Clinicopathological and prognostic significance of platelet-lymphocyte ratio (PLR) in gastric cancer: an updated meta-analysis

Background Pre-treatment PLR (platelet-lymphocyte ratio) was reported to be associated with the prognosis in gastric cancer (GC), but the results remain inconclusive. This meta-analysis aimed to investigate the prognostic potential of the pre-treatment PLR in gastric cancer. Methods We performed a systematic literature search in PubMed, Embase, and the Cochrane Library to identify eligible publications. The hazard ratio (HR)/odds ratio (OR) and its 95% confidence (CI) of survival outcomes and clinicopathological parameters were calculated. Results A total of 49 studies (51 cohorts), collecting data from 28,929 GC patients, were included in the final analysis. The pooled results demonstrated that the elevated pre-treatment PLR was significantly associated with poor overall survival (OS) (HR 1.37, 95% CI 1.26–1.49, p < 0.001; I2 = 79.90%, Ph < 0.001) and disease-free survival (DFS) (HR 1.52, 95% CI 1.22–1.90, p < 0.001, I2 = 88.6%, Ph < 0.001). Furthermore, the patients with the elevated PLR had a higher risk of lymph node metastasis (OR = 1.17, 95% CI 1.02–1.33, p = 0.023), serosal invasion (T3+T4) (OR = 1.34, 95% CI 1.10–1.64, p = 0.003), and increased advanced stage (III+IV) (OR = 1.20, 95% CI 1.06–1.37, p = 0.004). Conclusions An elevated pre-treatment PLR was a prognostic factor for poor OS and DFS and associated with poor clinicopathological parameters in GC patients.


Background
Gastric cancer (GC) is a kind of common malignant tumor and one of the main causes of cancer-related mortality and morbidity worldwide [1]. Majority of the patients are diagnosed at an advanced stage due to no symptoms in the early stage. Complete or partial resection is the only potential curative treatment. However, the high recurrence and metastasis after resection lead to the poor level of 5-year survival rate [2]. For individual patients with different disease status and physical conditions who should receive individualized therapeutic regimens, it is essential to identify different risk groups according to various biomarkers. Therefore, potential biomarkers are required and crucial for predicting the patient prognosis and designing therapeutic regimen and follow-up scheme.
The systemic inflammatory response (SIR), being associated with the outcome of a variety of tumor-related inflammation, is considered an important component of tumor progression [3]. Immune and inflammatory cells in peripheral blood, such as neutrophils, lymphocytes, platelets, and monocytes, play important roles in the tumor micro-environment and relate to invasion and metastasis of tumor cells [4]. Some indexes of the SIRrelated cells, such as neutrophil to lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), and monocyte to lymphocyte ratio (MLR), have been used to predict survival and recurrence of various cancers, including gastric cancer [5][6][7][8]. Among the indexes, PLR is considered a potential marker of endogenous residual anticancer pre-inflammatory and pre-coagulative response that arises in malignancies and is highly repeatable, cost-effective, and widely available [9,10]. The application of PLR in the diagnosis and prognosis of gastric cancer was also reported in a variety of studies but with controversial results. For example, Kim et al. found that elevated PLR predicted poor overall survival (OS) and disease-free survival (DFS) in GC patients after surgery [11]. However, some other studies did not detect the significant prognostic value of PLR for GC patients [12,13]. We conducted this meta-analysis to investigate the prognostic significance of pre-treatment PLR for OS and DFS, and the associations between PLR and clinicopathological features in GC patients.

Literature search
We performed a systematic literature search in PubMed, Embase, and the Cochrane Library. The search strategy terms are as follows: (PLR or "platelet lymphocyte ratio" or "platelet-to-lymphocyte ratio" or "platelet-lymphocyte ratio") and ("gastric cancer" or "gastric adenocarcinoma" or "gastric carcinoma" or "GC" or "gastric neoplasm" or "stomach tumor" or "stomach neoplasm"). The last search was updated to April 8, 2020, and studies published in English were included. This study was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement, and a flow chart of the systematic review is shown in Fig. 1. No ethical approval and patient consent are required in this study.

Inclusion and exclusion criteria
The predetermined inclusion and exclusion criteria were applied for the including of the articles in this study: Inclusion criteria are as follows: (1) the diagnosis must be confirmed by pathological examination; (2) HR and 95% CI for the OS and (or) DFS, the number of patients with various clinicopathological features are available; (3) PLR is the result of pre-treatment. Exclusion criteria are as follows: (1) conference abstracts, reviews, letters to the editor, and other nonclinical literature are not applied; (2) articles with insufficient data to estimate are not included; and (3) the articles with non-human research or non-English language are not included.

Data extraction and quality assessment
All studies were assessed independently by two authors according to the designed eligibility criteria. Any questions or disagreements were resolved by consulting another co-author. The extracted data included the following study information: first author, publication year, country, study design (retrospective or prospective), study period, treatment regimens, follow-up time, cutoff value of PLR, and the number of patients with various clinicopathological features, including tumor location, differentiation, size, depths of tumor invasion, lymph node metastasis, TNM stage, and HRs with 95% CIs of OS and DFS. The quality of each study was assessed according to the Newcastle-Ottawa Scale (NOS) by two authors [14] and a NOS score ≥ 6 were considered high-quality researches.

Statistical analysis
The pooled HRs were calculated based on HRs and their 95% CIs from each study to estimate prognostic role of PLR in GC patients. HRs and 95% CIs for OS and DFS were obtained directly from each study if available or were calculated from the related data according to the methods published by Tierney et al. [15]. Cochran's Q test and Higgins I-squared statistic were used to evaluate the heterogeneity of pooled results. A p value < 0.1 for the Q test or Ι 2 > 50% indicate significant heterogeneity among studies, and the random-effects model (DerSimonian-Laird method) was performed to calculate the pooled HRs. Otherwise, the fixed-effects model (Mantel-Haenszel method) was applied [16]. Odds ratios (OR) and 95% CI were used to analyze the relationship between PLR and clinicopathological factors. Publication bias of the literature was evaluated by Begg's funnel plot and Egger's linear regression tests, and p > 0.05 indicated that there was no significant publication bias. Sensitivity analysis was performed by removing each single study in turn to validate the stability of the pooled results. All statistical analyses were performed using STATA software version 14.0 (STATA Corporation, College Station, TX, USA). Results with p < 0.05 were considered statistically significant, and all the results were two sided.

Study characteristics
A total of 49 studies (51 cohorts) [7,[11][12][13] with 28,929 GC patients were included in the final metaanalysis. As in Fan Feng's study [37], the GC patients were included in a training set and a validation set independently; therefore, the two cohorts were extracted separately and named as Fan Feng(1) and Fan Feng (2). As in Aldemir's study, GC patients were divided into the surgery group and chemotherapy group. So we named the two groups as Aldemir(1) and Aldemir(2) [60]. The selection process of the included studies according to the PRISMA guidelines was shown in Fig. 1. We summarized the characteristics of the studies in Table 1. Among them, 10 studies were from Europe and the USA and 41 studies from Asia. The patients from 27 studies received surgery treatment, the patients with an advanced stage from 8 studies received chemotherapy strategy, and the patients from other 6 studied received mixed treatment (including chemotherapy, surgery, radiotherapy, targeted therapy, and supportive care). The cut-off values of PLR among the studies varied from 10.1 to 350. Therefore, we selected PLR = 150 to divide the studies in subgroup analysis. All studies with NOS scores ≥ 6 were regarded as high-quality studies.

PLR and clinicopathological parameters of GC
To further explore the impact of PLR on the clinicopathological parameters in GC, we extracted the number of patients from parts of studies in PLR-high and PLR-low groups according to the TNM stage, tumor differentiation, depth of invasion, tumor size, tumor location, and lymph node metastasis. As shown in Table 3, in comparison to low PLR groups, the high PLR groups had a higher risk of lymph node metastasis (n = 15, OR = 1.17, 95% CI 1.02-1.33, p = 0.023), serosal invasion (T3+T4) (n = 13, OR =

Sensitivity analysis
We performed sensitivity analysis for the OS by removing one single study at a time to check if individual study influenced the results. The corresponding pooled HRs are consistent, indicating stable and robust results in this meta-analysis (Fig. 4).

Publication bias
Begg's funnel plot and the Egger's linear regression test were performed to assess publication bias. The figure of the Begg's funnel plot showed obvious asymmetry (Fig.  5) and Egger's tests (p = 0.004) indicated significant publication bias. However, our finding that elevated PLR is associated with lower OS did not change after the adjustment for publication bias using the trim and fill method [62].

Discussion
The current meta-analysis was designed to investigate the prognostic value of elevated PLR for DFS and OS in GC patients. Pooled results demonstrated that elevated PLR was associated with poor OS and DFS. Moreover, elevated PLR was correlated with lymph node metastasis, serosal invasion, and advanced TNM stage with GC. Despite the development of new surgical techniques and the use of chemotherapy and radiotherapy, gastric cancer still remains one of the main causes of cancerrelated mortality and morbidity worldwide [63]. Because individual GC patients present with different conditions, including different degrees of invasion, differentiation, and TNM stages, the survival outcomes may vary. Therefore, identification of reliable prognostic factors, simple and low cost, to stratify patients into different risk groups, would contribute to the optimization of  [64][65][66]. The inflammation is involved in lymphocytopenia, neutrophilia, thrombocytosis, and leukocytosis [67,68]. The tumor-generated inflammatory reaction may contribute to tumor growth, progression, and metastasis through several mechanisms, including the upregulation of inflammatory mediators and cytokine, aberrant activation of immune regulatory cytokines, suppression of apoptosis, and DNA damage [65]. Recently, emerging evidence indicates that inflammatory reaction is an important factor for the initiation, progression, and prognosis of numerous cancers, including GC [69,70]. Helicobacter pylori infection in GC is characterized by an inflammatory infiltrate, consisting mainly of neutrophils and T cells [71]. Moreover, circulating lymphocytes were reported that could reflect patient's inflammatory status [72]. Thus, some inflammation-based parameters, such as lymphocyte count, systemic immuneinflammation index (SII), platelet-lymphocyte ratio (PLR), and neutrophil-lymphocyte ratio (NLR), have been used to predict survival and recurrence in cancer patients [44,[73][74][75][76].
The PLR, which combines platelet and lymphocyte counts, is a representative index of systemic inflammation and immune status [77,78]. Accumulating evidence indicates the correlation of PLR with different stages of tumor development, chemotherapeutic response, and prognostic survival outcomes of GC patients [38,42,78]. The specific mechanisms involved are complex and remain unclear. One potential explanation is that a decreased PLR may reflect tumor disadvantage status, such as inflammatory status, immune disorders, malnutrition, and a tendency for micro-vessel thrombosis [39,79]. Lymphocytes have an important role in cancer immune surveillance and preventing the development of malignancy [80]. A Table 3 Meta-analysis of the association between PLR and clinicopathological parameters of GC  pro-inflammatory status leads to compromised cellmediated immunity and impaired T-lymphocytic response via cytokines [81]. The decrease in CD4+ Thelper lymphocytes may result in a suboptimal lymphocyte-mediated immune response to tumor cells [82]. The T-lymphocytic cell-mediated malnutrition is a major cause of delayed wound healing [83,84].
Platelet count is an additional index of a systemic inflammatory response and potential micro-vessel thrombosis, which could inhibit wound healing via the deterioration of blood circulation in tissues [11,77,85]. Otherwise, aggregated platelets can promote tumor growth via releasing pro-angiogenic mediators within the micro-vasculature of tumors [86]. Platelets also  inhibit tumor cell extravasation by potentiating tumor cell-induced endothelial cell retraction, and enhance tumor cell adhesion and spreading across the extracellular matrix, which contribute to the promotion of tumor cell proliferation and metastasis [87]. Therefore, lymphocytopenia and thrombocytosis are considered negative prognostic markers in various cancers [88][89][90][91]. However, a decreased lymphocyte count or an increased platelet count alone may not reflect the host systemic inflammatory response and tumorigenesis process. Thus, the PLR, a biomarker combining platelet and lymphocyte counts, may better reflect the information of lymphocytopenia and thrombocytosis and predict the prognosis of GC patients. In addition, the value of PLR could be acquired from the routine laboratory tests, which provides clinical implications at a low cost. Accumulated studies have assessed the association between PLR and the diagnosis and prognosis of gastric cancer. Some studies showed that elevated PLR predicted poor OS and DFS in GC patients after surgery [22,24]. However, some other studies did not detect the significant prognostic value of PLR for GC patients [7,47]. Lian et al. reported that low PLR levels correlated with better clinicopathological features, including decreased depth of invasion, less lymph node metastasis, and early tumor stage [44]. Recently, a meta-analysis containing 8 studies comprising 4513 patients was conducted and showed that PLR was not a reliable predictor for OS in patients with GC, while another meta-analysis including 13 studies with 6280 patients indicated that elevated PLR could be a significant prognostic biomarker for poor OS [92,93]. Thus, the prognostic value of the PLR remains inconclusive in gastric cancer. So we conducted this updated meta-analysis to evaluate the prognostic role of the PLR in gastric cancer.
In the current study, including 49 studies (51 cohorts) with 28,929 GC patients, we not only investigated the prognostic value of PLR for OS and DFS, but also explored the associations between PLR and clinicopathological characteristics of GC. This analysis demonstrated that elevated PLR leads to a higher risk of lymph node metastasis, increased serosal invasion (T3+T4) risk, and advanced stage (III+IV) in patients with gastric cancer. Although the specific mechanism is still incompletely understood, our results are in accordance with other studies in various cancers, such as pancreatic ductal adenocarcinoma, hepatocellular carcinoma, and colorectal cancer [94][95][96][97][98].
Previous meta-analysis did not find significant association between PLR and OS or DFS in GC, maybe because of the limited studies included [92,93]. Our metaanalysis including much more studies suggested that elevated PLR might have powerful prognostic efficiency for poor OS in GC and could predict shorter DFS in GC. Subgroup analyses for OS revealed the similar result in Asian populations, but not in Caucasian populations. Moreover, we also eliminated the effect of different treatment methods on the prognostic value of the PLR. Our results showed that elevated PLR significantly predicted shorter OS in patients receiving surgery treatment, but did not have prognostic efficiency for patients receiving chemotherapy or mixed treatment. Except for the reason of too few studies included, another possible major reason is that the patients in the chemotherapy or mixed groups have huge differences in medical conditions and disease status, resulting in the inability to obtain significant results. To evaluate the effect of different cut-off values on the prognostic value of PLR in GC patients, subgroup analyses showed that patients with elevated PLR suffered worse OS than those with low PLR, regardless of the different cut-off values. The same effects were indicated in the subgroup analyses by different sample size of patients. These results might strengthen the possibility that PLR could act as a reliable prognostic biomarker in GC.
There were some limitations requiring to be addressed in this meta-analysis. First, the inclusion criteria of this meta-analysis were constrained to studies published in the English language only. So publication bias cannot be excluded. Second, almost the studies included were all retrospective, which could be more susceptible to some biases. Fortunately, the asymmetry in the funnel plots showed no significant publication bias, thus maintaining the substantial consistency of the results. Third, the different cut-off values of PLR used in each study could contribute to the heterogeneity. Subgroup analysis was conducted based on the different PLR cut-off values, while the results were not substantially changed. Further well-designed studies, especially randomized controlled trials (RCTs), are needed to determine the most appropriate cut-off value of PLR to predict the complication risks and survival outcomes in patients with GC.

Conclusions
In conclusion, elevated pre-treatment PLR is a prognostic factor for poor OS and DFS in GC patients. Furthermore, elevated PLR is correlated with a higher risk of serosal invasion, lymph node metastasis, and advanced TNM stage (III+IV) in gastric cancer. The present study suggests that the PLR could provide reliable information before treatment for patients with gastric cancer.