Prognostic and clinicopathological role of pretreatment systemic immune-inflammation index in patients with oral squamous cell carcinoma: a meta-analysis

Background There are many studies regarding the use of systemic immune-inflammation index (SII) to help predict oral squamous cell carcinoma (OSCC) prognosis, but findings have been inconsistent. The present meta-analysis was conducted to determine whether SII could contribute to predicting OSCC prognosis. Methods PubMed, Embase, Cochrane Library and Web of Science databases were thoroughly searched from their inceptions through August 20, 2023. The role of SII in predicting OSCC prognosis was determined through combined hazard ratios (HRs) with relevant 95% confidence intervals (CIs). Correlations of SII with clinicopathological characteristics of OSCC patients were analyzed based on combined odds ratios (ORs) with 95% CIs. Results This meta-analysis utilized 11 articles in total, involving 3,464 patients. According to the results, an elevated SII was markedly associated with dismal overall survival (OS) (HR=1.85, 95%CI=1.48-2.29, p<0.001) and poor disease-free survival (DFS) (HR=1.77, 95%CI=1.20-2.61, p=0.004) of OSCC. Moreover, a higher SII was markedly correlated with stage T3-T4 (OR=2.47, 95%CI=1.40-4.37, p=0.002), TNM stage III-IV (OR=2.29, 95%CI=1.53-3.44, p<0.001), and low differentiation (OR=1.74, 95%CI=1.25-2.43, p=0.001). Conclusion According to the present meta-analysis, an increased SII is significantly associated with dismal OS and DFS, advanced tumor stage and poor differentiation in OSCC. SII could be a potential and important biomarker for clinical management and predicting the prognosis of patients with OSCC. Systematic review registration https://inplasy.com/inplasy-2023-9-0033/), identifier INPLASY202390033.


Introduction
Head and neck cancer (HNC) is the sixth most common cancer across the world, affecting nearly 650,000 patients and contributing to 350,000 deaths every year (1,2).Oral squamous cell carcinoma (OSCC), has the highest morbidity in HNC and constitutes 48% of all HNC cases (3).Moreover, OSCC includes cancers that occur in the lips, gums, tongue, mouth, and palate (4).Although there have been improvements in multidisciplinary collaboration and comprehensive therapy, such as surgery, radiotherapy, and chemotherapy, OSCC has had a low 5-year survival rate (under 50%) over the past two decades (5).Nowadays, the tumor-nodemetastasis (TNM) classification system is widely used to guide the selection of treatment strategies and predict survival outcomes; however, patients of an identical TNM stage can have diverse disease courses (6).Therefore, identifying reliable and costeffective prognostic markers for OSCC patients is urgently needed to intervene treatment measures and improve overall prognosis.

Literature retrieval
Literature was retrieved from the PubMed, Embase, Cochrane Library and Web of Science databases, starting with the earliest possible date through August 20, 2023.The following terms were used to search and select literature for the meta-analysis: (systemic immune-inflammatory index or SII or systemic immuneinflammation index or systemic-immune-inflammation index) and (oral squamous cell carcinoma or OSCC or oral cancer or tongue cancer or mouth cancer or oral carcinoma or oral cavity cancer or lip cancer or gingiva cancer).More details about these search strategies are provided in Supplementary File 1.Only English publications were considered.Besides, references in each publication were manually retrieved to identify the possible relevant articles.

Study eligibility criteria
Included studies had the following features (1): pathological diagnosis of primary OSCC (2); explored a relationship between pre-treatment SII and OSCC prognosis (3); hazard ratios (HRs) with 95% confidence intervals (CIs) can be determined according to the available data (4); the threshold SII is identified; and (5) articles written in the English language.Exclusion criteria were as follows (1): meeting abstracts, reviews, letters, comments, and case reports (2); does not have sufficient or available data (3); contains overlapped patients; and (4) animal studies.

Data collection and quality evaluation
Qualified publications were evaluated by two independent reviewers (JZ, SD), who also extracted data.Any discrepancy was settled through negotiation until a consensus was reached.Data collected included, first author, publication year, study country/ region, sample size, age, gender, study center, study design, study period, tumor subsite, TNM stage, treatment, threshold, threshold determination approach, survival outcomes, survival analysis type, follow-up, HRs and 95% CIs.Our primary and secondary outcomes were overall survival (OS) and disease-free survival (DFS), separately.We employed the Newcastle-Ottawa Scale (NOS) for assessing study quality (30).The NOS contains three perspectives, selection (0-4 points), comparability (0-2 points), and outcome assessment (0-3 points), with a total score of 0-9 points.NOS scores ≥ 6 indicate high-quality.

Statistical analysis
Significance of SII in predicting OSCC prognosis was estimated based on combined HRs with 95% CIs.Additionally, I 2 statistics and Cochrane's Q test were utilized to evaluate inter-study heterogeneity.The random-effects model was utilized in the case of obvious heterogeneity (I 2 >50%, P<0.10), otherwise, a fixed-effects model was applied.The source of heterogeneity was detected by different factors-stratified subgroup analyses.Correlations of SII with clinicopathological characteristics of OSCC were evaluated through combined odds ratios (ORs) as well as 95% CIs.Sensitivity analysis was used to compare pooled effects, by eliminating one individual study in the sequence and observing any potential changes to the result, repeating the process for each study.We performed Egger's and Begg's tests for assessing publication bias, and conducted statistical analyses using Stata version 12.0 (Stata Corporation, College Station, TX, USA).P-values < 0.05 were defined as statistically significant differences.

Study screening
There were 117 articles obtained initially, among which 69 were retained following the removal of duplicates (Figure 1).Through titleand abstract-selection, 51 articles were then excluded due to irrelevance.Full-text review of the remaining 18 articles was conducted, among which, seven were eliminated for the following reasons, not focused on OSCC (n=3), no survival data provided (n=2), no cut-off value (n=1), and no report on SII (n=1).Ultimately, 11 articles were utilized for the remainder of the analysis, involving a total of 3,464 patients (18-28) (Figure 1, Table 1).

SII and OS of OSCC
Nine articles, involving 3,044 patients (18-23, 25-27), mentioned a significance of SII to predict OS in OSCC.Due to significant heterogeneity (I 2 = 50.2%,p=0.041), we selected the random-effects model.According to Figure 2 and Table 2, HR=1.85,95%CI=1.48-2.29,and p<0.001, which indicates that a higher SII was markedly related to the dismal OS of OSCC patients.According to subgroup analyses, sample size, study center, TNM stage, threshold, threshold determination method, and survival analysis type did not affect the significant role of SII to predict OS (p<0.05;Table 2).Moreover, higher SII still significantly predicted poor OS in the following subgroups: in Asian regions (p<0.001),tongue tumor site (p=0.004)or OSCC (p<0.001), and patients who received surgery (p<0.001) or RT (p=0.001) (Table 2).Forest plots on prognostic value of SII for overall survival in patients with OSCC.

Sensitivity analysis
Every article was removed individually during each sensitivity analysis.Results were recalculated each time, based on the remaining studies' OS and DFS.According to Figure 6, in the overall analysis of OS and DFS, there was no significant difference after eliminating each work, suggesting the reliability of our combined results.Forest plots on prognostic value of SII for disease-free survival in patients with OSCC.

Publication bias
Begg's funnel plots and the Egger's test were conducted to assess possible publication bias.The funnel plots observed in Figure 7 show symmetry, suggesting no significant publication bias for OS (p=0.175 and p=0.082 upon Begg's and Egger's tests, separately) or DFS (p=1 and p=0.542 upon Begg's and Egger's tests, separately).

Discussion
Previously, the effect of SII to predict OSCC prognosis has been explored, but no consistent findings have been reported (18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28).This work combined results from 11 articles involving 3,464 patients.According to our results, an elevated SII was remarkably related to dismal OS and inferior DFS of OSCC.Moreover, SII had a stable role when predicting prognosis, as examined by sensitivity, subgroup, and publication basis analyses.Higher SII was also evidently related to T3-T4, TNM III-IV, and poor tumor differentiation.Taken together, a higher SII significantly predicted the short-and long-term survival of OSCC, which was also dramatically related to tumor metastasis and poor differentiation.To our knowledge, this is the first meta-analysis investigating whether SII could be used to predict OSCC prognosis.
To understand the biological mechanism behind SII's prognostic value, it is necessary to understand the function of neutrophils, platelets, and lymphocytes.First, neutrophils release inflammatory mediators such as neutrophil elastase, interleukin-8 (IL-8) and matrix metalloproteinase-9 (MMP-9)  which enhance tumor cell growth, migration and invasion (31).
Increased neutrophil counts can also produce reactive oxygen species, nitric oxide, and arginase, resulting in disordered T cell activation (32).Consequently, the body loses its ability to target tumor cells, indirectly contributing to tumor progression (33).Second, platelets can protect cancer cells from natural killer cells and tumor necrosis factor-a (TNF-a) by using glycoprotein (GP) receptors and tumor cell integrin a vb-dependent pathway (34).Platelets also induce epithelialmesenchymal transition and support transendothelial migration in circulating tumor cells, ultimately protecting tumor cells from immune destruction and promoting distant metastasis (35, 36).Third, lymphocytes are responsible for the adaptive immune response and participate in cancer immunosurveillance and immunoediting.Tumor-infiltrating lymphocytes promote tumor cell apoptosis and remove dead cells by way of humoral and cellular immunity, and these processes are necessary for the host's immune defense and surveillance (37).Therefore, SII has a biological rationale for its role in predicting OSCC prognosis.Notably, a recent single study by Yoshimura et al. investigated the prognostic effect of multiple inflammation-nutrition parameters including NLR, PLR, LMR, CRP-albumin ratio (CAR), Glasgow prognostic score (GPS), modified GPS (mGPS), prognostic nutritional index (PNI), controlling nutrition status (CONUT), and modified CONUT (mCONUT) in patients with OSCC receiving surgery (38).They found that a low PNI was associated with shorter OS and DFS in patients with OSCC through multivariate analysis (38).Although that study did not include SII for analysis in OSCC, their results were important to investigate mechanisms (38).In peripheral blood analyses, inflammation-related markers were mainly composed of upregulated factors (neutrophils, platelets, monocytes, and CRP) and downregulated factors (lymphocytes, albumin, total cholesterol, and hemoglobin).Different combinations of these factors became prognostic indicators and the prognostic parameters were more stable than using a single element.
Many recent studies have also reported that SII could be used to predict the prognosis of different cancer types by conducting meta-analyses (39-43).A meta-analysis on 2,101 patients conducted by Zeng et al. found that elevated pretreatment SII was markedly associated with poor OS and progression-free survival (PFS) in esophageal squamous cell carcinoma (39).According to Wang et al., SII could independently predict OS and PFS of nasopharyngeal carcinoma patients through a meta-analysis that included nine studies (40).In the meta-analysis, which included 833 patients conducted by Salazar-Valdivia et al., indicated that high SII values are related to poor OS and PFS of testicular cancer (41).Moreover, a recent meta-analysis, including 1,426 patients, indicated that higher SII was significantly related to dismal OS and PFS in glioma patients (42).According to Zhang et al., a higher SII is linked dramatically to dismal OS and worse PFS/biochemical recurrence-free survival (bRFS) of prostate cancer in their meta-analysis enrolling 8,133 patients (43).The results of this SII focused meta-analysis mostly conforms to those obtained in additional cancer types.There were some limitations to be noted.First, every enrolled article had a retrospective design, which could introduce selection bias.Second, many enrolled articles were conducted in Asia (10 out of 11).Although the study region was not restricted, all included studies were published in English.
Therefore, the findings of this work may be more applicable in Asian OSCC populations.Third, threshold SII was not uniform across the included studies, so there could be differences to each conclusion.Due to these limitations, more multi-regional prospective trials are still necessary to further validate the utility of SII when predicting the prognosis of OSCC patients.

Conclusions
In conclusion, this meta-analysis demonstrates that higher SIIs are significantly related to dismal OS and DFS in OSCC.
SIIs are markedly related to advanced tumor stages and poor differentiation in OSCC.SII could be a potential and important biomarker for clinical management and prognosis prediction of OSCC patients.

TABLE 2
The subgroup analysis of the prognostic role of SII for OS in patients with OSCC.

TABLE 3
The subgroup analysis of the prognostic role of SII for DFS in patients with OSCC.