IVIG prevents infections without affecting mortality in DLBCL patients treated with CAR T-cell or bispecific antibody therapy: A real-world propensity-matched study

Abstract

Abstract Background Chimeric antigen receptor T-cell (CAR T) therapy and bispecific antibodies (BSAB) have transformed the treatment of relapsed/refractory diffuse large B-cell lymphoma (DLBCL). However, the prolonged B-cell aplasia associated with these immunotherapies leads to hypogammaglobulinemia and heightened susceptibility to infections. While intravenous immunoglobulin (IVIG) is commonly used to mitigate infection risk, robust real-world evidence regarding its impact in this setting is limited. We conducted a large, multicenter, real-world study to assess the clinical outcomes associated with IVIG in DLBCL patients treated with CAR T or BSAB therapies. Methods We utilized the TriNetX U.S. Collaborative Network, a federated database comprising 67 healthcare systems, to identify adult patients with DLBCL treated with CAR T-cell therapies (axicabtagene ciloleucel, tisagenlecleucel, lisocabtagene maraleucel) or BSABs (glofitamab, epcoritamab, mosunetuzumab) from 2013 to 2023. Patients were grouped based on receipt of IVIG within one year following immunotherapy initiation. Propensity score matching (1:1) was performed based on demographics and comorbidities, resulting in two balanced cohorts of 75 patients each. Primary outcomes included all-cause mortality and incident infections. Secondary outcomes were ICU admission, acute kidney injury (AKI), and other complications. Statistical analysis included Kaplan-Meier survival estimation, chi-square testing, and hazard ratio (HR) calculations. Results A total of 1,231 patients with diffuse large B-cell lymphoma (DLBCL) were included in the analysis, of whom 76 received intravenous immunoglobulin (IVIG) and 1,155 did not. After 1:1 propensity score matching, 75 patients remained in each group. Both cohorts were well-balanced in baseline characteristics, including age, sex, and comorbidity burden. All-cause mortality was identical between groups, with 13.3% of patients dying in each arm (HR 0.65; 95% CI, 0.16–2.67; p = 0.545), suggesting no survival advantage associated with IVIG administration. However, among patients without a documented prior infection, none in the IVIG cohort developed a new infection, compared to 76.9% (10 of 13) in the non-IVIG group (p < 0.001), representing a statistically and clinically significant reduction in infection risk (Table 1). ICU admissions occurred at equal rates in both cohorts (19.6%), and no differences were observed in mechanical ventilation or other intensive care needs. The incidence of AKI was higher in the IVIG group (27.8% vs. 18.5%), but this difference did not reach statistical significance (p = 0.301). Importantly, no cases of venous thromboembolism, hemolysis, meningitis, anaphylaxis, or mechanical ventilation were observed in either cohort. Median follow-up durations were comparable between groups (429 vs. 390 days). Discussion These results demonstrate that IVIG administration prevents post-treatment infections in DLBCL patients undergoing CAR T-cell or bispecific antibody therapy without increasing overall toxicity or altering survival outcomes. The magnitude of infection risk reduction and complete prevention of new infections in the IVIG cohort is notable and reinforces the potential role of IVIG as a critical supportive care intervention. Although not powered to detect small differences in secondary outcomes such as AKI, the data support the overall safety and clinical utility of IVIG in this high-risk population. Conclusion In a large, multicenter real-world analysis, IVIG use was associated with dramatically reducing new-onset infections among DLBCL patients treated with CAR T-cell or BSAB therapies, without any effect on all-cause mortality or ICU utilization. These findings support incorporating IVIG as a preventive strategy in the supportive care framework for patients receiving novel immunotherapies for aggressive lymphomas. Prospective validation is warranted to establish standardized IVIG protocols and optimize patient selection. Table 1. Clinical Outcomes After Propensity Score Matching (n = 75 per Group) Outcome IVIG Group (%) No IVIG Group (%) p-value All-cause mortality 13.3 13.3 0.545 New infection (no prior hx) 0.0 76.9 <0.001 ICU admission 19.6 19.6 1.000 Acute kidney injury (AKI) 27.8 18.5 0.301 Mechanical ventilation 0.0 0.0 — VTE, meningitis, hemolysis, etc. 0.0 0.0 —