Gene Deletions and Prognostic Values in B-Linage Acute Lymphoblastic Leukemia

Although pediatric-like treatment regimen has remarkably improved the survival rates of adults with acute lymphoblastic leukemia (ALL), the outcome of some adult patients is still poor owing to adverse genetic features. These molecular abnormalities, especially gene deletions, may be considered for the prognosis assessment for adult patients with ALL. In this study, using multiplex ligation-dependent probe amplification (MLPA) method, gene deletions were analyzed in from 211 adult B-ALL patients treated in our center. The data showed that 68.2% (144/211) adult B-ALL patients carried gene deletions, and the frequency is much higher in Ph+B-ALL patients. IKZF1 gene deletion is the most common gene deletion in adult B-ALL, followed by CDKN2A/B deletion. In Ph-B-ALL patients, the overall survival of patients with gene deletions is inferior to that of patients without any gene deletions. More obviously, patients with IKZF1 or CDKN2A/B deletion had a worse prognosis, whereas, allogeneic hematopoietic stem cell transplantation could improve OS in patients with IKZF1 deletion, but not in patients with CDKN2A/B deletion. Moreover, the outcome of Ph-B-ALL patients with double deletion of IKZF1and CDKN2A/B may be much worse than that of patients with IKZF1 or CDKN2A/B alone. Minimal residual disease (MRD) was also analyzed together with gene deletions and demonstrated that gene deletions have a negative impact on survival only in MRD positive Ph-B-ALL patients. In conclusion, gene deletions are closely related with the prognosis of adult Ph-B-ALL patients.


INTRODUCTION
Acute lymphoblastic leukemia (ALL) is a malignant transformation and proliferation of lymphoid progenitor cells in the bone marrow, blood and extramedullary sites, with more than 75% originating from B-linage (B-ALL) (1). Recent years have witnessed a significant increase of >90% in survival rates in pediatric ALL (2); however, the long-term outcome of adult ALL is much inferior to that of pediatric B-ALL, including many lowrisk patients with genetic heterogeneities and varied intensity of the treatments (3). Moreover, unlike adult acute myeloid leukemia, there is a lack of studies on adult ALL's molecular prognostic factors and a lack of consensus on molecular abnormalities in adult ALL risk stratification. Hence, the discovery of robust and assertive prognostic indicators is an urgent requirement for personalized clinical decisions.
Thus, the role of gene deletions in hematological malignancies and their prognostic values have garnered increasing research interest. A large number of studies have focused on extrapolating well-established molecular signatures in the prognostic assessment of pediatric ALL to adult ALL. For example, IKZF1 deletion, a hallmark of high-risk pediatric ALL, is related to poor prognosis in adult ALL (4)(5)(6)(7)(8)(9)(10). This phenomenon was consistent with our previous study, wherein IKZF1 deletion was associated with poor survival in adult B-ALL patients (11,12).
Stanulla et al. (13) found that IKZF1 deletion in patients those without ERG deletion was often accompanied by other gene deletions (CDKN2A/2B, PAX5, and PAR1), which was defined as IKZF1 plus gene abnormality. The study also demonstrated that IKZF1 plus presents a poor prognostic profile in pediatric B-ALL. However, large discrepancies are noted with respect to concomitant abnormalities of IKZF1 deletion and the prognostic value in different reports (14)(15)(16)(17)(18)(19), and whether IKZF1 plus achieves an equally superior prognostic performance in all B-ALL patients remains unknown.
In this study, multiplex ligation-dependent probe amplification (MLPA) method was used to detect 13 gene deletions in 211 adult B-ALL patients who were diagnosed and treated in our center from October 2009 to December 2018. The characteristics of patients with different gene deletions and gene deletions' prognostic impact stratified by disease subgroups were analyzed.
In the current study, we revealed the gene deletion frequency of IKZF1, CDKN2A/B, I&C and other genes in Chinese adult B-ALL patients using MLPA, and established gene deletions as prognostic indicators for these patients across various subgroups. These findings shed light on the roles of gene deletions in the initiation and progression of B-ALL, highlight the importance of shifting from individual risk factors to integrated prognostic profiles, and open an avenue towards the development of personalized treatment strategies for ALL.

Patients and Samples
The present study included 211 patients with newly diagnosed B-ALL in a prospective single arm clinical trial (ChiCTR-TNC-09000397) (20) at the Leukemia Center in the Institute of Hematology and Blood Diseases Hospital (Tianjin, China) between October 2009 and December 2018. All the experiments were performed in accordance with the Helsinki Declaration, and the study was approved by the institutional ethics committee. Informed consent was obtained from all patients.
All 211 patients were diagnosed by MICM mode, which included morphology, immunology, and molecular and genetic analysis. The median percentage of bone marrow leukemia blasts was 88.5% (range: 22.5-99.5%; Ph -B-ALL: 89%, 22.5-95%, Ph + B-ALL: 86%, 23.5-99.5%, 2 patients with blasts percentage: 20-30%). Patients with Ph -B-ALL were treated with the therapy schedule, according to the trial protocol. The induction regimen is VDCLP, and after reaching complete remission (CR), all patients received early intensive, late consolidation, and maintenance treatment. Ph + B-ALL patients were treated in the same manner with L-asparaginase omitted and a TKI (Imatinib or Dasatinib) added from the 8th day of induction chemotherapy. Among 211 patients, the CR rate was 95.7% (202/211). Nine (4.3%, Ph -B-ALL:7, Ph + B-ALL:2) patients did not achieve CR: 4 (1.9%) died of complications in induction therapy, 3 died of disease without remission, and data could not be evaluated in 2 patients; 8 patients harbored gene deletions, and 1 patient did not carry gene deletions (Supplementary Data 1). The data of minimal residual disease (MRD, detected using flow cytometry) after induction treatment and at 3 months postinitial treatment were obtained in 172 patients. The cutoff of flow cytometry result was 10 -4 . Patients with flow cytometry result <10 -4 at both test points were considered as MRD-negative. Patients with one or both results ≥10 -4 were considered as MRD-positive. High risk patients (according to NCCN risk stratification criteria), and patients with MRD positive at any test points were recommended to receive allo-HSCT. The time-point of HSCT was after 2 cycles of consolidation therapy.

MLPA
Bone marrow mononuclear cells of the patients were collected at the time of diagnosis, and DNA was extracted using the QIAamp DNA Blood Mini kit (catalog #51104; Qiagen GmbH, Hilden, Germany), according to the manufacturer's protocols. The MLPA assay was conducted using the SALSA MLPA P335 ALL-IKZF1 Probe Mix Kit (MRC Holland, Amsterdam, The Netherlands) to detect the deletion of genes related to ALL pathogenesis, according to the manufacturer's protocol (11). This assay contains 13 genes, including part or whole exons of IKZF1, EBF1, CDKN2A/B, PAX5, ETV6, BTG1, RB1, SHOX-AREA, CRLF2, CSF2RA, IL3RA and P2RY8 genes. The products were subjected to fragment analysis on an ABI-3730 genetic analyzer (Applied Biosystems, Carlsbad, CA, USA), and the MLPA data were analyzed using Coffalyser ® software.

Statistical Analysis
All statistical analyses were performed using SPSS 21.0 software (IBM Co., Armonk, NY, USA). The overall survival (OS) was defined as the time from diagnosis to the time of death (from any cause) or last follow-up, and the relapse-free survival (RFS) was defined as the time from achieving CR to relapse or death or the date of the last follow-up in CR. OS and RFS were evaluated by Kaplan-Meier analysis and compared by the log-rank test. Cox proportional hazard regression models were used to identify the prognostic factors. The values of the specific deleted gene were adjusted by age, white blood cell (WBC) count, MRD, and the allogeneic hematopoietic stem cell transplantation (allo-HSCT) and included in the model as a time-varying covariate. The other comparisons were performed using the chi-square or Fisher's exact test. P<0.05 indicated statistical significance.
48/82 patients with IKZF1 deletion underwent HSCT, while the remaining 34 patients did not receive HSCT. MRD results were obtained in 66 patients with IKZF1 deletion, 36 patients were MRD-negative, and 30 were MRD-positive. Patients with IKZF1 deletion showed higher MRD positive rate than IKZF1 wild-type patients (P=0.043) ( Table 2 and Supplementary Table 1).

IKZF1 Deletion Companied With Other Gene Deletions
IKZF1 PLUS was defined by Stanulla et al. (13) as IKZF1 deletion accompanied by at least one gene deletion in CDKN2A/B, PAX5, PAR1 in the absence of ERG deletion in pediatric B-ALL. In the current study, we did not assess ERG deletion and P2RY8-CRLF2 fusion in the whole cohort, thus we analyzed patients who carried IKZF1 deletion in combination with any of the other recurrent deletions (IKZF1+other deletion) instead of IKZF1 PLUS . 45/211 (21.3%, Ph + B-ALL: 26; Ph -B-ALL: 19) patients had this kind deletion. The specific types of IKZF1+other deletion were as follows: 36 cases of IKZF1 deletion combined with CDKN2A/B deletion (Ph + B-ALL: 19; Ph -B-ALL: 17), including 22 cases of single CDKN2A/B deletion. Among the 45 cases, 22 patients presented IKZF1 and PAX5 deletion simultaneously, 8 patients had single PAX5 deletion; 4 cases had IKZF1 and PAR1 deletion, and 1 case had PAR1 alone. Furthermore, 14 IKZF1 deletion cases also showed CDKN2A/B and PAX5 deletion together ( Figure 1B). IKZF1+other deletions rate in Ph + B-ALL was higher than that in Ph -B-ALL (30.6% vs. 14.8%, P<0.01). Of the 45 patients with IKZF1 +other deletions, 23 cases underwent HSCT, and 22 cases did not receive HSCT ( Table 2). A total of 36 patients with IKZF1+other deletions obtained MRD results (Supplementary Table 2).

CDKN2A/B Deletion Had Crucial Prognostic Significance in Ph -B-ALL
In Ph -B-ALL patients, 53/126 (42.1%) cases carried CDKN2A/B deletion; the survival of these patients was inferior to those without CDKN2A/B deletion (N=73). The 2-year OS: 54.7 ± 6.8% vs. 75.2 ± 5.1%, P=0.001; 2-year RFS: 39.8 ± 6.9% vs. 59.9 ± 6.9%, P=0.012 (Figures 4A, B). In the non-HSCT group, CDKN2A/B deletion patients (N=27) vs. no CDKN2A/B deletion patients (N=49) showed that the 2-year OS was 32.   Figures S3A, B). The prognosis of patients with I&C deletion or single CDKN2A/B deletion was worse than that of patients without any gene deletions in Ph-B-ALL. The data are shown in Supplementary Data 3.  Figures S5A, B). In MRD-negative Ph -B-ALL patients, the presence or absence of gene deletion had no significant effect on the prognosis (Supplementary Figure  S5C, D).
No obvious differences were noted in the survival with respect to PAX5, ETV6, BTG1, EBF1, or RB1 deletions in both Ph -B-ALL patients.

Prognosis of Gene Deletions in Ph + B-ALL
In the Ph + B-ALL cohort, the 2-year OS was 57.1 ± 5.4%, and the 2-year RFS was 47.8 ± 5.5%. Next, 70 patients had gene deletions, 15 patients had no gene deletions, and no statistical difference was observed in the survival between the two groups. In Ph + B-ALL patients, no clear prognostic significance of IKZF1, CDKN2A/B, I&C, PAX5, ETV6, BTG1, EBF1, or RB1 deletion was recorded. MRD results were obtained in 72 Ph + B-ALL patients (MRD-negative: 48, MRD-positive: 24), and no significant prognostic significance was found in MRD subgroup analysis ( Table 3).

COX Regression Analysis
The results of multivariate COX regression analysis were consistent with the univariate analysis. In Ph -B-ALL, the gene deletions constituted an independent prognostic factor after adjusted for age, WBC count and HSCT. IKZF1 deletion in Ph -B-ALL patients had an independent prognostic significance in the non-HSCT group. CDKN2A/B showed independent prognostic significance in the overall Ph -B-ALL cohort: HSCT and non-HSCT groups. I&C deletion also had independent prognostic significance in the overall Ph -B-ALL cohort and non-HSCT group ( Table 4). In addition, we also calculated MRD as the covariate in the COX regression, and found only CDKN2A/B deletion was the independent marker of OS and RFS after adjusting age, WB C count and MRD status (Supplementary Table 2). No independent prognostic factors were detected in Ph + B-ALL cohort.

DISCUSSION
Despite overlapping morphological and immunophenotypic characteristics, adult and pediatric B-ALL have distinct molecular drivers and divergent treatment outcomes. Gene deletions have been closely associated with the prognosis of B-ALL with a primary focus on pediatric B-ALL (5,(21)(22)(23). However, the prognostic significance of gene deletions in Chinese adult B-ALL has not yet been well-established.
Herein, we used the MLPA method to detect gene deletions of 13 genes in 211 adult B-ALL patients. The results showed a high frequency of IKZF1, CDKN2A/B, and PAX5 deletions in adult B-ALL patients. The incidence of IKZF1 gene deletion in Ph + B-ALL patients (65.9%) was significantly higher than that in the Ph -B-ALL group (20.3%). In the current study, the incidence of IKZF1+other deletions is 21.3%, which is significantly higher than IKZF1 PLUS that in pediatric BCP-ALL (6%) (13).
The frequency of CDKN2A/B deletion was significantly higher in the Ph -B-ALL group (39.7%) than in the Ph + B-ALL group (24.7%). CDKN2A/B deletion occurred in about 36-45% pediatric B-ALL patients (10,24,25), which is similar to that in adult B-ALL.
In the prognostic analysis of Ph -B-ALL patients, the prognosis of patients with gene deletions was significantly worse than in patients without any gene deletion. IKZF1 is one of the most common deleted gene in B-ALL (14), and its prognostic significance is controversial. Dorge et al. (15)(16)(17) considered that IKZF1 gene deletion has a significant impact on the prognosis of pediatric B-ALL. It can be incorporated in the risk stratification system and defined as the high-risk group. However, Stanulla et al. (13) and other studies (18,19) suggested that although the prognosis of pediatric B-ALL with IKZF1 deletion was poor, the OS of pediatric B-ALL with IKZF1 deletion was satisfactory, and hence, its prognostic significance was limited. Zhang et al. (26) conducted a meta-analysis on eight national and international studies about the prognosis of IKZF1 gene deletion in adult B-ALL and speculated that IKZF1 gene deletion indicates poor prognosis; however, this prognostic significance only exists in patients with Ph -B-ALL. Moreover, as Messina et al. reported, the outcome of IKZF1 deletion did not differ from that of IKZF1-WT patients both in pediatric and adult B-NEG ALL (B-ALL patients negative for the recurrent fusion transcripts, BCR-ABL, ETV6-RUNX1, TCF3-PBX1, KMT2A-rearrangement) (27). In the current analysis, which including 126 Ph -B-ALL patients, IKZF1 deletion was found to be related to poor OS in the group of patients who did not receive HSCT (2-year OS: 17.9 ± 11% vs. 62.6 ± 6.3%, P=0.037). Consequently, patients with IKZF1 deletion would be recommended to undergo HSCT, which demonstrated superior survival benefit compared to chemotherapy alone.
The prognostic significance of IKZF1 deletion in Ph + B-ALL was also not consistent: among the 83 adult Ph + B-ALL cases reported by Martinelli et al. (5), DFS was shorter in patients with IKZF1 deletion compared to those without IKZF1 deletion. However, in another study by Yao et al. on 43 Ph + B-ALL patients (22), those with IKZF1 deletion did not show any significant differences in non-relapse mortality, leukemia-free survival, or survival with patients without the deletion. In report of Pfeifer et al. (10), the prognostic significance of IKZF1 was weaker than CDKN2A/B, and IKZF1 deletion alone actually had a significant lower risk for relapse than with other deletions. In the current study, IKZF1 deletion did not show any prognostic significance in Ph + B-ALL patients.
The prognostic significance of CDKN2A/B deletion is controversial in pediatric B-ALL (24,25,28). In the study by Mirebeau et al. (29), CDKN2A/B deletions coincided with other HR factors in pediatric B-ALL but did not affect prognosis. However, Heyman et al. (30) found that CDKN2A/B was significantly associated with poor prognosis in 79 pediatric B-ALL patients. In study of Messina et al., CDKN2A/B deletion had impact of DFS in only adult cohort of B-NEG ALL (26). Owing to its high deletion incidence in adult B-ALL, we also analyzed the prognosis of patients with CDKN2A/B gene deletion and found that the deletion was an independent poor prognostic indicator in the Ph -B-ALL cohort, regardless of HSCT, and the single CDKN2A/B deletion was also associated with poor prognosis in each subgroup of Ph -B-ALL. In another study by Pfeifer et al. (10), CDKN2A/B deletion was detected in 45% Ph + B-ALL cases and indicative poor OS and DFS with a short remission duration. Nonetheless, in this study, the proportion of CDKN2A/B deletion in Ph + B-ALL was low (28.2%), and no statistical prognostic significance was detected in any subgroup. Stanulla et al. (13) defined IKZF1 plus and proposed it as an MDR-dependent, poor prognostic profile in pediatric B-ALL. In the current study, compared to PAX5 and PAR1, IKZF1 deletion occurred more frequently with CDKN2A/B deletion in adult B-ALL patients, especially in Ph -B-ALL patients. In this cohort, despite the small population size, patients with simultaneous deletion of I&C had worse OS and RFS than those without the deletion. Hence, we speculated that the prognostic significance of I&C in adult Ph -B-ALL is similar to that of IKZF1 plus in pediatric B-ALL, and hence, in this study, the prognostic significance of IKZF1+other gene deletions was not analyzed separately. In future studies, we would need to expand the sample size to substantiate these findings. In Ph + B-ALL, the prognostic significance of I&C deletions was not detected.
MRD is a critical index to evaluate the prognosis of ALL patients (31)(32)(33)(34)(35)(36). We analyzed the prognostic significance of gene deletions stratified by MRD status. In Ph -B-ALL patients with negative MRD, the gene deletion had no obvious effect on survival. In 38 MRD-positive Ph -B-ALL patients, those with gene deletions had worse OS and RFS than those without any gene deletions, indicating that gene deletions affected the prognosis of MRD-positive Ph -B-ALL patients. We also found that the incidence of CDKN2A/B deletion was significantly higher (52.6% vs. 33.9%, P=0.064) in these patients, and it was a poor prognostic factor in MRD-positive Ph -B-ALL, regardless of HSCT. However, our data demonstrated that MRD-positive patients without CDKN2A/B deletion could benefit from allo-HSCT, as the positive MRD is an adverse risk factor. On the other hand, the MRD-negative Ph -B-ALL cohort did not show any survival differences in different gene deletion groups, indicating that patients carrying CDKN2A/B deletion with a negative MRD could be potentially spared from HSCT. Taken together, these data suggested an MRD-dependent prognostic role of CDKN2A/B in allo-HSCT decision for Ph -B-ALL patients. In Ph + B-ALL patients, no significant effect of gene deletions was found on survival irrespective of the MRD status. Thus, in Chinese adult Ph + B-ALL, the Ph chromosome, not gene deletions plays a significant role in determining the prognosis of patients.
In conclusion, we proved that gene deletions play a major role in the prognosis of Chinese adult B-ALL patients, and we did the analyses in allotransplanted and MRD subsets in the current study. The overall survival of Ph -B-ALL patients with gene deletion is worse than that of patients without any gene deletion, Ph + B-ALL patients is not affected by gene deletion. In Ph -B-ALL, patients with IKZF1 or CDKN2A/B deletion showed poor prognosis; allo-HSCT could improve the OS in patients with IKZF1 deletion but could not improve the OS in patients with CDKN2A/B deletion. CDKN2A/B deletion also has a negative impact on survival only in MRD positive Ph -B-ALL patients

DATA AVAILABILITY STATEMENT
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary Material.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by the ethics committee of Institute of Hematology and Blood Diseases Hospital, Chinese Academy of Medical Sciences. Written informed consent to participate in this study was provided by the participants' legal guardian/next of kin.