Factors Affecting Outcomes in Patients With Stage III & IV Squamous Cell Carcinoma of Oropharynx: The Importance of p16 Status, BMI, and Race

Objective To identify racial disparities in survival outcomes among Stage III & IV patients with squamous cell carcinomas (SCCa) of the oropharynx treated with definitive radiation therapy (RT), with concurrent chemotherapy. Method This is a retrospective analysis of patients with stage III & IV SCCa of oropharynx treated with definitive RT at the State Academic Medical Center. All patients were treated to 70 Gy utilizing intensity-modulated radiation treatment (IMRT), and received concurrent chemotherapy with weekly cisplatin or cetuximab. Chi-square test was used to test the goodness of fit, overall survival (OS), and locoregional control (LRC) comparing races were generated by using Log-rank test & Kaplan-Meier method. The covariables associated with the OS and LRC were determined by the Cox regression model. A p-value of less than 0.05 was considered statistically significant. The SPSS 24.0 software (IBM Corp., Armonk, NY) was used. Results In the total 73 eligible patients, 54.8% were black, and 45.2% white patients. Stage distribution (per American Joint Committee on Cancer-AJCC 8th Ed) between black patients vs. white patients, Stage III (45.5% vs. 54.5%) and for Stage IV (56.5% vs. 43.5%); p=0.499. Median follow-up for the entire group was 41 months (range: 4-144 months). In the univariate analysis, variables p16 status, body mass index (BMI), alcohol history and tumor subsite were found to be significant. In the multivariate analysis, only BMI has shown to be significant. Three-year LRC for black patients was 37.8% vs.66.8% in white patients (p=0.354) and three-year OS for black patients was 51.8% vs. 80.9% for white patients (p=0.063), respectively. Five-year OS for p16 positive patients was 69.7% vs. 43% for p16 negative patients (p=0.034). Five-year OS for Stage IV black patients was 34% vs. 69.5% for Stage IV white patients (p=0.014). Conclusion Among all the co-variables examined, only BMI has shown affecting the OS outcomes; gender and BMI shown to be affecting the LRC. Racial factor appears to be significant in Stage IV patients.


Introduction
Oropharyngeal squamous cell carcinomas (SCCa) include cancers of the base of the tongue, the palatine tonsils, the soft palate, and the posterior oropharyngeal wall [1], and it is the sixth most commonly diagnosed cancers in the United States [2], accounting for 3% of all new cases of cancer in the United States [3]. In previous studies, the increased incidences of oropharyngeal cancers has been strongly associated with tobacco and alcohol exposure but it's been noted that incidence of oropharyngeal SCCa is high among middle-aged nonsmokers or light smokers male whites compared with older men with a significant history of drinking and smoking [4][5][6][7]. Approximately 75% of newly diagnosed oropharyngeal SCCa cases are related to human papillomavirus (HPV) [8,9]. According to the Center for Disease Control and Prevention (CDC), each year about 14,800 new HPV/p16-related oropharyngeal SCCa cases are diagnosed in men and 3,400 in women [10]. About 8.8 men and 1.8 women per 100,000 were diagnosed with HPV/p16related oropharyngeal SCCa among white people, and 6.6 men and 1.4 women per 100,000 among black people [10]. Black people are less likely to have HPV/p16-related oropharyngeal SCCa therefore, black patients have shown to be associated with poor overall survival (OS) when compared to white patients [11,12]. The racial disparities among oropharyngeal cancers reflect the interplay of age, gender, socioeconomic factors, culture, diet, stress, state of residence, the environment and biology [13,14].
The aim of this study is to investigate the OS and locoregional control (LRC) of advanced-stage oropharyngeal SCCa treated at the University of Mississippi Medical Center (UMMC), and whether these differences are related to patient race, tumor characteristics and or socioeconomic status (SES).
This article was previously presented as a scientific meeting abstract at the 2019 AACR-AHNS Head and Neck Cancer Conference at Austin, TX, USA on April 29, 2019 (DOI: 10.1158/1557-3265.AACRAHNS19-B21).

Patients
An institutional review board (IRB # 2010-0252) approval was obtained for retrospective analysis of oropharyngeal SCCa patients treated by definitive RT, with concurrent chemotherapy from 2011 to 2018 at UMMC, Jackson, MS, USA. We used the browser-based database tool Research electronic data capture (RedCap) to gather and store the patient's information in password-protected computers. All the patient's details for demographics, disease presentation, disease staging, treatments, and complications were recorded at the time of their enrollment in the database. Patient identifiers were removed before extracting the data. All the patient's histology is identified as SCCa by biopsy and selected only Stages III & IV. The variables race, gender, smoking and alcohol history were self-reported. The variables weight, body mass index (BMI) were measured before and after treatment. All the demographic variables, disease presentation, disease staging (AJCC stage 8th edition), treatments, and complications for each patient were recorded. Sample size (n=73) with 80% power to detect the factors affecting the survival in this study was estimated by using sample size calculator.

Treatment methods
All patients were treated to 70 Gy utilizing intensity-modulated radiation treatment (IMRT), and all patients received concurrent chemotherapy with weekly cisplatin, 40mg/m 2 (61.6%) or weekly cetuximab, 400 mg/m 2 (38.4%). All the treatment plans were evaluated by departmental peer review before the initiation [15]. HPV infection was confirmed to be positive, if 70% or more of the tumor cells show cytoplasmic and nuclear staining for p16 [16,17].

Statistical analysis
Fisher test is used to evaluate the racial comparison between black patients and white patients. OS for all the 73 patients in the cohort were calculated from the date of disease diagnosis and the patients last contact date. To evaluate the OS and LRC, Kaplan-Meier method was used and the comparison between black patients & white patients was measured by the log-rank test. The covariables associated with the OS and LRC were determined by the univariate and multivariate Cox regression model. Hazards ratio (HR) were estimated by time to event outcome with associated 95% confidence intervals (CIs) and p-values ≤ 0.05 were considered significant. The SPSS 24.0 software (IBM Corp., Armonk, NY) was used for data analysis.

Patient characteristics
This study includes a total of 73 patients with SCCa of the oropharynx treated from 2011 to 2018 at UMMC. The median follow-up was 41 months (range: 4 to 144 months). Among the total 73 eligible patients, 54.8% were black patients, and 45.2% were white patients ( Figure 1). Summary of patient's demographic and clinical description was represented in (Table 1). There was no significant difference in gender or sex, age, BMI, employment status, distance traveled to the treatment facility, marital status smoking and alcohol history between black patients and white patients.

Cox regression analysis for overall survival
In the univariate OS Cox regression analysis, variables like BMI, p16 status, alcohol history, and tumor subsites showed a statistically significant difference in disease prognosis (

Discussion
In this retrospective study of 73 oropharyngeal SCCa patients, we evaluated variables, socio-demographic individual characteristics (race/ethnicity, age, gender, BMI, marital status, health insurance); clinical characteristics (tumor site, tumor grade and p16 status); county-level SES attributes (age distribution, unemployment, income, distance traveled to treatment facility, smoking, and alcohol history).
We found the following important findings.
Our study found no statistically significant difference in survival outcomes between black and white patients' among Stage III & IV SCCa. This may be due to few numbers of patients in Stage III (15.1% vs. 84.9%) compared to Stage IV. However, racial factor appears to be significant in Stage IV patients, the black patients showed statistically significant worse 5-year OS compared to white patients, 34% vs. 69.5% (p=0.014).
Our study reports that overweight patients show statistically significant increased risk of death compared to normal weight patients, (HR 4.75, 95% CI, 1.31-17.24; p=0.018) Our study demonstrates that patients with alcohol history show a statistically significant worse survival compared to the patients without alcohol history, 42.3% vs. 76.9% (p=0.034).
Our study observed a trend of statistically significant worse survival for p16 negative patients compared to p16 positive patients, 43% vs. 69.7% (p=0.034).
In our study population, white patients present overall the best distribution according to the patient, tumor, socioeconomic and county-level attributes, similar to previous studies where black patients were more frequently presented with the least favorable distribution [18]. Many studies reported that the association of the black population with lower household income, lower employment rates are more likely associated with increased rates of smoking and heavy alcohol use [19,20]. Our study found no statistical difference in survival outcomes between the black patients and the white patients, because of few numbers of patients in Stage III compared to Stage IV. However, in Stage IV, black patients showed statistically significant worse survivals compared to their white counterparts.
The smoking and alcohol consumption rates were similar between white and black patients; however, in overall cohort, the patients with alcohol history demonstrated worse survival compared to the patients without alcohol history. Black oropharyngeal SCCa patients are more likely to reside in lower education and lower-income level and less likely to have insurance compared with the white study population [19,[20][21][22][23]. In our study, black patients (62.5%) were reported to be from low-income level compared to their white counterparts (36.4%).
A study conducted by Choi et al. showed that patients with the following characteristics, the race black, lowincome level, low educational level, advanced age, being single, smoking history, alcohol history, lower BMI, advanced clinical stage, and treated with chemotherapy were associated with poor prognosis [24,25]. In our study, the results observed for OS univariate analysis by Cox regression showed that having an overweight BMI, alcohol history, p16 negative status and BOT tumor subsite were associated with poor prognosis. However, in the multivariate model, we found that the BMI is the only important prognosis factor. It's been observed in other studies that black patients were higher at risk for death due to HN cancers. In our cohort, the observed differences with no statistical significance in OS considering race/ethnicity, indicates that equal clinical care (access to care and quality of care) provided at our state academic medical center and SES could bring more equity in racial/ ethnic differences observed for oropharyngeal SCCa survival rates.
Tumors induced with HPV have different biology than tobacco and alcohol-induced tumors [26][27][28]. HPV/p16 is one of the risk factor for oropharyngeal SCCa, but some studies showed that p16 positive tumors present a better prognosis than tobacco and alcohol-induced oropharynx tumors [29,30]. In our cohort, we observed a trend of worse OS for p16 negative (43%) compared to p16 positive patients (69.7%) which is inexplicable.
In our analysis of LRC, variables like gender, and BMI showed a statistically significant difference in prognosis. We observed that female patients presented a worse prognosis when compared to male patients, the overweight BMI patients presented twice the risk of relapsing oropharyngeal SCCa and p16 negative showed 19% decreased risk of relapse compared to p16 positive patients. In the multivariate analysis, gender and BMI are independent prognosis factors for LRC.
Equal access to healthcare is crucial to reduce health disparities along with lifestyle interventions such as dietary modification, smoking cessation, alcohol cessation and increased exercise should be incorporated into treatment guidelines and considered when defining research priorities.

Limitations
The number of patients in our retrospective study is very small and the income variable in this study is at the census tract level, so there might be a possibility of SES differences at the individual level confounding the racial/ ethnic differences reported.

Conclusions
Our retrospective study shows that p16 status, tumor subsites, BMI level, and alcohol history play important role in oropharyngeal SCCa survival rate. In the demographic and clinical analysis, the variables income quartile and p16 status shown statistical significance. In the univariate analysis, when adjusted for patients characteristics (BMI and alcohol history), tumor characteristics (p16 status and tumor subsite) showed a higher risk of death from oropharyngeal SCCa for our study population. In the subsequent multivariate analysis, the risk of death from oropharyngeal SCCa was no longer statistically significant for alcohol history, tumor subsite or p16 status but still high for BMI. In this retrospective study, we did not find statistical significant poor outcomes among any specific race, indicating an equal access to healthcare in our state academic medical center. In the future, we need to aim at the individual SES rather than census tract to better understand its effect on the racial disparity. In addition, studies should investigate the mechanisms surrounding the relationships between BMI and risk for HN cancer observed in this study as well as its implications on disease management and prevention.

Additional Information Disclosures
Human subjects: Consent was obtained or waived by all participants in this study. University of Mississippi Medical Center Institutional Review Board issued approval 2010-0252. Your Continuing Review was reviewed and approved by the Expedited Review process on 06/29/2020. You may continue this research. Animal subjects: All authors have confirmed that this study did not involve animal subjects or tissue.

Conflicts of interest:
In compliance with the ICMJE uniform disclosure form, all authors declare the following: Payment/services info: All authors have declared that no financial support was received from any organization for the submitted work. Financial relationships: All authors have declared that they have no financial relationships at present or within the previous three years with any organizations that might have an interest in the submitted work. Other relationships: All authors have declared that there are no