AUTHORS
Usman Ilyas, MD, James Foran, FRCPC2, Mohamed Kharfan-Dabaja, MD3, Hemant Murthy, MD4, Jennifer Jiang, M.D.5, Yao-Shan Fan, M.D., Ph.D.6, Rong He, MD7, David Viswanatha, MD8, Aref Al-Kali, M.D.9, Abhishek Avinash Mangaonkar, MBBS10, Antoine N. Saliba, M.D.11, Tariq Kewan, MD12, Cecilia Arana Yi, M.D.13, Lisa Sproat, M.D.14, Jeanne Palmer, MD15, Mark R. Litzow, MD16, Ayalew Tefferi, MD17, Naseema Gangat, MBBS18, Nathan Punwani, M.D.19, Mrinal M. Patnaik, MD, MBBS20, Talha Badar, MD21
AFFILIATIONS
1 Mayo Clinic, Phoenix, Arizona, United States
2 Mayo Clinic, Jacksonville, Florida, United States
3 Mayo Clinic, Jacksonville, Florida, United States
4 Mayo Clinic, Jacksonville, Florida, United States
5 Mayo Clinic, Jacksonville, Florida, United States
6 Mayo Clinic, Jacksonville, Florida, 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, Rochester, Minnesota, United States
13 Mayo Clinic, Phoenix, Arizona, United States
14 Mayo Clinic, Phoenix, Arizona, United States
15 Mayo Clinic, Phoenix, Arizona, 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, Phoenix, Arizona, United States
20 Mayo Clinic, Rochester, Minnesota, United States
21 Mayo Clinic, Jacksonville, Florida, United States
ABSTRACT
Background: Cohesin-complex mutations (MT) are commonly grouped together in myeloid neoplasms (MN), although their clinicogenomic and prognostic implications may differ by gene. We evaluated the clinicogenomic, treatment-response, and outcome features of STAG2-, RAD21-, and SMC3-mutated clonal precursor states and overt MN.
Methods: We identified 216 patients with STAG2, RAD21, or SMC3 MT from a Mayo Clinic cohort of >11,000 cases. Diagnoses were assigned according to ICC 2022 (Arber et al., Blood). For MDS, IPSS-M very-low, low, and moderate-low categories were grouped as lower-risk, and moderate-high, high, and very-high as higher-risk. An MDS-related co-MT signature was defined by ASXL1, RUNX1, BCOR, or a spliceosome MT. To minimize disease heterogeneity, treatment, response, allogeneic hematopoietic cell transplantation (alloHCT), progression-free survival (PFS), and overall survival (OS) were evaluated in patients with bone marrow (BM) blasts >10%. Survival was estimated by Kaplan-Meier methods and compared by log-rank testing; Cox regression evaluated PFS and OS, with alloHCT modeled as a time-dependent covariate.
Results: Among 216 patients, 177 (82%) had STAG2, 32 (15%) RAD21, and 7 (3%) SMC3 MT. Median age was 73 years (IQR, 66-77), and 148 (69%) were male. The most common diagnoses were MDS (n=103; among 101 with available IPSS-M, 28 were lower-risk and 73 higher-risk), AML (n=52), and MPN/MDS-MPN overlap (n=46); 15 had CHIP/CCUS. Disease distribution differed by gene (P<0.001), with STAG2 accounting for 76/78 (97%) high-risk MDS and 39/52 (75%) AML cases. Median MT VAF was 41%, 14%, and 21%, respectively (P<0.001). Among classifiable variants, truncating events accounted for 95%, 87%, and 100%, respectively. STAG2 was enriched for the MDS-related co-MT signature versus RAD21/SMC3 combined (88% vs 51%; odds ratio, 6.69; P<0.001). Complex cytogenetics occurred in 6/177 (3%), 3/32 (9%), and 2/7 (29%) (P=0.013), and TP53 MT in 5/177 (3%), 4/32 (13%), and 2/7 (29%) (P=0.004), respectively. Therapy-related MN occurred in 32/177 (18%), 4/32 (13%), and 1/7 (14%), respectively (P=0.85). Among 37 therapy-related cases, TP53 MT, complex cytogenetics, and monosomal cytogenetics occurred in 1/37 (3%), 2/37 (5%), and 1/37 (3%), respectively; the monosomal case was also complex. Among all 52 patients with ICC-defined AML, NPM1 MT occurred in 1/39 (3%) STAG2, 4/11 (36%) RAD21, and 1/2 (50%) SMC3 cases (P=0.003).
The BM blast >10% subgroup included 69 patients: STAG2 (n=58), RAD21 (n=9), and SMC3 (n=2); 48 (70%) had AML. Complex cytogenetics occurred in 4/58 (7%), 1/9 (11%), and 0/2 (0%) (P=0.59), and TP53 MT in 2/58 (3%), 0/9 (0%), and 0/2 (0%) (P=1.00), respectively. Treatment distribution did not differ significantly (P=0.10); intensive/combination chemotherapy was used in 15/58 (26%), 6/9 (67%), and 1/2 (50%), and HMA plus venetoclax in 17/58 (29%), 1/9 (11%), and 0/2 (0%), respectively. Among 56 patients with evaluable first-line response, recorded complete response rates were 24/47 (51%) for STAG2, 5/7 (71%) for RAD21, and 0/2 (0%) for SMC3 (P=0.261). AlloHCT was performed in 16/58 (28%), 7/9 (78%), and 2/2 (100%), respectively (P=0.002). Median PFS was 10.4, 22.6, and 0.8 months (P=0.051), and median OS was 19.9 months, not reached, and not reached (P=0.094), respectively.
In exploratory time-dependent Cox analysis for survival (n=69), secondary/therapy-related AML was independently associated with inferior OS (HR, 2.23; 95% CI, 1.08-4.61; P=0.029), while alloHCT was associated with improved OS (HR, 0.22; 95% CI, 0.07-0.70; P=0.010) and PFS (HR, 0.04; 95% CI, 0.004-0.28; P=0.002). Complex cytogenetics was associated with inferior PFS (HR, 3.20; 95% CI, 1.00-10.25; P=0.050). RAD21 status, intensive chemotherapy, and HMA plus venetoclax were not independently associated with OS or PFS; TP53 was omitted because only 2/69 high-blast cases were mutated.
Conclusions: Canonical cohesin-complex mutations define distinct clinicogenomic phenotypes rather than a uniform molecular category. STAG2 identifies a prevalent, truncating, MDS-related, TP53-sparse genomic background, whereas RAD21-mutated AML is enriched for NPM1 co-mutation. Survival did not significantly differ by gene in high-blast disease; exploratory analyses identified associations with disease ontogeny, complex cytogenetics, and time-dependent alloHCT. These findings support gene-level interpretation of cohesin-complex MT and warrant external validation.
DISCLOSURES
Usman Ilyas, MD: No relevant conflicts of interest to declare.