Introduction

T cell acute lymphoblastic leukemia/lymphoma (T-ALL/LBL) is a lymphocytic neoplasm of T cell lineage origin, which accounts for approximately 10–15% of childhood ALL and 25% of adult ALL [1, 2]. With improvements in treatment, the 5-year event-free survival (EFS) rate for childhood T-ALL/LBL has approached approximately 90%. However, the prognosis of adult T-ALL/LBL remains poor, especially in patients with relapsed/refractory (R/R) T-ALL/LBL. Using modern pediatric-like chemotherapy protocols that include risk stratification and possible allogeneic transplantation in first complete remission (CR), the 5-year overall survival (OS) and EFS rates are close to 65% and 60%, respectively, in patients under 60 years old [3]. However, the prognosis for R/R T-ALL/LBL and patients older than 60 remains poor, which highlights the urgent need for new therapeutic approaches. Deep understanding of the genetics of T-ALL/LBL offers new therapeutic avenues, particularly with the widespread use of next-generation sequencing [4]. In addition, there is significant biological and genetic heterogeneity in T-ALL/LBL, which poses a challenge to the development of widely used molecular targeted therapy. Currently, immunotherapies such as CD7 chimeric antigen T cell therapy (CD7 CAR T) have also been applied successfully in R/R T-ALL/LBL. In this review, we will describe the application of molecular targeted and immunotherapies developed in recent years in T-ALL/LBL.

Molecular targeted therapy

Through comprehensive cytogenetic and molecular studies, a large number of genetic biomarkers have been identified in T-ALL/LBL. Here, we present the preliminary results of a series of preclinical and clinical studies targeting these markers (Tables 1 and 2).

Table 1 Summary of relevant pre-clinical biological effects of molecular targeted therapies in T-ALL/LBL
Table 2 Results of recently published clinical trials in T-ALL patients

Targeted therapy for signaling pathways

NOTCH pathway

NOTCH1 activating mutations are detected in more than 70% of T-ALL/LBL cases [5]. Gamma secretase inhibitors (GSI) can block the cleavage and activation of intracellular NOTCH1 fragments, which has been investigated in T-ALL/LBL [6]. A patient with R/R ETP-ALL was reported to have NOTCH1 mutation. Ex vivo culture of primary blasts from this patient showed high levels of activated NOTCH1, which were repressed by GSI treatment. RNA-seq documented that GSIs downregulated multiple known NOTCH target genes [7]. The patient was treated with a GSI-BMS906024 at a dose of 6 mg once per week in 4-week cycles. After two cycles of BMS906024 treatment, a complete hematologic response with a deep molecular response was achieved, without other complications except for 1 week of withdrawal due to severe thrombocytopenia during the first cycle [7]. To determine the safe and tolerated dose of BMS-906024, the clinical study NCT01363817 was initiated, which has been completed and no results been reported.

Studies on GSI combined treatment have also been initiated. Pagliaro et al. found that GSI-resistant T-ALL cells were enriched of steroid synthesis and response pathways, indicating the possibility of GSI combined with glucocorticoid [8]. Crenigacestat (LY3039478) prevents cleavage of Notch proteins. A multicenter, phase 1 study included 31 patients with T-ALL and 5 patients with T-LBL, who were treated with crenigacestat and dexamethasone [9]. Crenigacestat was administered orally 3 times a week in combination with dexamethasone at 24 mg per day on days 1 to 5 every other week in a 28-day cycle. The maximum tolerated dose of crenigacestat was 75 mg. Twenty-eight patients (77.8%) experienced one or more treatment-emergent adverse events related to the treatment. The best overall response was a confirmed response, with one patient having a duration of response of 10.51 months. Six patients achieved stable disease, 12 patients experienced progressive disease, and 17 patients could not be assessed. The median EFS was only 1.18 months among all groups [9]. These results suggested limited clinical activity of crenigacestat in R/R T-ALL/LBL.

Anand et al. observed that resistance to NOTCH inhibitors in ETP-ALL was due to activation of the phosphatidylinositol 3-kinase (PI3K) pathway, implying the rationality of GSI combined with PI3K/AKT inhibitors [10]. This was demonstrated by the synergistic inhibitory effect of GSI and buparlisib (PI3K inhibitor) on proliferation of KOPT-K1 T-ALL cells [10].

CB-103 is a highly selective and potent inhibitor of NOTCH transcription complex. Medinger et al. treated a R/R T-ALL patient (with PTEN and NOTCH1 mutations) with CB-103 plus salvage venetoclax and decitabine. The patient achieved CR within 1 week of adding CB-103 and successfully underwent allogenic stem cell transplantation (allo-HSCT). CB-103 was also administrated throughout the conditioning, transplantation, and post-allo-HSCT period to control the clone harboring the NOTCH1 mutation. Adverse events associated with CB-103 were mild. A phase I/II multicenter clinical trial of CB-103 in T-ALL/LBL is currently underway (NCT03422679) [11].

JAK/STAT pathway

The JAK/STAT pathway is also a putative actionable target for T-ALL/LBL. Senkevitch et al. showed that ruxolitinib inhibited the proliferation and activation of D1_hIL7RP1 cells (the D1 thymocyte cell line transformed with a mutant human IL-7Rα derived from a patient sequence) and prolonged the survival of mice transplanted with D1_hIL7RP1 cells [12]. Ruxolitinib was also shown to reduce the number of blasts in the blood and spleen in ETP-ALL xenograft models [13]. Jaramillo et al. reported that ruxolitinib alone induced a lasting 5 months in a T-ALL patient with cutaneous relapse [14].

JAK inhibition may restore the sensitivity to glucocorticoids in T-ALL/LBL [15]. Tofacitinib is a JAK3 inhibitor that has been reported to show synergistic effects both in vitro and in vivo with prednisolone and dexamethasone in the treatment of T-ALL with JAK3 mutations [16]. Ruxolitinib or tofacitinib was also shown to have a synergistic effect combined with the BCL2 inhibitor venetoclax in inhibiting leukemic cell growth in mouse models [12, 17]. Ruxolitinib with lysine specific demethylase1 (LSD1) inhibitor SP2509 reversed the expression of anti-apoptotic BCL2 and proapoptotic BIM proteins in the ETP-ALL cell line LOUCY [18]. In addition, the above drug combination also inhibited the growth of leukemic cells in NSG mice inoculated human LOUCY cells [18]. Govaerts et al. reported that the viability was inhibited and the apoptosis was increased in a T-ALL cell line DND-41 cells by 7-day pretreatment with ruxolitinib and a PSEN1-selective GSI MRK-560 [19]. Degryse et al. reported that tofacitinib had synergistic effects with the mitogen-activated protein kinase inhibitors selumetinib and trametinib when treating JAK3mut PDX in vivo [17]. PIM1 is one of the effectors of the JAK/STAT pathway, which is a potential actionable target in approximately 30% of T-ALL/LBL [20]. A PIM inhibitor AZD1208 with ponatinib or chemotherapy improved survival of T-ALL cells in vivo with PDX model [20, 21].

The long-term efficacy of JAK inhibition in T-ALL/LBL is unknown. The response rate, duration of efficacy, and safety of JAK inhibitors monotherapy or combined therapy in T-ALL/LBL still need to be studied in the future. The phase I/II study of ruxolitinib plus l-asparaginase, vincristine, and prednisone in adults with relapsed or refractory early T-precursor acute lymphoblastic leukemia may provide insights (NCT03613428).

Targeted therapies for kinases

Rearrangements involving ABL1 are detected in approximately 5% of T-ALL/LBL patients. The BCR::ABL1 fusion has been occasionally reported in T-ALL/LBL. Ph-positive T-ALL has been observed to be sensitive to TKIs [22]. A phase III study evaluating the efficacy of imatinib in combination with chemotherapy is also enrolling T-ALL patients with BCR::ABL1 fusion (NCT03007147).

NUP214::ABL1 is the most common ABL1 rearrangement in T-ALL [23]. In rare cases, patients may have different NUP214::ABL1 fusions at the same time and be vulnerable to relapse due to widespread genomic instability [24]. NUP214::ABL1-induced cell proliferation is dependent on the activity of the SRC family [25]. This supports the use of dual ABL1/SRC inhibitors (e.g., dasatinib) as a new potential therapeutic approach for NUP214::ABL1-positive T-ALL. A patient with NUP214::ABL1-positive T-ALL was in complete hematologic and cytogenetic remission after 3 weeks of dasatinib monotherapy. Dasatinib was well tolerated throughout the subsequent treatment and no side effects were observed [26]. Although NUP214::ABL1 was positive in only 3 of 178 patients analyzed by RT-PCR (1.68%) [24], we still encourage screening for NUP214::ABL1 in T-ALL patients to explore the benefits of dual ABL1/SRC inhibitors in combination with chemotherapy and/or allo-HSCT.

Given that dasatinib can target SRC, it may have a broader clinical application. This was demonstrated in studies in which they detected responses to dasatinib in primary cells from 12 pediatric T-ALL patients who showed no typical ABL1 kinase translocations and a lack of activity against other typical ABL family kinase inhibitors (e.g., imatinib) [27]. Another SRC family kinase, lymphocyte-specific kinase (LCK), is usually highly expressed and might be a potential therapeutic target for T-ALL [28]. A systemic pharmacology analysis indicated that activation of preTCR-LCK is the driver of dasatinib sensitivity [29]. Based on these preclinical data, patients with SRC hyperphosphorylation and increased LCK expression may benefit from dasatinib treatment even in the absence of ABL1 abnormalities. Of note, dasatinib-sensitive primary T-ALL cells exhibit high BCL-XL and low BCL2 activity and venetoclax resistance, which is closely associated with T cell developmental arrest driving differential activation of LCK and BCL2 signaling [29]. The combination of dasatinib and venetoclax (or navitoclax) may have better efficacy in T-ALL.

Targeted therapies for epigenetic changes

HDAC inhibitor

The histone deacetylase (HDAC) participates in regulating chromatin structures [30]. Chidamide, a HDAC inhibitor, inhibits the expression of type I HDACs (HDAC1, HDAC2, HDAC3, and HDAC10) and further induces acetylation of histones H3 and H4 [31]. Chidamide can downregulate the expression of cyclins, arrest the cell cycle in G0/G1 phase, and affect a variety of signaling pathways in leukemia cell lines, such as NOTCH1-MYC, thereby inhibit the proliferation of leukemia cells [32, 33]. Chidamide in combination with chemotherapy showed higher overall response rate and better PFS than historical data in the chemotherapy group in R/R T-ALL patients [34]. A prospective clinical study (NCT03564704) on the efficacy and safety of chidamide in combination with the PDT-ALL-2016 protocol in T-ALL is ongoing.

BET inhibitor

The bromodomain and extraterminal (BET) proteins are epigenetic readers that detect acetyl-lysine side chains on histones, including the H3K27ac mark and transcription factors [35]. High expression of BRD4, a member of the BET protein family, is associated with poor prognosis for pediatric T-ALL [36]. The BRD4 inhibitor, ARV-825, inhibited cell proliferation in vitro by arresting the cell cycle and inducing apoptosis [36]. It consumes BET protein levels while inhibiting the H3K27Ac-Myc pathway and reducing c-Myc protein levels in T-ALL cells, thereby interfering with cell proliferation [36]. ARV-825 prolongs survival in mice with Notch1 mutation patient-derived T-ALL [37]. JQ1, a BRD4 inhibitor, combined with GSI significantly prolonged the survival of primary T-ALL xenograft mice [38].

PRC2 acts as an epigenetic regulator and transcription repressor by writing the H3K27me3 epigenetic mark [39]. When PRC2 is lost, an epigenetic switch is triggered toward global H3K27 reacetylation, which activates BET protein-dependent transcriptional programs [40]. As a result, BET inhibitors can potentially be used to target patients with T-ALL with altered PRC2. CRLF2 overexpression is a biomarker of poor prognosis that identifies a subset of high-risk T-ALL patients, who may benefit from therapies targeting the CRLF2 pathway [41]. Upregulation of CRLF2 is considered to be involved in the leukemogenesis of ETP, and its dysregulation may be associated with JAK3 mutation [42]. Maciel et al. found that epigenetic changes induced by loss of function of EZH2 and PRC2 were important for the regulation of CRLF2 in immature T-ALL [42]. Therefore, the possibility of using BET inhibitors for the treatment of T-ALL with high expression of CRLF2 was raised. And this strategy was observed to lead to a significant reduction of CRLF2 expression in LOUCY cells [42].

Targeted therapies for apoptosis

Different T-ALL/LBL subtypes showed distinct BCL2/BCL-XL expression profiles in in vitro studies. ETP-ALL cells are most sensitive to the BCL-2 inhibitor, venetoclax, while other T-ALL subtypes are more sensitive to treatment with the BCL-XL inhibitor-navitoclax [43]. A multicenter retrospective analysis showed that 23 R/R T-ALL (6 ETP-ALL) patients treated with venetoclax and navitoclax alone or in combination had an overall response rate of 35% after 28 days of treatment [44]. In Pullarkat et al.’s study, 10 of 18 patients (55.6%) with T-ALL responded to the treatment of venetoclax in combination with low-dose navitoclax. Among them, 8 were ETP-ALL (8/12, 66.7%) and 2 were non-ETP-ALL (2/6, 33.3%) [45].

Venetoclax can be used in combination with various targeted drugs for the treatment of T-ALL/LBL. In preclinical studies, the LSD1 inhibitor GSK2879552 [18], the MCL1-specific inhibitor S63845 [46], the BET inhibitor JQ1 [47], the HDAC inhibitor chidamide [48], and the JAK inhibitor ruxolitinib [12] all showed synergistic effects with venetoclax. A recent retrospective study of venetoclax in combination with chemotherapy for R/R T-ALL/LBL showed that 10 of 13 T-ALL/LBL patients (77%) achieved CR/CRi [49]. Case series of R/R T-ALL/LBL patients treated with venetoclax and azacitidine also showed a promising result. After 1 course of venetoclax and azacitidine treatment, 4 patients achieved CR and 1 patient achieved CRi. Three patients had measurable residual disease (MRD) negative CR after completion of the treatment [50]. Three ETP-ALL patients achieved hematologic complete or partial remission after a month of treatment with venetoclax and the proteasome inhibitor bortezomib, and stable cytogenetic remission after HSCT (follow-up >8 months) was also observed [51]. A refractory TP53-mutated ETP-ALL patient was reported to be successfully treated with venetoclax and low-dose decitabine [52]. Venetoclax plus HAG regimen (homoharringtonine, cytarabine, and G-CSF) also lead to CR in an ETP patient [53]. A phase I/II study (NCT03808610) of venetoclax in combination with low-dose decitabine for T-ALL/LBL is ongoing.

Targeted therapies for miRNA

Potentially relevant candidate oncogenic miRNAs for T-ALL/LBL can be screened by NGS, which are used to explore their oncogenic mechanisms in T-ALL/LBL [54, 55]. MiR-625-5p is a candidate oncogenic miRNA in T-ALL/LBL that is associated with negative regulation of apoptosis. Overexpression of miR-625-5p in T-ALL cells may result in post-transcriptional repression of the pro-apoptotic HRK gene [56]. And inhibition of miR-625-5p increases the rate of apoptosis in Jurkat cells [56]. Combination therapy with miR-34a and doxorubicin synergistically induced apoptosis in T cell acute lymphoblastic leukemia cell line Jurkat [57].

Immunol-targeted therapies

Immunotherapy offers an excellent strategy for improving the prognosis of patients with R/R T-ALL/LBL (Table 2). Overcoming the fratricide effect during CAR T cell preparation, optimizing efficacy, mitigating toxicity, and safely combined with conventional therapies are key considerations for the introducing of immunotherapy into the clinic. Immunol-targeted therapies for R/R T-ALL/LBL are summarized in Table 3.

Table 3 Ongoing clinical trials in T-ALL patients

Monoclonal antibodies

CD38 is expressed in the malignant cells in most T-ALL/LBL patients [58]. This makes anti-CD38 monoclonal antibody an option of targeted therapy for T-ALL. The anti-CD38 monoclonal antibody daratumumab has shown substantial efficacy in preclinical models for the treatment of T-ALL, and its efficacy in clinical trials is promising (NCT05289687, NCT03384654) [59]. Daratumumab is now recommended by the NCCN guidelines for the treatment of R/R T-ALL [60]. A patient with T-ALL who relapsed after allo-HSCT was reported to achieve long-term MRD negative remission with daratumumab [61]. Of note, although CD38 remains stable during chemotherapy [58], daratumumab treatment may result in false negative CD38 and requires monitoring when used. This may be the reason for some relapses after achieving second CR post daratumumab treatment [62]. Also, CD38 monoclonal antibody may be ineffective when used repeatedly [62], although isatuximab monotherapy, a novel anti-CD38 monoclonal antibody, has not been found to have satisfactory efficacy in clinical trial (NCT02999633). No patients achieved CR and most patients experienced disease progression [63]. However, it may have good results in combination with other drugs for the treatment of T-ALL/LBL. Currently, isatuximab is being evaluated in a phase II study (NCT03860844) in combination with cytotoxic chemotherapy in children with relapsed T-ALL.

More recently, a bispecific T cell recruitment antibody against CD3 and CD38 (XmAb 18968 CD3-CD38) is currently be investigated in a clinical trial for T-ALL/LBL (NCT05038644). Two novel Bispecific T-Cell Engager (BTCEs) targeted CD43/CD3ε and CD1a/CD3ε, respectively, have been developed [64, 65]. Both BTCEs showed significant inhibition of the proliferation of T-ALL cell lines with the corresponding targets and prolonged survival in mouse models of T-ALL [64, 65].

CD7-targeted CAR T therapy

CD7 is widely expressed in T-ALL/LBL, making it a promising therapeutic target for chimeric antigen receptor-modified T cell therapy (CAR T). Multiple clinical trials of CD7 CAR T cell therapy for T-ALL/LBL are ongoing (NCT04984356, NCT03690011, NCT04033302, NCT02742727). However, unlike the rapid development of CAR-T for B cell malignancies, CD7 CAR T cell therapy for T-ALL/LBL faces extremely high challenges. First, CD7 is shared among normal T cells, malignant T cells, and CAR T cells, which raises the problem of fratricide of CAR T cells or contamination of malignant T cells in the CD7 CAR T cells [66]. Second, it is difficult to obtain sufficient numbers of normal T cells from the patient to manufacture CAR T cells. Third, it takes a long time for preparation and quality control of autologous CD7 CAR T cells.

Ye et al. blocked CD7 antigen on the surface of T cells with a free anti-CD7 antibody containing the same binding domain as CAR to avoid self-incompatibility [67]. Pan et al. reported the first in-human phase I trial of donor-derived CD7 CAR T cell therapy (NCT04689659). The results showed that the rates of ORR and CR were 95% (19/20) and 85% (17/20) [68], respectively. And the median PFS and OS were 11.0(95% CI, 6.7–12.5) months and 18.3(95% CI, 12.5–20.8) months [69]. Lu et al. reported the first-in-human, phase I trial of naturally selected CD7 CAR T cell therapy for R/R T-ALL (NCT04572308). In this study, dual CD3-positive and CD4+CD8+ selection regimen was utilized to remove malignant T cells [70]. Nineteen of 20 patients (14 R/R T-ALL and 6 T-LBL) achieved MRD negativity in the bone marrow and 5 of 9 patients achieved extramedullary CR [70]. Zhang et al. reported the first autologous CD7 CAR T cell therapy for R/R T-ALL/LBL [71]. The CR rate was 87.5% (7/8) and one T-ALL patient achieved MRD-negative CR [71]. Four patients relapsed within 1 year post CAR T cell infusion [71]. Disease relapse after autologous CD7 CAR T cell therapy remains a considerable obstacle to improve survival [72]. Early loss or depletion of CAR T cells, selection of antigen-negative clones or downregulation of target expression, and spectrum switching in leukemia all contribute to relapse after CAR T cell therapy [73]. Bridging consolidation of allogeneic HSCT in patients treated with CD7 CAR T cell therapy has shown to bring durable remission and longer survival in several studies [74, 75].

Overall, the choice of autologous or allogeneic CD7 CAR T, as well as the timing of CAR T, is open to future research.

CD5 targeted CAR T cell therapy

CD5 is expressed in the majority of T-ALL patients and is an ideal target for CAR T therapy. CD5 CAR T cells only demonstrated mild fratricide in vitro, and their ability to kill CD5 positive T cells was not affected [76]. Dai et al. developed a humanized dual-epitope CD5KO CAR T cells containing FHVH1 and FHVH3, which can bind different epitopes of the CD5 antigen. The CD5KO FHVH3/VH1 CAR T cells have exhibited stronger and longer-lasting efficacy against T-ALL cells both in vitro and in vivo [77]. Several clinical studies on CD5 CAR-T therapy are underway (NCT03081910, NCT04594135, NCT05032599).

Central nervous system leukemia (CNSL) is frequently occurred in T-ALL. CD5-IL15/IL15sushi-CAR T cells, specifically target CD5, were reported to successfully treat a T-ALL patient with CNSL. The patient had nearly 80% of blasts in his cerebrospinal fluid, which were almost exclusively CD5 positive. One week after infusion of the CD5-IL15/IL15sushi-CAR T cells, blasts in the cerebrospinal fluid decreased to approximately 2% and continued to decrease over the next few weeks. [78] These results showed that CD5-IL15/IL15sushi-CAR may be an effective treatment for T-ALL with CNS involvement.

Since CD5 is also expressed in normal T cells, CAR-CD5 treatment may not only lead to T cell depletion but also enhance autoimmune responses [79]. Therefore, it is critical to mitigate the consequences of long-term T cell aplasia through controlling the expansion and activity of CAR T cells. Wada et al. evaluated the efficacy of co-expression CD5 and CD52 CAR T cells in preclinical trials for R/R T-ALL/LBL through applying an alemtuzumab-mediated CD52 safety switch. The results showed that the cytotoxicity and the expansion of CD5 CAR T cells in vitro were both not affected [80].

CD1a targeted CAR T cell therapy

CD1a is expressed exclusively in cortical T-ALL (coT-ALL) at diagnosis and is retained at relapse. CD1a is not expressed in either CD34-positive progenitor cells or mature T cells [81]. Expressions of CD1a were detected in 33% of R/R T-ALL patients [82]. In a preclinical study, CD1a-specific CAR was shown to exert potent and specific cytotoxicity both in vitro and in vivo [81]. There was no T cell fratricide when CD1a CAR T cells were expanded [81]. Therefore, CD1a CAR T cell therapy might also be an attractive immunotherapy for R/R T-ALL.

CCR9 and CD99 targeted CAR T cell therapy

CCR9 is expressed in a high proportion of R/R T-ALL/LBL cases, but in less than 5% of normal T cells [83]. Maciocia et al. found that CAR T cells targeting CCR9 were resistant to fratricide and demonstrated potent anti-leukemic activity both in vitro and in vivo [83].

CD99 is upregulated at both transcript and protein level compared to normal T cells [84]. CD99 is highly expressed in newly diagnosed T-ALL and has been used as a marker to detect MRD in T-ALL [85]. Anti-CD99 CAR T cells showed strong anti-leukemic effect in both in vitro and in vivo experiments and showed no obvious fratricide [84]. Two clinical studies on CD99 CAR T cell therapy are undergoing (www. chictr.org.cn, ChiCTR2100046764, ChiCTR2000033989).

Combination and bispecific CAR T cell therapy

Despite the excellent efficacy, relapse is the main reason for treatment failure of CAR T cell therapy targeting a single antigen for R/R T-ALL [68, 71]. The escape or incomplete expression of antigens has led to the search for additional antigens and proposals to target multiple antigens simultaneously (Fig. 1) [86]. Georgiadis et al. developed a combined CD3 and CD7 CAR treatment strategy. They orderly removed TCR/CD3 and CD7 ahead of lentiviral-mediated expression of CARs specific for CD3 (3CAR) or CD7 (7CAR), and co-cultured edited 3CAR and 7CAR products. The 3CAR/7CAR co-cultures resulted in “self-enrichment” yielding populations 99.6% of the TCR-/CD3-/CD7- [86]. The mixed 3CAR/7CAR treatment has shown profound anti-leukemic effects both in vitro and in vivo except in the CD3-CD7- cohort, and is expected to provide deep molecular remission prior to allo-HSCT in refractory T-ALL [86]. This strategy mixed two different monovalent CAR T cells which is challenging for product manufacture, and the possible disproportionate expansion of different CAR T cells in vivo may have an impact on efficacy. In addition, only a small fraction of T-ALL cells expresses CD3. Dai et al. developed CD5/CD7 bispecific CARs based on FHVHs [87]. This tandem CARs (Tan CARs) showed excellent anti-leukemic effects even after CD5 and CD7 were knocked down [87]. However, targeting two antigens simultaneously puts T cells at increased risk of immunodeficiency. Therefore, adding a suitable safety switch or bridging to allo-HSCT worth further investigation.

Fig. 1
figure 1

Representative targets for CAR T cell therapy in T-ALL/LBL. (a) T cells are genetically modified to express chimeric antigen receptor (CAR), which target antigens expressed on the surface of T-ALL cells, including CD7, CD5, CD1a, CD99, CCR9, etc., thereby killing T-ALL cells. (b) The bispecific CAR structure comes in two forms, tandem CAR and dual CAR. Tandem CAR shows superior therapeutic efficacy and is independent of the order of CD5CD7. (c) Combined CAR T cell therapy. CAR T cells are knocked down for CD3CD7 and then lentivirally transfected to form 3CAR (CD3-CD7-) and 7CAR (CD3-CD7-). CAR T: chimeric antigen T cell

CAR-NK immunotherapy

Unlike CAR T cells, CAR-NK cells do not cause immunodeficiency due to shared antigens between normal and malignant T cells [88]. CAR-NK cells have a lower risk of off-target effect. In addition, CAR-NK treatment also showed lower risk of cytokine release syndrome and neurotoxicity compared to CAR T cell therapy, because the former secreted cytokines such as GM-CSF, IL-3, and IFN-γ, whereas the latter secreted high levels of pro-inflammatory cytokines [89, 90]. Unresolved challenges for CAR-NK include the optimal source of NK cells, the optimal transduction vector system, the most biologically relevant signaling domains for CAR activation, and the preservation and persistence of CAR-NK cells [91]. It was shown that a subset of NK cells from CMV donors enhanced the ability to target T-ALL, especially when co-expressed with CD57 [92]. For CD5 NK-CAR cells, cells constructed from the intracellular structural domain of the NK cell-associated activation receptor 2B4 had greater anti-CD5 positive malignancy capacity than cells using the T cell–associated activation receptor-4-1BB [93]. Both CD7 CAR-NK cells and CD5 CAR-NK cells were used in preclinical studies and the results showed high efficacy in T-ALL [89, 94]. A clinical trial of CD5-CAR-NK is currently underway (NCT05110742).

Immune checkpoint inhibitors

Immune checkpoint inhibitors such as PD-1 act by blocking T cell downregulation pathways to enhance the activity of endogenous T cells or potentially CAR T cells. Xu et al. identified the PD-1 molecule as a marker of T-ALL leukemia stem cells (LSCs), and proposed a combination therapy scheme of targeting rapidly proliferating leukemia cells with conventional chemotherapy and targeting LSCs with PD-1 blocking [95]. Therefore, the role of PD-1/PD-L1 inhibition on the adjuvant or salvage strategy for T-ALL warrants to be explored. A phase 2 trial uses monoclonal anti-PD1 antibody pembrolizumab for the treatment of R/R T-ALL. However, the results for this trial were not satisfactory [96].

Conclusions

T-ALL/LBL is a highly aggressive and heterogeneous malignancy. Relapse occurs in a high percentage of patients when treated with conventional chemotherapy. Understanding the genetic aberrations at disease onset and recurrence using next-generation sequencing provides valuable clues for the precision diagnosis and treatment. Novel molecular and immune targeted therapies have shown tremendous efficacy in R/R T-ALL/LBL, and the application of these powerful therapeutic strategies will be translated into an improvement of the prognosis of T-ALL/LBL.