Эпидемиология и факторы риска, связанные с инфекцией у пациентов с множественными миеломами, которые лечились с помощью биконкретных антител.

Bispecific antibodies (BsAbs) represent a promising therapeutic strategy that has significantly improved outcomes for patients with multiple myeloma (MM). BsAb function as T-cell engagers, binding simultaneously to CD3 on the T-cell and to a surface antigen on the malignant plasma cell, thereby enhancing the immunological synapse and subsequent tumor cell lysis [ 1 ]. The primary targets of these agents include B-cell maturation antigen (BCMA) for teclistamab, elranatamab, linvoseltamab and etentamig, G protein-coupled receptor class C group 5 member D (GPRC5D) for talquetamab, and the membrane-proximal domain of Fc receptor homolog 5 (FcRH5) for cevostamab [ 2 ].

Despite their targeted mechanism of action and immunostimulatory effect, MM patients receiving BsAbs remain at increased risk of infection due to multiple factors [ 3 ]. These include immunosuppression from the underlying malignancy and prior lines of therapies, as well as BsAbs-induced humoral immunodeficiency in the form of hypogammaglobulinemia (HGG), which may be partially mitigated by intravenous immunoglobulin replacement (IVIg) therapy. Moreover, immune-related adverse events, such as cytokine release syndrome (CRS) and immune-effector-cell-associated neurotoxicity syndrome (ICANS), often require the use of immunosuppressive agents, including corticosteroids or anti-interleukin (IL)-6/IL-1 agents. Chronic T-cell activation may also result in T-cell exhaustion [ 4 , 5 ].

Randomized clinical trials (RCTs) and observational multicenter studies have reported cumulative rates of infection ranging from 40% to 76% [ 6–14 ]. BsAbs targeting BCMA have been associated with an increased risk, as have combination regimens involving anti-GPRC5D with other drugs [ 15 ]. Key risk factors include neutropenia, severe HGG, and corticosteroid therapy for CRS or ICANS [ 5 , 13 , 16 ]. Respiratory tract infections, predominantly viral, are the most frequently reported, with gastrointestinal bacterial infections and bacteremia also observed [ 13 ]. Grade 3–4 infection rates have been estimated at approximately 25% in RCTs and up to 50% in real-word studies with extended follow-up [ 5 , 13 , 15 ].

Despite this recognized risk, detailed characterization of infections associated to BsAbs for MM—including incidence, causative agents, clinical presentation, and temporal patterns—remains limited. To address this gap, we conducted a single-center study to evaluate the real-world use of BsAbs in MM patients. Our study provides a comprehensive clinical description of infection patterns, analyses predictive risk factors, and explores specific scenarios such as infections concurrent with CRS episodes.

MATERIALS AND METHODS

Study Population and Setting

We performed an observational cohort study including all adult patients with MM receiving BsAb therapy at the University Hospital “12 de Octubre” (Madrid, Spain), from January 2020 to May 2025. Patients were identified from a prospective registry of hematologic malignancies. Variables that were not already registered were retrospectively collected from the patient's electronic health records (EHRs) and entered into a standardized database. The study was performed in accordance with the ethical standards laid down in the Declaration of Helsinki and was approved by the local clinical ethics committee.

Study Design

All patients were enrolled at the initiation of BsAb therapy (day 0). Infectious events were systematically recorded from day 0 until one month after discontinuation of BsAbs or the initiation of a new line of therapy, whichever occurred first. Patients attended the outpatient hematology clinic at scheduled follow-up visits—ranging from weekly to monthly according to BsAbs dosing—or when clinically indicated.

The primary study aim was to characterize the epidemiology and identify risk factors for overall, viral, bacterial, and severe (grade ≥3) infections associated with BsAb therapy for MM. Secondary aims included the assessment of infections occurring during the first CRS episode and the evaluation of the impact of baseline immunosuppression, comorbidities, and IVIg therapy on the risk of infection.

Study Definitions

The characteristics of MM and the scoring system were defined according to the International Myeloma Working Group (IMWG) consensus [ 17 ]. MM cases refractory to one proteasome inhibitor (PI), one immunomodulatory drug (IMiD), and one anti-CD38 monoclonal antibody (mAb) were classified as triple-refractory . MM cases refractory to two different PIs, two different IMiDs and one anti-CD38 mAb were classified as penta-refractory . Severe HGG was defined by IgG level <400 mg/dL once the monoclonal component secondary to MM had been subtracted from the total IgG level.

The diagnosis of proven bacterial infection required the isolation of a potentially pathogenic microorganism from a clinical specimen in the presence of compatible local or systemic signs of infection. Probable bacterial infection required the presence of compatible signs and symptoms in the absence of an isolated pathogen, with complete resolution following the initiation of empiric antibiotic therapy. Febrile syndrome was not taken into account if no causative agent could be demonstrated and no antimicrobial treatment was needed to achieve clinical resolution, with the exception of episodes expected to resolve without antimicrobial therapy (eg, respiratory tract viral infections). Episodes of asymptomatic bacteriuria and lower urinary tract infection (ie, cystitis) were also excluded. Respiratory tract viral infections were also adjudicated as proven (laboratory-confirmed infection by nucleic acid amplification test or rapid antigen tests) or probable episodes (compatible signs and symptoms with no viral pathogen detected).

Frequencies and incidence rates of bacterial, viral, fungal, or parasitic infections were separately reported considering episodes with or without microbiological identification. The Supplementary Methods includes additional definitions of infectious syndromes, as well as information on grading by the Common Terminology Criteria for Adverse Events [ 18 ]. It also provides information on the definition, grading, and management of CRS and ICANS [ 19 ].

Antimicrobial Prophylaxis

Prophylaxis against Pneumocystis jirovecii pneumonia (trimethoprim-sulfamethoxazole [160/800 mg three times weekly] or monthly intravenous pentamidine) and herpesvirus infections (acyclovir [800 mg twice daily]) was prescribed for all patients according to institutional protocols. Cytomegalovirus DNAemia was not routinely monitored but was assessed according to the criteria of the attending physicians. Patients with severe HGG received IVIg replacement therapy (0.4 g/Kg monthly) ( Supplementary Methods ).

Statistical Analysis

Quantitative variables were shown as mean ± standard deviation (SD) or median with the interquartile range (IQR). Qualitative variables were expressed as absolute and relative frequencies. Normality was explored using the Shapiro-Wilk test. Categorical variables were compared using the χ 2 test or Fisher's exact test as appropriate, whereas the Student t test or Mann-Whitney U test was applied for continuous variables. Incidence rates for the study outcomes were calculated as episodes per BsAbs courses-year, with the corresponding 95% confidence intervals (CIs).

Factors associated with the occurrence of infection were analyzed using Cox regression models for recurrent events. Models were adjusted in a two-step process. First, a set of variables was assessed at the univariate level. Immunoglobulin therapy was analyzed as a time-dependent covariate. Only variables achieving univariate P -values <0.10 were next entered into the multivariate model as potential covariates. A multivariate model was performed using a backward stepwise procedure and confirmed with an all-subsets regression approach to find the model that minimizes the Akaike information criterion (AIC) and Bayesian information criterion (BIC). Multicollinearity was analyzed with the variance inflation factor (VIF), with VIF values <3 being considered acceptable. Associations were expressed as hazard ratios (HRs) with 95% CIs. A sensitivity analysis was also performed using a Fine-Gray competing-risk model to account for the competing risk of death, with associations expressed as subdistribution hazard ratios (SHRs) and corresponding 95% CIs. All significance tests were two-tailed. Statistical analysis was conducted using Stata® software version 18.0, and graphs were generated with Prism version 9.4.0 (GraphPad Software Inc., La Jolla, CA).

RESULTS

Study Population and Outcomes

Overall, we identified 92 MM patients who received a total of 98 BsAb courses, with a median time at risk of 316 (IQR: 86–567.3) days. Demographic and clinical characteristics as well as prior lines of treatment, type of BsAbs, prophylactic regimens, and laboratory values are shown in Table 1 . The median number of prior lines of treatment was 3 (IQR: 2–5), including 82 (83.7%) patients that had received a hematopoietic stem cell transplantation (HSCT), mostly autologous (93.9% [77/82]). Anti-BCMA was the most common BsAbs administered (74.5%), followed by anti-GPRC5D (23.4%).

Table 1. Demographic and Clinical Characteristics of BsAb Courses (n = 98 Courses) Variables Age, years [mean ± SD)] 63.3 ± 10.7 Gender, male [n (%)] 56 (57.1) Underlying MM Ig subtype [n (%)]  IgG 51 (52.0)  IgA 22 (22.5)  Light chain 24 (24.5)  Other 1 (1.0) Light chain subtype [n (%)]  Kappa 56 (57.1)  Lambda 42 (42.9) R-ISS [n (%)] a  One point 22 (30.1)  Two points 35 (48.0)  Three points 16 (21.9) Number of prior lines of treatment [median (IQR)] 3 (2–5) Prior lines of treatment [n (%)]  PI 96 (98.0)  IMiD 96 (98.0)  Anti-CD38 mAb 82 (83.7)  Antibody-drug conjugate 9 (9.2)  Other BsAbs 8 (8.2)  HSCT 82 (83.7)   Autologous 77 (78.6)   Allogeneic 5 (5.1)  CAR-T therapy 6 (6.1) Triple-refractory [n (%)] 67 (68.4) Penta-refractory [n (%)] 35 (35.7) ECOG at the initiation of BsAb therapy [n (%)] b  ECOG 0 64 (66.7)  ECOG 1 28 (29.2)  ECOG ≥2 4 (4.2) Type of BsAbs [n (%)]  Anti-BCMA 73 (74.5)   + PI 4 (4.1)   + anti-CD38 mAb 5 (5.1)   + IMiD 22 (22.4)   + immune checkpoint inhibitor 7 (7.1)   + nigoracestat 2 (2.0)  Anti-GPRC5D 23 (23.4)   + anti-CD38 mAb 8 (8.2)   + IMiD 5 (5.1)  Anti-FcRH5 1 (1.2)  BsAb combination …   Anti-BCMA + anti-GPRC5D 1 (1.2) PJP prophylaxis [n (%)] 100 (100.0)  Trimethoprim-sulfamethoxazole 94 (95.9)  Intravenous pentamidine 18 (19.4) Acyclovir prophylaxis [n (%)] 96 (98.0) IVIg replacement [n (%)] 57 (58.2) Laboratory values at BsAb therapy initiation  Absolute neutrophil count, cells/μL [median (IQR)] 2600 (1800–3700)  Absolute lymphocyte count, cells/μL [median (IQR)] 1000 (600–1800)  IgG level, mg/dL [median (IQR)] 356 (231–638)   HGG (IgG level <400 mg/dL) [n (%)] 53 (54.6) BCMA, B-cell maturation antigen; BsAbs, bispecific antibodies; CAR-T, chimeric antigen T-cell receptor; ECOG, Eastern Cooperative Oncology Group; FcRH5, Fc receptor-homolog 5; GPRC5D, G protein-coupled receptor class C group 5 member D; HGG, hypogammaglobulinemia; HSCT, hematopoietic stem cell transplantation; IMiD, immunomodulatory drug; IQR, interquartile range; IVIg, intravenous immunoglobulin replacement; mAb, monoclonal antibody; MM, multiple myeloma; PJP, Pneumocystis jirovecii pneumonia; PI, proteasome inhibitor; R-ISS, Revised International Staging System; SD, standard deviation.
a Data on R-ISS not available for 35 patients.
b Data on ECOG not available for 2 patients.

Demographic and Clinical Characteristics of BsAb Courses (n = 98 Courses)

BCMA, B-cell maturation antigen; BsAbs, bispecific antibodies; CAR-T, chimeric antigen T-cell receptor; ECOG, Eastern Cooperative Oncology Group; FcRH5, Fc receptor-homolog 5; GPRC5D, G protein-coupled receptor class C group 5 member D; HGG, hypogammaglobulinemia; HSCT, hematopoietic stem cell transplantation; IMiD, immunomodulatory drug; IQR, interquartile range; IVIg, intravenous immunoglobulin replacement; mAb, monoclonal antibody; MM, multiple myeloma; PJP, Pneumocystis jirovecii pneumonia; PI, proteasome inhibitor; R-ISS, Revised International Staging System; SD, standard deviation.

a Data on R-ISS not available for 35 patients.

b Data on ECOG not available for 2 patients.

Outcomes and laboratory values during the course of BsAb therapy are described in Table 2 . Discontinuation of BsAb therapy was required in 43 (43.9%) courses, at a median time of 103 days (IQR: 39–305). Most common causes included MM relapse (74.4% [32/43]) and BsAbs-associated toxicity (16.3% [7/43]).

Table 2. Outcomes of BsAb Courses (n = 98 Courses) Variables Time at risk, days [median (IQR)] 316 (86–567.3) CRS [n (%)] 68 (69.4)  Grade 1 49 (50.0)  Grade 2 17 (17.3)  Grade ≥3 2 (2.0) ICANS [n (%)] 5 (5.1)  Grade 1 0 (0.0)  Grade 2 2 (2.0)  Grade 3 2 (2.0)  Grade 4 1 (1.0) Treatment of first episodes of CRS/ICANS [n (%)]  Steroids 20 (20.4)  Tocilizumab 41 (41.8)  Anakinra 3 (3.1) Laboratory nadir values during BsAbs treatment  Lowest IgG level, mg/dL [median (IQR)] 200 (130–304)   HGG (IgG <400 mg/dL) [n (%)] 81 (85.3)  Lowest absolute neutrophil count, cells/μL [median (IQR)] 600 (300–1100)   Grade 4 neutropenia (<500 cells/μL) [n (%)] 40 (41.0)  Lowest absolute lymphocyte count, cells/μL [median (IQR)] 100 (100–200)   Grade 4 lymphopenia (<200 cells/μL) [n (%)] 61 (61.2) BsAbs discontinuation [n (%)] 43 (43.9)  Days until discontinuation [median (IQR)] 103 (39–305)  Reason for BsAb discontinuation …   MM relapse 32 (32.7)   Toxicity a 7 (7.1)   Other b 4 (4.0) MM relapse [n (%)] 32 (32.7)  Days until relapse [median (IQR)] 103 (33–251) All-cause mortality [n (%)] 13 / 92 (14.1)  Days until death [median (IQR)] 62 (43–281) BsAbs, bispecific antibodies; CRS, cytokine release syndrome; ICANS, immune effector cell-associated neurotoxicity syndrome; HGG, hypogammaglobulinemia; IQR, interquartile range; MM, multiple myeloma; SD, standard deviation.
a Includes infection (n = 3; COVID-19 [n = 2], Escherichia coli bacteremia [n = 1]), ICANS (n = 2), and neuropathy (n = 2).
b Progression of concomitant solid organ malignancy (n = 1), patient's decision (n = 1), patient moving to another city (n = 1), and lost to follow-up (n = 1).

Outcomes of BsAb Courses (n = 98 Courses)

BsAbs, bispecific antibodies; CRS, cytokine release syndrome; ICANS, immune effector cell-associated neurotoxicity syndrome; HGG, hypogammaglobulinemia; IQR, interquartile range; MM, multiple myeloma; SD, standard deviation.

a Includes infection (n = 3; COVID-19 [n = 2], Escherichia coli bacteremia [n = 1]), ICANS (n = 2), and neuropathy (n = 2).

b Progression of concomitant solid organ malignancy (n = 1), patient's decision (n = 1), patient moving to another city (n = 1), and lost to follow-up (n = 1).

Overall, 13 patients died (14.1%) at a median interval of 62 days (IQR: 43–281), resulting in a one-year overall survival rate of 89.1%. Infections accounted for 46.2% (6/13) of deaths. Fatal infections included pneumonia (n = 3, including two cases of coronavirus disease 2019 [COVID-19]), bacteremia (n = 2), and one case of probable cytomegalovirus (CMV) hepatitis.

Incidence and Clinical Syndromes

Overall, 352 episodes of infection were diagnosed during the follow-up period (overall incidence rate: 3.20 episodes per BsAbs course-year [95% CI: 2.88–3.55]), with ≥1 episode occurring in 81.6% (80/98) of BsAb courses. Incidence rates and grading and causative agents are shown in Table 3 . The median time to the first episode of infection was 39 days (IQR: 15–87). Most common clinical syndromes involved the respiratory (71.3% [251/352]) and digestive tracts (11.9% [42/352]) ( Figure 1 ). At least one episode of severe (grade ≥3) infection occurred in 45.9% (45/98) of BsAb courses, while 18.8% (66/352) of episodes were graded ≥3. Median time to the first episode of severe infection was 82 days (IQR: 33–165).

Table 3. Incidence Rates of Infection in MM Patients Receiving BsAb Therapy (n = 352 Episodes) Clinical Syndrome N (%) Incidence Rate Per BsAbs Course-Year (95% CI) Respiratory tract infection without pneumonia a 214 (60.8) 2.08 (1.8–2.39) Digestive tract infection a 42 (11.9) 0.40 (0.29–0.55) Pneumonia 37 (10.5) 0.35 (0.25–0.49) Eyes, ear, nose and throat infection 24 (6.8) 0.24 (0.16–0.36) Skin and soft tissue infection a 22 (6.3) 0.22 (0.15–0.33) Other bacteremia b 4 (1.1) 0.04 (0.02–0.11) CMV disease 4 (1.1) 0.04 (0.02–0.11) Pyelonephritis 3 (0.9) 0.03 (0.01–0.09) CNS infection 1 (0.3) 0.01 (0.00–0.07) Neutropenic fever of unknown origin 1 (0.3) 0.01 (0.00–0.07)
Grade of the episode (CTCAE)
Grade 1 42 (12.0) 0.42 (0.31–0.60) Grade 2 243 (69.2) 2.38 (2.10–2.71) Grade 3 54 (15.4) 0.49 (0.37–0.65) Grade 4 6 (1.7) 0.05 (0.02–0.12) Grade 5 6 (1.7) 0.05 (0.02–0.12)
Causative agent

c

Viral 221 (62.8) 2.14 (1.87–2.45) Bacterial 117 (33.2) 1.12 (0.93–1.35) Fungal 10 (2.8) 0.09 (0.05–0.17) Parasitic 2 (0.6) 0.03 (0.01–0.09) BsAbs, bispecific antibodies; CI, confidence interval; CMV, cytomegalovirus; CNS, central nervous system; CTCAE, Common Terminology Criteria for Adverse Events.
a One case of bronchiectasis infection, one skin and soft tissue infection and six episodes of digestive tract infection had associated bacteremia.
b Includes two episodes of vascular catheter-related bacteremia and two episodes of bacteremia of unknown origin.
c Includes episodes with and without microbiological confirmation.

Incidence Rates of Infection in MM Patients Receiving BsAb Therapy (n = 352 Episodes)

BsAbs, bispecific antibodies; CI, confidence interval; CMV, cytomegalovirus; CNS, central nervous system; CTCAE, Common Terminology Criteria for Adverse Events.

a One case of bronchiectasis infection, one skin and soft tissue infection and six episodes of digestive tract infection had associated bacteremia.

b Includes two episodes of vascular catheter-related bacteremia and two episodes of bacteremia of unknown origin.

c Includes episodes with and without microbiological confirmation.

Distribution of infectious syndromes in episodes of infection (n = 352).

The incidence rate peaked within the first 6 months of therapy (5.09 episodes per BsAbs course-year; 95% CI: 4.37–5.92), to significantly decrease during the subsequent periods between months 6 and 12 (2.88 episodes per BsAbs course-year, 95% CI: 2.29–3.62) and beyond month 12 (2.42 episodes per BsAbs course-year, 95% CI: 2.00–2.95; P -value <0.001) ( Supplementary Figure 1 ).

Microbiological confirmation was obtained in less than half of episodes (42.0% [148/352]), with a predominance of severe acute respiratory syndrome coronavirus type 2 (SARS-CoV-2), rhinovirus, and influenza among viral pathogens and Campylobacter spp. among bacteria ( Supplementary Table S and Figure 2 ). The relative frequency of SARS-CoV-2 infection peaked in 2022 and subsequently declined over the following years ( Supplementary Table 2 ). When proven and probable episodes were considered, the most common etiology was viral (62.8% [221/352]), followed by bacterial (33.2% [117/352]) and fungal (2.8% [10/352]).

Distribution of causative agents in episodes of infection (n = 352).

At least one episode of bacterial infection occurred in 62.2% (61/98) of BsAb courses, with an incidence rate of 1.12 per BsAbs course-year (95% CI: 0.93–1.35). Bacteremia occurred in 13.3% (13/98) of BsAb courses (mostly from abdominal source) and accounted for 3.9% (14/358) of all episodes (incidence rate of 0.10 per BsAbs course-year [95% CI: 0.05–0.19]). Fungal infection was documented in 8.2% (8/98) of BsAb courses, primarily as mild mucocutaneous infections, although there were two cases of invasive fungal disease (IFD). One patient had probable COVID-19-associated pulmonary aspergillosis (CAPA) which required mechanical ventilation and corticosteroids, with Aspergillus nidulans isolated in the bronchoalveolar fluid. The other patient developed probable IFD following treatment for CRS and ICANS with corticosteroids and anti-IL6, with A. fumigatus isolated in sputum cultures, a positive serum galactomannan antigen, and nodular opacities in a computed tomography (CT) scan. No cases of Pneumocystis jirovecii pneumonia (PJP) were detected.

Determination of CMV DNAemia during follow-up was performed at least once in 42 (42.9%) BsAb courses. There were 22 episodes of CMV infection, including asymptomatic reactivation (n = 18) and CMV disease (n = 4). Clinical forms of CMV disease included retinitis (two episodes in the same patient), hepatitis, and adrenalitis. Antiviral therapy was administered in 38.9% (7/18) episodes of asymptomatic CMV infection.

Description and Management of CRS/ICANS Episodes

At least one episode of CRS/ICANS occurred in 73 (74.5%) BsAb courses. Most of CRS episodes were grade 1 (72.1% [49/68]) and grade 2 (25.0% [17/68]). Recurrent CRS episodes occurred in 19 (19.4%) courses. Although the difference was not significant, the occurrence of CRS was more frequent with anti-GPRC5D than with anti-BCMA BsAbs (78.3% versus 67.1%; P- value = 0.408). Thirty-one episodes (42.5% [31/73]) were treated with tocilizumab only, 10 episodes (13.7% [10/73]) were treated with corticosteroids only, and 10 episodes (13.7% [10/73]) required the combination of both therapies. Regarding the first episode of CRS in each course, empiric antibiotic therapy was concomitantly initiated in 82.2% (60/73) of them, mainly with meropenem (55.0% [33/60]) ( Supplementary Table 3 ). Median time of antibiotic therapy was 6 days (IQR: 4–9). Infection was microbiologically documented in 6.8% (5/73) of these episodes: three cases of respiratory tract viral infection (adenovirus, parainfluenza, rhinovirus, and metapneumovirus coinfection), one case of mild chronic obstructive respiratory disease (COPD) exacerbation with isolation of Pseudomonas aeruginosa in sputum, and one case of P. aeruginosa bacteremia from a respiratory source. All these episodes resolved without complications.

Three out of five patients with ICANS (60.0%) developed an infectious complication, with a fatal course in two of them (one case of IFD and one case of bacteremia due to carbapenemase-producing Enterobacter cloacae due to ischemic/stercoral colitis).

Factors Associated With Overall, Viral and Bacterial Infection

Factors associated with overall infection are shown in Supplementary Table 4 . After adjusting for those covariates that have been previously proven to achieve univariate P -values <0.10, the previous receipt of other BsAb (adjusted HR [aHR]: 1.75; 95% CI: 1.23–2.50; P -value = 0.002), prior allogeneic HSCT (aHR: 1.59; 95% CI: 1.12–2.26; P -value = 0.010), and the development of ICANS (aHR: 3.23; 95% CI: 1.27–8.19; P -value = 0.001) were associated with overall infection. In addition, there was a trend toward a higher risk among triple refractory MM patients (aHR: 1.23; 95% CI: 0.97–1.56; P -value = 0.083) and those that developed grade 4 (<500 cells/μL) neutropenia during the course of BsAb therapy (aHR: 1.23; 95% CI: 0.97–1.55; P -value =0.084).

Prior allogeneic HSCT (aHR 1.94; 95% CI: 1.31–2.88; P -value = 0.001) was also identified as a risk factor for viral infection, whereas the development of grade 4 neutropenia during follow-up exhibited a near significant trend (aHR: 1.34; 95% CI: 1.02–1.76; P -value = 0.052) ( Supplementary Table 5 ). The use of IVIg therapy exerted a protective effect for overall (aHR: 0.68; 95% CI: 0.53–0.86; P -value = 0.002) and viral infection (aHR: 0.59; 95% CI: 0.44–0.80; P -value = 0.001).

Factors related to poorer MM staging and patient functional status at the initiation of BsAB therapy were predictive of bacterial infection: Revised International Staging System (R-ISS) (aHR [per one-point increase]: 1.32; 95% CI: 1.00–1.75; P- value = 0.054) and the Eastern Cooperative Oncology Group (ECOG) score (aHR [per one-point increase]: 1.49; 95% CI: 1.07–2.07; P -value = 0.018) ( Supplementary Table 6 ).

The risk of overall infection was higher with anti-BCMA compared to anti-GPRC5D therapy (aHR: 1.41; 95% CI: 1.02–1.94; P -value = 0.035), with no significant differences for viral, bacterial, or severe infection ( Supplementary Table 7 ).

We performed a sensitivity analysis based on a Fine-Gray model for competing risks, in which death was treated as the competing risk. Factors associated with overall infection were the prior use of anti-CD38 agents (SHR: 1.56; 95% CI: 1.02–2.40; P -value = 0.042), the concomitant therapy of immune checkpoint inhibitors (SHR: 9.97; 95% CI: 4.20–23.68; P -value = 0.001), G4 neutropenia (SHR: 2.08; 95% CI: 1.31–3.31; P -value = 0.002), and the use of tocilizumab for the treatment of CRS/ICANS (SHR: 1.81; 95% CI: 1.06–3.12; P -value = 0.031) ( Supplementary Table 8 ).

Factors Associated With Severe Infection

Independent predictors of severe infection (grade ≥3) were the ECOG score at the initiation of BsAbs (aHR [per one-point increase]: 1.51; 95% CI: 1.04–2.18; P -value = 0.030), prior therapy with anti-CD38 mAb (aHR: 2.86; 95% CI: 1.19–6.88; P -value = 0.019), and development of grade 4 neutropenia (aHR: 2.03; 95% CI: 1.13–3.66; P -value = 0.018) and ICANS (aHR: 5.92; 95% CI: 1.60–21.93; P -value = 0.008). A protective effect was observed for IVIg therapy (aHR: 0.42; 95% CI: 0.22–0.79; P -value = 0.007) ( Supplementary Table 9 ).

DISCUSSION

The use of BsAbs in MM patients is associated with a substantial infection burden, as suggested by evidence derived from pivotal RCTs and a limited number of retrospective series [ 6–14 , 20–23 ]. In this prospective cohort study, we provided a comprehensive characterization of the incidence rate, microbiological features, clinical presentation, and risk factors of infections associated with anti-BCMA, anti-GPRC5D and anti-FcRH5 BsAbs.

We observed an overall incidence rate of 3.20 infection episodes per BsAbs course-year and a cumulative frequency of 81.6%, in line with previous studies [ 6 , 8 , 9 , 11 , 12 ]. Notably, severe infections (grade ≥3) represented only 18.8% of episodes, reflecting the predominance of mild respiratory tract viral infections. Although fatal infections were uncommon (1.7% of all episodes), infections accounted for nearly half of all deaths during follow-up, with respiratory complications (ie, pneumonia or COVID-19) responsible for approximately half of these fatalities. These findings underscore the substantial morbidity and mortality in absolute terms associated with infections in this population. The incidence rate of infection significantly declined over time, likely reflecting incipient immune reconstitution and the favorable clinical course of patients who achieved disease remission and remained on BsAb therapy.

Respiratory tract infection of probable viral origin was the most frequent event. Identified pathogens included SARS-CoV-2, influenza virus, and rhinovirus, a pattern likely influenced by the COVID-19 pandemic but consistent with prior observations [ 13 , 15 , 20 , 22 ]. Bacterial infections were primarily caused by encapsulated organisms typically associated with HGG, such as Haemophilus influenzae or Streptococcus pneumoniae . A notable proportion of P. aeruginosa infections occurred in patients with COPD or bronchiectasis.

The gastrointestinal tract represented the second most common site of infection, with a remarkable incidence of Campylobacter spp. enterocolitis and associated bacteremia. Both hematologic malignancy and HGG are well-stablished risk factors for this complication (present in 26% and 14% of cases of Campylobacter bacteremia collected in a large multicenter study, respectively) [ 24 ]. Bilbao et al. have recently reported a high risk of Campylobacter infection, including bacteremia, in MM patients receiving BsAbs at their institution [ 16 ]. These comparable results from concurrent, independent cohorts may reflect local epidemiology or suggest an emerging complication as the use of BsAbs continues to expand. Clostridioides difficile infection was also common, reflecting the effect of cumulative antibiotic exposure and HGG.

In contrast, opportunistic infections related to impaired cell-mediated immunity were infrequent. Two cases of IFD due to Aspergillus spp. were diagnosed in patients with established risk factors, whereas one further patient with predisposing bronchiectasis developed non-tuberculous mycobacterial infection. No cases of PJP were detected, in contrast to the low but non-negligible incidence (1%–4%) reported in other cohorts receiving anti-BCMA therapies [ 8 , 16 , 22 ]. This discrepancy likely reflects the universal use of PJP prophylaxis in our center. Overall, these findings suggest that routine antifungal prophylaxis (other than PCP) may not be necessary in MM patients treated with BsAbs.

In the present experience, monitoring for CMV DNAemia was not conducted on a systematic basis but restricted to patients with compatible clinical manifestations. The cumulative incidence of CMV infection observed (13.3% of BsAb courses) is higher than rates reported from RCTs and observational studies lacking routine monitoring, but lower than cohorts with regular assessment for CMV DNAemia in asymptomatic patients [ 8 , 9 , 22 , 25 ]. Pei et al. found an incidence of CMV reactivation during teclistamab therapy as high as 52.2%, although symptomatic disease was uncommon [ 25 ]. The clinical relevance of CMV reactivation in this setting remains uncertain, and further studies are required to identify the patients that would benefit from tailored monitoring strategies or even prophylaxis.

Regarding infection factors for infection, our results align with previous studies [ 13 , 15 , 26 , 27 ]. Cumulative immunosuppression derived from previous MM therapies—in particular prior BsAb exposure, anti-CD38 agents, or allogeneic-HSCT—emerged as a major contributor. Our experience also aligns with the hypothesis that CRS/ICANS—and their management with immunosuppressive therapies—might increase susceptibility to infection, as previously reported [ 13 ]. In our multivariate analysis, the occurrence of ICANS was the only predictor directly associated to BsAb therapy identified, while the use of anti-IL6 agents was also identified as an independent risk factor in the Fine-Gray model accounting for the competing risk of death. We also found that higher R-ISS stage and poorer ECOG score at the initiation of therapy were associated with bacterial infection. Remarkably, concomitant therapy with immune checkpoint inhibitors was also associated to an increased risk of infection in the competing risk model, which could be explained by the subsequent occurrence of immune-related adverse events requiring immunosuppressive therapy or by accelerated T-cell exhaustion [ 28 ]. While this is an interesting area for further investigation, the clinical significance of this finding should be interpreted with caution given the small proportion of patients receiving these treatments and the possibility of residual confounding due to unadjusted baseline differences. Consistent with previous reports, our data confirms the protective effect of IVIg replacement therapy [ 4 , 29 ]. This finding highlights the importance of screening for HGG during and after BsAb therapy, as IVIg replacement remains the cornerstone of infection prevention in these patients.

Different BsAbs for MM have been associated with heterogeneous infection risks, with agents targeting molecules other than BCMA generally linked to lower rates [ 5 , 15 ]. In our cohort, anti-BCMA agents were associated with a higher overall infection risk compared with anti-GPRC5D. The incidence was also higher with combination regimens relative to BsAb monotherapy. Nevertheless, the heterogeneity in baseline characteristics, comorbidities, MM staging, and concurrent therapies limits direct comparability and precludes definitive conclusions about the intrinsic infectious risk across BsAb classes.

Another relevant aspect concerns the antimicrobial management of CRS episodes. Almost all patients that experienced this complication received broad-spectrum antibiotics for a median of 6 days, despite the low clinical suspicion of infection and lack of microbiological documentation. Notably, only two bacterial infections were identified within five days of CRS onset. Although profound immunosuppression and the clinical challenge of differentiating CRS from early infection justify the initiation of empirical antibiotic therapy, our experience is reassuring regarding the safety of limiting treatment duration according to clinical response and microbiological results. Local epidemiology should guide the choice of therapy, followed by early de-escalation when appropriate. In contrast, patients with severe CRS or any-grade ICANS require closer surveillance, since the rapid intensification of immunosuppression increases the risk of bacterial infection [ 13 , 30 ]. In the present cohort, two out of five patients with ICANS had fatal infections (IFD and bacteremia due to carbapenemase-producing Enterobacter ). Further studies regarding predictive factors or biomarkers of infection in this subgroup are needed, as the differentiation between severe CRS and bacterial sepsis remains a major challenge.

Our study has various limitations to be acknowledged. Firstly, although the primary aim was to characterize the incidence and patterns of infections associated to BsAbs for MM, meaningful comparisons between individual agents were not feasible due to the observational design and the heterogeneity of patient profiles. Data on individual BsAbs dose modifications due to concurrent infections were also not available. Secondly, the single-center design limits its external validity, as local epidemiology may influence the type of infection. Thirdly, incomplete availability of data in EHRs prevented the accurate determination of the duration of lymphopenia and neutropenia during individual BsAb courses. Since vaccination status was not collected, the same limitation applied for the analysis of the potential protective effect of vaccination practices against pneumococcal disease or respiratory viruses. Fourthly, the high frequency of upper respiratory tract viral infection diluted the relative proportion of more severe episodes. Although reflective of real-world practice, this may have underestimated the actual risk of clinically significant infections. Episodes of cystitis were excluded because of the inherent difficulty of retrospectively distinguishing symptomatic cystitis from asymptomatic bacteriuria, although this approach may have resulted in the omission of potentially relevant information. Fifthly, follow-up was censored at the time of BsAb discontinuation or initiation of a subsequent line of therapy. Consequently, late-onset infections potentially related to prior BsAb exposure may have been missed, leading to an underestimation of the actual incidence. Finally, the COVID-19 pandemic context in which the study was conducted may limit the generalizability to other time periods.

In conclusion, BsAb therapy in MM patients is associated with a high incidence of infection, particularly with respiratory and gastrointestinal involvement. Cumulative immunosuppression from prior therapies, underlying comorbidities, and the occurrence of ICANS are major determinants of infection risk, whereas IVIg therapy would confer a protective effect. Although these factors underscore the need for prompt and appropriate antimicrobial management, antibiotic stewardship remains feasible during early CRS episodes—when bacterial infections are uncommon—in contrast to later stages or in patients with severe CRS/ICANS, for which the infection risk becomes more clinically relevant.

Supplementary Material

Notes

Author contributions.
E. A. M., N. L. M., A. S. J. M., M. F. R., and I. R. G. participated in the study concept and design, clinical management of patients, data collection and analysis, and manuscript writing, drafting and reviewing. G. B. H. participated in data collection, and manuscript editing and reviewing. M. P. J. A., J. M. S. P., and M. C. participated in the study concept and design, clinical management of patients, data analysis, and manuscript editing and reviewing. J. B., F. L. M., R. S. J., J. T. S., J. M. L., and J. M. A. participated in the study concept and design, and manuscript editing and reviewing.

Data availability.
The data that support the findings of this study are available upon reasonable request to the corresponding author.

Financial support.
E. A. M. and J. B. hold “Río Hortega” research training contracts (CM25/00243 and CM25/00250), and I. R. G. holds a research contract “Juan Rodés” (JR24/00034), all from the Instituto de Salud Carlos III, Spanish Ministry of Science, Innovation and Universities.

Contributor Information

Eduardo Aparicio-Minguijón,
Unit of Infectious Diseases, Hospital Universitario “12 de Octubre”, Instituto de Investigación Sanitaria Hospital “12 de Octubre” (imas12), Madrid, Spain.

Nieves López-Muñoz,
Department of Hematology, Hospital Universitario “12 de Octubre”, Instituto de Investigación Sanitaria Hospital “12 de Octubre” (imas12), CNIO, Madrid, Spain.

Guillermo Bartolomé Herguedas,
Department of Internal Medicine, Hospital Universitario “12 de Octubre”, Madrid, Spain.

Adolfo Jesús Sáez Marín,
Department of Hematology, Hospital Universitario “12 de Octubre”, Instituto de Investigación Sanitaria Hospital “12 de Octubre” (imas12), CNIO, Madrid, Spain.

María Asunción Pérez-Jacoiste Asín,
Unit of Infectious Diseases, Hospital Universitario “12 de Octubre”, Instituto de Investigación Sanitaria Hospital “12 de Octubre” (imas12), Madrid, Spain;
Department of Medicine, School of Medicine, Universidad Complutense, Madrid, Spain;
Centro de Investigación Biomédica en Red de Enfermedades Infecciosas (CIBERINFEC), Instituto de Salud Carlos III, Madrid, Spain.

José María Sánchez-Pina,
Department of Hematology, Hospital Universitario “12 de Octubre”, Instituto de Investigación Sanitaria Hospital “12 de Octubre” (imas12), CNIO, Madrid, Spain.

María Calbacho,
Department of Hematology, Hospital Universitario “12 de Octubre”, Instituto de Investigación Sanitaria Hospital “12 de Octubre” (imas12), CNIO, Madrid, Spain.

Jorge Boán,
Unit of Infectious Diseases, Hospital Universitario “12 de Octubre”, Instituto de Investigación Sanitaria Hospital “12 de Octubre” (imas12), Madrid, Spain.

Francisco López-Medrano,
Unit of Infectious Diseases, Hospital Universitario “12 de Octubre”, Instituto de Investigación Sanitaria Hospital “12 de Octubre” (imas12), Madrid, Spain;
Department of Medicine, School of Medicine, Universidad Complutense, Madrid, Spain;
Centro de Investigación Biomédica en Red de Enfermedades Infecciosas (CIBERINFEC), Instituto de Salud Carlos III, Madrid, Spain.

Rafael San Juan,
Unit of Infectious Diseases, Hospital Universitario “12 de Octubre”, Instituto de Investigación Sanitaria Hospital “12 de Octubre” (imas12), Madrid, Spain;
Department of Medicine, School of Medicine, Universidad Complutense, Madrid, Spain;
Centro de Investigación Biomédica en Red de Enfermedades Infecciosas (CIBERINFEC), Instituto de Salud Carlos III, Madrid, Spain.

José Tiago Silva,
Unit of Infectious Diseases, Hospital Universitario “12 de Octubre”, Instituto de Investigación Sanitaria Hospital “12 de Octubre” (imas12), Madrid, Spain.

Joaquín Martínez-López,
Department of Hematology, Hospital Universitario “12 de Octubre”, Instituto de Investigación Sanitaria Hospital “12 de Octubre” (imas12), CNIO, Madrid, Spain;
Department of Medicine, School of Medicine, Universidad Complutense, Madrid, Spain.

José María Aguado,
Unit of Infectious Diseases, Hospital Universitario “12 de Octubre”, Instituto de Investigación Sanitaria Hospital “12 de Octubre” (imas12), Madrid, Spain;
Department of Medicine, School of Medicine, Universidad Complutense, Madrid, Spain;
Centro de Investigación Biomédica en Red de Enfermedades Infecciosas (CIBERINFEC), Instituto de Salud Carlos III, Madrid, Spain.

Mario Fernández-Ruiz,
Unit of Infectious Diseases, Hospital Universitario “12 de Octubre”, Instituto de Investigación Sanitaria Hospital “12 de Octubre” (imas12), Madrid, Spain;
Department of Medicine, School of Medicine, Universidad Complutense, Madrid, Spain;
Centro de Investigación Biomédica en Red de Enfermedades Infecciosas (CIBERINFEC), Instituto de Salud Carlos III, Madrid, Spain.

Isabel Rodríguez-Goncer,
Unit of Infectious Diseases, Hospital Universitario “12 de Octubre”, Instituto de Investigación Sanitaria Hospital “12 de Octubre” (imas12), Madrid, Spain;
Department of Medicine, School of Medicine, Universidad Complutense, Madrid, Spain;
Centro de Investigación Biomédica en Red de Enfermedades Infecciosas (CIBERINFEC), Instituto de Salud Carlos III, Madrid, Spain.

Supplementary Data

Supplementary materials are available at Open Forum Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.

References

Associated Data

Supplementary Materials