Beat pediatric ALL MRD: CD28 CAR T and transplant

In this issue of Blood, Curran et al report clinical trial results evaluating a second-generation CD19-directed chimeric antigen receptor (CAR) T-cell therapy in pediatric and young adult (AYA) relapsed or refractory B-cell acute lymphoblastic leukemia (B-ALL).¹  Autologous CAR T-cell immunotherapies are revolutionizing the treatment of B-cell malignancies. CAR T-cell treatment involves genetic engineering of a patient’s T cells by insertion of a CAR construct, which upon expression, recognizes CD19 on neoplastic cells and triggers T-cell activation and tumor killing.²  The paper is notable for several reasons. First, it confirms in a pediatric population the importance of intensive lymphodepletion before CAR T-cell infusion to achieve responses. Second, in children treated with a CD28-based CAR T cell achieving minimal residual disease (MRD)⁻ status, consolidative allogeneic transplant leads to acceptable long-term disease-free survival. Lastly, this report provides key additional clinical data on a CD28-based CART19 product whose commercial development was halted due to cerebral edema deaths during a separate pharma-sponsored trial.

Two CD19 CAR T-cell products, axicabtagene ciloleucel (axi-cel)³  and tisagenlecleucel (tisa-cel),⁴  are currently approved by the US Food and Drug Administration: tisa-cel for pediatric/young-adult patients with B-ALL and both axi-cel and tisa-cel for adult patients with diffuse large B-cell lymphoma. Each product uses a different costimulatory domain in the CAR structure to enhance T-effector functions: tisa-cel relies on 4-1BB that drives progressive CAR T-cell stimulation and persistence, whereas axi-cel uses CD28 to achieve stronger early T-cell activation with less long-term persistence. This is the first report of pediatric patients having received this CAR T-cell therapy, a CD28-CART similar to axi-cel, but differing in the anti-CD19 recognizing domain (SJ25C1 vs FMC63 scFv). Comparison across the major trials testing CD19 CAR T cell in pediatric/AYA ALL is extremely difficult due to differences in study design, patient characteristics, CAR construct, and manufacturing. However, there are striking similarities in complete response rates, despite key differences in disease burden and the number of patients undergoing allogeneic transplant before or after CAR T-cell therapy (see table). In this phase 1 trial, 25 pediatric/AYA patients with MRD⁺ (n = 15), or morphologic residual ALL (n = 10), were treated at 2 institutions. Although numbers are small, lymphodepletion with high-dose cyclophosphamide (Cy)3 g/m² was associated with higher response rates (93%, n = 14) compared with lower-dose Cy ≤1.5 g/m² (50%, n = 10), with similar toxicity. These results confirm the findings of others,^5,6  and although exact mechanisms in this case are unclear, more intense conditioning may increase homeostatic cytokines, decrease suppressive cell populations, remove immune cells that could reject CAR T cells, and cause direct tumor killing. There is ongoing debate about the optimal regimen for lymphodepletion, with many trials employing fludarabine (Flu) and Cy. This report suggests that if single-agent Cy is used, a higher dose is warranted.

Over one-half of the patients in this trial had MRD at the time of CAR T-cell therapy, and their outcome was superior to patients with morphologic disease. Similar trends were seen in other studies using both CD28-^6,7  and 4-1BB–based⁸  CAR T cells. Treating patients with lower tumor burden could possibly improve efficacy but also decrease toxicities. Interestingly, in this study, severe CRS was observed in a relatively limited number of patients (16%) possibly due to the low disease burden in most patients, although differences in the grading and management of CRS may likely have contributed.

The CAR T-cell product used in this trial was initially generated at Memorial Sloan Kettering Cancer Center and then developed by Juno Therapeutics as JCAR015. In 2017, JCAR015 clinical development was halted due to cerebral edema and death in 5 adult patients with B-ALL (although all under 30 years old).⁹  In the present trial, there was 1 cerebral edema event; however, the patient improved with aggressive management. Although there were no deaths due to neurotoxicity, the severe neurotoxicity rate was higher as compared with other CAR T-cell trials in pediatric/young adult patients with B-ALL, although lower than the adult ALL trials using the same construct (see table). It is still not clear why this specific construct caused high rates of cerebral edema, although patient, disease, and CAR T-cell product characteristics likely play a cumulative role. Recent studies suggest that CD8⁺ dose, cytokine production by T cells, early and rapid CAR T-cell proliferation, non-Ph⁺ type B-ALL,¹⁰  <30 years old, fewer prior therapies, intensive bridging therapies, high-intensity Flu/Cy, and high interleukin-15 levels may all contribute to endothelial breakdown, blood-brain barrier compromise, and the risk for cerebral edema.⁹ 

Last, significant debate remains in the field about the necessity of consolidation for patients with B-ALL in remission after CAR T cell. This study suggests that allo-SCT is a viable option for young patients with B-ALL without prior SCT receiving CD28-CART19. However, generalization beyond the specific conditions in this trial must be carefully considered. All patients were young; all had achieved MRD⁻ status after CAR T cell, and almost all had not had a prior allogeneic transplant. Furthermore, the durability of remissions without transplant with this CAR T cell is unknown. Multiple factors must be considered before choosing to perform consolidative transplant, including CAR costimulation, disease present, risk of CD19-negative escape, achievement of MRD after CAR T cell, age, and transplant features, such donor type, conditioning intensity, and graft-versus-host disease prophylaxis. These features must be further examined via observational and prospective research. Interestingly, in adult patients with B-ALL treated with the same CAR T-cell product, allo-SCT did not show clear improvement of survival.⁷  Alternatively, a different CD28-based CAR T-cell in pediatric/AYA B-ALL demonstrated better outcomes in patients receiving allo-SCT consolidation.⁶  The field should consider how age, and age-related disease differences, may impact the safety profile and efficacy of individual CAR T-cell therapies and subsequent treatments.

In summary, despite relatively few patients, this study implies that patients with lower tumor burden and treated with intensive lymphodepletion have better outcomes. Moreover, it verifies in the setting of CD28-based CART19 in pediatric B-ALL, that consolidation with allo-SCT can lead to acceptable results. Although this product is not being further developed, other CD28- and 4-1BB–based CART19 products are in continuous development, and it is likely that selection of patients with low tumor burden, and the use of intensive lymphodepletion, might increase responses rates while reducing toxicities.