Higher Rates of Severe Infection and Persistent Cytopenias in Long-Term CAR19 Responders Than after Autologous HCT: A Single Institution Study of 139 Subjects

Abstract

Introduction: Chimeric antigen receptor (CAR) T-cell therapy is a leading treatment for relapsed or refractory aggressive and indolent lymphomas. One major adverse effect of CAR T-cell therapy is prolonged risk for infection. To quantify infectious risk after CAR, we performed a single center analysis of patients after anti-CD19 CAR T-cell therapy axicabtagene ciloleucel (axi-cel, n=78) and compared these with chemo-responsive lymphoma patients concurrently undergoing autologous hematopoietic cell transplant (HCT) with cyclophosphamide, carmustine, and etoposide conditioning (n=61). Methods: All 139 patients were treated between 09/2017 and 10/2021, with 37/78 CAR and 32/61 HCT patients treated after the first report of COVID-19 in California (02/2020). Median follow up was 718 days (range 175-1650 days). We included only patients without documented relapse of their lymphoma to avoid the confounding impact of lymphoma relapse on study endpoints. There was no significant difference in age or sex between groups. The CAR-treated cohort had fewer non-CAR prior lines of therapy (mean 3.0 vs mean 3.6, p=0.014, Wilcoxon rank sum test). Nearly all patients in the CAR group (76/78) were treated for large B-cell lymphoma (LBCL) entities, with 2/78 having follicular lymphoma with a prior LBCL history. Similarly, a minority of patients in the HCT group were treated for mantle cell lymphoma (4/61) and follicular lymphoma (5/61). Of the CAR patients 19/78 had an HCT prior to CAR, and 1/61 HCT patients had prior CAR treatment. For each patient we catalogued and graded post-treatment infections after day +30 (D+30) to focus on post-discharge events. Endpoints analyzed were any infection as well as severe infections (grade ≥3). Death and treatment for a second malignancy were considered competing events. We additionally catalogued blood count measurements including hemoglobin, platelets, white blood cell count (WBC), absolute neutrophil count (ANC) and absolute lymphocyte count (ALC) at regular intervals from pre-treatment through D+800. There was no difference in baseline WBC or ANC between groups. CAR-treated patients had a higher baseline ALC and hemoglobin while HCT-treated patients had higher baseline platelets (p<0.05 for all values, Wilcoxon rank sum test). Results: There was no difference in overall survival between groups (p=0.18, log-rank test). We performed cumulative risk regression considering infections for 1200 days starting at D+30. When considering all infections there was no significant difference in the total incidence of infections between the CAR and HCT group. However, there was increased risk in CAR-treated patients for severe (grade >= 3) infections (Figure 1A, HR 2.4, CI 1.2-4.78, Gray's p=0.013). The increased risk for severe infection was sustained when considering only infections that occurred greater than one-year post-transplant (HR 3.06, CI 1.14-8.21, Gray's p=0.021). The increased risk for severe infection in the CAR-treated cohort remained if patients with prior HCT were excluded, or if COVID-19 cases were excluded. We used linear mixed effects modeling to assess cytopenic risk longitudinally from D+28 to D+800. Age, sex, and non-CAR lines of therapy were also included in the multivariate model. CAR-treated patients demonstrated significantly reduced WBC, ANC, and ALC relative to patients treated with HCT (Figure 1B, p-value <5E-7, <0.05, and <5E-5 respectively, restricted maximum likelihood with t-test using Satterthwaite's approximation). These values remained significant if CAR patients with prior HCT were excluded from analysis. WBC and ANC were significantly reduced relative to pre-lymphodepletion baseline through D+365 (p<0.05, Wilcoxon signed-rank test). There was no difference in hemoglobin level in CAR-treated vs HCT patients using the same methods. Platelet counts in CAR-treated patients resolved to the same level as following HCT by the end of the follow up period. Conclusion: Lymphoma patients without relapse after axi-cel CAR T-cell therapy have greater risk for severe infections than a concurrent group of patients treated with autologous HCT. This risk is associated with prolonged reduction in WBC, ANC, and ALC but not platelets or hemoglobin. Clinical surveillance for infections in these immune compromised CAR-treated patients therefore remains critical, with role of additional prophylaxis strategies needing further investigation. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal