Patients with TP53-mutated AML/MDS without complex or monosomal karyotype can achieve durable survival benefits with allogeneic HCT

Abstract

Abstract Introduction Whether allogeneic hematopoietic cell transplantation (HCT) is curative for acute myeloid leukemia (AML) or myelodysplastic syndrome (MDS) harboring TP53 mutations (TP53m) is under debate. TP53m often co-occur with complex (CK) and/or monosomal karyotype (MK), which promote chemotherapeutic resistance and result in poor overall (OS) and relapse-free survival (RFS). Absent CK/MK, the impact of TP53m on AML/MDS outcomes post-HCT is unclear. Here, we assess whether AML/MDS patients (pts) with either TP53m, CK/MK, or both have distinct outcomes after HCT. Methods We retrospectively reviewed 67 consecutive AML/MDS pts who underwent HCT at Stanford University from 2015-2024 and had available molecular and cytogenetic data. We categorized pts into 3 cohorts: 1) TP53m without CK/MK [TP53-only], 2) CK/MK without TP53m [CK/MK-only], or 3) Both TP53m and CK/MK [Both]. Cytogenetics and treatment responses were defined according to 2022 European LeukemiaNet guidelines. TP53 variant allele frequencies (VAF) were quantified using next-generation sequencing (NGS) assays. Biallelic TP53m was defined as ≥60% VAF or the presence of del(17p). Measurable residual disease (MRD) was detected using flow cytometry, cytogenetics, and/or NGS when available. Survival outcomes were estimated using Kaplan-Meier methods and compared using log-rank tests. Results Of 67 patients, 36 (54%) were in Both, 20 (30%) were in CK/MK-only, and 11 (16%) were in TP53-only. 57 pts (85%) had HLA-matched donors, 8 (12%) had 9/10 matched donors, 1 (1.5%) had umbilical cord blood, and 1 (1.5%) had a haploidentical donor. Donor distribution was balanced across cohorts. For conditioning, most pts received reduced intensity (RIC: TP53-only: 64% of cohort; CK/MK-only: 48%; Both: 44%) followed by myeloablative (MAC: TP53-only: 35%; CK/MK-only: 26%; Both: 36%), and nonmyeloablative (NMA: TP53-only: 1%; CK/MK-only: 26%; Both: 20%). In 45 evaluable pts, median TP53 VAF (mVAF) was 36% (range: 3%-87%). Missense mutations were the most common mutation type (75%). In TP53-only, 10/11 evaluable pts (91%) had monoallelic TP53 loss with mVAF 6%. In Both, 30/36 evaluable pts (83%) had biallelic TP53 loss with mVAF 38%. Both had significantly higher rates of biallelic TP53 loss (p<0.001) and mVAF (p<0.01) vs TP53-only. In total, 52 (77.6%) pts died: 81% from relapse and 18% from non-relapse mortality. Overall, median (mOS) and 1-year OS (OS1) were 19.2 months (mo) (10.2-37.3) and 55.2% (44.5%-68.5%), respectively. Median OS was 25.7 mo (11.1- NA) for TP53-only, 10.7 mo (6.7-NA) for CK/MK-only, and 20.3 mo (8.1-49.7) for Both, with no significant pairwise differences. OS1 was 63.6% (40.7%-99.5%) for TP53-only, 45.0% (27.7%-73.1%) for CK/MK-only, and 58.3% (44.3%-76.9%) for Both, with no significant pairwise differences. The median and 1-year cumulative incidence of relapse (CIR1) were 16.2 mo (95% CI 5.7-44.3) and 44.8% (32.5%-56.3%). TP53-only had significantly lower CIR1 of 18.2% (2.5%-45.6%) vs CK/MK-only of 65.0% (39.0%-82.1%, p <0.01). CIR1 for Both was 41.7% (25.3 - 57.2%), with no significant pairwise differences. Pts receiving RIC regimens had numerically longer mOS (27.4 mo, 7.2-NA) vs MAC (19.2 mo, 10.2-NA) and NMA (11.1, 5.3-NA) and lower CIR1 (RIC 35.5% [19.1%-52.3%] vs MAC 56.5% [33.5%-74.3%] and NMA 46.2% [17.5%-71%]). In AML, detectable MRD pre-HCT was associated with numerically shorter mOS (19.2 mo, 4.6-NA) compared to undetectable MRD (30.0 mo, 8.1-NA). MDS pts with active disease at HCT had numerically lower OS1 (47.1% [28.4%-77.9%] vs 55.6% [31.0-99.7%]) and mOS (11.1 mo [4.5-NA] vs 17.4 mo [7.3-NA]) compared to pts in remission at HCT. The presence of cGVHD as a time-varying covariate trended towards association with longer OS [HR 0.49 (0.22-1.11)] Discussion Pts with TP53m AML/MDS lacking CK/MK have decreased 1-year relapse rates after HCT compared to pts with CK/MK. OS parameters trended towards superiority for TP53-only vs other cohorts, which may reach significance in larger samples. Lower conditioning intensity also does not appear to adversely impact relapse or OS. Development of cGVHD may correlate with prolonged OS, reflecting graft-vs-leukemia effects. Even in the context of TP53m, CK/MK is likely the primary driver of poor outcomes after HCT, though larger studies are needed to confirm these findings. Pts with TP53m without CK/MK should be considered for HCT, while novel relapse reduction strategies are needed for patients with CK/MK.