Delayed Immune Reconstitution and Increased Infections in Allogeneic Transplant Recipients with B Cell Malignancies Treated Previously with CAR T-Cell Therapy

Background: Allogeneic hematopoietic stem cell transplant (alloHCT) and chimeric antigen receptor (CAR) T cell therapy remain potentially curative for several relapsed/refractory B-cell non-Hodgkin lymphomas (B-NHLs) and acute lymphoblastic leukemia (B-ALL). However, their success is compromised by ensuing impaired immune reconstitution and infections. Post-CAR T cytopenia and hypogammaglobulinemia are associated with a high risk of infections. Similarly, impaired immune reconstitution related to conditioning intensity, prophylaxis, and/or treatment of graft-versus-host disease (GVHD) results in increased infections after alloHCT. Whether the risk of infections is synergistically increased among recipients of CAR T-cell therapy followed by alloHCT remains unknown.

Methods: We studied 16 consecutive recipients of alloHCT after prior CAR T-cell therapy for B-NHL (N=10) and B-ALL (N=6) treated at our institution between 2015–2023. To estimate the rate of infections for each patient at risk, we calculated infection density (ID) as individual infectious events per 100 patient days within one year post-alloHCT. Laboratory and clinically diagnosed bacterial, fungal, and viral infections requiring treatment were included, whereas fever of unknown origin was an exclusion. Non-relapse mortality (NRM), disease-free (DFS), and overall survival (OS) estimates were calculated by the Kaplan-Meier method. Immune reconstitution parameters included serial serum immunoglobulins, CD19+ B cells, natural killer (NK) cells, CD4+, and CD8+ T cells, assessed by flow cytometry.

Results: Of 10 patients with B-NHL, 7 had large B-cell lymphoma, 1 marginal zone lymphoma, 1 follicular lymphoma, and 1 mantle cell lymphoma. The median patient age at alloHCT was 60 years (range, 22-72); 56% were male. CAR T-cell therapy included axicabtagene ciloleucel (44%), brexucabtagene autoleucel (44%), tisagenlecleucel (12%) and an experimental allogeneic CAR T product (12%). The median time from CAR T-cell therapy to alloHCT was 14 months (range, 2 – 29). Reduced-intensity conditioning (RIC) was used in 75% of patients, while all study patients received peripheral blood stem cell grafts (median 6.5x10⁶ CD34+ cells/kg). Donor type included matched unrelated (50%), haploidentical (38%), matched related (6%), and mismatched unrelated (6%) donors. The conditioning regimens included fludarabine (Flu)/cyclophosphamide/total body irradiation (TBI) (N=8, 50%), Flu/melphalan (N=4, 25%), Flu/busulfan 5300 (N=2, 12.5%), and Flu/TBI 1320 cGy (N=2, 12.5%). Post-transplant cyclophosphamide (PTCy)-based regimen (81%) was the predominant GVHD prophylaxis used, while 19% of patients received tacrolimus and methotrexate.

The ID was 2.9 within the first year (3.1 within the first 100 days) after alloHCT consisting of mainly bacterial (1.6), followed by viral (0.7) and fungal (0.6) etiologies. Bacterial infections tended to occur within the first 100 days (1.7) after alloHCT and were uncommon after day 100 (0.3). Our total infection frequency exceeded the internal historical benchmark (Khimani, TCT 2021) from the cohort of alloHCT recipients without prior CAR T (ID of 2.9 vs 0.5 per 100 patient days, p<0.0001). Similarly, our estimate for bacterial ID of 1.7 per 100 patient days was higher than the CIBMTR registry benchmark of 0.88 (Ustun, BMT 2024). Median IgG serum concentration at the first post-HCT check-up (median, 41 days) was 303 mg/dL (range, 179 – 601). Other immune reconstitution parameters were similarly lower within a median follow-up of 89 days post-transplant with the levels of CD4+ T cells of 0.17 K/uL (range, 0.02 – 0.40), CD8+ T cells of 0.19 K/uL (range, 0.01 – 2.55), NK cells of 0.14 K/uL (range, 0.092 – 0.326), and CD19+ B cells of 0 K/uL (range, 0 – 0.01). Absent CD19+ B cells in our study contrasted with average detectable levels of over 50 cells/uL within 90 days post-HCT compared to our historic benchmark. NRM, DFS, and OS at 1 year were estimated at 19%, 68%, and 75%, respectively in our cohort.

Conclusions: Our pilot case series of largely RIC and PTCy-based alloHCT after prior CAR T-cell therapy show an increased risk of infections (predominantly bacterial) in patients with B-NHL and B-ALL. Impaired post-HCT immune reconstitution with profound depletion of CD19+ B cells may account for an increased risk of infections. The next phase of conducting a matched case-control study is currently underway.