Role of systemic inflammatory markers (biomarkers) as important prognostic factors for survival in head and neck cancer

Introduction and Aim: Cancer related inflammation is recognized as a critical multifaceted player in tumour initiation and progression. The neutrophil to lymphocyte ratio (NLR) and platelet to lymphocyte ratio (PLR) are important markers of systemic inflammatory burden in malignancy. The present retrospective study aimed to evaluate the prognostic value of pre-treatment baseline (NLR) and (PLR) for survival in HNC patients. Materials and Methods: Analysis of data of 257 patients with head and neck cancer treated with definitive therapy over 3.5 yrs. Neutrophil, lymphocyte and platelet counts before treatment of all patients were collected and NLR and PLR were calculated. Results: Median value of NLR and PLR were 3.7 and 170.63 respectively. Receiver operator characteristics (ROC) curve analysis showed the predictive cut-off value of NLR and PLR as 3.9 and 158.3. Univariate analysis using Cox-regression model showed NLR (p < 0.001) and PLR (p = 0.001) significantly influenced the locoregional recurrence free survival (LRRFS) and overall survival (OS). The multivariate analysis showed NLR is the independent prognostic factors influencing LRRFS (p = 0.007) and OS (p = 0.002). Kaplan-Meier survival curve (Log rank test; Chi-square (χ 2 ) value) showed OS is significantly influenced by NLR group (LRRFS; χ 2 = 23.9 and OS; χ 2 = 33.7) and PLR group (LRRFS; χ 2 = 11.2 and OS; χ 2 = 19.3) in contrast to LRRFS. Conclusion: NLR can be strongly used as biomarker for prognostication for outcome and survival in head neck cancer. However, a well-designed, larger studies with longer follow-up is warranted.


INTRODUCTION
ead and neck cancers (HNCs) affect the upper aerodigestive tract and are one of the most common cancers worldwide (1). Nearly 60% of the population presents with locally advanced but non-metastatic disease. Radiotherapy along with concurrent chemotherapy has long been the standard nonsurgical therapy for locally advanced disease. Definitive chemo-radiotherapy is the main modality of treatment in locally advanced cancer but response varied depending upon both tumour characteristics and host related factors like age, sex, performance status etc., (2).
Inflammation is considered to have role in tumour development and progression (3). In the 19 th century, Rudolf Virchow showed leukocytes were present in tumours and suggested a possible link between inflammation and cancer and chronic inflammation together with the tissue injury enhance cell proliferation (4). Various studies have also showed close associations of systemic inflammation as an independent prognostic factor for treatment outcomes for many types of cancers, including colorectal cancer, renal cell cancer, pancreatic cancer, and head and neck squamous cell carcinoma (5). Neutrophil to lymphocyte ratio (NLR) may also predict the survival in HNCs (6). Host response to malignant tumours involved in systemic alterations, not only changes in the tumour microenvironment. An elevation of circulating neutrophil count is thought to be the result of tumour cells releasing cytokines, which stimulate the bone marrow to produce neutrophils (7).
Malignant tumours may also influence the number of platelets and alter their functioning. Several hypotheses have been suggested on pathophysiological relation between tumour progression and thrombocytosis (8). Platelets may also support tumour progression through facilitation of neo-angiogenesis, generation of adhesion molecules and rise of early metastatic niches. For this reason, platelet to lymphocyte ratio (PLR) has been proposed as an independent prognostic factor in different solid tumours (9). Although various clinical and molecular biomarkers of treatment outcomes after definitive radiotherapy (RT) for HNC have been investigated, no clear consensus regarding reliable H predictive indicators has been reached. This study aimed to prove NLR and PLR as a predictive biomarker for survival in HNCs.

MATERIALS AND METHODS
The clinical data of 257 patients treated either by definitive radiotherapy or by adjuvant radiotherapy at Institute of Post Graduate Medical Education and Research (IPGME&R), Kolkata, from February 2015 to August 2018 were collected. Staging was done based on the American Joint Committee on Cancer (AJCC) TNM staging system (7th edition, 2010) (10). The diagnosis had been validated by histopathological examination. The inclusion criteria for the patients were as follows: (i) age more than 18 years; (ii) complete clinical, laboratory, imaging, and follow-up data, (iii) no chronic infectious diseases, (iv) without any haematological disorders or any treatment within one month before radiotherapy leading to elevated NLR and PLR, like administration of hematopoietic drugs such as G-CSF, (v) no autoimmune disease or usage of steroids. The performance status of patients was categorised using Eastern cooperative oncology group (ECOG) (11). Routine blood counts were done within seven days prior to starting RT. NLR and PLR were calculated as the absolute neutrophil or platelet counts, respectively, divided by the absolute lymphocyte counts. After treatment completion, patients were followed up every 3 months. Local recurrence of the disease was referred as locoregional recurrence (LRR). The locoregional recurrence free survival (LRRFS) was defined as interval between completion of treatment (adjuvant radiation or definitive radiation) and locoregional recurrence. The overall survival (OS) was calculated as the interval between the completion of treatment to the date of death or the last visit.

Statistical analysis
The Statistical Package for the Social Sciences (IBM SPSS for Windows, version 25.0) was used for the statistical analysis. Descriptive statistics were used to characterize the patient population. Continuous variables were analysed using Student's t-test. Receiver operator characteristics (ROC) curve analysed using the NLR and PLR for locoregional recurrence and event at last follow-up to find out the predictive cut-off value of NLR and PLR. The assessment of the accuracy of NLR and PLR cut-off value is done by calculation of area under curve (AUC) and 95% confidence interval (CI). The Kaplan-Meier method was used to compute LRRFS and OS. Comparison of variables influencing LRRFS and OS were done using Logrank test. Univariate and multivariate analysis were performed using Cox regression model. The factors found statistically significant on univariate analysis were included in multivariate analysis for survival outcomes. A pvalue less than 0.05 considered statistically significant in all performed analysis.

Clinicopathological characteristics
Two hundred and fifty-seven patients were included in this study. The epidemiological, clinical, pathological and therapeutic data have been mentioned in Table-1.

ROC curve and optimal cut-off value of NLR and PLR
The optimal cut-off points of NLR was calculated for events (including locoregional recurrence and status at last follow-up) using ROC curve. The results showed that the optimal predictive cut-off value of NLR is 3.9 (area under ROC curve = 68.2%, p < 0.001), at which most significant difference was observed in events. The sensitivity and specificity was 66.4% and 69.5% respectively (fig 1). The patients were divided into two groups based on NLR cut-off points into low risk NLR group (NLR ≤ 3.9) and high risk NLR group (NLR > 3.9). The frequency of distribution of NLR risk group i.e., low risk NLR and high risk NLR were 52.9% and 47.1% respectively.

DISCUSSION
Inflammatory responses play major decisive roles at different stages of tumour development, including initiation, promotion, malignant conversion, invasion, and metastasis. Inflammation also affects immune surveillance and responses to therapy and elevated inflammatory mediators are related to poor prognosis in patients with cancer (12).
HNC occurs more frequently in older patients. The incidence of these tumours is low among the persons younger than 40 years. 9% of our patients were less than 40years.This rate matches with results of other studies around the world.Most of the patients (33.9%) in our study were aged between 51 to 60 years and 80% patients were between 41 to 70 years of age which approximates with other study (13). Most patients (83.3%) in current study were male. Large scale epidemiological studies and national cancer registries showed that male to female ratio varied depending of site of tumour from 2:1 to 15:1 (14). HNC in women has been increased significantly in recent days, probably due to changes in environmental exposure (15). In the present study, the most common involved site was larynx (37.7%) followed by oral cavity ca (27.6%). This finding is in agreement with other reports (16) but it is not matched with studies where oral cavity is the most affected site (17).
In the current study, the median duration of symptoms in most cases were 5 months. Delay in diagnosis may be due to illiteracy or poor financial status, or tendency of most patients to resort to home remedies before seeking medical help. A comparable time span has been reported by Shenoi et al., in their study (18). In the present study, microscopically, Grade II (moderately differentiated) tumours were the most common (64.4%) followed by Grade I (well differentiated) tumours (23.3%).
Normal NLR values are in the range of 0.78-3.53 in the general and NLR and PLR cut-off levels deployed in numerous studies varied in a wide range of PLR: 105.3-170 and NLR: 2.0 to 4.2 (20). In our study, the optimal threshold value, determined with the statistical method detailed above, was 3.9 for NLR and158.3 for PLR. The frequency of distribution of NLR risk group i.e., low risk NLR and high risk NLR were 52.9% and 47.1% respectively.
In the present study, we found NLR as one of the important prognostic factors for LRR and OS. Our findings matched with other studies (19,20 (25).
In contrast to this, there were many studies which were not matched with our result where PLR was not associated with outcome with no significant correlation between baseline PLR and OS or the association between PLR and OS or DFS was not statistically significant (21). Our results are therefore consistent with the concept that a high NLR contributes to poor disease control by suppressing the cytolytic activities of activated effector T cells and the peritumoral infiltration of immunosuppressive cells such as macrophages (26).
We further demonstrated that a higher PLR was associated with PFS in our multivariate analysis. Again, the significance of interactions between platelets and the tumour microenvironment remains somewhat unclear. The platelet count provides an additional index of systemic inflammation elicited by the tumour and degranulation. This inflammation with consequent release of platelet-derived proangiogenic mediators within the microvasculature of tumour serve as important factor for tumour growth (27).

CONCLUSION
Overall, the findings of the present study support that NLR and PLR are the potential biomarkers for host response to tumour aggressiveness, therapeutic response to chemo-radiotherapy and survival in HNC patients. NLR can be strongly used as biomarker for prognostication for outcome and survival of patients with head neck cancer. However, a well-designed, larger studies with longer follow-up is warranted.