Investigating epidemiologic trends and the geographic distribution of patients with anal squamous cell carcinoma throughout Canada cancer, anorectal cancer, anorectal adenocarcinoma, anal squamous cell carcinoma, geographic clustering, epidemiology, incidence in Canada

Background Anal cancer is a rare disease, constituting 0.5% of new cancer cases in the United States. The most common subtype is squamous cell carcinoma ( scc ). Studies in several developed nations have reported on an increasing incidence of anal cancer in recent decades, and various risk factors pertaining to the pathogenesis of the disease have been identified, including infection with the human papillomavirus, tobacco use, and immunosuppression. The epidemiology and distribution of anal scc throughout Canada remain poorly understood, however. Using 3 population-based cancer registries, a retrospective analysis of demographic data across Canada for 1992–2010 was performed. The incidence and mortality for anal scc was examined at the levels of provinces, cities, and the forward sortation area ( FsA ) component (first 3 characters) of postal codes. During 1992–2010, 3720 individuals were diagnosed with anal scc in Canada; 64% were women. The overall national incidence rate was 6.3 cases per million population per year, with an average age at diagnosis of 60.4 years. The incidence increased over time, with significantly higher incidence rates documented in British Columbia and Nova Scotia (9.3 cases per million population each). Closer examination revealed clustering of cases in various urban centres and self-identified lgbtq communities in Toronto, Montreal, and Vancouver. This study provides, for the first time, a comprehensive analysis of the burden of anal scc in Canada, identifying susceptible populations and shedding light onto novel avenues of research to lower the incidence of anal cancer throughout the country.


INTRODUCTION
Anal cancer is one of the least common malignancies worldwide, representing only 2.4% of all gastrointestinal cancers 1 . Cancers of the anal canal can be divided into several subtypes, the most common being squamous cell carcinoma (scc), which accounts for 85% of reported cases 2,3 . The precursor lesion to anal scc is an anal intraepithelial neoplasia 4 , which can be further classified as either a low-grade or a high-grade squamous intraepithelial neoplasm 4 . The most common symptom at presentation of anal cancer is anorectal bleeding 5 . Treatment of anal cancer depends heavily on the grade and stage of the neoplasm; a multimodal approach is frequently required 6 .
Various risk factors are implicated in the pathogenesis of this malignancy, including infection with the human papillomavirus (hpv), tobacco use, advanced age, and immunosuppression (such as in the case of hiv/Aids or receipt EPIDEMIOLOGIC TRENDS AND GEOGRAPHIC DISTRIBUTION OF ANAL SCC IN CANADA, Cattelan et al. of a solid organ transplant) 7,8 . In addition, risk factors related to sexual practices such as anal receptive intercourse and having multiple sexual partners have been linked to a higher incidence of anal scc 3 . A history of dysplasia or malignancy of the cervix, vagina, or vulva has also been found to increase the risk 3 . Current data demonstrate that, in approximately 90% of all cases of anal scc, positivity for hpv infection is seen, with most of those infections (approximately 79%) being attributed to hpv strains 16 and 18 3,9 . Previous epidemiologic analyses of data from the U.S. Surveillance, Epidemiology, and End Results program 10 found that anal cancer constitutes only 0.5% of new cancer cases in the United States 10 . The median age of diagnosis is reported to be approximately 60 years 5 . However, although this cancer is relatively uncommon, several epidemiologic studies have reported an increasing incidence in recent decades, particularly in developed countries such as Australia 11 , Scotland 12 , France 13 , Denmark 14 , and the United States 2,15 . In Canada, limited studies have reported rising rates of anal scc in Quebec and Alberta in recent years 16,17 . Because of a low anal cancer incidence rate, the epidemiology of this malignancy is generally poorly documented, and studies describing its epidemiologic characteristics are sparse. For the present study, we conducted an extensive epidemiologic analysis of the burden of anal scc across all provinces and territories in Canada between 1992 and 2010. The geographic distribution of patients in Canada was analyzed with the aim of better understanding the risk factors related to the pathogenesis of this neoplasm and of identifying communities at high risk for anal scc.

METHODS
Our study was conducted in accordance with the cissrdc-668035 and 13-ssh-mcg-3749-S001 protocols approved by, respectively, the Social Sciences and Humanities Research Council of Canada and the Québec Inter-University Center for Social Statistics. In accordance with institutional policy, the study also received an exemption from review by the McGill University Research Ethics Board. The data for anal scc incidence was examined using two distinct population-based cancer databases: the Canadian Cancer Registry and the Registre québécois du cancer for 1992-2010. Codes set out in the International Classification of Diseases for Oncology, 3rd edition, were used for 2 subtypes of anal scc as reported in previous studies by the authors  . To assess causes of mortality, we used the Canadian Vital Statistics database, applying topographic codes from the International Statistical Classification of Diseases and Related Health Problems, 9th revision (icd-9) for deaths occurring during 1992-1999 and 10th revision (icd-10) for deaths occurring during 2000-2010. Because of space limitations, detailed methods are presented in supplementary Appendix 1.

Demographic and Epidemiologic Characteristics of Patients with Anal SCC in Canada
The subtypes of anal scc examined using information gathered from the Canadian Cancer Registry and the Registre québécois du cancer included scc and verrucous carcinoma. Given that approximately 98% of cases in the database corresponded to the scc histology, a further breakdown of demographic data by subtype was not performed.
Table i presents a general overview of epidemiologic characteristics of anal scc in Canada for the selected time period. Of 3720 total patients diagnosed with anal scc, most (64%) were women, for a female:male incidence ratio of 1.8:1. The average age at diagnosis was 60.4 ± 1.6 years, with 74% of the patients being more than 50 years of age. Overall, the annual incidence of anal scc showed a steady upward trend [ Figure 1(A)]. In 1992, the incidence rate was 3.7 cases per million population per year; in 2010, it had risen to approximately 9.0 cases, representing an increase of 142% over 18 years. A notable increase in the incidence rate in 2007 corresponded to a rate of 9.7 cases per million population per year. When examining the annual incidence by sex, a general upward trend over the years was noted [ Figure 1(B)], with the highest incidence consistently being seen for women.
A relationship between anal scc incidence rates and socioeconomic status (ses) was evident. The incidence rates were significantly lower in the highest ses quintile compared with the lowest ses quintile (incidence rate ratio for ses Q5 vs. Q1: 0.20; 95% confidence interval: 0.11 to 0.37). With respect to ethnicity, incidence rates were higher in Q5 white (highest-ses white individuals) than in Q1 white (lower-ses white individuals), with an incidence rate ratio of 1.16. However, that increase was not statistically significant (95% confidence interval: 0.99 to 1.36).

Geographic Distribution of Anal SCC Cases in Canada
The average national incidence rate for anal scc during 1992-2010 was 6.3 cases per million population per year (Table i). Figure 2 presents a geographic analysis of trends in anal scc throughout the country. On the provincial level, Nova Scotia and British Columbia had incidence rates significantly higher than the national average, at 9.31 cases per million population each. In contrast, Manitoba, Saskatchewan, and Newfoundland and Labrador had significantly lower incidence rates of 5.0, 4.4, and 3.9 cases per million respectively. Prince Edward Island, New Brunswick, Quebec, Ontario, and Alberta had incidence rates that matched the national average. Although the frequencies of anal scc in the 3 Canadian territories were examined, those frequencies were not within a reportable range.
Incidence rates for Canadian cities (Table ii) corroborated the foregoing trends. In total, 120 cities with a reportable frequency of anal scc between 1992 and 2010 were identified. Of those cities, only 10 had incidence rates significantly higher than the national average. Strikingly, 7 of them (70%) were located in British Columbia, 2 (20%) in Ontario, and 1 (10%) in Nova Scotia. Conversely, of the 12 significantly low-incidence cities identified in Canada during the period of interest, 9 (75%) were located in Ontario; the remaining 3 (25%) were in Quebec.
The distribution within cities or rural areas of patients with anal scc was then analyzed by the forward sortation area Table iii identifies 3 statistically significant zeroincidence FsAs in Fredericton, New Brunswick (E3G), Stoneham-et-Tewkesbury, Quebec (G3C), and Sturgeon Country, Alberta (T8T). Of the 56 high-incidence FsAs identified, only 10 (18%) were located in rural areas across the country ( Table iv).

Analysis of Mortality from Anal Canal Cancer Across Canada
Table i presents a general overview of the epidemiologic characteristics of anal cancer mortality. Overall, 1110 patients died of anal cancer between 1992 and 2010 in Canada. Most were women (57% vs. 43% men), with a female:male incidence ratio of 1.3:1. The average age at death was 67.4 ± 3.2 years.   (A) Incidence rates. *Statistically significant lower incidence rate (p < 0.05) compared with Canada. **Statistically significant higher incidence rate (p < 0.05) compared with Canada. (B) Mortality rates. *Statistically significant lower mortality rate (p < 0.05) compared with Canada. **Statistically significant higher mortality rate (p < 0.05) compared with Canada.

DISCUSSION
In this report, we present extensive data on the epidemiology of anal scc throughout Canada during 1992-2010 at the provincial, municipal, and FsA levels. To the best of our knowledge, no previous study has assessed the distribution of anal scc across the entire country and to this level of detail. At the national level, we report a steadily increasing burden of anal scc, with a female predominance and an average incidence rate of 6.3 cases per million individuals per year. Similar trends have been observed in several studies conducted in Europe, Australia, and the United States 11,13,15,43,44 (Figures 4 and 5). The highest incidence of anal scc nationwide was located in the borough of Church and Wellesley, an area known as "The Village," a historic neighborhood for lgbtq communities 45,46 . Similarly, significant clusters were found in Montreal's Gay Village and Vancouver's Davie Village 47,48 . The others corresponded to the boroughs predominantly inhabited by low-income or immigrant populations [49][50][51][52][53] . Those findings are corroborated by our analysis of median income by FsA, which demonstrated a statistically significant relationship between low ses and a high incidence rate of anal scc. Further details concerning those FsA clusters are provided in supplementary Appendix 3.2.
We report a significantly higher incidence of anal scc in urban areas, with only 18% of cases found in rural FsAs. The same trend has been noted in Norway and the United States 44,54 . This predilection of anal scc for urban centres might reflect an increased concentration of individuals exposed to predisposing risk factors for the disease-namely, changes in sexual practices leading to increased hpv transmission, tobacco use, and hiv infection. Changes in sexual practices throughout the world, such as lower age at first intercourse, higher number of sexual partners, and increased practice of anal receptive intercourse, might be contributing to the increasing incidence of anal scc 55-57 . The increased incidence of this disease was observed in      Western countries 58,59 . Tobacco use has been cited as a risk factor for anal scc 60 ; however, the rate of tobacco use throughout Canada has been steadily decreasing 61 , and therefore might not be one of the most contributory risk factors in this population. Infection with hiv has also been cited as an important risk factor for the pathogenesis of anal scc 62 , and its prevalence throughout Canada has been increasing, particularly in men having sex with men (msm), representing 47% of national cases in 2011 40 . Clustering of hiv cases in msm has been noted in Ontario, Quebec, and British Columbia 40 , and in fact, we report clustering of anal scc cases in the FsAs that encompass the 3 largest self-identified lgbtq communities in the country, which are also located in those 3 provinces. Similar trends have been observed worldwide and have been attributed to higher proportions of msm, leading to an increased prevalence of anal receptive intercourse 63 . Data for the prevalence of lgbtq-identifying individuals by FsA were not available for our study. However, in parts of the world where that information has been available, similar trends have been observed; they are detailed in supplementary Appendix 3.3. The incidence of hpv-related cancers is increasing internationally 64 ; however, it appears that the frequency of anal scc is increasing the most dramatically of all-in both sexes and most ethnic groups 64 . Despite those observations, no national anal scc screening guidelines are in place 54 . Since 2010, widespread hpv vaccination programs have been instituted across Canada, leading to significant declines in the prevalences of hpv, anogenital warts, and cervical cancer in every province [65][66][67] . Cervical cancer and anal scc have both been strongly linked to the same particular strain, hpv 16 68,69 . Cost-benefit analyses of anal cancer screening in women with cervical cancer 70 and in msm with and without concomitant hiv infection [71][72][73][74] have shown that vaccine campaigns against hpv 16, regular digital rectal examination, and anal Pap smear screening could be beneficial in reducing morbidity and mortality from the disease. Given the increasing rate of anal scc and lack of knowledge about this disease in high-risk groups 72,75,76 , it might be useful to consider instituting screening or vaccination programs for susceptible populations in the Canadian urban centres with higher incidences.
It is often believed that routine screening for anal cancer could cause physical and psychological discomfort to patients, but a study performed in 2011 in Toronto found that only 15%-32% of 104 regularly screened patients reported negative psychological consequences associated with the screening process 77 . In fact, the largest barrier to care for these high-risk patient groups, both male and female, is the lack of knowledge and awareness about the disease, its risk factors, and the available prevention and screening resources [78][79][80] . Through this study, our group has provided data that can be used to target populations within each province affected by a high burden of anal scc, consequently providing increased opportunities and needs for intervention and outreach. The information we have compiled might help to guide community health ambassadors in developing targeted and culturally appropriate education and promotion campaigns to raise awareness of the risk factors for anal cancer, of screening methods, and of hpv vaccination. Primary health care workers within affected communities can be encouraged to incorporate screening questionnaires, relevant physical exams, and hpv vaccination more readily into their practice, and they in turn can encourage their patients to perform self-directed anal visual inspection regularly at home 81 . Our results could be further correlated with demographic information (such as education level, income, employment status, location and accessibility of medical specialists in the region) from other Canadian studies, to contribute to health care policy development and resource allocation.
Our study is not without limitations; large retrospective studies that use databases such as those used here present a risk of data omission or misclassification. When conducting our search, different sets of icd codes had to be used, depending on the year of interest. Data for the Canadian territories were generally sparse, and therefore not sufficiently statistically significant to comment on. Data concerning the prevalence of various sexual orientations and sexual practices by region were not available for our study, but might have strengthened the clustering trends observed. Finally, because of federal confidentiality regulations and the associated mandatory rounding and suppression of data at too-low frequencies, the data obtained could not be presented in their entirety.

CONCLUSIONS
This epidemiology study highlights areas of clustering of anal scc throughout Canada, localizing it to the 3 main urban centres: Toronto, Montreal, and Vancouver. We provide an overview of risk factors to consider, particularly in susceptible populations in the inner cities and in lgbtq-predominant communities. Future analyses might confirm the existence of those risk factors in those communities. Our report also provides the basis for highlighting areas of Canada that could benefit from more efforts in education, primary prevention, and screening for the disease, and more focused distribution of health care resources to lower the incidence, morbidity, and mortality relating to anal cancer throughout the country.