Introduction

B-cell acute lymphoblastic leukaemia (B-ALL) is a malignancy resulting from the transformation of a B-cell lineage progenitor cell [1]. The hallmark of B-ALL cases is the presence of genetic abnormalities, including chromosomal rearrangements, DNA copy number variations (CNV) and sequence mutations [2]. The 5th edition of the WHO classification divides B-ALL entity on the basis of refined diagnostic criteria and emphasis on therapeutically and/or prognostically actionable biomarkers [3]. The recently updated classification delineates a newly identified molecular subtype—B-ALL with BCR::ABL1-like features as a separate entity. It is characterized by a similar gene expression profile to the ALL with BCR::ABL1-fusion, but lacks the BCR::ABL1 fusion gene [4, 5]. It is exclusive of well-known drivers of B-ALL, including BCR::ABL1 fusion, KMT2A rearrangement, ETV6::RUNX1 and TCF3::PBX1 fusions [5]. The prevalence of the BCR::ABL1-like ALL is impacted by the age and ethnicity of distinct cohorts and the identification methodology. The incidence increases with age, with a peak in young adults population [6,7,8]. It is characterized by inferior outcomes due to a high rate of nonresponse to induction therapy, higher relapse risk, lower overall survival rates and the persistence of minimal residual disease (MRD) [9,10,11].

Diverse genetic alterations dysregulating kinase and receptor signaling are the hallmark of the BCR::ABL1-like ALL and can be divided into several classes: (1) alterations activating JAK-STAT pathway signaling (including rearrangements of cytokine receptor-like factor 2 (CRLF2) gene, Janus kinase 2 (JAK2) gene and erythropoietin receptor (EPOR) gene); (2) rearrangements of ABL-class genes (ABL1, ABL2, PDGFRα, PDGFRβ, CSF1R); (3) Ras pathway mutations (KRAS, NRAS, NF1, PTPN11) and other uncommon rearrangements [6, 7, 12].

The underlying molecular changes in the BCR::ABL1-like ALL remain of significant interest due to the possibility of incorporation of targeted therapy with tyrosine kinase inhibitors (TKI) and JAK inhibitors [13, 14]. Several ongoing clinical studies are evaluating the effectiveness of addition of targeted therapy to chemotherapy to improve the prognosis [15]. Current scientific efforts concentrate on the identification of molecular targets, and numerous algorithms have been proposed for the recognition of the BCR::ABL1-like ALL subtype, including targeted fusion testing, tiered algorithms and broad-based testing [16,17,18,19]. Nevertheless, the principal aim of the diagnostic approach is to recognize the underlying genetic feature, since they are determinative for the prognosis and targeted therapy. In smaller, real-world groups with constrained resources, the access to comprehensive sequencing strategy is limited. Hence, in those centers, the testing methods should be tailored.

Herein, we present our experience in the BCR::ABL1-like ALL diagnostics. We applied an integrated algorithm which allowed a cost-effective detection of this entity. The frequency and clinical outcome of BCR::ABL1-like ALL cases were analyzed and compared with the existing literature data, with a particular emphasis on the potential therapeutic options.

Materials and methods

Patients

The study was conducted at the Department of Hematology and Bone Marrow Transplantation of Poznan University of Medical Sciences. Adult patients diagnosed with B-cell ALL treated at our Department in the years 2008–2022 were included (n = 102). Thirty-one patients were excluded from further analysis due to the lack of cytogenetic material or essential clinical data. We performed a retrospective analysis of the clinical data, cytogenetic and molecular characteristics in patients treated in the years 2008–2020 (n = 63). Independently, a prospective analysis of cases diagnosed in the years 2020–2022 was performed (n = 8). This study was conducted in accordance with the Declaration of Helsinki. The study was approved by the Poznań University of Medical Sciences Bioethical Committee (Resolution No. 705/20). 63 patients (88.7%) enrolled in the study were treated with B-ALL protocols according to the Polish Adult Leukemia Group (PALG) guidelines. Remaining patients were treated according to hyper-CVAD protocol.

Methods

The expression of TSLPR (predictive of the rearrangement of the CRLF2) with an anti-TSLP antibody (Invitrogen™, clone 1F11/TSLPR PE) was performed using the 10-color multiparameter flow cytometry method (FCM; BD FacsCanto II Ilyric™) using the strategy of internal negative control.

The karyotype analysis was performed using G banding (GTG). The results were described according to the International System for Human Cytogenetic Nomenclature (ISCN). FISH studies were performed on the interface nuclei using break-apart probes for TCF3::PBX1, CRFL2, JAK2, EPOR, ABL1, ABL2 (Cytocell, Cambridge, UK) and for BCR::ABL1, KMT2A, and PDGFRb (Vysis, IL, USA) and, additionally, for IGH and P2RY8 in the CRLF2 rearranged (CRLF2-r) cases (Cytocell, Cambridge, UK). At least 200 interphase nuclei were scored for each probe by two independent examiners. The cut-off threshold for the BCR::ABL1-like FISH probes of > 10% of cells was established.

The analysis of the JAK2 exon 16 sequence was conducted using DNA extracted from whole-blood leukocytes at the time of diagnosis QIAmp DNA Mini Kit (Qiagen) and high resolution melt analysis (HRMA). For the variant type identification screened by HRMA, Sanger sequencing was applied using the BigDye Terminator v3.1 Cycle Sequencing kit (Applied Biosystems, Thermo Fisher Scientific) and the following primers—forward: 5ʹ-TGCTCCAGTACTTGTGGACTGA-3ʹ and reverse: 5ʹ-CCACTGCCCAAGTAAAGCTTAG-3ʹ.

Diagnostic algorithm

For the identification of BCR::ABL1-like ALL cases, we implemented a stepwise algorithm integrating all the above-mentioned techniques (Fig. 1).

Fig. 1
figure 1

A stepwise algorithm integrating multicolor flow cytometry and fluorescent in situ hybridization implemented in the study

Full size image

Firstly, the presence of the recurring cytogenetic features was evaluated. Patients with recurring cytogenetic lesions were excluded from screening for BCR::ABL1-like ALL. The expression of TSLPR was evaluated by FCM (in a prospective analysis only). Patients expressing TSLPR on leukemic blasts were enrolled for the FISH analysis with a CRLF2 break-apart probe. Patients lacking the TSLPR expression were recognized as non-CRLF2-rearranged (non-CRLF2-r) and subsequently proceeded to the analysis with remaining FISH probes (JAK2, EPOR, ABL1, ABL2, PDGFRb). In a case of retrospective analysis, the patients were primarily examined for the presence of CRLF2 rearrangement with a FISH probe. In CRLF2-r cases, the next step included an analysis with IGH and P2RY8 FISH probes to identify the fusion gene. Non-CRLF2-r cases proceeded to the analysis with the remaining FISH probes (JAK2, EPOR, ABL1, ABL2, PDGFRb).

Additionally, all patients with CRLF2-r B-ALL with available DNA were enrolled in the analysis of the JAK2 exon 16 mutational status.

Results

The median age of the patients at the time of initial diagnosis was 40 (range 18–69 years). Most of the individuals were diagnosed with the B-common phenotype (n = 50). 32 patients from the study group were recognized as B-ALL with recurring cytogenetic abnormalities. The remaining patients (n = 39) were screened for BCR::ABL1-like features. Out of the prospectively analyzed subjects (n = 8), we revealed high expression of CRLF2 in 3 cases in FCM. Second step analysis with FISH revealed CRLF2::IGH fusion in all 3 patients. In a retrospectively analyzed group, we revealed CRLF2-r in one patient and ABL-class genes rearrangements in 2 patients.

Interestingly, one of the patients was enrolled in the study due to a relapse of B-ALL within 13 years after the treatment with chemotherapy and an allogeneic hematopoietic stem cell transplantation (alloHSCT) in the first complete remission (CR1). Notably, we performed a retrospective analysis on the basis of cytogenetic material obtained at the time of the initial diagnosis, however, the rearrangement of CRLF2 was absent. The results of the cytogenetic analysis in the relapsed case is presented in Fig. 2.

Fig. 2
figure 2

The results of a diagnostic work-up of a patient with CRLF2-rearranged BCR::ABL1-like ALL which occurred during a relapse after a prolonged remission despite the absence of CRLF2 rearrangement at the initial diagnosis. Top left side: FISH analysis with CRLF2 break-apart probe (CytoCell®) on leukemic blasts at the initial diagnosis. In the normal cell, 2 fused red/green signals (2 R/G) or 2 yellow signals (2Y) are observed. Top right side: FISH analysis with CRLF2 break-apart probe (CytoCell®) on leukemic blasts at the relapse after prolonged remission (13 years). A translocation resulting in 1R, 1G, 1R/G. Bottom: Second step analysis with IGH probe (CytoCell®). In a normal cell, 2 fused red/green signals (2 R/G) or 2 yellow signals (2Y) are expected. Cells with 1R, 1G, 1R/G are indicative of IGH rearrangement. ALL, acute lymphoblastic leukemia; CRLF2, cytokine receptor-like factor 2; FISH, fluorescent in-situ hybridization

Full size image

The incidence of the BCR::ABL1-like ALL among patients lacking recurrent cytogenetic features was 15.4% and in the whole study group of B-ALL patients it was 8.5%. Most of the cases were CRLF2-r (n = 4; 66.7%). Overall, we distinguished 5 subtypes of B-ALL in the study group: BCR::ABL1 positive ALL, BCR::ABL1-like ALL, ALL with KMT2A-r, TCF3::PBX1 positive ALL and other B-ALL. In Table 1 we present a brief summary of clinical characteristics of distinguished cytogenetic subtypes of B-ALL. In Table 2 we present the clinical characteristics of BCR::ABL1-like ALL patients. The incidence of distinct entities is presented in Fig. 3.

Table 1 Clinical and laboratory characteristics of distinct cytogenetic subgroups of the studied patients with B-ALL (n = 71)
Full size table
Table 2 Characteristics and clinical outcome of patients with BCR::ABL1-like ALL
Full size table
Fig. 3
figure 3

Cytogenetic and molecular characterization of the studied B-ALL patients. The proportion of patients for a particular subgroup of the whole cohort (n = 71)

Full size image

Additionally, patients with CRLF2-r were enrolled in the analysis of the JAK2 exon 16 mutational status. HRMA revealed different melting profile in one studied sample (CRLF2-r case). We confirmed the presence of the variant LRG_612:c.2049A>C(p.Arg683Ser) using Sanger sequencing in this case (Fig. 4). Overall, the incidence of point mutation in the JAK2 exon 16 within CRLF2-r cases was 25%.

Fig. 4
figure 4

The high resolution melt analysis (HRMA)—top field, and Sanger sequencing result (bottom field) in the CRLF2-r patients. HRMA revealed abnormal melting profile in one studied sample, in contrast to the normal double-stranded DNA dissociation characteristics during heating in control samples (wild type, WT). In the presented case, Sanger sequencing study revealed the presence of the variant LRG_612:c.2049A>C(p.Arg683Ser). Reference transcript ID (RefSeq): NM_004972.4:c.2049A>C, NP_004963.1:p.(Arg683Ser)

Full size image

Discussion

Herein we present a strategy to identify cases with potentially targetable genomic lesions which can be applied in a limited resource setting. A similar approach integrating FISH and FCM has been implemented by Sharma and Virk [20, 21]. In the material from our Department, high expression of CRLF2 was indicative of the presence of CRLF2-r, similarly to observations from a larger cohort in the study of Virk et al. [21]. The frequency of BCR::ABL1-like cases in our material was 15.1% of B-ALL patients. The incidence of this entity in our cohort appears to be lower than in the literature data [6, 7]. It might be the result of both substantial number of cases excluded lacking adequate material and of limited techniques applied in the study. Sharma used a similar cost-effective approach in a larger cohort and revealed a slightly lower incidence of BCR::ABL1-like cases in − 11.4% of B-ALL in the screened group. Notably, this study group included adults, as well as the pediatric population, in which BCR::ABL1-like ALL is less frequently reported [20]. Among JAK-STAT pathway fusions, the rearrangements of CRLF2 account for the majority of cases [8, 9, 20]. The overexpression of CRLF2 observed in FCM, may be the result of either cryptic deletion of the pseudoautosomal region 1 of chromosomes X and Y leading to the gene fusion P2RY8::CRLF2, or the translocation resulting in the gene fusion IGH::CRLF2 [21, 22]. Approximately 50% of patients with CRLF2-r ALL harbor mutations in the JAK family genes, mainly in the JAK2 gene [6]. In our group, one patient harbored a point mutation within the exon 16 of the JAK2 gene, JAK2 c.2049A>C (p.R683S), accounting for 25% of CRLF2-r cases. Notably, the mutation occurred in the patient with a relapse after a prolonged remission post-alloHSCT. The relatively low frequency of JAK2 mutations in the study may be explained by the applied technique, which is less sensitive than the next generation sequencing implemented in numerous reports, a small study group and the fact that JAK2 mutations in the BCR::ABL1-like ALL may occur in other coding regions.

The optimal treatment strategy of the BCR::ABL1-like ALL is debatable. As far as the molecular background of this subtype is concerned, the combination of standard chemotherapy with TKI remains promising. Several preclinical studies and case studies reported safety, activity and efficacy of the JAK inhibitor, ruxolitinib, in BCR::ABL1-like ALL harboring JAK-STAT-activating aberrations and ABL-class inhibitors in cases with rearrangements of ABL-class genes [6, 10, 14, 23,24,25,26,27,28]. The studies by Steeghs et al., on the other hand, revealed that proliferation of JAK2 mutated ALL cells depended on several signaling pathways activity [29]. Hence, while JAK2-r leukemic cells were found to be susceptible to JAK inhibitors, both ruxolitinib and momelotinib, the efficacy of JAK specific therapy may be limited in JAK2 mutated cells. Similar results were observed by Schwartzman et al. [30]. Furthermore, the study of Steeghs et al. provides rationale for the hypothesis that JAK2 mutations may be secondary lesions in the leukemic process, while JAK2 rearrangements are leukemic drivers. Therefore, it is suggested to combine JAK inhibitors with Ras pathway inhibitors to avoid clonal selection [29,30,31,32]. A synergistic effect of combination of TKIs with antagonists of the BCL-2 anti-apoptotic protein, venetoclax and navitoclax, was also reported [24].

The role of alloHSCT in the first CR is also a subject of debate, since the prognostic impact of MRD negativity post-induction remains questioned [33, 34]. An analysis of Koller et al. suggests that alloHSCT may overcome the poor prognosis of CRLF2-r ALL [35]. It is postulated that patients with the presence of CRLF2-r and JAK2-r should be considered as candidates for alloHSCT [36]. On the other hand, relapses post-alloHSCT are often driven by CRLF2-r clones. These relapses occur irrespective of the MRD-negativity achievement, since CRLF2 fusions are considered early events in the leukemogenesis and CRLF2-r malignant clone may persist in a quiescence during the treatment, and eventually escape the immune system or gain a proliferative state trough acquired mutations [37,38,39]. Notably, herein we report a case of a patient with CRLF2-r which occurred during a relapse after a prolonged remission, despite the absence of CRLF2-r in at the initial diagnosis. Conversely, we could not exclude the possibility of overexpression of CRLF2 on leukemic blast at the original diagnosis, since it was not evaluated in flow cytometry at that time. Shah et al. described a similar case of an individual with a relapse of CRLF2-r ALL after a prolonged remission, however, the authors did not analyze the presence of CRLF2-r in the material from the initial diagnosis [39]. Studies by Aldoss et al. revealed that during a late relapse of ALL after alloHSCT, novel cytogenetic aberrations might occur as a manifestation of a genetic evolution of the disease or clonal selection, or even due to de novo secondary leukemogenesis related to former therapy [40].

Although our study provided valuable results regarding BCR::ABL1-like ALL diagnosis, it did have some drawbacks. The first is the small number of patients enrolled in the study and its mainly retrospective nature due to a relatively low incidence of B-ALL in the adult population. A substantial proportion of patients was eventually excluded as a result of lack of adequate cytogenetic material, which could impact the overall incidence of BCR::ABL1-like cases in the analyzed group. The second limitation is the use of standard diagnostic techniques which, however, are acceptable, if next generation techniques are unattainable [41]. Another limitation of the study is the fact that enrolled patients were treated over a long time period of time. Although most of the subjects were diagnosed and treated according to the guidelines of PALG, the therapeutic protocol evolved over the last decade, hence the patients were not uniformly treated. Finally, this is a single-center study, therefore, it presents a cohort which is not large enough to show significance. On the other hand, our results remain useful for future meta-analysis on BCR::ABL1-like ALL incidence and outcomes from real-world settings. Our study demonstrates that smaller centers can potentially provide useful information regarding BCR::ABL1-like ALL, regardless of the limited techniques employed. Our results remains also essential considering the potential advent of molecularly targeted therapy in BCR::ABL1-like patients.

Conclusions

The diagnostic strategy implementing widely available techniques enables the identification of high risk and therapeutically targetable cases of BCR::ABL1-like ALL. The presented approach may be particularly appropriable in settings with limited resources.