Low EVI1 expression at diagnosis identifies a high-risk subgroup in adult Ph-negative B-cell acute lymphoblastic leukemia

Background: The prognostic impact of EVI1 expression in B-cell acute lymphoblastic leukemia (B-ALL) remains to be explored.

Method: Bone marrow (BM) samples collected from 436 consecutive newly diagnosed adult Ph-negative B-ALL patients were tested for EVI1 transcript levels using real-time quantitative PCR.

Result: The median EVI1 transcript level in the whole cohort was 0.40% (range: 0.0030–94.3%). Low EVI1 expression defined by lower three quartiles (EVI1 transcript levels was 2.3%) was significantly related to poorer relapse-free survival (RFS) and overall survival (OS) (p = 0.0010 and < 0.0001) and was an independent adverse prognostic factor for RFS (HR (95% confidence interval): 2.3 (1.2–4.1), p = 0.0070) and OS (HR (95% CI): 2.5 (1.3–5.1), p = 0.0090) in the whole cohort. The optimal thresholds for EVI1 transcript levels in the fusion gene subgroups were determined individually, and low EVI1 expression was related to or tended to be related to poorer RFS in patients with TCF3::PBX1, Ph-like fusions, MEF2D fusions, and ZNF384 fusions groups (p = 0.0047, 0.025, 0.032, and 0.070) and was associated with poorer OS in ZNF384 fusions groups (p = 0.012), respectively. Furthermore, patients with TCF3::PBX1 or MEF2D fusion and high EVI1 expression had RFS and OS similar to those without the corresponding fusions, and patients with ZNF384 fusion and low EVI1 expression had RFS and OS comparable to those without ZNF384 fusions (all p > 0.05).

Conclusion: Low EVI1 transcript levels at diagnosis are related to poor prognosis in adult Ph-negative B-ALL, and EVI1 expression may improve fusion gene-defined risk stratification.

Introduction

The ecotropic viral integration site 1 (EVI1) is a regulatory transcription factor that plays an important role in hematopoiesis and self-renewal (1). It was first identified as a common ecotropic viral integration in the DNA of AKXD murine myeloid tumors (2). In human malignancies, EVI1 is frequently activated by chromosomal rearrangements at the 3q26.2 locus in acute myeloid leukemia (AML), myelodysplastic syndrome (MDS), and chronic myeloid leukemia (CML) (3, 4). The presence of 3q26.2 abnormalities in these diseases is consistently associated with adverse clinical outcomes (5–8). Notably, elevated EVI1 expression is also common in AML patients without 3q26.2 abnormalities and similarly confers an adverse prognosis (9). Previously, our center had reported that high EVI1 expression independently predicted poor outcomes in AML patients with intermediate cytogenetic risk (ICR-AML) receiving chemotherapy (10).

Compared with the extensive research on the expression and functional role of EVI1 in myeloid malignancies, fewer studies have shown examined its involvement in lymphoid malignancies. Unlike in AML, 3q26.2 rearrangements are fairly rare in acute lymphoblastic leukemia (ALL) and have been reported only in isolated cases of secondary ALL (11, 12). Nevertheless, dysregulated EVI1 expression has been observed in both pediatric and adult ALL (13–16). Konantz et al. have reported that EVI1 expression contributes to the leukemogenic potential and apoptosis resistance of ALL cells (17). However, our group has reported that low EVI1 expression was an independent adverse prognostic factor for relapse-free survival (RFS) and overall survival (OS) in pediatric B-ALL patients (18). In adult ALL, only two studies have been reported to date: one found no prognostic impact of either low or high EVI1 expression on the cumulative incidence of failure (CIF, failure meaning primary refractoriness or relapse) and event-free survival (EFS) in Ph-negative B-cell ALL (B-ALL) patients, while the other showed no significant impact of EVI1 expression levels on OS and disease-free survival (DFS) rates in ALL patients (including B- and T-ALL) (14, 16). With the recent identification of several new fusion genes that have refined molecular classification and risk stratification in B-ALL (19, 20), a comprehensive evaluation of the relationship between EVI1 expression patterns, molecular subtypes, and prognostic significance remains a significant unmet need.

In this study, we conducted a large-scale cohort study enrolling 436 consecutive adult Ph-negative B-ALL patients. All patients underwent comprehensive screening for both classical and new fusion genes. We retrospectively performed real-time quantitative RT-PCR (RQ-PCR) to detect EVI1 transcript levels in the bone marrow samples collected at diagnosis. Subsequently, we investigated the EVI1 expression pattern and the prognostic significance.

Patients and methods

Patients and treatment

A total of 436 newly diagnosed adult Ph-negative B-ALL patients were included in this study. They were consecutively diagnosed from December 2009 to April 2024 and received at least one course of induction chemotherapy in our institute. There were 205 (47.0%) male patients in total. The median age at diagnosis was 33 (range 16 to 65) years. The diagnosis was based on bone marrow morphology, immunophenotyping, chromosomal karyotyping, and molecular testing. The last follow-up was conducted in September 2024.

The chemotherapy procedures in our institute have been reported in detail in previous studies (21, 22). Induction therapy was performed using the CODP±L (cyclophosphamide and prednisone or dexamethasone, vincristine, and daunorubicin or idarubicin and L-asparaginase) regimen. Patients who achieved first complete remission (CR1) subsequently received post-remission therapy, which consisted of either hyper-CVAD-based (high-dose methotrexate and cytarabine therapy) chemotherapy only or chemotherapy followed by allogeneic hematopoietic stem cell transplantation (allo-HSCT). The indications for allo-HSCT, pretreatment protocols, donor selection, and prevention of graft-versus-host disease have been described previously (23). Minimal residual disease (MRD) was assessed in all patients by multiparameter flow cytometry after each cycle of chemotherapy and after HSCT.

Extraction of RNA and DNA

Bone marrow samples were collected from all patients at diagnosis for RNA and DNA extraction. Nucleated cells were obtained by treating fresh bone marrow samples with 0.144 M NH4Cl, 0.01 M NH4HCO3 to lyse the red cells. Total RNA and genomic DNA were individually extracted using Trizol and DNAzol reagents (Invitrogen, Carlsbad, CA, USA).

Detection of classical and new fusion genes and IKZF1 deletion

RNA was reverse transcribed into cDNA. TaqMan-based RQ-PCR was performed using cDNA to screen B-ALL-related classical fusion transcripts at diagnosis, including BCR::ABL1, TCF3::PBX1, TEL::AML1, and KMT2A rearrangement (KMT2A::AFF1, KMT2A-MLLT1, KMT2A::MLLT6, KMT2A::EPS15, and KMT2A::MLLT11. IKZF1) deletions were detected by RQ-PCR using DNA in 239 patients (24). Furthermore, all patients were retrospectively screened for new fusion transcripts, including ZNF384, MEF2D, and Ph-like fusion genes using TaqMan-based multiplex RQ-PCR, as previously reported (25).

Detection of EVI1 transcript levels

As reported in our previous study (10), TaqMan-based RQ-PCR was used to measure EVI1 transcript levels. The primers and probes for the EVI1 transcript were designed using Primer Express Software (Applied Biosystems, Foster City, USA) to detect all subtypes (1a, 1b, 1c, 1d, and 3 L), and the sequences were as follows:

Forward primer: 5’-CCCATGTGCCAGAGGAACTT-3′ (in exon 14).

Reverse primer: 5’-CAGTGACAGCATCATAGCATATGC-3′ (in exon 15).

Probe: 5’-FAM-CAGCCGTTACACAGAAAGTCCAAATCGC-TAMRA-3′ (in exon 14).

The primers and probe for ABL1 were based on a report from the Europe Against Cancer Program (26). The amplification efficiency of EVI1 and ABL1 was validated by constructing standard curves, both of which fell within the acceptable range of 90–110% with R² > 0.99 (Supplementary Figure S1). The EVI1 transcript level was calculated as EVI1 copies /ABL1 copies in percentage.

Definitions and statistical analysis

Complete remission (CR) was defined as the absence of circulating lymphoblasts and extramedullary disease, restoration of trilineage hematopoiesis (TLH) and <5% leukemic blasts, a neutrophil count of more than 1 × 10⁹/L, and a platelet count of more than 100 × 10⁹/L (20). Relapse was defined as the recurrence of more than 5% of BM blasts, reappearance of peripheral blasts in the blood, or in any extramedullary site after CR. Relapse-free survival (RFS) was measured from the date of achieving CR to relapse or to the last bone marrow testing. Overall survival (OS) was measured from the date of diagnosis to the date of death (regardless of cause) or the last follow-up. High-risk karyotype was defined based on the Eastern Cooperative Oncology Group (ECOG) 2,993 trial (27). Fusion transcript testing results by RQ-PCR were also considered.

The Mann–Whitney U test was performed on continuous variables, and the chi-squared or Fisher’s exact test was performed on categorical variables. The optimal cutoff levels were determined by dividing EVI1 transcript levels into quartiles or using receiver operating characteristic (ROC) curves based on different patient responses (achieving CR). Survival functions were estimated using the Kaplan–Meier method and were compared using the log-rank test. Variables with a p-value of < 0.05 from the univariate analyses were entered into the multivariate model using Cox proportional hazards models to identify the most statistically significant parameters associated with RFS and OS. The level of statistically significant difference was set at a p-value of < 0.05. SPSS software 26.0 (IBM Corporation, Armonk, NY) and GraphPad Prism 10 (GraphPad Software Inc., La Jolla, CA) were used for statistical analysis.

Results

Patient outcomes

In the entire cohort, 405 (92.9%) patients achieved CR after induction chemotherapy, and 136 patients subsequently relapsed with a median time of 5.5 (range: 0.67–69.5) months. Of the 405 patients who achieved CR, 148 received chemotherapy alone, while the remaining 257 patients underwent allo-HSCT after chemotherapy. The median follow-up duration for all 436 patients was 25.7 (range: 1.0–155.7) months. A total of 291 (66.7%) patients were alive at the last follow-up with a median follow-up time of 41.0 (range: 1.0–155.7) months. The 4-year RFS and OS rates of the entire cohort were 64.0% (95% confidence interval (CI): 58.5–68.9%) and 63.7% (95% CI: 58.4–68.5%), respectively.

Molecularly and cytogenetically defined patient groups

In the entire cohort, 186 (42.7%) patients were identified with fusion genes (Supplementary Table S1). A total of 23 (5.3%) patients had the TCF3::PBX1 fusion transcripts; 39 (8.9%) had an KMT2A rearrangement, including KMT2A::AFF1 (n = 35), KMT2A::MLLT1 (n = 3) and KMT2A::EPS15 (n = 1); 73 (16.7%) had ZNF384 fusion transcripts, including EP300::ZNF384 (n = 56), CREBBP::ZNF384 (n = 8), TAF15::ZNF384 (n = 4), TCF3::ZNF384 (n = 3), and EWSR1::ZNF384 (n = 2); 21 (4.8%) had MEF2D fusion transcripts, including MEF2D::BCL9 (n = 11), MEF2D::HNRNPUL1 (n = 8), MEF2D::DAZAP1 (n = 1), and MEF2D::FOXJ2 (n = 1); 19 (4.3%) had Ph-like fusion transcripts, including P2RY8::CRLF2 (n = 6), EBF1::PDGFRB (n = 3), RCSD1::ABL2 (n = 3), NUP214::ABL1 (n = 2), TEL::ABL1 (n = 1), EBF1::JAK2 (n = 1), PCM1::JAK2 (n = 1), PAX5::JAK2 (n = 1), and BCR::FGFR1 (n = 1); and 11 (2.5%) had hyperdiploidy karyotype. The remaining patients (n = 250, 57.3%) were defined as B-other in the current study. It should be noted that neither chromosome 3q26.2 abnormalities nor cryptic 3q26.2 rearrangements were found in the whole cohort.

EVI1 expression patterns in patients at diagnosis

The median EVI1 transcript levels for all patients were 0.40% (range:0.0030–94.3%). Compared with 27 normal bone marrow (NBM) samples collected from healthy donors, as we previously reported (10), Ph-negative B-ALL patients had significantly lower EVI1 transcript levels (p < 0.0001).

As shown in Figure 1 and Supplementary Table S1, the EVI1 transcript levels were varied significantly among the ZNF384 fusion, Ph-like fusion, TCF3::PBX1 fusion, hyperdiploidy, MEF2D fusion, KMT2A rearrangement, and B-other groups (p < 0.0001) with the median levels of 5.8% (range: 0.11–70.6%), 1.2% (range: 0.045–72.9%), 0.41% (range: 0.013–14.7%), 0.34% (range: 0.014–48.2%), 0.14% (range: 0.034–4.3%), 0.090% (range: 0.0040–0.50%), and 0.38% (range: 0.003–94.3%), respectively. The ZNF384 fusion group had significantly higher EVI1 transcript levels than TCF3::PBX1 fusion, MEF2D fusion, KMT2A rearrangement, and B-other groups (all p < 0.0001) and had no statistically significant difference compared to Ph-like fusion and hyperdiploidy groups (p = 0.34 and 0.18).

Figure 1

Figure 1. EVI1 expression patterns of molecularly and cytogenetically defined B-ALL groups.

Determining the optimal cutoff value of the EVI1 transcript levels for patient grouping

Patients were grouped into quartiles according to EVI1 transcript levels (1st quartile to fourth quartile, from low to high levels). As shown in Figure 2, both RFS and OS were significantly different among the four groups (p = 0.0050 and p = 0.0010, Figures 2A,B). The bottom three quartile groups demonstrated significantly poorer RFS and OS than the top quartile group (RFS: p = 0.012, 0.0004, and 0.014, OS: p = 0.0010, 0.0060, and 0.0001, respectively), while no significant differences were observed among the remaining three groups (RFS: p = 0.37, OS: p = 0.82). Therefore, we used the upper quartile (top 25%, EVI1 transcript levels were 2.3%) as the cutoff value to stratify the patients into EVI1-H (n = 109) and EVI1-L (n = 327) groups.

Figure 2

Figure 2. RFS and OS analysis in the entire cohort. Patients were grouped into quartiles according to EVI1 transcript levels (A,B); comparison of RFS (C) and OS (D) between EVI1-L and EVI1-H group.

Relationship between EVI1 expression and patient characteristics

As shown in Table 1, in the entire cohort, low EVI1 expression was significantly related to KMT2A rearrangement, TCF3::PBX1 fusion, ZNF384 fusion, IKZF1 deletion, high-risk karyotype, and fusion-defined high-risk group (p < 0.0001, 0.019, < 0.0001, 0.036, 0.0006, and < 0.0001), respectively. A non-significant trend was also observed for an association with lower platelet counts (p = 0.051). In contrast, EVI1 expression had no relationship with age, sex, WBC counts, hemoglobin, hyperdiploidy karyotype, MEF2D fusion, Ph-Like fusion, and immunophenotype-defined group (all p > 0.05).

Table 1

Table 1. Relationship between EVI1 expression and variables at diagnosis in adult Ph-negative B-ALL.

The impact of EVI1 expression at diagnosis on CR achievement

In the whole cohort, the CR rate after one course of induction therapy in the EVI1-L group was significantly higher than that in the EVI1-H group (88.1% versus 78.9%, p = 0.018); while the CR rate after two courses of induction therapy wase similar between them (90.4% versus 89.7%, p = 0.84).

Low EVI1 expression predicted poor outcome in the whole cohort

As shown in Figure 2, both RFS and OS were significantly poorer in the EVI1-L group than those in the EVI1-H group in the entire cohort (4-year RFS rate: 57.9% [95% confidence interval (CI) 51.2–63.6%] vs. 78.0% [95% CI 67.5–85.4%], p = 0.0012, Figure 2C; 4-year OS rate: 57.0% [95% CI 50.7–62.8%] vs. 82.7% [95% CI 73.6–88.9%], p < 0.0001, Figure 2D). In the chemotherapy only subgroup, EVI1-L patients showed a trend of poorer OS than EVI1-H patients, but no significant differences was observed in RFS between them (p = 0.092 and 0.40, Supplementary Figures S2A,B); while in allo-HSCT subgroup, the EVI1-L patients had significantly lower RFS and OS rate than EVI1-H patients (p = 0.026 and 0.0003, immunophenotyping defined group Supplementary Figures S2C,D).

Univariate and multivariate analyses of RFS and OS in the whole cohort

In the entire cohort, except for low EVI1 transcript levels, age ≥ 40 years, WBC ≥ 20×10⁹/L, platelet count < 67×10⁹/L, high-risk karyotype, fusion-defined high-risk group (28), treated with chemotherapy only, no CR after one course induction, and MRD > 0.01% after the first consolidation were all significantly associated with a lower RFS rate (all p < 0.05, Table 2). As shown in Figure 3, the multivariate analysis showed that EVI1-L (HR (95%CI): 2.3 (1.2–4.1), p = 0.0070), age ≥ 40 years (HR (95%CI): 1.6 (1.0–2.5), p = 0.039), fusion-defined high-risk group (HR (95%CI): 2.3 (1.2–4.0), p = 0.0070), treating with chemotherapy only (HR (95%CI): 4.4 (2.7–7.2), p < 0.0001), no CR after one course induction (HR (95%CI): 6.0 (2.8–13.0), p < 0.0001), and MRD > 0.01% after the first consolidation (HR (95%CI): 2.6 (1.6–4.2), p < 0.0001) were independent poor prognostic factors for RFS.

Table 2

Table 2. Univariate and multivariate analysis of RFS in the entire cohort.

Figure 3

Figure 3. Forest plots for multivariate analysis of RFS and OS.

As shown in Table 3, for OS, in addition to low EVI1 expression, age ≥ 40 years, WBC ≥ 20×10⁹/L, platelet count < 67×10⁹/L, high-risk karyotype, fusion-defined high-risk group, treated with chemotherapy only, and no CR after one course of induction were all associated with a lower OS rate (all p < 0.05). The multivariate analysis revealed that low EVI1 expression (HR (95%CI): 2.5 (1.3–5.1), p = 0.0090), age ≥ 40 years (HR (95%CI): 1.8 (1.1–3.0), p = 0.011), treating with chemotherapy only (HR (95%CI): 2.8 (1.7–4.5), p < 0.0001), and no CR after one course induction (HR (95%CI): 3.9 (1.9–8.0), p < 0.0001) were independent poor prognostic factors (Figure 3).

Table 3

Table 3. Univariate and multivariate analysis of OS in the entire cohort.

The impact of EVI1 expression on outcome within the individual fusion gene groups

We first performed survival analyses of patients with or without the fusion gene. KMT2A rearrangement and MEF2D fusions were significantly related to worse RFS (p < 0.0001 and 0.001) and ZNF384 fusions were significantly related to better RFS (p = 0.003), whereas TCF3::PBX1 and Ph-like had no relation to RFS (all p = 0.22). Moreover, KMT2A rearrangement is significantly related to worse OS (p = 0.012), and ZNF384 fusions are significantly related to better OS (p = 0.025).

As shown in Table 1, all patients (100%) with KMT2A rearrangement and the vast majority of patients (> 90%) with TCF3::PBX1 or MEF2D fusions were classified as EVI1-L if using the cutoff values (2.3%) applied to the whole cohort. Thus, we performed quartile stratification for each fusion gene subgroup and found no significant difference in RFS and OS among quartile groups (all p > 0.05). Then, we performed ROC curve analysis using relapse or not as a distinguishing variable to determine the cutoff value for EVI1 transcript levels within each fusion-defined group. As a result, 0.33, 0.05, 2.0, 0.5, and 0.9% were optimal thresholds for TCF3::PBX1, KMT2A rearrangement, ZNF384 fusion, MEF2D fusion, and Ph-like fusion groups, and patients with EVI1 levels greater than or equal to the cutoff value and those with levels less than the cutoff value were defined as the EVI1-H and EVI1-L, respectively. We further evaluated the impact of EVI1 expression on RFS and OS within the individual group. The RFS of patients in the EVI1-L group was or tended to be significantly inferior to that of patients in the EVI1-H group within TCF3::PBX1, Ph-like fusions, MEF2D fusions, and ZNF384 fusions cohorts (p = 0.0047, 0.025, 0.032, and 0.07), respectively. The OS of the EVI1-L group was significantly poorer than that of the EVI1-H group in the ZNF384 fusions cohort (p = 0.012). There was no difference in both RFS and OS between the EVI1-L and the EVI1-H groups in KMT2A rearrangement cohort.

We further made comparisons in the whole cohort, in which patients were categorized into the following three groups: EVI1-H with the fusion gene, EVI1-L with the fusion gene, and without the corresponding fusion gene. As shown in Figure 4, patients in the EVI1-H with TCF3::PBX1 or MEF2D fusion group had similar RFS and OS in both cohorts compared to those without the corresponding fusions (TCF3::PBX1: p = 0.14 and 0.89; MEF2D: p = 0.23 and 0.17), while patients in the EVI1-L with the TCF3::PBX1 group exhibited a significantly poorer RFS or tended to have significantly poorer OS (p < 0.0001 and 0.061) compared to those without TCF3::PBX1, and patients in the EVI1-L with MEF2D fusion group exhibited a significantly poorer RFS and similar OS (p < 0.0001 and 0.25) compared to those without MEF2D fusion. In Ph-like cohort, patients in the EVI1-L group had similar RFS and tended to have poorer OS (p = 0.66 and 0.054), and those in the EVI1-H group tended to have better RFS but similar OS (p = 0.055 and = 0.93) than those without the Ph-like fusions. In addition, patients in the EVI1-L with ZNF384 fusion group had similar RFS and OS (p = 0.54 and 0.85), and EVI1-H with ZNF384 fusion group had significantly better RFS and OS (p = 0.008 and 0.0024), compared to those without ZNF384 fusion. In summary, EVI1 transcript level serves as a critical prognostic modifier within specific fusion-defined subgroups. High EVI1 expression identified a lower-risk subgroup within traditionally high-risk categories defined by TCF3::PBX1, MEF2D, and Ph-like fusions. Conversely, low EVI1 expression defines a higher-risk subgroup within the typically more favorable ZNF384 fusion group.

Figure 4

Figure 4. RFS and OS in fusion-defined groups. RFS (A) and OS (F) in the TCF3::PBX1 group; RFS (B) and OS (G) in the MEF2D fusions group; RFS (C) and OS (H) in the ZNF384 fusions group; RFS (D) and OS (I) in the Ph-like fusions group; RFS (E) and OS (J) in the KMT2A rearrangement group. The optimal EVI1 expression cutoff was individually determined for each genetic subgroup using receiver operating characteristic (ROC) curve analysis.

Discussion

In this study, we thoroughly evaluated EVI1 expression patterns at diagnosis and its prognostic significance in the consecutive 436 adult Ph-negative B-ALL patients. We found that low EVI1 expression, defined as the lower three quartiles of the quartile distribution, was an independent poor prognostic factor for RFS and OS in the entire cohort. Furthermore, the subgroup analysis showed that low EVI1 expression was related to poorer RFS within the TCF3::PBX1, MEF2D fusions and Ph-like fusions groups, and related to poorer OS in the ZNF384 group.

In both adult and pediatric AML, those with high expression of EVI1 have poor outcomes (9, 10, 29, 30). In contrast, this is not universally observed in ALL. The current findings suggested that the consistent prognostic significance of EVI1 expression in adults is consistent with pediatric Ph-negative B-ALL patients as we previously reported (18), wherein lower transcript levels being associated with poorer prognosis, despite differences in cutoff values for EVI1. Moreover, we found that low EVI1 expression was an independent poor prognostic factor for both RFS and OS in the current cohort. Abbal et al. assessed EVI1 gene expression in a cohort of 138 adult Ph-negative B-ALL patients and found no prognostic impact for either low (0.05%) or high (1.65%) EVI1 expression (14). Similarly, Nabil et al. reported no significant impact of EVI1 expression on survival in a cohort of 71 adult ALL patients (including both B-ALL and T-ALL) (16). We propose that several methodological factors may account for these discrepant results. First, the criteria for patient inclusion are critical. ALL is a heterogeneous disease, and the inclusion of T-ALL, a subtype with distinct molecular pathogenesis, may obscure subtype-specific prognostic signals. Unlike the above two studies, our analysis was confined to patients with Ph-negative B-ALL, enhancing the homogeneity of our cohort. Second, statistical power is a key consideration. Our study enrolled 436 adult patients, representing the largest cohort to date for this specific investigation, which improves the reliability of the findings. Finally, the methodology for defining EVI1 expression groups differed substantially. Abbal et al. utilized the 1st and 99th percentiles of a control group, a highly stringent approach that may classify very few patients as having “low” or “high” expression. In contrast, we employed a quartile-based cutoff, defining the lower three quartiles and upper quartile as the low and high expression groups, respectively. This approach provides a more balanced distribution for robust statistical comparison and may be more sensitive for detecting a continuous relationship between expression levels and clinical outcome.

The pathogenic mechanisms of the EVI1 gene in AML have been partially explored, implicating it in several key signaling pathways. For instance, EVI1 is known to relieve cellular growth inhibition mediated by transforming growth factor beta (TGF-β) (31) and promotes the proliferation of myeloid preleukemic cells by upregulating Spi1 in encoding PU.1. (32). In contrast, the mechanisms of EVI1 in lymphoid malignancies remain poorly characterized, and only Konantz et al. reported the in vitro study results on ALL to date. They showed that the apoptosis rates were significantly elevated after knockdown of EVI1 expression in ALL cell lines NALM-16 and REH, and found that EVI1 modulates expression of several apoptosis-related genes (such as BCL2, BCL-x, XIAP, NOXA, PUMA, TRAIL-R1) (17). Our research team is currently conducting in vitro study to investigate the mechanism of EVI1 expression on B-ALL cells.

In AML, the hallmark of 3q26.2 rearrangement is high EVI1 expression, primarily driven by the hijacking of the GATA2 distal hematopoietic enhancer (4, 33, 34). In contrast, 3q26.2 rearrangement is exceptionally rare in ALL, having been reported only in isolated cases of secondary ALL (11, 12). Consistent with this finding, our analysis of this Ph-negative B-ALL cohort identified no evidence of 3q26.2 chromosomal abnormalities or cryptic 3q26.2 rearrangements. This absence strongly suggests that the mechanism underlying EVI1 overexpression in this context is independent of 3q26.2 rearrangement (9, 34, 35).

We observed no significant differences between the low and high EVI1 expression groups in adult Ph-negative B-ALL patients across multiple clinical parameters, including sex, age, WBC count, and hemoglobin. These findings are consistent with those of other studies. Konantz et al. showed that EVI1 expression is not associated with age and WBC count in ALL (17), a result corroborated by Abbal et al. in their adult Ph-negative B-ALL cohort (14). In contrast, unlike our research on adult ALL, Steven et al. discovered that EVI1 expression is age-related and hypothesized that TGF-β might be a significant factor influencing-related gene expression in pediatric ALL (13). Moreover, significant differences emerged in the distribution of specific molecular and cytogenetic alterations. The EVI1 expression groups were distinct in their prevalence of KMT2A rearrangements, TCF3::PBX1 fusions, ZNF384 fusions, IKZF1 deletions, and certain cytogenetic categories. This finding, however, diverges from the report by Abbal et al., who did not observe significant associations between EVI1 expression and IKZF1 deletions or complex karyotypes (14). We posit that several factors may account for these discrepant results. Key considerations include differences in cohort composition (e.g., the specific distribution of genetic subtypes), statistical power due to varying sample sizes, and potentially the statistical methodologies employed.

Differing from a significant correlation between high EVI1 expression and KMT2A rearrangement established in both adult and pediatric AML patients (9, 10, 13, 17, 36), our observations in adult Ph-negative B-ALL patients indicate an association between KMT2A rearrangement and uniformly fairly low EVI1 expression levels (median, 0.090%, range: 0.0040–0.50%). This finding is consistent with our previous findings in pediatric B-ALL (18) and are further supported by literatures performed by Abbal et al. and Konantz et al., in which a significant downregulation of EVI1 gene expression were also found in adult Ph-negative ALL patients with the KMT2A::AFF1+/ t(4;11) rearrangement (14, 17).

In this study, all patients were tested for not only classical but also new fusion genes, all of which serve as critical indicators for genetic stratification in B-ALL. Currently, risk stratification by fusion gene in B-ALL remains inconsistent. The 2024 NCCN guidelines classify ZNF384 fusions as high-risk and TCF3::PBX1 as standard-risk, while the 2024 ELN recommendations categorize them as intermediate-risk and favorable-risk, respectively (19, 20). A multinational RNA-sequencing study of 1,223 B-ALL cases found that ZNF384 fusions confer an intermediate-risk and TCF3::PBX1 a high-risk profile in adult B-ALL patients (37). In line with this study, our center’s consecutive decade-long cohort study identified ZNF384 fusions as standard-risk and TCF3::PBX1 as high-risk (25, 28). Similarly, a Japanese study reported the favorable prognosis of ZNF384 in adult B-ALL patients (38). Regarding the Ph-like subtype in our cohort, our previous study showed that it did not have inferior RFS but poor OS, which was caused by low 1-course CR rate (28). In addition, the Ph-like subtype sample size is limited (n = 19), which is consistent with data from Taiwan but lower than those from Western populations (19, 20, 39). These discrepancies may stem from ethnic variations and differences in treatment strategies. In particular, the use of allo-HSCT may significantly affect the prognosis of adult Ph-negative B-ALL.

To our knowledge, no prior study has integrated EVI1 transcript levels with both classical and new fusion genes to assess their combined prognostic significance. In the present cohort, the highest level of EVI1 expression was observed in ZNF384 fusions group. Given the favorable prognosis associated with ZNF384 fusions in our previous reports (25, 28), this may be one reason for the good prognosis of the high EVI1 expression. In order to clarify the prognostic role of EVI1 transcript levels, we evaluated the prognostic significance of EVI1 within each fusion-defined subgroup. We found that low EVI1 expression is related to poorer RFS in the TCF3::PBX1, MEF2D fusion and Ph-like fusion groups, with poorer OS—and tended to related to poorer RFS—in the ZNF384 group. No significant difference in both RFS and OS between the EVI1-L and the EVI1-H subgroups in KMT2A rearrangement cohort was observed.

Furthermore, we attempted to combine EVI1 expression with a fusion gene for further risk stratification of the entire cohort. Among patients with high-risk fusion (TCF3::PBX1, MEF2D, and Ph-like), those with high EVI1 expression exhibited favorable outcomes comparable to those without such fusions. Conversely, among patients with ZNF384 fusions, low EVI1 expression identified a subgroup with outcomes similarly unfavorable to those of high-risk fusion carriers. It illustrated that EVI1 expression may improve fusion-defined risk stratification in adult Ph-negative B-ALL.

In conclusion, our study demonstrates that EVI1 transcript levels at diagnosis are highly variable and exhibit a distinct association with specific fusion gene subtypes in adult Ph-negative B-ALL. Critically, low EVI1 expression served as a powerful, independent predictor of adverse outcomes, not only in the overall cohort but also within key molecular subgroups, including those with TCF3::PBX1, Ph-like, MEF2D, and ZNF384 fusions. Notably, the integration of EVI1 expression with fusion gene status refined risk stratification, effectively identifying standard-risk patients within traditionally high-risk fusion groups and high-risk patients within otherwise favorable-risk groups. These findings underscore the clinical utility of incorporating EVI1 transcript quantification into the initial diagnostic workup of adult Ph-negative B-ALL. We acknowledge the limitations inherent in our retrospective study design and the heterogeneity in treatment protocols. Therefore, prospective validation in independent, multi-center cohorts is essential to confirm the generalizability of our observations and to further elucidate the prognostic and potential therapeutic implications of EVI1 expression in the context of modern, molecularly defined ALL therapy.

Data availability statement

The original contributions presented in the study are included in the article/Supplementary material, further inquiries can be directed to the corresponding author.

Ethics statement

The studies involving humans were approved by Ethics Committee of the Peking University People’s Hospital (2023PHB301). The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

Author contributions

SK: Data curation, Formal analysis, Methodology, Writing – original draft, Writing – review & editing, Project administration. XW: Data curation, Formal analysis, Methodology, Writing – original draft, Writing – review & editing, Project administration. W-MC: Software, Supervision, Project administration, Validation, Writing – review & editing. L-DL: Project administration, Software, Supervision, Validation, Writing – review & editing. YH: Project administration, Software, Supervision, Validation, Writing – review & editing. J-YL: Methodology, Project administration, Supervision, Validation, Writing – review & editing. D-HX: Methodology, Project administration, Supervision, Validation, Writing – review & editing. Z-YL: Methodology, Project administration, Supervision, Validation, Writing – review & editing. Y-YL: Project administration, Resources, Validation, Visualization, Writing – review & editing. HJ: Project administration, Resources, Validation, Visualization, Writing – review & editing. QJ: Project administration, Resources, Validation, Visualization, Writing – review & editing. Y-ZQ: Conceptualization, Funding acquisition, Investigation, Project administration, Resources, Visualization, Writing – review & editing, Formal analysis.

Funding

The author(s) declared that financial support was received for this work and/or its publication. This study was supported by National Natural Science Foundation of China, Grant Number: 82370160.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that Generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fmed.2025.1701539/full#supplementary-material

References

1. Goyama, S, Yamamoto, G, Shimabe, M, Sato, T, Ichikawa, M, Ogawa, S, et al. Evi-1 is a critical regulator for hematopoietic stem cells and transformed leukemic cells. Cell Stem Cell. (2008) 3:207–20. doi: 10.1016/j.stem.2008.06.002,

PubMed Abstract | Crossref Full Text | Google Scholar

2. Mucenski, ML, Taylor, BA, Ihle, JN, Hartley, JW, Morse, HC 3rd, Jenkins, NA, et al. Identification of a common ecotropic viral integration site, Evi-1, in the DNA of AKXD murine myeloid tumors. Mol Cell Biol. (1988) 8:301–8. doi: 10.1128/mcb.8.1.301-308.1988,

PubMed Abstract | Crossref Full Text | Google Scholar

3. Ogawa, S, Mitani, K, Kurokawa, M, Matsuo, Y, Minowada, J, Inazawa, J, et al. Abnormal expression of Evi-1 gene in human leukemias. Hum Cell. (1996) 9:323–32.

Google Scholar

4. Hinai, AA, and Valk, PJ. Review: aberrant EVI1 expression in acute myeloid leukaemia. Br J Haematol. (2016) 172:870–8. doi: 10.1111/bjh.13898

Crossref Full Text | Google Scholar

5. Grimwade, D, Hills, RK, Moorman, AV, Walker, H, Chatters, S, Goldstone, AH, et al. Refinement of cytogenetic classification in acute myeloid leukemia: determination of prognostic significance of rare recurring chromosomal abnormalities among 5876 younger adult patients treated in the United Kingdom Medical Research Council trials. Blood. (2010) 116:354–65. doi: 10.1182/blood-2009-11-254441,

PubMed Abstract | Crossref Full Text | Google Scholar

6. Cui, W, Sun, J, Cotta, CV, Medeiros, LJ, and Lin, P. Myelodysplastic syndrome with inv(3)(q21q26.2) or t(3;3)(q21;q26.2) has a high risk for progression to acute myeloid leukemia. Am J Clin Pathol. (2011) 136:282–8. doi: 10.1309/AJCP48AJDCKTHUXC

Crossref Full Text | Google Scholar

7. Akiyama, H, Kantarjian, H, Jabbour, E, Issa, G, Haddad, FG, Short, NJ, et al. Outcome of 3q26.2/MECOM rearrangements in chronic myeloid leukemia. Int J Hematol. (2024) 120:203–11. doi: 10.1007/s12185-024-03787-z,

PubMed Abstract | Crossref Full Text | Google Scholar

8. Lugthart, S, Gröschel, S, Beverloo, HB, Kayser, S, Valk, PJM, van Zelderen-Bhola, SL, et al. Clinical, molecular, and prognostic significance of WHO type inv(3)(q21q26.2)/t(3;3)(q21;q26.2) and various other 3q abnormalities in acute myeloid leukemia. J Clin Oncol. (2010) 28:3890–8. doi: 10.1200/JCO.2010.29.2771

Crossref Full Text | Google Scholar

9. van Barjesteh Waalwijk Doorn-Khosroi, S, Erpelinck, C, van Putten, WL, Valk, PJ, van der Poel-van de Luytgaarde, S, Hack, R, et al. High EVI1 expression predicts poor survival in acute myeloid leukemia: a study of 319 de novo AML patients. Blood. (2003) 101:837–45. doi: 10.1182/blood-2002-05-1459

Crossref Full Text | Google Scholar

10. Qin, YZ, Zhao, T, Zhu, HH, Wang, J, Jia, JS, Lai, YY, et al. High EVI1 expression predicts poor outcomes in adult acute myeloid leukemia patients with intermediate cytogenetic risk receiving chemotherapy. Med Sci Monit. (2018) 24:758–67. doi: 10.12659/MSM.905903,

PubMed Abstract | Crossref Full Text | Google Scholar

11. Liu, D, Chen, S, Pan, J, Zhu, M, Wu, N, Zhu, F, et al. Acquired EVI1 rearrangement involved in the transformation from 5q- syndrome to pre-B lymphocytic leukemia in a Chinese patient. Int J Hematol. (2012) 96:806–9. doi: 10.1007/s12185-012-1185-8,

PubMed Abstract | Crossref Full Text | Google Scholar

12. Wang, L, Tang, J, Feng, J, Huang, Y, Cheng, Y, Xu, H, et al. Case report: a rare case of coexisting Waldenstrom Macroglobulinemia and B-cell acute lymphoblastic leukemia with KMT2D and MECOM mutations. Front Immunol. (2022) 13:1001482. doi: 10.3389/fimmu.2022.1001482,

PubMed Abstract | Crossref Full Text | Google Scholar

13. Stevens, A, Hanson, D, de Leonibus, C, Whatmore, A, Donn, R, White, DJ, et al. EVI1 expression in childhood acute lymphoblastic leukaemia is not restricted to MLL and BCR/ABL rearrangements and is influenced by age. Blood Cancer J. (2014) 4:e179. doi: 10.1038/bcj.2013.76,

PubMed Abstract | Crossref Full Text | Google Scholar

14. Abbal, C, Abdelali, RB, Lafage, M, Ben Abdelali, R, Huguet, F, Leguay, T, et al. Value of EVI1 gene expression level in adult acute lymphoblastic leukemia (ALL): a study from the group for research on adult ALL (GRAALL). Blood. (2014) 124:1081–1. doi: 10.1182/blood.V124.21.1081.1081

Crossref Full Text | Google Scholar

15. Jia, M, Hu, BF, Zhang, JY, Xu, LY, and Tang, YM. Clinical features and prognostic implications of ecotropic viral integration site 1 (EVI1) in childhood acute lymphoblastic leukemia. Pediatr Hematol Oncol. (2023) 40:371–81. doi: 10.1080/08880018.2022.2117881,

PubMed Abstract | Crossref Full Text | Google Scholar

16. Nabil, R, Abdellateif, MS, Gamal, H, Hassan, NM, Badawy, RH, Ghareeb, M, et al. Clinical significance of EVI-1 gene expression and aberrations in patient with de-novo acute myeloid and acute lymphoid leukemia. Leuk Res. (2023) 126:107019. doi: 10.1016/j.leukres.2023.107019,

PubMed Abstract | Crossref Full Text | Google Scholar

17. Konantz, M, André, MC, Ebinger, M, Grauer, M, Wang, H, Grzywna, S, et al. EVI-1 modulates leukemogenic potential and apoptosis sensitivity in human acute lymphoblastic leukemia. Leukemia. (2013) 27:56–65. doi: 10.1038/leu.2012.211,

PubMed Abstract | Crossref Full Text | Google Scholar

18. Yang, L, Dao, FT, Lu, AD, Chen, WM, Li, LD, Long, LY, et al. Low EVI1 expression at diagnosis predicted poor outcomes in pediatric Ph-negative B cell precursor acute lymphoblastic leukemia patients. Pediatr Hematol Oncol. (2022) 39:97–107. doi: 10.1080/08880018.2021.1939818,

PubMed Abstract | Crossref Full Text | Google Scholar

19. Gökbuget, N, Boissel, N, Chiaretti, S, Dombret, H, Doubek, M, Fielding, A, et al. Diagnosis, prognostic factors, and assessment of ALL in adults: 2024 ELN recommendations from a European expert panel. Blood. (2024) 143:1891–902. doi: 10.1182/blood.2023020794,

PubMed Abstract | Crossref Full Text | Google Scholar

20. Shah, B, Mattison, RJ, Abboud, R, Abdelmessieh, P, Aldoss, I, Burke, PW, et al. Acute lymphoblastic leukemia, version 2.2024, NCCN clinical practice guidelines in oncology. J Natl Compr Cancer Netw. (2024) 22:563–76. doi: 10.6004/jnccn.2024.0051,

PubMed Abstract | Crossref Full Text | Google Scholar

21. Yang, S, Wang, J, Zhao, T, Jia, JS, Zhu, HH, Jiang, H, et al. CD20 expression sub-stratifies standard-risk patients with B cell precursor acute lymphoblastic leukemia. Oncotarget. (2017) 8:105397–406. doi: 10.18632/oncotarget.22207,

PubMed Abstract | Crossref Full Text | Google Scholar

22. Li, Z, Lai, Y, Zhang, X, Xu, L, Liu, K, Wang, Y, et al. Monosomal karyotype is associated with poor outcomes in patients with Philadelphia chromosome-negative acute lymphoblastic leukemia receiving chemotherapy but not allogeneic hematopoietic stem cell transplantation. Ann Hematol. (2020) 99:1833–43. doi: 10.1007/s00277-020-04155-7,

PubMed Abstract | Crossref Full Text | Google Scholar

23. Zhang, XH, Chen, J, Han, MZ, Huang, H, Jiang, EL, Jiang, M, et al. The consensus from the Chinese Society of Hematology on indications, conditioning regimens and donor selection for allogeneic hematopoietic stem cell transplantation: 2021 update. J Hematol Oncol. (2021) 14:145. doi: 10.1186/s13045-021-01159-2,

PubMed Abstract | Crossref Full Text | Google Scholar

24. Caye, A, Beldjord, K, Mass-Malo, K, Drunat, S, Soulier, J, Gandemer, V, et al. Breakpoint-specific multiplex polymerase chain reaction allows the detection of IKZF1 intragenic deletions and minimal residual disease monitoring in B-cell precursor acute lymphoblastic leukemia. Haematologica. (2013) 98:597–601. doi: 10.3324/haematol.2012.073965,

PubMed Abstract | Crossref Full Text | Google Scholar

25. Qin, YZ, Jiang, Q, Xu, LP, Wang, Y, Jiang, H, Dao, FT, et al. The prognostic significance of ZNF384 fusions in adult Ph-negative B-cell precursor acute lymphoblastic leukemia: a comprehensive cohort study from a single Chinese center. Front Oncol. (2021) 11:632532. doi: 10.3389/fonc.2021.632532,

PubMed Abstract | Crossref Full Text | Google Scholar

26. Beillard, E, Pallisgaard, N, van der Velden, VH, Bi, W, Dee, R, van der Schoot, E, et al. Evaluation of candidate control genes for diagnosis and residual disease detection in leukemic patients using 'real-time' quantitative reverse-transcriptase polymerase chain reaction (RQ-PCR) - a Europe against cancer program. Leukemia. (2003) 17:2474–86. doi: 10.1038/sj.leu.2403136

Crossref Full Text | Google Scholar

27. Moorman, AV, Harrison, CJ, Buck, GA, Richards, SM, Secker-Walker, LM, Martineau, M, et al. Karyotype is an independent prognostic factor in adult acute lymphoblastic leukemia (ALL): analysis of cytogenetic data from patients treated on the Medical Research Council (MRC) UKALLXII/eastern cooperative oncology group (ECOG) 2993 trial. Blood. (2007) 109:3189–97. doi: 10.1182/blood-2006-10-051912,

PubMed Abstract | Crossref Full Text | Google Scholar

28. Sun, K, Wang, J, Chen, WM, Xu, N, Long, LY, Wang, X, et al. Clinical features and outcomes of fusion gene defined adult Ph-negative B-cell precursor acute lymphoblastic leukemia patients: a single institutional report. Biomol Biomed. (2023) 23:298–309. doi: 10.17305/bjbms.2022.7851,

PubMed Abstract | Crossref Full Text | Google Scholar

29. Lugthart, S, van Drunen, E, van Norden, Y, van Hoven, A, Erpelinck, CAJ, Valk, PJM, et al. High EVI1 levels predict adverse outcome in acute myeloid leukemia: prevalence of EVI1 overexpression and chromosome 3q26 abnormalities underestimated. Blood. (2008) 111:4329–37. doi: 10.1182/blood-2007-10-119230,

PubMed Abstract | Crossref Full Text | Google Scholar

30. Balgobind, BV, Lugthart, S, Hollink, IH, Arentsen-Peters, STJCM, van Wering, ER, de Graaf, SSN, et al. EVI1 overexpression in distinct subtypes of pediatric acute myeloid leukemia. Leukemia. (2010) 24:942–9. doi: 10.1038/leu.2010.47,

PubMed Abstract | Crossref Full Text | Google Scholar

31. Hata, A, and Chen, YG. TGF-β signaling from receptors to Smads. Cold Spring Harb Perspect Biol. (2016) 8:2061. doi: 10.1101/cshperspect.a022061,

PubMed Abstract | Crossref Full Text | Google Scholar

32. Ayoub, E, Wilson, MP, McGrath, KE, Li, AJ, Frisch, BJ, Palis, J, et al. EVI1 overexpression reprograms hematopoiesis via upregulation of Spi1 transcription. Nat Commun. (2018) 9:4239. doi: 10.1038/s41467-018-06208-y,

PubMed Abstract | Crossref Full Text | Google Scholar

33. Morishita, K, Parganas, E, William, CL, Whittaker, MH, Drabkin, H, Oval, J, et al. Activation of EVI1 gene expression in human acute myelogenous leukemias by translocations spanning 300-400 kilobases on chromosome band 3q26. Proc Natl Acad Sci USA. (1992) 89:3937–41. doi: 10.1073/pnas.89.9.3937,

PubMed Abstract | Crossref Full Text | Google Scholar

34. Yamazaki, H, Suzuki, M, Otsuki, A, Shimizu, R, Bresnick, EH, Engel, JD, et al. A remote GATA2 hematopoietic enhancer drives leukemogenesis in inv(3)(q21;q26) by activating EVI1 expression. Cancer Cell. (2014) 25:415–27. doi: 10.1016/j.ccr.2014.02.008,

PubMed Abstract | Crossref Full Text | Google Scholar

35. Langabeer, SE, Rogers, JR, Harrison, G, Wheatley, K, Walker, H, Bain, BJ, et al. EVI1 expression in acute myeloid leukaemia. Br J Haematol. (2001) 112:208–11. doi: 10.1046/j.1365-2141.2001.02569.x

Crossref Full Text | Google Scholar

36. Liu, XX, Pan, XA, Gao, MG, Kong, J, Jiang, H, Chang, YJ, et al. The adverse impact of ecotropic viral integration site-1 (EVI1) overexpression on the prognosis of acute myeloid leukemia with KMT2A gene rearrangement in different risk stratification subtypes. Int J Lab Hematol. (2023) 45:195–203. doi: 10.1111/ijlh.13987,

PubMed Abstract | Crossref Full Text | Google Scholar

37. Li, JF, Dai, YT, Lilljebjörn, H, Shen, SH, Cui, BW, Bai, L, et al. Transcriptional landscape of B cell precursor acute lymphoblastic leukemia based on an international study of 1,223 cases. Proc Natl Acad Sci USA. (2018) 115:E11711–e11720. doi: 10.1073/pnas.1814397115,

PubMed Abstract | Crossref Full Text | Google Scholar

38. Yasuda, T, Nishijima, D, Kojima, S, Kawazu, M, Ueno, T, Tsuzuki, S, et al. Genomic and clinical characterization of adult Ph-negative B-cell acute lymphoblastic leukemia. Blood. (2018) 132:2821–1. doi: 10.1182/blood-2018-99-115716

Crossref Full Text | Google Scholar

39. Hsu, YC, Yu, CH, Chen, YM, Roberts, KG, Ni, YL, Lin, KH, et al. Philadelphia chromosome-negative B-cell acute lymphoblastic leukaemia with kinase fusions in Taiwan. Sci Rep. (2021) 11:5802. doi: 10.1038/s41598-021-85213-6,

PubMed Abstract | Crossref Full Text | Google Scholar