Introduction

Histone-lysine N-methyltransferase 2A (KMT2A), also known as the mixed-lineage leukemia (MLL) gene, is located at 11q23 and has been reported to have more than 100 different types of fusion partners in the literature, with t(4;11)(q21;q23) being the most common [1]. Overall, KMT2A rearrangement can be found in more than 70% of infantile B-lineage acute lymphoblastic leukemia (B-ALL), 5–6% of pediatric case, and about 15% of adult cases [2, 3]. The incidence of t(4;11)(q21;q23) translocation, which leads to the KMT2A::AFF1 fusion gene, in adult B-ALL is approximately 5%; however, its association with unique characteristics in cancer biology and clinical presentation has raised a great deal of interest. Bispecific T-cell engagers (BiTEs) has become an important therapeutic option for B-ALL and usually takes the form of a monoclonal antibody targeting CD19 and CD3 on B and T cells, respectively, causing B lymphoblasts to be directly lysed by cytotoxic T cells. Inotuzumab ozogamicin, an anit-CD22 monoclonal antibody-drug conjugate, is another established treatment option for relapsed and refractory B-ALL. To date, a few cases of KMT2A-rearranged ALL that show a lineage-switch into mixed phenotype acute leukemia (MPAL) after BiTE therapy have been reported [4,5,6]. In this report, we present the case of a 36-year-old woman who was diagnosed with B-ALL with t(4;11)(q21;q23), which transformed into MPAL soon after two lines of B lymphoblast-targeted therapies.

Clinical history

A 36-year-old woman who had been previously healthy visited a local hospital presented with fever and exertional dyspnea. Complete peripheral blood (PB) counts showed pancytopenia with circulating blasts. Bone marrow (BM) examination revealed an extremely hypercellular BM with more than 70% blasts that were positive for CD19 and cytoplasmic CD79a, CD45, and CD22, and negative for CD34 (majority), CD10, CD20, CD117, and CD33 (Fig. 1a, Supplementary Fig. 1). Cytogenetic analysis revealed t(4;11)(q21;q23) in 18 of the 20 analyzed cells (Fig. 1b). The patient then received induction chemotherapy with a hyperCVAD regimen. However, after two courses of chemotherapy, follow-up BM aspiration revealed refractory disease, and the patient was referred to our hospital for further evaluation and treatment, including allogeneic hematopoietic stem cell transplantation.

Fig. 1
figure 1

a Immunophenotyping of the patient’s bone marrow specimen at initial diagnosis. The B-ALL lymphoblasts were CD19+, cyCD79a+, CD34-(majority). Red, CD34+ lymphoblasts; orange, CD34− lymphoblasts; green, granulocytes; pink, residual normal mature B-cells. SSC (side scatter). b Cytogenetic analysis of the patient’s bone marrow at initial diagnosis revealed presence of t(4;11)(q21;q23) (red arrow). Detailed karyotypes: 46,XX,t(4;11)(q21;q23)[13]/45,XX, t(4;11),dic(17;19)(p11.2:p13.2)[2]/47,XX,+X,t(4;11), dic(17;19),+19[3]/46,XX[2]

Full size image

Material and methods

Cell morphology was evaluated with a bone marrow aspiration smear stained with Liu’s stain (a modified Romanowsky stain). Myeloperoxidase (MPO) and chloroacetate esterase /α-naphthyl butyrate esterase (ANBE) staining were performed to facilitate lineage assignment. Immunophenotyping was performed on bone marrow aspirates using an eight-color assay with internal controls to determine positive and negative patterns. Cytogenetic analysis was performed using a conventional G-banding technique.

Results

After admission, the BOMES regimen (BCNU, vincristine, methotrexate etoposide, and methylprednisolone) was administered for re-induction; however, follow-up BM examination still showed refractory disease with leukemic cells accounting for 72.37% of total nucleated cells (Fig. 2a); all of which expressed CD19 and CD22, but were negative for CD34 (majority), CD10, and CD20 (Fig. 2b). Blinatumomab was administered as a salvage treatment, but the disease persisted. Therefore, inotuzumab ozogamicin was administered because her lymphoblasts tested positive for CD22. However, all treatments failed to induce remission, and elevated PB blasts were still observed one week later. Interestingly, we noticed that the morphology of the PB blasts at this time became atypical for lymphoblasts, with most blasts having a monocytoid appearance with abundant cytoplasm and folded nuclei. Therefore, BM examination was performed again to ascertain the patient’s disease status. The BM smear revealed the presence of two blast populations: a smaller population had a lymphoblast-like morphology and stained negative for MPO and ANBE, whereas a larger blast population exhibited a monoblast-like morphology and negatively stained for MPO but positively stained for ANBE (Fig. 3a). Immunophenotyping using flow cytometry also confirmed that the two blast clones were lymphoblasts, 15.23% (CD34−, CD19+, CD10−, and CD20−) and monoblasts, 67.35% (CD34−, HLA-DR−, partial CD117+, partial CD11b+, CD64+, partial CD14+, CD33+, CD15+, and CD56+). Notably, both clones expressed neuron-glial antigen-2 (NG2), a marker that is strongly associated with KMT2A rearrangement (Fig. 3b). Moreover, cytogenetic examination showed persistent t(4;11)(q21;q23) (Fig. 3c). Therefore, disease evolution from B-ALL to MPAL (B/myeloid) was diagnosed. The ENOP (Etoposide, MitoXANtrone, VinCristine, 6-MP) regimen was administered as salvage therapy. Unfortunately, her disease did not respond either, and the patient quickly succumbed to rapidly progressive leukemia and superimposed infections.

Fig. 2
figure 2

Follow-up bone marrow examination of the patient. a The patient’s marrow was filled with monotonous lymphoblasts with immature chromatin morphology and high N/C ratio (Liu’s stain, 1000×). The myeloperoxidase (MPO) stain was negative. b Flow cytometric immunophenotyping revealed that leukemic cells expressed CD19+, CD22+, CD10−, CD20−, CD34- (majority), CD38+. Red, lymphoblasts; green, granulocytes; blue, mature lymphocytes. SSC (side scatter)

Full size image
Fig. 3
figure 3

a Morphological examination of the patient’s bone marrow post blinatumomab and inotuzumab ozogamicin treatment. In the Liu’s stain, two populations of abnormal cells are observed, one resembles monoblasts (red arrows), and the other resembles lymphoblasts (black arrows). The monoblasts are myeloperoxidase negative and ANBE positive, while the lymphoblasts are negative for both. b Immunophenotyping of the patient’s bone marrow post blinatumomab and inotuzumab ozogamicin treatment. The lymphoblastic clone (red) are CD19+, CD34−, CD10−, CD38+, HLA-DR+; while the monoblastic clone (light green) are CD19−, CD64+, partial CD14+, partial 11b+, CD33+, CD56+, CD38+, CD15+, HLA-DR−. Both leukemic clones expressed NG2, a biomarker of KMT2A rearrangement. SSC (side scatter). c Follow-up cytogenetics in peripheral blood specimen showed persistent existence of t(4;11) (q21;q23) (red arrow). Detailed karyotypes: 45-52,XX,+der(4)t(4;11)(q21;q23),t(4;11),+dup(7)(q22q32),+8,+12,+13,+19[cp20]

Full size image

Discussion

B-ALL with t(4;11)(q21;q23) comprises about 5% of adult B-ALL and is categorized as B-lymphoblastic leukemia/lymphoma with KMT2A rearrangement in the updated 5th edition WHO Classification, and as B-ALL with t(v;11q23.3)/KMT2A rearranged in the 2022 International Consensus Classification (ICC) classification [7, 8]. This category of leukemia is characterized by mixed-lineage leukemia and potential for lineage switching during the treatment course. Immunophenotypically, B-ALL with t(4;11) usually expresses CD19, CD34, and TdT, but not CD10 or CD20, which corresponds to very immature B precursors (pro-B cell immunophenotype). Clinically, patients with B-ALL with t(4;11) usually present with leukocytosis and extramedullary involvement (including central nervous system and hepatosplenic) and have a grave prognosis. In terms of molecular pathogenesis, t(4;11) (q21;q23) translocation results in KMT2A::AFF1 fusion (previously known as MLL-AF4), which causes aberrant overexpression of a number of genes, including the anti-apoptotic BCL2 and the proto-oncogene MYC, ultimately driving leukemogenesis [9].

The lineage switch is a unique phenomenon that may occur in B-ALL. Although KMT2A-rearranged B-ALL is the most commonly documented disease that has the potential of lineage switch, ZNF384 translocations, especially TCF3-ZNF384 or TAF15-ZNF384 had also been reported to be implicated in lineage switch process [10, 11].

According to the literature, the lineage switch from B-ALL with KMT2A::AFF1 fusion is usually bi-lineal, with the myeloid component most commonly being monocytic [12,13,14]. However, the exact mechanisms leading to the lineage switch remain unclear. B-ALL with KMT2A::AFF1 fusion has unique characteristics at presentation, including a high WBC count, propensity for CNS involvement, splenomegaly, and poor clinical outcomes. Immunophenotypically, this type of B-ALL is usually CD10− and NG2+ and is classified as Pro-B-ALL [15].

The patient presented in this report was initially diagnosed with B-ALL as evidenced by the typical features on morphological examination and immunophenotyping. However, the patient demonstrated resistance to many lines of B-ALL-directed chemotherapeutic and targeted agents and the B-ALL evolved into MPAL. This case was particularly intriguing because the MPAL transformation occurred soon after two potent targeted therapies, blinatumomab and inotuzumab ozogamicin, were administered. Maybe there was a very small clone of monocytic blasts at initial diagnosis, which became the dominant leukemic clone during disease progression. However, when the patient was transferred to our hospital, the flow cytometry study collected 500,000 events and thus could reached a theoretical limit of sensitivity of 1 × 10−4, so the likelihood of this hypothesis was rather low. Furthermore, we reason that it was more likely that the harsh selection pressure of multiple rounds of B-ALL directed chemotherapy and targeted agents eradicated the bulk of the patient’s initial leukemic burden. However, precursor leukemic cells harboring the KMT2A::AFF1 fusion could theoretically possess lineage plasticity that were later genetically reprogrammed and gave rise to the main monoblastic clone, which lacked CD19 and CD22 expressions, to evade therapeutic killing. In addition, we demonstrated that B lymphoblasts and monoblasts are clonally related, because follow-up cytogenetics of the PB cells showed persistence of t(4;11) (q21;q23), and both populations expressed NG2, as shown using flow cytometry. Therefore, we argue that, for our patient, the lineage switch occurred after B cell-targeted therapy.

Anti-CD19 CAR-T therapy, while being a potent treatment for CD19+ B-ALL, has been reported to induce the lineage switch phenomenon in B-ALL [16]. In a recent study, Jacoby et al. used single-cell technologies to confirm whether leukemic cells with lineage plasticity could undergo cellular reprogramming to switch lineage characteristics following anti-CD19 CAR-T therapy. They further used murine models to demonstrate that a number of key transcription factors, such as PAX5 and ERBF1 are important regulators in this process, because CRISPR/Cas9 mediate ablation of Pax5 or Ebf1 recapitulates the lineage redirection phenomenon under the selection pressure of CD19 CAR [17], BiTEs, such as blinatumomab, efficiently kill CD19+ cancer cells by activating the cytotoxic potential of the patient’s own T cells to lyse tumor cells; thus, the findings presented in the aforementioned study remain relevant in patients with B-ALL undergoing blinatumomab treatment. In the literature, it had been reported in multiple occasions that infantile KMT2A rearranged B-ALL could lineage switch to AML following blinatumomab treatment [6, 18, 19]. In particular, our patient was administered both blinatumomab and inotuzumab ozogamicin, an anti-CD22 antibody-drug conjugate. Therefore, the selection pressure on the B lymphoblast clone was even more intense than that reported in the literature, resulting in the transformation into MPAL with a major monoblastic dominance.

In summary, we present a patient who had been diagnosed with B-ALL, with t(4;11)(q21;q23), treated with multiple lines of chemotherapy and two lines of B cell-directed immunotherapies, and later developed a disease consisting of both B lymphoblasts and monoblasts. We suggest that when managing patients with B-ALL with KMT2A rearrangement, especially the KMT2A::AFF1 fusion, re-evaluation of the nature of leukemic blasts is warranted if the disease becomes refractory to multiple lines of B-ALL-directed therapy. Close monitoring of bone marrow cell morphology and immunophenotyping should be considered at regularly spaced time intervals, for instance every three months, to capture lineage switch at an early stage. Although the prognosis of KMT2A-rearranged B-ALL is dismal, early detection of disease transformation should prompt physicians to consider myeloid leukemia-directed therapy and even stem cell transplantation to successfully salvage these patients.