AUTHORS
xueshuang zhang, MD, Yanli Zhao2, Jianping Zhang, MD3, Min Xiong4, Xingyu Cao, MD5, Deyan Liu6, Ruijuan Sun7, Zhijie Wei8, Jiarui Zhou9, Peihua (Peggy) Lu, MD10, Jing Long11, Yue Lu, PhD12
AFFILIATIONS
1 Hebei Yanda Ludaopei hospital, Langfang, China
2 Hebei Yanda Ludaopei hospital, Langfang, China
3 Hebei Yanda Ludaopei hospital, Langfang, China
4 Hebei Yanda Ludaopei hospital, Langfang, China
5 Hebei Yanda Ludaopei hospital, Langfang, China
6 Hebei Yanda Ludaopei hospital, Langfang, China
7 Hebei Yanda Ludaopei hospital, Langfang, China
8 Hebei Yanda Ludaopei hospital, Langfang, China
9 Hebei Yanda Ludaopei hospital, Langfang, China
10 Hebei Yanda Ludaopei hospital, Langfang, China
11 Hebei Yanda Ludaopei hospital, Langfang, China
12 Hebei Yanda Ludaopei hospital, Langfang, China
ABSTRACT
Background
Persistent molecular measurable residual disease (MRD) before allogeneic hematopoietic stem cell transplantation (allo-HSCT) is generally regarded as an indication of residual leukemia and an adverse prognostic marker in acute lymphoblastic leukemia. However, ZNF384-rearranged (ZNF384-r) B-ALL frequently exhibits persistent fusion transcript positivity despite deep immunophenotypic remission, raising an unresolved clinical question of whether isolated molecular persistence truly represents transplant-resistant disease or should delay allo-HSCT.
Methods
We retrospectively analyzed 60 consecutive patients with ZNF384-r B-ALL who underwent allo-HSCT in complete remission between April 2020 and December 2024. Fusion partners, immunophenotypic characteristics, co-occurring genomic alterations, and pre-transplant MRD assessed by multiparameter flow cytometry (MFC) and RT-qPCR were correlated with post-transplant outcomes. Overall survival (OS), leukemia-free survival (LFS), cumulative incidence of relapse (CIR), and non-relapse mortality (NRM) were analyzed.
Results
The cohort represented the biological diversity of ZNF384-r B-ALL. The median age was 21.5 years (IQR, 11.0–31.0), and the predominant fusion partners were EP300::ZNF384 (43.3%) and TCF3::ZNF384 (31.7%), while the remaining patients harbored diverse rare ZNF384 fusions, including TAF15, ARID1B, CREBBP, MEF2D, EWSR1, and LIMD1. Consistent with the characteristic phenotype of ZNF384-r B-ALL, aberrant myeloid-associated antigen expression was frequent, with CD13 and CD33 detected in 70.6% and 74.1% of evaluable patients, respectively. Genomic profiling revealed substantial molecular heterogeneity, with kinase/RAS pathway alterations identified in 58.3% of patients and additional recurrent alterations involving WT1 (31.0%), IKZF1 (19.0%), KMT2A/D (17.2%), and ETV6 (15.5%).
Before transplantation, 55 patients (91.7%) achieved MFC-MRD negativity, whereas molecular MRD remained detectable in 12 patients (20.0%). Among them, 9 patients (15.0%) exhibited isolated MFC-negative/molecular-positive (MFC−/Mol+) MRD discordance, observed across different ZNF384 fusion subgroups. This finding highlighted a clinically relevant discrepancy between immunophenotypic remission and molecular persistence in this disease subtype.
After a median follow-up of 28.2 months (range, 5.7–74.1), patients achieved favorable transplant outcomes, with 3-year OS, LFS, CIR, and NRM rates of 81.9%, 78.7%, 10.0%, and 11.3%, respectively. Notably, molecular MRD persistence did not translate into inferior clinical outcomes. Compared with patients with molecular MRD negativity, those with molecular positivity showed comparable 3-year OS (83.3% vs. 81.8%, P=0.806), LFS (66.7% vs. 80.5%, P=0.516), and CIR (25.0% vs. 7.3%, P=0.255). Strikingly, all nine patients with isolated MFC−/Mol+ MRD remained alive at last follow-up, achieving a 3-year OS of 100% and LFS of 80.0%, comparable to patients negative by both assays.
Furthermore, allo-HSCT appeared to overcome the prognostic impact of underlying biological heterogeneity. Outcomes were comparable among patients with EP300::ZNF384, TCF3::ZNF384, and other ZNF384 fusion partners, with no significant differences in OS, LFS, CIR, or NRM (all P>0.05). Similarly, recurrent genomic alterations, including kinase/RAS pathway mutations, IKZF1, ETV6, and KMT2A/D abnormalities, were not associated with inferior post-transplant survival (all P>0.05). Importantly, these findings remained unchanged after multivariable adjustment. Persistent molecular MRD was not independently associated with OS (HR 0.92, P=0.925) or LFS (HR 1.21, P=0.794), while no conventional clinical, molecular, or genomic variable independently predicted post-transplant survival.
Conclusions
Persistent ZNF384 fusion transcript positivity before allo-HSCT does not necessarily represent biologically meaningful residual leukemia capable of driving post-transplant relapse. Rather, allo-HSCT largely abrogates the prognostic impact of persistent molecular MRD and underlying genetic heterogeneity, supporting timely transplantation once MFC-MRD negativity is achieved rather than delaying allo-HSCT solely to eradicate residual fusion transcripts.
Key Words: ZNF384 rearrangement; B-ALL; allo-HSCT; Genetic heterogeneity; molecular MRD
DISCLOSURES
Yue Lu, PhD: No relevant conflicts of interest to declare.