Genetic Prole and Clinical Implications of PDGFRB Fusion in Adult B-Cell Acute Lymphoblastic Leukemia: A Retrospective Analysis

Background:To investigate the pathogenesis, genetic changes, treatment and prognosis of adult B-cell acute lymphoblastic leukemia (B-ALL) with platelet-derived growth factor receptor B (PDGFRB) fusion. Method (cid:0) Clinical characteristics, treatment process and prognosis of 7 adult B-ALL patients with PDGFRB fusion were presented respectively, and the PDGFRB fusion was conrmed by uorescence in situ hybridization, RNA sequencing and Polymerase Chain Reaction. Results: 7 adult B-ALL patients with PDGFRB fusion were presented respectively. There were 5 males and 2 females, with a median age of 21 (range, 17-39) years and a median white blood cell count of 7.85 (range, 2.30-111.25) ×10^9/L. The fusion partners were EBF1, SMIM3 and TERF2. Three of 7 cases received tyrosine kinase inhibitor (TKI) targeted therapy, and 4 of 7 cases received chimeric antigen receptor T -cell (CAR-T) therapy. A total of 6 (85.71%) cases underwent haploid hematopoietic stem cell transplantation (Haplo-HSCT). Overall response rate was 71.43% (4/7), and the common adverse reactions during induction were febrile neutropenia and nausea. Seventy-ve percent (3/4) and 100% (6/6) of patients were minimal residual disease negative after CAR-T and Haplo-HSCT. To the last follow-up, 2 of 7 cases relapsed in 2 and 7 months after remission, respectively, and the estimated 36 months of Event-free survival and Overall survival was 75.0% and 66.7%. Conclusions: intensive therapy combined with CAR-T or TKI is needed to achieve deep remission, and sequential Haplo-HSCT may improve the prognosis of adult B-ALL with PDGFRB fusion. Ph-like is associated with poor prognosis. 2 Allogeneic hematopoietic stem cell transplantation (allo-HSCT) was recommend and may improved this poor outcome. 3 Fusion of platelet-derived growth factor receptor B (PDGFRB) has been emphasized in myeloproliterative neoplasms, and has become a therapeutic target of tyrosine kinase inhibitor (TKI). 4 However, belonging to the ABL-class of fusion kinase category of Ph-like B-ALL, the fusion oncogene of PDGFRB are rarely reported in B-ALL. 5 PDGFRB rearrangement in conventional karyotype examination is prone to low detection rate and missed diagnosis, 6 and a rather proportion of B-ALL with PDGFRB rearrangement were refractory to conventional therapy. 7 Given that patients with PDGFRB fusions are amenable to targeted therapy of TKI, 8 it is necessary to identify these patients by using a comprehensive screening strategy. To provide a better comprehension of the clinical and pathological features, we performed a retrospective observational study in 7 cases of B-ALL with PDGFRB fusion in the First Aliated Hospital of Soochow University. CART therapy. Case 2 relapsed rapidly in 1.5 months after dasatinib administration. In case 5, the time-point of CR and CR MRD− was 61 and 139 days; even if CR was achieved after induction therapy, MRD remained positive and then turned negative after CAR-T therapy. As same as adult Ph + ALL, consisting of a TKI-based induction/consolidation 29,30 may be the suitable treatment for patients with PDGFRB fusion. It suggested that the targeted monotherapy is still inadequate for patients with PDGFRB fusion. Intensive therapy combined with CAR-T or TKI is needed to achieve deep remission for these patients. In our results, 75% (3/4) and 100% (6/6) of patients maintain MRD negative after CAR-T and Haplo-HSCT. It seem to be that sequential allo-HSCT after deep remission is expected to improve the prognosis of such patients, and we highly recommend the allo-HSCT as the consolidation if possible. Future studies in vitro and on large patients are very necessary to conrm this perspective, and this result may contribute to establish the current standard treatment paradigm for PDGFRB fusion in adult B-cell ALL. Our results suggested that intensive therapy combined with CAR-T or TKI is needed to achieve deep remission, and sequential allo-HSCT may improve the prognosis of


Background
Disease progression remains a primary cause of mortality in B cell acute lymphoblastic leukemia (B-ALL), especially for Philadelphia-like (Ph-like) ALL, a highrisk subtype of B-ALL. 1 Owning a similar gene expression pro le of Philadelphia chromosome-positive (Ph+) ALL, Ph-like ALL is associated with poor prognosis. 2 Allogeneic hematopoietic stem cell transplantation (allo-HSCT) was recommend and may improved this poor outcome. 3 Fusion of platelet-derived growth factor receptor B (PDGFRB) has been emphasized in myeloproliterative neoplasms, and has become a therapeutic target of tyrosine kinase inhibitor (TKI). 4 However, belonging to the ABL-class of fusion kinase category of Ph-like B-ALL, the fusion oncogene of PDGFRB are rarely reported in B-ALL. 5 PDGFRB rearrangement in conventional karyotype examination is prone to low detection rate and missed diagnosis, 6 and a rather proportion of B-ALL with PDGFRB rearrangement were refractory to conventional therapy. 7 Given that patients with PDGFRB fusions are amenable to targeted therapy of TKI, 8 it is necessary to identify these patients by using a comprehensive screening strategy. To provide a better comprehension of the clinical and pathological features, we performed a retrospective observational study in 7 cases of B-ALL with PDGFRB fusion in the First A liated Hospital of Soochow University.

Method Patients and Study design
This was a retrospective, observational and descriptive study in the First A liated Hospital of Soochow University, which was approved by Ethics Committee of hospital, and informed consent form was provided by all patients. From 2015 to 2020, a total of 423 patients were diagnosed as Philadelphia chromosomenegative (Ph-) B-ALL in our center. 7 cases of which (1.6%) were found with PDGFRB fusion, and included in our study. The diagnosis of ALL with PDGFRB fusion was based on morphology, immunotyping, chromosome karyotype, molecular genetics and detection of somatic mutation (illumina nextseq550). Polymerase Chain Reaction (PCR), uorescence in situ hybridization (FISH) and RNA sequencing (RNA seq, novaseq 6000) were integrated as the methods for testing PDGFRB fusion.

Treatment protocols
Induction chemotherapy: IVP (Idarubicin, Vincristine and Prednisone)-like regimen, such as VILP (L: L-Asparaginasum/pegaspargase), VDCP (D: doxorubicin, C: Cyclophosphamide), CIVP and VDLP was executed as the treatment protocols of induction therapy. Only one case (case 3) received the regimen of hyperfractionated cyclophosphamide, vincristine, doxorubicin, and dexamethasone (Hyper-CVAD)-A, 9 and another case (case 5) took Dasatinib orally for inductive treatment ( Table 2). The minimal residual disease (MRD) was analyzed in the bone marrow by ow cytometry (Beckman Coulter Ireland, Navios, Flow 10-color antibody panels). Consolidation regimen: Consolidation chemotherapy was based on high-dose methotrexate (MTX) and cytarabine (Ara-C), 10 and 3 cases were combined with TKI therapy ( Table 2). Chimeric antigen receptor-modi ed T cell (CAR-T) therapy: CAR-T cells were derived from autologous lymphocyte. Production of CAR-T cell was described in detail according to the previous report. 11 Allogeneic hematopoietic stem cell transplantation (Allo-HSCT): The conditioning regimen for patients received Allo-HSCT was modi ed BUCY (semustine, cytarabine, busulfan, and cyclophosphamide, BuCy). 12 All patients received the transfusion support of umbilical cord blood (UCB) pre-transplantation. 13 Cyclosporin A, short-term methotrexate, and mycophenolate mofetil were used for graft-versus-host disease (GvHD) prophylaxis. 14 Moreover, Anti-Thymocyte Globulin (ATG, Genzyme, Cambridge, MA, USA) was also administrated in vivo at 2.5 mg/kg on days -5 to -2 for preventing GvHD after haploidentical allo-HSCT.

De nitions and evaluations
De nition: Complete remission (CR), partial remission (PR), and relapsed disease were de ned according to 2017 ELN recommendations. 15 Overall response rate (ORR) was de ned as (CR + PR)/ (all of the patients). Negativity or remission of MRD was de ned as a level of MRD lower than 0.0001. CR without MRD Cytokine Release Syndrome (CRS) and Neurologic Toxicity was graded respectively according to the consensus. 16 The methods of diagnosis and treatment for acute GVHD (aGVHD) and chronic GVHD (cGVHD) were previously described in detail. 17,18,19 Results

Patients' characteristics
The characteristics of 7 patients with PDGFRB fusion were showed in Table 1, including 5 males and 2 females, with a median onset age of 21 (range, 17 to 39) years. There were 5 cases of initial diagnosis and other 2 cases of recurrence ALL. The median white blood cell (WBC) count at diagnosis was 7.85 (range, 2.3 to 111.25)×10^9/L, and median platelet count was 51 (range, 28 to 191)×10^9/L. The fusion partners were EBF1 (n=4), SMIM3 (n=1), TERF2 (n=1) and only 1 case with a unknown partner gene. For patients with EBF1-PDGFRB fusion, corresponding mutations were found with BCORL1 germline mutation, and somatic mutation of PAX5 and RUNX1, respectively. Case 3 was combined with SMIM3/PDGFRB fusion and TP53 somatic mutation. 4 of 7 cases showed anemia and most of patients (6/7) presented as normal karyotype at diagnosis. No eosinophils were found in bone marrow morphology, and there was no 5q31-33 rearrangement in karyotype in all patients (Table 1).

Consolidation, Relapse and Survival
Three cases undertook a comprehensive therapy combined with TKI; TKI was administrated after CR for case 1, after MRD-for case 2, and in the middle of induction chemotherapy for case 3, respectively. Four cases received CAR-T therapy (2 for eliminating MRD, and 2 for re-induction therapy) and 6 cases underwent allo-HSCT from haploid-related donors. The recurrence rate was 28.57% (2/7). Case 2 and 7 experienced disease recurrence before allo-HSCT, with    (Table 3). Allo-HSCT A total of 6 patients underwent Allo-HSCT from haploid-related donors (Haplo-HSCT). A total of 3 cases underwent CART bridging transplantation, and case 2 received CAR-T therapy as re-induction therapy before transplantation. Case 7 underwent Hyper-CVAD, VDCP, VTCP, AAG (cytarabinoside + acracycin + G-CSF) regimen and did not achieve CR, and then received Salvage transplantation. The median time was 5.5 (range, 4-16) months from diagnosis to transplant. Side effects were mainly for the mild skin GVHD (grade 1) and diarrhea. No cases have disease recurrence after transplantation. Table 4.

Discussion
The PDGFRB gene located at chromosome 5q32/5q33 is a member of the receptor tyrosine kinase family. 7,20 The occurrence of chromosomal microdeletions during PDGFRB fusion resulted in di cult identi cation and missed diagnosis in routine chromosome karyotype analysis, which was also con rmed by the fact that no abnormality of chromosome was found in 5 out of 7 patients in this study. PDGFRB fusions result in the synthesis of constitutively activated tyrosine kinase fusion proteins, usually associated with continuous cell proliferation. 21 Expression of PDGFRB fusions resulted in cytokine-independent proliferation and activation of phosphorylated STAT5. 5 Cell lines and human leukemic cells expressing PDGFRB fusions were sensitive in vitro to dasatinib. 5 Several signalling cascades, including the Ras/mitogen-activated protein kinase, phosphatidylinositol 3′-kinase and phospholipase-γpathways, are key downstream mediators of PDGFRB signalling. 6 Haematopoietic neoplasms with PDGFRB fusions are very sensitive to tyrosine kinase inhibitors (TKIs). 22 To date, 12 fusion partners of PDGFRB have been reported in adult B-cell acute lymphoblastic leukemia, 21,23−26 with EBF1-PDGFRB being the most common. 7 EBF1-PDGFRB fusion accounts for 0.5% of B-cell precursor ALL and 2.7% of the B-other subtype. 7 The EBF1 (EBF1) gene located at chromosome 5q33, is a nuclear transcription factor required for normal B-lineage speci cation, commitment, and development. 27 EBF1-PDGFRB fusion results in loss of EBF1 function, multimerization and autophosphorylation of the fusion protein, activation of STAT5 signaling, and gain of IL-7-independent cell proliferation. 27 Deregulation and loss of EBF1 function is critically dependent on the nuclear export activity of the TM domain of PDGFRB. 27 EBF1-PDGFRB synergizes with loss of IKAROS function in a fully penetrant B-ALL in vivo. 27 In this study, fusion of EBF1-PDGFRB was con rmed in 4 of the 7 patients. Schwab et al. 7 reported the results of 15 cases of pediatric ALL patients with positive EBF1-PDGFRB, with more females than males (11/15). In this study, there was a predominance of males (3/4), heterogeneity of disease between adults and children may need to be studied to con rm this phenomenon.
In this study, two novel fusion partners of PDGFRB were identi ed: SMIM3 and TERF2. In SMIM3-PDGFRB fusion, the SMIM3 breaking site was exon 1, with no coding promoter region and no coding fusion protein, while the PDGFRB gene breaking site was exon 12 in the catalytic region of tyrosine kinase. Therefore, this case should not be classi ed as ph-like ALL. In TERF2-PDGFRB fusion, exon 1-8 of TERF2 gene is fused with exon 9-23 of PDGFRB gene, which loses the DNA binding domain of TERF2 gene and retains the tyrosine kinase domain of PDGFRB gene, which is expected to be involved in the proliferation and cell cycle differentiation of leukemia cells. We are studying and verifying the related functions of TERF2-PDGFRB fusion gene.
Hiroto Inaba and Charles G. Mullighan reported that the CR rate was approximately 98% in pediatric patients. 28 However, the CR rate at the end of induction was only 57.14% in this group. The recurrence rate (28.57% vs. 46.67%) of patients in this group were lower than literature reports, 7 suggesting that intensive chemotherapy combined with other treatment methods (target therapy and CAR-T) may bene t the patients in this group. Case 3 with TP53 mutation was classi ed as high-risk but achieved early deep remission after receiving hyper-CVAD A regimen, which may support this assumption. Most of all, 6 of 7 cases in our group received allo-HSCT, which was still a curative therapeutic option for adults ALL patients with PDGFRB. For patients received targeted therapy (Case 1, 2 and 5), case 1 achieved CR soon after administrating the imatinib, however, MRD remained at 0.001 level for a long time even after CART therapy.
Case 2 relapsed rapidly in 1.5 months after dasatinib administration. In case 5, the time-point of CR and CR MRD− was 61 and 139 days; even if CR was achieved after induction therapy, MRD remained positive and then turned negative after CAR-T therapy. As same as adult Ph + ALL, consisting of a TKI-based induction/consolidation 29,30 may be the suitable treatment for patients with PDGFRB fusion. It suggested that the targeted monotherapy is still inadequate for patients with PDGFRB fusion. Intensive therapy combined with CAR-T or TKI is needed to achieve deep remission for these patients. In our results, 75% (3/4) and 100% (6/6) of patients maintain MRD negative after CAR-T and Haplo-HSCT. It seem to be that sequential allo-HSCT after deep remission is expected to improve the prognosis of such patients, and we highly recommend the allo-HSCT as the consolidation if possible. Future studies in vitro and on large patients are very necessary to con rm this perspective, and this result may contribute to establish the current standard treatment paradigm for PDGFRB fusion in adult B-cell ALL.

Conclusion
Our results suggested that intensive therapy combined with CAR-T or TKI is needed to achieve deep remission, and sequential allo-HSCT may improve the prognosis of these patients.