Abstract 817: Mendelian randomization and mediation analysis of 5p15.33, telomere length and lung cancer risk
Bibliographic record
Abstract
Abstract Background: Telomere length (TL) is a predictor of lung cancer risk, but the direction of this association differs between and prospective and case-control studies. This discrepancy may be attributed to reverse causation in the latter, due to disease-related changes in TL that is measured after diagnosis or treatment. To overcome these limitations and characterize the relationship between TL and lung cancer risk we carried out observational and mediation analyses, and a 2-stage Mendelian Randomization (MR) analysis, where we developed novel genetic instruments for TL and tested the association with lung cancer using 20 OncoArray studies in the Transdisciplinary Research in Cancer of the Lung group of the International Lung Cancer Consortium. Methods: The observational analysis examined TL measured using qPCR in 1128 cases and 928 controls. Odds ratios (OR) for TL were adjusted for age, sex and cigarette pack-years. Mediation analysis was used to estimate the% of the lung cancer association in 5p15 that operates through TL. To develop novel TL instruments, variants identified through deep sequencing of the 5p15 locus were genotyped in 900 controls. Variants that met MR criteria were combined into a single instrumental variable (IV), and its association with TL was estimated. We also used 6 previously identified TL predictors (p<5×10-8) as genetic instruments: rs10165485 (ACYP2), rs10936599 (TERC), rs11100479 (NAF1), rs9420907 (OBFC1), rs6028466 (DHX35), rs755017 (RTEL1). For these SNPs association estimates for TL were obtained from the literature. To estimate the association with lung cancer risk, we tested all 7 IVs using data from 14324 lung cases and 10783 controls in OncoArray. Lastly, to obtain a summary estimate for the causal effect of TL on lung cancer risk, ORs from OncoArray were combined with â-TL estimates using a likelihood-based MR model. Results: The observational analysis suggested that longer TL is associated with decreased lung cancer risk (OR = 0.94, p = 0.04). This was more pronounced for squamous carcinoma (OR = 0.77, p = 1.1×10-4). We also showed that TL mediates up to 8% (p<0.05) of the lung cancer signal in 5p15. In the first stage of the MR analysis, we identified 8 5p15 SNPs that were associated with TL (p<5×10-3), including 6 novel rare variants, not previously associated with TL. Together these variants were reliably associated with TL (â = 0.15, p = 1.8×10-7) and explained 2.3% of variance in TL. Using this new instrument and 6 other SNPs as IVs, our MR analysis showed that longer TL is a risk factor for lung cancer (OR = 1.77, 95% CI: 1.32-1.54), especially adenocarcinoma (OR = 2.02, 95% CI: 1.29-3.71). Conclusions: We developed novel genetic instruments for TL, and confirmed that genetically predicted longer TL is associated with increased lung cancer risk. These findings suggest that previously reported associations of long TL with decreased risk were likely due to residual confounding by smoking, age and/or reverse causation. Citation Format: Linda Kachuri, George Davey Smith, Geoffrey Liu, Maria Teresa Landi, David C. Christiani, Neil E. Caporaso, James D. McKay, Xifeng Wu, Melinda C. Aldrich, Gad Rennert, Dawn Teare, Chu Chen, Gary E. Goodman, Jennifer A. Doherty, John K. Field, Lambertus A. Kiemeney, Adonina Tardón, Aage Haugen, Stephen Lam, Loic Le Marchand, Matthew B. Schabath, Angeline S. Andrew, Mattias Johansson, Jonas Manjer, Philip Lazarus, Susanne Arnold, Gordon Fehringer, Xuchen Zong, Paul Brennan, Stig E. Bojesen, Christopher I. Amos, Rayjean J. Hung. Mendelian randomization and mediation analysis of 5p15.33, telomere length and lung cancer risk. [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 817.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.031 | 0.054 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.003 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.010 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".