CD24 expression for better outcome prediction in newly diagnosed multiple myeloma patients

Abstract

This study investigates the clinical prognostic value of CD24 antigen expression level in patients with newly diagnosed multiple myeloma. A retrospective cohort study design was used to quantify the intensity of CD24 membrane surface expression in bone marrow specimens of 54 patients with primary diagnosis of multiple myeloma (MM) by flow cytometry, and the cohort was divided into a high-expression group (n = 24) and a low-expression group (n = 30) using the median expression value (5.05%) as the threshold. Baseline clinical characteristics of patients in the 2 groups were systematically collected, including age, Durie–Salmon stage, ISS stage, β2-microglobulin, serum lactate dehydrogenase, and other parameters, and the survival curves were plotted using the Kaplan–Meier method and compared with the differences in overall survival by the log-rank test. The Cox proportional risk regression model was further applied to adjust for potential confounders. The median overall survival in the CD24 high-expression group was significantly better than in the low-expression group (50 months vs 24 months, hazard ratio = 0.41, 95% confidence interval = 0.21–0.97, P = .04). Multifactorial analysis showed that high CD24 expression could be used as an independent prognostic factor, and its prognostic value was independent of ISS staging and cytogenetic risk stratification. Differential expression of CD24 in patients with a primary diagnosis of MM has a significant prognostic stratification value, and the high-expression profile suggests a superior survival expectancy, this biomarker may influence the disease process by modulating plasma cell adhesion-mediated drug resistance mechanisms, providing a new laboratory basis for precise prognostic assessment in MM.

1. Introduction

Multiple myeloma (MM) is a clonal plasma cell malignant hematological systemic tumor with significant molecular heterogeneity, accounting for approximately 10% of hematologic malignancies.^[1] Its biological behavior and clinical progression are driven by complex genomic instability. It exhibits a highly diversified trajectory of clonal evolution, and its clinical prognosis is closely related to the cytogenetic features, clonal evolutionary patterns and bone marrow microenvironmental remodeling.^[2]

Although recent breakthroughs in novel therapeutic strategies, including the clinical application of proteasome inhibitors (bortezomib, carfilzomib), immunomodulatory drugs (lenalidomide, pomalidomide), CD38 monoclonal antibodies (dalteplumab), and BCMA-targeted CAR-T cell therapies, have dramatically improved overall remission rates (ORR ≥ 90%) and complete remission rates, MM is still characterized by incurability.^[3,4] Studies have shown that approximately 20% to 30% of patients have primary resistance due to high-risk cytogenetic abnormalities (t(4;14), del(17p)) or secondary plasma cell leukemia, and that clonal evolution-driven acquired resistance is a significant barrier to long-term survival.^[5]

Studies have shown that the genetic abnormalities of MM present multilayered features, with approximately 60% of patients having IgH translocations (t(11;14), t(4;14)) and 40% carrying chromosome 1q gain/amplification (1q+) at the time of initial diagnosis, and this molecular heterogeneity directly leads to significant prognostic variability of MM patients, with a low-risk group (hyperdiploid or t(11;14)) having median overall survival (OS) can reach more than 10 years, while the 5-year OS rate of patients with high-risk markers such as del(17p)/amp(1q32) is <40%, highlighting the importance of accurate prognostic stratification.^[6] Currently, international prognostic evaluation systems (revised international staging system [R-ISS], R2-ISS) have achieved quantitative stratification of survival outcomes in MM patients by integrating serum β2 microglobulin (β2-MG), lactate dehydrogenase, and cytogenetic risk factors (del(17p), t(4;14), 1q+, etc).^[7] Current risk stratification systems based on cytogenetic methodologies—primarily fluorescence in situ hybridization and karyotyping—constitute a cornerstone of clinical management in MM. However, their inherent limitations increasingly preclude sole reliance for precise risk stratification. Key constraints include: tumor heterogeneity and sampling bias leading to clonal underrepresentation; temporal clonal dynamics and disease evolution; and complex genetic interaction patterns extending beyond singular molecular markers. However, there are still blind spots in the analysis of tumor microenvironment (infiltration of immunosuppressive cells) and epigenetic regulators in existing models, so there is an urgent need to explore novel biomarkers to improve the prognostic assessment system.

In recent years, CD24, a glycosylphosphatidylinositol-anchored membrane protein, has attracted attention for its potential role in the regulation of plasma cell adhesion-mediated drug resistance and immune checkpoints such as the sialic-acid-binding Ig-like lectin 10 (Siglec-10)/CD24 signaling axis,^[8] CD24, an early B-cell differentiation marker, may influence MM progression by modulating bone marrow microenvironment interactions.^[9] However, significant controversy remains regarding the correlation between CD24 expression and MM prognosis: Gross Even-Zohar study indicated that high CD24 expression (5%) was associated with prolonged progression-free survival (PFS) (hazard ratio [HR] = 0.975, 95% confidence interval [CI] = 0.960–0.991), and significantly correlated with a better OS (P = .017).^[10] Elina Alaterre study found that CD24 significantly downregulated in malignant plasma cells from MM patients (ratio = 0.40, FDR < 0.05).^[11] CD24 + MM cells were found to exhibit lower tumorigenicity, such as decreased ability to migrate and form clones, along with increased apoptosis, which suggests that CD24 may suppress the malignant phenotype of MM through regulating the microenvironment on its expression of CXCR4 and CD38.^[9] However, Minjie Gao study showed that MM patients with a high initial percentage of CD24 + MM cells had inferior PFS (HR = 3.81, 95% CI = 5.66–18.34, P < .001) and overall survival (HR = 3.87, 95% CI = 16.61–34.39, P = .002).^[12] Sun, F study finds that CD24 + MM cells are not sensitive to bortezomib and dexamethasone, and correlated with a high proportion of CD24 + cells in samples from relapsed patients, CD24-positive cells causing immune-cell inhibition by modulating chemokines and macrophage polarization, leading to treatment resistance and relapse.^[8] Therefore, it is crucial to investigate the clinical prognostic significance of CD24 antigen expression levels in newly diagnosed multiple myeloma (NDMM), aiming to establish a novel laboratory-based approach for precise prognostic evaluation of MM.

2. Methods

2.1. Patient samples

The clinical data of 54 patients NDMM admitted to the Department of Hematology, Affiliated Hospital of Qinghai University for initial diagnosis between December 2019 and June 2020 were retrospectively collected. To be eligible, The diagnosis of MM was established following Criteria in the Chinese Guidelines for the Diagnosis and Treatment of Multiple Myeloma (Revised 2022), and high-risk cytogenetic abnormalities, including del(17p), t(4;14), and/or t(14;16), were identified based on the R-ISS criteria.^[13] Before enrollment, none of the patients had undergone any relevant treatment, and all had completed pretreatment bone marrow examinations and flow cytometric analysis of bone marrow samples, studies were conducted under the Declaration of Helsinki. They were approved by the institutional review boards at the Ethics Committee of Qinghai University Hospital (P-SL-2022109), with written informed consent obtained from all participants.

2.2. Data collection

Data abstraction encompassed baseline demographics (gender, age, ethnicity, and residential altitude), disease characteristics (subtype, ISS stage), laboratory parameters (hemoglobin, platelets, albumin, creatinine, serum calcium, and β2-MG), therapeutic interventions, and survival outcomes (OS). Electronic medical record extraction and comparative analysis were conducted following institutional review board protocols.

Gene expression profiling: GEO accession numbers are GSE4581 (http://www.ncbi.nlm.nih.gov/geo), which includes 414 patients with MM. The correlation between CD24 expression and survival in MM patients was analyzed using the “survival” and “survminer” packages.

2.3. Flow cytometry analysis of MM samples

Flow cytometric analysis of 54 fresh bone marrow specimens was performed using a multicolor antibody panel (CD24) for immunophenotypic characterization.

2.4. Statistical analysis

Statistical analyses were conducted using SPSS 22.0. Categorical variables were presented as percentages and analyzed using χ² tests or Fisher exact tests, while continuous variables were expressed as mean ± standard deviation and compared via t tests. Survival outcomes were evaluated through Kaplan–Meier methodology with log-rank tests for curve comparisons. Cox proportional hazards regression models were employed to assess quantitative predictors of overall survival. Bivariate correlations were examined using Pearson coefficient. All statistical inferences were based on 2-tailed tests with the α-level set at 0.05 for significance determination.

3. Results

3.1. CD24 expression stratification and clinicopathological correlations

Fifty-four patients with NDMM were enrolled in this study and categorized into CD24 high-expression group (n = 24) and low-expression group (n = 30) based on the median value of CD24 expression (5.05%) detected by flow cytometry. The overall cohort was 60.5% (32/54) male and 48.1% (26/54) aged >65 years. There were no significant differences between the 2 groups in terms of gender (males: CD24-high 58.3% vs CD24-low 63.3%), age (>65 years old: CD24-high 45.8% vs CD24-low 50.0%), race, place of residence, R-ISS staging, treatment regimen, and degree of bone destruction (P > .05). Notably, cytogenetic characteristics showed statistically significant differences between the 2 groups (P < .05) (Table 1). The expression of CD24 may be related to abnormal cytogenetic indicators, but due to the small number of samples, whether this is relevant or not should be investigated further.

Table 1.

The clinical parameters of NDMM in CD24-high/low group.

Characteristics	CD24-high	CD24-low	Total	χ 2	P
Gender
Male	15 (62.50%)	23 (76.67%)	38 (70.37%)	1.28	.26
Female	9 (37.50%)	7 (23.33%)	16 (29.63%)
Age
≥65	10 (41.67%)	16 (53.33%)	26 (48.15%)	0.73	.39
<65	14 (58.33%)	14 (46.67%)	28 (51.85%)
Ethnic groups
Han Chinese	18 (75.00%)	20 (66.67%)	38 (70.37%)	0.44	.51
Ethnic minorities	6 (25.0%)	10 (33.33%)	16 (29.63%)
Region of residence
High altitude	10 (42.67%)	9 (30.00%)	19 (35.19%)	0.80	.37
Low altitude	14 (58.33%)	21 (70.00%)	35 (64.81%)
R-ISS
I/II	11 (45.83%)	11 (36.67%)	22 (40.74%)	0.46	.50
III	13 (54.17%)	19 (63.33%)	32 (59.26%)
Cytogenetics
Standard risk	5 (20.83%)	14 (46.67%)	19 (35.19%)	6.20	.04*
High risk	7 (29.17%)	10 (33.33%)	17 (31.48%)
Missing	12 (50.00%)	6 (20.00%)	18 (33.33%)
M-protein
IgA	10 (41.67%)	13 (43.34%)	23 (42.59%)	0.94	.82
IgG	9 (37.50%)	9 (30.00%)	18 (33.33%)
LC only	4 (16.67%)	7 (23.33%)	11 (20.37%)
Nonsecretory	1 (4.16%)	1 (3.3%)	2 (3.71%)
Bone destruction
≥3	10 (41.67%)	14 (46.67%)	24 (44.44%)	0.14	.71
<3	14 (58.33%)	16 (53.33%)	30 (55.56%)
Treatment
Chemotherapy	19 (79.20%)	27 (90.00%)	46 (85.19%)	0.53	.47
Chemotherapy + ASCT	5 (20.80%)	3 (10.00%)	8 (14.81%)

High altitude: ≥2500 m, low altitude:<2500 m.

ASCT = autologous stem cell transplantation, IgA = immunoglobulin, LC = light chain, R-ISS = revised international staging system, IgG = immunoglobulin G.

^*P < .05.

3.2. CD24 expression and biomarker profile correlations

Comparative analysis revealed no significant intergroup differences in leukocyte count, platelet levels, hemoglobin, albumin, lactate dehydrogenase, serum calcium, or β2-MG (P > .05; Fig. 1A–H). Notably, the CD24-high group exhibited reduced creatinine and β2-MG levels relative to CD24-low counterparts.

Figure 1.

Association between CD24 expression with clinical parameters in MM patients (mean ± SD) n₁ = 24, n₂ = 30. ALB = serum albumin, Cr = creatinine, Hb = hemoglobin, LDH = lactate dehydrogenase, PLT = blood platelet, WBC = white blood cell, β2-MG = β2-microglobulin, n₁ = CD24-high group, n₂ = CD24-low group.

3.3. Survival analysis

Survival analysis was conducted on 54 patients completing ≥ 2 Bortezomib-based therapy with dexamethasone chemotherapy regimens (CD24-high: n = 24; CD24-low: n = 30), with follow-up performed via telephone to ascertain survival status and minimize attrition. Follow-up concluded on February 28, 2025, with durations ranging from 1 to 70 months. OS, defined as the interval from MM diagnosis to death or last follow-up, demonstrated significant intergroup disparity: CD24-high patients exhibited prolonged OS compared to CD24-low counterparts (median OS: P < .05; HR = 0.41, 95% CI 0.21–0.79), as shown in Figure 2.

Figure 2.

Overall survival stratification by CD24 expression status.

Furthermore, analysis of the GSE4581 dataset (n = 414) revealed a significant association between elevated CD24 expression and prolonged overall survival in MM patients (P < .05; Fig. 3). CD24-high patients demonstrated superior survival outcomes compared to CD24-low counterparts.

Figure 3.

Survival analysis of MM patients with CD24 expression based on GEO database.

3.4. Correlation analysis

Pearson correlation analysis demonstrated significant positive associations between CD24 expression and survival time, with inverse correlations observed for β2-MG and creatinine (Fig. 4) (P < .05).

Figure 4.

CD24 expression and serum biomarker correlations.

Furthermore, we conducted 2 multivariate Cox regression analyses. The first included the cytogenetics and CD24 and showed that the latter did not retain its significance as an OS predictor (P = .533), although there was a statistical trend towards significance when the cytogenetics and CD24 OS were assessed (P = .04, HR = 0.41 with a CI of 0.21–0.97). The second analysis included all variables that were found to be significant in the univariate analysis (Table 2).

Table 2.

Multivariate analysis showing the statistically significant factors correlating with poorer OS and CD24%.

	Hazard ratio	Confidence interval	P-value
Low-risk cytogenetics	0.107	0.025–0.451	.002*
ALB	0.21	0.044–1.009	.051
Cr	6.629	1.230–35.736	.028*
CD24%	0.726	0.265–1.990	.533

ALB = albumin, Cr = creatinine, OS = overall survive.

^*P < .05.

4. Discussion

MM, an incurable hematological malignancy, originates from the clonal expansion of aberrant plasma cells (PCs) within bone marrow microenvironments.^[14] While normal PCs emerge through antigen-driven polyclonal differentiation of B lymphocytes, malignant transformation drives PCs to secrete monoclonal immunoglobulins (M-proteins), inducing pathognomonic complications including osteolytic lesions, renal impairment, and hypercalcemia.^[10] Although therapeutic advances such as proteasome inhibitors (bortezomib) and immunomodulatory drugs (lenalidomide) have extended median survival from 3 to over 8 years, marked interpatient heterogeneity persists in treatment responses and clinical outcomes. To address this prognostic variability, we conducted a comprehensive evaluation of CD24 a glycosylphosphatidylinositol-anchored glycoprotein implicated in tumor immune evasion and drug resistance—utilizing multiparametric flow cytometry. This study pioneers the systematic investigation of CD24 expression patterns, their correlation with R-ISS criteria, and prognostic stratification potential in 54 treatment-naïve MM patients.

CD24, a glycosylphosphatidylinositol-anchored glycoprotein, is expressed on hematopoietic, neuronal, muscular, and epithelial cells, serving as a differentiation marker during B-cell development with downregulation upon germinal center entry.^[15] Within the bone marrow microenvironment, CD24 demonstrates functional duality: in solid tumors, hypoxia activates HIF-1α to bind the CD24 promoter, enhancing transcriptional activity.^[16] This hypoxic induction facilitates tumor immune evasion via CD24–Siglec-10 interactions^[17] and promotes metabolic reprogramming through glycolytic pathway regulation.^[18] Our cohort analysis revealed no significant altitude-dependent CD24 expression variation in MM patients (P > .05), contrasting with established hypoxic induction mechanisms in solid tumors. Further investigation is warranted to clarify: altitudinal hypoxia effects on CD24 in hematological malignancies and conserved versus divergent CD24 regulatory mechanisms between solid/hematologic tumors.

CD24 has emerged as a dual-functional biomarker in oncology, serving both as a diagnostic/prognostic marker in solid tumors^[19] and a cancer stem cell indicator.^[20] It mediates immune evasion through proliferation/metastasis promotion and immune checkpoint inhibition (CD24–Siglec-10).^[21] In MM, current evidence presents conflicting findings, Gross Even-Zohar et al reported CD24 mRNA downregulation in MM PCs correlating with poor survival (HR = 1.89),^[10] Another study also showed that CD24 + MM PCs are less tumorigenic phenotypes than CD24‐ MM PCs^[11] through regulating the microenvironment on its expression of CXCR4 and CD38.^[9] Our data confirm this trend (CD24 > 5%: median OS 50 vs 24 months, P = .04). Both studies demonstrate a favorable prognostic role of CD24 in multiple myeloma (MM), with elevated expression correlating with improved survival outcomes. Grossi Even-Zohar et al conducted a single-center retrospective study analyzing 124 NDMM patients receiving first-line bortezomib-based therapy. In a separate investigation, Alaterre et al incorporated CD24 into a multigene prognostic scoring system, with gene expression profiling demonstrating differential CD24 expression between normal and malignant plasma cells. In our current study, we systematically evaluated CD24 expression patterns in primary multiple myeloma specimens. Although inter-study methodological and sample variations may lead to subtle differences, the principal conclusions show consistent directional agreement.

However, Gao et al conversely associated CD24 + PCs > 2% with advanced ISS, elevated β2-MG, and reduced PFS/OS (HR = 2.17).^[12] Our results demonstrated a significant inverse correlation between CD24 expression levels and both creatinine and β2-MG concentrations.

Sun et al demonstrate that CD24 + multiple myeloma cells exhibit therapeutic resistance to bortezomib/dexamethasone, with elevated CD24 + cell frequencies observed in relapsed patients; these cells mediate immune suppression through chemokine modulation and macrophage polarization, contributing to treatment resistance and disease recurrence. CD24 + PCs dominate residual disease post-BCMA–CAR-T therapy. Preclinical studies show dual-targeting (BCMA–CD24–CAR-T) enhances macrophage phagocytosis via CD24–Siglec-10 blockade, achieving 92% tumor clearance versus 68% with single-target therapy.^[8] The role of CD24 expression in multiple myeloma remains controversial, possibly due to multi-investigator study methods, sample size and treatment differences, and further sample expansion and comparative studies at different stages of the disease are needed.

5. Conclusion

CD24 expression heterogeneity serves as an independent prognostic determinant in NDMM, demonstrating significant stratification capacity through its correlation with prolonged OS. While these findings position CD24 as a dual-functional biomarker-therapeutic target for risk-adapted management, 3 critical limitations constrain clinical extrapolation: Underpowered cohort (n = 54) increases type II error likelihood and limits population generalizability; Uncontrolled therapeutic variability introduces endpoint confounding, mandating treatment-adjusted survival analyses; Geographically confined Caucasian-dominant sampling obscures ethnic-specific pathobiological insights, particularly concerning documented interracial disparities in myeloma genomics. Multicenter prospective trials employing ethnically stratified cohorts are essential to verify translational applicability.

Acknowledgments

The authors wish to express their heart-felt gratitude for the facility support by the Affiliated Hospital of Qinghai University, China.

Author contributions

Data curation: Hui Geng.

Formal analysis: Lu Wang, Qichao Yin.

Methodology: Youliang Li.

Project administration: Lu Wang.

Supervision: Hui Geng, Jie Ma.

Writing – original draft: Lu Wang.

Writing – review & editing: Yan Jiang.