetermined significance with no dysplasia; MDS, myelodysplastic syndromes; VAF, variant allele frequency. *Data adapted from Jajosky et al.1 †Bold font indicates statistically significant p values.
Molecular feature (Data presented as % unless stated otherwise)	CCUS-ND n = 24	CCUS-D n = 25	MDS n = 49	p value
				CCUS-ND vs CCUS-D	CCUS-ND vs MDS	CCUS-D vs MDS
Non-TET2/DNMT3A/ASXL1 variants	21	60	86	0.004	<0.0001	0.027
Co-mutation of TET2/DNMT3A/ASXL1 with ≥1 other gene(s)	4	28	33	0.024	<0.001	0.69
Spliceosome (SF3B1, SRSF2, ZRSR2, or U2AF1) variants	13	48	51	0.006	<0.001	0.81
IDH1, IDH2, RUNX1, or CBL variants	0	24	16	0.011	0.004	0.46
TP53 variant	8	4	27	0.54	0.038	0.004
>2 variants	0	24	29	0.011	<0.0001	0.68
≥1 variant with VAF	33	68	78	0.015	<0.001	0.40
Average number variants per patient, n (range)	1.3 (1–2)	1.8 (1–4)	1.9 (0–5)	0.08	0.007	0.70
Average VAF per variant (range)	14 (2–47)	31 (2–57)	37 (2–97)	<0.0001	<0.0001	0.06

Smoking and molecular alterations

Smoking was associated with non-TET2/DNMT3A/ASXL1 variants (p = 0.020), RUNX1 variant (p = 0.043), and co-mutation of TET2/DNMT3A/ASXL1 along with other genes (p = 0.013) compared to non-smokers within the total CCUS cohort. However, the overall mutational burden did not differ between CCUS smokers versus non-smokers. No associations were noted for smokers versus non-smokers in the MDS patient group.

Conclusion

The study demonstrates that CCUS with dysplasia bears a closer resemblance to MDS than CCUS-ND. Clonal expansion and acquisition of further mutations may result in a linear progression from CHIP to CCUS-ND then to CCUS-D, which may be a precursor to MDS. Therefore, distinguishing between CCUS-ND and CCUS-D maybe a useful measure to monitor disease progression and clinical outcomes. A long-term follow-up in larger cohort of patients is warranted to determine whether patients with CCUS-D are at increased risk of disease progression.