PURPOSE:Cancer outcomes in sub-Saharan Africa are driven by delayed diagnosis and treatment initiation. We evaluated the magnitude and determinants of diagnostic and treatment delays among cancer patients in Kinshasa, Democratic Republic of the Congo (DRC). METHODS:We conducted a hospital-based cross-sectional study of 460 adults with confirmed cancer at Nganda Hospital Center in Kinshasa, DRC. Two outcomes were assessed: delay from symptom onset to diagnosis and delay from diagnosis to treatment initiation. Log-normal regression models were fitted for each outcome to estimate adjusted geometric mean ratios (aGMRs) and 95% confidence intervals (CIs). Covariates included demographic, socioeconomic, clinical, behavioral, and stigma-related factors. RESULTS:The median age was 55 years, and 76.2% of participants were women. Overall, 55.0% of participants experienced symptom-to-diagnosis delays >6 months, and 49.4% experienced diagnosis-to-treatment delays >3 months. Older age was associated with longer diagnostic delay (aGMR 1.55, 95% CI 1.03-2.31) and treatment delay (1.51, 1.07-2.14). Unemployment was strongly associated with both diagnostic delay (1.68, 1.15-2.47) and treatment delay (2.27, 1.54-3.33), as was hepatitis B co-infection (1.88, 1.06-3.34 and 2.42, 1.15-5.11, respectively). Longer diagnostic delay was additionally associated with informal trading (1.99, 1.21-3.28), taxi or motorbike transport (1.92, 1.25-2.94), and smoking history (2.25, 1.03-4.91), while high cancer-stereotype stigma was associated with longer treatment delay (1.56, 1.04-2.34). CONCLUSION:Substantial delays exist across the DRC cancer care continuum, driven by socioeconomic vulnerability, transport barriers, hepatitis B co-infection, and cancer-related stigma. These findings highlight the need for integrated interventions to improve timely diagnosis and treatment initiation, including strengthening financial protection, decentralizing cancer services, and reducing stigma in cancer care.

The relationship between body mass index (BMI) and melanoma and other skin cancers remains unclear. The objective of this study was to employ the Mendelian randomization (MR) approach to evaluate the effects of genetically predicted childhood adiposity on the risk of developing skin cancer later in life. Two-sample MR analyses were conducted using summary data from genome-wide association study (GWAS) meta-analyses of childhood BMI, melanoma, cutaneous squamous cell carcinoma (cSCC), and basal cell carcinoma (BCC). We used the inverse-variance-weighted (IVW) methods to obtain a pooled estimate across all genetic variants for childhood BMI. We performed multiple sensitivity analyses to evaluate the potential influence of various assumptions on our findings. We found no evidence that genetically predicted childhood BMI was associated with risks of developing melanoma, cSCC, or BCC in adulthood (OR, 95% CI: melanoma: 1.02 (0.93-1.13), cSCC 0.94 (0.79-1.11), BCC 0.97 (0.84-1.12)). Our findings do not support the conclusions from observational studies that childhood BMI is associated with increased risks of melanoma, cSCC, or BCC in adulthood. Intervening on childhood adiposity will not reduce the risk of common skin cancers later in life.

Assessing the risk of cancer among people living with HIV (PLHIV) in the current era of antiretroviral therapy (ART) is crucial, given their increased susceptibility to many types of cancer and prolonged survival due to ART exposure. Our study aims to compare the association between HIV infection and specific cancer sites in Rwanda. Population-based cancer registry data were used to identify cancer cases in both PLHIV and HIV-negative persons. A probabilistic record linkage approach between the HIV and cancer registries was used to supplement HIV status ascertainment in the cancer registry. Associations between HIV infection and different cancer types were evaluated using unconditional logistic regression models. We performed several sensitivity analyses to assess the robustness of our findings and to evaluate the potential impact of different assumptions on our results. From 2007 to 2018, the cancer registry recorded 17,679 cases, of which 7% were diagnosed among PLHIV. We found significant associations between HIV infection and Kaposi's Sarcoma (KS) (adjusted odds ratio [OR]: 29.1, 95% CI: 23.2-36.6), non-Hodgkin lymphoma (NHL) (1.6, 1.3-2.0), Hodgkin lymphoma (HL) (1.6, 1.1-2.4), cervical (2.3, 2.0-2.7), vulvar (4.0, 2.5-6.5), penile (3.0, 2.0-4.5), and eye cancers (2.2, 1.6-3.0). Men living with HIV had a higher risk of anal cancer (3.1, 1.0-9.5) than men without HIV, but women living with HIV did not have higher risk than women without HIV (1.0, 0.2-4.3). Our study found that in an era of expanded ART coverage in Rwanda, HIV is associated with a broad range of cancers, particularly those linked to viral infections.

Male-pattern baldness (MPB) is related to dysregulation of androgens such as testosterone. A previously observed relationship between MPB and skin cancer may be due to greater exposure to ultraviolet radiation or indicate a role for androgenic pathways in the pathogenesis of skin cancers. We dissected this relationship via Mendelian randomization (MR) analyses, using genetic data from recent male-only meta-analyses of cutaneous melanoma (12,232 cases; 20,566 controls) and keratinocyte cancers (KCs) (up to 17,512 cases; >100,000 controls), followed by stratified MR analysis by body-sites. We found strong associations between MPB and the risk of KC, but not with androgens, and multivariable models revealed that this relationship was heavily confounded by MPB single nucleotide polymorphisms involved in pigmentation pathways. Site-stratified MR analyses revealed strong associations between MPB with head and neck squamous cell carcinoma and melanoma, suggesting that sun exposure on the scalp, rather than androgens, is the main driver. Men with less hair covering likely explains, at least in part, the higher incidence of melanoma in men residing in countries with high ambient UV.

OBJECTIVES:To determine the proportions of newly diagnosed melanomas treated by different medical specialist types, to describe the types of excisions performed, and to investigate factors associated with treating practitioner specialty and excision type.DESIGN, SETTING:Prospective cohort study; analysis of linked data: baseline surveys, hospital, pathology, Queensland Cancer Register, and Medical Benefits Schedule databases.PARTICIPANTS:Random sample of 43 764 Queensland residents aged 40-69 years recruited during 2011, with initial diagnoses of in situ or invasive melanoma diagnosed to 31 December 2019.MAIN OUTCOME MEASURES:Treating practitioner type and treatment modality for first incident melanoma; second and subsequent treatment events for the primary melanoma.RESULTS:During a median follow-up of 8.4 years (interquartile range, 8.3-8.8 years), 1683 eligible participants (720 women, 963 men) developed at least one primary melanoma (in situ melanoma, 1125; invasive melanoma, 558), 1296 of which (77.1%) were initially managed in primary care; 248 were diagnosed by dermatologists (14.8%), 83 by plastic surgeons (4.9%), 43 by general surgeons (2.6%), and ten by other specialists (0.6%). The most frequent initial procedures leading to histologically confirmed melanoma diagnosis were first excision (854, 50.7%), shave biopsy (549, 32.6%), and punch biopsy (178, 10.6%); 1339 melanomas (79.6%) required two procedures, 187 (11.1%) three. Larger proportions of melanomas diagnosed by dermatologists (87%) or plastic surgeons (71%) were in people living in urban areas than of those diagnosed in primary care (63%); larger proportions of melanomas diagnosed by dermatologists or plastic surgeons than of those diagnosed in primary care were in people with university degrees (45%, 42% v 23%) or upper quartile clinical risk scores (63%, 59% v 47%).CONCLUSIONS:Most incident melanomas in Queensland are diagnosed in primary care, and nearly half are initially managed by partial excision (shave or punch biopsy). Second or third, wider excisions are undertaken in about 90% of cases.

Observational studies have suggested that smoking may increase the risk of cutaneous squamous cell carcinoma (cSCC) while decreasing the risks of basal cell carcinoma (BCC), and melanoma. However, it remains possible that confounding by other factors may explain these associations. The aim of this investigation was to use Mendelian randomization (MR) to test whether smoking is associated with skin cancer, independently of other factors. Two-sample MR analyses were conducted to determine the causal effect of smoking measures on skin cancer risk using genome-wide association study (GWAS) summary statistics. We used the inverse-variance-weighted estimator to derive separate risk estimates across genetic instruments for all smoking measures. A genetic predisposition to smoking initiation was associated with lower risks of all skin cancer types, although none of the effect estimates reached statistical significance (OR 95% CI BCC 0.91, 0.82–1.01; cSCC 0.82, 0.66–1.01; melanoma 0.91, 0.82–1.01). Results for other measures were similar to smoking initiation with the exception of smoking intensity which was associated with a significantly reduced risk of melanoma (OR 0.67, 95% CI 0.51–0.89). Our findings support the findings of observational studies linking smoking to lower risks of melanoma and BCC. However, we found no evidence that smoking is associated with an elevated risk of cSCC; indeed, our results are most consistent with a decreased risk, similar to BCC and melanoma.

Journal of the European Academy of Dermatology and VenereologyVolume 37, Issue 6 p. e792-e795 LETTER TO THE EDITOR Cholesterol-lowering genetic variants are not associated with the risk of skin cancer Jean Claude Dusingize, Corresponding Author Jean Claude Dusingize [email protected] orcid.org/0000-0001-5721-100X Departments of Population Health and Computational Biology, QIMR Berghofer Medical Research Institute, Brisbane, Queensland, Australia Correspondence Jean Claude Dusingize, Departments of Population Health and Computational Biology, Cancer Control Group, QIMR Berghofer Medical Research Institute, Locked Bag 2000, Royal Brisbane and Women's Hospital, Queensland 4029, Australia. Email: [email protected]Search for more papers by this authorCatherine M. Olsen, Catherine M. Olsen Departments of Population Health and Computational Biology, QIMR Berghofer Medical Research Institute, Brisbane, Queensland, Australia Faculty of Medicine, The University of Queensland, Brisbane, AustraliaSearch for more papers by this authorMatthew H. Law, Matthew H. Law Departments of Population Health and Computational Biology, QIMR Berghofer Medical Research Institute, Brisbane, Queensland, Australia Faculty of Medicine, The University of Queensland, Brisbane, Australia School of Biomedical Sciences, Faculty of Health, Queensland University of Technology, Brisbane, Queensland, AustraliaSearch for more papers by this authorNirmala Pandeya, Nirmala Pandeya Departments of Population Health and Computational Biology, QIMR Berghofer Medical Research Institute, Brisbane, Queensland, Australia School of Public Health, University of Queensland, Brisbane, AustraliaSearch for more papers by this authorRachel E. Neale, Rachel E. Neale Departments of Population Health and Computational Biology, QIMR Berghofer Medical Research Institute, Brisbane, Queensland, Australia School of Public Health, University of Queensland, Brisbane, AustraliaSearch for more papers by this authorStuart MacGregor, Stuart MacGregor Departments of Population Health and Computational Biology, QIMR Berghofer Medical Research Institute, Brisbane, Queensland, Australia Faculty of Medicine, The University of Queensland, Brisbane, AustraliaSearch for more papers by this authorDavid C. Whiteman, David C. Whiteman Departments of Population Health and Computational Biology, QIMR Berghofer Medical Research Institute, Brisbane, Queensland, Australia School of Public Health, University of Queensland, Brisbane, AustraliaSearch for more papers by this authorJue-Sheng Ong, Jue-Sheng Ong Departments of Population Health and Computational Biology, QIMR Berghofer Medical Research Institute, Brisbane, Queensland, AustraliaSearch for more papers by this author Jean Claude Dusingize, Corresponding Author Jean Claude Dusingize [email protected] orcid.org/0000-0001-5721-100X Departments of Population Health and Computational Biology, QIMR Berghofer Medical Research Institute, Brisbane, Queensland, Australia Correspondence Jean Claude Dusingize, Departments of Population Health and Computational Biology, Cancer Control Group, QIMR Berghofer Medical Research Institute, Locked Bag 2000, Royal Brisbane and Women's Hospital, Queensland 4029, Australia. Email: [email protected]Search for more papers by this authorCatherine M. Olsen, Catherine M. Olsen Departments of Population Health and Computational Biology, QIMR Berghofer Medical Research Institute, Brisbane, Queensland, Australia Faculty of Medicine, The University of Queensland, Brisbane, AustraliaSearch for more papers by this authorMatthew H. Law, Matthew H. Law Departments of Population Health and Computational Biology, QIMR Berghofer Medical Research Institute, Brisbane, Queensland, Australia Faculty of Medicine, The University of Queensland, Brisbane, Australia School of Biomedical Sciences, Faculty of Health, Queensland University of Technology, Brisbane, Queensland, AustraliaSearch for more papers by this authorNirmala Pandeya, Nirmala Pandeya Departments of Population Health and Computational Biology, QIMR Berghofer Medical Research Institute, Brisbane, Queensland, Australia School of Public Health, University of Queensland, Brisbane, AustraliaSearch for more papers by this authorRachel E. Neale, Rachel E. Neale Departments of Population Health and Computational Biology, QIMR Berghofer Medical Research Institute, Brisbane, Queensland, Australia School of Public Health, University of Queensland, Brisbane, AustraliaSearch for more papers by this authorStuart MacGregor, Stuart MacGregor Departments of Population Health and Computational Biology, QIMR Berghofer Medical Research Institute, Brisbane, Queensland, Australia Faculty of Medicine, The University of Queensland, Brisbane, AustraliaSearch for more papers by this authorDavid C. Whiteman, David C. Whiteman Departments of Population Health and Computational Biology, QIMR Berghofer Medical Research Institute, Brisbane, Queensland, Australia School of Public Health, University of Queensland, Brisbane, AustraliaSearch for more papers by this authorJue-Sheng Ong, Jue-Sheng Ong Departments of Population Health and Computational Biology, QIMR Berghofer Medical Research Institute, Brisbane, Queensland, AustraliaSearch for more papers by this author First published: 20 January 2023 https://doi.org/10.1111/jdv.18886Read the full textAboutPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onEmailFacebookTwitterLinkedInRedditWechat REFERENCES 1Lomas A, Leonardi-Bee J, Bath-Hextall F. A systematic review of worldwide incidence of nonmelanoma skin cancer. Br J Dermatol. 2012; 166(5): 1069–80. 10.1111/j.1365-2133.2012.10830.x CASPubMedWeb of Science®Google Scholar 2Yang K, Marley A, Tang H, Song Y, Tang JY, Han J. Statin use and non-melanoma skin cancer risk: a meta-analysis of randomized controlled trials and observational studies. Oncotarget. 2017; 8(43): 75411–7. 10.18632/oncotarget.20034 PubMedGoogle Scholar 3Bonovas S, Nikolopoulos G, Filioussi K, Peponi E, Bagos P, Sitaras NM. Can statin therapy reduce the risk of melanoma? A meta-analysis of randomized controlled trials. Eur J Epidemiol. 2010; 25(1): 29–35. 10.1007/s10654-009-9396-x CASPubMedWeb of Science®Google Scholar 4Burgess S, Butterworth A, Thompson SG. Mendelian randomization analysis with multiple genetic variants using summarized data. Genet Epidemiol. 2013; 37(7): 658–65. 10.1002/gepi.21758 PubMedWeb of Science®Google Scholar 5Willer CJ, Schmidt EM, Sengupta S, Peloso GM, Gustafsson S, Kanoni S, et al. Discovery and refinement of loci associated with lipid levels. Nat Genet. 2013; 45(11): 1274–83. 10.1038/ng.2797 CASPubMedWeb of Science®Google Scholar 6Liyanage UE, Law MH, Han X, An J, Ong JS, Gharahkhani P, et al. Combined analysis of keratinocyte cancers identifies novel genome-wide loci. Hum Mol Genet. 2019; 28(18): 3148–60. 10.1093/hmg/ddz121 CASPubMedWeb of Science®Google Scholar 7Landi MT, Bishop DT, MacGregor S, Machiela MJ, Stratigos AJ, Ghiorzo P, et al. Genome-wide association meta-analyses combining multiple risk phenotypes provide insights into the genetic architecture of cutaneous melanoma susceptibility. Nat Genet. 2020; 52(5): 494–504. 10.1038/s41588-020-0611-8 CASPubMedWeb of Science®Google Scholar 8Dickerman BA, García-Albéniz X, Logan RW, Denaxas S, Hernán MA. Avoidable flaws in observational analyses: an application to statins and cancer. Nat Med. 2019; 25(10): 1601–6. 10.1038/s41591-019-0597-x CASPubMedWeb of Science®Google Scholar 9Yu WY, Hill ST, Chan ER, Pink JJ, Cooper K, Leachman S, et al. Computational drug repositioning identifies statins as modifiers of prognostic genetic expression signatures and metastatic behavior in melanoma. J Invest Dermatol. 2021; 141(7): 1802–9. 10.1016/j.jid.2020.12.015 CASPubMedWeb of Science®Google Scholar 10von Schuckmann LA, Khosrotehrani K, Ghiasvand R, Hughes MCB, van der Pols JC, Malt M, et al. Statins may reduce disease recurrence in patients with ulcerated primary melanoma. Br J Dermatol. 2020; 183(6): 1049–55. 10.1111/bjd.19012 PubMedWeb of Science®Google Scholar Volume37, Issue6June 2023Pages e792-e795 ReferencesRelatedInformation

BACKGROUND Skin screening is associated with higher melanoma detection rates, a potential indicator of overdiagnosis, but it remains possible that this effect is due to confounding by genetic risk. OBJECTIVES To compare melanoma incidence among screened vs. unscreened participants within tertiles of genetic risk. METHODS We investigated melanoma incidence in the QSkin Study, a prospective cohort study which for this analysis comprised 15,283 participants aged 40-69 years with genotype data and no prior history of melanoma. We calculated a polygenic score (PGS) for melanoma. We first calculated age-standardised rate (ASR) of melanoma within PGS tertiles, and then measured the association between skin examination and melanoma detection by calculating the hazard ratio (HR) and 95% confidence interval (95% CI), overall and within PGS tertiles. RESULTS Melanoma incidence increased with PGS (ASR/100000/yr) tertile 1: 442; tertile 2: 519; tertile 3: 871). We found that the hazard ratios for all melanomas (i.e. in situ and invasive) associated with skin examination differed slightly across PGS tertiles (age- and sex-adjusted tertile 1 HR 1.88, 95% CI 1.26-2.81; tertile 2 HR 1.70, 95% CI 1.20-2.41; tertile 3 HR 1.96, 95% CI 1.43-2.70; fully adjusted tertile 1 HR 1.14, 95% CI 0.74-1.75; tertile 2 HR 1.21, 95% CI 0.82-1.78; tertile 3 HR 1.41, 95% CI 1.00-1.98) but these differences were not statistically significant. Hazard ratios for in situ melanoma associated with skin examination were similar across PGS tertiles. For invasive melanomas, the point estimates appeared highest in PGS tertile 3 in both minimally adjusted (age, sex) and fully adjusted models, however these apparent differences were also not statistically significant. CONCLUSIONS Genetic risk predicts subsequent melanoma incidence, and is weakly associated with screening behaviour, but does not explain the higher rate of melanoma detection between screened and unscreened people.

Numerous epidemiologic studies have reported positive associations between higher nevus counts and internal cancers. Whether this association represents a true relationship or is due to bias or confounding by factors associated with both nevus counts and cancer remains unclear. We used germline genetic variants for nevus count to test whether this phenotypic trait is a risk-marker for cancer. We calculated polygenic risk scores (PRS) for nevus counts using individual-level data in the UK Biobank (n = 394 306) and QSkin cohort (n = 17 427). The association between the nevus PRS and each cancer site was assessed using logistic regression adjusted for the effects of age, sex and the first five principal components. In both cohorts, those in the highest nevus PRS quartile had higher risks of melanoma than those in the lowest quartile (UK Biobank odds ratio [OR] 1.42, 95% confidence interval [CI]: 1.29-1.55; QSkin OR 1.58, 95% CI: 1.29-1.94). We also observed increases in risk of basal cell carcinoma (BCC) and squamous cell carcinoma (SCC) associated with higher nevus PRS quartiles (BCC UK Biobank OR 1.38, 95% CI: 1.33-1.44; QSkin OR 1.20, 95% CI: 1.05-1.38 and SCC UK Biobank OR 1.41, 95% CI: 1.28-1.55; QSkin OR 1.44, 95% CI: 1.19-1.77). We found no consistent evidence that nevus count PRS were associated with risks of developing internal cancers. We infer that associations between nevus counts and internal cancers reported in earlier observational studies arose because of unmeasured confounding or other biases.

Nevi (moles) are collections of nondendritic melanocytes in the skin that form darkly pigmented spots. Nevi-prone individuals are at a higher risk of melanoma. Mean telomere length is a highly heritable trait that has been associated with many diseases, including melanoma (Butt et al., 2010Butt H.Z. Atturu G. London N.J. Sayers R.D. Bown M.J. Telomere length dynamics in vascular disease: a review.Eur J Vasc Endovasc Surg. 2010; 40: 17-26Abstract Full Text Full Text PDF PubMed Scopus (77) Google Scholar; Han et al., 2009Han J. Qureshi A.A. Prescott J. Guo Q. Ye L. Hunter D.J. et al.A prospective study of telomere length and the risk of skin cancer.J Invest Dermatol. 2009; 129: 415-421Abstract Full Text Full Text PDF PubMed Scopus (146) Google Scholar; Rachakonda et al., 2018Rachakonda S. Srinivas N. Mahmoudpour S.H. Garcia-Casado Z. Requena C. Traves V. et al.Telomere length and survival in primary cutaneous melanoma patients [published correction appears in Sci Rep 2018;8:17963].Sci Rep. 2018; 8: 10947Crossref PubMed Scopus (17) Google Scholar). Observational studies have found an association between shorter telomere length and reduced nevus count, whereas others report null findings (Bataille et al., 2007Bataille V. Kato B.S. Falchi M. Gardner J. Kimura M. Lens M. et al.Nevus size and number are associated with telomere length and represent potential markers of a decreased senescence in vivo.Cancer Epidemiol Biomarkers Prev. 2007; 16: 1499-1502Crossref PubMed Scopus (95) Google Scholar; Han et al., 2009Han J. Qureshi A.A. Prescott J. Guo Q. Ye L. Hunter D.J. et al.A prospective study of telomere length and the risk of skin cancer.J Invest Dermatol. 2009; 129: 415-421Abstract Full Text Full Text PDF PubMed Scopus (146) Google Scholar; Li et al., 2016Li X. Liang G. Du M. De Vivo I. Nan H. No association between telomere length-related loci and number of cutaneous nevi.Oncotarget. 2016; 7: 82396-82399Crossref PubMed Scopus (4) Google Scholar). To resolve the uncertainty around a role for telomeres in nevi development, we used Mendelian randomization (MR). MR uses single-nucleotide variations (SNVs) that have been shown by GWASs to be associated with the trait of interest as instrumental variables (IVs) to infer causality between two traits. To determine whether genetically predicted telomere length is associated with nevus count, the IVs' exposure effect sizes (i.e., β) are compared with their effect size on the outcome. Causality can be inferred with MR analysis primarily because the direction of effect is clear–from IV to trait and not the reverse (Davey Smith and Hemani, 2014Davey Smith G. Hemani G. Mendelian randomization: genetic anchors for causal inference in epidemiological studies.Hum Mol Genet. 2014; 23: R89-R98Crossref PubMed Scopus (1369) Google Scholar). Valid MR requires certain assumptions to be met; these are detailed in the Supplementary Materials and Methods. To construct telomere length IVs, we used GWAS data, consisting of 472,174 individuals (of European descent) from the UK Biobank (Codd et al., 2021Codd V. Wang Q. Allara E. Musicha C. Kaptoge S. Stoma S. et al.Polygenic basis and biomedical consequences of telomere length variation.Nat Genet. 2021; 53: 1425-1433Crossref PubMed Scopus (33) Google Scholar). Telomere length measurements have been described in detail previously (Codd et al., 2021Codd V. Wang Q. Allara E. Musicha C. Kaptoge S. Stoma S. et al.Polygenic basis and biomedical consequences of telomere length variation.Nat Genet. 2021; 53: 1425-1433Crossref PubMed Scopus (33) Google Scholar). For nevus count, we obtained β and standard errors from a 2018 GWAS meta-analysis of nevus density (Duffy et al., 2018Duffy D.L. Zhu G. Li X. Sanna M. Iles M.M. Jacobs L.C. et al.Novel pleiotropic risk loci for melanoma and nevus density implicate multiple biological pathways [published correction appears in Nat Commun 2019;10:299].Nat Commun. 2018; 9: 4774Crossref PubMed Scopus (51) Google Scholar) that included a total of 52,506 individuals of European descent. We also included 12,930 unrelated participants from the QSkin Sun and Health Study (Olsen et al., 2012Olsen C.M. Green A.C. Neale R.E. Webb P.M. Cicero R.A. Jackman L.M. et al.Cohort profile: the QSkin Sun and Health Study.Int J Epidemiol. 2012; 41 (929–929i)Crossref Scopus (103) Google Scholar) and 314 additional members of the Brisbane Twin Nevus Morphology Study (Lee et al., 2016Lee S. Duffy D.L. McClenahan P. Lee K.J. McEniery E. Burke B. et al.Heritability of naevus patterns in an adult twin cohort from the Brisbane Twin Registry: a cross-sectional study.Br J Dermatol. 2016; 174: 356-363Crossref PubMed Scopus (15) Google Scholar), boosting the total sample size to 65,777. To meet the strong instrument MR assumption (assumption 1; Supplementary Materials and Methods), 197 SNVs associated with telomere length that exceeded the genome-wide significance threshold (P = 5 × 10–8) were selected from Codd et al., 2021Codd V. Wang Q. Allara E. Musicha C. Kaptoge S. Stoma S. et al.Polygenic basis and biomedical consequences of telomere length variation.Nat Genet. 2021; 53: 1425-1433Crossref PubMed Scopus (33) Google Scholar. After linkage disequilibrium clumping, filtering, and identification of suitable proxy SNVs (Supplementary Results), 114 SNVs remained that were used as IVs in this MR analysis; these IVs captured 2.8% of telomere length variance (Supplementary Table S1). All the five MR regression models used (adjusting for varying levels of pleiotropy; Supplementary Materials and Methods) indicated that genetically predicted longer telomeres were significantly associated with higher nevus count (P < 0.0002; Table 1 and Figure 1). The MR Egger intercept was not significantly different from null (intercept = –0.0009, 95% confidence interval = –0.0019 to +0.00007, P = 0.072), suggesting that there was little evidence for pleiotropy that would otherwise violate the third assumption of MR. There was no evidence for heterogeneity among the βs of the IVs (Cochran's Q statistic = 125, P = 0.2).Table 1MR Effect Estimates for Effect of Telomere Length on Nevus CountMethodβSEP-ValueMR Egger0.060.0210.0058Weighted median0.0210.0080.0073Inverse variance weighted0.0230.0060.0002Simple mode0.130.0440.0035Weighted mode0.020.0070.0055Abbreviations: IV, instrumental variable; MR, Mendelian randomization; SE, standard error.The table shows the individual effect estimates (i.e., β), SE, and P-values for each MR method (Hemani et al., 2018Hemani G. Zheng J. Elsworth B. Wade K.H. Haberland V. Baird D. et al.The MR-base platform supports systematic causal inference across the human phenome.Elife. 2018; 7: e34408Crossref PubMed Google Scholar); these include inverse variance weighted; MR Egger, which adjusts for horizontal pleiotropic effects for the IVs; weighted median, which adjusts for up to 50% of the IVs having a pleiotropic effect; simple mode, which is the mode of the Wald-type estimates used to calculate the inverse variance weighted; and weighted mode, which assigns a SE-based weighting to the simple mode. Open table in a new tab Abbreviations: IV, instrumental variable; MR, Mendelian randomization; SE, standard error. The table shows the individual effect estimates (i.e., β), SE, and P-values for each MR method (Hemani et al., 2018Hemani G. Zheng J. Elsworth B. Wade K.H. Haberland V. Baird D. et al.The MR-base platform supports systematic causal inference across the human phenome.Elife. 2018; 7: e34408Crossref PubMed Google Scholar); these include inverse variance weighted; MR Egger, which adjusts for horizontal pleiotropic effects for the IVs; weighted median, which adjusts for up to 50% of the IVs having a pleiotropic effect; simple mode, which is the mode of the Wald-type estimates used to calculate the inverse variance weighted; and weighted mode, which assigns a SE-based weighting to the simple mode. To assess whether the genes affecting nevus count affected telomere length, potentially driving a false positive result, 13 genome-wide significant SNVs were selected from the nevus count GWAS as IVs and used in an MR performed as stated earlier; this showed no significant association (inverse variance‒weighted β = 0.037, standard error = 0.028, P = 0.19). Because this reverse MR result is null, it provides reassurance that the effect of telomere length on nevus count is unidirectional (telomere influence nevi and not the reverse) and also that the causal association is not due to a pleiotropic effect whereby another (possibly unknown) factor affects both nevus count and telomere length. We used MR to explore the hypothesis that genetically predicted longer telomere length causes an increase in nevus count. Accounting for the large size of the telomere GWAS, this MR analysis is equivalent to the size of an observational study of ∼13,220 individuals with telomere measurements. We are confident that our analysis satisfies the three core assumptions of MR (Supplementary Materials and Methods). Furthermore, both the sensitivity analysis (Supplementary Table S2) and the reverse test are reassuring that the results are not reflecting a type 1 error. To convert this to a more recognizable scale, we used a subset of QSkin participants who reported a count of moles >2 mm on their upper left arm as a benchmark; in this subset, a 1 SD change was 7 moles. When the MR analysis was performed with only this QSkin subset, β was 0.13 (Supplementary Table S3), indicating that a 1 SD increase in genetically predicted telomere length results in approximately 0.91 additional moles >2 mm on the upper left arm. Approximately 80% of melanocytes carry a key oncogenic BRAF or NRAS mutation (Ghanadan et al., 2021Ghanadan A. Yousefi T. Kamyab-Hesari K. Azhari V. Nasimi M. Prevalence and main determinants of BRAF V600E mutation in dysplastic and congenital nevi.Iran J Pathol. 2021; 16: 51-56Crossref PubMed Scopus (3) Google Scholar). There has been speculation that individuals with longer telomeres exhibit a delayed cell senescence period, during which time in the absence of cell senescence signal and B-Raf/NRAS oncogenic signaling, nondendritic melanocytes can proliferate, increasing both the size and number of nevi (Han et al., 2009Han J. Qureshi A.A. Prescott J. Guo Q. Ye L. Hunter D.J. et al.A prospective study of telomere length and the risk of skin cancer.J Invest Dermatol. 2009; 129: 415-421Abstract Full Text Full Text PDF PubMed Scopus (146) Google Scholar). It has also been suggested that delayed senescence may also reduce the apoptotic ability of melanocytes, prohibiting the natural reduction of nevi with age (Bataille et al., 2007Bataille V. Kato B.S. Falchi M. Gardner J. Kimura M. Lens M. et al.Nevus size and number are associated with telomere length and represent potential markers of a decreased senescence in vivo.Cancer Epidemiol Biomarkers Prev. 2007; 16: 1499-1502Crossref PubMed Scopus (95) Google Scholar). There are limitations to be considered in this study. First, the nevus count GWAS was derived from a range of studies with differing sample sizes and different methods of measuring nevus count, leading to technical as well as sample size‒related sources of variation. Although this will likely not lead to a false positive in the results presented in this paper, it may influence the magnitude of the effect estimate. Another limitation is that our nevus count GWAS was a meta-analysis of a meta-analysis; however, we obtained consistent results when using the smaller QSkin nevus count GWAS (Supplementary Table S3), indicating that the combination of the initial 2018 meta-analysis with additional QSkin samples improved power rather than leading to a type 1 error. We have shown evidence supportive of a causal association between genetically predicted longer telomeres and an increased number of nevi. Although the observational evidence was inconsistent, our MR analysis allows us to confirm the causative direction of this association. Telomere length and nevus count are both risk factors for melanoma; whether telomere length and nevi act synergistically or independently on melanoma risk is the focus of our ongoing research. See Supplementary Table S1 for all SNVs used in this analysis, along with effect sizes, and standard errors. This study primarily used publicly available data. This study uses data from previous published studies and biobanks. Below is a brief overview detailing data collection procedures in each study. QSkin: Participants signed consent forms giving permission for approved researchers to use information provided in the survey and permitted data linkage to cancer registries, pathology laboratories, public hospital databases and Medicare Australia (Olsen et al., 2012Olsen C.M. Green A.C. Neale R.E. Webb P.M. Cicero R.A. Jackman L.M. et al.Cohort profile: the QSkin Sun and Health Study.Int J Epidemiol. 2012; 41 (929–929i)Crossref Scopus (103) Google Scholar). The study protocol was approved by QIMR Berghofer ethics committee. UK Biobank: All participants gave broad consent for the use of data and samples for any health-related research and for UK Biobank to access their health-related records. All data is anonymised. Ethics approval was granted by the UK Biobank ethics committee. Brisbane Twin Nevus Morphology Study: Data from twins and their siblings from Brisbane, Australia, was collected between the years 1992 to 2016. All participants gave informed consent for participation in this study, and the study protocol was approved by QIMR Berghofer ethics committee (Duffy et al., 2010Duffy D.L. Iles M.M. Glass D. Zhu G. Barrett J.H. Höiom V. et al.IRF4 variants have age-specific effects on nevus count and predispose to melanoma.Am J Hum Genet. 2010; 87: 6-16Abstract Full Text Full Text PDF PubMed Scopus (101) Google Scholar). Nathan Ingold: http://orcid.org/0000-0002-9668-6275 Jean Claude Dusingize: http://orcid.org/0000-0001-5721-100X Racheal E. Neale: http://orcid.org/0000-0001-7162-0854 Catherine M. Olsen: http://orcid.org/0000-0003-4483-1888 David C. Whiteman: http://orcid.org/0000-0003-2563-9559 David L. Duffy: http://orcid.org/0000-0001-7227-632X Stuart MacGregor: http://orcid.org/0000-0001-6731-8142 Matthew H. Law: http://orcid.org/0000-0002-4303-8821 DCW and SM were supported by grants from the National Health and Medical Research Council of Australia. SM, DCW, JCD and CMO were supported by the Australian National Health and Medical Research Council (NHMRC) Fellowships - grants APP1123248, APP1116360, APP1150144, APP1023911, APP1185416 and APP1155413. The QSkin Study is supported by grants from the NHMRC [APP1185416] [APP1073898] [APP1063061]. NI received philanthropic support from the Laurence Edward Wilkins Foundation PhD Scholarship. For acknowledgements for the nevus count GWAS and funding see Duffy et al., 2018Duffy D.L. Zhu G. Li X. Sanna M. Iles M.M. Jacobs L.C. et al.Novel pleiotropic risk loci for melanoma and nevus density implicate multiple biological pathways [published correction appears in Nat Commun 2019;10:299].Nat Commun. 2018; 9: 4774Crossref PubMed Scopus (51) Google Scholar. We would like to acknowledge the work of John Pearson, Scott Wood, and Scott Gordon for their technical and computational support. This work used the UK Biobank Resource (application number 25331) as an LD reference panel. This work was carried out in Brisbane, Australia. Conceptualization: NI, JCD; Data Curation: REN, DCW, CMO, DLD; Formal Analysis: NI; Supervision: SM, MHL; Writing - Original Draft Preparation: NI; Writing - Review and Editing: NI, JCD, REN, DCW, CMO, DLD, SM, MHL To meet the strong instrument Mendelian randomization (MR) assumption (assumption 1; see Supplementary Materials and Methods), 197 SNVs associated with telomere length that exceed the genome-wide significance threshold (P = 5 × 10–8) were selected from Codd et al., 2021Codd V. Wang Q. Allara E. Musicha C. Kaptoge S. Stoma S. et al.Polygenic basis and biomedical consequences of telomere length variation.Nat Genet. 2021; 53: 1425-1433Crossref PubMed Scopus (78) Google Scholar. A total of 133 of these SNVs were also present in Duffy et al., 2018Duffy D.L. Zhu G. Li X. Sanna M. Iles M.M. Jacobs L.C. et al.Novel pleiotropic risk loci for melanoma and nevus density implicate multiple biological pathways [published correction appears in Nat Commun 2019;10:299].Nat Commun. 2018; 9: 4774Crossref PubMed Scopus (63) Google Scholar nevus count GWAS (Supplementary Materials and Methods). We identified additional 25 SNVs (Supplementary Table S1) that were present in the nevus count GWAS with a linkage disequilibrium (LD) r2 > 0.98 to use as proxies for missing SNVs. Clumping (r2 = 0.01, kb = 1,000) removed 28 SNVs. Five SNVs with allele frequencies close to 0.5 could not be unambiguously matched for strand owing to their alleles (A/T, G/C) and were removed. A total of 11 pleiotropic SNVs were identified by MR Pleiotropy RESidual Sum and Outlier and were removed. This left a total of 114 SNVs to be used as instrumental variables (IVs) in the MR analysis. The IVs were selected to surpass genome-wide significance (P < 5 × 10–8) for telomere length (MR assumption 1). Ensuring that both cohorts used were filtered to only individuals of European descent and appropriate adjustments were made to account for residual population stratification (e.g., fitting principal components) reduced the possibility of a false-positive result (second assumption of MR). MR Egger, a pleiotropically robust model, revealed an intercept that was not significantly different from zero, revealing no significant pleiotropic effect (third assumption of MR). In this analysis, we used the same 114 telomere length IVs as in the main analysis. However, in this analysis, we used the subset of nevus count GWAS to only include the QSkin partition of the GWAS meta-analysis. This QSkin data were from 13,692 individuals who provided a self-reported count of moles on the upper left arm. The distribution of this trait was inverse rank normalized before analysis. There was one SNV, which was present in the telomere IVs but not in the QSkin nevus GWAS; it was removed, leaving a total of 113 IVs. MR results found Supplementary Table S3. Using a GWAS of nevus count on a known scale helps to translate the MR effect size into a reliable quantitative measure of the number of moles increased per SD increase in telomere length. However, the power in this analysis is greatly reduced owing to the small sample size relative to that of the full nevus GWAS meta-analysis. Owing to the inverse variance‒ranked transformation performed on both the main GWAS meta-analysis used in the primary analysis and that used in the QSkin subset used in this study, we can be confident that both GWAS have a similar distribution of nevi and are therefore on a similar scale. In this analysis, we used a different telomere GWAS to select IVs. To extract the genetic IVs for our analysis, we used (Li et al., 2020Li C. Stoma S. Lotta L.A. Warner S. Albrecht E. Allione A. et al.Genome-wide association analysis in humans links nucleotide metabolism to leukocyte telomere length.Am J Hum Genet. 2020; 106: 389-404Abstract Full Text Full Text PDF PubMed Scopus (82) Google Scholar) GWAS of telomere length in leukocytes. This study consisted of 78,592 individuals of European descent. This study measured telomere length using a qPCR method, previously described (Cawthon, 2002Cawthon R.M. Telomere measurement by quantitative PCR.Nucleic Acids Res. 2002; 30: e47Crossref PubMed Scopus (2619) Google Scholar). There were 18 telomere-associated SNVs that were genome-wide significant from this GWAS (clumped LD r2 = 0.01, kb = 1,000) that were also present in the nevus count GWAS. Five SNVs were identified by MR Pleiotropy RESidual Sum and Outlier as potential pleiotropic outliers and hence were removed from the analysis; this left 13 SNVs to construct the IVs. MR results found Supplementary Table S2 To calculate the telomere length variance explained by each IV used in this analysis, the following equation was used:R2=2β2EAF1−EAF2β2EAF1−EAF+se22NEAF1−EAF where β is the effect of the SNV, EAF is the effect allele frequency, N is the total sample size, and se is the standard error of the beta. For MR to be valid, three key assumptions are required to be met, listed as follows:1.The IVs are associated with the risk factor of interest.2.The IVs share no common cause with the outcome.3.The IVs do not affect the outcome except through the risk factor. Assumption 1 can be met using SNVs that surpass genome-wide significance (P < 5 × 10–8) from GWAS. Assumption 2 can refer to an unmeasured artifact driving a false-positive association between exposure and outcome. For example, this could be population stratification within both cohorts used that have a shared genetic architecture different from that of the rest of the cohort; therefore, it is important to use cohorts that have been filtered to contain only one genetic ancestry, for example, European descent. Assumption 3 is referring to the pleiotropic effect of SNVs and whether the SNVs used are also acting through an alternative pathway, which may be a confounding factor affecting both exposure and outcome (Burgess et al., 2013Burgess S. Butterworth A. Thompson S.G. Mendelian randomization analysis with multiple genetic variants using summarized data.Genet Epidemiol. 2013; 37: 658-665Crossref PubMed Scopus (1546) Google Scholar). For the MR analysis, five main estimators were calculated: (i) inverse variance weighted, which is a multiplicative random-effects model that combines a Wald-type estimator for each IV; (ii) MR Egger, which although similar to the inverse variance weighted, additionally adjusts for the horizontal pleiotropic effect of the IVs; (iii) weighted median, which adjusts for up to 50% of the IVs having a pleiotropic effect; (iv) simple mode, which is the mode of the Wald-type estimates used to calculate the inverse variance weighted; and (v) weighted mode, which is similar to the simple mode but assigns a standard error‒based weighting to each IV (Bowden et al., 2015Bowden J. Davey Smith G. Burgess S. Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression.Int J Epidemiol. 2015; 44: 512-525Crossref PubMed Scopus (2608) Google Scholar; Burgess et al., 2013Burgess S. Butterworth A. Thompson S.G. Mendelian randomization analysis with multiple genetic variants using summarized data.Genet Epidemiol. 2013; 37: 658-665Crossref PubMed Scopus (1546) Google Scholar). MR was performed using R-3.6.2, with R packages TwoSampleMR and MRInstruments from the MR-Base platform (Hemani et al., 2018Hemani G. Zheng J. Elsworth B. Wade K.H. Haberland V. Baird D. et al.The MR-base platform supports systematic causal inference across the human phenome.Elife. 2018; 7: e34408Crossref PubMed Scopus (1999) Google Scholar). Filtering (known informally as clumping) of the lead telomere SNVs from Codd et al., 2021Codd V. Wang Q. Allara E. Musicha C. Kaptoge S. Stoma S. et al.Polygenic basis and biomedical consequences of telomere length variation.Nat Genet. 2021; 53: 1425-1433Crossref PubMed Scopus (78) Google Scholar for strict independence with minimal LD was performed in PLINK.1.90 beta (Purcell et al., 2007Purcell S. Neale B. Todd-Brown K. Thomas L. Ferreira M.A. Bender D. et al.PLINK: a tool set for whole-genome association and population-based linkage analyses.Am J Hum Genet. 2007; 81: 559-575Abstract Full Text Full Text PDF PubMed Scopus (20981) Google Scholar). An LD reference panel of 4,990 individuals from the UK Biobank and 40 million SNVs was used. Any SNV that was either within 1,000 kb or had an LD r2 > 0.01 with another SNV was removed. The method of ascertaining nevus count in Duffy et al., 2018Duffy D.L. Zhu G. Li X. Sanna M. Iles M.M. Jacobs L.C. et al.Novel pleiotropic risk loci for melanoma and nevus density implicate multiple biological pathways [published correction appears in Nat Commun 2019;10:299].Nat Commun. 2018; 9: 4774Crossref PubMed Scopus (63) Google Scholar meta-analysis varied between studies: continuous or categorical rating measurements, self-report, or clinically recorded, with measures from either selected regions or across the whole body. Within the QSkin sample, nevus count was self-reported, categorical nevus count (none, few, some, and many). Details of nevus count analysis and confirmation that the differing methods can be combined in a GWAS meta-analysis have been previously reported (Duffy et al., 2018Duffy D.L. Zhu G. Li X. Sanna M. Iles M.M. Jacobs L.C. et al.Novel pleiotropic risk loci for melanoma and nevus density implicate multiple biological pathways [published correction appears in Nat Commun 2019;10:299].Nat Commun. 2018; 9: 4774Crossref PubMed Scopus (63) Google Scholar).Supplementary Table S1All SNVs Included Were Used to Construct the IVs in the AnalysisSNVA1A2Exposure: Telomere LengthOutcome: Nevus CountFreqBetaSEP-ValueFreqβSEP-Valuers66731853GA0.6830.0180.0023.30E–160.6860.00180.0060.769rs17185038CG0.9350.0250.0041.20E–090.9460.01410.0090.105rs6669563GA0.562–0.0180.0038.90E–190.5620.00380.0070.565rs3768321GT0.810.0150.0035.00E–090.8180.00560.0080.462rs41269079TA0.811–0.0150.0033.60E–090.812–0.00280.0080.715rs139795227キAC0.986–0.0610.0094.90E–120.984–0.10510.0380.006rs4498805GT0.453–0.0150.0021.00E–130.483–0.00190.0040.616rs11579626AC0.915–0.0250.0041.00E–110.9080.00790.0110.469rs61818036GA0.1760.020.0033.40E–120.19–0.00410.0080.617rs932002CT0.8490.0390.0032.40E–400.8320.02350.0090.006rs56178008#TA0.563–0.0140.0033.80E–120.574–0.00180.0020.455rs12615793GA0.86–0.0450.0031.00E–510.872–0.02280.0080.005rs12613375CT0.862–0.0190.0031.40E–100.8580.00690.0070.303rs869785TC0.3280.0150.0029.20E–120.335–9.00E–040.0030.781rs78491606AC0.9820.0730.0071.70E–220.9810.00280.0350.936rs13062095TC0.672–0.0140.0025.70E–100.685–9.00E–040.0010.076rs6776756GA0.4020.0180.0022.70E–170.3840.00030.0060.963rs41272947GA0.4580.0160.0021.90E–150.4570.00620.0040.119rs146546514CA0.983–0.0820.0093.80E–220.982–0.01790.0360.619rs871134CT0.4310.0180.0022.20E–180.447–0.00410.0060.517rs13129697TG0.721–0.0170.0025.20E–130.7080.00270.0060.656rs4695407AG0.492–0.0140.0032.40E–120.521–0.00250.0030.344rs7705526CA0.673–0.0790.0029.10E–2720.6710.00190.0080.813rs115451758GA0.990.0850.0103.70E–170.992–0.04080.0480.398rs61748181CT0.9710.0590.0061.40E–220.9670.01140.0250.652rs79717857キCA0.974–0.0510.0073.20E–140.982–0.09670.030.001rs72801474GA0.9120.0210.0042.60E–090.9230.00280.0110.805rs34255404キGA0.943–0.0380.0051.10E–160.945–0.05010.0160.002rs80324517GA0.952–0.040.0051.10E–160.943–0.00410.0020.089rs154979CA0.971–0.0440.0063.90E–130.9820.0070.0210.744rs9398196AG0.480.0140.0021.10E–110.480.00210.0020.241rs13230646キTC0.7510.0170.0032.30E–120.749–2.00E–040.00010.047rs11769630TA0.9280.0260.0044.80E–110.9270.00770.0120.511rs2538745TC0.3970.0120.0024.80E–090.3910.00210.0020.239rs2056726GA0.7860.0230.0031.70E–190.7840.01280.0070.081rs7790856CT0.7110.0430.0021.80E–800.7120.02320.0070.0005rs117811540GA0.992–0.1450.0121.30E–330.994–0.05960.0420.153rs4731541CG0.3750.020.0021.60E–210.3820.00310.0060.63rs11556924CT0.624–0.0130.0025.60E–100.6370.00510.0050.303rs1985369AG0.1320.0310.0032.40E–230.131–1.00E–040.0150.995rs2306646#GC0.4410.0210.0024.20E–240.4430.00020.0060.972rs762679TA0.143–0.0310.0034.30E–260.1380.00770.010.448rs7012816GA0.87–0.0170.0034.40E–080.857–0.00330.0090.708rs10112752GA0.570.0290.0029.90E–440.5760.00170.0020.473rs1023767GA0.7620.0180.0024.30E–140.7240.00830.0070.255rs4742448#CG0.531–0.0160.0025.50E–140.519–9.00E–040.0070.894rs11557154CT0.870.0350.0033.60E–300.8570.00420.010.678rs4743037CT0.769–0.0160.0031.40E–100.768–0.00330.0040.393rs12572897GA0.870.0340.0034.00E–270.8720.01150.0090.215rs11190184GC0.7160.0170.0033.50E–130.727–0.00150.0010.198rs4919611CA0.1130.0270.0031.00E–160.130.01390.0090.136rs9419958TC0.1390.0810.0036.40E–1550.1550.03040.0090.001rs939916GA0.33–0.0240.0026.30E–260.327–0.00940.0070.167rs1609812GA0.16–0.0460.0036.30E–590.179–0.00140.0010.152rs10840270CG0.344–0.0150.0021.40E–110.3823.00E–040.0030.931rs2293579GA0.6140.0130.0022.90E–100.618–0.00840.0060.183rs611646#TA0.5910.0380.0031.30E–710.565–0.01440.0060.022rs6590343AG0.484–0.0120.0026.50E–090.4770.00520.0060.354rs10845387GA0.6470.0140.0024.60E–110.6650.00550.010.565rs12369950TC0.8590.0180.0032.50E–090.839–0.00880.0090.322rs79977579CA0.904–0.0290.0044.80E–170.907–0.00640.0120.585rs1907702キGA0.233–0.0140.0031.20E–080.225–0.04490.0084.8E-09rs10774625AG0.476–0.0160.0024.30E–140.462–0.01320.0070.043rs76666449TC0.899–0.030.0041.20E–180.901–7.00E–040.00040.079rs4758644キAC0.2690.0170.0036.90E–130.277–9.00E–040.0010.082rs1727302GA0.259–0.020.0033.50E–170.266–0.00980.0070.165rs35017269GA0.984–0.0650.0091.10E–140.985–0.01940.0350.578rs3093888GA0.9490.0290.0051.10E–090.945–0.00420.0130.739rs73581419CT0.893–0.0240.0037.60E–130.8920.01670.010.087rs12884911キCT0.4970.0140.0023.00E–110.486–0.00090.00050.072rs762810CA0.6480.020.0023.90E–200.6660.0030.0060.638rs1957937AT0.84–0.020.0037.80E–130.842–0.00150.0090.861rs17677991CG0.658–0.0220.0027.80E–240.6540.00220.0070.74rs1980240AC0.595–0.0130.0024.70E–100.61–0.00370.0030.211rs5742915TC0.554–0.020.0027.80E–230.573–0.00230.0060.719rs80116508GA0.9380.0330.0052.00E–140.9360.00220.0020.174rs11646283TC0.589–0.0150.0022.40E–120.586–0.00320.0070.628rs182059586TC0.9750.0590.0078.60E–180.9760.00830.0380.828rs8053839GT0.4580.0150.0025.90E–130.4850.00290.0020.241rs76219171GA0.942–0.0360.0044.70E–160.9430.02240.0150.132rs28711261キAG0.867–0.0180.0035.90E–090.864–0.03570.010.0003rs34003787CT0.9120.0240.0041.80E–100.9190.01630.0110.151rs183553155GA0.989–0.0720.0112.00E–120.99–0.02070.0470.662rs11866592GA0.858–0.0340.0033.70E–310.8680.00950.0090.282rs2303262CT0.2230.0470.0031.80E–810.2150.01630.0070.025rs7218033CT0.7470.0230.0035.10E–210.7240.00690.0070.298rs4724GA0.8830.0560.0032.10E–640.875–0.00720.0090.446rs75664430CG0.7520.0230.0032.30E–220.7550.00960.0070.175rs111527438TC0.649–0.0130.0021.70E–090.6750.00320.0080.686rs12941945AG0.8320.0260.0037.10E–210.821–0.00190.0090.826rs2069536AG0.250.0150.0032.20E–100.245–0.00190.0070.787rs144204502CT0.9870.1010.0093.40E–270.9850.05470.0380.153rs3891167AG0.7470.0440.0033.70E–710.7540.03180.0090.0003rs116863223GA0.9880.0820.0094.10E–180.990.01090.0420.795rs78694226GA0.9920.0660.0121.60E–080.992–0.02740.060.65rs79824385TA0.87–0.0320.0034.30E–260.8590.00040.010.968rs8088824CT0.237–0.0250.0035.60E–250.230.00310.0070.671rs2276182#CG0.597–0.0240.0022.60E–300.573–0.01680.0060.007rs6565924AG0.639–0.0120.0024.10E–080.651–0.00750.0060.231rs35601737キCG0.7040.0140.0021.80E–100.7190.01920.0110.083rs8105767AG0.705–0.0330.0031.90E–480.705–7.00E–040.0040.874rs4530278GT0.402–0.0130.0029.60E–100.3860.0010.0020.677rs429358TC0.846–0.0170.0032.30E–090.859–0.00460.0090.604rs11084431GA0.3960.0120.0022.60E–080.4110.00030.0060.962rs8102497GA0.5680.0150.0028.00E–130.568–0.00370.0040.358rs1291143AC0.151–0.0490.0031.20E–630.156–0.0060.0090.498rs2259797TC0.9070.0830.0042.50E–1150.8950.01570.0130.243rs41308088CT0.919–0.0470.0041.60E–340.913–0.00920.0160.567rs35640778GA0.9790.210.0077.00E–1900.9830.03790.0190.052rs28502153CA0.6220.0220.0033.60E–240.629–0.00740.0070.286rs17464525∗GA0.820.0330.0032.90E–350.8150.00710.0070.339rs9752727∗CT0.5710.0150.0022.40E–120.5490.00490.0050.347rs2370835∗CT0.4090.0270.0024.20E–380.4170.01030.0060.097rs9999831∗TC0.8860.0240.0047.30E–120.867–0.00210.010.837rs9999362∗AG0.6110.0140.0021.00E–100.627–0.00460.0060.47rs28840093∗TC0.235–0.0530.0032.50E–1020.2290.00250.0020.311rs7737331∗CT0.853–0.0230.0031.30E–140.8460.00280.010.768rs2763979∗CT0.6410.0280.0024.20E–380.613–0.00710.0070.309rs688242∗CA0.618–0.0130.0025.50E–100.62–0.00250.0030.435rs7818283∗TC0.1150.030.0041.20E–190.111–0.0040.010.689rs10795541∗CT0.3990.0190.0027.30E–190.4010.02540.0064.E-05rs9532725∗TC0.18–0.0240.0032.10E–170.185–5.00E–040.00020.011rs59591876∗CT0.9–0.0450.0041.50E–400.897–0.00420.0090.638rs145723797∗AT0.7610.0340.0032.60E–440.7730.00820.0070.258rs3785074∗AG0.711–0.0230.0032.40E–240.74–6.00E–040.0060.925rs9675571∗TC0.1380.0190.0039.30E–100.1460.00990.0090.281rs2350929∗AC0.644–0.0150.0031.70E–120.6270.00350.0070.6rs125∗CT0.6170.0210.0021.50E–220.4650.00730.0060.221rs34101249∗GT0.75–0.0180.0021.90E–120.827–0.01390.010.184rs73153168∗AG0.9–0.0230.0042.90E–100.879–0.02460.010.016Abbreviations: Freq, effect-allele frequency; IV, instrumental variable; RSID, reference SNP cluster ID; SE, standard error.The RSIDs (build 37) for all the IVs used in the main analysis along with the effect allele (A1), noneffect allele (A2), Freq, effect size (β), SE, and P-values are reported for telomere length and nevus count. SNVs marked with an ∗ were used as proxy SNV for an SNV that was present in one GWAS but not in the other. Proxy SNVs had an r2 with the original ≥0.98, calculated with PLINK. SNVs marked with a キ were removed because they were identified as pleiotropic outliers; SNVs marked with a # were removed because they were identified as palindromic. Open table in a new tab Supplementary Table S2MR Using Li et al., 2020Li C. Stoma S. Lotta L.A. Warner S. Albrecht E. Allione A. et al.Genome-wide association analysis in humans links nucleotide metabolism to leukocyte telomere length.Am J Hum Genet. 2020; 106: 389-404Abstract Full Text Full Text PDF PubMed Scopus (82) Google Scholar GWAS Telomere SNVs as IVsMethodβSEP-ValueMR Egger–0.0370.0970.708Weighted median0.0090.0070.179IVW0.0090.0070.158Simple mode–0.0070.0310.820Weighted mode0.0090.0060.157Abbreviations: IV, instrumental variable; IVW, inverse variance weighted; MR, Mendelian randomization; SE, standard error.In this table, we have effect sizes (β), SEs, and P-values for five MR tests performed using 13 SNVs to construct the telomere IVs. Although nonsignificant (likely owing to power), the IVW effect size is consistent with the effect seen in the main analysis (which has greater power). Open table in a new tab Supplementary Table S3MR with 113 Telomere IVs Tested in a Subset of only QSkin Nevus Count SampleMethodβSEP-ValueMR Egger0.240.090.01Weighted median0.120.070.10IVW0.130.050.01Simple mode0.150.170.38Weighted mode0.160.100.13Abbreviations: IV, instrumental variable; IVW, inverse variance weighted; MR, Mendelian randomization; SE, standard error.In this table, we have effect sizes (β), SEs, and P-values for five MR tests performed using 113 telomere IVs from the main analysis but in a subset of just the QSkin nevus count data. The effect sizes from IVW have been used to infer a translatable scale for the number of moles increased per one SD change in telomere length (see the main Discussion). Open table in a new tab Abbreviations: Freq, effect-allele frequency; IV, instrumental variable; RSID, reference SNP cluster ID; SE, standard error. The RSIDs (build 37) for all the IVs used in the main analysis along with the effect allele (A1), noneffect allele (A2), Freq, effect size (β), SE, and P-values are reported for telomere length and nevus count. SNVs marked with an ∗ were used as proxy SNV for an SNV that was present in one GWAS but not in the other. Proxy SNVs had an r2 with the original ≥0.98, calculated with PLINK. SNVs marked with a キ were removed because they were identified as pleiotropic outliers; SNVs marked with a # were removed because they were identified as palindromic. Abbreviations: IV, instrumental variable; IVW, inverse variance weighted; MR, Mendelian randomization; SE, standard error. In this table, we have effect sizes (β), SEs, and P-values for five MR tests performed using 13 SNVs to construct the telomere IVs. Although nonsignificant (likely owing to power), the IVW effect size is consistent with the effect seen in the main analysis (which has greater power). Abbreviations: IV, instrumental variable; IVW, inverse variance weighted; MR, Mendelian randomization; SE, standard error. In this table, we have effect sizes (β), SEs, and P-values for five MR tests performed using 113 telomere IVs from the main analysis but in a subset of just the QSkin nevus count data. The effect sizes from IVW have been used to infer a translatable scale for the number of moles increased per one SD change in telomere length (see the main Discussion).

Cutaneous melanomas are common cancers in white‐skinned populations, and early detection is promoted as a means of reducing morbidity and mortality. There is concern that increased skin screening is leading to overdiagnosis of indolent melanomas with low risk of lethality. The extent of melanoma overdiagnosis associated with screening is unknown.

Because little is known about cataract in pilots, we estimated prevalence by anonymously ascertaining all commercial airline pilots diagnosed with cataract 2011-2016 using the electronic Medical Records System of the Australian Civil Aviation Safety Authority. Of 14,163 Australian male commercial pilots licensed in 2011, 1286 aged >= 60 had biennial eye examinations showing a cataract prevalence of 11.6%. Among 12,877 pilots aged <60, based on compulsory eye examinations only when first licensed, prevalence was 0.5%. There was no significant difference by ambient ultraviolet (UV) radiation levels in state of residence though lowest prevalence was seen in the low-UV state of Victoria. Most cataract in pilots >= 60 years was bilateral and of mild severity, while cataract in pilots <60 were more likely to be unilateral and of greater severity.

Abstract Background A personal history of keratinocyte carcinoma (KC) has been reported as a risk factor for developing subsequent primary cutaneous and non-cutaneous malignancies. However most evidence to date stems from observational studies which are prone to bias, confounding and reverse causation. Our aim was to examine this association using different Mendelian randomization (MR) approaches. Methods We performed a one-sample MR analysis using individual-level data from the UK Biobank (n = 394,306). This analysis was then validated in an independent dataset in the QSkin cohort (n = 16,896). Using 64 independent genetic variants known to be associated with KC, we generated a polygenic risk score (PRS) for each participant in the UK Biobank and the QSkin cohort. We then performed two-sample MR analyses using genome-wide association study (GWAS) summary statistics. We tested the association between genetically predicted KC and risk of subsequent cancer using logistic regression. Results Results from one-sample MR analyses in the UK Biobank indicated that a personal history of KC was significantly associated with cancer overall (excluding melanoma) (OR: 1.15, 95% CI: 1.10-1.20, per doubling the prevalence of KC). The results from the two-sample MR corroborate the findings from the one-sample MR, although the risk estimate was lower (OR: 1.05, 95% CI: 1.03-1.07). Conclusions Our MR analyses suggest that genetically predicted KC is a risk factor for developing subsequent primary malignancies. Key messages A personal history of KC may serve as a proxy marker of inherited cancer risk.

Prospectively identifying people at increased risk for melanoma is a key prerequisite to designing interventions aiming to detect melanoma early when treatment has a high probability of being curative. Whereas population-based screening for melanoma is not recommended because there is no evidence that it decreases mortality (Wernli et al., 2016Wernli K.J. Henrikson N.B. Morrison C.C. Nguyen M. Pocobelli G. Blasi P.R. Screening for skin cancer in adults: updated evidence report and systematic review for the US Preventive Services Task Force.JAMA. 2016; 316: 436-447Crossref PubMed Scopus (87) Google Scholar), medical authorities in the United States, Australia, and elsewhere advocate targeted screening of people at a high risk. What is unknown, however, is whether people can accurately assess their own personal risk of developing melanoma. Self assessments in European populations have been shown to be only moderately accurate in identifying people at a high risk, correlating poorly with dermatologists’ evaluations on melanoma risk factors (Carli et al., 2003Carli P. De Giorgi V. Palli D. Maurichi A. Mulas P. Orlandi C. et al.Dermatologist detection and skin self-examination are associated with thinner melanomas: results from a survey of the Italian Multidisciplinary Group on Melanoma.Arch Dermatol. 2003; 139: 607-612Crossref PubMed Scopus (144) Google Scholar; Harbauer et al., 2003Harbauer A. Binder M. Pehamberger H. Wolff K. Kittler H. Validity of an unsupervised self-administered questionnaire for self-assessment of melanoma risk.Melanoma Res. 2003; 13: 537-542Crossref PubMed Scopus (25) Google Scholar; Richtig et al., 2008Richtig E. Santigli E. Fink-Puches R. Weger W. Hofmann-Wellenhof R. Assessing melanoma risk factors: how closely do patients and doctors agree?.Public Health. 2008; 122: 1433-1439Crossref PubMed Scopus (9) Google Scholar). Moreover, there is evidence to suggest that people are not confident in assessing their future risk of melanoma and can be reluctant to seek advice from their treating doctors (Eiser et al., 2000Eiser J.R. Pendry L. Greaves C.J. Melia J. Harland C. Moss S. Is targeted early detection for melanoma feasible? Self assessments of risk and attitudes to screening.J Med Screen. 2000; 7: 199-202Crossref PubMed Scopus (29) Google Scholar). We examined whether the self-assessed risk of developing melanoma accorded with predicted risk, as measured by a risk prediction tool developed and validated in a large, population-based, prospective cohort of Queensland men and women (QSkin) aged 40–69 years at recruitment in 2011 (n = 41,936) (Olsen et al., 2012Olsen C.M. Green A.C. Neale R.E. Webb P.M. Cicero R.A. Jackman L.M. et al.Cohort profile: the QSkin Sun and Health Study.Int J Epidemiol. 2012; 41 (929–929i)Crossref Scopus (73) Google Scholar). As reported previously, we derived the risk prediction tool from a set of 28 candidate factors that were all self reported on the baseline survey (Olsen et al., 2018Olsen C.M. Pandeya N. Thompson B.S. Dusingize J.C. Webb P.M. Green A.C. et al.Risk stratification for melanoma: models derived and validated in a purpose-designed prospective cohort.J Natl Cancer Inst. 2018; 110: 1075-1083Crossref PubMed Scopus (24) Google Scholar). The most parsimonious model for predicting future risk of melanoma (invasive and in situ cases) included 12 terms: age, sex, ethnicity, family history of melanoma, tanning ability, number of nevi at age 21 years, hair color, sunscreen use, number of excisions for skin cancers, number of previous nonsurgical treatments for actinic lesions, history of skin checks by a doctor, and private health insurance. The model showed high discrimination and good calibration in both the development and internal validation samples. Thus, we generated a risk prediction score for each participant and then assigned them to one of the five categories of predicted risk on the basis of quintile cut points of the overall risk distribution (category 1 [very much below average risk] through to category 5 [very much above average risk]). In addition to questions on medical and phenotypic factors outlined above, the baseline survey also included the following question about self-assessed risk: “Compared to other Queenslanders, how likely do you think it is that you will get melanoma at some time in the future?” Possible answers were “Highly unlikely,” “Somewhat unlikely,” “About the same as other Queenslanders,” “Somewhat more likely,” and “Highly likely.” We examined possible discordance between self-assessed risk (as captured by the question above) and predicted risk (as measured by the validated prediction tool comprising 12 factors) of developing melanoma during follow-up by cross classifying people according to their categories of self-assessed versus predicted risk level. Discordance was defined as a difference between self-assessed and predicted risk by two or more categories. We examined the factors associated with discordance using univariable and multivariable regression analyses adjusting for age and sex; these included demographic, phenotypic, and lifestyle factors as well as medical history related to the diagnosis and treatment of skin lesions. The cohort for these analyses included 41,936 participants; we excluded participants with a melanoma diagnosis prior to baseline. Statistical significance was inferred at P < 0.05. All analyses were conducted using SAS 9.4 software (SAS Institute, Cary, NC). The Human Research Ethics Committee at the QIMR Berghofer Medical Research Institute (Brisbane, Queensland, Australia) approved the study, and all QSkin participants gave their written informed consent to take part. Most participants (18,995 or 45.3%) self assessed their risk of developing melanoma as “about the same as other Queenslanders,” 13.1% as “highly unlikely,” 27.7% as “somewhat unlikely,” 10.6% as “somewhat more likely,” and 3.3% as “highly likely.” Most participants (28,593; 68.2%) were not discordant by more than one category; however, 9,375 (22.4%) participants assessed their risk to be at least two categories lower than their risk as predicted by the validated tool, whereas 3,968 (9.5%) participants overestimated their risk category (Figure 1). The characteristics of these three groups are provided in Supplementary Table S1. We focused our analysis on the 22.4% of the sample who underestimated their risk of melanoma because these would be the likely targets for public health interventions. Underestimators were older than participants who scored their risk more accurately (i.e., those who were not discordant by more than one category) (Table 1); those aged >65 years were more likely to be underestimators (risk ratio [RR] = 1.6, 95% confidence interval [CI] = 1.5–1.7). After adjustment for age and sex, underestimators were more likely to have European ancestry (RR = 17.4, 95% CI = 12.1–25.0), to have a sun-sensitive phenotype (i.e., a skin type that does not tan [RR for not tan vs. tan deeply = 3.9, 95% CI = 3.5–4.2], high nevus counts [RR for many vs. none = 3.6, 95% CI = 3.4–3.9], and red or auburn hair color [RR = 2.7, 95% CI = 2.5–2.9]) and to be sunscreen users (RR = 3.7, 95% CI = 3.4–3.9) than participants who were not discordant by more than one category (Table 1). They were also slightly more likely to be university educated (RR = 1.2, 95% CI = 1.1–1.3) and to have a history of excisions for skin cancer and nonsurgical treatments for actinic lesions (Table 1).Table 1Selected Sociodemographic, Lifestyle, and Phenotypic Characteristics and their Association with Underestimation of Level of Melanoma Risk among 41,936 Men and Women in the QSkin Study CohortCharacteristicUnderestimators1Participants defined as underestimators assessed their risk to be two or more categories lower than their predicted level of risk, on a scale of 1–5. n (%)Reasonable Estimators2The comparison group (reasonable estimators) scored their risk more accurately (i.e., those who were not discordant by more than one category). n (%)Adjusted3Adjusted for age and sex; RRs were calculated using log-binomial regression models. RR (95% CI)Age, y <45853 (9.1)3,494 (12.2)Reference 45 to <501,192 (12.7)4,451 (15.6)1.08 (1.00–1.17) 50 to <551,570 (16.7)5,345 (18.7)1.16 (1.08–1.25) 55 to <601,798 (19.2)5,651 (19.8)1.24 (1.15–1.33) 60 to <651,887 (20.1)5,084 (17.8)1.39 (1.29–1.49) ≥652,075 (22.1)4,568 (16.0)1.61 (1.50–1.72)Sex Male4,163 (44.4)13,023 (45.5)Reference Female5,212 (55.6)15,570 (54.5)1.07 (1.03–1.10)Ethnicity Non-Caucasian29 (0.3)1,967 (6.9)Reference Caucasian9,346 (99.7)26,626 (93.1)17.42 (12.13–25.01)Education School certificate2,311 (24.7)7,083 (24.8)Reference High school certificate/Trade/Diploma4,489 (47.9)14,246 (49.8)1.06 (1.02–1.11) University2,575 (27.5)7,264 (25.4)1.19 (1.13–1.25)Smoking status Never smoker5,577 (59.5)15,565 (54.4)Reference Past smoker3,221 (34.4)10,222 (35.8)0.89 (0.86–0.92) Current smoker577 (6.2)2,806 (9.8)0.67 (0.62–0.72)Alcohol consumption per wk Never or <1 drink3,096 (33.0)10,455 (36.6)Reference 1–6 drinks3,130 (33.4)8,886 (31.1)1.17 (1.12–1.22) >6 drinks3,149 (33.6)9,252 (32.4)1.14 (1.09–1.19)Tanning ability Tan deeply717 (7.7)7,272 (25.4)Reference Tan moderately4,716 (50.3)14,407 (50.4)2.73 (2.54–2.94) Tan a little3,068 (32.7)5,263 (18.4)4.09 (3.79–4.41) Not tan874 (9.3)1,651 (5.8)3.85 (3.52–4.20)Hair color Black495 (5.3)3,207 (11.2)Reference Brown or Blonde7,694 (82.1)24,274 (84.9)1.84 (1.69–2.00) Red or Auburn1,186 (12.7)1,112 (3.9)2.73 (2.54–2.94)Moles at age 21 y None1,796 (19.2)8,815 (30.8)Reference A few4,397 (46.9)15,705 (54.9)1.40 (1.34–1.48) Some2,524 (26.9)3,482 (12.2)2.75 (2.61–2.89) Many658 (7.0)591 (2.1)3.60 (3.38–3.85)Sunscreen use (past year) Never618 (6.6)6,551 (22.9)Reference Ever8,757 (93.4)22,042 (77.1)3.65 (3.38–3.94)Number of skin cancers excised None3,222 (34.4)19,543 (68.3)Reference 11,820 (19.4)3,552 (12.4)2.36 (2.25–2.48) 2 or more4,333 (46.2)5,498 (19.2)3.09 (2.97–3.22)Number of actinic keratoses treated None1,608 (17.2)14,842 (51.9)Reference 1–52,538 (27.1)7,853 (27.5)2.51 (2.37–2.66) 6–203,258 (34.8)3,597 (12.6)4.98 (4.72–5.25) >201,971 (21.0)2,301 (8.0)4.89 (4.62–5.18)Family history of melanoma No6,506 (69.4)21,420 (74.9)Reference Yes2,869 (30.6)7,173 (25.1)1.26 (1.21–1.31)Skin check (past 3 years) Never421 (4.5)8,960 (31.3)Reference Once2,446 (26.1)9,397 (32.9)4.62 (4.18–5.11) ≥2 times6,508 (69.4)10,236 (35.8)8.53 (7.75–9.38)Abbreviations: CI, confidence interval; RR, risk ratio.1 Participants defined as underestimators assessed their risk to be two or more categories lower than their predicted level of risk, on a scale of 1–5.2 The comparison group (reasonable estimators) scored their risk more accurately (i.e., those who were not discordant by more than one category).3 Adjusted for age and sex; RRs were calculated using log-binomial regression models. Open table in a new tab Abbreviations: CI, confidence interval; RR, risk ratio. In summary, we have estimated that almost one quarter of the study population seriously underestimated their future risk of melanoma. Of particular concern, 66% of the underestimators reported a past history of keratinocyte cancer, suggesting that this group would benefit from counseling from their treating clinician at the time of first skin cancer diagnosis about their future risk of developing melanoma. Our finding that underestimators were more likely to report sunscreen is likely due to confounding by indication, whereby the determinants of sunscreen use overlap with the risk factors for melanoma (i.e., sun-sensitive phenotype and sun exposure). Our findings are consistent with previous research reporting modest accuracy of self-assessed risk (Carli et al., 2003Carli P. De Giorgi V. Palli D. Maurichi A. Mulas P. Orlandi C. et al.Dermatologist detection and skin self-examination are associated with thinner melanomas: results from a survey of the Italian Multidisciplinary Group on Melanoma.Arch Dermatol. 2003; 139: 607-612Crossref PubMed Scopus (144) Google Scholar; Harbauer et al., 2003Harbauer A. Binder M. Pehamberger H. Wolff K. Kittler H. Validity of an unsupervised self-administered questionnaire for self-assessment of melanoma risk.Melanoma Res. 2003; 13: 537-542Crossref PubMed Scopus (25) Google Scholar; Richtig et al., 2008Richtig E. Santigli E. Fink-Puches R. Weger W. Hofmann-Wellenhof R. Assessing melanoma risk factors: how closely do patients and doctors agree?.Public Health. 2008; 122: 1433-1439Crossref PubMed Scopus (9) Google Scholar). The strengths of our study were the large sample size and prospective design. The self-assessed risk was measured at baseline using a numerical rating scale, and predicted risk was measured prospectively using a validated risk stratification tool. QSkin study participants were slightly more likely to report having white European ancestry than the general population (93% vs. 86%) (Australian Bureau of StatisticsAustralian Bureau of Statistics. Census. Community profiles, https://www.abs.gov.au/websitedbs/censushome.nsf/home/communityprofiles?opendocument&navpos=230; 2016 (accessed 16 March 2020).Google Scholar; Olsen et al., 2012Olsen C.M. Green A.C. Neale R.E. Webb P.M. Cicero R.A. Jackman L.M. et al.Cohort profile: the QSkin Sun and Health Study.Int J Epidemiol. 2012; 41 (929–929i)Crossref Scopus (73) Google Scholar) and may have been more likely to take part in the study owing to their higher innate risk of skin cancer. Knowledge of skin cancer risk factors among QSkin study participants may also be higher than among the general population, and thus, our findings may not generalize to populations of different ethnic composition and where awareness of risk factors for skin cancer is lower. We repeated our analyses in a large independent population sample (N = 178,434) who used the online QSkin risk tool and again answered the question about self-assessed risk (see Supplementary Materials and Methods). The validation cohort was less likely to underestimate their risk level than the QSkin cohort (11.0% compared with 22.4%) (Supplementary Table S2 and Supplementary Figure S1); however, the factors associated with the underestimating were similar across cohorts (Supplementary Table S3). In the absence of population screening, targeted early detection for melanoma depends on those at a high risk presenting for physician skin checks. We have shown that a subgroup of the population (likely approximately 11% but possibly up to 22% in some segments of the population) underestimates their level of melanoma risk as determined by a validated risk prediction tool. We have previously shown that people who self assess their risk of melanoma as higher than average are more likely to undergo physician skin checks and to use sun protection measures (Olsen et al., 2015Olsen C.M. Thompson B.S. Green A.C. Neale R.E. Whiteman D.C. QSkin Sun and Health Study GroupSun protection and skin examination practices in a setting of high ambient solar radiation: a population-based cohort study.JAMA Dermatol. 2015; 151: 982-990Crossref PubMed Scopus (17) Google Scholar), suggesting that this group is receptive to primary and secondary prevention messages. Thus, in addition to the benefits in terms of secondary prevention, the correct assessment of future risk of melanoma also has the potential to motivate positive behaviors in relation to sun protection, which may prevent melanomas from developing. In summary, we have shown that at least 1 in 10 Australians are ill-equipped to self assess their risk of melanoma, undermining those early detection efforts that depend on those at the highest risk referring themselves for screening. Risk can be assessed accurately by using publicly available risk prediction tools. Data can be made available to external parties upon written formal requests, subject to ethical and legal approvals. Catherine M. Olsen: http://orcid.org/0000-0003-4483-1888 Nirmala Pandeya: http://orcid.org/0000-0003-1462-4968 Jean Claude Dusingize: http://orcid.org/0000-0001-5721-100X Bridie S. Thompson: http://orcid.org/0000-0003-0316-9293 David C. Whiteman: http://orcid.org/0000-0003-2563-9559 The authors state no conflicts of interest. The QSkin study is supported by National Health and Medical Research Council of Australia grants APP1073898 and APP1058522 . DCW is supported by a Research Fellowship from the National Health and Medical Research Council of Australia ( APP1155413 ). Conceptualization: DCW, CMO, NP; Data Curation: CMO, NP, BST, JCD; Formal Analysis: NP; Funding Acquisition: DCW, CMO; Investigation: DCW, CMO, NP, BST, JCD; Methodology: DCW, CMO, NP; Project Administration: DCW, CMO; Supervision: DCW; Writing - Original Draft Preparation: CMO, DCW; Writing - Review and Editing: DCW, CMO, NP, BST, JCD Download .pdf (.54 MB) Help with pdf files Supplementary Data

BACKGROUND Epidemiological studies have consistently documented an increased risk of developing primary non-cutaneous malignancies among people with a history of keratinocyte carcinoma (KC). However, the mechanisms underlying this association remain unclear. We conducted two separate analyses to test whether genetically predicted KC is related to the risk of developing cancers at other sites. METHODS In the first approach (one-sample), we calculated the polygenic risk scores (PRS) for KC using individual-level data in the UK Biobank (n = 394 306) and QSkin cohort (n = 16 896). The association between the KC PRS and each cancer site was assessed using logistic regression. In the secondary (two-sample) approach, we used genome-wide association study (GWAS) summary statistics identified from the most recent GWAS meta-analysis of KC and obtained GWAS data for each cancer site from the UK-Biobank participants only. We used inverse-variance-weighted methods to estimate risks across all genetic variants. RESULTS Using the one-sample approach, we found that the risks of cancer at other sites increased monotonically with KC PRS quartiles, with an odds ratio (OR) of 1.16, 95% confidence interval (CI): 1.13-1.19 for those in KC PRS quartile 4 compared with those in quartile 1. In the two-sample approach, the pooled risk of developing other cancers was statistically significantly elevated, with an OR of 1.05, 95% CI: 1.03-1.07 per doubling in the odds of KC. We observed similar trends of increasing cancer risk with increasing KC PRS in the QSkin cohort. CONCLUSION Two different genetic approaches provide compelling evidence that an instrumental variable for KC constructed from genetic variants predicts the risk of cancers at other sites.

Keratoacanthomas are common keratinocyte skin tumours. However, there is little community-based data published on the clinical features of keratoacanthoma. The aim of this study was to describe the patient and tumour characteristics of keratoacanthomas, as well as their treatment patterns. Data were obtained from the QSkin Sun and Health study, a prospective cohort of 40,438 randomly sampled and consented participants aged 40-69 years in Queensland, Australia. In 2010, a baseline survey collected data, including demography, phenotype, ultraviolet radiation exposure, medical history and lifestyle. Histopathological reports of keratoacanthomas arising until 30 June 2014 were reviewed. In total, 584 participants developed 738 keratoacanthomas; 18% of participants developed multiple tumours. Common patient characteristics were male sex (58%), age ≥60 years (76%), fair skin (80%), and previous history of actinic keratoses/keratinocyte cancers (89%). Keratoacanthomas were commonly located on the legs/feet (48%), and rarely on the the head/neck (7%). Excision was the most frequently used surgical method (71%) Evidence of histopathological regression was reported in 67% of keratoacanthomas, suggesting a potential for spontan-eous resolution in a significant proportion of keratoacanthomas.

Importance Keratoacanthoma (KA) is a common and generally benign keratinocyte skin tumor. Reports of the incidence rates of KA are scant. In addition, the risk factors for KA are not well understood, although associations with UV radiation exposure and older age have been described. Objective To investigate the incidence rate of KA and the risk factors for developing KA. Design, Setting, and Participants The study included data from 40 438 of 193 344 randomly selected residents of Queensland, Australia, who participated in the QSkin Sun and Health (QSkin) prospective population-based cohort study. All participants completed a baseline survey between 2010 and 2011 and were ages 40 to 69 years at baseline. Histopathologic reports of KA were prospectively collected until June 30, 2014, through data linkage with pathologic records. Cox proportional hazards models were used to identify risk factors associated with KA while controlling for potential confounding variables. Data were analyzed from January 2 to April 8, 2020. Exposures Demographic characteristics, phenotypes, UV radiation exposure, medical history, and lifestyle. Results Among 40 438 participants (mean [SD] age, 56 [8] years; 18 240 men [45.1%]), 596 individuals (mean [SD] age, 62 [6] years; 349 men [58.6%]) developed 776 KA tumors during a median follow-up period of 3.0 years (interquartile range, 2.8-3.3 years). The person-based age-standardized incidence rate for KA in the age-restricted cohort was 409 individuals per 100 000 person-years (based on the 2001 Australian population). Risk factors after adjustment for potential confounders were older age (age ≥60 years vs age <50 years; hazard ratio [HR], 6.38; 95% CI, 4.65-8.75), male sex (HR, 1.56; 95% CI, 1.33-1.84), fair skin (vs olive, dark, or black skin; HR, 3.42; 95% CI, 1.66-7.04), inability to tan (vs ability to tan deeply; HR, 1.69; 95% CI, 1.19-2.40), previous excisions of keratinocyte cancers (ever had an excision vs never had an excision; HR, 6.28; 95% CI, 5.03-7.83), current smoking (vs never smoking, HR, 2.02; 95% CI, 1.59-2.57), and high alcohol use (≥14 alcoholic drinks per week vs no alcoholic drinks per week; HR, 1.42; 95% CI, 1.09-1.86). Conclusions and Relevance This is, to date, the first large prospective population-based study to report the incidence rate and risk factors for KA. The high person-based incidence rate (409 individuals per 100 000 person-years) highlights the substantial burden of KA in Queensland, Australia. Furthermore, the study's findings suggest that older age (≥60 years), male sex, UV radiation-sensitive phenotypes, indications of high sun exposure (eg, previous keratinocyte cancer excisions), smoking, and high alcohol use are independent risk factors for the development of KA.

PURPOSE Keratinocyte cancers are exceedingly common in high-risk populations, but accurate measures of incidence are seldom derived because the burden of manually reviewing pathology reports to extract relevant diagnostic information is excessive. Thus, we sought to develop supervised learning algorithms for classifying basal and squamous cell carcinomas and other diagnoses, as well as disease site, and incorporate these into a Web application capable of processing large numbers of pathology reports. METHODS Participants in the QSkin study were recruited in 2011 and comprised men and women age 40-69 years at baseline (N = 43,794) who were randomly selected from a population register in Queensland, Australia. Histologic data were manually extracted from free-text pathology reports for participants with histologically confirmed keratinocyte cancers for whom a pathology report was available (n = 25,786 reports). This provided a training data set for the development of algorithms capable of deriving diagnosis and site from free-text pathology reports. We calculated agreement statistics between algorithm-derived classifications and 3 independent validation data sets of manually abstracted pathology reports. RESULTS The agreement for classifications of basal cell carcinoma (κ = 0.97 and κ = 0.96) and squamous cell carcinoma (κ = 0.93 for both) was almost perfect in 2 validation data sets but was slightly lower for a third (κ = 0.82 and κ = 0.90, respectively). Agreement for total counts of specific diagnoses was also high (κ > 0.8). Similar levels of agreement between algorithm-derived and manually extracted data were observed for classifications of keratoacanthoma and intraepidermal carcinoma. CONCLUSION Supervised learning methods were used to develop a Web application capable of accurately and rapidly classifying large numbers of pathology reports for keratinocyte cancers and related diagnoses. Such tools may provide the means to accurately measure subtype-specific skin cancer incidence.

ABSTRACT Aim To compare occupational flying hours (a surrogate for occupational exposure to radiation) of commercial pilots subsequently diagnosed with melanoma, with those without melanoma. Methods Nested case-control study of de-identified male commercial pilots in Australia 2011-2016, ascertained through the Civil Aviation Safety Authority (CASA). Cases were pilots diagnosed with melanoma 2011-2016; controls were randomly-selected pilots age-matched 1:2 with invasive cases. Total flying hours and hours flown in the last 6 months in 2011, date of birth and state of residence were also obtained. We estimated the association between total flying hours (in tertile groups), and melanoma by odds ratios adjusted for age and state (ORsadj; 95% confidence intervals (CIs)). Results During 2011-2016, 51 pilots developed invasive melanoma and 63, in situ (mean ages 47 and 49 years, respectively). Their median cumulative flying hours in 2011 were 6,108 and 6,900 respectively, compared with 7,500 for 102 control pilots (mean age 48.6). Risk of invasive melanoma did not increase per 1000 total hours flown (ORadj=1.00) nor did risk increase in pilots with highest vs lowest total flying hours (ORadj=1.18, 95% CI 0.44-3.15). Total flying hours were inversely associated with invasive melanoma development in pilots aged < 50 (ORadj=0.37, not significant), and not associated with melanoma on exposed sites. Recent flying hours were not associated with melanoma. Results were unchanged with inclusion of in situ cases. Conclusion Risk of melanoma in Australian commercial pilots is unrelated to cumulative or recent occupational exposure to radiation as indicated by total and recent flying hours.