Donor-derived CD7 CAR-T in T-ALL/LBL: promise and pitfalls Free

In this issue of Blood, Pan et al report high efficacy of donor-derived, CD7-targeted chimeric antigen receptor T-cell therapy (CD7 CAR-T) in both pediatric and adult patients with relapsed/refractory (R/R) T-cell acute lymphoblastic leukemia/lymphoblastic lymphoma (T-ALL/LBL).¹ 

Significant progress has been achieved in T-ALL/LBL. This improvement is mostly attributed to optimized frontline regimens and enhanced risk stratification incorporating genetic profiling and measurable residual disease (MRD). Consequently, outcomes for newly diagnosed T-ALL now approach those observed for B-cell acute lymphoblastic leukemia (B-ALL).² However, the prognosis of patients with R/R T-ALL remains dismal,² underscoring a critical unmet clinical need.

Although immune and cellular therapies have revolutionized B-ALL treatment, their application in T-ALL therapy has been significantly more challenging. This difficulty is largely due to shared surface antigens between malignant and normal T cells, the same cells that are harnessed in immune-based therapies such as CAR-T. This overlap creates obstacles in distinguishing malignant T cells from healthy ones during leukapheresis, increasing the risk of product contamination with malignant cells. In addition, targeting antigens that are expressed on both effector and malignant T cells can result in fratricide, a phenomenon in which CAR T cells attack each other, thereby compromising therapeutic efficacy.

Innovative CAR engineering has sought to overcome these challenges through several strategies. First, to avoid CAR T-cell contamination, investigators have explored the use of donor-derived T cells for CAR-T transduction and manufacturing.^3-5 Although this approach mitigates the risk of product contamination, it introduces the potential for graft-versus-host disease (GVHD) in HLA-mismatched recipients. Second, to address this issue, allogeneic CARs now incorporate T-cell receptor gene editing or deletion, thereby preventing host antigen recognition and diminishing GVHD risk.^3,4 Finally, to circumvent fratricide, various strategies have been explored to eliminate target genes on CAR-expressing T cells.^3-6 

Here, Pan et al report results from a single-center phase 2 trial investigating donor-derived CD7 CAR-T for R/R T-ALL, incorporating a combination of these strategies.¹ The study enrolled 70 pediatric and younger adults, among which 55 were ultimately infused with CD7 CAR T cells (dose = 1 × 10⁶/kg to 5 × 10⁶/kg) after lymphodepletion (LD) with fludarabine and cyclophosphamide. Patients who had failed earlier hematopoietic cell transplantation (HCT) received CD7 CAR-T using T cells manufactured from their original HCT donor (arm A; n = 38). The remaining patients were transplant-naïve and received CD7 CAR-T using T cells from related donors (arm B; n = 17). Patients were heavily pretreated with a median of 4 prior lines of therapy and 69% of them had extramedullary disease (EMD) before LD. Most donors (85%) were haploidentical.

Donor-derived CD7 CAR T cells were effective at clearing T-ALL/LBL, with 84% achieving complete response (CR; n = 13) or CR with incomplete count recovery (n = 33) without detectable MRD. Responses were decreased among patients with EMD. Despite encouraging early responses, limited median event-free survival (EFS) and overall survival (OS) of 5 and 8.5 months, respectively, were attained. However, among the 18 (33%; cohort A = 7, cohort B = 11) patients who underwent remission consolidation with allogeneic HCT, median EFS (14.2 months) and OS (20.2 months) were significantly superior. Antigen escape was identified as a major mechanism of CAR resistance with 50% of relapses occurring with CD7 loss.

Although the early CD7 CAR T-cell–mediated toxicity was mild and manageable, with 11% of the patients having grade 3 cytokine release syndrome and no high-grade neurologic toxicity, late toxicity proved problematic. Moreover, 10 (35%) of the patients who did not undergo consolidative HCT, died of complications in the absence of relapse, including 5 from infection, 2 from thrombotic microangiopathy, 2 from pulmonary GVHD, and 1 from hepatic failure. Prolonged cytopenias were frequent and may have contributed to the overall high rate of infectious complications.

Therefore, the donor-derived CD7 CAR T-cell approach described in this trial represents a promising therapeutic option for remission in patients with advanced T-ALL/LBL in the absence of alternative treatments. For some patients, CD7 CAR-T may present an otherwise unattainable opportunity to successfully bridge to curative allogeneic HCT. However, the broader application of this strategy is limited by several key challenges, including logistical barriers to suitable donor identification, short remission duration in the absence of subsequent HCT, and undue risk for late and serious toxicities. Further research is required to establish CD7 CAR-T as a stand-alone, safe, and curative cellular therapy for T-ALL.

Accordingly, the field is rapidly advancing to meet this challenge. Reports of base editing of allogeneic CD7 CARs to eliminate TRBC, CD7, and CD52⁵; autologous CD7 CARs modified to block CD7 protein expression and downregulate CD7⁶; and off-the-shelf CD7 CARs using CRSPR-Cas9 editing to delete CD7 and TRAC⁴ have all shown promising results. In addition, ongoing early-phase investigations are exploring CAR-T directed against other T-cell antigens such as CD1a, CD2, CD5, and CD21, offering promising avenues for the treatment of patients who relapse after CD7-directed therapy.

The advent of CD7 CAR-T has generated long-awaited enthusiasm for extending cellular therapy to patients with T-ALL/LBL. However, cellular immunotherapy for T-ALL still lags far behind the incredible progress observed in B-ALL. Future efforts can improve outcomes by focusing on extending durability, reducing toxicity, and addressing CD7-negative relapses, an increasingly recognized mechanism of CD7 CAR resistance.

Conflict-of-interest disclosure: I.A. served as a consultant for Autolus, Amgen, Pfizer, AstraZeneca, Kite, Jazz Pharmaceuticals, Syndax, AbbVie, Takeda, Servier, Acentage Pharma, and Adaptive Biotechnologies and received research funding from Jazz Pharmaceuticals and AbbVie. L.M. served as a consultant for Kite, Autolus, Incyte, and Astellas and received research funding from Wugen, Adaptive Biotechnologies, and Kite.