Temporal trends in the incidence and survival of cancers of the upper aerodigestive tract in Ontario and the United States
Version of Record online: 23 APR 2009
Copyright © 2009 UICC
International Journal of Cancer
Volume 125, Issue 9, pages 2159–2165, 1 November 2009
How to Cite
Gupta, S., Kong, W., Peng, Y., Miao, Q. and Mackillop, W. J. (2009), Temporal trends in the incidence and survival of cancers of the upper aerodigestive tract in Ontario and the United States. Int. J. Cancer, 125: 2159–2165. doi: 10.1002/ijc.24533
- Issue online: 25 AUG 2009
- Version of Record online: 23 APR 2009
- Accepted manuscript online: 23 APR 2009 12:00AM EST
- Manuscript Accepted: 7 APR 2009
- Manuscript Received: 29 JUL 2008
- Ontario Cancer Research Network/Ontario Institute for Cancer Research
- Cancer Care Ontario
- Canadian Institutes of Health Research
- head and neck cancer;
- cancer of the oropharynx;
- relative survival
The Ontario Cancer Registry (OCR) and the Surveillance, Epidemiology and End Results (SEER) databases were used to describe temporal trends in the incidence and survival of squamous cancers of the upper aerodigestive tract (UADT) in Ontario and the US between 1984 and 2001. Between the 1984–86 and 1999–01 periods, the age-adjusted incidence rate of all first primary cancers of the UADT decreased from 11.6 (11.2–12.0) to 8.8 (8.5–9.1) in Ontario and 13.0 (12.7–13.3) to 10.2 (10.0–10.4) in the US. Significant decreases in incidence were observed in many UADT sites but there was no significant change in the incidence of cancer of the oropharynx in either the US or Canada. Over the same period, the 5-year relative survival for all UADT cancers increased from 49.2% (47.2–51.2%) to 57.1%(55.0–59.1%) in Ontario and from 48.1% (46.9–49.3%) to 52.4% (51.2–53.6%) in the US. This significant improvement in the outcome of UADT cancer was largely due to a dramatic increase in the 5-year relative survival for cancers of the oropharynx from 31.1% (27.1–35.1%) to 53.6% (49.3–57.9%) in Ontario and from 35.3% (32.9–37.8%) to 51.0% (48.7–53.3%) in the US. Smaller increases in survival were observed in cancers of the oral cavity, nasopharynx, and hypopharynx, but there was no evidence of any increase in survival for cancer of the larynx. These results are consistent with the hypothesis that there has been a major change in the etiology of cancer of the oropharynx in Canada and the US and a concomitant change in its response to therapy. © 2009 UICC
Cancer of the UADT remains a significant problem in North America.1 Until recently, the use of tobacco and alcohol were considered to be the principal causes of UADT cancers in the developed countries.2 The decline in smoking rates in recent decades has been associated with a reduction in the overall incidence of cancer of the UADT in the US. However, the incidence of cancer of the oropharynx has not decreased in parallel with that of other UADT cancers.3, 4 This has led to the hypothesis that a real decrease in the incidence of smoking-related cancers of the oropharynx may have been obscured by a simultaneous increase in the incidence of a new variant of the disease caused by human papilloma virus (HPV).3–9
In a previous population-based study of UADT cancer over the period from 1982 to 1994, Skarsgard, et al.10 showed that the incidence of UADT cancers in Ontario, Canada was remarkably similar to that in the US. Smoking rates have been decreasing in Canada11 as they have in the US,3 but it is not yet known whether this has resulted in a decrease in the incidence of UADT cancer in Canada. The first objective of this study was to compare trends in the incidence of UADT cancers in Ontario and the US to determine: (a) whether the overall incidence of UADT cancer in Ontario has decreased as it has in the US; and (b) whether the temporal trend in the incidence of oropharyngeal cancer in Ontario deviates from that of other UADT sites, as it does in the US.
Over the last two decades, there have also been changes in the outcomes of cancer of the UADT that may, in part, reflect the changes in its etiology described above.4, 12 Several investigators have examined trends in the survival of UADT cancers at the population level in the US.4, 13, 14 Two groups have demonstrated a dramatic improvement in survival in cancer of the oropharynx in the US.4, 13 Increases in survival were seen in some other UADT sites, but there has been no improvement in the outcome of cancers of the oral cavity or the larynx; three groups have, in fact, reported a small decrease in the survival of patients with laryngeal cancer.4, 13, 14 The management of UADT cancer has changed over the last two decades. There has been a well-documented increase in the use of combined modality treatment involving radiotherapy and surgery.13 More recently, clinical trials have provided clear evidence of improvements in survival when radiotherapy is combined with chemotherapy.15–17 However, there is no evidence of any changes in treatment unique to oropharyngeal cancer, which might readily explain the exceptional improvement in survival observed in this disease.
In a previous study, Skarsgard, et al.10 showed that the survival of UADT cancers was almost identical in Ontario and the US. The second objective of this study was to compare trends in the relative survival of cancers of the UADT in Ontario and the US and to determine whether the divergent, site-specific, temporal changes in survival that have been documented in the US have also occurred in Canada.
Material and methods
This was a retrospective, population-based study of the incidence and relative survival of cancers of the UADT diagnosed in Ontario and the United States from 1984 to 2001.
The study population included all patients with squamous cell cancers (SCC) of the UADT diagnosed between January 1, 1984 and December 31, 2001 in the Canadian province of Ontario and in the regions of the US covered by the Surveillance, Epidemiology and End Results (SEER) program. For the analysis of survival, incident cases were grouped into 3-year cohorts (1984–86 through 1999–01). To facilitate comparisons of trends in survival with trends in incidence, incidence rates were reported for the same 3-year periods.
Sources of data
SEER is an active, population-based cancer registry supported by the US National Cancer Institute (NCI). It captures cancer cases occurring in the metropolitan areas of San Francisco-Oakland, Detroit, Atlanta, and Seattle-Puget Sound, and the states of Connecticut, Hawaii, Iowa, New Mexico, and Utah.18 These areas include ∼ 30 million people or 9% of the US population.18
The Ontario Cancer Registry (OCR) is a passive, population-based cancer registry operated by Cancer Care Ontario. It captures cancer cases occurring in the Canadian province of Ontario, which has a population of ∼ 11.6 million. The operation of the OCR has been described in detail elsewhere.19, 20 The registry identifies cancer cases by linking electronic records from multiple sources including pathology reports, cancer centre records, hospital separation records, and death certificates. The OCR uses a system of probabilistic records linkage to minimize the occurrence of double entries in the registry.20 The algorithms used are conservative, and this may lead to under-reporting of second primary cancers; to trigger the creation of a second primary in an individual patient, the registry must receive a record that includes both a different ICD-10 site code and a different ICD-O morphology code from those associated with the first primary.20 This highly conservative rule is only over-ridden when specific, high-quality information is available to confirm that there are indeed two separate primaries. As a result, only 0.9% of UADT cancer patients in the OCR had a record of more than one primary cancer of the UADT compared with 4.2% in SEER. We, therefore, decided to include in our analyses only the first head and neck cancer recorded in an individual patient; subsequent head and neck primaries were excluded. This permits a more valid comparison between OCR and SEER, but of course leads to a slight underestimation of the overall burden of these diseases. The OCR and SEER each provided information about the primary site of the cancer in terms of the International Classification of Disease (ICD-10),21 and its morphology in terms of the International Classification of Disease for Oncology (ICD-O) histology codes.22 The registries also provided date of diagnosis, sex, date of birth, and date of death.
Classification of UADT cancer
The ICD-10 codes corresponding to the UADT were grouped into clinical sites and subsites as described previously.10 The clinical cancer sites and subsites were defined as follows based on ICD-10 codes: Glottis: C32.0; Supraglottis: C32.1; Oral Cavity: C00.3, C00.4, C00.5, C02 (except C02.4), C03, C04, C05 (except C05.1, C05.2), C06; Oropharynx: C01.9, C02.4, C05.1, C05.2, C09, C10 (except C10.1); Nasopharynx: C11; Hypopharynx: C12, C13; and Sinus, Nasal Cavity, Middle Ear, Auditory Tube, Mastoid (SNCE): C30, C31.
For the purposes of this study, SCC was defined to include tumours with ICD-O histology codes 8050, and 8052–8084.22 This group included 83.7% of all cancers at UADT sites in the OCR, and 83.8% in SEER. The proportion of all UADT cases defined as SCC changed very little over the study period.
Incidence rates were generated based on census/intercensal estimates of the population of Ontario and the SEER areas of the US. Age-standardized incidence rates were calculated by adjusting to the age distribution of the 1991 Canadian population using the direct method.23
The incidence for each 3-year period was calculated as follows:
Three-year incidence rate = total number of incident cases within a 3 year period/number of person-years at risk. The standard error of the standardized incidence rate was calculated as described by Armitage et al23:
Standardized rate is as follows:
Standard error of standardized rate is as follows:
where pi—the crude rate at ith stratum.
ni—the number of persons at risk at ith stratum.
Ni is the number of persons in the ith stratum of standard population.
Analysis of survival
A few cases, 0.6% in the OCR and 1.0% of those in SEER, were diagnosed only at the time of death (by death certificate only or by autopsy) and were excluded from the survival analysis. Relative survival, calculated as described by Ederer et al.,24 was used to permit the international comparison of cancer outcomes, controlling for differences in mortality due to other causes. Expected-life tables were generated using US and Canadian mortality data. Expected survival was then assigned to each case based on age at diagnosis, gender, and the year of diagnosis. Race was also considered in the assignment of expected survival in the SEER population. The OCR does not collect information about race. Relative survival was calculated by dividing the observed survival by the expected survival.
Extent of Disease: SEER Historic Stage was used as a measure of the extent of the disease at diagnosis. This simple staging system classifies tumours as “local,” “regional,” or “distant.” Definitions for these categories are available from SEER.25 Overall, 5.9% of cases were unstaged over the whole study period with the proportion of unstaged cases decreasing slightly from 6.9% in the 1984–86 period to 3.3% in the 1999–01 period.
SEER records information on tumour grade at diagnosis based on the ICD-O-322 classifications of Grade 1, well differentiated, Grade 2, moderately differentiated, Grade 3, poorly differentiated, and Grade 4, undifferentiated tumours. Overall, 18.6% of cases diagnosed between 1984 and 2001 had no record of tumour grade in the registry. This proportion decreased from 23.5% in the 1984–86 period to 17.0% in the 1999–01 period. Due to differences in reporting between pathology labs, tumours classified as Grades 3 and 4 were combined into a single “high-grade” group.
Treatment information was only available for SEER cases. The OCR compiles no information about treatment. We determined whether radiotherapy, surgery, both were used in each patient's initial management. Details of rules used for classifying treatment are included in SEER's coding and staging manuals.25, 26 Over the whole study period, the proportion of patients with no treatment record remained approximately constant at 9%.
Incidence rates and relative survival were calculated using SEER*Stat 6.4.4 and SAS 9.1.3. Changes in incidence over time were tested for statistical significance using the Joinpoint method.27 Changes in relative survival over time were tested for significance using the test for trend described by Klein and Moeschberger,28 and the score statistics developed by Brown.29
For reasons described in the Methods, only first primary UADT cancers were included in this study. Between 1984 and 2001, we identified 19,573 first primary UADT cancers in Ontario, and 50,396 in the SEER population of the US. For the entire 1984–01 period, the mean age at diagnosis for first primary UADT cancers was 63.5 years in Ontario and 62.3 years in the US. Between the 1984–86 and 1999–01 periods, the mean age for all UADT cancers changed very little (from 63.1 to 63.6 years in Ontario and from 62.4 years to 61.5 years in the US). Over the same period, the mean age at diagnosis for cancer of the oropharynx decreased slightly from 62.2 to 61.4 in Ontario and from 61.0 to 58.6 in the US.
For the entire 1984–2001 period, the ratio of male to female patients (M:F ratio) with first primary UADT cancers was 2.9 in Ontario, and 2.6 in the US. Over the study period, the M:F ratio for all UADT sites remained constant at 2.9 in Ontario and decreased slightly from 2.7 to 2.6 in the US. However, there were some significant site-specific trends; the M:F ratio for cancer of the oral cavity decreased from 2.0 to 1.7 (p < 0.01) in Ontario, and 1.7 to 1.5 (p < 0.01) in the US, whereas the M:F ratio for cancer of the oropharynx increased from 2.4 to 3.0 (p < 0.01) in Ontario and from 2.2 to 3.6 (p < 0.01) in the US.
For the entire study period between 1984 and 2001, the age-standardized incidence rates for first primary SCC of the UADT was 10.2/105 in Ontario, and 11.6/105 in the US. Between the 1984–86 and 1999–01 periods, the age-adjusted incidence rate of all first primary cancers of the UADT decreased from 11.6/105 (11.2–12.0) to 8.8/105 (8.5–9.1) in Ontario and from 13.0/105 (12.7–13.3) to 10.2/105 (10.0–10.4) in the US. Table I shows that there were also significant decreases in incidence for many individual cancer sites, including the glottis, supraglottis, oral cavity, and hypopharynx. However, there was no statistically significant change in incidence rate of cancer of the oropharynx in either Ontario or the US.
|Site||Registry||Year of Diagnosis||Trend|
|Glottis||OCR (N = 3959)||2.4 (2.2–2.6)||2.2 (2.0–2.3)||2.3 (2.1–2.5)||2.1 (2.0–2.3)||1.9 (1.7–2.0)||1.7 (1.6–1.9)||p < 0.01|
|SEER (N = 9487)||2.5 (2.4–2.7)||2.4 (2.3–2.6)||2.4 (2.2–2.5)||2.2 (2.0–2.3)||1.9 (1.8–2.0)||1.8 (1.7–1.9)||p < 0.01|
|Supraglottis||OCR (N = 1930)||1.1 (1.0–1.2)||1.2 (1.0–1.3)||1.2 (1.1–1.4)||1.0 (0.9–1.1)||0.9 (0.8–1.0)||0.7 (0.6–0.8)||p = 0.04|
|SEER (N = 5876)||1.55 (1.46–1.65)||1.55 (1.46–1.65)||1.4 (1.3–1.5)||1.24 (1.16–1.32)||1.3 (1.2–1.4)||1.26 (1.18–1.34)||p = 0.11|
|Oral Cavity||OCR (N = 5538)||3.1 (2.9–3.4)||2.9 (2.7–3.1)||3.0 (2.8–3.2)||2.6 (2.4–2.8)||2.7 (2.5–2.9)||2.6 (2.4–2.7)||p = 0.02|
|SEER (N = 13714)||3.6 (3.4–3.7)||3.4 (3.3–3.6)||3.2 (3.1–3.3)||3.1 (3.0–3.2)||2.9 (2.8–3.0)||2.7 (2.6–2.8)||p < 0.01|
|Oropharynx||OCR (N = 3982)||2.3 (2.1–2.5)||2.1 (2.0–2.3)||2.0 (1.8–2.2)||2.1 (1.9–2.2)||2.1 (1.9–2.3)||2.1 (1.9–2.3)||p = 0.31|
|SEER (N = 10858)||2.5 (2.4–2.6)||2.6 (2.4–2.7)||2.5 (2.4–2.6)||2.4 (2.3–2.5)||2.5 (2.4–2.6)||2.5 (2.4–2.6)||p = 0.54|
|Nasopharynx||OCR (N = 834)||0.4 (0.3–0.5)||0.5 (0.4–0.6)||0.47 (0.39–0.54)||0.5 (0.4–0.6)||0.4 (0.3–0.5)||0.37 (0.30–0.43)||p = 0.30|
|SEER (N = 1906)||0.49 (0.44–0.55)||0.44 (0.39–0.49)||0.47 (0.42–0.52)||0.46 (0.41–0.51)||0.42 (0.38–0.47)||0.38 (0.34–0.43)||p = 0.04|
|Hypopharynx||OCR (N = 1448)||1.1 (0.9–1.2)||0.8 (0.7–0.9)||0.8 (0.7–0.9)||0.8 (0.7–0.9)||0.8 (0.6–0.9)||0.6 (0.5–0.7)||p = 0.01|
|SEER (N = 3848)||1.14 (1.06–1.23)||1.0 (0.9–1.1)||0.94 (0.87–1.01)||0.87 (0.81–0.94)||0.8 (0.7–0.9)||0.65 (0.59–0.71)||p < 0.01|
|SNCE||OCR (N = 605)||0.33 (0.25–0.40)||0.32 (0.25–0.39)||0.35 (0.28–0.42)||0.27 (0.21–0.32)||0.3 (0.2–0.36)||0.22 (0.17–0.27)||p = 0.06|
|SEER (N = 1438)||0.37 (0.33–0.42)||0.36 (0.31–0.40)||0.32 (0.28–0.37)||0.30 (0.26–0.34)||0.35 (0.31–0.39)||0.28 (0.24–0.31)||p = 0.08|
|All UADT1||OCR (N = 19573)||11.6 (11.2–12.0)||10.9 (10.5–11.3)||10.9 (10.6–11.3)||10.0 (9.6–10.4)||9.5 (9.2–9.9)||8.8 (8.5–9.1)||p < 0.01|
|SEER (N = 50396)||13.0 (12.7–13.3)||12.6 (12.4–12.9)||12.1 (11.9–12.4)||11.4 (11.1–11.6)||10.9 (10.6–11.1)||10.2 (10.0–10.4)||p < 0.01|
Panel A of Figure 1 compares the temporal trend in the incidence of cancer of the oropharynx with the trend observed in other UADT sites. In both Ontario and the US, the incidence of other UADT cancers decreased by >20% over the study period. In contrast, the incidence of cancer of the oropharynx remained more or less constant in the SEER population of the US throughout the study period. In Ontario, the incidence of cancer of the oropharynx decreased in the 1980's, but remained constant after 1991. Panels A and B of Figure 2 show that the difference in the trends in incidence between cancer of the oropharynx and the other UADT cancers was greater in people under 60 years of age than in those older than 60. Panels A and B of Figure 3 show that the divergence of the trend in incidence for cancer of the oropharynx from the trend for other UADT cancers was only observed in men; in women, the incidence of cancer of the oropharynx decreased in parallel with the decrease observed in other UADT cancers.
For the entire study period between 1984 and 2001, the relative survival for first primary UADT cancer cases was 53.0% in Ontario, and 49.7% in the US. Between the 1984–86 and 1999–01 periods, the 5-year relative survival for all UADT cancers increased from 49.2% (47.2–51.2%) to 57.1% (55.0–59.1%) in Ontario and from 48.1% (46.9–49.3%) to 52.4% (51.2–53.6%) in the US. As shown in Table II, the overall improvement in the outcome of UADT cancer was largely due to a dramatic increase in the 5-year relative survival of cancer of the oropharynx from 31.1% (27.1–35.1%) to 53.6% (49.3–57.9%) in Ontario and from 35.3% (32.9–37.8%) to 51.0% (48.7–53.3%) in the US. Smaller increases in relative survival were observed in cancers of the oral cavity and nasopharynx, but there was no evidence of any increase in relative survival in cancer of the larynx. In the early years of this study, the relative survival for most head and neck cancer sites was somewhat lower in Ontario than in the US, but that gap narrowed over the period of the study.
|Site||Registry||Year of Diagnosis||Trend|
|Glottis||OCR (N = 3959)||76.0 (71.6–80.3)||79.9 (75.5–84.3)||77.0 (72.8–81.2)||77.1 (72.9–81.3)||77.5 (73.2–81.8)||74.3 (69.6–79.0)||p = 0.92|
|SEER (N = 9487)||74.3 (71.6–76.9)||75.7 (73.0–78.3)||77.3 (74.7–79.9)||75.9 (73.2–78.6)||74.1 (71.3–76.9)||73.8 (70.9–76.6)||p = 0.75|
|Supraglottis||OCR (N = 1930)||45.9 (39.6–52.3)||38.4 (32.5–44.4)||47.6 (41.8–53.3)||46.4 (40.3–52.6)||45.0 (39.0–51.1)||44.5 (37.4–51.6)||p = 0.84|
|SEER (N = 5876)||48.0 (44.6–51.3)||45.9 (42.6–49.1)||47.5 (44.1–50.8)||40.7 (37.2–44.1)||44.8 (41.4–48.2)||43.8 (40.5–47.1)||p = 0.05|
|Oral Cavity||OCR (N = 5538)||51.4 (47.5–55.3)||53.6 (49.7–57.6)||55.1 (51.3–58.8)||55.8 (52.0–59.7)||57.4 (53.8–61.1)||61.9 (58.1–65.7)||p < 0.01|
|SEER (N = 13714)||51.3 (49.0–53.5)||48.2 (45.9–50.4)||50.1 (47.8–52.3)||51.5 (49.3–53.8)||48.9 (46.7–51.2)||54.0 (51.7–56.3)||p = 0.19|
|Oropharynx||OCR (N = 3982)||31.1 (27.1–35.1)||35.6 (31.4–39.9)||39.3 (34.9–43.7)||46.4 (42.1–50.6)||44.0 (39.9–48.1)||53.6 (49.3–57.9)||p < 0.01|
|SEER (N = 10858)||35.3 (32.9–37.8)||38.1 (35.6–40.5)||39.1 (36.7–41.5)||44.3 (41.9–46.8)||49.0 (46.7–51.4)||51.0 (48.7–53.3)||p < 0.01|
|Nasopharynx||OCR (N = 834)||52.9 (42.9–63.0)||54.6 (45.7–63.5)||60.0 (51.3–68.7)||57.9 (49.5–66.2)||59.5 (50.2–68.7)||60.2 (51.2–69.2)||p = 0.60|
|SEER (N = 1906)||44.7 (38.9–50.4)||42.3 (36.5–48.1)||45.3 (39.6–50.9)||55.9 (50.3–61.5)||51.4 (45.7–57.1)||56.0 (50.0–61.9)||p = 0.01|
|Hypopharynx||OCR (N = 1448)||18.3 (13.4–23.3)||22.5 (16.7–28.3)||27.7 (21.3–34.1)||20.4 (14.8–26.1)||21.8 (16.1–27.6)||24.7 (17.3–32.2)||p = 0.69|
|SEER (N = 3848)||23.3 (20.1–26.5)||23.2 (19.9–26.5)||27.8 (24.2–31.5)||29.2 (25.5–32.9)||26.1 (22.5–29.7)||25.3 (21.3–29.2)||p = 0.46|
|SNCE||OCR (N = 605)||49.8 (37.5–62.1)||43.9 (32.5–55.3)||42.0 (31.1–52.9)||46.5 (34.4–58.6)||53.8 (43.1–64.6)||59.0 (46.5–71.4)||p = 0.87|
|SEER (N = 1438)||51.8 (44.8–58.8)||51.5 (44.5–58.5)||49.6 (42.6–56.7)||47.2 (40.1–54.3)||49.9 (43.3–56.6)||47.6 (40.4–54.9)||p = 0.92|
|All UADT1||OCR (N = 19573)||49.2 (47.2–51.2)||50.7 (48.7–52.7)||53.2 (51.2–55.2)||54.5 (52.5–56.5)||53.8 (51.8–55.8)||57.1 (55.0–59.1)||p < 0.01|
|SEER (N = 50396)||48.1 (46.9–49.3)||47.9 (46.7–49.0)||49.3 (48.1–50.4)||50.2 (49.0–51.4)||50.2 (49.1–51.4)||52.4 (51.2–53.6)||p < 0.01|
Panel B of Figure 1 compares the temporal trends in relative survival for cancer of the oropharynx with those observed in other UADT cancers. Between the 1984–86 and 1999–01 periods, the 5-year relative survival of cancer of the oropharynx increased by >15% in both Ontario and the US cohort, while the relative survival of other UADT cancers remained relatively constant in both countries. Panels C and D of Figure 2 show that the improvement in survival in cancer of the oropharynx was greater in patients under the age of 60 compared with the older group. Panels C and D of Figure 3 show that the improvement in survival was greater in men than in women.
We looked for changes in case mix with respect to tumour stage and grade in an attempt to explain the differential trends in survival described above. Only the SEER cases could be included in this analysis because there was no information about stage or grade available in the OCR (see Methods).
Over the course of the study, the proportion of all staged UADT cases that were classified as “local” decreased from 41.0% to 36.9%; the proportion classified as “regional” increased from 49.1% to 53.5%; and the proportion classified as “distant” metastases remained more or less constant (9.9%–9.6%). The proportion of cases of cancer of the oropharynx classified as “local” decreased from 25.2% to 15.9%; the proportion classified as “regional” increased from 58.7% to 71.8%; and the proportion classified as “distant” decreased slightly from 16.0% to 12.2%.
Among the cases with grade assigned, the proportion of all UADT cases classified as well-differentiated decreased from 24.9% to 16.2%; the proportion classified as moderately differentiated increased slightly from 48.4% to 52.0%; and the proportion classified as poorly differentiated increased from 26.7% to 31.8%. For cancer of the oropharynx, the proportion of well-differentiated cases decreased from 15.8% to 7.1%; the proportion of moderately differentiated cases decreased slightly from 49.4% to 47.5%; and the proportion of poorly differentiated cases increased from 34.8% to 45.4%.
Among those with a record of treatment, the proportion of all UADT cases receiving surgery alone decreased from 29.3% to 24.5%; the proportion receiving radiotherapy alone remained constant at ∼32%; and the proportion receiving both surgery and radiotherapy increased from 26.1% to 34.2%. The proportion of patients with cancer of the oropharynx who were treated with surgery alone decreased from 21.2% to 12.7%; the proportion receiving radiotherapy alone decreased slightly from 39.1% to 36.1%; and the proportion receiving radiotherapy and surgery increased from 28.0% to 42.8%.
The main finding of this study is that trends in the incidence and survival of UADT cancers observed in Ontario over the last 20 years are essentially identical to those observed in the SEER population of the US. In both Ontario and the US, there has been a substantial decrease in the overall incidence of UADT cancer, whereas the incidence of cancer of the oropharynx has remained more or less constant. Over the same period, there has been a dramatic improvement in the 5-year relative survival of cancer of the oropharynx, whereas the survival of other UADT cancers has increased much less, or not at all. In both Ontario and the US, the divergence of the incidence and survival trends for oropharyngeal cancer from those observed for other UADT sites is primarily observed in men under the age of 60. In women and older men, the incidence of oropharyngeal cancer has decreased along with that of other UADT sites. Similar changes in the incidence and survival of cancer of the UADT have previously been reported in the US.3, 4, 9, 13 The similarity in the trends in incidence and survival of UADT cancers in Ontario and the US is striking, but consistent with previous comparative studies that have shown that the incidence and survival of UADT cancer in Ontario and the US is very similar.10, 30, 31
Overall, our findings are consistent with the hypothesis that a decrease in the incidence of smoking-related oropharyngeal cancer has been offset by an increase in the incidence of a new HPV-related variant of the disease.3, 5, 7, 9, 13 There is now strong molecular evidence that implicates human papillomavirus (HPV) in the causation of cancer of the oropharynx.5 There is also increasing epidemiological evidence of a strong association between HPV exposure and the risk of oropharyngeal cancer in the US.7 In recent reports from the US, at least 50% of oropharyngeal cancers have been positive for HPV DNA,8, 9 and the proportion of HPV-positive cases has increased greatly over the last two decades.9
It appears probable that the exceptional improvement in the survival of oropharyngeal cancer is due to the increasing proportion of HPV-related cases. There is increasing evidence that HPV-positive cancers of the UADT have a better prognosis than comparable HPV-negative cases.12 A meta-analysis of 23 retrospective studies showed that survival was significantly higher in HPV-positive cases when compared with HPV-negative cases of oropharyngeal cancer.12 There was no significant difference in prognosis between HPV-positive and HPV-negative cases at other UADT sites.12 A recent analysis of the impact of HPV-positivity on the outcome of radiochemotherapy for head and neck cancer, in the context of a prospective Phase 2 clinical trial, demonstrated large differences in both response rate and survival, in favour of the HPV-positive cases of oropharyngeal cancer.8
It remains possible that changes in the treatment of cancers of the UADT over the period of study may have been partly responsible for the large improvement in the survival of oropharyngeal cancer, as well as the more modest improvements observed in other sites. Our analysis of treatment in the US, which only included the use of radiotherapy and surgery, revealed an increased use of combined modality treatment in cancer of the oropharynx, but a similar change in practice was observed across the entire spectrum of UADT cancers. Other changes in the management of cancers of the UADT may have occurred over the period of study. Several clinical trials have found concomitant radiochemotherapy to be a more effective treatment for cancers of the head and neck than RT alone.15–17 Recent studies in the US using the National Cancer Database (NCDB) have shown that the use of radiochemotherapy has increased in the general cancer population of the US32, 35; Chen et al. found a 12.5% increase in the use of radiochemotherapy for laryngeal cancer32 compared with a 14% increase in oropharyngeal cancer.35 The small difference in the rate of adoption of this approach is clearly insufficient to explain the differential trends in survival reported here.
We have focused here on the extraordinary trends in incidence and survival in cancer of the oropharynx, where the evidence for causal association with HPV-16 is strongest. However, we recognize that HPV-16 has also been implicated in the etiology of cancer of the oral cavity, and that there is also some evidence for the involvement of HPV-6 and HPV-11 in the causation of cancer of the larynx. For a comprehensive review, readers are referred to a recent IARC monograph on the carcinogenicity of HPV.36
It is fortunate that the HPV-associated variant of oropharyngeal cancer appears to be more curable than the previously more common variant of the disease. Nonetheless, not every patient is cured, and the morbidity of treatment is often very high. Thus, the continuing increase in the incidence of this disease constitutes a serious public health problem. Our study suggests that Canada is now faced with an epidemic of HPV-related oropharyngeal cancer comparable in scale to the better-documented epidemic now in progress in the United States.3–7, 9, 13
It would clearly be preferable to prevent the occurrence of HPV-associated oropharyngeal cancer rather than incurring the substantial health and economic costs of treatment. Epidemiological studies show that the risk of HPV-associated oropharyngeal cancer is increased with multiple sexual partners and in particular multiple oral sex partners.7 However, lifestyle-related risk factors are not easy to modify and, therefore, the potential role of HPV-vaccines in the prevention of this disease is currently being discussed.33, 34 Further research into the natural history, transmission characteristics, and carcinogenity of HPV with respect to the UADT is required to permit modelling of the potential benefits of HPV vaccinations for boys as well as girls.
Mr. Shlok Gupta was supported by an Undergraduate Student Research Award from the Natural Sciences and Engineering Research Council of Canada. The authors thank Mr. Daniel Sutton for helpful discussions, and Ms. Beverley Shortt for her skill and patience in the preparation of the manuscript.
- 18NCI. SEER cancer statistics review, 1975-2005. Bethesda, MD: NCI, 2008.
- 20Cancer registration in Ontario: a computer approach. In: JensenOM,ParkinDM,MacLennanR,MuirCS,SkeetRG, eds. Cancer registration principles and methods. Lyon, France: International Agency for Research in Cancer, 1991. 246–257., , .
- 21WHO. The international statistical classification of diseases and health related problems, ICD-10, 2nd edn. Geneva: WHO, 2004.
- 22FritzAG,PercyC,JackA,SobinLH,ParkinMD, eds. International classification of diseases for oncology, ICD-O,3rd edn. Geneva: WHO, 2000.
- 23Statistical methods for epidemiology. In: Statistical methods in medical research, 4th edn. Oxford: Blackwell Publishing, 2002. 659–67., , .
- 25NCI. SEER Summary Staging Manual 2000: Codes and Coding Instructions. Bethesda, Maryland: NCI 2001, Report No.: NIH Pub No. 01-4969.
- 26NCI. SEER Program Coding and Staging Manual 2007. Bethesda, Maryland: NCI 2007, Report No.: NIH Pub. No. 07-5581.
- 36IARC. IARC monographs on the evaluation of carcinogenic risks to humans, human papillomaviruses, Vol. 90. Lyon: International Agency for Research on Cancer, 2007.