Globally, pulmonary malignant tumors is the major source of cancer-derived deaths for the last several decades(1). With the advent of IT and TT, a large proportion of patients with advanced lung cancer also have effective treatment strategies and have reaped good outcomes(38). However, the treatment options for patients who have been clinically or histologically diagnosed with BM from lung cancer remain inconclusive and still need to be determined based on the patient's prognosis estimate(39). This often results in delayed treatment or waste of medical resources, therefore, the early use of simple hematological indicators to determine patient prognosis is of great significance in clinical work. Our meta-analysis confirms that NLR and PLR before treatment were statistically associated with OS in lung cancer patients with BM.
NLR and PLR are very common hematological indicators in laboratory tests. Many studies have confirmed the association among these indicators and the outcome of solid tumors(40). According to a retrospective study conducted by Liu et al. High NLR and PLR signified shorter PFS (For NLR: HR = 2.17, 95%CI 1.01-4.55, P = 0.048; For PLR: HR = 2.56, 95%CI 1.06‐5.88, P = 0.025) and shorter OS (For NLR: HR = 5.00, 95%CI 1.61‐16.67, P = 0.002; For PLR: HR = 5.00, 95%CI 1.37‐16.67, P = 0.008) in advanced NSCLC, suggesting the adverse prognosis of high NLR and PLR in cancer patients(41). For lung cancer that has progressed to the appearance of BM, the study by Cho et al. also demonstrated that OS was shorter in the NLR-increased group than NLR-not-increased group (HR: 1.817; 95%CI = 1.301–2.539, p < 0.001), and the same results were found in the PLR group (HR: 1.654; 95%CI = 1.178–2.322; p = 0.004)(23). This also confirms that in patients with BM, these hematological indicators have similar significance.
The microenvironment of brain tumors has a distinctive cellular composition, histological and anatomical structure, immune environment, and various metabolic constraints that differ significantly from extracranial lesions, exerting unique and profound selective pressures on tumor cells and influencing the process of metastatic and therapeutic response(42). The brain parenchyma contains cell types not found in extracranial organs, such as neurons, astrocytes, microglia, and oligodendrocytes. BBB and BCSFB isolate the brain into a relatively independent environment by sealing the tight connection between brain capillary endothelial cells or specific epithelial cells and meningeal cells, protecting the brain from various pathogenic factors(43).
But when brain metastasis occurs, the situation is different from normal. The brain has long been considered an organ with "immune privilege", meaning that surrounding immune cells are restricted by BBB from entering the brain in large numbers, making the immune environment of the brain very different from that of the external parts of the brain. However, recent research suggests that that the brain is not a complete immune sanctuary and that immune cells from the periphery can enter the brain in certain ways, and this is particularly evident in patients who have developed BM(44). During inflammation, immune cells in the peripheral blood such as leukocytes can make the BBB more permeable by secreting IL-1β or by the adhesion proteins ICAM1, VCAM1 or E-selectin(45). Moreover, when BM occur, the tumor compromises the structural integrity of the BBB, leading to the creation of a BTB characterized by inhomogeneous permeability and active efflux of molecules(46). From this, we can reasonably speculate that the indicators of peripheral blood circulation in patients with BM can also reflect intracranial lesions to some extent.
We can explain their relationship with patient prognosis by analyzing the composition of NLR and PLR. Neutrophils, platelets, and lymphocytes can represent three physiological or pathological response systems of the body: acute inflammatory response, coagulation response, and acquired immune response, and the tremendous impact of these various cells on the development of tumors is also widely documented. Early in the tumor process, these cells generate an attractive environment for tumor growth, promote genomic instability and foster angiogenesis(47). Neutrophils regulate inflammation through the production of reactive intermediates such as ROS and RNS. They also promote angiogenesis, progression and invasion of tumor by releasing NE and matrix metalloproteinases MMP8/9, which reconfigure the extracellular matrix in the tumor microenvironment. These proteases degrade pro-inflammatory cytokines and reposition the microenvironment of tumor, increasing the oncogenic potential of tumor cells in vivo and vitro, as well as the metastasis initiation potential of cancer cells to enhance tumor progression and aid metastasis(48, 49). Platelets are important for hematogenous metastatic dissemination. Platelets provide large amounts of secreted proteins and alpha-particle contents to the adjacent areas, all of which contribute to the initiation and acceleration of the host inflammatory response on the one hand, and on the other hand, its activation creates a tumor-friendly microenvironment that protects tumor cells from shear and NK cell attack, prompting platelet embolization of tumor cells to lodge in the vessel wall. Then, by producing growth factors, tumor cells acquire a mesenchymal-like phenotype and expand the gap between capillary endothelial cells, accelerating extravasation to other organs(50). Lymphocytes can exert extrinsic tumor suppressive effects through cancer immunosurveillance function. The role that immunity plays in the complex interaction between tumor and host has been termed "cancer immune editing(51)”. There is growing evidence that finding TILs in a cancer patient's tumor often suggests a better outcome for that patient. Several seminal studies involving patients with primary or metastatic melanoma(52), squamous cell carcinoma or adenocarcinoma of the esophagus(53), and patients with advanced ovarian adenocarcinoma(54) have established strong relationships between the presence of TILs and patient outcome, and has analyzed the prognostic value of various T-cell subsets infiltrating in the lesions.
This meta-analysis involved 11 publications including 1977 patients with lung cancer who developed BM, with no language restrictions. According to the findings of this study, OS was shorter in both the high NLR and high PLR groups, that implies pre-treatment NLR and PLR may be a very promising prognostic predictor for patients with BM from lung cancer. However, according to the subgroup analysis, no statistically significant relationships were found between NLR and OS in BM patients whose primary cancer type was SCLC in the NLR group. The analysis also suggests that the heterogeneity might originate from East Asia, sample size larger than 150, cut-off value less than 3, primary lung cancer type of SCLC and studies with a single-center experimental design. In the PLR group, the association between PLR and OS was also not statistically significant when the primary cancer type was BM with SCLC and the cutoff point was less than 200 when the study type was single-center. The results of the analysis suggest that heterogeneity may arise from studies with cut-off value 200 or less, with primary lung cancer type of SCLC, and with a single-center experimental design. By further analysis, we found that in all studies that included only SCLC patients, clinicians did not apply immunotherapy to treat patients. However, recent research has shown that the prognostic significance of NLR and PLR for NSCLC needs to be predicated on the treatment of immune checkpoint inhibitors, and high level of NLR and PLR suggest a poor prognosis for such patients(55). Therefore, no statistically significant association was found between PLR, NLR and OS in SCLC patients in the subgroup analysis. Gu et al. found that PLR was significantly associated with PFS and OS at a cutoff value of 180 in patients with Caucasian NSCLC; however, PLR values above 200 were associated with lower prognostic values in Asians(56). While the included literature that analyzed the prognostic value of PLR were from Asia. Therefore, in the subgroup analysis when the PLR as of value is less than 200, there is no statistically significant association between PLR and OS. This suggests that if PLR is used to assess the prognosis of Asian lung cancer patients with BM, then a cut-off value higher than 200 is a better choice. Of course, this should be validated by prospective research with a larger sample capacity. Finally, the limitations of single-center studies also affect the significance in statistical terms of the relationship between PLR and patient OS.
Our study also has other limitations. Firstly, all included studies were retrospective, resulting in less credibility of the evidence than in clinical randomized controlled trials. Secondly, some of the included literature did not mention a clear follow-up time, which may have resulted in bias. Thirdly, each study used a different treatment strategy, which increased the heterogeneity of the included studies. Finally, the cut-off points of NLR and PLR were different in every study, which increased the heterogeneity among studies and made the clinical application more difficult, follow-up studies should establish a better generalizable cut-off value system to guide research and clinical practice.