CAR T-cell approaches for the treatment of patients with B- and T-cell lymphomas

As of February 2020, there are two approved chimeric antigen receptor (CAR) T-cell therapies for the treatment of relapsed or refractory (R/R) large B-cell lymphoma, axicabtagene ciloleucel and tisagenlecleucel. Other CAR T-cell products are in clinical development, including lisocabtagene maraleucel, which is awaiting approval. During the 2^nd CAR T meeting, Sitges, ES, Martin Pule from University College London, UK, presented an overview of the CAR T-cell approaches for B- and T-cell lymphomas.¹

CAR T-cell therapies for diffuse large B-cell lymphoma (DLBCL)

Although the two approved CAR T-cell therapies targeting CD19 are highly effective, not all patients respond, and others have a relatively short duration of the response, with CD19 antigen loss and CAR T exhaustion due to programmed death ligand 1 (PDL1) upregulation, cited as potential relapse mechanisms.^2,3 In addition, these therapies are associated with unique toxicities of cytokine release syndrome (CRS) and neurotoxicity. Therefore, researchers are aiming to design products that could be manufactured on a large scale, have a high sustained complete response (CR) rate, and a safety profile that will enable treatment in the outpatient setting.

AUTO3¹

One of the novel CAR T therapies is AUTO3, which contains two independent CARs targeting CD19 and CD22 antigens delivered in a bicistronic vector, with OX40/4BB co-stimulatory domains and humanized binders. The therapy is being evaluated in patients with R/R DLBCL in a single-arm, open-label multi-center phase I/II study, ALEXANDER (NCT03287817). The phase I dose-escalation part of the study has a rolling 6 design with patients receiving CD19/CD22 CAR T-cell infusions at doses between 50x10⁶ and 900x10⁶ cells/kg after preconditioning with fludarabine and cyclophosphamide with or without pembrolizumab. The phase II part follows Simon’s 2-stage optimal design:

• Cohort 1 – patients with DLBCL not otherwise specified (NOS), high-grade B-cell lymphoma or DLBCL transformed from follicular lymphoma (FL) with >2 prior therapy lines
• Cohort 2 – primary mediastinal large B-cell lymphoma or  DLBCL transformed from other than indolent non-Hodgkin lymphoma (NHL) with >2 prior therapy lines

The product delivery platform is semi-automated and allows for parallel processing of multiple samples.

Results (data cut-off: 21 Jan 2020)¹

AUTO3 product was successfully manufactured for all patients

Adverse events of special interest included:

• Grade 1/2 CRS in 38.9% of patients (7/18), no grade 3 or 4 events
• Grade 3/4 neurotoxicity in 5.6% of patients (1/18 patients receiving the lowest dose without pembrolizumab)
• Grade 4 lung infection caused by para-influenza virus-induced pneumonia in 5.6% of patients (1/18)

No prophylactic measures were used. Two patients received tocilizumab for CRS and one patient received steroids for neurotoxicity.

Low levels of cytokines were consistent with incidences of low-grade CRS.

Overall response rate (ORR) among all patients was 61% (11/18), including a complete response (CR) rate of 44% (8/18):

• Amongst patients who received a dose of 450x10⁶ cells/kg, 71% (5/7) patients responded with a CR rate of 57% (4/7)

Importantly, CR is ongoing in 87% of patients (7/8) after a median follow-up of 6 months (1–18).

AUTO3 in combination with a single dose of pembrolizumab on Day 1 is being evaluated further after demonstrating positive results:

• Best CR at 55% with all 7/7 responses ongoing at a median follow-up of 2 months (1–12)
• No CRS≥ grade 3, or neurotoxicity of any grade

Next generation AUTO3 (AUTO3NG)¹

Researchers are investigating whether the AUTO3 function in the immunosuppressive environment could be improved by the addition of truncated dSHP2 modules to block PD-1/PD-L1 axis to restore T-cell function and dnTBR2 to prevent the activation of the immunosuppressive TGFβ/SMAD signaling pathway. This approach is designed to inhibit TGFβ signaling and simultaneously block multiple inhibitory receptors involved in immune checkpoint signaling, including 2B4, BTLA, TIGIT, and CTLA4.

Moreover, different costimulatory combinations are being tested to further optimize the cytotoxic and proliferative capability of AUTO3. Moreover, AUTO3NG shows enhanced tumor clearance at suboptimal doses, in particular, the CARs with the costimulatory domains OX40z/41BBz and 41BBz/CD28z.

CAR T-cell therapies for T-cell lymphoma¹

T-cell lymphomas are a heterogeneous group of aggressive diseases, with limited therapy options and a poor prognosis.⁴ Therefore, there is an urgent need for new, targeted therapies, in particular for patients with R/R disease. CAR T therapies are one of the most promising novel cancer treatments but so far, the development of CARs for T-cell lymphoma has proved to be more problematic than for B-cell lymphomas. One of the reasons for that is the lack of malignant T-cell-specific targets. The investigated approaches include targeting CD3, CD4, CD5, CD7, CD30, CD37, TRBC1, and CCR4.⁵

In this talk, the speaker focused on TRBC1 and TRBC2 as targets for peripheral T-cell lymphoma (PTCL). He described the difficulties of targeting TRBC1, which are caused by the similarity to TRBC2, as well as the fact that peripheral blood T-cells contain a mixture of cells expressing both of the T-cell receptors.

AUTO4¹

Next, he outlined the strategy that led to the identification of a TRBC1 antibody and the recognized epitope, which was found to be expressed by 22 out of 56 (38%) screened patients with different types of primary T-cell cancers.⁶ With some of the subtypes having an even higher frequency of positivity:

• Adult T-cell leukemia/lymphoma – 2/2 (100%)
• PTCL, NOS – 8/18 (44%)
• Anaplastic large cell lymphoma – 5/12 (42%)
• Angio-immunoblastic T-cell lymphoma – 2/5 (40%)

After validating the target, researchers generated AUTO4, a humanized anti-TRBC1 CAR, which showed selectivity against TRBC1 and anti-tumor activity in an in vivo animal model.¹

The preclinical results demonstrated the utility of targeting TCR β chain isoforms that in theory could be effective against a range of different subtypes of T-cell lymphoma. A single arm, open label, multi-centre, phase I/II study (NCT03590574) evaluating the safety and efficacy of AUTO4, in patients with R/R TRBC1 positive T-cell NHL, including PTCL NOS, angioimmunoblastic T cell lymphoma, and anaplastic large cell lymphoma, is currently ongoing.

AUTO5¹

Next, the speaker presented data on AUTO5, an anti-TRBC2 CAR T-cell therapy, able to specifically target and eliminate TRBC2 positive cells in an in vivo model. AUTO5 is progressing towards receiving an investigational new drug status.

Accompanying diagnostic test¹

As mentioned above, it is estimated that around 40% of T-cell lymphomas are TBCR1 positive and around 60% are TBCR2 positive. Therefore, it is important to be able to stratify patients into those that could potentially respond to AUTO4 or AUTO5. In the final stage of the presentation, Martin Pule described the development of a diagnostic test for such patient stratification. The test is based on a next-generation sequencing clonality assay.

In summary, AUTO4 and AUTO5 are designed to avoid severe immunosuppression associated with targeting T-cells in patients with T-cell lymphomas. The initial report from the AUTO4 clinical trial is expected in the second half of 2020. The start of the AUTO5 clinical study is also expected at the end of 2020, and the development of the accompanying diagnostic test is on schedule.