Fus::ERG fusion represents adverse-plus risk group in AML patients: Evidence from a registry-based matched cohort Study

Abstract Background The Allo-HCT refined ELN 2022 classification identifies Adverse-Plus acute myeloid leukemia (AML) as the highest-risk subgroup among patients undergoing allogeneic hematopoietic stem cell transplantation (allo-HSCT). This group exhibits significantly inferior post-transplant outcomes, with a 5-year overall survival (OS) rate of 32.3%, a 5-year leukemia-free survival (LFS) rate of 24.3%, and 5-year cumulative incidence of relapse (CIR) of 64.3% (Blood Cancer J. 2025). FUS::ERG, resulting from the translocation t(16;21)(p11;q22), is a rare but recurrent cytogenetic abnormality observed in approximately 1% of de novo AML cases. Although emerging evidence suggests an association with poor clinical outcomes, FUS::ERG-positive AML is not currently recognized as a distinct adverse-risk genetic entity in the ELN 2022 classification. Methods We conducted a registry-based matched cohort study using data from transplantation centers participating in the Chinese Blood and Marrow Transplantation Registry Group (CBMTRG). Patients with FUS::ERG-positive AML were matched to FUS::ERG-negative controls at a 1:3 ratio using a combination of stratified exact matching for categorical variables—such as disease status prior to allo-HSCT (CR1 vs. ≥CR2), ELN 2022 risk stratification(Int vs. Adv), donor type (haploidentical, matched sibling donor [MSD], or matched unrelated donor [MUD])—and caliper matching (±5 years) for continuous variables including patient age and year of transplantation. Statistical analyses were conducted using R software (version 4.2.1). Results Between March 2009 and March 2025, a total of 189 patients were included in the final analysis, comprising 47 patients with FUS::ERG-positive AML and 142 matched FUS::ERG-negative controls at a 1:3 ratio. The median follow-up for the entire cohort was 36.6 months after allogeneic HSCT. Baseline characteristics were balanced between groups, with no significant differences in sex (male: 59.9% vs. 57.4%, P=0.77) or pre-HSCT measurable residual disease (MRD) status by FCM (MRD-positive: 27.7% vs. 21.8%, P=0.16),etc. The 3-year OS and LFS were significantly lower in the FUS::ERG-positive cohort compared to controls (OS: 45.9% [95% CI, 30.9–68.1%] vs. 87.7% [82.1–93.6%], P<0.001; LFS: 23.4% [12.6–43.3%] vs. 78.2% [71.5–85.4%], P<0.001). The 3-year CIR was markedly higher in the FUS::ERG-positive group (63.1% [44.1–77.2%] vs. 15.4% [9.9–22.0%], P<0.001), while non-relapse mortality (NRM) did not differ significantly (13.9% [5.5–26.2%] vs. 6.5% [3.2–11.4%], P=0.30). In multivariate analysis adjusting for age, ELN 2022 risk, MRD status, remission status, donor type, and GVHD severity, FUS::ERG positivity remained an independent risk factor for inferior OS (HR=6.40, 95% CI: 2.95–13.91, P<0.001), LFS (HR=4.11, 95% CI: 2.10–8.06, P<0.001), and higher CIR (HR=3.65, 95% CI: 1.79–7.44, P<0.001). Subgroup analyses within the ELN 2022 intermediate-risk population demonstrated significantly inferior outcomes for FUS::ERG-positive patients compared to controls (3-year OS: 52.7% vs. 85.1%, P<0.001; LFS: 29.1% vs. 76.9%, P<0.001; CIR: 63.5% vs. 13.1%, P<0.001). Similarly, within the ELN 2022 adverse-risk group, FUS::ERG-positive patients showed markedly worse 3-year OS (35.1% vs. 87.2%,P<0.001), DFS (13.1% vs. 72.8%,P<0.001), and higher CIR (65.2% vs. 22.8%,P<0.001). Among the FUS::ERG-positive patients (n=47), 38 had available transcript level data. Multivariate analysis identified pre-HSCT FUS::ERG transcript levels >1% as the only independent predictor of inferior OS (HR=7.61, 95% CI: 1.21–47.76, P=0.03) and LFS (HR=9.25, 95% CI: 2.43–35.16, P=0.001). Conclusion Our registry-based matched cohort study demonstrates that FUS::ERG fusion is independently associated with significantly inferior post-transplant survival in patients with AML. The markedly low LFS of 23.4% and high CIR of 63.1% are comparable to the dismal outcomes observed in Adverse-Plus AML following transplantation. These findings support the inclusion of FUS::ERG-positive AML as an adverse-risk entity in future guidelines. Furthermore, pre-transplant FUS::ERG transcript levels exceeding 1% were independently predictive of poor outcomes, suggesting potential utility as a biomarker for therapeutic decision-making.