Low Disease Burden Pre-CAR-T Cell Therapy for Children and AYA with B-ALL Is Associated with an Impaired Outcome When Obtained through Intensification of the Bridging Therapy

Rationale. The role of bridging therapy, i.e., treatment between apheresis and lymphodepletion (LD) preceding tisagenlecleucel (tisa-cel) infusion for relapsed/refractory B cell acute lymphoblastic leukemia (R/R BCP-ALL) is still unclear. We report a retrospective study on our experience on 81 children, adolescents, and young adults with R/R BCP-ALL who underwent apheresis with an intent to receive tisa-cel infusion, between March 2016 and October 2020.

Methods. Bridging therapy intensity was classified into two intensity groups. A high intensity bridging (HIB) regimen, used in 23 patients (pts), was defined as the use of more than 4 intravenous anti-leukemic drugs per week and/or the need of secondary intensification of bridging therapy, the latter being the case in 19 pts (83%). A low-intensity bridging (LIB), in general a combination of vincristine, steroids +/-asparaginase, was used in 58 pts. The aim of the study was to evaluate the impact of HIB versus LIB on tisa-cel efficacy. Factors associated with a failure to infuse tisa-cel were also explored.

Results. Patients had a median age of 15 years (range: 1-19). Tisa-cel indication was a primary refractory disease for 5 pts (6%), a first post-allo-HSCT relapse for 47 pts (58%) and a second relapse or more for 29 pts (36%). Patients who received a HIB therapy were more likely to have high-risk genetic features (39% vs 12%, p= .01) and/or a primary refractory disease (17% vs 2%, p=.02). Grade 3+ non-infectious and infectious adverse events during bridging therapy were also more likely in this group, albeit in a non-statistically significant way (35% vs 16 %, p= .07 and 52% vs 34 %, p= .21; respectively).

Nine patients (11%) were not infused (disease progression: 3 pts; toxicity: 5 pts (sinusoidal obstruction syndrome: 1 pt, encephalopathy: 2 pts, fungal infection; 1 pt, hemoptysis: 1 pt); and CD19 negativation: 1 pt). A high pre-apheresis tumor burden ( p=.01), the occurrence ( p=.01), and the number of grade ≥ 3 adverse events during bridging period ( p<.01) were associated with a higher risk of tisa-cel non-infusion, while HIB was not ( p=.73). The 72 remaining patients (89%) finally received tisa-cel infusion, 52 (64%) after a LIB therapy and 20 (25%) after a HIB therapy. Despite a comparable BM disease burden at apheresis between the LIB and the HIB groups, patients presenting a high disease burden prior to LD (bone marrow (BM) blasts > 50%) were more frequent in the HIB group ( p< .01). When compared to patients who received a LIB, patients treated with a HIB had a worst, albeit non-statistically different cumulative incidence of relapse (CIR) (SHR 1.94, 95% CI [0.95-3.91], p=.07), an inferior 18-month event-free survival (EFS) (26.9% vs 54.1%; p=.02) and a lower 18-month overall survival (OS) (50.9% vs 88.8%; p<.01). Considering a threshold at 50% BM blasts, patients who presented a high disease burden prior to LD had poorer early response rates (D28 minimal residual disease (MRD) negativity: 85% vs 69%, p< .01), a higher, albeit non-statistically different CIR (SHR 2.27, 95% CI [0.98-5.27], p=.06), an inferior 18-month EFS (26.9% vs 51.3%; p=.02) and a lower 18-month OS (22.6% vs 91.5%; p<.01).

We then asked if the bridge intensity, aimed at achieving a low disease burden before LD, could influence patient outcomes. Median percentages of BM blast infiltration were similar after LIB (0%) and HIB therapy (1%, p=.87) in patients with ≤50% BM blast at LD. In these patients, a prior HIB regimen was associated with a higher rate of detectable MRD at D28 (40% vs 8%, p=.02), an inferior 18-month EFS (18.0% vs 57.5%; p=.02) and a higher, albeit non-statistically different CIR (SHR 2.05, 95%CI [0.87-4.88], p=.1). The 18-month OS did not differ between groups (74.0% vs 94.6%; p=.15), suggesting that these patients may respond to further lines of treatment after relapse. The outcome of patients presenting a low disease burden at LD after a HIB regimen was thus similar to those with a high disease burden at LD.

Conclusion. This study confirms the prognostic value of a high disease burden pre-LD but also reveals that a low disease burden prior tisa-cel therapy is associated with an impaired outcome when obtained through intensification of the bridging therapy. Additional post-tisa-cel therapies should be considered in this group of patients.