- Email: [email protected]
Rising incidence of oral tongue cancer among white men and women in the United States, 1973–2012
Oral Oncology 67 (2017) 146–152
Contents lists available at ScienceDirect
Oral Oncology journal homepage: www.elsevier.com/locate/oraloncology
Rising incidence of oral tongue cancer among white men and women in the United States, 1973–2012 Joseph E. Tota a,⇑, William F. Anderson a, Charles Coffey b, Joseph Califano b, Wendy Cozen c, Robert L. Ferris d, Maie St. John e, Ezra E.W. Cohen b, Anil K. Chaturvedi a a
National Cancer Institute, Division of Cancer Epidemiology and Genetics, Rockville, MD, USA UC San Diego, Moores Cancer Center, La Jolla, CA, USA University of Southern California, Departments of Preventive Medicine and Pathology, Los Angeles, CA, USA d University of Pittsburgh, Department of Otolaryngology, Pittsburgh, PA, USA e UCLA, Department of Head and Neck Surgery, UCLA Head and Neck Cancer Program, Los Angeles, CA, USA b c
a r t i c l e
i n f o
Article history: Received 23 January 2017 Received in revised form 18 February 2017 Accepted 20 February 2017
Keywords: Oral tongue cancer Incidence Temporal trends Age-period-cohort analysis
a b s t r a c t Background: Despite significant reductions in tobacco use in the US, oral tongue cancer incidence has reportedly increased in recent years, particularly in young white women. We conducted age-periodcohort analyses to identify birth cohorts that have experienced increased oral tongue cancer incidence, and compared these with trends for oropharyngeal cancer, a cancer caused by human papillomavirus (HPV) that has also recently increased. Methods: We utilized cancer incidence data (1973–2012) from 18 registries maintained by the NCI SEER Program. Incidence trends were evaluated using log-linear joinpoint regression and age-period-cohort modeling was utilized to simultaneously evaluate effects of age, calendar year, and birth year on incidence trends. Results: Incidence of oral tongue cancer increased significantly among white women during 1973–2012 (0.6% annual increase, p < 0.001) and white men during 2008–2012 (5.1% annual increase, p = 0.004). The increase was most apparent among younger white individuals (<50 years; annual increase of 0.7% for men [p = 0.02] and 1.7% for women [p < 0.001] during 1973–2012). Furthermore, the magnitude of the increase during 1973–2012 was similar between young white men and women (2.3 vs. 1.8 cases per million, respectively). Incidence trends for oropharyngeal cancer were similar to trends for oral tongue cancer and similar birth cohorts (born after the 1940s) experienced rising incidence of these cancers (p-value: white men = 0.12, white women = 0.42), although the magnitude of increase was greater for oropharyngeal cancer. Conclusions: The incidence of oral tongue and oropharyngeal cancer has significantly increased among young white men and women within the same birth cohorts in the US. Published by Elsevier Ltd.
Introduction Cancers of the oral cavity and pharynx have traditionally been considered a single etiologic entity. However, recent studies show significant differences in the etiology and epidemiology for individual anatomic sites [1–3]. Reduction in smoking in the United States has resulted in significant declines in the incidence of most oral cavity cancers, including lip, gum, floor of mouth, hard palate, buccal mucosa, and vestibule cancers [4–8]. In contrast, the ⇑ Corresponding author at: National Cancer Institute, Division of Cancer Epidemiology and Genetics, Infections and Immunoepidemiology Branch, 9609 Medical Center Drive, Room 6E220, Rockville, MD 20850, USA. E-mail address: [email protected] (J.E. Tota). http://dx.doi.org/10.1016/j.oraloncology.2017.02.019 1368-8375/Published by Elsevier Ltd.
incidence of oral tongue cancers (anterior 2/3 of the tongue) has reportedly increased in recent years [7–12]. The incidence of oropharyngeal cancers has also increased, and descriptive and molecular epidemiologic studies have identified human papillomavirus (HPV) infection as the cause [2,3]. However, the reasons for the increase in oral tongue cancers are unknown, and molecular studies indicate that HPV does not play a major etiologic role [13–16]. Descriptive studies of oral tongue cancer trends have reported that incidence has significantly increased among young (ages 18–44 years), white individuals, and primarily in women [7,8,10– 12]. It is, however, unclear whether oral tongue cancer incidence has increased in similar magnitude among young, white men [7–12]. Furthermore, prior studies have not identified the specific
147
J.E. Tota et al. / Oral Oncology 67 (2017) 146–152
birth cohorts that have experienced an increase in oral tongue cancer incidence, which could provide important etiologic clues. For example, the identification of birth cohorts of men that went through the sexual revolution as the primary demographic subgroup that has experienced the rise in oropharyngeal cancer incidence, in part, enabled the identification of HPV as the cause [2,3,5,17,18]. The main objective of this study was to identify the demographic subgroups and birth cohorts that have experienced an increase in oral tongue cancer in the United States. We also aimed to compare and contrast our findings for oral tongue cancer with oropharyngeal cancer to determine if similar birth cohorts have experienced an increase in these cancers. This comparison may provide clues regarding the possible role of a sexually transmitted infection or other environmental exposure in the development of oral tongue cancer. Methods Data source Cancer incidence information from the US National Cancer Institute’s Surveillance, Epidemiology, and End Results (SEER) program database were obtained for the years 1973–2012 [19–21]. We included data from the SEER 9 Registries Database [19] for the years 1973–1991 (Atlanta, Connecticut, Detroit, Hawaii, Iowa, New Mexico, San Francisco-Oakland, Seattle-Puget Sound, and Utah), SEER 13 Registries Database [20] for the years 1992–1999 (SEER 9 regions plus Los Angeles, San-Jose Monterey, Rural Georgia, and Alaska Native Tumor Registry), and the SEER 18 Registries Database [21] for the years 2000–2012 (SEER 13 regions plus Greater California, Kentucky, Louisiana, New Jersey, and Greater Georgia), covering 14%, 17%, and 28% of the US population, respectively. Classification of anatomic sites For all analyses, oral cavity cancer sites were subdivided into two groups: oral tongue cancer, including the dorsal surface, border, ventral surface, and anterior 2/3 of the tongue (International Classification of Diseases for Oncology version-3 [ICD-O-3] topography codes C020-023) and other oral cavity cancer sites, including lip, gum, floor of mouth, and other/unspecified parts of the mouth (ICD-O-3 codes C000-009, C030-039, C040-050, and C060-069). HPV-associated oropharyngeal cancers were also evaluated, and these sites included base of tongue, lingual tonsil, soft palate and uvula, tonsil, oropharynx, and Waldeyer ring (ICD-O-3 codes C019, C024, C051-052, C090-099, C100-109, and C142) [5]. Cancers classified as tongue, not otherwise specified (ICD-O-3 code: C028) were excluded due to potential ambiguity regarding origin within the oral tongue vs. base of tongue. All histologic subtypes were included in our primary analysis for each cancer site. This study did not use personal identifying information from the SEER data or involve interaction with human subjects, therefore informed consent and institutional review was not required. Statistical analysis All analyses were conducted separately by sex and race: white (both non-Hispanic and Hispanic), black, and other races (American Indians and Alaska Natives, Asian/Pacific Islanders, and unknown). For each site, annual incidence rates (1973–2012; age-standardized to the US 2000 standard population) were obtained using SEER⁄Stat version 8.2.1 software. Temporal trends were evaluated using log-linear joinpoint regression, incorporating cubic regression splines selected based on the Akaike information
criterion. Trends are expressed as the annual percentage change (APC) in incidence [22]. For years with zero cases for a particular stratum, the annual incidence rate and standard error were taken from the preceding year, or closest year (preceding or following) when there were consecutive years with zero cases. In addition to presenting the APC (a measure of relative change in incidence), we also calculated the incidence rate difference (a measure of the absolute change in incidence) between 1973 and 2012, and difference in the average incidence rate between late (1973–77) and recent (2008–12) time periods. To simultaneously evaluate the effects of age, calendar year/ period, and birth year/cohort on incidence rates, age-periodcohort modeling was applied using 5-year age groups (20 to 24, . . ., 80 to 84), 5-year calendar periods (1973 to 1977, . . ., 2008 to 2012), and 5-year birth cohorts (1893, . . ., 1988). These models may be useful for detecting changes in exposure experiences (indicated by cohort effects), or changes in reporting, coding, screening and/or diagnostic practices (indicated by period effects) [23–25]. We specifically focused on two aspects of birth cohort effects in these models – cohort deviations and age-specific temporal trends (or net drifts) – that can enable identification of birth cohorts and age groups that have experienced significant changes in incidence rates. Differences in age-specific net drifts and cohort deviations were compared across cancer sites for the same sex (oral tongue cancer vs. oropharyngeal cancer and oral tongue cancer vs. other oral cavity cancer) and across sex for the same site using a 1-df Wald test. For oral tongue cancer and oropharyngeal cancer, temporal trends were then reevaluated focusing on young and older age groups separately, with the age cut point identified based on the observed age-specific net drifts. Given the relative rarity of oral tongue cancers, we combined the three SEER registries for enhanced statistical power. Nonetheless, to ensure robustness of our results as well as potential comparability of incidence rates across the SEER databases, we conducted two sensitivity analyses. First, we reevaluated all incidence trends using SEER 9 registries. Second, for overlapping calendar years (2000–2012), we compared incidence trends for oral tongue and oropharyngeal cancers across SEER 9, 13, and 18. We also conducted sensitivity analyses for oral tongue cancers restricted to squamous cell histologies.
Table 1 Characteristics of cases of oral tongue, oropharyngeal, and other oral cavity cancers during 1973–2012. Patient characteristic
Oral tongue (n = 16,206) no. (%)
Age at diagnosis, yearsa Mean 60.2 Standard 13.5 deviation
Oropharyngeal (n = 67,789) no. (%)
Other oral cavity (n = 56,168) no. (%)
60.0 11.5
63.0 13.1
Sex Male Female
9483 (58.5) 6723 (41.5)
50,748 (74.9) 17,041 (25.1)
37,074 (66.0) 19,094 (34.0)
Race/ethnicity White Black Other Unknown
13,772 (85.0) 1014 (6.2) 1292 (8.0) 128 (0.8)
56,977 (84.0) 7858 (11.6) 2611 (3.9) 343 (0.5)
49,556 (88.2) 3867 (6.9) 2117 (3.8) 628 (1.1)
Year of diagnosis 1973–1982 1983–1992 1993–2002 2003–2012
1742 2043 4400 8021
5906 (8.7) 7571 (11.2) 16,363 (24.1) 37,949 (56.0)
9360 (16.7) 9944 (17.7) 15,170 (27.0) 21,694 (38.6)
(10.7) (12.6) (27.2) (50.5)
a Restricted to individuals aged 0–84 years at time of cancer diagnosis (oral tongue: n = 15,350; oropharyngeal: n = 65,923; other oral cavity: n = 51,524).
148
J.E. Tota et al. / Oral Oncology 67 (2017) 146–152
Fig. 1. Incidence trends for oral tongue, oropharyngeal, and other oral cavity cancers in whites, stratified by gender (1973–2012). The two panels show a significant increase in the incidence of oral tongue and oropharyngeal cancer, and a significant decrease in other oral cavity cancers among both men and women (panels A and B, respectively). APC, annual percent change; OCC, oral cavity cancer; OPC, oropharyngeal cancer; OTC, oral tongue cancer. *Joinpoint regression comparison P < 0.05.
Results During 1973–2012, 16,206 oral tongue cancers, 67,789 oropharyngeal cancers, and 56,168 other oral cavity cancers occurred (Table 1). The mean age of patients diagnosed with oral tongue and oropharyngeal cancer was slightly lower (60.2 and 60.0 years, respectively) compared with patients diagnosed with other oral cavity cancer (63.0 years). The majority of cases across sites occurred among males (58.5–74.9%) and white individuals (84.0– 88.2%); however, females and ‘‘other” race/ethnic groups
accounted for a relatively high proportion of oral tongue cancers (41.5% and 8.0%, respectively). Due to the introduction of new registries being added to the SEER database, the proportion of cases across sites was higher in more recent decades. Overall incidence trends (1973–2012) The incidence of other oral cavity cancers significantly declined in recent years among most race and gender subgroups (Fig. 1A and B, Supplementary Fig. 1 A–D). In contrast, the
Fig. 2. Age-specific temporal trends (net drifts) in incidence of oral tongue, oropharyngeal, and other oral cavity cancers among white men and women (1973–2012). Wald P values are presented for evaluating differences in age-specific net drifts and cohort deviations across cancer sites (men and women separately), and across sex for the same cancer site. In panels A (men) and B (women), the overall increase in oral tongue cancer is shown to be heterogeneous by age (most apparent in younger individuals; p < 0.001 for both men and women), with a stronger increase in women than in men (p = 0.007). OCC = oral cavity cancer; OPC = oropharyngeal cancer; OTC = oral tongue cancer.
J.E. Tota et al. / Oral Oncology 67 (2017) 146–152
incidence of oral tongue cancers significantly increased among white women during 1973–2012 (APC = 0.6%, p < 0.001) as well as among white men during 2008 to 2012 (APC = 5.1%, p = 0.004) (Fig. 1A and B). The magnitude of the absolute increase in oral tongue cancer incidence during 1973–2012 was similar between white men and women (an increase of 2.5 vs. 2.6 cases per million). No increase in oral tongue cancer incidence was observed among men and women of other racial/ethnic groups (Supplementary Fig. 1 A–D). Results were similar when analyses were stratified by anatomic subsite (border of tongue, dorsal surface of tongue, and ventral surface of the tongue; data not shown). Significant increases in the incidence of oropharyngeal cancer were observed among both white men (APC = 0.8% during 1973–
149
1998, and 3.8% during 1998–2009; Fig. 1A) and white women (APC = 0.7% during 1995–2012; Fig. 1B). However, the magnitude of the increase (absolute change in incidence) for oropharyngeal cancer during 1973–2012 was substantially greater among white men than among white women (35.3 vs. 0.4 cases per million increase). Significant increases in oropharyngeal cancer were also observed among black men (APC = 4.2% during 1973–1984), black women (APC = 6.5% during 1973–1982), and men of other races (APC = 1.0% during 1973–2012). Because only whites experienced a statistically significant increase in both the incidence of oral tongue and oropharyngeal cancers, subsequent analyses were restricted to whites. However, Results from additional age-stratified comparisons for oral tongue,
Fig. 3. Incidence trends for oral tongue cancer and oropharyngeal cancer among white men and women, stratified by age (1973–2012). Panels A and B show a higher relative increase in oral tongue cancer among young women (reflected by higher APC) but a similar absolute increase in young men and women. Among older men and women, significant increases in oral tongue cancer are restricted to the most recent calendar years. Panels C and D show that both the relative and absolute increase in oropharyngeal cancer is higher among men (both young and older age groups), and that among older individuals, a significant increase occurred only in men and only in the most recent calendar years (1997–2012). *Joinpoint regression comparison P < 0.05. APC = annual percent change; OPC = oropharyngeal cancer.
150
J.E. Tota et al. / Oral Oncology 67 (2017) 146–152
oropharyngeal, and other oral cavity cancers for blacks and other races are provided in the Supplement (Supplementary Table 1). Age-specific incidence trends among white men and women (1973– 2012) The overall increase in oral tongue cancer incidence observed among white men and white women was heterogeneous by age, with the increases being most apparent among younger individuals (<50 years of age) (Fig. 2A and B). Additionally, the pattern of increase in young individuals was stronger in women than men (Wald p-value = 0.007). When stratified by age (<50 years and 50 years), oral tongue cancer incidence significantly increased during 1973–2012 in white men and white women ages <50 years (APC = 0.7% among men [p = 0.02] and APC = 1.7% among women [p < 0.001]) (Fig. 3A and B). Consistent with significant gender differences in age-specific net drifts (Fig. 2), the estimated APC for oral tongue cancer was higher among young white women compared with white men. However, the magnitude of the absolute increase during 1973–2012 was similar in men and women (2.3 vs. 1.8 cases per million increase; Fig. 3A and B). Among older white men and women (50 years of age), significant increases in oral tongue cancer were restricted to the most recent calendar years (APC = 5.1% during 2008–2012 among men [p = 0.005] and APC = 2.9% during 2005–2012 among women [p = 0.005]) (Fig. 3A and B). In individuals younger than 50 years, oropharyngeal cancer incidence increased significantly during 1973–2004 among white
men and during 1973–2012 among white women (Fig. 3C and D). Both the relative increase and the absolute increase in oropharyngeal cancer incidence during 1973–2012 was greater among young men than young women (APC = 3.2% vs. 0.7% and absolute increase = 10.6 vs. 0.0 cases per million increase). In individuals 50 years or older, oropharyngeal cancer incidence significantly increased among white men during 1997–2012 (3.7% increase per year) whereas no significant increase (or decrease) in incidence was observed among white women (Fig. 3C and D). Age-specific incidence rates by birth cohort To identify the specific birth cohorts that have experienced rising incidence, we conducted age-period-cohort analyses comparing cohort deviations between oral tongue and oropharyngeal cancer sites by gender, and cohort deviations across gender by cancer site. Similar patterns were observed in the age-specific incidence of oral tongue cancer and oropharyngeal cancer by birth cohort, with higher incidence rates among more recent birth cohorts of white men and women (Fig. 4A–D). Specifically, the same birth cohorts of white men and women (cohorts born after the 1940s) experienced rising incidence of both oral tongue and oropharyngeal cancers (p-value for birth cohort comparisons for oral tongue vs. oropharynx for white men = 0.12 and white women = 0.42). Across gender, the same birth cohorts of men and women experienced rising incidence of oral tongue cancers (pvalue for birth cohort comparisons in men vs. women for oral tongue cancers = 0.41). However, the birth cohort increases for
Fig. 4. Age-specific incidence rates for oral tongue cancers and oropharyngeal cancers among white men and women across birth cohort. Focusing on men and women separately, similar birth cohorts (born after the 1940s) experienced rising incidence of oral tongue and oropharyngeal cancers (men, panels A and C, p-value for birth cohort comparisons = 0.12; women, panels B and D, p-value = 0.42) and while birth cohort increases for oral tongue cancer were not significantly different across gender (p-value = 0.41), for oropharyngeal cancer, the magnitude and duration of increase was significantly greater among men (p < 0.001). OPC = oropharyngeal cancer.
J.E. Tota et al. / Oral Oncology 67 (2017) 146–152
oropharyngeal cancer were longer in duration and stronger in magnitude among men than among women (p < 0.001). In sensitivity analyses, the key observations of rising oral tongue and oropharyngeal cancer incidence during 1973–2012 in young, white men and women were qualitatively similar in analyses restricted to SEER 9 registries (Supplementary Table 2). Likewise, during the years 2000–2012, incidence trends for oral tongue and oropharyngeal cancer were similar across SEER 9, SEER 13, and SEER 18 registries (Supplementary Fig. 1). Results were not meaningfully different when comparing single-year and 5-year average incidence rate differences (Supplementary Table 3). Results were also similar in analyses restricted to squamous cell histologies (data not shown).
Discussion In this analysis of US cancer registry data spanning a period of four decades, we observed significant increases in oral tongue cancer incidence among white men and women, but not men and women of other racial/ethnic groups. As reported in numerous recent studies, the incidence of oropharyngeal cancers also increased significantly in recent decades in white men. Further, age-period-cohort analyses indicated that similar recent birth cohorts of white men and women have experienced a rise in the incidence of both oral tongue and oropharyngeal cancers in recent decades in the United States. Nonetheless, the magnitude of increase was appreciably greater for oropharyngeal cancers when compared to tongue cancers. Our finding of rising incidence for oral tongue cancers in young white individuals are similar to prior studies within the US as well as in Nordic countries [7–12,26,27]. However, based on differences by gender in annual percent change, most have suggested that the increase has primarily occurred among women [7,8,12]. In contrast, we show that the increase in oral tongue cancer incidence has occurred in both men and women to a similar extent, as measured by the number of cases. Further, the observation that similar birth cohorts of men and women have experienced the rise in oral tongue cancers suggests a similar etiologic exposure in both genders. Several observations argue against artefactual explanations or period effects for the rising tongue cancer trends, including consistent increases in incidence in young individuals over the entire duration since 1973, the percolation of the rising incidence into older individuals in recent years, and the similarity in trends across SEER 9, 13, and 18 registries. Further, there has not been an increase in screening for oral cavity cancers or oral tongue cancers in recent years, which would manifest as period effects. These observations collectively suggest birth cohort effects of changing exposure in white individuals. Tobacco and alcohol are unlikely to explain this rise in tongue cancer incidence given recent declines in smoking and per-capita alcohol consumption in the US population [4–8,28]. Likewise, the prevalence of chewing tobacco, another strong risk factor for oral tongue cancers as well as other oral cavity cancers, has remained historically low in the United States [29]. Indeed consistent with these risk factor trends, the incidence of most oral cavity cancers has significantly declined in recent years. The reasons for the increase in oral tongue cancer incidence are currently unclear, and as noted herein, most studies to date have been negative. Our observation that similar birth cohorts have experienced an increase in oral tongue and oropharyngeal cancers could suggest a potential role for HPV infection. However, recent molecular studies indicate that oncogenic alpha-HPV types (such as HPV16), the principal cause of oropharyngeal cancers, do not play a major etiologic role in oral tongue cancers [13–16].
151
Nonetheless, a recent US study reported that certain beta- and gamma-HPV types (such as HPV b-5) are associated with increased risk of head and neck cancers, including oral cavity cancers [30]. It is as yet unclear whether this association is specific to cancers of the oral tongue in the presence/absence of smoking, and if the prevalence of beta- and gamma-HPV types has increased in recent birth cohorts. The similar birth cohort experience for oral tongue and oropharyngeal cancers could also point to another environmental exposure or sexually transmitted virus/infection. Marijuana use, a suspected risk factor for oral cavity cancers, has significantly increased in recent birth cohorts. Nevertheless, a recent analytic study within the International Head and Neck Cancer Epidemiology Consortium reported that marijuana use was associated with a paradoxical decrease in risk of oral tongue cancers [31]. Further, two recent US studies that evaluated tumor specimens from young non-smoking oral tongue cancer patients (n = 7 and n = 53) failed to find evidence of viral involvement using next-generation exome or RNA sequencing methods [16,32]. Based on genomic similarity between oral tongue tumors obtained from young non-smokers and older smoking individuals, it has also recently been suggested that another as yet unidentified risk factor that induces inflammation or that disrupts the oral microbiome could increase oral tongue cancer risk [33]. Recently, Brouwer and colleagues utilized multistage clonal expansion models in combination with ageperiod-cohort models to estimate tumor initiation, promotion, and malignant conversion rates for different head and neck cancer subsites [34]. Promotion rates for oral tongue cancer were lower in comparison with other sites (other tobacco/alcohol-related oral cavity cancers and HPV-related oropharyngeal cancers), suggestive of longer sojourn time and a unique etiology [34]. We acknowledge both strengths and limitations of our study. The combination of multiple SEER databases provided us with an adequate sample to perform age-period-cohort modeling to compare cohort effects across gender and related cancer sites. However, we had limited sample sizes (power) for some subgroup analyses, which may explain the lack of statistically significant trends for most cancers among non-whites. Prior studies have arbitrarily selected age cut-points for defining ‘young’ versus ‘older’ individuals in making their comparisons of oral tongue cancer trends, whereas we evaluated age-specific net drifts to identify and empirically define our cut-point. Importantly, given the registry-based nature of our analyses, we did not have information on traditional risk factors, such as smoking and alcohol. We thus relied on ecologic comparisons across cancer sites to rule out the potential effect of these risk factors on oral tongue cancer incidence. There is a possibility that misclassification of HPV-related base of tongue cancers as anterior tongue cancers could have partly contributed to the rise in incidence of oral tongue cancers, but it is unlikely that such misclassification is restricted only to certain age and racial subgroups (e.g. young whites). To avoid such misclassification, we excluded overlapping lesions of the tongue as well as those cancers without anatomic tongue site specification. Further, the increase in oral tongue cancer incidence has occurred in both men and women, while the rise in HPV-related base of tongue cancers has primarily occurred in men, which also argues against misclassification of base of tongue cancers as a potential explanation. The risk factors that are contributing to the rise in oral tongue cancer cases observed among young white men and women in the United States are not currently known. Thus, additional studies are needed to generate and confirm hypotheses concerning risk factors for oral tongue cancers in the presence and absence of the traditional risk factors, such as smoking. These studies could potentially focus on suspected or emerging oral cancer risk factors whose prevalence may have increased in cohorts of white individuals born after the 1940s.
152
J.E. Tota et al. / Oral Oncology 67 (2017) 146–152
Conflict of interest Ezra E. W. Cohen reports consulting fees from AstraZeneca, Bristol-Myers Squibb, EMD Serono, Human Longevity Inc., Merck, and Pfizer outside the submitted work. The other authors do not have any conflict of interest to report.
Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.oraloncology. 2017.02.019.
References [1] Mork J, Lie AK, Glattre E, Hallmans G, Jellum E, Koskela P, et al. Human papillomavirus infection as a risk factor for squamous-cell carcinoma of the head and neck. N Engl J Med 2001;344:1125–31. [2] Chaturvedi AK, Anderson WF, Lortet-Tieulent J, Curado MP, Ferlay J, Franceschi S, et al. Worldwide trends in incidence rates for oral cavity and oropharyngeal cancers. J Clin Oncol: Off J Am Soc Clin Oncol 2013;31:4550–9. [3] Gillison ML, Chaturvedi AK, Anderson WF, Fakhry C. Epidemiology of human papillomavirus-positive head and neck squamous cell carcinoma. J Clin Oncol: Off J Am Soc Clin Oncol 2015;33:3235–42. [4] CDC. Cigarette smoking among adults and trends in smoking cessation – United States, 2008. MMWR Morb Mortal Wkly Rep. 2009; 58: 1227–32. [5] Chaturvedi AK, Engels EA, Anderson WF, Gillison ML. Incidence trends for human papillomavirus-related and -unrelated oral squamous cell carcinomas in the United States. J Clin Oncol: Off J Am Soc Clin Oncol 2008;26:612–9. [6] Shiboski CH, Shiboski SC, Silverman Jr S. Trends in oral cancer rates in the United States, 1973–1996. Commun Dent Oral Epidemiol 2000;28:249–56. [7] Patel SC, Carpenter WR, Tyree S, Couch ME, Weissler M, Hackman T, et al. Increasing incidence of oral tongue squamous cell carcinoma in young white women, age 18 to 44 years. J Clin Oncol: Off J Am Soc Clin Oncol 2011;29:1488–94. [8] Brown LM, Check DP, Devesa SS. Oral cavity and pharynx cancer incidence trends by subsite in the United States: changing gender patterns. J Oncol 2012;2012:649498. [9] Shemen LJ, Klotz J, Schottenfeld D, Strong EW. Increase of tongue cancer in young men. JAMA 1984;252:1857. [10] Myers JN, Elkins T, Roberts D, Byers RM. Squamous cell carcinoma of the tongue in young adults: increasing incidence and factors that predict treatment outcomes. Otolaryngol Head Neck Surg 2000;122:44–51. [11] Shiboski CH, Schmidt BL, Jordan RC. Tongue and tonsil carcinoma: increasing trends in the U.S. population ages 20–44 years. Cancer 2005;103:1843–9. [12] Joseph LJ, Goodman M, Higgins K, Pilai R, Ramalingam SS, Magliocca K, et al. Racial disparities in squamous cell carcinoma of the oral tongue among women: a SEER data analysis. Oral Oncol 2015;51:586–92. [13] Herrero R, Castellsague X, Pawlita M, Lissowska J, Kee F, Balaram P, et al. Human papillomavirus and oral cancer: the International Agency for Research on Cancer multicenter study. J Natl Cancer Inst 2003;95:1772–83. [14] Lingen MW, Xiao W, Schmitt A, Jiang B, Pickard R, Kreinbrink P, et al. Low etiologic fraction for high-risk human papillomavirus in oral cavity squamous cell carcinomas. Oral Oncol 2013;49:1–8. [15] Kruger M, Pabst AM, Walter C, Sagheb K, Gunther C, Blatt S, et al. The prevalence of human papilloma virus (HPV) infections in oral squamous cell carcinomas: a retrospective analysis of 88 patients and literature overview. J Cranio-Maxillo-Facial Surgery: Off Publ Eur Assoc Cranio-Maxillo-Facial Surgery 2014;42:1506–14.
[16] Bragelmann J, Dagogo-Jack I, El Dinali M, Stricker T, Brown CD, Zuo Z, et al. Oral cavity tumors in younger patients show a poor prognosis and do not contain viral RNA. Oral Oncol 2013;49:525–33. [17] Chaturvedi AK, Engels EA, Pfeiffer RM, Hernandez BY, Xiao W, Kim E, et al. Human papillomavirus and rising oropharyngeal cancer incidence in the United States. J Clin Oncol: Off J Am Soc Clin Oncol 2011;29:4294–301. [18] Ndiaye C, Mena M, Alemany L, Arbyn M, Castellsague X, Laporte L, et al. HPV DNA, E6/E7 mRNA, and p16INK4a detection in head and neck cancers: a systematic review and meta-analysis. Lancet Oncol 2014;15:1319–31. [19] SEER-9. Surveillance, Epidemiology, and End Results (SEER) Program (www. seer.cancer.gov) SEER⁄Stat Database: Incidence – SEER 9 Regs Research Data, Nov 2014 Sub (1973–2012) Katrina/Rita Population Adjustment – Linked To County Attributes – Total U.S., 1969–2013 Counties, National Cancer Institute, DCCPS, Surveillance Research Program, Surveillance Systems Branch, released April 2015, based on the November 2014 submission. [20] SEER-13. Surveillance, Epidemiology, and End Results (SEER) Program (www. seer.cancer.gov) SEER⁄Stat Database: Incidence – SEER 13 Regs Research Data, Nov 2014 Sub (1992–2012) Katrina/Rita Population Adjustment - Linked To County Attributes – Total U.S., 1969–2013 Counties, National Cancer Institute, DCCPS, Surveillance Research Program, Surveillance Systems Branch, released April 2015, based on the November 2014 submission. [21] SEER-18. Surveillance, Epidemiology, and End Results (SEER) Program (www. seer.cancer.gov) SEER⁄Stat Database: Incidence – SEER 18 Regs Research Data + Hurricane Katrina Impacted Louisiana Cases, Nov 2014 Sub (2000–2012) Katrina/Rita Population Adjustment – Linked To County Attributes – Total U.S., 1969–2013 Counties, National Cancer Institute, DCCPS, Surveillance Research Program, Surveillance Systems Branch, released April 2015, based on the November 2014 submission. [22] Kim HJ, Fay MP, Feuer EJ, Midthune DN. Permutation tests for joinpoint regression with applications to cancer rates. Stat Med 2000;19:335–51. [23] Holford TR. Understanding the effects of age, period, and cohort on incidence and mortality rates. Annu Rev Public Health 1991;12:425–57. [24] Tarone RE, Chu KC. Evaluation of birth cohort patterns in population disease rates. Am J Epidemiol 1996;143:85–91. [25] Rosenberg PS, Anderson WF. Age-period-cohort models in cancer surveillance research: ready for prime time? Cancer Epidemiol Biomarkers Prev 2011;20:1263–8. [26] Annertz K, Anderson H, Biorklund A, Moller T, Kantola S, Mork J, et al. Incidence and survival of squamous cell carcinoma of the tongue in Scandinavia, with special reference to young adults. Int J Cancer 2002;101:95–9. [27] Annertz K, Anderson H, Palmer K, Wennerberg J. The increase in incidence of cancer of the tongue in the Nordic countries continues into the twenty-first century. Acta Otolaryngol 2012;132:552–7. [28] Caetano R, Clark CL. Trends in alcohol consumption patterns among whites, blacks and Hispanics: 1984 and 1995. J Stud Alcohol 1998;59:659–68. [29] Nelson DE, Mowery P, Tomar S, Marcus S, Giovino G, Zhao L. Trends in smokeless tobacco use among adults and adolescents in the United States. Am J Public Health 2006;96:897–905. [30] Agalliu I, Gapstur S, Chen Z, Wang T, Anderson RL, Teras L, et al. Associations of Oral alpha-, beta-, and gamma-human papillomavirus types with risk of incident head and neck cancer. JAMA Oncol 2016. [31] Marks MA, Chaturvedi AK, Kelsey K, Straif K, Berthiller J, Schwartz SM, et al. Association of marijuana smoking with oropharyngeal and oral tongue cancers: pooled analysis from the INHANCE consortium. Cancer Epidemiol Biomarkers Prev 2014;23:160–71. [32] Li R, Faden DL, Fakhry C, Langelier C, Jiao Y, Wang Y, et al. Clinical, genomic, and metagenomic characterization of oral tongue squamous cell carcinoma in patients who do not smoke. Head Neck 2015;37:1642–9. [33] Pickering CR, Zhang J, Neskey DM, Zhao M, Jasser SA, Wang J, et al. Squamous cell carcinoma of the oral tongue in young non-smokers is genomically similar to tumors in older smokers. Clin Cancer Res 2014;20:3842–8. [34] Brouwer AF, Eisenberg MC, Meza R. Age effects and temporal trends in HPVrelated and HPV-unrelated oral cancer in the United States: a multistage carcinogenesis modeling analysis. PLoS ONE 2016;11:e0151098.
Contents lists available at ScienceDirect
Oral Oncology journal homepage: www.elsevier.com/locate/oraloncology
Rising incidence of oral tongue cancer among white men and women in the United States, 1973–2012 Joseph E. Tota a,⇑, William F. Anderson a, Charles Coffey b, Joseph Califano b, Wendy Cozen c, Robert L. Ferris d, Maie St. John e, Ezra E.W. Cohen b, Anil K. Chaturvedi a a
National Cancer Institute, Division of Cancer Epidemiology and Genetics, Rockville, MD, USA UC San Diego, Moores Cancer Center, La Jolla, CA, USA University of Southern California, Departments of Preventive Medicine and Pathology, Los Angeles, CA, USA d University of Pittsburgh, Department of Otolaryngology, Pittsburgh, PA, USA e UCLA, Department of Head and Neck Surgery, UCLA Head and Neck Cancer Program, Los Angeles, CA, USA b c
a r t i c l e
i n f o
Article history: Received 23 January 2017 Received in revised form 18 February 2017 Accepted 20 February 2017
Keywords: Oral tongue cancer Incidence Temporal trends Age-period-cohort analysis
a b s t r a c t Background: Despite significant reductions in tobacco use in the US, oral tongue cancer incidence has reportedly increased in recent years, particularly in young white women. We conducted age-periodcohort analyses to identify birth cohorts that have experienced increased oral tongue cancer incidence, and compared these with trends for oropharyngeal cancer, a cancer caused by human papillomavirus (HPV) that has also recently increased. Methods: We utilized cancer incidence data (1973–2012) from 18 registries maintained by the NCI SEER Program. Incidence trends were evaluated using log-linear joinpoint regression and age-period-cohort modeling was utilized to simultaneously evaluate effects of age, calendar year, and birth year on incidence trends. Results: Incidence of oral tongue cancer increased significantly among white women during 1973–2012 (0.6% annual increase, p < 0.001) and white men during 2008–2012 (5.1% annual increase, p = 0.004). The increase was most apparent among younger white individuals (<50 years; annual increase of 0.7% for men [p = 0.02] and 1.7% for women [p < 0.001] during 1973–2012). Furthermore, the magnitude of the increase during 1973–2012 was similar between young white men and women (2.3 vs. 1.8 cases per million, respectively). Incidence trends for oropharyngeal cancer were similar to trends for oral tongue cancer and similar birth cohorts (born after the 1940s) experienced rising incidence of these cancers (p-value: white men = 0.12, white women = 0.42), although the magnitude of increase was greater for oropharyngeal cancer. Conclusions: The incidence of oral tongue and oropharyngeal cancer has significantly increased among young white men and women within the same birth cohorts in the US. Published by Elsevier Ltd.
Introduction Cancers of the oral cavity and pharynx have traditionally been considered a single etiologic entity. However, recent studies show significant differences in the etiology and epidemiology for individual anatomic sites [1–3]. Reduction in smoking in the United States has resulted in significant declines in the incidence of most oral cavity cancers, including lip, gum, floor of mouth, hard palate, buccal mucosa, and vestibule cancers [4–8]. In contrast, the ⇑ Corresponding author at: National Cancer Institute, Division of Cancer Epidemiology and Genetics, Infections and Immunoepidemiology Branch, 9609 Medical Center Drive, Room 6E220, Rockville, MD 20850, USA. E-mail address: [email protected] (J.E. Tota). http://dx.doi.org/10.1016/j.oraloncology.2017.02.019 1368-8375/Published by Elsevier Ltd.
incidence of oral tongue cancers (anterior 2/3 of the tongue) has reportedly increased in recent years [7–12]. The incidence of oropharyngeal cancers has also increased, and descriptive and molecular epidemiologic studies have identified human papillomavirus (HPV) infection as the cause [2,3]. However, the reasons for the increase in oral tongue cancers are unknown, and molecular studies indicate that HPV does not play a major etiologic role [13–16]. Descriptive studies of oral tongue cancer trends have reported that incidence has significantly increased among young (ages 18–44 years), white individuals, and primarily in women [7,8,10– 12]. It is, however, unclear whether oral tongue cancer incidence has increased in similar magnitude among young, white men [7–12]. Furthermore, prior studies have not identified the specific
147
J.E. Tota et al. / Oral Oncology 67 (2017) 146–152
birth cohorts that have experienced an increase in oral tongue cancer incidence, which could provide important etiologic clues. For example, the identification of birth cohorts of men that went through the sexual revolution as the primary demographic subgroup that has experienced the rise in oropharyngeal cancer incidence, in part, enabled the identification of HPV as the cause [2,3,5,17,18]. The main objective of this study was to identify the demographic subgroups and birth cohorts that have experienced an increase in oral tongue cancer in the United States. We also aimed to compare and contrast our findings for oral tongue cancer with oropharyngeal cancer to determine if similar birth cohorts have experienced an increase in these cancers. This comparison may provide clues regarding the possible role of a sexually transmitted infection or other environmental exposure in the development of oral tongue cancer. Methods Data source Cancer incidence information from the US National Cancer Institute’s Surveillance, Epidemiology, and End Results (SEER) program database were obtained for the years 1973–2012 [19–21]. We included data from the SEER 9 Registries Database [19] for the years 1973–1991 (Atlanta, Connecticut, Detroit, Hawaii, Iowa, New Mexico, San Francisco-Oakland, Seattle-Puget Sound, and Utah), SEER 13 Registries Database [20] for the years 1992–1999 (SEER 9 regions plus Los Angeles, San-Jose Monterey, Rural Georgia, and Alaska Native Tumor Registry), and the SEER 18 Registries Database [21] for the years 2000–2012 (SEER 13 regions plus Greater California, Kentucky, Louisiana, New Jersey, and Greater Georgia), covering 14%, 17%, and 28% of the US population, respectively. Classification of anatomic sites For all analyses, oral cavity cancer sites were subdivided into two groups: oral tongue cancer, including the dorsal surface, border, ventral surface, and anterior 2/3 of the tongue (International Classification of Diseases for Oncology version-3 [ICD-O-3] topography codes C020-023) and other oral cavity cancer sites, including lip, gum, floor of mouth, and other/unspecified parts of the mouth (ICD-O-3 codes C000-009, C030-039, C040-050, and C060-069). HPV-associated oropharyngeal cancers were also evaluated, and these sites included base of tongue, lingual tonsil, soft palate and uvula, tonsil, oropharynx, and Waldeyer ring (ICD-O-3 codes C019, C024, C051-052, C090-099, C100-109, and C142) [5]. Cancers classified as tongue, not otherwise specified (ICD-O-3 code: C028) were excluded due to potential ambiguity regarding origin within the oral tongue vs. base of tongue. All histologic subtypes were included in our primary analysis for each cancer site. This study did not use personal identifying information from the SEER data or involve interaction with human subjects, therefore informed consent and institutional review was not required. Statistical analysis All analyses were conducted separately by sex and race: white (both non-Hispanic and Hispanic), black, and other races (American Indians and Alaska Natives, Asian/Pacific Islanders, and unknown). For each site, annual incidence rates (1973–2012; age-standardized to the US 2000 standard population) were obtained using SEER⁄Stat version 8.2.1 software. Temporal trends were evaluated using log-linear joinpoint regression, incorporating cubic regression splines selected based on the Akaike information
criterion. Trends are expressed as the annual percentage change (APC) in incidence [22]. For years with zero cases for a particular stratum, the annual incidence rate and standard error were taken from the preceding year, or closest year (preceding or following) when there were consecutive years with zero cases. In addition to presenting the APC (a measure of relative change in incidence), we also calculated the incidence rate difference (a measure of the absolute change in incidence) between 1973 and 2012, and difference in the average incidence rate between late (1973–77) and recent (2008–12) time periods. To simultaneously evaluate the effects of age, calendar year/ period, and birth year/cohort on incidence rates, age-periodcohort modeling was applied using 5-year age groups (20 to 24, . . ., 80 to 84), 5-year calendar periods (1973 to 1977, . . ., 2008 to 2012), and 5-year birth cohorts (1893, . . ., 1988). These models may be useful for detecting changes in exposure experiences (indicated by cohort effects), or changes in reporting, coding, screening and/or diagnostic practices (indicated by period effects) [23–25]. We specifically focused on two aspects of birth cohort effects in these models – cohort deviations and age-specific temporal trends (or net drifts) – that can enable identification of birth cohorts and age groups that have experienced significant changes in incidence rates. Differences in age-specific net drifts and cohort deviations were compared across cancer sites for the same sex (oral tongue cancer vs. oropharyngeal cancer and oral tongue cancer vs. other oral cavity cancer) and across sex for the same site using a 1-df Wald test. For oral tongue cancer and oropharyngeal cancer, temporal trends were then reevaluated focusing on young and older age groups separately, with the age cut point identified based on the observed age-specific net drifts. Given the relative rarity of oral tongue cancers, we combined the three SEER registries for enhanced statistical power. Nonetheless, to ensure robustness of our results as well as potential comparability of incidence rates across the SEER databases, we conducted two sensitivity analyses. First, we reevaluated all incidence trends using SEER 9 registries. Second, for overlapping calendar years (2000–2012), we compared incidence trends for oral tongue and oropharyngeal cancers across SEER 9, 13, and 18. We also conducted sensitivity analyses for oral tongue cancers restricted to squamous cell histologies.
Table 1 Characteristics of cases of oral tongue, oropharyngeal, and other oral cavity cancers during 1973–2012. Patient characteristic
Oral tongue (n = 16,206) no. (%)
Age at diagnosis, yearsa Mean 60.2 Standard 13.5 deviation
Oropharyngeal (n = 67,789) no. (%)
Other oral cavity (n = 56,168) no. (%)
60.0 11.5
63.0 13.1
Sex Male Female
9483 (58.5) 6723 (41.5)
50,748 (74.9) 17,041 (25.1)
37,074 (66.0) 19,094 (34.0)
Race/ethnicity White Black Other Unknown
13,772 (85.0) 1014 (6.2) 1292 (8.0) 128 (0.8)
56,977 (84.0) 7858 (11.6) 2611 (3.9) 343 (0.5)
49,556 (88.2) 3867 (6.9) 2117 (3.8) 628 (1.1)
Year of diagnosis 1973–1982 1983–1992 1993–2002 2003–2012
1742 2043 4400 8021
5906 (8.7) 7571 (11.2) 16,363 (24.1) 37,949 (56.0)
9360 (16.7) 9944 (17.7) 15,170 (27.0) 21,694 (38.6)
(10.7) (12.6) (27.2) (50.5)
a Restricted to individuals aged 0–84 years at time of cancer diagnosis (oral tongue: n = 15,350; oropharyngeal: n = 65,923; other oral cavity: n = 51,524).
148
J.E. Tota et al. / Oral Oncology 67 (2017) 146–152
Fig. 1. Incidence trends for oral tongue, oropharyngeal, and other oral cavity cancers in whites, stratified by gender (1973–2012). The two panels show a significant increase in the incidence of oral tongue and oropharyngeal cancer, and a significant decrease in other oral cavity cancers among both men and women (panels A and B, respectively). APC, annual percent change; OCC, oral cavity cancer; OPC, oropharyngeal cancer; OTC, oral tongue cancer. *Joinpoint regression comparison P < 0.05.
Results During 1973–2012, 16,206 oral tongue cancers, 67,789 oropharyngeal cancers, and 56,168 other oral cavity cancers occurred (Table 1). The mean age of patients diagnosed with oral tongue and oropharyngeal cancer was slightly lower (60.2 and 60.0 years, respectively) compared with patients diagnosed with other oral cavity cancer (63.0 years). The majority of cases across sites occurred among males (58.5–74.9%) and white individuals (84.0– 88.2%); however, females and ‘‘other” race/ethnic groups
accounted for a relatively high proportion of oral tongue cancers (41.5% and 8.0%, respectively). Due to the introduction of new registries being added to the SEER database, the proportion of cases across sites was higher in more recent decades. Overall incidence trends (1973–2012) The incidence of other oral cavity cancers significantly declined in recent years among most race and gender subgroups (Fig. 1A and B, Supplementary Fig. 1 A–D). In contrast, the
Fig. 2. Age-specific temporal trends (net drifts) in incidence of oral tongue, oropharyngeal, and other oral cavity cancers among white men and women (1973–2012). Wald P values are presented for evaluating differences in age-specific net drifts and cohort deviations across cancer sites (men and women separately), and across sex for the same cancer site. In panels A (men) and B (women), the overall increase in oral tongue cancer is shown to be heterogeneous by age (most apparent in younger individuals; p < 0.001 for both men and women), with a stronger increase in women than in men (p = 0.007). OCC = oral cavity cancer; OPC = oropharyngeal cancer; OTC = oral tongue cancer.
J.E. Tota et al. / Oral Oncology 67 (2017) 146–152
incidence of oral tongue cancers significantly increased among white women during 1973–2012 (APC = 0.6%, p < 0.001) as well as among white men during 2008 to 2012 (APC = 5.1%, p = 0.004) (Fig. 1A and B). The magnitude of the absolute increase in oral tongue cancer incidence during 1973–2012 was similar between white men and women (an increase of 2.5 vs. 2.6 cases per million). No increase in oral tongue cancer incidence was observed among men and women of other racial/ethnic groups (Supplementary Fig. 1 A–D). Results were similar when analyses were stratified by anatomic subsite (border of tongue, dorsal surface of tongue, and ventral surface of the tongue; data not shown). Significant increases in the incidence of oropharyngeal cancer were observed among both white men (APC = 0.8% during 1973–
149
1998, and 3.8% during 1998–2009; Fig. 1A) and white women (APC = 0.7% during 1995–2012; Fig. 1B). However, the magnitude of the increase (absolute change in incidence) for oropharyngeal cancer during 1973–2012 was substantially greater among white men than among white women (35.3 vs. 0.4 cases per million increase). Significant increases in oropharyngeal cancer were also observed among black men (APC = 4.2% during 1973–1984), black women (APC = 6.5% during 1973–1982), and men of other races (APC = 1.0% during 1973–2012). Because only whites experienced a statistically significant increase in both the incidence of oral tongue and oropharyngeal cancers, subsequent analyses were restricted to whites. However, Results from additional age-stratified comparisons for oral tongue,
Fig. 3. Incidence trends for oral tongue cancer and oropharyngeal cancer among white men and women, stratified by age (1973–2012). Panels A and B show a higher relative increase in oral tongue cancer among young women (reflected by higher APC) but a similar absolute increase in young men and women. Among older men and women, significant increases in oral tongue cancer are restricted to the most recent calendar years. Panels C and D show that both the relative and absolute increase in oropharyngeal cancer is higher among men (both young and older age groups), and that among older individuals, a significant increase occurred only in men and only in the most recent calendar years (1997–2012). *Joinpoint regression comparison P < 0.05. APC = annual percent change; OPC = oropharyngeal cancer.
150
J.E. Tota et al. / Oral Oncology 67 (2017) 146–152
oropharyngeal, and other oral cavity cancers for blacks and other races are provided in the Supplement (Supplementary Table 1). Age-specific incidence trends among white men and women (1973– 2012) The overall increase in oral tongue cancer incidence observed among white men and white women was heterogeneous by age, with the increases being most apparent among younger individuals (<50 years of age) (Fig. 2A and B). Additionally, the pattern of increase in young individuals was stronger in women than men (Wald p-value = 0.007). When stratified by age (<50 years and 50 years), oral tongue cancer incidence significantly increased during 1973–2012 in white men and white women ages <50 years (APC = 0.7% among men [p = 0.02] and APC = 1.7% among women [p < 0.001]) (Fig. 3A and B). Consistent with significant gender differences in age-specific net drifts (Fig. 2), the estimated APC for oral tongue cancer was higher among young white women compared with white men. However, the magnitude of the absolute increase during 1973–2012 was similar in men and women (2.3 vs. 1.8 cases per million increase; Fig. 3A and B). Among older white men and women (50 years of age), significant increases in oral tongue cancer were restricted to the most recent calendar years (APC = 5.1% during 2008–2012 among men [p = 0.005] and APC = 2.9% during 2005–2012 among women [p = 0.005]) (Fig. 3A and B). In individuals younger than 50 years, oropharyngeal cancer incidence increased significantly during 1973–2004 among white
men and during 1973–2012 among white women (Fig. 3C and D). Both the relative increase and the absolute increase in oropharyngeal cancer incidence during 1973–2012 was greater among young men than young women (APC = 3.2% vs. 0.7% and absolute increase = 10.6 vs. 0.0 cases per million increase). In individuals 50 years or older, oropharyngeal cancer incidence significantly increased among white men during 1997–2012 (3.7% increase per year) whereas no significant increase (or decrease) in incidence was observed among white women (Fig. 3C and D). Age-specific incidence rates by birth cohort To identify the specific birth cohorts that have experienced rising incidence, we conducted age-period-cohort analyses comparing cohort deviations between oral tongue and oropharyngeal cancer sites by gender, and cohort deviations across gender by cancer site. Similar patterns were observed in the age-specific incidence of oral tongue cancer and oropharyngeal cancer by birth cohort, with higher incidence rates among more recent birth cohorts of white men and women (Fig. 4A–D). Specifically, the same birth cohorts of white men and women (cohorts born after the 1940s) experienced rising incidence of both oral tongue and oropharyngeal cancers (p-value for birth cohort comparisons for oral tongue vs. oropharynx for white men = 0.12 and white women = 0.42). Across gender, the same birth cohorts of men and women experienced rising incidence of oral tongue cancers (pvalue for birth cohort comparisons in men vs. women for oral tongue cancers = 0.41). However, the birth cohort increases for
Fig. 4. Age-specific incidence rates for oral tongue cancers and oropharyngeal cancers among white men and women across birth cohort. Focusing on men and women separately, similar birth cohorts (born after the 1940s) experienced rising incidence of oral tongue and oropharyngeal cancers (men, panels A and C, p-value for birth cohort comparisons = 0.12; women, panels B and D, p-value = 0.42) and while birth cohort increases for oral tongue cancer were not significantly different across gender (p-value = 0.41), for oropharyngeal cancer, the magnitude and duration of increase was significantly greater among men (p < 0.001). OPC = oropharyngeal cancer.
J.E. Tota et al. / Oral Oncology 67 (2017) 146–152
oropharyngeal cancer were longer in duration and stronger in magnitude among men than among women (p < 0.001). In sensitivity analyses, the key observations of rising oral tongue and oropharyngeal cancer incidence during 1973–2012 in young, white men and women were qualitatively similar in analyses restricted to SEER 9 registries (Supplementary Table 2). Likewise, during the years 2000–2012, incidence trends for oral tongue and oropharyngeal cancer were similar across SEER 9, SEER 13, and SEER 18 registries (Supplementary Fig. 1). Results were not meaningfully different when comparing single-year and 5-year average incidence rate differences (Supplementary Table 3). Results were also similar in analyses restricted to squamous cell histologies (data not shown).
Discussion In this analysis of US cancer registry data spanning a period of four decades, we observed significant increases in oral tongue cancer incidence among white men and women, but not men and women of other racial/ethnic groups. As reported in numerous recent studies, the incidence of oropharyngeal cancers also increased significantly in recent decades in white men. Further, age-period-cohort analyses indicated that similar recent birth cohorts of white men and women have experienced a rise in the incidence of both oral tongue and oropharyngeal cancers in recent decades in the United States. Nonetheless, the magnitude of increase was appreciably greater for oropharyngeal cancers when compared to tongue cancers. Our finding of rising incidence for oral tongue cancers in young white individuals are similar to prior studies within the US as well as in Nordic countries [7–12,26,27]. However, based on differences by gender in annual percent change, most have suggested that the increase has primarily occurred among women [7,8,12]. In contrast, we show that the increase in oral tongue cancer incidence has occurred in both men and women to a similar extent, as measured by the number of cases. Further, the observation that similar birth cohorts of men and women have experienced the rise in oral tongue cancers suggests a similar etiologic exposure in both genders. Several observations argue against artefactual explanations or period effects for the rising tongue cancer trends, including consistent increases in incidence in young individuals over the entire duration since 1973, the percolation of the rising incidence into older individuals in recent years, and the similarity in trends across SEER 9, 13, and 18 registries. Further, there has not been an increase in screening for oral cavity cancers or oral tongue cancers in recent years, which would manifest as period effects. These observations collectively suggest birth cohort effects of changing exposure in white individuals. Tobacco and alcohol are unlikely to explain this rise in tongue cancer incidence given recent declines in smoking and per-capita alcohol consumption in the US population [4–8,28]. Likewise, the prevalence of chewing tobacco, another strong risk factor for oral tongue cancers as well as other oral cavity cancers, has remained historically low in the United States [29]. Indeed consistent with these risk factor trends, the incidence of most oral cavity cancers has significantly declined in recent years. The reasons for the increase in oral tongue cancer incidence are currently unclear, and as noted herein, most studies to date have been negative. Our observation that similar birth cohorts have experienced an increase in oral tongue and oropharyngeal cancers could suggest a potential role for HPV infection. However, recent molecular studies indicate that oncogenic alpha-HPV types (such as HPV16), the principal cause of oropharyngeal cancers, do not play a major etiologic role in oral tongue cancers [13–16].
151
Nonetheless, a recent US study reported that certain beta- and gamma-HPV types (such as HPV b-5) are associated with increased risk of head and neck cancers, including oral cavity cancers [30]. It is as yet unclear whether this association is specific to cancers of the oral tongue in the presence/absence of smoking, and if the prevalence of beta- and gamma-HPV types has increased in recent birth cohorts. The similar birth cohort experience for oral tongue and oropharyngeal cancers could also point to another environmental exposure or sexually transmitted virus/infection. Marijuana use, a suspected risk factor for oral cavity cancers, has significantly increased in recent birth cohorts. Nevertheless, a recent analytic study within the International Head and Neck Cancer Epidemiology Consortium reported that marijuana use was associated with a paradoxical decrease in risk of oral tongue cancers [31]. Further, two recent US studies that evaluated tumor specimens from young non-smoking oral tongue cancer patients (n = 7 and n = 53) failed to find evidence of viral involvement using next-generation exome or RNA sequencing methods [16,32]. Based on genomic similarity between oral tongue tumors obtained from young non-smokers and older smoking individuals, it has also recently been suggested that another as yet unidentified risk factor that induces inflammation or that disrupts the oral microbiome could increase oral tongue cancer risk [33]. Recently, Brouwer and colleagues utilized multistage clonal expansion models in combination with ageperiod-cohort models to estimate tumor initiation, promotion, and malignant conversion rates for different head and neck cancer subsites [34]. Promotion rates for oral tongue cancer were lower in comparison with other sites (other tobacco/alcohol-related oral cavity cancers and HPV-related oropharyngeal cancers), suggestive of longer sojourn time and a unique etiology [34]. We acknowledge both strengths and limitations of our study. The combination of multiple SEER databases provided us with an adequate sample to perform age-period-cohort modeling to compare cohort effects across gender and related cancer sites. However, we had limited sample sizes (power) for some subgroup analyses, which may explain the lack of statistically significant trends for most cancers among non-whites. Prior studies have arbitrarily selected age cut-points for defining ‘young’ versus ‘older’ individuals in making their comparisons of oral tongue cancer trends, whereas we evaluated age-specific net drifts to identify and empirically define our cut-point. Importantly, given the registry-based nature of our analyses, we did not have information on traditional risk factors, such as smoking and alcohol. We thus relied on ecologic comparisons across cancer sites to rule out the potential effect of these risk factors on oral tongue cancer incidence. There is a possibility that misclassification of HPV-related base of tongue cancers as anterior tongue cancers could have partly contributed to the rise in incidence of oral tongue cancers, but it is unlikely that such misclassification is restricted only to certain age and racial subgroups (e.g. young whites). To avoid such misclassification, we excluded overlapping lesions of the tongue as well as those cancers without anatomic tongue site specification. Further, the increase in oral tongue cancer incidence has occurred in both men and women, while the rise in HPV-related base of tongue cancers has primarily occurred in men, which also argues against misclassification of base of tongue cancers as a potential explanation. The risk factors that are contributing to the rise in oral tongue cancer cases observed among young white men and women in the United States are not currently known. Thus, additional studies are needed to generate and confirm hypotheses concerning risk factors for oral tongue cancers in the presence and absence of the traditional risk factors, such as smoking. These studies could potentially focus on suspected or emerging oral cancer risk factors whose prevalence may have increased in cohorts of white individuals born after the 1940s.
152
J.E. Tota et al. / Oral Oncology 67 (2017) 146–152
Conflict of interest Ezra E. W. Cohen reports consulting fees from AstraZeneca, Bristol-Myers Squibb, EMD Serono, Human Longevity Inc., Merck, and Pfizer outside the submitted work. The other authors do not have any conflict of interest to report.
Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.oraloncology. 2017.02.019.
References [1] Mork J, Lie AK, Glattre E, Hallmans G, Jellum E, Koskela P, et al. Human papillomavirus infection as a risk factor for squamous-cell carcinoma of the head and neck. N Engl J Med 2001;344:1125–31. [2] Chaturvedi AK, Anderson WF, Lortet-Tieulent J, Curado MP, Ferlay J, Franceschi S, et al. Worldwide trends in incidence rates for oral cavity and oropharyngeal cancers. J Clin Oncol: Off J Am Soc Clin Oncol 2013;31:4550–9. [3] Gillison ML, Chaturvedi AK, Anderson WF, Fakhry C. Epidemiology of human papillomavirus-positive head and neck squamous cell carcinoma. J Clin Oncol: Off J Am Soc Clin Oncol 2015;33:3235–42. [4] CDC. Cigarette smoking among adults and trends in smoking cessation – United States, 2008. MMWR Morb Mortal Wkly Rep. 2009; 58: 1227–32. [5] Chaturvedi AK, Engels EA, Anderson WF, Gillison ML. Incidence trends for human papillomavirus-related and -unrelated oral squamous cell carcinomas in the United States. J Clin Oncol: Off J Am Soc Clin Oncol 2008;26:612–9. [6] Shiboski CH, Shiboski SC, Silverman Jr S. Trends in oral cancer rates in the United States, 1973–1996. Commun Dent Oral Epidemiol 2000;28:249–56. [7] Patel SC, Carpenter WR, Tyree S, Couch ME, Weissler M, Hackman T, et al. Increasing incidence of oral tongue squamous cell carcinoma in young white women, age 18 to 44 years. J Clin Oncol: Off J Am Soc Clin Oncol 2011;29:1488–94. [8] Brown LM, Check DP, Devesa SS. Oral cavity and pharynx cancer incidence trends by subsite in the United States: changing gender patterns. J Oncol 2012;2012:649498. [9] Shemen LJ, Klotz J, Schottenfeld D, Strong EW. Increase of tongue cancer in young men. JAMA 1984;252:1857. [10] Myers JN, Elkins T, Roberts D, Byers RM. Squamous cell carcinoma of the tongue in young adults: increasing incidence and factors that predict treatment outcomes. Otolaryngol Head Neck Surg 2000;122:44–51. [11] Shiboski CH, Schmidt BL, Jordan RC. Tongue and tonsil carcinoma: increasing trends in the U.S. population ages 20–44 years. Cancer 2005;103:1843–9. [12] Joseph LJ, Goodman M, Higgins K, Pilai R, Ramalingam SS, Magliocca K, et al. Racial disparities in squamous cell carcinoma of the oral tongue among women: a SEER data analysis. Oral Oncol 2015;51:586–92. [13] Herrero R, Castellsague X, Pawlita M, Lissowska J, Kee F, Balaram P, et al. Human papillomavirus and oral cancer: the International Agency for Research on Cancer multicenter study. J Natl Cancer Inst 2003;95:1772–83. [14] Lingen MW, Xiao W, Schmitt A, Jiang B, Pickard R, Kreinbrink P, et al. Low etiologic fraction for high-risk human papillomavirus in oral cavity squamous cell carcinomas. Oral Oncol 2013;49:1–8. [15] Kruger M, Pabst AM, Walter C, Sagheb K, Gunther C, Blatt S, et al. The prevalence of human papilloma virus (HPV) infections in oral squamous cell carcinomas: a retrospective analysis of 88 patients and literature overview. J Cranio-Maxillo-Facial Surgery: Off Publ Eur Assoc Cranio-Maxillo-Facial Surgery 2014;42:1506–14.
[16] Bragelmann J, Dagogo-Jack I, El Dinali M, Stricker T, Brown CD, Zuo Z, et al. Oral cavity tumors in younger patients show a poor prognosis and do not contain viral RNA. Oral Oncol 2013;49:525–33. [17] Chaturvedi AK, Engels EA, Pfeiffer RM, Hernandez BY, Xiao W, Kim E, et al. Human papillomavirus and rising oropharyngeal cancer incidence in the United States. J Clin Oncol: Off J Am Soc Clin Oncol 2011;29:4294–301. [18] Ndiaye C, Mena M, Alemany L, Arbyn M, Castellsague X, Laporte L, et al. HPV DNA, E6/E7 mRNA, and p16INK4a detection in head and neck cancers: a systematic review and meta-analysis. Lancet Oncol 2014;15:1319–31. [19] SEER-9. Surveillance, Epidemiology, and End Results (SEER) Program (www. seer.cancer.gov) SEER⁄Stat Database: Incidence – SEER 9 Regs Research Data, Nov 2014 Sub (1973–2012) Katrina/Rita Population Adjustment – Linked To County Attributes – Total U.S., 1969–2013 Counties, National Cancer Institute, DCCPS, Surveillance Research Program, Surveillance Systems Branch, released April 2015, based on the November 2014 submission. [20] SEER-13. Surveillance, Epidemiology, and End Results (SEER) Program (www. seer.cancer.gov) SEER⁄Stat Database: Incidence – SEER 13 Regs Research Data, Nov 2014 Sub (1992–2012) Katrina/Rita Population Adjustment - Linked To County Attributes – Total U.S., 1969–2013 Counties, National Cancer Institute, DCCPS, Surveillance Research Program, Surveillance Systems Branch, released April 2015, based on the November 2014 submission. [21] SEER-18. Surveillance, Epidemiology, and End Results (SEER) Program (www. seer.cancer.gov) SEER⁄Stat Database: Incidence – SEER 18 Regs Research Data + Hurricane Katrina Impacted Louisiana Cases, Nov 2014 Sub (2000–2012) Katrina/Rita Population Adjustment – Linked To County Attributes – Total U.S., 1969–2013 Counties, National Cancer Institute, DCCPS, Surveillance Research Program, Surveillance Systems Branch, released April 2015, based on the November 2014 submission. [22] Kim HJ, Fay MP, Feuer EJ, Midthune DN. Permutation tests for joinpoint regression with applications to cancer rates. Stat Med 2000;19:335–51. [23] Holford TR. Understanding the effects of age, period, and cohort on incidence and mortality rates. Annu Rev Public Health 1991;12:425–57. [24] Tarone RE, Chu KC. Evaluation of birth cohort patterns in population disease rates. Am J Epidemiol 1996;143:85–91. [25] Rosenberg PS, Anderson WF. Age-period-cohort models in cancer surveillance research: ready for prime time? Cancer Epidemiol Biomarkers Prev 2011;20:1263–8. [26] Annertz K, Anderson H, Biorklund A, Moller T, Kantola S, Mork J, et al. Incidence and survival of squamous cell carcinoma of the tongue in Scandinavia, with special reference to young adults. Int J Cancer 2002;101:95–9. [27] Annertz K, Anderson H, Palmer K, Wennerberg J. The increase in incidence of cancer of the tongue in the Nordic countries continues into the twenty-first century. Acta Otolaryngol 2012;132:552–7. [28] Caetano R, Clark CL. Trends in alcohol consumption patterns among whites, blacks and Hispanics: 1984 and 1995. J Stud Alcohol 1998;59:659–68. [29] Nelson DE, Mowery P, Tomar S, Marcus S, Giovino G, Zhao L. Trends in smokeless tobacco use among adults and adolescents in the United States. Am J Public Health 2006;96:897–905. [30] Agalliu I, Gapstur S, Chen Z, Wang T, Anderson RL, Teras L, et al. Associations of Oral alpha-, beta-, and gamma-human papillomavirus types with risk of incident head and neck cancer. JAMA Oncol 2016. [31] Marks MA, Chaturvedi AK, Kelsey K, Straif K, Berthiller J, Schwartz SM, et al. Association of marijuana smoking with oropharyngeal and oral tongue cancers: pooled analysis from the INHANCE consortium. Cancer Epidemiol Biomarkers Prev 2014;23:160–71. [32] Li R, Faden DL, Fakhry C, Langelier C, Jiao Y, Wang Y, et al. Clinical, genomic, and metagenomic characterization of oral tongue squamous cell carcinoma in patients who do not smoke. Head Neck 2015;37:1642–9. [33] Pickering CR, Zhang J, Neskey DM, Zhao M, Jasser SA, Wang J, et al. Squamous cell carcinoma of the oral tongue in young non-smokers is genomically similar to tumors in older smokers. Clin Cancer Res 2014;20:3842–8. [34] Brouwer AF, Eisenberg MC, Meza R. Age effects and temporal trends in HPVrelated and HPV-unrelated oral cancer in the United States: a multistage carcinogenesis modeling analysis. PLoS ONE 2016;11:e0151098.