AUTHORS
Tariq Kewan, MD, Antoine N. Saliba, M.D.2, Kristen McCullough, PharmD3, Aasiya Matin, MBBS4, Mehrdad Hefazi Torghabeh, MD5, Abhishek Avinash Mangaonkar, MBBS6, Naseema Gangat, MBBS7, William Hogan, M.B., B.Ch.8, Rong He, MD9, Dragan Jevremovic, MD PhD10, Patricia Greipp, DO11, Cecilia Arana Yi, M.D.12, Yael Kusne, MD, PhD13, Talha Badar, MD14, James Foran, FRCPC15, Mark R. Litzow, MD16, Kebede Begna, M.D.17, Mithun Shah, MD, PhD18, Hassan B Alkhateeb, M.D.19, Mrinal M. Patnaik, MD, MBBS20, Aref Al-Kali, M.D.21
AFFILIATIONS
1 Mayo Clinic, Rochester, Minnesota, United States
2 Mayo Clinic, Rochester, Minnesota, United States
3 Mayo Clinic, Rochester, Minnesota, United States
4 Mayo Clinic, Rochester, Minnesota, United States
5 Mayo Clinic, Rochester, Minnesota, United States
6 Mayo Clinic, Rochester, Minnesota, United States
7 Mayo Clinic, Rochester, Minnesota, United States
8 Mayo Clinic, Rochester, Minnesota, United States
9 Mayo Clinic, Rochester, Minnesota, United States
10 Mayo Clinic, Rochester, Minnesota, United States
11 Mayo Clinic, Rochester, Minnesota, United States
12 Mayo Clinic, Phoneix, Arizona, United States
13 Mayo Clinic, Phoneix, Arizona, United States
14 Mayo Clinic, Jacksonville, Florida, United States
15 Mayo Clinic, Jacksonville, Florida, United States
16 Mayo Clinic, Rochester, Minnesota, United States
17 Mayo Clinic, Rochester, Minnesota, United States
18 Mayo Clinic, Rochester, Minnesota, United States
19 Mayo Clinic, Rochester, Minnesota, United States
20 Mayo Clinic, Rochester, Minnesota, United States
21 Mayo Clinic, Rochester, Minnesota, United States
ABSTRACT
Introduction
Clonal cytopenia of uncertain significance (CCUS) is associated with variable risk of progression to myeloid neoplasia (MN). Although recurrent somatic mutations (MTs) related to progression have been identified, it remains unclear whether evolution to MN is driven by the acquisition of new molecular subtypes or by clonal expansion and genomic evolution within preexisting clones.
Methods
Patients (pts) with CCUS diagnosis within the Mayo Clinic system between 2015 to 2025 were included. We applied stringent criteria to exclude MN and only pts with bone marrow examination at time of CCUS diagnosis were included. Unsupervised k-mean machine learning clustering using MTs variant allele frequency (VAF) of the CCUS pts was used to identify molecular clusters (MC). The optimal number of MC was determined using silhouette value. MC signature was determined using enrichment analysis. The same MC signature was applied to a cohort of myelodysplastic syndrome (MDS) pts. We took into consideration that MTs in SF3B1 and TP53 and del5q are MDS defining. CCUS and MDS cases within the same MC were compared for the acquisition of new MTs, rates of cytogenetic abnormalities, or clonal size as estimated by VAF.
Results
We identified 208 pts with CCUS, the median age was 72 years, 70% were males and 30% were females. MTs in TET2 (36%), SRSF2 (21%), ASXL1 (21%), and ZRSR2 (17%) were the most common. Only 61 (29%) pts had abnormal karyotypes. Overall, 64 (31%) pts progressed to MN and 59 (28%) died during a median follow up time of 45 (95%CI: 41-53) months (mo).
Using unsupervised clustering, 5 MCs (n/%, signature) identified: MC1 (n = 28/13%, TET2), MC2 (n = 15/7%, TET2 and SRSF2), MC3 (n = 21/10%, SRSF2, IDH1/2), MC4 (n= 117/56%, U2AF1, DNMT3A and other mutations), and MC5 (n = 27/13%, ZRSR2 and TET2). Significant differences in baseline features were observed across the MCs. Age (p<0.001), gender (p<0.001), hemoglobin (p=0.04), platelets (p=0.01), and MN progression rate (p=0.02) differed significantly between clusters. Notably, MC5 consisted exclusively of male pts (100%), while MC3 had the highest progression rate (57%). Median overall survival (OS) was different between MC1 (70 mo), MC2 (not reached [NR]), MC3 (65 mo), MC4 (NR), and MC5 (50), p= 0.02.
We then applied the molecular signatures on the MDS cohort and identified cases with similar profiles (n/%): MC1 (59/12%), MC2 (24/5%), MC3 (46/9%), MC4 (258/51%), and MC5 (20/4%). Only 98 (19%) of the MDS cases were not assigned to a MC. MDS pts had higher rates of TP53 (24% vs. 6%, p<0.001), SF3B1 (16% vs. 4%, p<0.001), RUNX1 (11% vs. 4%, p=0.003), and BCOR (7% vs. 3%, p=0.03) MTs while CCUS pts had higher rates of TET2 (36% vs.19%, p<0.001), ZRSR2 (17% vs.4%, p<0.001) and IDH1 (6% vs.1%, p<0.001) MTs.
After that, we compared CCUS and MDS pts assigned to the same MC. Within MC1, MDS cases had higher rates SF3B1 MT (32% vs.11%, p=0.03) and lower rates of normal karyotype (79% vs.49%, p=0.01). In MC2, no signficant changes observed between CCUS and MDS pts. MDS pts within MC3 had higher rates of ASXL1 (52% vs. 24%, p=0.04) and STAG2 (30% vs. 0%, p=0.003) but lower rates of IDH1 (7% vs. 38%, p=0.003) MTs. SF3B1 (23% vs. 4%, p<0.001), RUNX1 (9% vs. 3%, p=0.04), STAG2 (5% vs. 0%, p<0.001) and TP53 (43% vs. 9%, p<0.001) MTs were signficantly higher in the MDS pts within MC4. ZRSR2 (6% vs. 1%, p=0.01), TET2 (9% vs. 0%, p<0.001), and SRSF2 (3% vs. 0%, p<0.001) MTs and normal KT (62% vs. 33%, p<0.001) were significantly higher in the CCUS pts. Similar to MC4, RUNX1 MT (30% vs. 4%, p=0.03) was higher in the MDS pts in MC5.
Finally, we compared the median VAF of the signature MTs within each MC between MDS and CCUS cases. The median VAF of TET2 MT in MDS was higher compared to CCUS pts within MC1 (57% v. 45%, p=0.04) and MC5 (49% vs. 37%, p=0.001). In MC4, the median VAF of DNMT3A MT in MDS pts was higher compared to CCUS pts (32% vs. 12%, p=0.002). No significant differences between other MTs VAF observed between CCUS and MDS pts within the same clusters.
Conclusions
Progression from CCUS to MDS is largely driven by clonal expansion, acquisition of cooperating high-risk lesions (RUNX1, ASXL1, and TP53) , and cytogenetic evolution within established molecular frameworks, rather than by the emergence of entirely novel molecular subtypes.These findings support a stepwise model of MN evolution and could facilitate early disease interception in high risk CCUS.
DISCLOSURES
Tariq Kewan, MD: No relevant conflicts of interest to declare.