Background In the developed countries, uveal melanoma is the most common primary intraocular malignancy in adults. Little is known about the epidemiological and geographical distribution of uveal melanoma in Canada.
Methods To determine the incidence patterns and geographical distribution of uveal melanoma cases in Canada, we conducted the first comprehensive, population-based national study of this malignancy across all Canadian provinces and territories during 1992–2010 years. We examined two independent population-based registries: the Canadian Cancer Registry and Le Registre Québécois du Cancer using corresponding International Classification of Diseases for Oncology-3rd edition codes for all histological subtypes of uveal melanoma.
Results We report that 2215 patients were diagnosed with uveal melanoma, of which 52.1% were males. The average -annual incidence rate of uveal melanoma in Canada was 3.75 cases per million individuals per year (95% CI 3.60 to 3.91). Overall, we report a steady increase in uveal melanoma incidence with an annual increase of 0.074 cases per million individuals per year. Significant differences in the incidence rates of uveal melanoma between Canadian provinces and territories were noted, where the highest crude incidence was in British Columbia and Saskatchewan with rates of 6.38 and 5.47 cases per million individuals per year, respectively.
Conclusions This work, for the first time, defines the disease burden of uveal melanoma in Canada and highlights important longitudinal, geographical and spatial differences in the distribution of uveal melanoma in Canada.
- uveal melanoma
- epidemiology in Canada
- incidence in Canada
- patient clusters
Statistics from Altmetric.com
It is estimated that approximately 8–11 individuals per million population per year are diagnosed with primary tumours that involve the eye and adnexal tissues every year in Canada. Among the ocular tumours, ocular melanoma is the most common primary intraocular malignancy in adults and is the second most common type of melanoma after the cutaneous type. The majority (80%–95%) of ocular melanomas arise within the uveal tract, and hence are known as uveal melanoma. The mean age-adjusted incidence rate of uveal melanoma in the USA between 1973 and 1997 was reported to be 4.3 cases per million individuals per year.1 In Europe, the incidence of uveal melanoma from 1983 to 1994 ranged from over eight cases per million individuals per year in Norway, Scotland and Denmark, to less than two cases per million individuals per year in Spain and Southern Italy.2
Several internal and external risk factors for uveal melanoma have been highlighted by previous reports.3 Intrinsic factors include Caucasian race, fair hair/skin complexion and light iris colour,4 family history of uveal melanoma,5 specific genetic abnormalities,6 oculodermal melanocytosis (nevus of Ota), diagnosis of familial atypical mole syndrome and previous history of cutaneous malignant melanoma. Extrinsic factors include exposure to artificial UV and/or blue light from welding and sunlamps7 and occupational exposure to chemicals.8 Viruses, such as herpesviruses and togaviruses, were also implicated as possible risk factors for uveal melanoma.9 Notably, while the role of ultraviolet (UV) radiation as a primary risk factor for the development of cutaneous melanoma is well established, conflicting data were reported by different studies regarding the role of solar ultraviolet (UV) radiation in the pathogenesis of uveal melanoma.10
In this study, we define the disease burden of uveal melanoma in Canada and highlight important longitudinal, geographical and spatial differences in the distribution of this cancer in Canada.
We examined the data on the incidence of uveal melanoma using two population-based cancer registries: the Canadian Cancer Registry (CCR) and Le Registre Québécois du Cancer (LRQC) for the period of 1992–2010 using International Classification of Diseases for Oncology-3rd edition (ICD-O-3) codes for all histological subtypes of uveal melanoma, in a similar manner as previously reported for other malignancies.11–15 The CCR is a dynamic registry of Canadian residents from 12 Canadian provinces and territories (excluding Québec), who have been diagnosed with primary tumours from 1992 to 2013, alive or dead. Data for Québec patients were obtained from LRQC and wereavailable only from 1992 to 2010. Data on new cases of uveal melanoma were obtained from the CCR (2014 version) which spanned the period of 1992–2013. Since the data from LRQC registry for Québec were only available up to 2010, we chose for this study to analyse the data from 1992 to 2010 to include all Canadian provinces for the same period. The CCR/LRQC registries provide demographic and geographical information including patients’ sex, year of diagnosis, age at the time of diagnosis, province/city/postal code of residence and the ICD-O-3 code of the tumour. The CCR/LRQC registries did not collect data on clinical disease stage at the time of diagnosis or ethnic background of patients.
Cases of uveal melanoma were searched and obtained from the registries by the primary site (topography) of uvea (denoted by the codes C69.2-C69.4), where C69.2 represents malignant neoplasm of retina, C69.3 malignant neoplasm of choroid and C69.4 malignant neoplasm of ciliary body (eyeball, iris, lens, sclera, uveal tract). The uveal melanoma subtypes that were searched and analysed were defined based on the following ICD-O-316 codes: 8720 (malignant melanoma, not otherwise specified [NOS]), 8721 (nodular melanoma), 8730 (amelanotic melanoma), 8741 (malignant melanoma in precancerous melanosis) and 8743 (superficial spreading melanoma).
For incidence analyses, data on population counts nationally, per province, city and forward sortation area (FSA: geographical region in which all postal codes start with the same three characters) were obtained from the Canadian Census of Population for 1996, 2001, 2006 and 2011 years from Statistics Canada. In Canada, postal codes consist of letters and numbers (eg, H3G 1A4), where the first three entries (eg, FSA) define a region in the country, with more than 1600 FSAs across Canada.
Mandatory data rounding
A number of confidentiality rules apply to CCR and LRQC data prior to its publication, such as random rounding of variables instead of presenting the exact number of cases. In regard to random rounding of tabular data, the SSHRC/Statistics Canada requires to round each cell count, independently of other cells, to a lower or higher multiple of 5 using an unbiased rounding scheme. No number ≥1 and <5 cases can be released as per the SSHRC regulation to protect patient confidentiality.
Statistical and mapping analyses
Unless otherwise specified, analysis of the complete data on all patients with uveal melanoma across Canada for the period 1992–2010 is presented throughout the paper. All presented rates were depicted per million individuals per year. Crude as well as age-adjusted incidence rates and 95% CIs were calculated and reported overall, by the year of diagnosis and specific regions that were identified by the mapping analysis. CIs were based on exact Poisson distributions. Incidence rates were plotted and linear regression models were used to assess trends over time. Data on population counts nationally, per province and city, were obtained from the Canadian Census of Population Programme from Statistics Canada.
Geographical maps of Canada which are divided per FSA postal code indicating the residence of patients with uveal melanoma recorded by the CCR and LRQC registries were generated using Microsoft Excel mapping software. In mapping of the CCR and LRQC results, only FSA with populations of at least 5000 individuals based on 1996, 2001, 2006 and 2001 census data were selected to reduce erroneous false-positive hits, where few cases of uveal melanoma occurring within a scarcely populated (<5000 residents) FSA might have artificially inflated the incidence rate. Similarly, cities with population of ≥50 000 were analysed.
Demographic characteristics of Canadian patients with uveal melanoma
There were 2215 patients who were diagnosed with uveal melanoma, accounting for ~2.9% of all melanoma cases in Canada within the aforementioned period (table 1), with skin accounting for ~94.2% of all melanomas in Canadian patients.12 13 The two main histological subtypes of uveal melanoma were malignant melanoma, not otherwise (NOS) (98.9% of all cases) and amelanotic melanoma (0.9%).
The average annual incidence rate of uveal melanoma during 1992–2010 was 3.75 cases per million individuals per year (95% CI 3.60 to 3.91) (figure 1A). The age-standardised incidence rate for uveal melanoma, against the world population, for the entire period was calculated to be 3.34 cases per million individuals per year (95% CI 3.20 to 3.47). Notably, rates of uveal melanoma were on the rise with the slope of 0.07 cases per million individuals per year (p=0.003). The overall incidence of uveal melanoma cases was slightly higher in males (52.6% vs 47.4% in females—sex adjusted), with male to female incidence rate ratio of 1.11. Average incidence rates for males and females were 3.93 and 3.54, respectively. For both sexes, there was a yearly incidence increase of 0.069 cases per million individuals for males vs 0.079 for females. The mean age ±SD of diagnoses for uveal melanoma in Canada was 61.49±14.21 years, which did not significantly differ between the two sexes (61.12±13.55 for males vs 61.89±14.89 for females). Analysis of the age distribution revealed that 58.2% of patients with uveal melanoma in Canada during 1992–2010 were 60 years of age or older, while 34.6% of patients were 40–59 years old. Only 7.2% of patients were younger than 40 years of age.
Incidence and geographical distribution of uveal melanoma cases in Canada during 1992–2010
The incidence rates for the Canadian provinces and cities revealed notable trends (online supplementary table 1 and figures 2–3). The provinces of British Columbia and Saskatchewan had significantly higher incidence than the national average, with rates of 6.38 and 5.47 cases per million individuals per year, respectively. On the other hand, several provinces such as Alberta, Ontario and Newfoundland had lower incidence rates, ranging from 2.06 to 3.32 cases per million individuals per year (Figure 2 and online Supplementary table 1). The analysis of uveal melanoma incidence by city and FSA corroborated the provincial trends (figures 3–4 and online supplementary table 2). Uveal melanoma incidence in five cities in British Columbia: North Vancouver, Port Coquitlam, Nanaimo, Victoria and Saanich was significantly higher than the Canadian average (figure 4A–L and online supplementary table 1). On the other hand, several cities in the remaining provinces had statistically significant lower rates of uveal melanoma than the rest of the country (Figures 4 and online Supplementary table 1).
Twenty-seven FSAs with statistically significant higher incidence of uveal melanoma were identified, listed in online supplementary table 2 and are geographically depicted in different shades of red (depending on the incidence rate) in figure 4A–L. More than half (63%) of theses high incidence FSAs (17/27) were in the provinces of British Columbia and Saskatchewan. Four high incidence FSAs were in Ontario, 3 in Eastern Québec, 2 in Alberta and one in Manitoba (online supplementary table 2). We also conducted incidence rate analysis to identify areas that were spared (ie, zero cases) by uveal melanoma during 1992–2010. This analysis documented six FSAs that were statistically significant as having low incidence of uveal melanoma. Perhaps not surprising, none of these communities were in British Columbia or Saskatchewan (online supplementary table 3). Also, of note, the majority (four out of six) of these FSAs were in Ontario and Alberta, the provinces with statistically significant low incidence of uveal melanoma.
Importantly, in several cases there were multiple high-incidence FSAs that were located geographically side by side. Specifically, in British Columbia, the following FSAs were contiguous: (V0K, V0E, V0X, V2C, V0H, V0G) spanning an area from Kamloops to Kelowna and beyond, (V7L, V3B, V3H) in Port Moody, Port Coquitlam and North Vancouver. Also, FSAs V9W, V0R, V8P, V9T covered a large region of the Vancouver Island (figure 4D). In Saskatchewan, FSAs (S0G and S0K) highlighted neighbourhoods in the two major cities (Saskatoon and Regina) and farming communities located in between (figure 4A). Other significant FSAs are presented in figure 4A–L.
We dissected in this study the epidemiology of uveal melanoma in Canada and presented a detailed account of uveal melanoma incidence on the provincial, municipal and FSA levels. The epidemiological results detailed in this study closely mirror the findings for uveal melanoma in the USA and other developed countries. For instance, the uvea is the primary site for the majority of ocular melanoma cases in Canada, comprising 85% of all ocular melanoma cases (82.5% in the USA).17 The mean age-standardised (against world standard) incidence rate of uveal melanoma in Canada from 1992 to 2010 was 3.34 cases per 1 million individuals per year (vs 4.2 in the USA from 1996 and 2000) with slight male predilection (male to female rate ratio of 1.11 in Canada, vs 1.29 in the USA).17 Also, similar to the USA, in Canada uveal melanoma mainly affects individuals who are >60 years of age. The mean age of uveal melanoma diagnosis in Canada of 61.5 years was also consistent with the reported numbers around the world. The mean ages of individuals affected by this cancer in the USA, Austria and Spain were 60, 61.7 and 60.09 years at the time of diagnosis, respectively.18 19
Our geographical analysis identified two provinces with high uveal melanoma incidence (British Columbia and Saskatchewan) and a large proportion of the high-incidence FSAs (63%) were documented in these provinces. The identification of areas of high incidences raises a question whether there are putative external and potentially preventable aetiological triggers that are contributing to uveal melanoma pathogenesis in these areas? Identifying and decreasing exposure to these triggers may help prevent death and suffering from this devastating malignancy in the future.
Several internal and external risk factors for uveal melanoma have been described in literature.3 These include Caucasian race, fair hair/skin complexion, light iris colour, family history of OM and coexistence of specific genetic abnormalities.6 Ambient exposure to artificial UV and/or blue light from arc welding and sunlamps were also reported as risk factors.7 20 Other occupational exposures to chemicals21 have been implicated. Sun exposure was described as an independent risk factor for choroidal and ciliary body melanomas in Australia.22 Indeed, long-term sun exposure and occupational sun exposure increases the risk of uveal melanoma. Farming activities in particular were significantly associated with iris melanoma.22 Different explanations may account for the increased incidence of uveal melanoma in certain regions in Canada. Long-term sun exposure by residents of Canada’s westernmost province, British Columbia, may at least partially account for the increased incidence. A case–control study conducted in the USA reported increased risks of uveal melanoma in sailors, ship officers and fishermen.23 Further, agriculture, forestry and fishing were associated with the elevated relative risk of uveal melanoma.24 British Columbia has one of the largest fishing industries in Canada. It is possible that occupational risks as well as habitual ones that increase total exposure to solar light (sunset viewing, reflection of the sunlight from water, etc) may be contributing to the increased incidence of uveal melanoma in British Columbia. These trends are interesting as more than half of ambient UV exposure comes from light being scattered from water surfaces and clouds. Some authorities question an association between sun exposure and uveal melanoma.22 However, McLaughlin et al reported that iris and ciliary body melanomas were more common in Southern and coastal states than in Northern and land-locked states.17 On the other hand, Saskatchewan’s agriculture and farming industry are significant sources of economic prosperity in the province that earned it the title of the ‘Bread Basket of Canada’. It is possible that exposure to sun in farmers in Saskatchewan may account for increased incidence of uveal melanoma. These explanations remain a speculation, and more research is necessary to corroborate these risk factors in the Canadian population.
This study had several limitations. There are no available data in the examined registries on uveal melanoma patients’ ethnicity and clinical stage at the time of diagnosis. It is established that uveal melanoma incidence in Caucasians is significantly higher than in African-American, Asian and Hispanic individuals. Since we did not have this information, we could not confirm these trends in Canada.
In conclusion, we conducted a detailed epidemiological analysis of uveal melanoma in the Canadian population and for the first time demonstrated previously unrecognised variations in disease rates between different provinces and cities with impressive statistically significant findings of increased incidence in British Columbia and Saskatchewan and geographical clustering of cases as well as areas/cities that had lower uveal melanoma rates. This work is fundamental to the understanding of uveal melanoma in North America and further indicates that environmental factors may play an important role in the pathophysiology of this ocular cancer. Identification of these potentially preventable triggers for uveal melanoma in Canada and around the world may enable us to decrease incidence, suffering and deaths from this malignancy.
We thank all the staff that administer the data registries included in this analysis. The interpretation, analysis and reporting of these data are the sole responsibility of the authors.
Contributors FMG, RD and IVL: Collected, plotted and analysed data, prepared the figures and wrote the article. AZ, LM, JVB, DS and MNB: analysed and interpreted the data and co-wrote the article. ML, FMG and IVL: prepared the revised version of the manuscript. ER: Performed statistical analyses. MNB, DS and IVL: Designed and supervised the study.
Funding This work was supported by the Cole Foundation Grant to IVL, Canadian Dermatology Foundation research grants to DS and IVL and the Fonds de la recherche en santé du Québec (FRSQ# 34753 and 36769) research grants to IVL.
Disclaimer No funding bodies had any role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests None declared.
Patient consent for publication Not required.
Ethics approval The presented study was conducted in accordance with protocols approved by the Social Sciences and Humanities Research Council of Canada (SSHRC) and the Québec Inter-University Centre for Social Statistics (QICSS), respectively, protocol numbers: CISS-RDC-668035 and 13-SSH-MCG-3749-S001. Further, in accordance with the institutional policy, this study received an exemption from the McGill University Research Ethics Board review.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.