studies br studies br Human herpesvirus
Human herpesvirus, type 8 IFA Necessary cause
Primary effusion lymphoma (9678) IFA Part of diagnostic
Human T-cell lymphotropic virus, type 1
PCR Necessary cause
Adult T-cell leukemia and lymphoma (9827)
Abbreviations: EBER ISH = EBV-encoded RNA in situ hybridization, PCR = polymerase chain reaction, LMP1 = latent member protein 1,
IHC = immunohistochemistry, CI = confidence interval, PAR = population attributable risk, ENKTL = extranodal natural killer T-cell lymphoma,
IFA = immunofluorescent assays.
d Included studies can be found in the supplement under their respective infection and cancers.
hepatocellular carcinomas were attributable to chronic HBV infection (Supplement, Table S2). Chronic HCV had higher attributable percen-tages than HBV, ranging from 2.5 to 30.0% (Supplement, Table S4). However, the percent of non-Hodgkin lymphoma attributable to HCV was negligible (< 0.7%) for each age and sex group.
The proportion of attributable cancers in Canada in 2015 ranged from a low of 0.4% for HCV in non-Hodgkin lymphoma to a high of 100.0% for HPV in cervical cancer. Cervical cancer was one of five cancers where all cases are attributable to an infection. With few ex-ceptions (HCV in non-Hodgkin lymphoma, and HPV in the oral cavity and larynx), all the calculated PARs exceeded 25.0%, thereby demon-strating the important role that AUY-922 infections play in certain malignancies.
We found that the burden of infection-caused cancers was higher among women (4.0%) than men (3.5%), largely because of HPV's role in cervical and other anogenital cancers. Estimates for the United Kingdom also demonstrated a higher attributable proportion among women than men (3.7% versus 2.5%, respectively) in 2011 (Parkin, 2011) and a similar finding was found in Australia where 2.4% of cancers diagnosed among men in 2010 were attributed to infections and 3.7% among women (Antonsson et al., 2015). In contrast, an analysis for the USA found that 3.3% among both men and women were attri-butable to infections in 2014 (Islami et al., 2018).
As PAR estimates assume causality between the exposure and out-come, we included only established carcinogens and cancers where the evidence for the role of the infection was deemed ‘sufficient’ by the IARC (except for HPV16 in laryngeal cancer). Yet, there is increasing
Summary of the number of cases and proportion of cancers attributable to infections in Canada in 2015
Infection, cancer(s) Total
Obs casesa ACb % Attributablec Obs cases AC % Attributable Obs cases AC % Attributable
Stomach, MALT lymphoma
Adult T-cell leukemia and lymphoma
Abbreviations: Obs = observed, AC = attributable cases, MALT = mucosa-associated lymphoid tissue, ENKTL = extranodal natural killer T-cell lymphoma.
a Cancer incidence data for the year 2015 from the Canadian Cancer Registry. Quebec's cancer incidence was estimated. Hepatocellular carcinoma incidence was estimated by applying the proportion 0.797 to liver cancer incidence. b Number of cancer cases at individual cancer sites that can be attributed to infection. c Proportion attributable was calculated by dividing the number of cases attributable to infection by the number of the associated cancer cases. It differs from PAR which for some cancers varied by age and/or sex.
d All associated cancers includes all cancers known to be associated with infections listed in the table. g All cancers includes all incident cancer cases in Canada among those 18 and older in 2015. h Includes the base of the tongue and tonsils.
evidence that other infection cancer associations including EBV in gastric carcinoma, HBV in non-Hodgkin lymphoma and HCV in cho-langiocarcinoma, among others, may also cause cancer. If these asso-ciations were included, the impact of infections on cancer incidence would have been higher than what we reported here.
The combined impact of the hepatitis viruses resulted in 27.4% of hepatocellular carcinoma incidence being attributable to HBV/HCV. Since HBV can be avoided with vaccination that began in Canada in the early 1980s, and because HCV can be prevented through a variety of behavioral interventions and treated with direct-acting antivirals, the future burden of hepatocellular carcinoma has the potential to decrease by reducing the prevalence of these viruses.
Globally, H. pylori was responsible for 89% of non-cardia gastric cancers (Plummer et al., 2015). We calculated that 68.8% of incident non-cardia gastric cancers in Canada were due to this infection. We estimated PARs based on elimination of the infection. This information is helpful for understanding the impact of infections on cancer in-cidence; however, in practice, elimination may not be entirely feasible. For example, H. pylori can be treated with quadruple antibiotic therapy, but challenges in the scalability of screening for the infection and concerns over antibiotic resistance limit the prospect of eliminating the infection at the population level (Bourke et al., 2005; Hunt et al., 2004; Fallone et al., 2016).