Paper Details

ABSTRACT

Background: Treatment with BCMA-directed T-cell engagers (TCEs) is associated with humoral and cellular immunodeficiency and an increased risk of serious. In clinical trials, the rate of grade ≥3 infections ranges from 41.5% to 44.8%. Guidelines for infection mitigation vary in their recommendations and complexity and lack prospective validation. We implemented a system-wide simplified infection mitigation protocol (Mohan et al., 2023; PMID: 37287117) with the intent of prospectively assessing the effectiveness of these expert-based recommendations.

Methods: We analyzed consecutive patients receiving BCMA-directed TCEs as their first TCE therapy beginning in January 2023, corresponding to the introduction of system-wide infection mitigation measures at a referral myeloma center. Infection mitigation consisted of Pneumocystis jirovecii pneumonia (PJP) prophylaxis with trimethoprim-sulfamethoxazole (TMP-SMX, B) 3x/week starting with the first TCE dose (with alternatives for patients with sulfa allergy), universal varicella-zoster virus/herpes simplex virus (VZV/HSV) prophylaxis with Acyclovir, oral Levofloxacin for the entire first cycle of therapy (28 days), and intravenous immunoglobulin (IVIG) at a dose of 0.4 g/kg beginning on day 1 of cycle 2 and repeated on day 1 of each subsequent cycle (BALI protocol). BALI was administered irrespective of laboratory parameters, including absolute neutrophil count (ANC), absolute lymphocyte count (ALC), lymphocyte subpopulations, or immunoglobulin levels. Adherence was verified by medical record review and patient interview at each visit. There was no advanced monitoring for infection in the absence of symptoms, including no monitoring of CMV viremia and no routine cultures, including for fever occurring within the expected window for cytokine release syndrome (CRS; typically the first 10 days of therapy). Patients with fever beyond the expected CRS window and those with other signs of infection regardless of timing underwent a comprehensive infectious workup and often received empiric broad-spectrum antibiotic therapy. Interventions continued for at least the duration of TCE therapy. We collected baseline patient and disease characteristics, duration of therapy, adherence to BALI, occurrence of grade ≥3 infections (per CTCAE v5.0), and infection characteristics and outcomes.

Results: A total of 101 consecutive patients received BCMA-directed TCEs, contributing 962.7 patient-months of observation. The median age was 71 years (IQR, 64–78), the median number of prior lines of therapy was 4 (IQR, 2–5), 48% had high-risk cytogenetics, 80% had triple-class refractory multiple myeloma, and 18% had received prior chimeric antigen receptor (CAR) T-cell therapy. The BCMA-directed TCE used was teclistamab in 73% of patients, elranatamab in 21%, and other agents in 6%. At TCE initiation, the median ANC was 2,660/μL (IQR, 2,075–3,820), the median ALC was 1,200/μL (IQR, 675–1,650), and the median IgG level was 360 mg/dL (IQR, 215–600). The median IgG level at the start of cycle 2, before the first IVIG dose, was 240 mg/dL (IQR, 160–355). Adherence to BALI was 90%; nonadherence was driven primarily by patient reluctance or intolerance to IVIG. There were six episodes of grade ≥3 infection, with a cumulative 12-month incidence of 8.4% (95% CI, 1.7%–15.1%). No grade ≥3 infections occurred beyond 6.5 months of TCE therapy. Of the six episodes, three occurred among the 10 patients who did not receive IVIG. Four episodes were pneumonia (one pneumococcal pneumonia in a patient not receiving IVIG and three with no identified pathogen; all patients recovered), one was Escherichia coli bacteremia (patient receiving IVIG; recovered), and one was a soft tissue infection caused by Klebsiella pneumoniae and Achromobacter spp in a patient not receiving IVIG, levofloxacin, or TMP-SMX; this infection was fatal. Among the 91 patients adherent to BALI, the 12-month cumulative incidence of grade ≥3 infection was 4.6% (95% CI, 0%–9.9%). In multivariable analysis, nonadherence to BALI was the only factor associated with an increased risk of grade ≥3 infection (HR, 7.49; 95% CI, 1.51–37.15; P = 0.014). There were no cases of grade ≥3 viral illness, PJP, Clostridium difficile infection, or invasive fungal infection.