Multiple myeloma accounts for 1% of all cancers and about 10% of all hematological malignant tumors. It is characterized by a monoclonal proliferation of plasma cells and is associated with insufficient production of complete or incomplete immunoglobulins (1, 2). Unlike other malignant tumors that metastasize to bone, there is no new bone formation in osteolytic bone lesions of multiple myeloma (3). In the past 15 years, overall survival (OS) of multiple myeloma has improved significantly with the emergence of thalidomide, bortezomib and lenalidomide (4–7). More recently, drugs such as carfilzomib, pomalidomide and daratumumab have been approved for the treatment of relapsed multiple myeloma, further improving the efficacy (8, 9). With the continuous improvement of treatment level (including autologous hematopoietic stem cell transplantation and the emergence of new drugs), the life expectancy of MM patients is gradually prolonging (10, 11). However, almost all MM patients will relapse eventually, so we urgently need to find valuable prognostic indicators to assess the risk of patients and guide more active treatment, so as to delay the progress of the disease.

The International Staging System (ISS) divides the risk into three grades based on the concentrations of β2-microglobulin and serum albumin; which is related to the prognosis of patients with MM (12). In addition, the deletion of 17p13 (the locus for the tumor-suppressor gene, p53) leads to the loss of heterozygosity of TP53, which is considered to be a high-risk feature in MM (13, 14). Other high-risk chromosomal abnormalities in MM are characterized by structural changes, and several studies have confirmed that patients with t (4;14) and t (14;16) have a poor prognosis (15–17). However, MM patients with the same ISS stage also have different prognosis, and fluorescence in situ hybridization (FISH) is very expensive to evaluate the prognosis of MM patients. Therefore, we need to explore some convenient and easily available prognostic indicators. Inflammatory cells in the blood and systemic inflammatory response (SIR) have a considerable impact on the tumor microenvironment and the progression of malignant diseases, and are related to the prognosis of tumor patients (18, 19). A series of inflammation-related indexes derived from peripheral whole blood cell count (WBCC), including neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), lymphocyte-to-monocyte ratio (LMR) or monocyte-to-lymphocyte ratio (MLR), are easily available and inexpensive, which are considered to be a kind of potential biomarkers.

At present, some studies of WBCC-derived indicators in MM have been reported, but the results of all publications are not consistent. Meta-analysis can overcome the difference of sample size between studies and calculate the best estimated value, which is a powerful statistical tool. The purpose of this study is to quantitatively describe the prognostic value of NLR, PLR and LMR/MLR in MM by meta-analysis.

Multiple myeloma (MM) is a malignancy of plasma cells, which is currently the second most common hematological malignant tumor (1). For newly diagnosed MM patients, the combination of proteasome inhibitors and/or immunomodulatory drugs combined with autologous stem cell transplantation (ASCT) can significantly improve the prognosis of MM patients (10, 11). However, multiple myeloma is still an incurable disease, and the clinical course is so highly variable. Therefore, we need to accurately and effectively evaluate the potential indicators which may be used to predict the prognosis of patients with multiple myeloma.

Multiple myeloma is strongly dependent on the bone marrow microenvironment, which can support the proliferation and survival of myeloma cells and is related to their drug resistance (48). With the deepening of the understanding of the tumor inflammatory microenvironment, we found that inflammation plays an important role in the occurrence, growth and development of tumors (18). Previous studies have shown that inflammatory markers are associated with the prognosis of different tumors, including non-small cell lung cancer, gallbladder cancer, diffuse large B-cell lymphoma, etc. (49–51). At present, we know that some inflammatory cells (including macrophages, dendritic cells, etc.) are involved in the coordination of MM microenvironment (52). Therefore, systemic inflammatory markers (NLR, PLR and LMR/MLR) derived from peripheral whole blood cell count (WBCC) have recently received close attention in MM.

Our meta-analysis showed that elevated NLR, decreased LMR/elevated MLR could predict poor OS/PFS in patients with multiple myeloma, but there was no significant correlation between PLR and prognosis of MM patients. The heterogeneity among the included studies may be partially contributed by study participants, cut-off value, survival analysis method and sample size. Since the number of studies included in the NLR group>10, we conducted a subgroup analysis. The results show that the effect of NLR on PFS may be related to cut-off value and sample size (there may be exist false positive results). However, the effect of NLR on OS was unaffected by the above factors. Although different treatments for MM patients may affect the outcome of OS/PFS in a single study, patients were grouped according to the pretreatment indicators. Therefore, the differences between treatment schemes should not be sufficient to have a significant impact on the results of meta-analysis.

Our findings inferred that NLR has significant prognostic value in MM patients, which is consistent with the studies of Mu et al. and Zeng et al. (53, 54). Our meta-analysis also showed that elevated NLR was closely related to advanced ISS staging. We know that ISS stage is based on serum albumin and β2-microglobulin, which is mainly used by the World Health Organization (WHO) to determine the prognosis of MM patients. High NLR reflects a decrease in the number of lymphocytes and an increase of neutrophils in tumor microenvironment. Absolute neutrophil count might serve as an important marker of systemic inflammation, which provides favorable environment for the occurrence and development of malignant tumors. On the contrary, the absolute count of lymphocytes reflects immunosuppression, which is associated with poor prognosis in a number of solid and hematological malignancies (55). In MM patients, cytokines including IL-6, IL-17, IL-21, IL-22 and IL-23 were detected at a high level (56). These inflammatory factors can combine with accessory cells in milieu, protect tumor cells from immune escape and simultaneously promote the development of tumors (57). These alterations of inflammatory components in the tumor microenvironment could be reflected by WBCC in a certain extent. Therefore, the elevated NLR generates a favorable immune microenvironment, which promotes vascular invasion and host immunosuppression, thereby correlating to poor prognosis of patients (58).

In our meta-analysis, 9 studies involving 2141 patients revealed the prognostic value of LMR/MLR in MM. The results showed that low LMR/high MLR predicted poor prognosis in MM patients. Studies have shown that circulating monocytes can produce a kind of tumor-associated macrophages (TAMs), which constitute an important proportion of tumor-related inflammatory cells (59). It is reported that TAMs are associated with poor prognosis of classical Hodgkin lymphoma (HL), follicular lymphoma (FL) and MM (60, 61). TAMs can be recruited to the tumor site by tumor-derived chemokines, thus affecting the number of peripheral blood monocytes. Therefore, the absolute monocyte count (AMC) in peripheral blood may reflect the recruitment degree of TAM to some extent (62). On the other hand, monocyte lineage cells are essential for innate immune response. A number of genes expressed by peripheral blood monocytes, whose products take part in immunologic response-2, and their expression levels are related to cancer prognosis. In addition, monocytes secrete TNF α and IL-1, thus monocytes play an important role in tumor microenvironment and could be used as markers of tumor load (63).

Our study showed that there was no significant correlation between pre-treatment PLR level and prognosis in MM patients. Solmaz et al. and Shi et al. both suggested that lower PLR is associated with poor survival in MM patients (36, 38). On the contrary, Szudy-Szczyrek et al. demonstrated that elevated PLR results in shorter OS (30). However, other studies suggested that PLR has not yet shown prognostic value in MM patients (41, 44). Current studies have confirmed that platelets can interact with platelet-derived growth factor and platelet factor to stimulate cell proliferation and metastasis through complex mechanisms of hemostasis activation as well as cell signal transduction, thus promoting tumor progression (64, 65). PLR combined with platelet and lymphocyte count index can more accurately reflect the pro-tumor efficacy and anti-tumor capacity of the host (66). The increase of PLR suggests that the balance of inflammation and anti-inflammation in the tumor may be disrupted, and the inflammatory response promotes tumor formation and is associated with a poor prognosis of some cancers (49, 67–69). This difference in multiple myeloma may be due to the pathology of the disease. Some studies have found that patients with MM show evidence of platelet activation, and the activated platelets secrete many kinds of cytokines that are required for the growth of myeloma cells, including IL-6, VEGF, SDF-1α and IGF-1, suggesting that platelets may affect the microenvironment (70). On the other hand, with the diffusion of malignant plasma cells in the bone marrow of MM patients, its aggregation can inhibit normal thrombopoiesis, and M protein produced in MM may adhere to the platelet surface, resulting in the decrease in the platelet activation and function (71). Based on our meta-analysis, we think that PLR could not predict the prognosis of MM patients at present. Though the treatment regimens of MM patients in the included studies are not consistent, which may lead to differences in the results between studies; in addition, the insufficient number of studies may lead to unstable results.

This meta-analysis still has some limitations that need to be carefully considered. Firstly, there is a significant heterogeneity between the studies included in NLR and PLR group. Despite the utility of sensitivity analysis and meta regression, the origin of heterogeneity could not be fully traced. Secondly, most of the studies are retrospective studies with low quality of evidence compared with other types of studies. Thirdly, the cut-off value of each study is not uniform, if each index can have a unified cut-off value standard, we may be able to better compare the differences between the studies. Finally, only few studies investigated PLR and LMR/MLR, so the statistical data of meta-analysis of the latter two indicators are under-powered. It is hoped that more with reasonably designed, high-quality, multicenter studies can be reported in the future to further enrich our results.

In summary, our meta-analysis validated that NLR and LMR/MLR derived from WBCC could be used as prognostic biomarkers to predict the survival outcome of MM patients, but the prognostic value of PLR is open to question. This series of indicators are inexpensive and easy to obtain, which help to assess the risk of MM patients and guide more active treatment, thus delaying the progress of the disease.

The data analyzed in this study is subject to the following licenses/restrictions: The original contributions presented in the study are included in the article/supplementary material. Further inquiries can be directed to the corresponding authors. Requests to access these datasets should be directed to CH, huangchunlan@swmu.edu.cn.

XZ, JD, and ZW designed the project. XZ, JD, and ZW performed data extraction and analysis. Quality assessment was conducted by HX, XC, YL, and KL. XZ and CH made critical changes to the manuscript and oversaw the project. All the authors read and approved the final manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.