Highlight
In a large cohort of 1786 patients with multiple myeloma across disease stages and treatment settings, infection risk was associated with a distinct immune phenotype characterized by lower percentages of CD27+ B cells, lower CD27- natural killer cells, and a higher CD27-/CD27+ T-cell ratio in bone marrow.
These immune risk factors were validated in three independent data sets, strengthening confidence that the observed signal is reproducible rather than cohort-specific.
An immune score based on the presence of at least two of these features identified patients with substantially higher infection incidence than those with zero or one risk factor (60% vs 35%, P< .001).
Because immune populations in bone marrow and peripheral blood were significantly correlated, the study suggests that minimally invasive blood-based monitoring may be feasible in routine practice.
Background and Clinical Context
Infection remains one of the most important non-relapse complications in multiple myeloma. The reasons are multifactorial. The disease itself impairs humoral and cellular immunity through marrow infiltration, immunoparesis, and chronic immune dysregulation. Treatment adds further immunosuppression, including corticosteroids, proteasome inhibitors, anti-CD38 antibodies, bispecific antibodies, chimeric antigen receptor T-cell strategies, and other targeted therapies. Age, frailty, renal dysfunction, prior lines of therapy, and cumulative marrow damage further shape susceptibility.
Clinically, infection risk is not evenly distributed across the myeloma population. Some patients experience recurrent bacterial and viral events, whereas others tolerate intensive therapy with relatively few infectious complications. This heterogeneity is only partially captured by conventional variables such as disease stage, neutropenia, or treatment class. A more biologically grounded method of risk estimation could help guide prophylaxis, surveillance, timing of vaccination, and perhaps treatment selection in vulnerable individuals.
Flow cytometric immune profiling is attractive in this setting because it can quantify multiple immune cell compartments simultaneously and may reflect the cumulative impact of host factors, disease burden, and therapy. However, despite wide recognition that myeloma is an immunocompromised state, there has been limited large-scale evidence linking specific immune signatures to clinically meaningful infection outcomes. The study by Zabaleta and colleagues addresses this gap using next-generation flow cytometry in both bone marrow and peripheral blood.
Study Design
This was a large immune profiling study involving 1786 patients with multiple myeloma at different disease stages and under different treatment scenarios. According to the abstract, the investigators performed next-generation flow cytometry on bone marrow and peripheral blood samples. The main objective was to identify immune biomarkers associated with increased infection risk and to validate those findings across independent cohorts.
The study population appears intentionally broad, which is an important strength. It reflects real-world heterogeneity in myeloma, including variation in disease status and therapy exposure. This design increases the potential clinical usefulness of the findings, because infection risk in myeloma is dynamic and strongly influenced by treatment phase.
The principal candidate biomarkers derived from immune profiling were then related to the subsequent occurrence of infection. The investigators also developed an immune score based on the number of adverse immune features present. Multivariable analysis was used to assess whether the immune score added information beyond standard clinical factors and treatment exposures.
Although the abstract does not specify every endpoint definition, timing window, or microbiologic classification, the core clinical endpoint was infection incidence. The investigators also examined the relationship between bone marrow and peripheral blood immune subsets to assess whether less invasive monitoring could approximate marrow-derived risk signals.
Key Findings
Specific immune cell abnormalities were associated with infection
Patients who developed infection had significantly lower percentages of CD27+ B cells in the bone marrow. CD27 is generally regarded as a marker of memory differentiation in B cells. Reduced CD27+ B-cell representation may therefore reflect impaired humoral memory and reduced capacity for efficient antibody-mediated defense. In myeloma, where immunoparesis is already common, this observation is biologically plausible and clinically intuitive.
Patients with infection also had lower percentages of CD27- natural killer cells. Natural killer cells are central to innate antiviral and antitumor immunity, but their subsets differ in maturation state and functional capacity. The study’s identification of a CD27-defined NK-cell alteration suggests that innate immune remodeling in myeloma may carry infection relevance beyond simple total lymphocyte counts.
A third signal was an increased CD27-/CD27+ T-cell ratio in bone marrow. This ratio likely captures an imbalance in T-cell differentiation and immune competence, though the precise functional interpretation depends on the phenotype of the underlying T-cell subsets. Broadly, a relative expansion of CD27-negative T cells can be consistent with chronic antigen exposure, exhaustion, terminal differentiation, or immune senescence. In a myeloma setting, such skewing could signify reduced adaptive flexibility when pathogens are encountered.
Validation across independent data sets
A notable strength is that these immune risk factors were validated in three independent data sets. Reproducibility is critical for biomarker development, especially when dealing with high-dimensional immune phenotyping that may otherwise yield spurious associations. External or independent validation makes the reported findings more credible and increases the likelihood that they reflect a real biology of infection susceptibility rather than a center-specific artifact.
An immune score stratified clinically meaningful infection risk
The investigators translated the three biomarker signals into a practical immune score. Patients were grouped according to whether they had zero or one risk factor versus two or more risk factors. This simple stratification clearly separated infection incidence: 35% in the lower-risk group versus 60% in the higher-risk group, with P< .001.
From a clinical standpoint, this is an important result. A score based on just three immunophenotypic features is easier to implement than a complex multidimensional algorithm. The magnitude of separation is also meaningful. An absolute difference of 25 percentage points in infection incidence suggests that the score may identify a subgroup meriting enhanced preventive strategies, particularly if confirmed prospectively in treatment-specific cohorts.
The immune score remained independently associated with infection
In multivariable analysis, the immune score was independently associated with infection incidence, with an odds ratio of 2.31 and P< .001. This indicates that the score retained predictive value even after accounting for other relevant clinical variables. In addition, disease stage and exposure to CD38-, B-cell maturation antigen-, or G protein-coupled receptor class C group 5 member D-targeted therapy were independently associated with infection incidence.
These findings fit current clinical experience. Advanced disease often coincides with deeper immune dysfunction, higher tumor burden, and more prior therapy exposure. Likewise, modern immunotherapies have transformed myeloma outcomes but can produce profound infectious vulnerability through plasma cell depletion, hypogammaglobulinemia, T-cell redirection effects, or prolonged immune reconstitution. The fact that immune profiling still added independent signal beyond these factors argues for its clinical relevance.
Bone marrow and peripheral blood immune compartments were significantly correlated
All cell types detectable in bone marrow and peripheral blood were significantly correlated. This is potentially one of the study’s most practice-changing observations. Bone marrow sampling is invasive and not suitable for frequent monitoring solely for infection risk assessment. If peripheral blood can serve as a reliable surrogate for marrow immune architecture, then serial immune surveillance becomes much more feasible in routine laboratories.
That said, correlation does not always mean interchangeability. Future work will need to determine whether blood-based thresholds can replicate the predictive performance of marrow-derived cutoffs and whether they remain robust across therapies that redistribute immune cells differently. Still, the translational potential is obvious.
Biological Interpretation
The study’s three-component signature points to a combined defect in humoral memory, innate surveillance, and T-cell homeostasis. This is conceptually appealing because infection susceptibility in myeloma is rarely explained by a single immune compartment. Instead, risk emerges from layered dysfunction: impaired antibody production, altered antigen presentation, T-cell exhaustion or senescence, dysregulated NK-cell activity, mucosal barrier injury, and treatment-related cytopenias.
CD27+ B-cell depletion may mark the erosion of normal memory B-cell reserves in a marrow dominated by malignant plasma cells and treatment effects. NK-cell changes may indicate weakened first-line defense against viral reactivation and certain opportunistic processes. A high CD27-/CD27+ T-cell ratio may reflect chronic immune stress and impaired adaptive responsiveness. Together, these markers plausibly summarize cumulative immune injury better than traditional laboratory tests alone.
This also aligns with a broader shift in oncology and hematology: moving from crude measures of immunosuppression toward quantitative immune phenotyping. Similar approaches are being explored in transplantation, chronic lymphocytic leukemia, and solid-tumor immunotherapy, where immune state can predict both toxicity and infection.
Clinical Implications
The immediate clinical message is not that every patient with myeloma should undergo intensive immune profiling tomorrow, but that infection risk assessment may be improved by incorporating immunologic context rather than relying only on stage, blood counts, and treatment class.
Potential applications include identifying patients who may benefit from closer monitoring during vulnerable treatment periods, earlier or more sustained antimicrobial prophylaxis in selected settings, lower thresholds for intravenous immunoglobulin consideration when recurrent infections coexist with hypogammaglobulinemia, and more strategic timing of vaccination or revaccination. The score may also prove useful in counseling patients receiving CD38-, BCMA-, or GPRC5D-directed therapies, where infection prevention is already a major operational concern.
The observed association with targeted therapies is particularly timely. Anti-CD38 antibodies, BCMA-targeted bispecific antibodies and CAR T-cell therapies, and GPRC5D-targeted agents have delivered substantial efficacy, but infectious complications remain a major challenge. A validated immune score could eventually support pre-treatment baseline risk assessment and dynamic reassessment during therapy.
Another practical implication is laboratory accessibility. Next-generation flow cytometry is increasingly available in hematology centers, especially those already using standardized minimal residual disease platforms. If the relevant immune subsets can be harmonized across laboratories, implementation barriers may be lower than for more complex transcriptomic or functional assays.
Strengths and Limitations
The study has several notable strengths: a very large sample size, inclusion of diverse disease stages and treatment contexts, use of next-generation flow cytometry, validation in three independent data sets, and translation of biologic findings into a simple risk score. These features make the work more clinically relevant than many exploratory immune studies.
Several limitations should also be acknowledged. First, the abstract does not provide full details on infection definitions, grading, microbiology, timing relative to sampling, or the distinction between bacterial, viral, and fungal events. Different pathogens may map to different immune vulnerabilities. Second, observational biomarker studies can demonstrate association but not causation. Third, heterogeneity in treatment status, prior therapy exposure, and supportive care practices could introduce residual confounding even after multivariable adjustment.
Fourth, although marrow and blood compartments were correlated, correlation alone does not prove that a blood-only assay can replace marrow measurements for clinical decision-making. Fifth, the report as summarized does not provide calibration metrics, discrimination statistics, or decision-curve analyses that would help clinicians judge whether the score materially improves prediction over existing models. Finally, the score will need prospective validation, ideally with predefined thresholds and clinically actionable interventions linked to risk categories.
How This Fits With Existing Evidence
Current myeloma guidelines already emphasize infection prevention, especially early in therapy and during treatment with highly immunosuppressive regimens. Prior studies have established the importance of hypogammaglobulinemia, disease burden, corticosteroid exposure, neutropenia, and treatment class, but no universally adopted immunophenotypic infection score is used in routine myeloma care. The present study adds granularity by identifying specific lymphoid signatures associated with risk.
The findings also complement prior observations that myeloma is marked by profound immune remodeling in both marrow and blood. Anti-CD38 therapy is known to affect normal plasma cells and immune subsets, while BCMA- and GPRC5D-directed approaches can produce prolonged defects in humoral immunity. In this context, a biomarker framework that integrates host and therapy effects is highly relevant.
Research Priorities
The next steps are clear. Prospective studies should test whether the immune score predicts pathogen-specific infection, severe infection, hospitalization, and infection-related mortality. It will also be important to determine how stable these markers are over time and how rapidly they change after treatment initiation, response, relapse, or stem cell transplantation.
Interventional studies are especially needed. A biomarker is most useful when it changes management. Future trials could evaluate whether high-risk patients, as defined by this score, benefit from intensified prophylaxis, immunoglobulin replacement strategies, modified vaccination schedules, or altered monitoring frequency. Standardization of flow cytometry panels and thresholds across centers will be essential if the score is to move beyond academic use.
Finally, peripheral blood-based versions of the score deserve particular attention. If validated, they could support serial monitoring with much less patient burden and broader adoption across community and academic practices.
Funding and Trial Registration
The abstract and citation provided do not report funding details or a ClinicalTrials.gov registration number. Readers should consult the full Blood publication for funding sources, disclosures, ethics approval details, and any registry information if applicable.
Conclusion
Zabaleta and colleagues provide compelling evidence that infection risk in multiple myeloma can be captured by a focused immune signature rather than by clinical variables alone. Lower CD27+ B cells, lower CD27- NK cells, and a higher CD27-/CD27+ T-cell ratio in bone marrow identified patients at increased risk, and a simple score based on these features stratified infection incidence in a clinically meaningful way. The fact that this signal was validated in three independent data sets and that marrow and blood immune compartments were significantly correlated makes the study especially relevant for translational practice.
For clinicians, the work reinforces a central point: infection in myeloma is a biologic consequence of layered immune dysfunction, not merely an unavoidable side effect of treatment. For researchers, it opens a path toward prospective, biomarker-guided prevention. If future studies confirm that blood-based immune monitoring can guide prophylaxis and supportive care, this approach could become an important part of precision supportive oncology in multiple myeloma.
References
1. Zabaleta A, Tamariz-Amador LE, Kostopoulos IV, Sklavenitis-Pistofidis R, Smits F, Rodriguez-Otero P, Roncal C, Aranha MP, Zihala D, Machu M, Tsakirakis N, Bakouros P, Tsitsilonis O, Solia I, Moreno C, Maia C, Martin-Sanchez E, Perez JJ, Encinas C, Rios-Tamayo R, Oriol A, Blanchard MJ, de Arriba F, Gonzalez ME, Lakhwani S, Sureda A, Cabañas V, Escalante F, Carrillo-Cruz E, Pérez-Montaña A, Ocio EM, Bargay J, Orfao A, Jelinek T, Ghobrial IM, Mutis T, Zweegman S, Terpos E, Kastritis E, Martinez-Lopez J, Lahuerta JJ, Fernandez de Larrea C, Rosiñol L, Bladé J, Mateos MV, San-Miguel J, Cedena MT, Puig N, Paiva B. Immune biomarkers of increased infection risk in multiple myeloma. Blood. 2026-Apr-30;147(18):2081-2088. PMID: 41610428.
2. NCCN Clinical Practice Guidelines in Oncology. Multiple Myeloma. Current publicly available versions recommend infection prevention measures, vaccination, and therapy-specific supportive care in patients with myeloma.
3. Ludwig H, Sonneveld P, Davies F, et al. European perspective on multiple myeloma treatment strategies and supportive care, including infection prevention. Published guideline and consensus literature in myeloma supportive care provides broader clinical context for interpreting infection risk.