Background
Squamous cell carcinomas of the head and neck (SCCHNs) are increasing in incidence worldwide.1,2 Contemporary reports have identified HPV infection as a leading risk factor for development of SCCHN, particularly in oropharyngeal carcinoma.3,4 Although the incidence of HPV-negative oropharyngeal carcinoma in the United States has been declining because of decreased smoking, HPV-positive disease is increasing.5
Social determinants of health for HPV-positive and HPV-negative SCCHN are poorly characterized—available data are limited to case reports or retrospective analyses of prospective trials.5–15 It has been suggested that HPV-negative SCCHN is more common among nonwhite, urban, and poorly insured individuals due to higher smoking rates in these populations.7,14 Furthermore, the inferior outcomes observed in patients of lower socioeconomic status (SES) with SCCHN have largely been attributed to a predominance of HPV-negative disease.7,14 However, these hypotheses have not been tested in the prospective setting, given that clinical trials tend to include a disproportionate number of patients who are white and of higher SES.16 Furthermore, no large population-based studies have examined the burden of HPV-associated disease and cancer-related survival as affected by socioeconomic factors.
Therefore, we sought to characterize the interplay between HPV status, race, socioeconomic factors, and outcomes in SCCHN.
Methods
Study Cohort
We used the custom SEER Head and Neck with HPV Status Database to identify 4,735 adult patients (aged >18 years) with primary nonmetastatic (M0) SCCHN and known HPV status diagnosed in 2013 through 2014.17 These data are not yet public; we obtained access via a proposal to the SEER custom data group, where the a priori analyses described were determined to be appropriate use of the data. The study inclusion period of 2013 through 2014 represents the years in which HPV status was collected and reviewed for quality assurance.
SEER cancer registries code primary cancer site and histology based on ICD-O-3. Squamous cell carcinoma histology of the head and neck was identified using the following ICD-O-3 histologic type codes: 8050, 8051, 8054, 8070 through 8076, 8083, and 8094. TNM staging was determined using the 7th edition of the AJCC Cancer Staging Manual, as provided by SEER (8th edition staging is not yet reported by the database). Race was classified as non-Hispanic white versus nonwhite/Hispanic. Small-area estimates for percent ever-smoker were linked to SEER via estimates developed from the Behavioral Risk Factor Surveillance System (BRFSS) and the National Health Interview Survey (NHIS).
Statistical Analyses
Comparison of Baseline Patient Characteristics by HPV Status
The Wilcoxon rank-sum and Mantel-Haenszel chi-square tests compared distributions of continuous and categorical covariates, stratified by HPV status among patients with M0 disease (N=4,735). Multivariable logistic regression defined adjusted odds ratios (aORs) and 95% confidence intervals (CIs) between patient characteristics and HPV status.
aORs for Presentation With Advanced Disease
Table 1 shows baseline patient characteristics. In addition, we performed multivariable logistic regressions to calculate the aORs of presenting with HPV-positive disease. The following variables pertaining to SES were included in the model: race (white [referent] or nonwhite), insurance status (private or other insurance [referent], uninsured, Medicaid, unknown), and marital status (married [referent], single, divorced/widowed/separated, unknown). Demographic and clinical variables included in the model were age at diagnosis (continuous), sex (female [referent], male), tumor stage (T1 [referent], T2, T3, T4, unknown), nodal stage (N0 [referent], N2a, N2b, N2c, N2 not otherwise specified [NOS], N3, unknown), oropharynx versus nonoropharynx SCCHN subsite (nonoropharynx included nasopharynx, hypopharynx, and other pharyngeal SCCHN), smoking propensity, household income (continuous county variable), education level (continuous county variable), and initial definitive treatment (none [referent], surgery, radiotherapy, chemotherapy). In the SEER database, smoking propensity, education level, and household income are population-level variables abstracted from county data. In Table 2, multivariable logistic regressions defined aORs and associated 95% CIs for odds of presenting with advanced disease (defined as T3–T4 or N2–N3). Given that advanced stage was the outcome of interest in this analysis, T and N characteristics were not included in the model separately.
Baseline Patient Characteristics
Multivariable aOR of Presenting With Advanced Disease (T3–T4 or N2–N3)
Cancer-Specific Mortality Estimates by HPV Status, Tumor, and SES Characteristics
STATA/SE, version 14.2 (StataCorp LLP) was used for survival analyses of patients with M0 disease with at least 1 month of follow-up (n=4,476), in which the primary endpoint was cancer-specific mortality (CSM). Patients were stratified into the following 4 subgroups: HPV-positive oropharynx, HPV-negative oropharynx, HPV-positive nonoropharynx, and HPV-negative nonoropharynx. For each of these subgroups, multivariable Fine-Gray competing risks regression was used to estimate hazard ratios for the socioeconomic and clinical factors described earlier.
To ascertain whether there was a differential prognosis of race by HPV status and disease site, a Fine-Gray competing risks regression model for CSM included a race (nonwhite vs white) * disease subsite (nonoropharynx vs oropharynx) * HPV status (positive vs negative) interaction term. Once it was established that this interaction was statistically significant, we performed further multivariable analyses of the 4 individual subgroups that included the original factors in the model, excluding the interaction term, HPV status, and disease subsite.
Using the defined subgroups of HPV status and disease subsite, cumulative incidence plots for CSM were generated for the purposes of illustration along with estimation of 20-month CSM for each subgroup. Adjusted hazard ratios (aHRs) with associated 95% CIs and P values were calculated for all covariates. Statistical testing was 2-sided, with level of significance set at P=.05. The Dana-Farber/Harvard Cancer Center Institutional Review Board approved this study.
Results
Comparison of Baseline Patient Characteristics by HPV Status
As shown in Table 1, there were significant clinical and socioeconomic differences between HPV-positive and HPV-negative subgroups of patients with SCCHN. An oropharyngeal primary (aOR, 1.90; 95% CI, 1.62–2.24; P<.001), nodal involvement (aOR, 1.62; 95% CI, 1.16–2.27; P=.004 for N3 vs N0), male sex (aOR, 1.59; 95% CI, 1.33–1.89; P<.01), and higher education (aOR, 1.02; 95% CI, 1.00–1.03; P=.01) were associated with greater odds of having HPV-positive SCCHN. In contrast, bulky primary tumors (aOR, 0.68; 95% CI, 0.54–0.85; P=.001 for T4 vs T1), older age at diagnosis (aOR, 0.92; 95% CI, 0.98–0.99; P<.001), no health insurance (aOR, 0.53; 95% CI, 0.37–0.77; P=.001 for uninsured vs private/other insurance), single marital status (aOR, 0.67; 95% CI, 0.56–0.80; P<.001), and nonwhite race (aOR, 0.40; 95% CI, 0.31–0.52; P=.01) were associated with higher odds of having HPV-negative SCCHN.
aOR for Presentation With Advanced Disease
Associations between patient characteristics and advanced stage at presentation are shown in Table 2. On multivariable analysis, advanced stage at presentation was positively associated with oropharyngeal subsite (aOR, 2.31; 95% CI, 1.91–2.80; P<.01), male sex (aOR, 1.50; 95% CI, 1.26–1.78; P<.01), Medicaid insurance (aOR, 1.33; 95% CI, 1.06–1.66; P=.01), single marital status (aOR, 1.28; 95% CI, 1.06–1.54; P=.01), and divorced/widowed/separated marital status (aOR, 1.38; 95% CI, 1.15–1.66; P=.001), and was negatively associated with white race (aOR, 0.68; 95% CI, 0.57–0.81; P<.001).
CSM Estimates by HPV Status, Tumor, and SES Characteristics
After a median follow-up of 10 months, there were 339 cancer-related deaths and 109 competing deaths. White patients with HPV-positive oropharyngeal SCCHN had a significantly lower risk of CSM compared with their nonwhite peers (aHR, 0.55; 95% CI, 0.34–0.88; 20-month CSM, 5.6% vs 11.2%; P=.01; Figure 1A). Among those with HPV-negative oropharyngeal SCCHN, no statistically significant difference in CSM was seen between white and nonwhite patients (aHR, 1.11; 95% CI, 0.74–1.67; 20-month CSM, 22.9% vs 20.9%; P=.60; Figure 1B). Similarly, there were no significant differences in CSM between white and nonwhite patients for nonoropharyngeal SCCHN, regardless of HPV status (Figure 1C and D, Table 3). Accordingly, there was a significant interaction between race, disease subsite, and HPV status (Pinteraction<.001; see Table 3).
Kaplan-Meier cancer-specific survival curves comparing white and nonwhite patients with (A) HPV-positive oropharyngeal, (B) HPV-negative oropharyngeal, (C) HPV-positive nonoropharyngeal, and (D) HPV-negative nonoropharyngeal squamous cell carcinoma of the head and neck.
Citation: Journal of the National Comprehensive Cancer Network J Natl Compr Canc Netw 18, 2; 10.6004/jnccn.2019.7356
Kaplan-Meier cancer-specific survival curves comparing white and nonwhite patients with (A) HPV-positive oropharyngeal, (B) HPV-negative oropharyngeal, (C) HPV-positive nonoropharyngeal, and (D) HPV-negative nonoropharyngeal squamous cell carcinoma of the head and neck.
Citation: Journal of the National Comprehensive Cancer Network J Natl Compr Canc Netw 18, 2; 10.6004/jnccn.2019.7356
Kaplan-Meier cancer-specific survival curves comparing white and nonwhite patients with (A) HPV-positive oropharyngeal, (B) HPV-negative oropharyngeal, (C) HPV-positive nonoropharyngeal, and (D) HPV-negative nonoropharyngeal squamous cell carcinoma of the head and neck.
Citation: Journal of the National Comprehensive Cancer Network J Natl Compr Canc Netw 18, 2; 10.6004/jnccn.2019.7356
Multivariable aHR for CSM Among Patients With M0 SCCHN and at Least 1 Month of Follow-Up (N=4,476)
In addition to risk of CSM based on race, uninsured patients with HPV-positive oropharyngeal SCCHN had a higher risk of CSM compared with their privately insured peers (aHR, 3.12; 95% CI, 1.19–8.13; P=.02); notably, such differences were not seen in other HPV-negative or nonoropharyngeal subgroups. Marital status did not have a significant effect on CSM in patients with HPV-positive oropharyngeal SCCHN, but among patients with HPV-negative oropharyngeal SCCHN, those who were single had a worse risk for CSM (aHR, 1.66; 95% CI, 1.04–2.65; P=.04).
Discussion
This report is, to our knowledge, the largest population-based analysis of the socioeconomic factors affecting outcomes in HPV-associated SCCHN. We found that nonwhite patients with HPV-positive oropharyngeal SCCHN were at higher risk of CSM than white patients, whereas both white and nonwhite patients with HPV-negative or nonoropharyngeal HPV-positive SCCHN had similarly poor outcomes (Pinteraction<.001). In addition, uninsured patients with HPV-positive oropharyngeal SCCHN had an increased risk of CSM compared with their privately insured peers.
These findings are significant in that they contradict earlier reports and hypotheses suggesting that the worse CSM seen in nonwhite populations with SCCHN is driven by a higher likelihood of HPV-negative disease among this subgroup6 or that nonwhite patients with HPV-negative disease and lower SES have disproportionately worse outcomes.14,18 Differences in CSM between white and nonwhite populations with HPV-associated oropharyngeal SCCHN persisted even after adjusting for relevant potential socioeconomic confounders, such as insurance status, age, sex, smoking propensity, education level, and household income. As such, these results raise concern that nonwhite patients may not be achieving the full potential for improved outcomes seen in HPV-associated SCCHN in the modern era. Similar concerns exist regarding patients without insurance coverage.
A plausible explanation for these findings is that nonwhite and uninsured patients may have inferior access to care, such as supportive therapies during intensive treatment of SCCHN. Others have hypothesized that worse outcomes in SCCHN may result in part from insufficient access to screening programs that allow detection of cancer at earlier and more treatable stages.19,20 Indeed, earlier work has demonstrated that African Americans are less likely to receive definitive treatment, with a consequent increased risk of CSM.19 Finally, there may be undescribed genomic differences that result in worse outcomes in nonwhite patients with HPV-associated SCCHN. On the other hand, known HPV status may itself be an indicator of access to appropriate cancer care and good healthcare delivery.
Interestingly, despite the known association between sexual activity and HPV infection, married patients in this study were more likely than single patients to present with HPV-positive SCCHN. We hypothesize that these differences may result from HPV coinfection among couples or, alternatively, from uncaptured risk factors, such as smoking and alcohol consumption, that may increase the likelihood of HPV-negative disease in single patients. We also found that single patients with HPV-negative SCCHN had a higher risk of CSM than married patients with HPV-negative SCCHN, which is consistent with findings of prior studies.20–22
Because of the retrospective nature of SEER, with incompletely captured data on smoking habits and the use of county-level indexes to estimate economic and educational factors, caution should be taken when interpreting these results. In particular, given the well-established effect of smoking status on outcome among patients with oropharyngeal cancers, the lack of individual-level patient data on smoking in the SEER database poses an ongoing challenge. Nonetheless, this study represents the first large report describing the effect of socioeconomic factors on HPV-associated SCCHN and CSM for those patients. Prior studies have relied on small case series of patients from academic centers and retrospective reviews of prospective studies, which are not representative of the general population.16,23 By contrast, SEER registries capture a wider cross-section of the population, including a more representative racial and economic balance from both academic and community medical centers in urban and rural areas.17
Conclusions
This study highlights the socioeconomic differences between patients with HPV-positive and HPV-negative SCCHN and identifies striking racial disparities among individuals with HPV-positive oropharyngeal SCCHN. This work suggests a potentially missed opportunity for cure among nonwhite and uninsured populations with HPV-positive oropharyngeal SCCHN and an unmet need for access to high-quality cancer care. Further work is urgently needed to reduce socioeconomic disparities in SCCHN.
Acknowledgments
We would like to thank the Longitudinal Research and Health Outcomes Forum for valuable discussion.
References
- 1.↑
Chaturvedi AK, Anderson WF, Lortet-Tieulent J, et al.. Worldwide trends in incidence rates for oral cavity and oropharyngeal cancers. J Clin Oncol 2013;31:4550–4559.
- 2.↑
Jemal A, Simard EP, Dorell C, et al.. Annual report to the nation on the status of cancer, 1975-2009, featuring the burden and trends in human papillomavirus (HPV)-associated cancers and HPV vaccination coverage levels. J Natl Cancer Inst 2013;105:175–201.
- 3.↑
Gillison ML, Zhang Q, Jordan R, et al.. Tobacco smoking and increased risk of death and progression for patients with p16-positive and p16-negative oropharyngeal cancer. J Clin Oncol 2012;30:2102–2111.
- 4.↑
Wilting SM, Steenbergen RDM. Molecular events leading to HPV-induced high-grade neoplasia. Papillomavirus Res 2016;2:85–88.
- 5.↑
Schache AG, Powell NG, Cuschieri KS, et al.. HPV-related oropharynx cancer in the United Kingdom: an evolution in the understanding of disease etiology. Cancer Res 2016;76:6598–6606.
- 6.↑
Settle K, Posner MR, Schumaker LM, et al.. Racial survival disparity in head and neck cancer results from low prevalence of human papillomavirus infection in black oropharyngeal cancer patients. Cancer Prev Res (Phila) 2009;2:776–781.
- 7.↑
Fakhry C, Westra WH, Wang SJ, et al.. The prognostic role of sex, race, and human papillomavirus in oropharyngeal and nonoropharyngeal head and neck squamous cell cancer. Cancer 2017;123:1566–1575.
- 8.↑
Chaturvedi AK, Engels EA, Anderson WF, et al.. Incidence trends for human papillomavirus-related and -unrelated oral squamous cell carcinomas in the United States. J Clin Oncol 2008;26:612–619.
- 9.↑
D’Souza G, Kreimer AR, Viscidi R, et al.. Case-control study of human papillomavirus and oropharyngeal cancer. N Engl J Med 2007;356:1944–1956.
- 10.↑
Gillison ML, D’Souza G, Westra W, et al.. Distinct risk factor profiles for human papillomavirus type 16-positive and human papillomavirus type 16-negative head and neck cancers. J Natl Cancer Inst 2008;100:407–420.
- 11.↑
Sinha P, Logan HL, Mendenhall WM. Human papillomavirus, smoking, and head and neck cancer. Am J Otolaryngol 2012;33:130–136.
- 12.↑
Frisch M, Goodman MT. Human papillomavirus-associated carcinomas in Hawaii and the mainland U.S. Cancer 2000;88:1464–1469.
- 13.↑
Stenmark MH, Shumway D, Guo C, et al.. Influence of human papillomavirus on the clinical presentation of oropharyngeal carcinoma in the United States. Laryngoscope 2017;127:2270–2278.
- 14.↑
Worsham MJ, Stephen JK, Chen KM, et al.. Improved survival with HPV among African Americans with oropharyngeal cancer. Clin Cancer Res 2013;19:2486–2492.
- 15.↑
Rubin SJ, Kirke DN, Ezzat WH, et al.. Marital status as a predictor of survival in patients with human papilloma virus-positive oropharyngeal cancer. Am J Otolaryngol 2017;38:654–659.
- 16.↑
Murthy VH, Krumholz HM, Gross CP. Participation in cancer clinical trials: race-, sex-, and age-based disparities. JAMA 2004;291:2720–2726.
- 17.↑
National Cancer Institute; Surveillance, Epidemiology, and End Results Program. Head and Neck with HPV Status Database (2010–2016). Available at: https://seer.cancer.gov/seerstat/databases/hpv/index.html. Accessed September 18, 2019.
- 18.↑
Chu KP, Habbous S, Kuang Q, et al.. Socioeconomic status, human papillomavirus, and overall survival in head and neck squamous cell carcinomas in Toronto, Canada. Cancer Epidemiol 2016;40:102–112.
- 19.↑
Mahal BA, Inverso G, Aizer AA, et al.. Impact of African-American race on presentation, treatment, and survival of head and neck cancer. Oral Oncol 2014;50:1177–1181.
- 20.↑
Inverso G, Mahal BA, Aizer AA, et al.. Marital status and head and neck cancer outcomes. Cancer 2015;121:1273–1278.
- 21.↑
Aizer AA, Chen MH, McCarthy EP, et al.. Marital status and survival in patients with cancer. J Clin Oncol 2013;31:3869–3876.
- 22.↑
Schaefer EW, Wilson MZ, Goldenberg D, et al.. Effect of marriage on outcomes for elderly patients with head and neck cancer. Head Neck 2015;37:735–742.
- 23.↑
Posner MR, Lorch JH, Goloubeva O, et al.. Survival and human papillomavirus in oropharynx cancer in TAX 324: a subset analysis from an international phase III trial. Ann Oncol 2011;22:1071–1077.
