MRD-Negative Remission Induced in EP300-ZNF384 Positive B-ALL Patients by Tandem CD19/CD22 CAR T-Cell Therapy Bridging to Allogeneic Stem Cell Transplantation

Introduction

B-cell acute lymphoblastic leukemia (B-ALL) represents a group of highly heterogeneous hematological malignancies. Approximately 55% of B-ALL patients present with normal cytogenetics at diagnosis and are stratified into a standard risk group, many of whom, however, experienced relapse under conventional treatment strategies.¹ With the application of whole-transcriptome sequencing, a growing number of cryptic aberrations have been revealed, which facilitate diagnosis, risk stratification, and targeted therapy in B-ALL patients with a normal karyotype.

EP300-ZNF384 is generated by a cryptic t(12;22)(p13;q13) chromosome translocation and can be detected in 4.3–5.7% of B-ALL patients.^2,3 EP300-ZNF384-positive B-ALL patients often have a high expression of CD19 and CD22, weak expression of CD20 and aberrant expression of CD13 and/or CD33.⁴ These patients also have a distinct gene signature with concurrent upregulation of the JAK-STAT pathway, reduced expression of B-cell regulators and poor DNA repair capacity.³ Based on small-scale retrospective studies, these patients are sensitive to chemotherapy and have a favorable clinical outcome. Because of its low incidence, there is a shortage of large-sample clinical studies focusing on this entity, and studies on salvage therapy for patients in whom conventional treatments fail are especially rare.⁵ Chimeric antigen receptor (CAR) T-cell therapy has shown an encouraging efficacy in refractory and relapsed (R/R) B-ALL.⁶ Here, we report the successful application of CAR T-cell therapy using a tandem CD19/CD22 CAR in two R/R B-ALL patients with the EP300-ZNF384 fusion. The schematic of the CD19/CD22 CAR structure is shown in Figure 1A.

Figure 1 Clinical characteristics of Case 1. (A) The structure of tandem CD19/CD22 CAR. (B) Flow cytometry analysis of BM samples before and after CAR T-cell infusion. (C) The infusion schedule of Case 1, with FC conditioning regimen and 3 consecutive days CAR T-cells infusion. (D) The changes in temperature and the WBC count post CAR T-cell infusion. Case 1 suffered high fever at day 3 after infusion. With the increase in the WBC count and supportive treatment, the patient’s temperature returned to normal. (E) CAR T-cell copies detected by qPCR post infusion.

Case Reports

Case 1 was a 30-year-old male diagnosed in February 2017 with B-ALL in the standard risk group. He received induction chemotherapy with daunorubicin, cyclophosphamide, vincristine, prednisone and L-asparaginase and achieved complete remission (CR), which he maintained under consolidation and maintenance chemotherapy for 3 consecutive years according to the GRAAPH-2003 protocol.⁷ In April 2020, hematologic relapse was detected in a regular follow-up. BM aspirates showed 7% blasts, with negative peroxidase staining. Flow cytometry revealed that the blasts were positive for CD13, CD19, CD22, CD33, CD34, CD38, CD79a, HLA-DR and negative for CD20, consistent with B-ALL. A normal karyotype was detected. No fusion genes were detected using a multiplex RT-PCR panel covering 43 fusion genes commonly seen in acute leukemia (Supplementary Table 1). No mutations were detected using a targeted next-generation sequencing (NGS) panel covering 172 genes (Supplementary Table 2). The EP300-ZNF384 fusion was detected firstly using RNA sequencing and further confirmed with Sanger sequencing (Supplementary Figure 1A), with fusion of exon 6 of EP300 to exon 3 of ZNF384. The patient received chemotherapy (mitoxantrone 10 mg on day 1, vincristine 4 mg on day 1, and dexamethasone 14 mg on days 1–7) but he still had 82% of blasts. Considering very weak expression of CD20 (0.13%) and high expression of CD19 (98.3%) and CD22 (95.2%) (Figure 1B) in the blasts, he was enrolled in a phase I/II study involving tandem anti-CD19 and CD22 CAR-T cells in R/R B-ALL (NCT03614858). Chemotherapy (fludarabine 30 mg/m² and cyclophosphamide 300 mg/m²) was administered 5, 4 and 3 days before the first infusion of CAR T-cells (June 27, 2020), which was followed by three infusions administered for the next 3 days. The effective anti-CD19/CD22 CAR T-cells totaled 1 × 10⁷/kg (Figure 1C). He developed high fever, hypotension and elevated IL-6. Grade 3 cytokine release syndrome (CRS) was determined according to ASTCT Consensus⁸ (Figure 1D). The CRS was relieved by dexamethasone, high-dose norepinephrine and symptomatic treatments. No signs of immune effector cell-associated neurotoxicity syndrome (ICANS) were observed. At 28 days post CAR T-cell therapy, the BM smear revealed no blasts, and flow cytometry analysis showed 4.3 × 10⁻⁴ blasts (Table 1). Amplification of CAR T-cell as detected by qPCR peaked at day 12 and remained detectable before HSCT (Figure 1E). At the same time, he tested negative for the EP300-ZNF384 fusion. He underwent haploidentical allogeneic stem cell transplantation (allo-HSCT) with his sister as donor one month after CAR T-cell therapy. BM evaluation at a month after HSCT showed <2.4 × 10⁻⁵ of blasts by flow cytometry and negative EP300-ZNF384 with qPCR. Following HSCT, he exhibited grade IV acute graft-versus-host disease (GVHD) involving the intestine and liver according to the European Bone Marrow Transplantation consensus,⁹ which was relieved with tacrolimus, monoclonal antibody against CD25, etanercept, fecal microbiota transplantation and supportive care. At present, he has maintained continuous measurable residual disease (MRD)-negative remission (<0.01% of bone marrow blasts by flow cytometry) for 14 months post HSCT.

Table 1 Clinical Characteristics of Patients at Relapse

Case 2 was a 25-year-old female diagnosed in January 2020 with B-ALL in the standard risk group. White blood count at diagnosis was 18.93 × 10⁹/L. CR was achieved after chemotherapy with idarubicin, vincristine, pegaspargase and prednisone. After consolidation chemotherapy with the hyper-CVAD regimen, hematologic relapse was detected in June 2020. BM morphology showed that 60% of blasts were positive for CD13, CD19, CD22, CD33, CD34, CD38, cCD79a, and HLA-DR and negative for CD20 (Figure 2A). A normal karyotype was detected. The results of multiplex PCR were negative. The EP300-ZNF384 fusion gene was also detected with RNA sequencing, with fusion of exon 6 of EP300 to exon 3 of ZNF384 (Supplementary Figure 1B). The NGS results showed KRAS Glu63_Ala66delinAspProThr and KMD6A Leu231fs mutations. After reinduction chemotherapy (idarubicin 10 mg on day 1, vincristine 4 mg on day 1, and dexamethasone 14 mg on days 1–7), the blasts remained at 20.5%. Due to the weak expression of CD20 (1.13%) and the high expression of CD19 (100%) and CD22 (67.3%), she was enrolled in the same clinical trial and received the same chemotherapy 5, 4 and 3 days before the first infusion of CAR T-cells, which was followed by three infusions administered for the next 3 days. The effective anti-CD19/CD22 CAR T-cells totaled 2 × 10⁷/kg (Figure 2B). She suffered grade 2 CRS and presented with fever and hypotension, which was relieved by symptomatic treatments and low-dose norepinephrine (Figure 2C). No symptoms of ICANS were observed. Amplification of CAR T-cell as detected by qPCR peaked at day 10 and dropped detectable before HSCT (Figure 2D). At 28 days post CAR T-cell therapy, BM evaluation showed <1 × 10⁻⁴ of blasts by flow cytometry and negative EP300-ZNF384 with qPCR, which demonstrated that MRD-negative CR was achieved (Table 1). In addition, results for the EP300-ZNF384 fusion and KRAS and KMD6A mutations all became negative. She underwent a haploidentical allo-HSCT with her sister as donor at a month post CAR T-cell therapy. No signs of GVHD were observed, and she remained in MRD-negative CR for 13 months after CAR-T therapy.

Figure 2 Clinical characteristics of Case 2. (A) Flow cytometry analysis of BM samples before and after CAR T-cell infusion. (B) The infusion schedule of case 2. The patient got 4 days CAR T-cells infusion with 2 × 107/kg. (C) The changes in temperature and the WBC count post CAR T-cell infusion. Case 2 suffered high fever at day 4 after infusion. With the increase in the WBC count and supportive treatment, the patient’s temperature returned to normal. (D) CAR T-cell copies detected by qPCR post infusion.

Discussion

EP300-ZNF384-positive B-ALL is a rare subtype with a characteristic immunophenotype and is more prevalent in adolescents and young adults according to the limited number of cases reported thus far.^3,4 Both patients presented here were young adults, showing abnormal expression of CD13 and CD33 and weak expression of CD20 at relapse, in accordance with previous studies.³ Both EP300 and ZNF384 are located within close proximity of the telomeres of their respective chromosomes, which makes EP300-ZNF384 difficult to identify by conventional cytogenetic analysis. Therefore, patients with R/R B-ALL and a normal karyotype should be screened for EP300-ZNF384 using transcriptome sequencing. EP300-ZNF384 is a newly identified fusion gene in B-ALL, with only 53 cases reported hitherto (Table 2). Most patients harboring the EP300-ZNF384 fusion achieved CR and had 62.5% survival with conventional chemotherapy in a previous study.^3,10,11 However, some patients with EP300-ZNF384 also suffered relapses and were refractory to salvage chemotherapy (Supplementary Table 3).⁵ Both patients reported here had weak expression of CD20 and therefore presented no target for anti-CD20 monoclonal antibody or CD20-targeted CAR T-cell therapy. As reported in most cases, exon 6 of EP300 was fused with exon 3 of ZNF384 in both cases.¹¹ The EP300-ZNF384 fusion protein lacks the histone acetyltransferase (HAT) domain of EP300, and retained the complete ZNF384 protein.¹¹ EP300-ZNF384 without the HAT domains may be involved in the development of ALL by deregulating acetylation-mediated inactivation of the BCL6 oncoprotein and activation of the p53 tumor suppressor.¹³ In addition, the upregulation of the JAK-STAT pathway and reduced DNA repair capacity have also been identified in patients with EP300-ZNF384.³ However, there have been no in vitro or in vivo studies on the application of histone acetyltransferase regulators, JAK inhibitors or drugs targeting DNA repair pathways in this subtype.

Table 2 Summary of Genetic Details of Patients with EP300-ZNF384

CAR T-cell therapy has shown unprecedented success in R/R B-ALL.⁶ At present, only two EP300-ZNF384-positive B-ALL patients have undergone CAR T-cell therapy; however, both patients died.⁵ No details about the efficacy and safety of CAR T-cells for these patients were reported. Despite the high response rate of CD19 CAR T-cell therapy, approximately 50% of R/R B-ALL patients relapsed after CD19 CAR T-cell therapy, which is partially due to CD19 antigen loss.⁶ CD19 and CD22 dual-targeted CAR T-cells have demonstrated potential in reducing the relapse rate of R/R B-ALL patients with antigen loss in clinical trials.¹⁴ Because of the concurrently high expression of CD19 and CD22, we applied tandem CD19/CD22 CAR T-cell therapy in both patients to avoid relapse. Both patients reported here were chemo-resistant and had high tumor burden before infusion of CAR T cells. At day 28, in the BM evaluation, both patients achieved CR, which demonstrated the potent antileukemic activity of the CD19/CD22 CAR T-cells. No symptoms of ICANS were observed. All toxicities related to CRS were relieved with symptomatic treatments and supportive care. Hence, this bispecific CD19/CD22 CAR T-cell infusion appears safe and effective in EP300-ZNF384-positive patients with a high tumor burden. With CAR T-cell therapy bridging to allo-HSCT, the 1-year overall survival and event-free survival for R/R B-ALL were shown to increase to 87.7% and 73%, respectively in a large cohort.¹⁵ Therefore, allo-HSCT was subsequently performed in both patients to prevent relapse and achieve sustainable remission. Both patients had maintained CR at the last follow-up.

To our knowledge, this is the first report of EP300-ZNF384-positive B-ALL patients achieving continuous deep remission through tandem CD19/CD22 CAR T-cell therapy bridging to allo-HSCT. Due to the lower incidence of EP300-ZNF384-positive R/R B-ALL and the limited number of patients in this study, the outcome of such a strategy in EP300-ZNF384-positive B-ALL needs to be investigated in a larger, multiple-center clinical trial.