HPV indicates human papillomavirus.
eFigure. Study design for case-cohort study examining chronic sinusitis and cancer risk
eTable. Associations of medical conditions and nasal endoscopy with select HNCs
Customize your JAMA Network experience by selecting one or more topics from the list below.
Beachler DC, Engels EA. Chronic Sinusitis and Risk of Head and Neck Cancer in the US Elderly Population. JAMA Otolaryngol Head Neck Surg. 2017;143(1):25–31. doi:10.1001/jamaoto.2016.2624
Is chronic sinusitis involved in the etiology of head and neck cancer?
This study of SEER-Medicare data found that chronic sinusitis was associated with subsequent nasopharyngeal cancer, human papillomavirus–related oropharyngeal cancer, and nasal cavity and paranasal sinus cancer, but the strength of these associations became attenuated with longer follow-up time.
Sinusitis-related inflammation and/or immunodeficiency may contribute to the etiology of head and neck cancer, but this effect is at most modest.
Chronic sinusitis may be involved in the etiology of certain head and neck cancers (HNCs), due to immunodeficiency or inflammation. However, the risk of specific HNCs among people with chronic sinusitis is largely unknown.
To evaluate the associations of chronic sinusitis with subsequent HNC, including nasopharyngeal cancer (NPC), human papillomavirus–related oropharyngeal cancer (HPV-OPC), and nasal cavity and paranasal sinus cancer (NCPSC), in an elderly US population.
Design, Setting, and Participants
We used the Surveillance, Epidemiology, and End Results (SEER)-Medicare database to conduct a case-cohort study of US individuals aged 65 years or older during 2004 through 2011. The study included 483 546 Medicare beneficiaries from SEER areas in a 5% random subcohort, and 826 436 from the entire source population who developed cancer (including 21 716 with HNC).
Main Outcomes and Measures
Incidence of HNCs including NPC, HPV-OPC, and NCPSC.
Most individuals were female (57.7%), and the mean (SD) age at entry was 72.6 (8.0) years. Chronic sinusitis was associated with risk of developing HNC (adjusted hazard ratio [aHR], 1.37; 95% CI, 1.27-1.48), particularly NPC (aHR, 3.71; 95% CI, 2.75-5.02), HPV-OPC (aHR, 1.33; 95% CI, 1.13-1.57), and NCPSC (aHR, 5.49; 95% CI, 4.56-6.62). Most of this increased risk was limited to risk within 1 year of the chronic sinusitis diagnosis, as associations were largely attenuated 1 year or more after chronic sinusitis (NPC: aHR, 1.60; 95% CI, 0.96-2.65; HPV-OPC: aHR, 1.07; 95% CI, 0.86-1.32; NCPSC: aHR, 2.47; 95% CI, 1.84-3.31). All 3 HNC subtypes had cumulative incidence of less than 0.07% 8 years after chronic sinusitis diagnosis.
Conclusions and Relevance
Chronic sinusitis is associated with certain HNCs, particularly NPC and NCPSC. These HNCs are rare, and most of the increased HNC risk is limited to within 1 year of chronic sinusitis diagnosis, consistent with surveillance or detection bias. The associations were weaker over longer intervals, suggesting at most a modest role for sinusitis-related inflammation and/or immunodeficiency.
Acute sinusitis is a common inflammatory condition often caused by viral or bacterial infections. The condition is considered chronic when the episode persists longer than 12 weeks. Chronic sinusitis can be attributed to various infectious or noninfectious agents.1 Individuals with chronic sinusitis may have an increased risk of head and neck cancer (HNC) due to local inflammatory damage caused by sinusitis. Alternatively, sinusitis may be a marker of a local immunodeficiency. Thus, whereas the infectious agents that cause sinusitis are unlikely to directly induce cell transformation, individuals with chronic sinusitis may be prone to other chronic oncogenic viruses. For example, sinusitis may mark an increased susceptibility to Epstein-Barr virus (EBV), which can cause nasopharyngeal cancer (NPC); and oral human papillomavirus (HPV), which can cause oropharyngeal cancer (HPV-OPC).
Whereas the risk of most cancer types among people with chronic sinusitis has not been extensively examined, 2 recent studies from Taiwan observed a more than 2-fold higher risk of NPC in individuals with chronic sinusitis.2,3 In addition, a recent cross-sectional study found that women with self-reported chronic sinusitis were 3 times more likely to have an oral HPV infection, even after adjusting for tobacco use and sexual behavior.4 To our knowledge, associations between chronic sinusitis and HPV-OPC (or most other cancers) have not been assessed longitudinally. If immunosuppression related to sinusitis plays an etiologic role, HNCs may be expected to occur years after indications of chronic sinusitis, given that cancers develop many years after EBV or HPV acquisition and persistence. Distinguishing between short-term and long-term associations is important because a number of artifacts are likely to induce short-term associations between chronic sinusitis and HNCs, whereas longer-term associations may better reflect the chronic effects of sinusitis related to inflammation or immunosuppression.
Therefore, we examined the risks of HNCs in individuals with and without chronic sinusitis using the Surveillance, Epidemiology, and End Results (SEER)-Medicare linked database.5,6 SEER-Medicare is a large longitudinal database of US individuals older than 65 years, a group of interest because more than 40% of HNC diagnoses occur after age 65 years.7 The inclusion of cancer registry data in SEER-Medicare provides the ability to evaluate uncommon cancer outcomes.
This study used the SEER-Medicare database, which links US Medicare claims with data from SEER cancer registries.6 Medicare is a government insurance program that provides hospital, medical, and surgical benefits for almost all US citizens aged 65 years or older.5 SEER is composed of 18 cancer registries that cover roughly 28% of the US population and has reliable and detailed ascertainment of cancer diagnoses.5,6,8 This study uses deidentified data provided in the National Cancer Institute’s SEER-Medicare database. The National Institute of Health’s Office of Human Subjects Research Protections considers this research not to involve human subjects and that it is therefore exempt from human subjects review.
This study’s source population comprised approximately 9.4 million Medicare beneficiaries aged 65 years or older and living in SEER areas between 2004 and 2011. The SEER-Medicare database includes Medicare claims data for 100% of cancer cases in this source population, identified by SEER, as well as a random 5% sample of the entire source population. We used these data to conduct a case-cohort study,9 which maximally uses the available data as described previously.10,11
Chronic sinusitis was identified using International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) diagnosis codes 473.0 to 473.9 in hospital, physician, and outpatient claims. We required at least 1 hospital or 2 physician or outpatient claims at least 30 days apart to define chronic sinusitis to increase the measure’s positive predictive value. Other procedures or medical conditions of interest were defined using the following ICD-9-CM and Current Procedural Terminology, 4th Edition (CPT-4), codes: nasal/sinus endoscopy (CPT-4: 31231-31297, 31299), otitis media (ICD-9-CM: 381.X, or 382.X), tonsillitis (ICD-9-CM: 463, 474.00, or 474.02), and pharyngitis (ICD-9-CM: 462, 472.1).
Cancers were categorized by anatomic site using the SEER site recode variable. We identified HPV-OPCs by using the following ICD-Oncology (version 3) topographic codes: base of tongue (C01.9), tonsil (C02.4, C09.0-9.9), or oropharynx (C10.0, 10.2-10.9, 14.0, 14.2).12 Recent estimates suggest that roughly 70% of these OPCs are HPV positive, although HPV prevalence may be modestly lower in older adults with OPC.13 We categorized cancers of the nasal cavity, accessory sinuses, and middle ear as nasal cavity and paranasal sinus cancers (NCPSCs: C30.0, 30.1, 31.0-31.9), and NPCs were selected using nasopharynx codes (C11.0-11.9). Based on limited evidence, roughly 80% of NPCs are attributable to EBV in low-incidence areas such as the United States.14 We also examined overall cancer, the 5 most common individual cancers in the United States, and other cancers with an infectious and/or immune-related etiology (cervix, anus, liver, stomach, non-Hodgkin lymphoma [NHL, including lymphocytic leukemia], Hodgkin lymphoma [HL], and Kaposi sarcoma).
Characteristics of individuals in the 5% subcohort, with and without chronic sinusitis during 2004 through 2011, were compared using χ2 tests and t tests. Follow-up time for these analyses started at the latest of the following: an individual’s 65th birthday; January 1, 2004; first date living in a SEER area; start of cancer registry coverage; beginning of Medicare part A and B coverage outside of a health maintenance organization (because health maintenance organizations do not provide claims to Medicare); and the first Medicare claim. Chronic sinusitis was evaluated as a time-dependent exposure, as we updated the chronic sinusitis status if a person developed chronic sinusitis during follow-up. However, once an individual received a diagnosis of chronic sinusitis, he or she was considered “exposed” for the rest of follow-up. Individuals were followed up until they developed a cancer of interest, migrated out of a SEER area, discontinued Medicare Parts A/B coverage or entered a health maintenance organization, died, or survived through the end of follow-up on December 31, 2011.
We considered risk for overall and specific types of HNC, along with cancer overall and other individual cancer types. In our analysis of overall cancer risk, we counted only a person’s first cancer diagnosis during follow-up. People who had a cancer diagnosed prior to the start of follow-up were included in this analysis because they were still at risk for other cancer types. However, in analyses for specific cancer types, individuals with the cancer of interest diagnosed prior to baseline were excluded, and the remaining individuals were followed for each specific cancer type without censoring for other types of cancer.
Weighted Cox proportional hazards regression (SURVEYPHREG procedure, SAS 9.3) was used to estimate hazard ratios (HRs) assessing the association between chronic sinusitis and cancer risk. The weights incorporated the case-cohort sampling fractions (ie, weight = 1 for cancer cases because 100% were sampled, and weight = 20 for cancer-free individuals in the subcohort because 5% were sampled).15 Regression models were adjusted for age at study entry, sex, race, and calendar year at start of follow-up.
The relationship between chronic sinusitis and HNC risk may be prone to surveillance bias (ie, people with sinusitis may be evaluated closely for other related conditions, leading to a diagnosis of cancer), reverse causation (ie, an early HNC could cause sinusitis), or HNC symptoms may be initially misdiagnosed as sinusitis. These potential biases could increase the association of sinusitis and HNC over short time intervals. Therefore, we assessed Cox models in which HRs were estimated for different intervals following chronic sinusitis diagnosis. We categorized time into less than 1 year or 1 year or more after chronic sinusitis diagnosis, and divided follow-up at least 1 year after chronic sinusitis into periods 1 to 2 years, 2 to 3 years, or 3 to 8 years after chronic sinusitis in further analyses. We also conducted sensitivity analyses examining the risk of certain HNCs (NPC, HPV-OPC, and NCPSC) among individuals with other upper airway conditions (otitis media, pharyngitis, and tonsillitis), and among individuals who received a nasal endoscopy at or after their diagnosis of chronic sinusitis.
Cumulative incidence curves were used to depict the absolute risk of subtypes of HNC among individuals with chronic sinusitis, as a function of time since chronic sinusitis diagnosis. The cumulative incidence analyses accounted for competing risks of death,10,16(pp1-8) and all analyses used a 2-sided α of .05 for significance.
There were 483 546 Medicare beneficiaries in our 5% subcohort including 18 759 (3.9%) individuals who developed chronic sinusitis during follow-up. The population was mostly white (81.6%), a majority were female (57.7%), and the mean (SD) age at entry was 72.6 (8.0) years. The individuals with chronic sinusitis in the 5% subcohort were more likely to be younger, female, white, and have other upper airway conditions than individuals who did not receive a diagnosis of chronic sinusitis during follow-up (P < .001) (Table 1). Overall, this study included 826 436 cases of cancer (ie, cases both within and outside the 5% subcohort), including 21 716 HNCs (eFigure in the Supplement).
Risk of HNC was modestly elevated among individuals with prior chronic sinusitis compared with people without chronic sinusitis in both unadjusted and adjusted analyses (adjusted HR [aHR], 1.37; 95% CI, 1.27-1.48) (Table 2). Chronic sinusitis was specifically associated with infection-related HNCs: NPC (aHR, 3.71; 95% CI, 2.75-5.02) and HPV-OPC (aHR, 1.33; 95% CI, 1.13-1.57), and some non–infection-related HNCs, including NCPSC (aHR, 5.49; 95% CI, 4.56-6.62) and larynx (aHR, 1.54; 95% CI, 1.35-1.76). Other HNCs including lip, tongue, and mouth cancer were not associated with chronic sinusitis (Table 2).
We further examined 3 HNC types of interest (NPC, HPV-OPC, and NCPSC) given our a priori hypotheses and/or their strong associations in adjusted analyses (Table 3). The associations with these 3 cancers were largely driven by associations within 1 year following chronic sinusitis diagnosis (Table 3). Specifically, incidence of both NPC and HPV-OPCs were significantly increased in the 1-year period after chronic sinusitis diagnosis (NPC: aHR, 8.88; 95% CI, 6.30-12.50; OPC: aHR, 1.99; 95% CI, 1.56-2.53), but not 1 or more years after chronic sinusitis diagnosis (NPC: aHR, 1.60; 95% CI, 0.96-2.65; HPV-OPC: aHR, 1.07; 95% CI, 0.86-1.32). There was also a stronger association for NCPSC when restricted to less than 1 year compared with more than 1 year after chronic sinusitis diagnosis (<1 year: aHR, 12.53; 95% CI, 10.15-15.48; ≥1 year: aHR, 2.47; 95% CI, 1.84-3.31).
The cumulative incidence of NPC, HPV-OPC, and NCPSC among people with chronic sinusitis is shown in the Figure. Within 1 year after chronic sinusitis diagnosis, approximately 0.01% of individuals had a diagnosis of NPC; 0.02%, HPV-OPC; and 0.03%, NCPSC. Thereafter, each of these cancers developed at a rate of less than 0.02% per year, and the cumulative incidence of each was less than 0.07% after 8 years of follow-up (Figure).
In additional analyses, we examined the risk of the 3 selected HNC subtypes among individuals who received a nasal endoscopy at the time of or after their chronic sinusitis diagnosis. Results were similar to the main results, although the associations were particularly stronger for NCPSC (aHR, 11.39; 95% CI, 8.29-15.66) (eTable in the Supplement). With respect to other upper airway conditions, otitis media was significantly associated with an increased risk of HPV-OPC, NCPSC, and particularly NPC (aHR, 11.96; 95% CI, 9.46-15.12), whereas pharyngitis and tonsillitis were significantly associated only with an increased risk of HPV-OPC (eTable in the Supplement).
We also examined associations between chronic sinusitis and other cancers. The risk of any cancer type was 8% higher in people with chronic sinusitis than in individuals without chronic sinusitis (aHR, 1.08; 95% CI, 1.06-1.10). The most common cancers were not strongly associated with chronic sinusitis (aHRs between 1.14 and 0.86) (Table 2), although the positive associations with lung and prostate cancers and the inverse association with colorectal cancer were statistically significant (Table 2). Cancers with an infectious and/or immune-related etiology were not positively associated with chronic sinusitis, except for NHL (aHR, 1.32; 95% CI, 1.26-1.39) and HL (aHR, 1.59; 95% CI, 1.26-2.00).
In this large longitudinal population-based study, chronic sinusitis was strongly associated with certain HNCs, particularly NPC and NCPSC. In addition, this is the first study to suggest that chronic sinusitis is modestly associated with subsequent risk of HPV-OPC. Much of the increased risk for these cancers was limited to a short time interval after chronic sinusitis diagnosis (ie, the elevation was greatest within 1 year following the diagnosis of chronic sinusitis). In contrast, the associations over longer intervals were weaker, suggesting that sinusitis-related inflammation and/or immunodeficiency play, at most, a minor role in the development of these cancers.
The HNC type with the strongest association with chronic sinusitis was NCPSC, which is not known to have an infection-related etiology. Notably, we found that most of the increased risk for NCPSC (as well as for NPC and HPV-OPC) appears to occur within 1 year of diagnosis of chronic sinusitis, which may reflect several artifacts. For example, an individual who receives a chronic sinusitis diagnosis may be more likely to undergo medical evaluation of the head and neck, including diagnostic imaging, which would increase the probability of identifying an indolent HNC tumor (surveillance bias). It is also possible that an undiagnosed HNC could be initially misdiagnosed as chronic sinusitis or cause chronic sinusitis (reverse causation). Symptoms such as nasal obstruction and/or congestion, rhinorrhea, and facial pressure commonly occur in both chronic sinusitis and HNC. Patients with HNC might receive a diagnosis of and be treated provisionally for sinusitis prior to undergoing imaging to identify the neoplasm, given the higher incidence of chronic sinusitis. These artifacts would be most likely to explain HNCs diagnosed soon after chronic sinusitis. For NCPSC (but not for the other HNCs), we still observed a significant association up to 3 years after chronic sinusitis diagnosis. One possibility is that chronic sinusitis, especially if present over an extended period, could induce genetic damage to the epithelium lining these spaces, which, when cumulative, could promote carcinogenesis.
Immunodeficiency or inflammation related to sinusitis has been hypothesized to increase infection-related HNC risk. A recent study suggested that oral HPV infection is more common in people with chronic sinusitis.4 Moreover, individuals with chronic sinusitis have an elevated prevalence of humoral immunodeficiency syndromes, such as low IgG or IgA levels,17 and they may respond inadequately to vaccination against pneumococcal infection.18 These observations suggest that chronic sinusitis may be a marker of an immunodeficiency. If so, the potential effect of an immunodeficiency would be expected to become evident years after the diagnosis of chronic sinusitis, particularly for infection-related HNCs, which develop decades after HPV or EBV acquisition and persistence. Because our study found only modest and nonsignificant associations with the infection-related HNCs (HPV-OPC and NPC) over intervals longer than 1 year, our results suggest that sinusitis-related inflammation and immunodeficiency does not play a large role in promoting the early or intermediate carcinogenic stages of infection-related HNC.
The association that we observed between chronic sinusitis and NPC in this study is similar to that reported in 2 recent studies from Taiwan.2,3 However, both Taiwanese studies found an association between chronic sinusitis and NPC over intervals longer than 1 year. Our study suggests that there could be a modestly elevated NPC risk more than 1 year after chronic sinusitis, but the risk was not significant (aHR, 1.60; 95% CI, 0.96-2.65). Some geographic factors could explain these differences. Asian countries such as Taiwan have a higher incidence of NPC than seen in the United States.19 Furthermore, NPCs in Asia may be more strongly linked to EBV infection than NPCs in the United States,14 and chronic sinusitis may have a stronger effect on EBV-related cases.
We found similar associations between chronic sinusitis and the selected HNCs among individuals who received a nasal endoscopy at the time of or after their chronic sinusitis diagnosis. Nasal endoscopies are often performed on patients with persistent chronic sinusitis to confirm their diagnosis, and the presence of claims for this procedure thus should reduce misclassification of chronic sinusitis. Conversely, because nasal endoscopy would facilitate a diagnosis of HNC, the strong associations could reflect the aforementioned surveillance bias.
In addition to chronic sinusitis, we found that other upper airway conditions were associated with elevated risk of HNCs. Otitis media was associated with all 3 examined HNC subtypes, while pharyngitis and tonsillitis were only associated with HPV-OPC. These associations are prone to and may reflect similar biases (such as surveillance bias) as seen for chronic sinusitis. Otitis media in particular was most strongly associated with NPC; this association could be due to reverse causation, as NPCs can impinge on the middle ear.
We also examined the associations of chronic sinusitis with common cancers and other cancers known to have an infection and/or immune-related etiology. Chronic sinusitis was not strongly associated with any common cancer type, although we observed modest positive associations with lung and prostate cancer and an inverse association with colon cancer. We were able to observe these associations given the large sample of this study, and these weak associations could be attributable to multiple factors including confounding by smoking. While chronic sinusitis was not associated with every infection and/or immune-related cancer, it was associated with increased risk of both NHL and HL. This mirrors what has been seen in a previous study20 and could potentially be related to the presence of immunodeficiency prior to lymphoma diagnosis.
The strengths of this study include its large sample size, population-based design, and use of validated cancer outcomes from SEER cancer registries, which have high standards for completeness and accuracy.6 The results of our analyses of Medicare beneficiaries are likely generalizable to the US elderly population. In addition, the longitudinal nature of this study allowed for assessment of HNC risk according to time since chronic sinusitis diagnosis and for calculations of the cumulative incidence of HNC in this population.
We also acknowledge several limitations. First, we lacked data on tobacco use, which may be associated with the risk of developing chronic sinusitis (with relative risks of ~1.1-1.9).21,22 However, chronic sinusitis was not strongly associated with tobacco-related cancers (eg, lung and bladder cancers) in our study, suggesting that the associations with HNCs were likely not substantially confounded by unmeasured tobacco use. We were also limited by the lack of data on the duration of chronic sinusitis and on the viral status of HNCs. We considered all cases of HPV-OPC and NPC to be virus related, but assuming that chronic sinusitis may play a role only in cases actually caused by HPV or EBV, our associations could be attenuated. Because OPC cases linked to HPV tend to occur in younger individuals than HPV-unrelated cases, the lack of middle-aged adults in our study is a limitation.23 Also, we could not assess the association of chronic sinusitis with inverted papilloma of the nose and paranasal sinuses, which also occurs more commonly at younger ages.24
Despite the fact that people with chronic sinusitis have an increased risk for certain subtypes of HNCs, the absolute risk of these cancers is low. The cumulative incidence of NPC, HPV-OPC, and NCPSC was less than 0.10% after 8 years of follow-up after a chronic sinusitis diagnosis. There are currently no general US guidelines for HNC screening, but given the low absolute risk, our findings do not support a need for HNC screening in individuals with chronic sinusitis.
This is one of the first studies to demonstrate associations of chronic sinusitis with HPV-OPC, NPC, and NCPSCs. The HNC risks within 1 year of chronic sinusitis diagnosis support that these associations can be explained in large part by artifacts such as surveillance bias, reverse causation, or diagnostic confusion. Whereas an etiologic contribution of sinusitis-related inflammation or immunodeficiency to HNC cannot be excluded, this study suggests that they do not play a large role in promoting the early or intermediate carcinogenic stages of HNC.
Accepted for Publication: July 13, 2016.
Corresponding Author: Daniel C. Beachler, PhD, MHS, Infections and Immunoepidemiology Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute, 9609 Medical Center Dr, Room 6-E220, Bethesda, MD 20892 (email@example.com).
Published Online: September 8, 2016. doi:10.1001/jamaoto.2016.2624
Author Contributions: Dr Beachler had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Both authors.
Acquisition, analysis, or interpretation of data: Both authors.
Drafting of the manuscript: Beachler.
Critical revision of the manuscript for important intellectual content: Both authors.
Statistical analysis: Beachler.
Study supervision: Engels.
Conflict of Interest Disclosures: Both authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.
Funding/Support: This work was supported by the Intramural Research Program of the National Cancer Institute.
Role of the Funder/Sponsor: The National Cancer Institute had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Additional Contributions: The authors acknowledge the efforts of the Applied Research Program, National Cancer Institute; the Office of Research, Development, and Information, Centers for Medicare and Medicaid Services; Information Management Services, Inc; and the Surveillance, Epidemiology, and End Results (SEER) Program tumor registries in the creation of the SEER-Medicare database. Winnie Ricker, BS, Information Management Services, Rockville, Maryland, provided assistance with database management, and Elizabeth Yanik, PhD, National Cancer Institute, provided programming assistance particularly with the development of the figures. They were not compensated beyond their salaries for their contributions.
Create a personal account or sign in to: