Multiple myeloma (MM) is a hematologically malignant clonal plasma cell disease that originates from terminally differentiated B cells. It is characterized by abnormal clonal plasma cells that produce monoclonal immunoglobulins, leading to skeletal destruction, renal impairment, anemia, and hypercalcemia
[1]. The multiparameter flow cytometry approach, which utilizes multiple parameters to assess the prognosis, monitor minimal residual disease, and evaluate the extent of complete response in bone marrow, has made immunophenotypic markers increasingly important
[2]. Combining relevant studies on these immunophenotypic features can provide evidence for improving risk stratification in MM.
CD200 (OX-2 membrane glycoprotein) is a cell membrane protein that is widely expressed on tumor cells, while its receptor, CD200R, is predominantly expressed on immune cells, such as macrophages and dendritic cells. CD200 plays a crucial role in the tumor microenvironment and is recognized as a key factor in tumor-induced immunosuppression. Its immuno-modulatory function is mediated by inhibiting the anti-tumor immune responses through CD200R expression, positioning it as a promising target for immune checkpoint blockade therapies
[3]. CD200 has been confirmed to be a diagnostic and prognostic marker for hematological diseases. A study
[4] has proven that CD200 is correlated with the prognosis of acute myeloid leukemia and lymphoid malignancies. While some studies
[5-8] have evaluated the correlation between CD200 and MM prognosis. However, the results remain inconsistent.
CD45, also known as Ly-5 or leukocyte common antigen, is a 180-220 kD (1 D=1 u) receptor-type protein tyrosine phosphatase expressed on the surface of nucleated hematopoietic cells and their precursors
[9]. A high level of CD45 expression has been shown to be associated with a high cumulative relapse incidence in acute lymphoblastic leukemia
[10]. However, the role of CD45 in the pathogenesis and prognosis of MM remains unclear.
This study found that CD45 and CD200 can serve as prognostic biomarkers for newly diagnosed MM (NDMM) patients. It is not known whether the combination of 2 biomarkers can help predict the prognosis of NDMM better than each biomarker separately. This study aims to explore the correlation between the expression of CD45/CD200 and the prognosis of MM patients, as well as to determine whether the combination of CD45 and CD200 serves as a better predictor of the prognosis of MM than CD45 or CD200 alone.
1 Materials and methods
1.1 Ethics statement
The study complied with the Declaration of Helsinki and was approved by the Ethics Committee of the Shengjing Hospital of China Medical University (approval number: 2024PS148K). Written informed consent was obtained from all participants.
1.2 Patient selection and data collection
A total of 123 patients (66 male; 57 female) newly diagnosed with MM at Shengjing Hospital of China Medical University from July 2015 to August 2019 were included in this observational retrospective study.
The inclusion criteria were: 1) Patients who were newly diagnosed with MM; 2) flow cytometry was performed during the first visit. Exclusion criteria were: 1) Diagnosis of asymptomatic myeloma, solitary plasmacytoma, plasma cell leukemia, reactive plasmacytosis, macroglobulinemia, polyneuropathy, organmegaly, endocrinopathy, M-protein, skin changes (POEMS) syndrome, or amyloidosis; 2) concurrent other malignant diseases.
Each patient was diagnosed with MM according to the criteria prescribed by the International Myeloma Working Group and was staged according to both the Durie-Salmon (DS) and the International Staging System (ISS)
[11]. Collected data included gender, age, DS stage, ISS stage, revised ISS (R-ISS) stage, fluorescence in situ hybridization (FISH) results, white blood cell (WBC) count, hemoglobin level, platelet count, albumin level, creatinine level, serum calcium level, serum β2-microglobulin (β2-MG) level, and serum lactate dehydrogenase (LDH) level. Treatment regimens affect the patient’s prognosis. In this study, 91 patients received bortezomib-based chemotherapy [bortezomib-cyclophosphamide-dexamethasone (VCD), bortezomib-lenalidomide-dexamethasone (VRD), bortezomib- doxorubicin-dexamethasone (PAD), bortezomib-dexamethasone (VD), bortezomib-thalidomide、dexamethason (VTD) regimens] and 32 patients received non-bortezomib-based chemotherapy [cyclophosphamide-thalidomide-dexamethason (CTD) and thalidomide-dexamethasone (TD) regimens]. The process of patient selection is shown in Supplementary
Figure 1 (
https://doi.org/10.57760/sciencedb.28338).
Overall survival (OS) is the time from diagnosis to death from any cause, or the last follow-up time. Progression-free survival (PFS) is defined as the time from diagnosis to the occurrence of disease progression or relapse. Patients were followed up for 2 to 65 months (median 29 months) and follow-up continued until December 31, 2020.
1.3 Flow cytometry
Fresh bone marrow specimens were anticoagulated using ethylenediaminetetraacetic acid and processed within 12 h after collection for immunophenotyping analysis. Red blood cells were lysed with lysis buffer (Becton, Dickinson and Company, USA). After antibody labeling, 30 000 cells were obtained with Cell Quest software of a 4-color flow cytometer. The cell population to be tested was gated by CD45/SSC or CD200/SSC, and antigen expression was analyzed for each group. The fluorescently labeled mouse anti-human monoclonal antibodies used [CD19-fluorescein isothiocyanate (FITC), CD20-FITC, human leucocyte antigen (HLA)-DR-FITC, CD13-FITC, CD16-FITC, CD5-FITC, CD33-FITC, anti-kappa-FITC, CD138-phycoerythrin (PE), CD22-PE, CD34-PE, CD117-PE, CD56-PE, CD10-PE, CD3-PE, anti-lambda-PE, CD45-Perc P, and CD38-allophycocyanin (APC)] were from Becton, Dickinson and Company, USA. In the results, cells with positive expression accounted for more than 80% of the cell population as strong positive expression, between 20% and 80% as weak positive expression, and less than 20% as negative expression, as previously reported
[6]. Based on the expression levels of CD45 and CD200, the patients were divided into 3 groups (Group A: CD45 and CD200 both positive; Group B: CD45 positive or CD200 positive; Group C: CD45 and CD200 both negative). In subsequent analyses, Group B and Group C were combined into a Group B+C and compared against Group A.
1.4 Random survival forest analysis
Random survival forest (RSF) algorithm was used to rank the importance of immunophenotypic markers. As a learning technique, RSF generates numerous decision trees and outputs the classifications (in the case of classification) generated by each tree
[12]. Based on majority voting, the predicted class of the input instance is determined. A bootstrap training set was used for each tree, which represented approximately 2/3 of the discovery cohort with replacement. Different models were trained using the 70% training data and the 30% test data. Five hundred trees were used to train the model for the training set, and the accuracy of the prediction was determined using the test set. We selected the model with the best performance based on the results. Next, the model was optimized according to the number of variables selected for each tree, and validation of the classifier was performed using the test set. The receiver operating characteristic (ROC) curve analysis was used to evaluate the performance.
1.5 Statistical analysis
Statistical analyses were performed using R version 4.0.3. Survival analysis was performed with the log-rank test. Univariate and multivariate Cox regression analyses were conducted to identify independent factors associated with MM. Based on the multivariate Cox models, a nomogram was established to predict the 2-, 3-, and 4-year survival of NDMM patients. Finally, the calibration curve was used to evaluate the accuracy and resolution of the nomogram. For categorical variables, the data were recorded as numbers and percentages and we used the Chi-square test and Fisher’s exact test. For continuous variables, the results were expressed as mean±standard deviation (SD) and were analyzed with the Mann-Whitney U and Kruskal-Wallis tests. P<0.05 was considered statistically significant.
2 Results
2.1 Clinical features
According to Supplementary
Table 1 (
https://doi. org/10.57760/sciencedb.28338), there were 123 NDMM patients with a gender ratio of 1.16 (male/female). The mean age was 60 (47 to 69) years. FISH results, hemoglobin, and platelet levels differed significantly among groups A (CD45-positive and CD200-positive,
n=24), B (CD45-positive or CD200-positive,
n=17), and C (CD45-negative and CD200-negative,
n=82), and the hemoglobin level was significantly different between Group A and Group B+C (all
P<0.05). Moreover, none of the other factors were significantly different between the groups (all
P>0.05).
2.2 Immunophenotype analysis and survival prediction
Through flow cytometry, we obtained the expression levels of various immunophenotypic markers, including CD45, CD200, CD38, CD138, CD20, CD27, CD117, CD56, et al.
Figure 1A illustrates the training process for RSF model. As the number of decision trees increased, the error rate of the model tended to stabilize.
Figure 1B shows the 5 most important immunophenotypic markers, which are CD38, CD138, CD45, CD56, and CD200. The model was then further validated. A significant difference was found in prediction scores between the survival and death groups in the training set (
P=1.2e-14,
Figure 1C). In the test set, we also observed a significant difference in the prediction scores of survival and death groups (
P=0.015,
Figure 1E). The prediction model achieved overall ranking ability of 100% in the training cohort (
Figure 1D) and an overall ranking ability of 83.375% in the validation cohort (
Figure 1F) , calculated by area under the curve (AUC).
2.3 Correlation between immunophenotypic markers’ expression and MM patients’ survival
Univariate Cox regression analysis and hazard ratio (HR) analysis were used to obtain the top 5 immunophenotypic markers associated with survival, including CD38 (
HR=1.019, 95%
CI 1.007 to 1.032;
P=0.002), CD138 (
HR=1.019, 95%
CI 1.007 to 1.032;
P=0.002), CD45 (
HR=1.018, 95%
CI 1.004 to 1.031;
P=0.009), CD56 (
HR=1.019, 95%
CI 1.005 to 1.033;
P=0.007), and CD200 (
HR=1.021, 95%
CI 1.006 to 1.036;
P=0.006;
Table 1). Five immunophenotypic markers obtained via univariate Cox regression analysis were consistent with the results of RSF model. Kaplan-Meier (K-M) curves analysis further demonstrated the effect of different immunophenotypic markers on OS and PFS. In CD38-positive group compared with CD38-negative group, the OS and PFS were both significantly shorter (both
P<0.001, Figure
2A and
2B). In CD138-positive group, the OS and PFS tended to be shorter than those of CD138-negative group (both
P<0.001, Figure
2C and
2D). CD56-positive group had shorter OS (
P<0.001) and PFS (
P=0.005) than CD56-negative group (Figure
2E and
2F). Patients were divided into 2 groups based on the expression levels of CD45. CD45-positive group had shorter OS and PFS than CD45-negative group (both
P<0.001, Figure
3A and
3B). Further, in CD200-positive group, OS and PFS tended to be shorter than those noted for CD200-negative group (both
P=0.001, Figure
3C and
3D). We focused on CD45 and CD200, which are relatively less reported according to previous research. OS (
P<0.001) and PFS (
P=0.001) were significantly different among groups A, B, and C (Figure
3E and
3F). The median follow-up time of Group A, Group B, and Group C was 33 months, 33 months, and 32 months, respectively. Since positive CD45 and CD200 both indicated a poor prognosis, we further combined the 2 factors to observe whether they had an impact on the prognosis. The results indicated that OS and PFS were significantly different between Groups A and B+C (both
P=0.001, Figure
4A and
4B), which further illustrated that patients positive for expression of both CD45 and CD200 had worse prognosis. To assess whether bortezomib could improve the prognosis of patients in Group A and Group B+C, we performed a survival analysis of 91 patients treated with bortezomib. The results showed that there were significant differences in OS and PFS between Group A and Group B+C (both
P=0.015, Figure
4C and
4D). In addition, OS and PFS were not significantly different between patients in Group A receiving bortezomib and those in Group A who did not receive it (both
P>0.05, Figure
4E and
4F). The above results indicated that Group A had the worst prognosis among patients treated with bortezomib, and bortezomib could not improve the long-term prognosis of patients in Group A.
2.4 Univariate and multivariate Cox regression analyses of independent prognostic factors
Univariate Cox regression analyses (
Table 2) revealed that group (
HR=2.876, 95%
CI 1.470 to 5.626;
P=0.002), ISS stage (
HR=2.638, 95%
CI 1.274 to 5.462;
P=0.009), R-ISS stage (
HR=2.616, 95%
CI 1.356 to 5.049;
P=0.004), FISH results (
HR=2.070, 95%
CI 1.078 to 3.975;
P=0.029), levels of creatinine (
HR=2.862, 95%
CI 1.400 to 5.850;
P=0.004), and β2-MG (
HR=2.621, 95%
CI 1.322 to 5.195;
P=0.006), were associated with OS. After adjusting for potential confounders by multivariate Cox regression analysis, Group A (CD45 and CD200 both positive,
HR=2.178, 95%
CI 1.048 to 4.529;
P=0.037) was an independent prognostic factor for OS. Moreover, common factors associated with NDMM PFS were analysed (
Table 3). Group (
HR=2.942, 95%
CI 1.494 to 5.795;
P=0.002), R-ISS stage (
HR=3.199, 95%
CI 1.607 to 6.368;
P=0.001), and FISH results (
HR=2.635, 95%
CI 1.316 to 5.276;
P=0.006), as well as levels of creatinine (
HR=2.660, 95%
CI 1.298 to 5.452;
P=0.008), were associated with PFS. Group A (CD45 and CD200 both positive) (
HR=2.146, 95%
CI 1.027 to 4.485;
P=0.042) was an independent prognostic factor for PFS. Other relevant factors examined are provided in Supplementary Table
1 and
2 (
https://doi.org/10.57760/sciencedb.28274.)
2.5 Construction and validation of prognostic model
Based on the results of multivariate Cox regression analyses, group, ISS stage, R-ISS stage, FISH, creatinine level, and β2-MG level were included in a prognostic nomogram (Figure
5A and
5B). For each factor, the corresponding score can be found in the graph. The 2-, 3-, and 4-year survival probability of patients were predicted based on the total score. To validate the nomogram model internally, 200 bootstrap resamplings were performed. At 2, 3, and 4 years, calibration curves showed that the model was in good agreement (concordance index=0.706; 95%
CI 0.661 to 0.751) (Figure
5C-
5E). From the risk factor map constructed from the multivariate Cox regression analysis results, it can be observed that the number of deaths in high-risk groups was higher than that in low-risk groups. Additionally, high-risk group expressed higher levels of CD45 and CD200 than low-risk group (
Figure 5B). Compared with R-ISS and the Myeloma Prognostic Score System (MPSS), our model showed a higher C-index than either R-ISS or MPSS (
Table 4).
2.6 Subgroup analysis
We conducted subgroup analyses to determine whether Group A remained a prognostic factor in certain subgroups (
Figure 6). The survival curves showed that Group A had worse prognosis in the following subgroups: Albumin<35 g/L (
P<0.001), creatinine< 177 μmol/L (
P<0.001), DS stage Ⅲ (
P=0.003), female (
P=0.040), FISH standard risk (
P=0.008), hemoglobin<100 g/L (
P=0.021), ISS stage Ⅲ (
P=0.040), platelet≥100×10
9/L (
P=0.001), β2-MG<5.5 mg/L (
P=0.003), serum calcium<2.8 mmol/L (
P=0.002), WBC between 4×10
9-10×10
9/L (
P<0.001), and male (
P=0.009).
3 Discussion
In this study, OS and PFS among the 3 groups (CD45-negative and CD200-negative, CD45-positive and CD200-negative, and CD45-positive and CD200-positive) were compared in MM patients. Our results suggested that the group that was both positive for CD45 and CD200 had the worst survival rates.
RSF was used to select the 5 most important immunophenotypic markers (CD38, CD138, CD45, CD56, and CD200) from a variety of immuno-phenotypes. In addition, we performed survival analysis for these 5 immunophenotypic markers, and found that compared with CD38, CD138, CD45, and CD56 negative patients, CD38, CD138, CD45, and CD56 positive patients had poor prognosis. CD38 is a type II transmembrane glycoprotein that regulates migration, receptor-mediated adhesion, and signaling events
[13]. The expression level of CD38 in plasma cells and MM cells is higher than in myeloid cells, lymphoid cells, and some non-hematopoietic tissues, which is one of the reasons why CD38 has become a target for MM targeted therapy. Daratumumab, an antibody targeting CD38, has been approved for clinical treatment
[13]. Most MM cells express syndecan-1 (CD138), which is a cell adhesion molecule that maintains cell morphology and their interaction with the microenvironment. The abnormal expression of CD138 antigen is related to tumor cell proliferation, invasion, and drug resistance
[14]. CD56 is a subtype of neural cell adhesion molecule and a membrane glycoprotein. CD56 is usually expressed by natural killer cells in the nervous system and the immune system, and remains constant throughout the progression of MM, which is significantly associated with the disease progression
[15]. CD38 and CD138 are classic indicators in myeloma. In MM patients, most tumor cells express CD138 and CD38. As mentioned above, many studies have focused on CD38, CD138, CD56, and MM, and even targeted drugs for CD38 have been invented. Based on previous literature and the sample size of each immunophenotype, we selected CD45 and CD200, which are relatively less reported according to previous research, as the immuno-phenotypes for subsequent studies.
The correlation between CD45 expression and CD200 expression and their impact on prognosis for patients with MM remain unclear. Alapat, et al
[16] showed CD200-negative status in MM patients to be more associated with clinically aggressive disease than CD200-positive status. Based on the results of another study, CD200 expression was not significantly correlated with OS or PFS in all patients or newly diagnosed patients
[7]. In addition, results from a study
[6] suggest that being CD200-negative predicts better prognosis in NDMM patients and that being CD200-negative is an independent predictor of OS. Downregulation of CD200 expression during treatment predicts better treatment response and better prognosis during disease
[6, 8]. Similar to CD200, the correlation between CD45 and prognosis is inconclusive
[17-19]. A clinical study
[17] conducted in 2004 noted the proportion of CD45-positive MM patients to be 68.5%, and the prognosis of CD45-positive MM patients was better. However, several recent clinical studies
[19-21] have shown that the proportion of CD45-positive MM patients is 28%, 23%, and 35%, and the prognosis of CD45-positive MM patients is worse. In this study, the proportion of CD45-positive patients was 33.3%, which is concordant with findings of recent studies. Recent studies
[16, 22, 23] on CD200 have shown that the proportion of CD200-positive patients exceeds 70%. In our study, the proportion of CD200-positive patients was 19.5%, which is significantly lower than the proportion of CD200-positive patients in recent studies. This may be because earlier studies used 3- or 4-color flow cytometry to detect CD45, CD38, and CD138, and the lack of gating on CD19 may have resulted in the inclusion of nonclonal plasma cells in the analysis; however, recent studies
[22-23] frequently used 4- or 8-color flow cytometry, in addition to using SSC, CD38, CD138, CD19, and CD45 to jointly set gates, for dividing myeloma cells more accurately. The grouping bias caused by the detection method may be one of the reasons for the conflicting conclusions regarding CD45 and CD200 expression as the prognostic factors in different periods. Furthermore, recent study
[8] on CD200 may be biased in case selection. Since CD200 is the main research object, it may lead to the selection of a greater number of CD200-positive patients for the study, thereby increasing the proportion of CD200-positive patients. In this study, a 4-color flow cytometer was used to accurately classify myeloma cells through joint gating, which avoided the grouping bias of myeloma cells to some extent. In addition, we collected clinical information from all eligible patients to avoid selection bias as much as possible.
However, the mechanisms through which the expression levels of CD45 and CD200 exert a negative impact on prognosis remains unclear. CD45 exhibits structural features, including multiple extracellular immunoglobulin-like domains, which enable it to interact with other cell surface molecules and regulate T cell signaling
[24]. It plays a crucial role in various biological functions, particularly in immune system regulation. Research has demonstrated that CD45 is involved in modulating T-cell receptor (TCR) signaling, influencing T cell activation and proliferation
[25]. Additionally, CD45 facilitates the differentiation process of memory B cells into antibody-secreting cells by enhancing B-cell receptor (BCR) signaling
[26]. Moreover, CD45-positive tumor cells are closely associated with cancer progression and metastasis, suggesting their potential role in immune evasion
[27]. CD200 is widely expressed on tumor cells, while its receptor, CD200R, is predominantly expressed on immune cells, including macrophages and dendritic cells. CD200 plays a critical role in the tumor microenvironment and is regarded as a key factor in tumor immunosuppression. Its immunomodulatory function is mediated by inhibiting anti-tumor immune responses via CD200R, positioning it as a potential target for immune checkpoint inhibition therapy
[3].Therefore, CD45 and CD200 mainly affect the survival of MM patients through immune evasion. In this study, co-expression of CD45 and CD200 resulted in additive effects, with double-positive patients having the worst prognosis.
Our study found that a majority of patients in Group A were classified as R-ISS stage II. This finding carries significant clinical implications. According to the R-ISS staging system, stage II patients typically have a medium-risk prognosis. However, the poor prognosis observed in Group A suggested that additional biological factors or mechanisms, not fully captured by the R-ISS system, may contribute to the adverse outcomes in this group. The R-ISS primarily considers β2-MG, albumin levels, and high-risk genetic abnormalities, while the expression of CD45 and CD200 may reflect the immunoregulatory features of the myeloma microenvironment. Since traditional R-ISS markers do not account for immune-related phenotypes, this may explain why a majority of patients in Group A, despite being in R-ISS stage II, had a poor prognosis. These findings suggest potential limitations of the R-ISS system. Therefore, although these patients are classified as R-ISS stage II, their poor prognosis appears to be closely linked to the expression of CD45 and CD200, suggesting immune evasion mechanism or more aggressive disease features in myeloma.
In our study, bortezomib treatment did not significantly improve OS or PFS in CD45/CD200-positive patients, indicating that bortezomib may have limited efficacy for this patient population. This finding has important clinical implications, suggesting that single-agent therapies, particularly those based on bortezomib, may not be suitable for patients with this specific immunophenotype. We hypothesize that CD45/CD200-positive patients may have a more complex immune microenvironment, where immune evasion mechanisms limit the effectiveness of bortezomib. Besides, researches
[22, 28] have shown that the immune evasion mechanisms associated with CD200 and CD45 in the tumor microenvironment can diminish the therapeutic effectiveness of bortezomib. Therefore, our results align with these findings, further emphasizing the complex role of the immune microenvironment in shaping therapeutic responses. Although bortezomib did not significantly improve OS or PFS in CD45/CD200-positive patients, these results provide valuable insights for future clinical studies. They suggest that treating this subset may require combination immunotherapy or other therapeutic strategies aimed at overcoming immune evasion mechanisms to evasion therapeutic efficacy.
Compared to R-ISS and MPSS, our model achieved a higher C-index. This is the first model to incorporate CD45/CD200 co-expression as an independent prognostic factor for MM. This inclusion may better reflect the tumor microenvironment or immune evasion mechanisms, and addresses the limited focus on immunophenotype in existing systems. However, the implementation of this model requires additional testing for CD45 and CD200 expression, which could increase laboratory costs and workflow complexities, whereas R-ISS and MPSS rely solely on conventional laboratory indicators. Moreover, MPSS is based on a larger sample size, while the smaller sample size of this study may impact the model’s stability and necessitate external validation. The MPSS also incorporates cytogenetic data, such as 1q21 amplification, which is not included in this model. Future work should focus on multi-center external validation, prospective study design, functional experiments, and the development of standardized testing protocols to further validate and apply this model.
Emphasis on the concept of minimal residual disease in recent study
[29] has led to multiparametric flow cytometry immunophenotypic features being highly recommended not only as independent post-diagnostic prognostic factors, but also as markers for guided risk-based treatment strategies, in various malignancies. Individual indicators are not enough to meet the needs of diagnosing diseases and predicting prognosis. Therefore, the use of a combination of multiple indicators is a novel direction that is currently being explored. In this study, we identified relevant immunophenotypes using a random forest algorithm and analyzed their correlation with prognosis in MM. It was determined that the combined expression of CD45 and CD200, as an independent prognostic factor, could predict patient prognosis. In addition, the prognostic model further illustrated the effects of CD45 and CD200 expression on prognosis. Subgroup K-M curves specifically illustrated the correlation between CD45 and CD200 expression and prognosis. We analyzed the association of common phenotypes with prognosis in MM and applied 2 biomarkers with different functions in the same stage of MM. These approaches gave us more opportunities to avoid the occurrence of heterogeneity in our study. In addition, the positive correlation between CD45 expression and CD200 expression may indicate an additive effect on prognosis.
Some limitations exist in this study. Firstly, the sample size was relatively small, with only 123 patients enrolled. Secondly, the CD45-positive and CD200-positive group contained 24 patients, which likely contributed to the lack of statistical power to detect FISH an independent prognostic factor for OS and PFS. Thirdly, although we performed internal validation step, the nomogram model needs further external validation in larger myeloma cohort to test its efficacy. Lastly, most of the patients included were DS stage Ⅲ patients, which might have impacted subsequent analysis. Future studies should aim to increase sample sizes, incorporate prospective clinical trial designs, and potentially combine experimental approaches to generate more robust evidence relevant to the treatment of CD45-positive and CD200-positive MM patients.
Our study suggested that CD38-, CD138-, CD56-, CD45-, and CD200-positive NDMM patients were associated with shorter OS and PFS, and CD45/CD200 positive represented an independent prognosis factor for NDMM patients. The combined use of CD45 and CD200 may provide better prognostic stratification for MM patients compared to the use of these markers alone.
the National Natural Science Foundation, China(81870166)
©Journal of Central South University (Medical Science). All rights reserved.
,
京公网安备11010202008535号
PDF
(3017KB)
