Comparative prognostic value of different preoperative complete blood count cell ratios in patients with oral cavity cancer treated with surgery and postoperative radiotherapy

Abstract Background We sought to compare the prognostic significance of different preoperative complete blood count cell ratios in patients with oral cavity squamous cell carcinoma (OSCC) treated with surgery and postoperative radiotherapy (PORT). Methods We retrospectively reviewed the clinical records of 890 patients with OSCC who were treated with surgery and PORT. The following preoperative complete blood count cell ratios were collected: neutrophil‐to‐lymphocyte ratio (NLR), platelet‐to‐lymphocyte ratio (PLR), and lymphocyte‐to‐monocyte ratio (LMR). Overall survival (OS), local control, regional control, and distant control (DC) served as the main outcomes of interest. Results The results of multivariate analysis in the entire study cohort revealed that a low NLR was the only independently favorable marker of both OS (adjusted hazard ratio [HR]: 0.794, 95% confidence interval (CI): 0.656–0.961, bootstrap p = 0.028) and DC (adjusted HR: 0.659, 95% CI: 0.478–0.909, bootstrap p = 0.015). Both LMR and PLR were not retained in the model as independent predictors. Subgroup analyses in high‐risk patients (i.e., those bearing T4 disease, N3 disease, or poor differentiation) revealed that a high NLR was a significant adverse risk factor for both OS and DC (all p < 0.03)—with a borderline significance being evident for DC in patients with T4 disease (p = 0.058). Conclusions A high pretreatment NLR was an independent unfavorable risk factor for both OS and DC in patients with OSCC who underwent surgery and PORT. No other preoperative complete blood count parameters and cell ratios were found to have prognostic significance.


| INTRODUCTION
Inflammation has been shown to promote tumor initiation and progression, whereas escape from immune surveillance may favor cancer invasiveness and distant spread. [1][2][3] There is consistent evidence that a high tumor infiltration by neutrophils and macrophages has an adverse prognostic significance. 4,5 In contrast, tumor-infiltrating lymphocytes portend more favorable outcomes. [6][7][8] Various preoperative complete blood count cell ratios-including neutrophilto-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), and lymphocyte-to-monocyte ratio (LMR)-have been extensively investigated in relation to prognosis in patients with different solid malignancies, including oral squamous cell carcinoma (OSCC). [9][10][11][12] However, the comparative value of NLR, PLR, and LMR for predicting clinical outcomes in patients with OSCC remains unclear. Further, most published studies have been focused on overall survival (OS).
Starting from these premises, we designed this study to specifically compare the prognostic significance of different preoperative complete blood count cell ratios in patients with OSCC treated with surgery and postoperative radiotherapy (PORT). Besides OS, the ratios were investigated in relation to other clinical endpoints-including local control (LC), regional control (RC), and distant control (DC).

| Patients
We retrospectively reviewed the clinical records of patients who had undergone radical surgery and PORT (either with or without chemotherapy) at our hospital between January 2005 and December 2012 (n = 1055). Patient staging was performed according to the 2018 American Joint Committee on Cancer TNM staging system. Exclusion criteria were as follows: (a) unavailability of official pathological reports (n = 97), (b) not squamous cell carcinoma (n = 7), (c) presence of a second primary cancer occurring in the three years preceding or following treatment for the primary tumor (n = 45), (d) equivalent dose in 2 Gy fractions (EQD2) <60.0 Gy (n = 15), and 5) age <18 years (n = 1). Figure 1 depicts the flow of patients through the study. Data collection was performed by a radiation oncologist and an experienced nurse. The study protocol followed the tenets of the Helsinki declaration and was granted ethics approval by F I G U R E 1 Patient flow chart of the study the Institutional Review Board Committee of our hospital. Owing to the retrospective nature of the study, the need for informed consent was waived.

| Calculation of pretreatment blood count cell ratios
A pretreatment complete blood count was obtained in the 14 days preceding radical surgery. The three ratios of interest (NLR, PLR, and LMR) were calculated from absolute counts of neutrophils, lymphocytes, platelets, and monocytes, as appropriate.

| Variable definitions
In keeping with the American Centers for Disease Control and Prevention classification system, cigarette smoking was dichotomized as yes (subjects who smoked ≥100 cigarettes in their lifetime) vs. no (subjects who smoked <100 cigarettes in their lifetime and not currently smoking). Alcohol drinking (current or former drinkers vs. non-drinkers) and betel quid chewing (current or former chewers vs. nonchewers) were similarly considered as dichotomous variables. Pretreatment BMI-calculated as pretreatment weight in kilograms divided by height in meters squared-was dichotomized as underweight or normal (BMI <25 kg/m 2 ) vs. overweight (BMI ≥25 kg/m 2 ).

| Statistical analyses
The primary outcome measure was OS, whereas LC, RC, and DC served as secondary endpoints. Survival was calculated as the time elapsed (in years) from the start of PORT to the event of interest. The optimal cutoff points for NLR, PLR, and LMR were based on where the Youden index (sensitivity +specificity -(a) was maximal via timedependent receiver operating characteristic (TDROC) curve analysis taking the overall survival (OS) at 5 years from the start of PORT as the endpoint of interest. Patients were divided into two groups (high vs. low) according to the optimal cutoff values. Intergroup differences were assessed with the Student's t-test (continuous variables) or the chi-square tests (categorical variables). Survival curves were constructed using the Kaplan-Meier method (log-rank test). Cox proportional hazard regression models were used to assess the impact of each variable on the study endpoints. Results are expressed as hazard ratios (HRs) with their 95% confidence intervals (CIs). Two-tailed p values <0.05 were considered statistically significant. The bootstrap method (1000 resamples) was used for internal validation. TDROC curve analysis was performed in the R environment using the "timeROC" package (R Foundation for Statistical Computing, Vienna, Austria) through inverse probability of censoring weighting (IPCW) approach to estimate timedependent ROC curve and AUC for censored events with competing risks. All other calculations were carried out with SPSS, version 22.0 (IBM).

| Preoperative complete blood count cell ratios and overall survival
The following variables were investigated for their associations with OS in univariate analyses: age, sex, tumor subsites,   clinicopathologic factors, BMI, risky oral habits, and complete blood count cell ratios (Table 2 and Figure 2D). Variables with univariate associations at a p < 0.2 level were entered as covariates in multivariate analyses (Table 3 and Figure 3). Significant univariate adverse risk factors for OS were older age (p < 0.01), high T stage (p = 0.001), high N stage (p < 0.001), high clinical stage (p < 0.001), lower BMI (p < 0.001), alcohol drinking (p = 0.008), treatment with chemotherapy (p < 0.001), high NLR (p = 0.001) ( Figure 4A), and high PLR (p = 0.025). In multivariate analyses, a low NLR was independently associated with a more favorable OS

| Preoperative complete blood count cell ratios and local, regional, and distant control
The following variables were investigated for their associations with LC, RC, and DC in univariate analyses: age, sex, tumor subsites, clinicopathologic factors, BMI, risky oral habits, and complete blood count cell ratios (Table 2 and Figure 2A, B, and C). Variables with univariate associations at a p < 0.2 level were entered as covariates in multivariate analyses (Table 3 and Abbreviations: AJCC, American Joint Committee on Cancer; BMI, body mass index; CI, confidence interval; HR, hazard ratio; LMR, lymphocyte-to-monocyte ratio; NLR, neutrophil-to-lymphocyte ratio; PLR, platelet-tolymphocyte ratio.

| Subgroup analyses in high-risk patients
Because a high NLR was the blood count cell ratio most consistently associated with adverse outcomes, we performed subgroup analyses of this variable in different subgroups of patients bearing risk factors for distant metastases or death (T4 disease, N3 disease, and poor differentiation). A high NLR was significantly associated with a less favorable OS in all high-risk subgroups (all p < 0.02; Figure 4C, E, and G). Similarly, a high NLR was associated with less favorable DC in patients with N3 disease or poor differentiation (all p < 0.03; Figure 4F and H) but not in those with T4 disease (p = 0.058; Figure 4D).

| DISCUSSION
In this study, we compared the prognostic value of different preoperative complete blood count cell ratios in patients with OSCC who were treated with radical surgery and PORT. Our results indicate that NLR was superior to both LMR and PLR in the prediction of OS and DC. Notably, NLR retained its statistical significance even in specific subgroups of high-risk patients, suggesting that it may further refine prognostic stratification with respect to traditional risk factors for poor OS and DC (T4 disease, N3 disease, and poor differentiation).

F I G U R E 3
Forest plot based on multivariate analysis adjusted hazard ratios (HRs) from Cox regression for local control, regional control, distant control, and overall survival in the entire cohort (n = 890). Abbreviations: AJCC, American Joint Committee on Cancer; BMI, body mass index; CI, confidence interval; HR, hazard ratio; NLR, neutrophil-to-lymphocyte ratio The exact mechanisms whereby NLR has a higher predictive value than LMR and PLR remain to be established. In general, white blood cell and platelet counts reflect an individual's systemic and/or local inflammatory status. Neutrophils are known to produce cytokines, chemokines, and growth factors that may promote angiogenesis as well as tumor cell proliferation and migration. 13 Numerous studies have consistently shown that an increased neutrophil count predicts adverse outcomes in patients with different solid cancers. [14][15][16] In contrast, lymphocytes are responsible for antitumorspecific immune response-including T-lymphocyte tumor infiltration 17 and cytotoxic T-lymphocyte-mediated antitumor activity. 18 Notably, a low lymphocyte count is a poor prognostic factor in patients with malignancies. 19,20 Platelets produce growth factors that promote cancer growth and its distant spread. 21,22 A high platelet count predicts unfavorable outcomes in patients with head and neck malignancies, and antiplatelet agents may have a therapeutic antitumor potential. 23 Finally, monocytes-which can differentiate into tumor-infiltrating macrophages and dendritic cells-produce proinflammatory molecules involved in carcinogenesis and tumor metastasis. 4,24 In this regard, a high monocyte count has an adverse prognostic significance in patients with oral cavity cancer. [25][26][27] Based on these observations, it is not surprising that high NLR and PLR and a low LMR have been related to increased cancer-related mortality and recurrence rates. [28][29][30][31] Albeit being the most widely applied tool for predicting prognosis in patients with OSCC, the TNM staging system is a static instrument that solely relies on tumor-related characteristics. In this scenario, there is an urgent need for reliable prognostic tools grounded on simple preoperative variables. Our results clearly indicate that the preoperative NLR is a simple and effective index that warrants further scrutiny in