Изучение многочисленных патобиологии миэломы и сигнатур иммунных клеток: свидетельства из двунаправленного генетического анализа.

1. Introduction

Multiple myeloma (MM) is a hematological malignancy characterized by the proliferation of abnormal plasma cells. It represents one of the most prevalent types of blood cancer, accounting for 1% to 1.8% of all cancer cases. [ 1 ] The estimated annual incidence of MM in Europe ranges from 4.5 to 6.0 per 100,000 individuals, [ 2 ] with age-standardized incidence and mortality rates of 1.78 and 1.14 per 100,000 individuals, respectively. [ 3 ]

Current therapeutic approaches for MM encompass chemotherapy, targeted therapy, immunomodulation, stem cell transplantation, bispecific T cell engager therapy, chimeric antigen receptor T cell therapy, and supportive care. [ 4 , 5 ] Despite significant advances in treatment modalities that have improved patient survival rates, only 10% to 15% of patients achieve or exceed the expected survival duration. [ 4 , 6 ] This limited success rate underscores the necessity for further investigation into the underlying pathogenesis of MM to develop more effective therapeutic strategies.

The immune system plays a pivotal role in the development and progression of MM. [ 7 ] Immune cells, including lymphocytes (B and T cells), macrophages, monocytes, and granulocytes, constitute essential components of the body’s defense mechanism and participate in various inflammatory responses. [ 8 ] In MM, immune dysfunction is a hallmark of the clinically active form of the disease. The interaction between malignant plasma cell clones and immune cells within the bone marrow microenvironment facilitates the formation of a protumor niche. [ 7 , 9 ]

Recent single-cell analyses have provided unprecedented insights into the immune microenvironment of MM, revealing complex immune cell interactions and heterogeneity within the tumor microenvironment. Studies have demonstrated that MM progression is associated with significant alterations in immune cell composition and function at the single-cell level, including the emergence of distinct immune phenotypes that correlate with disease progression and patient outcomes. [ 10 ] Furthermore, bioinformatics analyses have identified specific immune dysregulation patterns and prognostic gene signatures that characterize different stages of MM evolution, highlighting the dynamic nature of immune-tumor interactions. [ 11 , 12 ] These findings emphasize that immune dysfunction in MM extends beyond simple cell count alterations to encompass complex transcriptional and functional changes that create an immunosuppressive microenvironment conducive to disease progression. [ 13 ]

Characteristic abnormalities in the composition of bone marrow immune cells have been observed in MM patients, with increased numbers of activated CD4 + T cells, CD8 + T cells, and natural killer (NK) cells. [ 14 , 15 ] As MM progresses, the immune cell population undergoes significant alterations, including decreased memory and immature CD4 + T cells and increased CD8 + effector T cells and regulatory T cells (Tregs). [ 16 ] The abundance of specific immune cell subtypes, such as CD8 + effector memory T cells, may also influence therapeutic outcomes. [ 17 ] Interventions targeting immune cells (T cells, NK cells, and monocytes/macrophages) have demonstrated promising therapeutic efficacy in MM management. [ 14 ]

Although alterations in immune cell profiles correlate closely with MM onset and progression, existing research has predominantly relied on observational cohort correlation analyses, which have inherent limitations in establishing causal relationships. While recent studies have characterized immune phenotypes and their prognostic significance through comprehensive bioinformatics approaches, the question remains whether these immune alterations represent causative factors or consequences of the disease process. Consequently, while specific immune cell alterations have been associated with MM, it remains unclear whether these changes represent causative factors or consequences of the disease. Furthermore, the causal relationship between MM onset and diverse immune cell profiles has not been comprehensively investigated.

Mendelian randomization (MR) provides a robust methodological framework for inferring causal effects by utilizing genetic variants associated with specific exposures (instrumental variables (IVs)) on clinical outcomes. [ 18 ] This approach is less susceptible to common confounding factors, such as postnatal environment and lifestyle, thereby reducing the risk of reverse causation. [ 18 , 19 ] Recent advances in genome-wide association studies (GWASs) and molecular pathway characterization have strengthened the application of MR in research.

To date, no study has comprehensively explored the causal relationship between various immune cell types and MM using MR methods. Existing studies have primarily examined associations involving single or specific immune cells with MM, leaving potential links between MM and a broader range of immune cell phenotypes largely unexplored. In this study, we leverage GWAS data and employ an MR approach to investigate the bidirectional causal relationships between different immune cells (encompassing 731 immunophenotypes) and MM at the genetic level, aiming to enhance our understanding of MM pathogenesis and potentially identify novel therapeutic targets.

2. Materials and methods

2.1. Immunity-wide GWAS data sources

GWAS summary statistics for immune traits were accessed through the GWAS Catalog, with the accession numbers GCST0001391–GCST0002121. [ 20 ] The dataset comprised 731 immunophenotypes, including absolute cell (AC) counts ( n = 118), median fluorescence intensities (MFIs, representing surface antigen levels, n = 389), morphological parameters (MPs, n = 32), and relative cell (RC) counts ( n = 192). The MFI, AC, and RC datasets provide phenotypic information on various immune cell types, such as B and classical dendritic cells (cDCs); maturation stages of T cells, monocytes, and myeloid cells; a combined T, B, and natural killer cell (TBNK) panel; and Tregs. The MP dataset pertains to cDCs and the TBNK panel. The initial GWAS dataset of immune traits utilized a cohort of 3757 European individuals without overlap.

2.2. MM GWAS data sources

The MM data were obtained from the integrative epidemiology unit (IEU) Open GWAS Project ( https://gwas.mrcieu.ac.uk/ ) and FinnGen ( https://www.finngen.fi/en ). All analytical procedures were based on the summarized data presented in Table S1 (Supplemental Digital Content, https://links.lww.com/MD/P862 ). The data exclusively comprised European populations. The included studies were subjected to ethical review and approval from their respective committees.

2.3. Study design

The MR study design (Fig. 1 ) fulfilled the 3 assumptions [ 21 ] and adhered to the STROBE-MR guidelines. [ 22 ] The 3 key MR hypotheses are strong correlation, independence, and exclusivity. The IVs were screened using the following standard criteria. The IV must be highly correlated with the exposure, and P < 1 × 10 −5 indicates a strong correlation (relevance hypothesis). [ 23 , 24 ] The IV must be independent of confounding factors (independence hypothesis). The IV has no direct relationship with the outcome, influencing it solely through the exposure variable (exclusivity hypothesis). The intercept of the MR-Egger regression model ( P > .05) further supported the absence of genetic pleiotropy.

2.4. Selection of IVs

Significant single-nucleotide polymorphisms (SNPs) were identified from the GWAS summary data of immune cells using a threshold of P < 1 × 10 −5 . A linkage disequilibrium coefficient ( r 2 ) of 0.001 and region width of 10,000 kb were employed to ensure the independence of each SNP and mitigate the impact of linkage disequilibrium on the findings. Single SNPs associated with potential confounders and outcomes were excluded using PhenoScanner.

The significant SNPs were extracted from the GWAS summary data of MM with a minimum r 2 > 0.8. The 2 datasets were summarized and SNPs directly related to MM ( P < 1 × 10 −5 ) were removed. F -statistics > 10 were used as criteria for eliminating weak IVs. [ 25 ] The same steps were used to screen relevant SNPs in reverse causal relationships.

2.5. MR analysis

This study used 2 regression models, MR-Egger regression and the random-effects inverse variance weighted (IVW) method, to verify the causal relationship between immune cells and MM. IVW analysis was used as the primary MR methodology, which omits an intercept term in the regression and is dependent on outcome variance for estimation in the absence of directed pleiotropy in the IVs. [ 26 ]

The Cochran Q test was used to determine the heterogeneity of SNPs. [ 27 ] The results were considered heterogeneous at P < .05. I 2 is another measure of heterogeneity that represents the proportion of variation caused by heterogeneity in the total variation. I 2 has a range of 0% to 100%, with I 2 >50% indicating a certain degree of heterogeneity, and is calculated as I 2 = ( Q − Q _df)/ ( Q × 100%).

We estimated the MR-Egger intercept to assess whether the observed associations are confounded by the horizontal pleiotropic effects of other traits. [ 28 ]

Leave-one-out analysis was performed on the IVs to assess the sensitivity of the results. A non-zero intercept in the MR-Egger regression model ( P > .05) indicates that the IVs do not have multiple effects. Each SNP was gradually removed, and the remaining SNPs were reanalyzed to observe the impact of each on the results.

All statistical analyses were performed in R (version 4.1.0, the R Foundation for Statistical Computing, Vienna, Austria), using the TwoSampleMR package.

2.6. Meta-analysis

The intersections of significant immune cells from different MM datasets and their MR estimates were used in the meta-analysis and tested for heterogeneity using the I 2 statistic, Cochran Q test, and τ 2 test.

The meta package was used for the analyses in R (version 4.1.0).

3. Results

3.1. SNPs associated with immune cells and MM

Following extensive screening, 33,878 SNPs associated with MM were included in the immune cell-related dataset. Basic information on the SNPs is presented in Table S2 (Supplemental Digital Content, https://links.lww.com/MD/P862 ) The individual SNPs had F -statistics of 19.548 to 2435.818 (mean 33.641), suggesting a reduced likelihood of weak IV bias affecting causality inference.

The final immune cell-related dataset associated with MM comprised 21,551 SNPs. Table S3 (Supplemental Digital Content, https://links.lww.com/MD/P862 ) displays the basic information for a subset of these SNPs. The individual SNPs had F -statistics of 19.535 to 27.549 (mean 21.330), implying a decreased potential for causality to be influenced by weak IV bias.

3.2. Causal effect of immunophenotypes on MM risk

Using immune cells as the exposure factor in the MR analysis of the MM IEU dataset revealed 42 immune cells with significant causal relationships to MM. Elevated and decreased levels of 18 and 24 immune cell types, respectively, were associated with an increased risk of MM (Fig. 2 ).

MR analysis of the FinnGen dataset identified 42 immune cells with potential causal links to MM. Elevated and reduced levels of 15 and 27 immune cell types, respectively, were correlated with an elevated risk of MM (Fig. 3 ).

Both datasets revealed 11 shared immune cells. These 11 immune cells within the 3 immune profiles (AC, RC, and MFI) were distributed in B cells (3 cells), cDCs (1 cell), the maturation stages of T cells (1 cell), myeloid cells (3 cells), TBNK (2 cells), and Treg panels (1 cell, Table S4, Supplemental Digital Content, https://links.lww.com/MD/P862 ).

IVW analysis of the IEU dataset identified 3 immune cell characteristics associated with an increased risk of MM, namely, CD39 + CD8br %CD8br [odds ratio (OR) = 1.152, 95% confidence interval (CI): 1.017–1.305, P = .026], CD20 on CD20 − CD38 − cells (OR = 1.281, 95% CI: 1.050–1.563, P = .015), and CD38 on transitional cells (OR = 1.192, 95% CI: 1.040–1.365, P = .011). Eight immune cell characteristics were identified with protective effects against MM, namely, CD11c + monocyte AC (OR = 0.830, 95% CI: 0.703–0.980, P = .028), Im myeloid-derived suppressor cell (MDSC) %CD33 dim HLA-DR − CD66b − (OR = 0.882, 95% CI: 0.785–0.990, P = .034), CD8 dim %T cell (OR = 0.798, 95% CI: 0.666–0.956, P = .014), CD8 dim %leukocyte (OR = 0.810, 95% CI: 0.696–0.941, P = .006), CD24 on CD24 + CD27 + cells (OR = 0.705, 95% CI: 0.508–0.977, P = .036), CD4 on naive CD4 + cells (OR = 0.827, 95% CI: 0.700–0.976, P = .025), CD11b on Mo MDSCs (OR = 0.892, 95% CI: 0.803–0.990, P = .032), and HLA-DR on CD33 dim HLA-DR + CD11b − cells (OR = 0.859, 95% CI: 0.759–0.973, P = .017).

IVW analysis of the FinnGen dataset identified 3 immune cell characteristics associated with an increased risk of MM, namely, CD39 + CD8br %CD8br (OR = 1.207, 95% CI: 1.073–1.356, P = .002), CD20 on CD20 − CD38 − cells (OR = 1.212, 95% CI: 1.006–1.459, P = .043), and CD38 on transitional cells (OR = 1.158, 95% CI: 1.014–1.323, P = .030). Eight immune cell characteristics were identified with protective effects against MM, namely, CD11c + monocyte AC (OR = 0.829, 95% CI: 0.700–0.981, P = .029), Im MDSC %CD33 dim HLA-DR − CD66b − (OR = 0.870, 95% CI: 0.768–0.987, P = .030), CD8 dim %T cell (OR = 0.800, 95% CI: 0.659–0.972, P = .025), CD8 dim %leukocyte (OR = 0.780, 95% CI: 0.671–0.907, P = .001), CD24 on CD24 + CD27 + cells (OR = 0.704, 95% CI: 0.506–0.981, P = .038), CD4 on naive CD4 + cells (OR = 0.770, 95% CI: 0.653–0.909, P = .002), CD11b on Mo MDSCs (OR = 0.887, 95% CI: 0.799–0.984, P = .023), and HLA-DR on CD33 dim HLA-DR + CD11b − cells (OR = 0.875, 95% CI: 0.784–0.977, P = .018).

Analyses of the IEU and FinnGen datasets revealed no heterogeneity between the 11 immune cell types and SNPs related to MM (Table 1 ). Scatter and funnel plots for both datasets demonstrated a symmetrical distribution of all the included SNPs in immune cells, suggesting a low likelihood for the causal relationships to be skewed by potential biases (Figs. S1 and S2, Supplemental Digital Content, https://links.lww.com/MD/P861 for IEU; Figs. S3 and S4, Supplemental Digital Content, https://links.lww.com/MD/P861 for FinnGen). The MR-Egger results from both datasets indicated no significant difference between the intercept term and zero (Table 1 ), implying a lack of horizontal pleiotropy among the SNPs. The robustness of the results was further confirmed by the leave-one-out tests. After sequentially removing SNPs, the results of the remaining SNPs remained consistent with those including all SNPs, and no individual SNP was found to significantly influence the causal estimation (Fig. S5 for IEU; Fig. S6 for FinnGen, Supplemental Digital Content, https://links.lww.com/MD/P861 ).

3.3. Meta-analysis of MR estimates for positive causality

Immune cells were used as exposure factors, and the MR estimates of immune cells from the intersections of the 2 MM datasets were used for the meta-analysis. The meta results identified 3 immune cell characteristics associated with an increased risk of MM, namely, CD39 + CD8br %CD8br ( P < .001), CD20 on CD20 − CD38 − cells ( P = .002), and CD38 on transitional cells ( P < .001). Eight immune cell characteristics were identified with protective effects against MM, namely, CD11c + monocyte AC ( P = .002), Im MDSC %CD33 dim HLA-DR − CD66b − ( P = .002), CD8 dim %T cell ( P < .001), CD8 dim %leukocyte ( P < .001), CD24 on CD24 + CD27 + cells ( P = .003), CD4 on naive CD4 + cells ( P < .001), CD11b on Mo MDSCs ( P = .002), and HLA-DR on CD33 dim HLA-DR + CD11b − cells ( P < .001). Cochran Q in the heterogeneity test indicated that the effect estimates of the different datasets were not heterogeneous (Fig. 4 ).

3.4. Causal effect of MM onset on immunophenotypes

We investigated the potential causal effects of MM on the immunophenotypes through the two-sample IVW-based MR analysis. The IEU and FinnGen datasets revealed 29 and 36 putative immunophenotypes, respectively (Figs. 5 and 6 ). Analysis of both datasets revealed 2 common immune cells comprising 1 B cell and 1 monocyte subtype. Analysis of the IEU dataset indicated that MM onset may elevate CD19 expression on CD24 + CD27 + B cells (OR = 1.035, 95% CI: 1.001–1.069, P = .042). Similarly, elevated CD19 expression on CD24 + CD27 + B cells was observed in the FinnGen dataset (OR = 1.045, 95% CI: 1.002–1.090, P = .041). CD16 expression on CD14 + CD16 + monocytes was increased in patients with MM in the IEU dataset (OR = 1.048, 95% CI: 1.014–1.084, P = .006). Similarly, the FinnGen dataset showed increased CD16 expression in CD14 + CD16 + monocytes (OR = 1.039, 95% CI: 1.002–1.078, P = .037; Table S5, Supplemental Digital Content, https://links.lww.com/MD/P862 ).

The sensitivity analysis results confirmed the robustness of the observed causal associations (Table 2 ). Specifically, the intercept of the MR-Egger model confirmed the absence of horizontal pleiotropy (Table 2 ). Scatter and funnel plots also indicated good stability in the results (Fig. S7, Supplemental Digital Content, https://links.lww.com/MD/P861 ).

3.5. Meta-analysis of MR estimates for reverse causality

MM was used as the exposure factor to merge CD19 with CD24 + CD27 + B cell estimates in a meta-analysis. Cochran Q in the heterogeneity test ( P = .71) indicated that the effect estimates of the different datasets were not heterogeneous. The random-effects model, indicated a significant causal relationship between MM and CD19 in CD24 + CD27 + B cells (OR = 1.038, 95% CI = 1.012–1.066, P = .004; Fig. 7 A). MM was used as the exposure factor to merge CD16 with CD14 + CD16 + monocyte estimates in a meta-analysis. Cochran Q in the heterogeneity test ( P = .74) indicated that the effect estimates of the different datasets were not heterogeneous. The random-effects model indicated a significant causal relationship between MM and CD16 expression in CD14 + CD16 + monocytes (OR = 1.044, 95% CI = 1.019–1.070, P < .001; Fig. 7 B).

4. Discussion

Our study employed a two-sample bidirectional MR approach to investigate causal relationships between immune cell traits and MM risk. To our knowledge, this represents the first comprehensive investigation of causal associations between a wide range of immune cell phenotypes (731 immunophenotypes) and MM susceptibility. We identified 11 immune cell traits distributed across 3 immune profiles – AC, RC, and MFI – that exhibited significant causal effects on MM risk. Furthermore, our reverse MR analysis revealed that MM development may causally influence the expression of specific markers on B cells and monocytes, suggesting a bidirectional relationship between immune cell phenotypes and MM pathobiology.

4.1. Immune cell phenotypes causally associated with MM risk

Our findings identified 3 immune cell characteristics associated with increased MM risk: CD39 + CD8br %CD8br, CD20 on CD20 − CD38 − cells, and CD38 on transitional cells. These results provide important insights into potential immunological mechanisms underlying MM development.

The association between elevated CD39 + CD8br %CD8br and increased MM risk is particularly noteworthy. CD39 (ENTPD1) is an ectonucleotidase that hydrolyzes extracellular ATP and ADP to AMP, which is subsequently converted to adenosine by CD73. [ 29 ] This adenosinergic pathway creates an immunosuppressive microenvironment that can facilitate tumor escape from immune surveillance. [ 30 ] CD39 expression on CD8 + T cells has been associated with an exhausted phenotype and diminished anti-tumor functionality. [ 31 ] Our findings suggest that individuals with genetically determined higher proportions of CD39 + CD8 + T cells may have compromised immunosurveillance against premalignant plasma cells, potentially facilitating MM development. This aligns with previous studies demonstrating that CD39 expression on T cells correlates with poor prognosis in various malignancies, [ 32 ] although this specific relationship had not been previously established in MM.

The causal relationship between CD20 expression on CD20 − CD38 − cells and MM risk presents an intriguing finding. This seemingly paradoxical result – CD20 expression on ostensibly CD20-negative cells – likely reflects low-level CD20 expression that falls below traditional binary classification thresholds but is still detectable by sensitive flow cytometry measurements. This subtle CD20 expression may identify a specific B cell subset with altered functionality that contributes to MM pathogenesis. Previous studies have demonstrated that CD20 expression patterns on B cells influence their activation status, antigen presentation capacity, and cytokine production, [ 33 ] which could modulate the pre-MM microenvironment.

The association between CD38 expression on transitional B cells and MM risk is particularly relevant given the role of CD38 in MM pathophysiology and therapeutics. CD38 is highly expressed on malignant plasma cells and represents a key therapeutic target for MM. [ 34 ] Our findings suggest that genetically determined higher CD38 expression on transitional B cells may influence the developmental trajectory of these cells, potentially facilitating their transformation into malignant plasma cells or altering their interaction with emerging clonal plasma cells. This observation provides genetic evidence supporting the importance of CD38-related pathways in MM pathogenesis beyond its role as a marker of malignant cells.

4.2. Protective immune cell phenotypes against MM development

Our MR analysis identified 8 immune cell characteristics with protective effects against MM development. The protective association of CD11c + monocyte absolute count (AC) suggests that these cells may play an important role in tumor immunosurveillance. CD11c + monocytes can differentiate into dendritic cells with potent antigen-presenting capabilities, [ 35 ] potentially enhancing immune recognition of premalignant plasma cells. Our findings provide genetic evidence supporting the importance of this cell population in preventing MM development, which aligns with previous observational studies showing reduced dendritic cell function in MM patients. [ 36 ]

The protective effect of immature myeloid-derived suppressor cells (Im MDSC %CD33 dim HLA-DR − CD66b − ) against MM is somewhat unexpected, given that MDSCs are typically associated with immunosuppression and tumor progression. [ 37 ] However, this finding suggests that the functional heterogeneity within MDSC populations may be more complex than previously recognized, with certain immature MDSC phenotypes potentially exhibiting anti-tumor properties. This warrants further investigation into the functional characteristics of specific MDSC subsets in the context of MM pathogenesis.

The protective associations of CD8 dim %T cell and CD8 dim %leukocyte with MM risk highlight the importance of CD8 + T cell populations in MM immunosurveillance. CD8 dim T cells represent a distinct subset with unique functional properties compared to conventional CD8 bright T cells. [ 38 ] Our findings suggest that these cells may possess enhanced anti-tumor activity against premalignant plasma cells, potentially through cytotoxic mechanisms or cytokine production that inhibits the expansion of abnormal plasma cell clones. This relationship provides genetic evidence supporting the role of specific CD8 + T cell subsets in preventing MM development.

The protective effect of CD24 expression on CD24 + CD27 + cells against MM risk is notable given that CD24 + CD27 + B cells represent a memory B cell population with distinct functional properties. [ 39 ] Higher CD24 expression on these cells may enhance their regulatory functions, potentially limiting the expansion of abnormal plasma cell clones. This finding suggests that memory B cell phenotypes may influence MM susceptibility, adding to our understanding of the complex interplay between various B cell subsets and MM pathogenesis.

The protective association of CD4 expression on naive CD4 + cells with MM risk suggests that appropriate CD4 signaling in naive T cells may be crucial for maintaining effective immunosurveillance against premalignant plasma cells. CD4 + T cells orchestrate immune responses through cytokine production and help to other immune cells, [ 40 ] and our findings provide genetic evidence that proper CD4 expression on naive T cells may be protective against MM development.

The protective effects of CD11b expression on monocytic MDSCs and HLA-DR expression on CD33 dim HLA-DR + CD11b − cells further emphasize the importance of myeloid cell phenotypes in MM pathogenesis. CD11b (integrin αM) mediates various cellular functions including adhesion, migration, and phagocytosis, [ 41 ] while HLA-DR is crucial for antigen presentation to CD4 + T cells. [ 42 ] These findings suggest that appropriate expression of these markers on specific myeloid cell populations may enhance immunosurveillance against premalignant plasma cells, potentially through improved antigen presentation or enhanced phagocytic capacity.

4.3. Bidirectional relationship: causal effects of MM on immune cell phenotypes

Our reverse MR analysis revealed that MM development may causally influence the expression of CD19 on CD24 + CD27 + B cells and CD16 on CD14 + CD16 + monocytes. These findings provide genetic evidence for MM-induced immune dysregulation, suggesting that once established, MM can alter specific immune cell phenotypes, potentially creating a more permissive environment for disease progression.

The causal effect of MM on CD19 expression in CD24 + CD27 + B cells is particularly interesting. CD19 is a crucial co-receptor in B cell activation, [ 43 ] and its altered expression on memory B cells (CD24 + CD27 + ) may reflect MM-induced dysregulation of B cell function. This finding aligns with previous observational studies showing abnormal B cell phenotypes in MM patients, [ 44 ] but uniquely establishes a causal relationship rather than mere association. This MM-induced alteration in CD19 expression might contribute to the well-documented humoral immune deficiency in MM patients, [ 45 ] potentially creating a feed-forward loop that promotes disease progression.

Similarly, the causal effect of MM on CD16 expression in CD14 + CD16 + monocytes suggests that MM can directly influence monocyte phenotypes. CD16 (FcγRIII) mediates antibody-dependent cellular cytotoxicity and phagocytosis, [ 46 ] and its altered expression on monocytes may impair these functions, potentially compromising anti-tumor immunity. This finding provides a genetic basis for the observed monocyte dysfunction in MM patients [ 47 ] and suggests that MM-induced changes in monocyte phenotypes may contribute to disease pathogenesis.

4.4. Clinical implications and future directions

Our genetic findings complement recent single-cell and bioinformatics analyses of MM immune dysregulation. [ 10 , 11 ] While previous studies characterized bone marrow immune dysfunction at the cellular level, our MR analysis establishes causal relationships between peripheral immune traits and MM susceptibility, suggesting systemic immune alterations may precede localized bone marrow dysfunction.

The protective effect of CD11c + monocytes identified in our study aligns with observations of reduced dendritic cell function in MM patients, [ 10 ] while our identification of CD39 + CD8 + T cells as risk factors supports previous findings of T cell exhaustion signatures in the MM microenvironment. [ 12 ] Furthermore, our findings suggest that inflammatory gene signatures predicting MM evolution [ 13 ] may originate from genetically determined peripheral immune traits, as evidenced by the bidirectional relationships we identified between immune cell phenotypes and MM development.

These causal relationships offer several translational applications. First, the identification of immune cell traits causally associated with MM risk provides potential targets for preventive interventions in high-risk individuals, such as those with monoclonal gammopathy of undetermined significance or smoldering MM. For instance, therapeutic strategies targeting CD39 on CD8 + T cells or modulating CD38 expression on transitional B cells might help prevent MM development in these high-risk populations. Second, our results highlight potential biomarkers for MM risk stratification. Measuring the identified immune cell traits in peripheral blood might help identify individuals at increased risk for MM development, enabling more targeted surveillance strategies. Furthermore, monitoring these immune parameters in patients with precursor conditions might provide insights into disease progression risk. Third, the identification of protective immune cell phenotypes offers opportunities for immunomodulatory interventions. Strategies aimed at increasing CD11c + monocyte counts or enhancing CD24 expression on CD24 + CD27 + B cells might help prevent MM development or progression. These approaches could complement existing therapeutic strategies focused on targeting malignant plasma cells directly. Lastly, our finding of a bidirectional relationship between immune cell phenotypes and MM emphasizes the importance of considering immune-tumor interactions in MM management. The causal effects of MM on specific immune cell traits suggest that early intervention might be crucial to prevent MM-induced immune dysregulation that could facilitate disease progression.

Future research should focus on validating these findings in prospective cohort studies and elucidating the molecular mechanisms underlying the identified causal relationships. Functional studies investigating how the implicated immune cell populations interact with premalignant plasma cells could provide valuable insights into MM pathogenesis. Additionally, exploring the therapeutic potential of modulating these immune cell traits in preclinical models of MM would be an important step toward translating these findings into clinical applications.

4.5. Strengths and limitations

Our study has several strengths. The use of a bidirectional MR approach allows for robust causal inference while minimizing confounding and reverse causation concerns inherent in observational studies. The large-scale analysis of 731 immunophenotypes provides a comprehensive assessment of immune cell traits potentially involved in MM pathogenesis. Furthermore, the consistency of findings across 2 independent datasets (IEU and FinnGen) and the validation through meta-analysis strengthen the reliability of our results.

However, there are limitations to consider. First, while MR establishes causal relationships at the genetic level, it does not elucidate the precise biological mechanisms underlying these associations. Second, our analysis was restricted to individuals of European ancestry, potentially limiting the generalizability of our findings to other populations. Third, the immune cell measurements were derived from peripheral blood, which may not fully reflect the immune microenvironment in the bone marrow where MM develops. Fourth, although we employed various sensitivity analyses to assess the robustness of our findings, the possibility of residual pleiotropy cannot be entirely ruled out.

5. Conclusion

Our bidirectional MR analysis provides genetic evidence for causal relationships between specific immune cell phenotypes and MM risk, as well as the causal effects of MM on certain immune cell traits. These findings enhance our understanding of the complex interplay between immune dysregulation and MM pathogenesis, offering insights into potential preventive and therapeutic strategies.

Acknowledgments

We acknowledge the participants and investigators of the FinnGen and IEU studies.

Author contributions

Conceptualization: Yue Zhang, Wenming Chen.

Data curation: Wenming Chen.

Formal analysis: Ying Tian, Wenming Chen.

Methodology: Ying Tian, Wenming Chen.

Software: Ying Tian.

Supervision: Yue Zhang, Wenming Chen.

Validation: Yue Zhang, Wenming Chen.

Visualization: Ying Tian.

Writing – original draft: Ying Tian.

Writing – review & editing: Yue Zhang, Wenming Chen.

Supplementary Material

Abbreviations: AC absolute count br bright cDC dendritic cell CI confidence interval CM central memory dim diminished DN double negative EM effector memory GWAS genome-wide association studies HSC hematopoietic stem cell Im immature IV instrumental variable IVW inverse variance weighted MDSC myeloid-derived suppressor cell MFI median fluorescence intensity MGUS monoclonal gammopathy of undetermined significance MM multiple myeloma MP morphological parameter MR Mendelian randomization NK natural killer OR odds ratio PB/PC plasmablast/plasma cell RC relative cell SNP single-nucleotide polymorphism TBNK T, B, and natural killer cells TD terminally differentiated Treg regulatory T cell

All data generated or analyzed during this study are included in this published article and its supplementary information files. GWAS summary statistics for each immune cell are publicly available from the GWAS Catalog (accession numbers GCST0001391–GCST0002121), and MM-related GWAS data can be obtained from the IEU Open GWAS project ( https://gwas.mrcieu.ac.uk/ ) and FinnGen Consortium ( https://www.finngen.fi/en ).

The views expressed in this article are those of the authors and do not necessarily represent the official positions of their affiliated institutions or of the data custodians (GWAS Catalog, MRC IEU OpenGWAS, and FinnGen). Any errors are the authors’ own.

Ethical approval is not applicable for this study.

The authors have no funding and conflicts of interest to disclose.

Supplemental Digital Content is available for this article.

How to cite this article: Tian Y, Chen W, Zhang Y. Exploring multiple myeloma pathobiology and immune cell signatures: Evidence from bidirectional genetic analysis. Medicine 2025;104:36(e44276).

References

[1] Siegel RL Miller KD Jemal A . Cancer statistics, 2020. CA Cancer J Clin . 2020 ; 70 : 7 – 30 . 31912902 10.3322/caac.21590

[2] Ferlay J Colombet M Soerjomataram I . Cancer incidence and mortality patterns in Europe: estimates for 40 countries and 25 major cancers in 2018. Eur J Cancer . 2018 ; 103 : 356 – 87 . 30100160 10.1016/j.ejca.2018.07.005

[3] Fitzmaurice C Abate D Abbasi N ; Global Burden of Disease Cancer Collaboration . Global, regional, and national cancer incidence, mortality, years of life lost, years lived with disability, and disability-adjusted life-years for 29 cancer groups, 1990 to 2017: a systematic analysis for the global burden of disease study. JAMA Oncol . 2019 ; 5 : 1749 – 68 . 31560378 10.1001/jamaoncol.2019.2996 PMC6777271

[4] Kumar SK Rajkumar V Kyle RA . Multiple myeloma. Nat Rev Dis Primers . 2017 ; 3 : 17046 . 28726797 10.1038/nrdp.2017.46

[5] Mikhael J Ismaila N Cheung MC . Treatment of multiple myeloma: ASCO and CCO joint clinical practice guideline. J Clin Oncol . 2019 ; 37 : 1228 – 63 . 30932732 10.1200/JCO.18.02096

[6] Palumbo A Anderson K . Multiple myeloma. N Engl J Med . 2011 ; 364 : 1046 – 60 . 21410373 10.1056/NEJMra1011442

[7] Pratt G Goodyear O Moss P . Immunodeficiency and immunotherapy in multiple myeloma. Br J Haematol . 2007 ; 138 : 563 – 79 . 17686051 10.1111/j.1365-2141.2007.06705.x

[8] Nicholson LB . The immune system. Essays Biochem . 2016 ; 60 : 275 – 301 . 27784777 10.1042/EBC20160017 PMC5091071

[9] Manier S Sacco A Leleu X Ghobrial IM Roccaro AM . Bone marrow microenvironment in multiple myeloma progression. J Biomed Biotechnol . 2012 ; 2012 : 157496 . 23093834 10.1155/2012/157496 PMC3471001

[10] Schinke C Poos AM Bauer M . Characterizing the role of the immune microenvironment in multiple myeloma progression at a single-cell level. Blood Adv . 2022 ; 6 : 5873 – 83 . 35977111 10.1182/bloodadvances.2022007217 PMC9647426

[11] Fang CF Li Y Yang C Fang H Li C . Bioinformatics analysis of intrinsic drivers of immune dysregulation in multiple myeloma to elucidate immune phenotypes and discover prognostic gene signatures. Sci Rep . 2025 ; 15 : 15662 . 40325058 10.1038/s41598-025-00074-7 PMC12053621

[12] Zhao Z Zhao Z Lin Z . Decoding multiple myeloma: single-cell insights into tumor heterogeneity, immune dynamics, and disease progression. Front Immunol . 2025 ; 16 : 1584350 . 40406148 10.3389/fimmu.2025.1584350 PMC12095158

[13] Botta C Di Martino MT Ciliberto D . A gene expression inflammatory signature specifically predicts multiple myeloma evolution and patients survival. Blood Cancer J . 2016 ; 6 : e511 . 27983725 10.1038/bcj.2016.118 PMC5223153

[14] Guillerey C Harjunpää H Carrié N . TIGIT immune checkpoint blockade restores CD8(+) T-cell immunity against multiple myeloma. Blood . 2018 ; 132 : 1689 – 94 . 29986909 10.1182/blood-2018-01-825265

[15] Minnie SA Kuns RD Gartlan KH . Myeloma escape after stem cell transplantation is a consequence of T-cell exhaustion and is prevented by TIGIT blockade. Blood . 2018 ; 132 : 1675 – 88 . 30154111 10.1182/blood-2018-01-825240

[16] Kawano Y Zavidij O Park J . Blocking IFNAR1 inhibits multiple myeloma-driven Treg expansion and immunosuppression. J Clin Invest . 2018 ; 128 : 2487 – 99 . 29558366 10.1172/JCI88169 PMC5983341

[17] Bailur JK McCachren SS Doxie DB . Early alterations in stem-like/resident T cells, innate and myeloid cells in the bone marrow in preneoplastic gammopathy. JCI Insight . 2019 ; 5 : e127807 . 31013254 10.1172/jci.insight.127807 PMC6629164

[18] Smith GD Ebrahim S . ‘Mendelian randomization’: can genetic epidemiology contribute to understanding environmental determinants of disease? Int J Epidemiol . 2003 ; 32 : 1 – 22 . 12689998 10.1093/ije/dyg070

[19] Little M . Mendelian randomization: methods for using genetic variants in causal estimation. J R Stat Soc Ser A: Stat Soc . 2018 ; 181 : 549 – 50 .

[20] Orrù V Steri M Sidore C . Complex genetic signatures in immune cells underlie autoimmunity and inform therapy. Nat Genet . 2020 ; 52 : 1036 – 45 . 32929287 10.1038/s41588-020-0684-4 PMC8517961

[21] Gagliano Taliun SA Evans DM . Ten simple rules for conducting a mendelian randomization study. PLoS Comput Biol . 2021 ; 17 : e1009238 . 34383747 10.1371/journal.pcbi.1009238 PMC8360373

[22] Skrivankova VW Richmond RC Woolf BAR . Strengthening the reporting of observational studies in epidemiology using Mendelian randomization: the STROBE-MR statement. JAMA . 2021 ; 326 : 1614 – 21 . 34698778 10.1001/jama.2021.18236

[23] Bowden J Holmes MV . Meta-analysis and Mendelian randomization: a review. Res Synth Methods . 2019 ; 10 : 486 – 96 . 30861319 10.1002/jrsm.1346 PMC6973275

[24] Li W Xu JW Chai JL . Complex causal association between genetically predicted 731 immunocyte phenotype and osteonecrosis: a bidirectional two-sample Mendelian randomization analysis. Int J Surg . 2024 ; 110 : 3285 – 93 . 38498404 10.1097/JS9.0000000000001327 PMC11175804

[25] Burgess S Thompson SG ; CRP CHD Genetics Collaboration . Avoiding bias from weak instruments in Mendelian randomization studies. Int J Epidemiol . 2011 ; 40 : 755 – 64 . 21414999 10.1093/ije/dyr036

[26] Burgess S Davey Smith G Davies NM . Guidelines for performing Mendelian randomization investigations: update for summer 2023. Wellcome Open Res . 2019 ; 4 : 186 . 32760811 10.12688/wellcomeopenres.15555.1 PMC7384151

[27] Burgess S Small DS Thompson SG . A review of instrumental variable estimators for Mendelian randomization. Stat Methods Med Res . 2017 ; 26 : 2333 – 55 . 26282889 10.1177/0962280215597579 PMC5642006

[28] Burgess S Thompson SG . Interpreting findings from Mendelian randomization using the MR-Egger method. Eur J Epidemiol . 2017 ; 32 : 377 – 89 . 28527048 10.1007/s10654-017-0255-x PMC5506233

[29] Allard B Longhi MS Robson SC Stagg J . The ectonucleotidases CD39 and CD73: Novel checkpoint inhibitor targets. Immunol Rev . 2017 ; 276 : 121 – 44 . 28258700 10.1111/imr.12528 PMC5338647

[30] Antonioli L Pacher P Vizi ES Haskó G . CD39 and CD73 in immunity and inflammation. Trends Mol Med . 2013 ; 19 : 355 – 67 . 23601906 10.1016/j.molmed.2013.03.005 PMC3674206

[31] Gupta PK Godec J Wolski D . CD39 expression identifies terminally exhausted CD8+ T cells. PLoS Pathog . 2015 ; 11 : e1005177 . 26485519 10.1371/journal.ppat.1005177 PMC4618999

[32] Li XY Moesta AK Xiao C . Targeting CD39 in cancer reveals an extracellular ATP- and inflammasome-driven tumor immunity. Cancer Discov . 2019 ; 9 : 1754 – 73 . 31699796 10.1158/2159-8290.CD-19-0541 PMC6891207

[33] Kuijpers TW Bende RJ Baars PA . CD20 deficiency in humans results in impaired T cell-independent antibody responses. J Clin Invest . 2010 ; 120 : 214 – 22 . 20038800 10.1172/JCI40231 PMC2798692

[34] van de Donk NWCJ Richardson PG Malavasi F . CD38 antibodies in multiple myeloma: back to the future. Blood . 2018 ; 131 : 13 – 29 . 29118010 10.1182/blood-2017-06-740944

[35] Collin M Bigley V . Human dendritic cell subsets: an update. Immunology . 2018 ; 154 : 3 – 20 . 29313948 10.1111/imm.12888 PMC5904714

[36] Leone P Berardi S Frassanito MA . Dendritic cells accumulate in the bone marrow of myeloma patients where they protect tumor plasma cells from CD8+ T-cell killing. Blood . 2015 ; 126 : 1443 – 51 . 26185130 10.1182/blood-2015-01-623975 PMC4592278

[37] Kumar V Patel S Tcyganov E Gabrilovich DI . The nature of myeloid-derived suppressor cells in the tumor microenvironment. Trends Immunol . 2016 ; 37 : 208 – 20 . 26858199 10.1016/j.it.2016.01.004 PMC4775398

[38] Caruso A Licenziati S Corulli M . Flow cytometric analysis of activation markers on stimulated T cells and their correlation with cell proliferation. Cytometry . 1997 ; 27 : 71 – 6 . 9000587 10.1002/(sici)1097-0320(19970101)27:1<71::aid-cyto9>3.0.co;2-o

[39] Klein U Rajewsky K Küppers R . Human immunoglobulin (Ig)M+IgD+ peripheral blood B cells expressing the CD27 cell surface antigen carry somatically mutated variable region genes: CD27 as a general marker for somatically mutated (memory) B cells. J Exp Med . 1998 ; 188 : 1679 – 89 . 9802980 10.1084/jem.188.9.1679 PMC2212515

[40] Zhu J Paul WE . CD4 T cells: fates, functions, and faults. Blood . 2008 ; 112 : 1557 – 69 . 18725574 10.1182/blood-2008-05-078154 PMC2518872

[41] Solovjov DA Pluskota E Plow EF . Distinct roles for the alpha and beta subunits in the functions of integrin alphaMbeta2. J Biol Chem . 2005 ; 280 : 1336 – 45 . 15485828 10.1074/jbc.M406968200

[42] Roche PA Furuta K . The ins and outs of MHC class II-mediated antigen processing and presentation. Nat Rev Immunol . 2015 ; 15 : 203 – 16 . 25720354 10.1038/nri3818 PMC6314495

[43] Wang K Wei G Liu D . CD19: a biomarker for B cell development, lymphoma diagnosis and therapy. Exp Hematol Oncol . 2012 ; 1 : 36 . 23210908 10.1186/2162-3619-1-36 PMC3520838

[44] Pojero F Flores-Montero J Sanoja L ; EuroFlow group . Utility of CD54, CD229, and CD319 for the identification of plasma cells in patients with clonal plasma cell diseases. Cytometry B Clin Cytom . 2016 ; 90 : 91 – 100 . 26130131 10.1002/cyto.b.21269

[45] Tete SM Bijl M Sahota SS Bos NA . Immune defects in the risk of infection and response to vaccination in monoclonal gammopathy of undetermined significance and multiple myeloma. Front Immunol . 2014 ; 5 : 257 . 24917865 10.3389/fimmu.2014.00257 PMC4042361

[46] Guilliams M Bruhns P Saeys Y Hammad H Lambrecht BN . The function of Fcγ receptors in dendritic cells and macrophages. Nat Rev Immunol . 2014 ; 14 : 94 – 108 . 24445665 10.1038/nri3582

[47] Petty AJ Yang Y . Tumor-associated macrophages: implications in cancer immunotherapy. Immunotherapy . 2017 ; 9 : 289 – 302 . 28231720 10.2217/imt-2016-0135 PMC5619052