Professional Documents
Culture Documents
(2017) GHAZAWI, Feras M. Et Al. Distribution and Clustering of Cutaneous (CTCL) - Cases in Canada During 1992 To 2010 PDF
(2017) GHAZAWI, Feras M. Et Al. Distribution and Clustering of Cutaneous (CTCL) - Cases in Canada During 1992 To 2010 PDF
research-article2017
CMSXXX10.1177/1203475417745825Journal of Cutaneous Medicine and SurgeryGhazawi et al
Abstract
Background: Clustering of patients with cutaneous T-cell lymphoma (CTCL) was reported in several jurisdictions around the
world. This rare cancer is known to affect spouses and in some cases multiple members of the same family. These combined
results suggest the existence of external disease triggers/promoters. We recently conducted the first comprehensive analysis
of CTCL incidence and mortality in Canada, which revealed case clustering in several regions.
Objectives: To extend our previous analysis on CTCL incidence across Canada and to provide all the collected data on
CTCL patient incidence in Canada during the period of 1992 to 2010.
Methods: Clinical parameters for patients with CTCL in Canada were analyzed using 2 independent population-based cancer
registries: Canadian Cancer Registry and Le Registre Québécois du Cancer. The CTCL incidence rates were examined on
different geographical levels, including provinces/territories, cities, and forward sortation areas.
Results: Our findings further corroborate our earlier observations of higher CTCL incidence in Newfoundland and
Labrador, maritime provinces (Nova Scotia and New Brunswick), and prairie provinces (Manitoba and Saskatchewan).
Also, most cities with high CTCL incidence were located in these provinces. Extensive mapping of high-incidence postal
codes supports case clustering in a number of communities that are located in the proximity of industrial centres and
seaports.
Conclusions: Detailed analysis of CTCL incidence in Canada is critical to fully understand the burden of this disease in our
country, to begin the search for a possible external trigger for this lymphoma, and to reform how health care resources are
distributed throughout the country to better serve Canadian patients with CTCL.
Keywords
cutaneous T-cell lymphoma (CTCL), incidence, geographic clustering, industrial exposure, transportation, seaports
discovery process, epidemiologic studies were proven to be viral agents were implicated as possible triggers of CTCL, the
a powerful method that can estimate the occurrence of dis- precise nature of this cancer initiator(s)/promoter(s) remains
eases in a given population. There are several types of epide- largely unknown.16
miologic studies, including cohort, case-control, ecologic, We recently published for the first time a comprehensive
and cluster studies. Notably, cluster studies identify a non- investigation on CTCL epidemiology in Canada, which illus-
random distribution of patients and can implicate environ- trated trends that were similar to the ones observed in the
mental factors as potential causes of a disease. United States and also identified nonrandom geographic clus-
Skin, lungs, and gastrointestinal tract are the 3 main organ tering for this malignancy in our country.17 The incidence
systems that are constantly interfacing with the environment. rates of all CTCL types in Canada were on the rise during
Therefore, it is not surprising that many malignancies in 1992 to 1998 and then stabilized at 11.32 cases per 1 million
these organs are triggered by external and often preventable individuals per year (95% confidence interval [CI], 11.05-
causes such as smoking and asbestos responsible for lung 11.59), consistent with the US rate of ~10.2 cases per 1 mil-
cancers and mesotheliomas, respectively. Similarly, expo- lion population per year.18 Patients’ clinical characteristics
sure to polycyclic aromatic hydrocarbons was implicated as (eg, mean age of ~59.4 years at the time of diagnosis and male
a cause for colon cancers, and Helicobacter pylori is known disease predominance) were also confirmed by this study.
to cause gastric cancers. In recent years, research has estab- Notably, the geographic analysis of CTCL distribution
lished a definitive link between several etiologic agents and across the Canadian provinces revealed nonuniform distribu-
skin cancers. Ultraviolet radiation was proven to cause mela- tion across the country, where Newfoundland and Labrador,
nomas, and viruses including human papillomavirus (HPV) Nova Scotia, New Brunswick, Manitoba, and Saskatchewan
and Merkel cell polyomavirus (MCPyV) were shown to had higher incidence rates than the average rate in Canada,
cause squamous cell and Merkel cell cancers, respectively. while Ontario had the lowest incidence rate for CTCL.
Indeed, skin is a highly complex and dynamic immune organ Furthermore, these trends were also consistent at the city
and contains lymphocytes that protect the body from exter- level, where Moncton, St John’s, Saint John, Winnipeg, and
nal pathogens. Disruption of molecular pathways in skin Regina displayed the highest incidence rates of CTCL in
lymphocytes by bacterial, viral, or environmental factors is Canada, while Ontario/Québec administrative cities, includ-
known to lead to cutaneous lymphomas.2 Not surprisingly, ing Ottawa, Gatineau, and Québec City, reported signifi-
for a number of these malignancies, bacterial or viral agents cantly lower disease rates. Analysis of incidence rates by
were shown to drive carcinogenesis and cancer progression. postal codes (forward sortation areas or FSAs) identified
For instance, recent evidence indicates that primary cutane- several regions with considerably high CTCL incidence.
ous marginal zone B-cell lymphomas can be triggered by These FSAs in many cases were covering industrial regions
Borrelia burgdorferi.2 Also, for a number of systemic T-cell in Canadian cities. In contrast, Ottawa (the city with minimal
lymphomas that often manifest with skin lesions such as industrial presence) contained 3 of 8 in the country statisti-
adult T-cell leukemia/lymphoma or natural killer T-cell lym- cally significant populous FSAs with zero CTCL incidence.
phoma of nasal type, viral etiology and involvement of In the initial study describing this data set, we have pub-
human T-cell lymphotropic virus type 1 (HTLV-1) or Epstein- lished only the statistically significant findings,17 which
Barr virus (EBV), respectively, have been extensively excluded a significant amount of data that may be helpful to
documented.3,4 practicing dermatologists across Canada. In the present
Cutaneous T-cell lymphoma (CTCL) is the most common report, we provide all the collected data on CTCL patient
skin lymphoma and represents a heterogeneous group of incidence in Canada during 1992 to 2010.
non-Hodgkin lymphomas, yet the precise pathogenesis by
which CTCL develops remains poorly elucidated.5-10 Patients
Methods
with CTCL typically present with thickened plaques or
patches that may clinically mimic more common benign der- Methods in the current study are identical to the methods
matoses such as eczema, psoriasis, and drug eruptions.5-10 As described in our previous publication.17 Briefly, we exam-
the malignancy advances, involvement of local and distal ined the data on CTCL incidence using 2 distinct population-
lymph nodes, blood, and other visceral organs may occur, based cancer databases, Canadian Cancer Registry (CCR)
and outcomes can be devastating. and Le Registre Québécois du Cancer (LRQC), for the period
Notably, epidemiologic studies reported nonrandom distri- of 1992 to 2010 using the third edition of the International
bution of patients with CTCL in different parts of the world, Classification of Diseases for Oncology (ICD-O-3) codes for
including Sweden,11 Pittsburgh,12 and Texas.13,14 This cancer 10 CTCL subtypes, as previously reported.17 To conduct this
was reported to occur in married couples and cluster in research, a research ethics board waiver was obtained from
families.15 The nonrandom distribution of patients with CTCL McGill University Health Centre. We interpreted and pre-
suggests that external and potentially preventable risk factors sented the data on patient reporting province, city, and FSA
may exist for this malignancy. While several agents such as (geographical region where postal codes start with the same
hydrochlorothiazide diuretics as well as some bacterial and 3 characters) as previously described.17
Ghazawi et al 3
Figure 2. Changing incidence rates (cases per 1 million individuals per year) for each cutaneous T-cell lymphoma (CTCL) subtype
during 1992 to 2010. Analysis of the CTCL subtype incidence rates over time is shown for the 10 subtypes of CTCL. This figure was
developed based on the data from our previous report.17 For definitions of abbreviations, see Figure 1.
Ghazawi et al 5
Figure 3. Incidence rates of cutaneous T-cell lymphoma (CTCL) (per 1 million individuals per year) in the Canadian provinces during
1992 to 2010. *statistically significant lower incidence rate (P < .05) compared to Canada. **Statistically significant higher incidence rates
(P < .05) compared to Canada. This figure was developed based on the data from our previous report.17
(95% CI, 4.20-8.88) cases per million per year. Complete highlighted,17 3 of the zero-incidence FSAs clustered in
information on CTCL incidence rates for all major Canadian Ottawa, the nation’s capital and main administrative city.
cities is provided in Table 1. In our previous analysis of the distribution of patients
Our previous analysis by FSAs (regions where all postal with CTCL in Canada,17 we noted that in several cases, as
codes start with the same 3 characters) identified select described for the cities of Montréal, Hamilton, Oakville, and
regions in the country where several high-incidence FSAs Winnipeg, statistically significant high-incidence FSAs were
were adjacent. Furthermore, in a number of instances, we located in industrial regions of these cities and/or near impor-
observed an important additional clustering of patients within tant transportation junctions. Furthermore, in several other
the identified FSAs, where CTCL cases were not occurring in instances such as in the town of Englehart, Ontario, patients
a random, “buckshot” distribution but clustered in specific with this rare cancer were residing in a proximity of major
regions within each high-incidence FSA. At that time, only railway stations and transportation ports.17 In this analysis,
the statistically significant FSAs were reported.17 However, we extend our findings and correlate distribution of high-
we believe it is important to provide incidence data for all incidence FSAs near important seaports. As noted for the cit-
analyzed FSAs to the Canadian practitioners (Suppl. Table ies of Charlottetown in Prince Edward Island, Halifax in
S1). Mapping geographically all FSAs that had ≥2-fold higher Nova Scotia, Gaspé in Quebec, Hamilton in Ontario, and
incidence of CTCL further strengthens the clustering trends Vancouver/Vancouver Island in British Columbia, the high-
that were identified in our previous report.17 FSAs with an lighted high-incidence FSAs were located in close proximity
increased incidence of CTCL are geographically depicted in (or overlapping) with marine or ferry terminals (Suppl.
different shades of brown (depending on the fold increase in Figure S1).
incidence rate) in Figure 4. Notably, many of these FSAs
were contiguous as in Newfoundland and Labrador (A0C,
Discussion
A0H, and A0J; A1A, A1C, A1L, and A1M), New Brunswick
(E4P and E4K), Québec (H9S and H4W), Manitoba (R3B In this report, we provided complete data on CTCL incidence
and R3C), and British Columbia (V9K and V9P, V7S and across Canada and reported important trends that further
V7R, and V6L and V6M). strengthen our earlier presented statistically significant clus-
Similarly, we also conducted an incidence rate analysis to tering findings.17 We presented a detailed account of CTCL
identify areas that were spared (ie, zero cases) by CTCL dur- incidence on the provincial, city, and FSA levels and high-
ing 1992 to 2010. In our study, 8 FSAs were reported as sta- lighted concordant trends, where provinces with a high inci-
tistically significant, and most (6/8) were in Ontario.17 Here dence of CTCL had a higher incidence of CTCL in their major
we present all the FSAs with the population of ≥10 000 resi- cities and accounted for a significant number of high-incidence
dents that had zero cases of this skin lymphoma during 1992 FSAs. In many cases, these FSAs clustered together and
to 2010 (Table 2). Consistent with our previously reported painted important major industrial areas in these cities. We
trends, 11 of 22 of these communities were in Ontario, the also highlight that in a number of instances, these high-inci-
province with the lowest CTCL incidence rate; 6 of 22 were dence FSAs were located near seaports and transportation
in Québec; and 3 of 22 were in Alberta. As previously junctions.
6 Journal of Cutaneous Medicine and Surgery 00(0)
Figure 4. (continued)
8 Journal of Cutaneous Medicine and Surgery 00(0)
Figure 4. (continued)
Ghazawi et al 9
Figure 4. Geographic maps illustrating case clustering (low-incidence and high-incidence rates) by forward sortation area (FSA: the first
3 letters and numbers in the postal code). High-incidence FSAs are highlighted in different shades of brown, and low-incidence FSAs are
highlighted in gray. Data are presented in 19 maps (A-S) showing different regions of Canada.
10 Journal of Cutaneous Medicine and Surgery 00(0)
Table 2. List of Populous Forward Sortation Areas (FSAs) in Canada With Zero Incidence of Cutaneous T-Cell Lymphoma During
1992 to 2010.
quantities such as carbon monoxide, sulfur oxides, nitrogen Importantly, this nonrandom distribution of CTCL cases
oxides, volatile hydrocarbons, and low molecular weight does not appear to be correlated to the high density of derma-
polyaromatic hydrocarbons, according to the US National tologists in a given province as many of the provinces with
Toxicology Program.21 Indeed, such emissions from diesel high CTCL incidence have very few dermatologists. In fact,
engines have been directly linked to a number of cancers, the CTCL incidence rates in the provinces of Ontario and
including bladder cancer, leukemia, and non-Hodgkin lym- Québec, where more than 65% of the Canadian dermatologist
phomas.22 Also, in many cases, major railroad lines and workforce is concentrated, were relatively low compared to
highways connect to ports, thereby increasing traffic of haz- the rest of Canada. Similarly, there appears to be no correlation
ardous materials through a given area. The implication in between the presence of multidisciplinary specialized CTCL
carcinogenesis of particulate matters in ambient air adjacent clinics, which are available in Ottawa, Toronto, Québec City,
to industrial regions, seaports, and major highways was eval- Montréal, Calgary, Edmonton, and Vancouver, and increased
uated in a number of studies. Specifically, in Hamilton, incidence of CTCL cases in these cities. In fact, to our knowl-
Ontario, mice that were exposed to ambient air adjacent to edge, such specialized clinics are not yet organized in New
the industrial part of Hamilton, which includes steel mills, a Brunswick, Newfoundland and Labrador, Saskatchewan, or
highway, and a seaport, were shown to develop many heri- Manitoba, where an unusually high incidence of CTCL was
table mutations at repetitive DNA loci.23,24 These mutations observed. These data bring to the forefront the need to redis-
were not observed in similar mice that were reared 30 km tribute available health care resources to ensure that patients
away in rural Ontario or in mice housed in the same facilities and physicians, who treat them, have adequate access to spe-
but whose surrounding air was purified by a high-efficiency cialized diagnostic tools (eg, flow cytometry, immunopheno-
particulate air (HEPA) filtration system. Therefore, it is pos- typing, T-cell receptor clonality analysis) and treatments (eg,
sible that ambient pollutant air exposure from one or more extracorporeal photophoresis, psoralen and ultraviolet A
sources can be directly implicated in lymphomagenesis. It is [PUVA] therapy) in these cities. We must also ensure that indi-
likely, however, that a combination of several factors, includ- viduals in these underserved communities have adequate edu-
ing environmental exposures, among others, may increase a cation and support resources to battle their skin lymphomas.
person’s chance of developing CTCL, and it is important to It is notable that the average age of CTCL diagnosis in
investigate the identity of such factors. African American and Hispanic patients is significantly
Ghazawi et al 11
younger than in white individuals.25-29 Furthermore, African 3. Suzuki R. Pathogenesis and treatment of extranodal natural
American men typically present with a higher clinical stage killer/T-cell lymphoma. Semin Hematol. 2014;51(1):42-51.
at the time of diagnosis and follow a more aggressive clinical 4. Tsukasaki K, Tobinai K. Human T-cell lymphotropic virus type
course.13,25,29 Unfortunately, we were not able to analyze data I–associated adult T-cell leukemia-lymphoma: new directions
in clinical research. Clin Cancer Res. 2014;20(20):5217-5225.
on ethnic background of Canadian patients as the CCR/
5. Litvinov IV, Cordeiro B, Fredholm S, et al. Analysis of STAT4
LRQC databases do not provide/release these data.
expression in cutaneous T-cell lymphoma (CTCL) patients and
Finally, parallels can be drawn in medicine and oncology patient-derived cell lines. Cell Cycle. 2014;13(18):2975-2982.
for other rare cancers, where, due to increased incidence in 6. Litvinov IV, Cordeiro B, Huang Y, et al. Ectopic expression of
certain populations and/or geographic areas, it was possible cancer testis antigens in cutaneous T-cell lymphoma (CTCL)
to identify a definitive trigger. For instance, after studying patients. Clin Cancer Res. 2014;20(14):3799-3808.
the prevalence of mesotheliomas in asbestos mines of South 7. Litvinov IV, Jones DA, Sasseville D, Kupper TS.
Africa and in Québec, Canada, asbestos was established as a Transcriptional profiles predict disease outcome in patients
critical factor responsible for this deadly disease. Similarly, a with cutaneous T-cell lymphoma. Clin Cancer Res.
study of a small arsenic mining town in Prussia in 1898, 2010;16(7):2106-2114.
where chronic poisoning took place through the use of con- 8. Litvinov IV, Netchiporouk E, Cordeiro B, et al. The use of
transcriptional profiling to improve personalized diagnosis
taminated drinking water, helped establish the link between
and management of cutaneous T-cell lymphoma (CTCL). Clin
arsenic and the occurrence of arsenical keratoses and skin
Cancer Res. 2015;21(12):2820-2829.
squamous cell carcinomas. Therefore, it appears likely that 9. Litvinov IV, Netchiporouk E, Cordeiro B, et al. Ectopic
future epidemiologic studies that are similar to this one will expression of embryonic stem cell and other developmen-
reveal regional clustering of patients and may lead to an tal genes in cutaneous T-cell lymphoma. Oncoimmunology.
identification of environmental triggers for CTCL and poten- 2014;3(11):e970025.
tially other rare cancers. 10. Litvinov IV, Tetzlaff MT, Thibault P, et al. Gene expression
analysis in cutaneous T-cell lymphomas (CTCL) highlights
Acknowledgments disease heterogeneity and potential diagnostic and prognostic
indicators. Oncoimmunology. 2017;6(5):e1306618.
We thank all the staff that administered the data registries included
11. Gip L, Nilsson E. Clustering of mycosis fungoides in the
in this analysis. The interpretation, analysis, and reporting of these
County of Vasternorrland. Lakartidningen. 1977;74(12):1174-
data are the sole responsibility of the authors.
1176.
12. Moreau JF, Buchanich JM, Geskin JZ, Akilov OE, Geskin LJ.
Declaration of Conflicting Interests Non-random geographic distribution of patients with cutane-
The author(s) declared no potential conflicts of interest with respect ous T-cell lymphoma in the Greater Pittsburgh Area. Dermatol
to the research, authorship, and/or publication of this article. Online J. 2014;20(7).
13. Litvinov IV, Tetzlaff MT, Rahme E, et al. Identification of
geographic clustering and regions spared by cutaneous T-cell
Funding lymphoma in Texas using 2 distinct cancer registries. Cancer.
The author(s) disclosed receipt of the following financial support 2015;121(12):1993-2003.
for the research, authorship, and/or publication of this article: This 14. Litvinov IV, Tetzlaff MT, Rahme E, et al. Demographic
work was supported by the new investigator funding program from patterns of cutaneous T-cell lymphoma incidence in Texas
the Ottawa Hospital Research Institute to Dr Litvinov and the based on two different cancer registries. Cancer Med.
Canadian Dermatology Foundation research grants to Dr Sasseville 2015;4(9):1440-1447.
and Dr Litvinov, Joan Sealy Trust Cancer Research to Dr Litvinov, 15. Hodak E, Klein T, Gabay B, et al. Familial mycosis fungoides:
and the Fonds de la recherche en santé du Québec (FRSQ) research report of 6 kindreds and a study of the HLA system. J Am Acad
grant to Dr Sasseville (FRSQ 22648) and to Dr Litvinov (FRSQ Dermatol. 2005;52(3, pt 1):393-402.
34753 and 36769). 16. Litvinov IV, Shtreis A, Kobayashi K, et al. Investigating
potential exogenous tumor initiating and promoting factors for
ORCID iD cutaneous T-cell lymphomas (CTCL), a rare skin malignancy.
Oncoimmunology. 2016;5(7):e1175799.
Steven Glassman http://orcid.org/0000-0002-9699-303X
17. Ghazawi FM, Netchiporouk E, Rahme E, et al. Comprehensive
Kevin Pehr http://orcid.org/0000-0002-2950-4603
analysis of cutaneous T-cell lymphoma (CTCL) incidence and
mortality in Canada reveals changing trends and geographic clus-
References tering for this malignancy. Cancer. 2017;123(18):3550-3567.
1. Anand P, Kunnumakkara AB, Sundaram C, et al. Cancer is 18. Korgavkar K, Xiong M, Weinstock M. Changing incidence
a preventable disease that requires major lifestyle changes. trends of cutaneous T-cell lymphoma. JAMA Dermatol.
Pharm Res. 2008;25(9):2097-2116. 2013;149(11):1295-1299.
2. Bogle MA, Riddle CC, Triana EM, Jones D, Duvic M. 19. Ziyath AM, Egodawatta P, Goonetilleke A. Build-up of toxic
Primary cutaneous B-cell lymphoma. J Am Acad Dermatol. metals on the impervious surfaces of a commercial seaport.
2005;53(3):479-484. Ecotoxicol Environ Saf. 2016;127:193-198.
12 Journal of Cutaneous Medicine and Surgery 00(0)
20. Thurston GD, Ito K, Lall R. A source apportionment of U.S. 26. Jenni D, Karpova MB, Seifert B, et al. Primary cutaneous
fine particulate matter air pollution. Atmospheric Environ. lymphoma: two-decade comparison in a population of 263
2011;45(24):3924-3936. cases from a Swiss tertiary referral centre. Br J Dermatol.
21. Sharma DC. Ports in a storm. Environ Health Perspect.
2011;164(5):1071-1077.
2006;114(4):A222-A231. 27. Kim YH, Liu HL, Mraz-Gernhard S, Varghese A, Hoppe
22. Boffetta P, Dosemeci M, Gridley G, Bath H, Moradi T,
RT. Long-term outcome of 525 patients with mycosis
Silverman D. Occupational exposure to diesel engine emis- fungoides and Sezary syndrome: clinical prognostic fac-
sions and risk of cancer in Swedish men and women. Cancer tors and risk for disease progression. Arch Dermatol.
Causes Control. 2001;12(4):365-374. 2003;139(7):857-866.
23. Somers CM, McCarry BE, Malek F, Quinn JS. Reduction of 28. Agar NS, Wedgeworth E, Crichton S, et al. Survival out-
particulate air pollution lowers the risk of heritable mutations comes and prognostic factors in mycosis fungoides/Sezary
in mice. Science. 2004;304(5673):1008-1010. syndrome: validation of the revised International Society for
24. Somers CM, Yauk CL, White PA, Parfett CL, Quinn JS. Air Cutaneous Lymphomas/European Organisation for Research
pollution induces heritable DNA mutations. Proc Natl Acad and Treatment of Cancer staging proposal. J Clin Oncol.
Sci USA. 2002;99(25):15904-15907. 2010;28(31):4730-4739.
25. Wilson LD, Hinds GA, Yu JB. Age, race, sex, stage, and inci- 29. Sun G, Berthelot C, Li Y, et al. Poor prognosis in non-Cauca-
dence of cutaneous lymphoma. Clin Lymphoma Myeloma sian patients with early-onset mycosis fungoides. J Am Acad
Leuk. 2012;12(5):291-296. Dermatol. 2009;60(2):231-235.