Search

Background: Lung cancer is the leading cause of cancer death worldwide. We explored the relationship between socioeconomic status and lung cancer incidence across world regions, using educational level as a proxy for socioeconomic status. Methods: We analyzed the harmonized database of the Lung Cancer Cohort Consortium (LC3). The current study included data from 18 prospective cohorts from 15 countries in the US, Europe, Asia, and Australia. Separately for participants who never or currently/formerly smoked, we estimated the association between educational level and incident lung cancer using Cox proportional hazards models. Information on education was harmonized using the International Standard Classification of Education and then modeled as an ordinal variable in 4 categories. Models were adjusted for age, sex, and for participants who currently/formerly smoked, smoking duration, cigarettes per day, and time since cessation. Results: Among 2.6 million participants from 15 countries, 62,645 developed lung cancer during follow-up (median follow-up = 12.6 years). Among current/former smokers, increased educational level was associated with decreased lung cancer incidence in most cohorts after adjustment for age, sex, and detailed smoking information, with HRs ranging from 0.77 (95%CI: 0.42-1.41) per 1-unit increase in educational level in the Iranian Golestan Cohort Study to 1.02 (95%CI: 0.95-1.09) in the Australian Melbourne Collaborative Cohort Study. When grouping by world region, the association between education and lung cancer incidence among currently/formerly smoking participants was similar for the US (HRpooled=0.88, 95%CI: 0.87-0.89), Europe (HRpooled=0.89, 95%CI: 0.88-0.91), and Asia (HRpooled=0.91, 95%CI: 0.86-0.96), but attenuated in the Australian cohort (HR=1.02, 95%CI: 0.95-1.09). Among never smokers, after adjustment for age and sex, there was no statistically significant association between educational level and lung cancer incidence (p-trend>0.05 in all cohorts), with the exception of the US Southern Community Cohort Study, which comprises primarily African-Americans and showed a HR of 0.75 (95%CI: 0.62-0.90). Conclusion: Among currently and formerly smoking individuals, higher educational level showed a strikingly consistent decreased risk of incident lung cancer across cohorts from 4 continents, after detailed adjustment for smoking. In contrast, among people who never smoked, there was no association between education and lung cancer incidence in any cohort, with the exception of the Southern Community Cohort Study. Further research is needed to clarify the mechanisms, either related or unrelated to smoking, that contribute to the association between education and lung cancer risk. Citation Format: Justina U. Onwuka, Hana Zahed, Xiaoshuang Feng, Karine Alcala, Mattias Johansson, Hilary A. Robbins, Lung Cancer Cohort Consortium. Socioeconomic status and lung cancer incidence: An analysis of data from 15 countries in the Lung Cancer Cohort Consortium [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 1950.

BACKGROUND: In the course of screening, individual risk of disease evolves based on screening results. We calculated how individual lung cancer risk changes based on screening CT findings using data from the National Lung Screening Trial (NLST), which conducted annual screening for 3 years. METHODS: We calculated lung cancer risks by combining CT findings with individual predicted 1-year “pre-screening risk.” Pre-screening risk r(x) was predicted in the absence of screening using a validated risk model (Katki et al., JAMA 2016) with covariates (x): age, education, sex, race, smoking intensity/duration/quit-years, body mass index, family history of lung cancer, and self-reported emphysema. We used log-binomial regression to calculate the risk of an “interval” lung cancer (within 1 year of a negative screen) or a “screen-detected” cancer detected at the next annual screen. For each, covariates included log-transformed 1-year pre-screening risk and CT findings including classification as negative or false-positive and other specific features. RESULTS: The median 1-year pre-screening risk at the first NLST screen was 0.32% (interquartile range 0.19-0.53%). Among CT-negatives, risk over the next year was substantially reduced as r(x)1.32 (median interval cancer risk 0.05%), but risk at the next screen reverted to pre-screening risk as r(x)1. Risk at the next screen was higher for those whose CT noted either emphysema (r(x)0.95, median risk 0.53%) or consolidation (r(x)0.76, median risk 1.6%). Among CT-false-positives, overall risk at the next screen increased as r(x)0.74 (median risk 1.5%). Risk was higher among those with nodule(s) that were larger, had spiculated margins (median risk 4.1%), were located in the upper lobes (median risk 1.4%), or grew during the most recent screening interval (median risk 7.9%), while nodules with smooth margins indicated lower risk (median risk 0.71%). Those with a smooth-margins nodule and no risk-increasing factors essentially reverted to their pre-screening risk at the next screen as r(x)1.01 (median risk 0.29%), as if they had screened negative. Overall, only the immediately prior screen result, and not earlier screens, predicted lung cancer risk (all p>0.2). Exponents were similar for each interval and at each screen (all p>0.07). CONCLUSIONS: CT-negatives experienced reduced lung cancer risk over the next year, but reverted to their pre-screening risk at the next screen. CT-false-positives experienced substantially increased lung cancer detection at the next annual screen, with most risks exceeding 1%. These risk increases were explained by specific CT features including nodule size, location, margins, and growth. Citation Format: Hilary A. Robbins, Christine D. Berg, Li C. Cheung, Anil K. Chaturvedi, Hormuzd A. Katki. Effect of screening CT results and features on lung cancer risk prediction within the National Lung Screening Trial [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 5291. doi:10.1158/1538-7445.AM2017-5291

Background: Women living with HIV (WLHIV) have increased cervical cancer risk. Guidelines suggest screening WLHIV with negative cervical cytology annually, and managing WLHIV with ASC-US cytology by general population (GP) guidelines (1 year return if HPV testing unavailable). Methods: We used risk benchmarking to compare cervical precancer risk in WLHIV to the GP. For WLHIV ages 21-65 years in the Women's Interagency HIV Study (WIHS), we evaluated the first cytology result from 2000 or later. We used parametric survival models to calculate HSIL risks after negative or ASC-US cytology, overall and by CD4 cell count. Separately, we synthesized HSIL risk estimates among GP women in 13 published studies using mixed-effects models. We then benchmarked risks in WLHIV to the 3-year GP risks, which for negative cytology represent the threshold for re-screening. Results: Among 2,653 WLHIV, 1,982 (75%) had negative cytology and 377 (14%) had ASC-US cytology. Most WLHIV with negative (72%) and ASC-US (52%) cytology had CD4≥350 cells/μL. We observed 95 cases of HSIL (CIN2+; 44 were CIN3+ including 1 cancer) within 5 years of cytology. After negative cytology, 3-year GP risk “benchmarks” were 0.83% (CIN2+) and 0.60% (CIN3+). WLHIV with CD4≥350 met both benchmarks by 2 years (1.1% and 0.68% respectively), while WLHIV with CD4 Conclusions: For WLHIV with CD4≥350, these data suggest that the interval for re-screening after negative cervical cytology can be lengthened from 1 to 2 years, and that inclusion in GP guidelines for managing ASC-US cytology is appropriate. In contrast, WLHIV with CD4 Citation Format: Hilary A. Robbins, L. Stewart Massad, Christopher B. Pierce, Lisa Flowers, Teresa M. Darragh, Howard Minkoff, Lisa Rahangdale, Marla J. Keller, Joel Milam, Margaret Fischl, Sadeep Shrestha, Christine Colie, Howard Strickler, Gypsyamber D’Souza. Optimizing cervical cancer screening for HIV-infected women: A risk-based approach. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 2580.