Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Open Access

Peer-reviewed

Research Article

Causal role of metabolites in non-small cell lung cancer: Mendelian randomization (MR) study

  • Qian Li ,

    Roles Conceptualization, Data curation, Writing – original draft

    liqian20210703@icloud.com

    Affiliation Department of Cardiac Surgery, The First Affiliated Hospital of Kunming Medical University, Kunming, 650000, China

    ⨯
  • Zedong Wei,

    Roles Data curation, Project administration, Visualization

    Affiliation Department of Thoracic Surgery, Qianjiang Central Hospital, Qianjiang, 433100, China

    ⨯
  • Yonglun Zhang,

    Roles Validation, Writing – review & editing

    Affiliation Department of Thoracic Surgery, Qianjiang Central Hospital, Qianjiang, 433100, China

    ⨯
  • Chongqing Zheng

    Roles Supervision, Writing – review & editing

    Affiliation Department of Thoracic Surgery, Qianjiang Central Hospital, Qianjiang, 433100, China

    ⨯

Abstract

On a global scale, lung cancer(LC) is the most commonly occurring form of cancer. Nonetheless, the process of screening and detecting it in its early stages presents significant challenges. Earlier research endeavors have recognized metabolites as potentially reliable biomarkers for LC. However, the majority of these studies have been limited in scope, featuring inconsistencies in terms of the relationships and levels of association observed.Moreover, there has been a lack of consistency in the types of biological samples utilized in previous studies. Therefore, the main objective of our research was to explore the correlation between metabolites and Non-small cell lung cancer (NSCLC).Thorough two-sample Mendelian randomization (TSMR) analysis, we investigated potential cause-and-effect relationships between 1400 metabolites and the risk of NSCLC.The analysis of TSMR revealed a significant causal impact of 61 metabolites on NSCLC.To ensure the reliability and validity of our findings, we perform FDR correction for P-values by Benjaminiand Hochberg(BH) method, Our results indicate that Oleate/vaccenate (18:1) levels and Caffeine to paraxanthine ratio may be causally associated with an increased risk of NSCLC [Oleate/vaccenate(18:1)levels: OR = 1.171,95%CI: 1.085–1.265, FDR = 0.036; Caffeine to paraxanthine ratio: OR = 1.386, 95%CI:1.191–1.612,FDR = 0.032].

Citation: Li Q, Wei Z, Zhang Y, Zheng C (2024) Causal role of metabolites in non-small cell lung cancer: Mendelian randomization (MR) study. PLoS ONE 19(3): e0300904. https://doi.org/10.1371/journal.pone.0300904

Editor: Abeer El Wakil, Alexandria University, EGYPT

Received: January 16, 2024; Accepted: March 6, 2024; Published: March 22, 2024

Copyright: © 2024 Li et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: The GWAS data for non-small cell lung cancer comes from the database (http://gwas.mrcieu.ac.uk), and can be downloaded via the GWAS ID "finn-b-C3_LUNG_NONSMALL". The GWAS data for 1,400 metabolites comes from the database (https://www.ebi.ac.uk/gwas/), and can be downloaded using the European GWASs registration number: GCST90199621-90201020. The results of the Mendelian randomization analysis related to the article can be found in the supplementary files.

Funding: The author(s) received no specific funding for this work.

Competing interests: The authors declare that they have no competing interests.

Introduction

LC is a prevalent form of cancer, accounting for 12.9% of global cancer-related fatalities and causing approximately 1.3 million deaths annually [1–3]. Approximately 80.0 to 85.0% of LC cases are categorized as NSCLC [3–5]. The late diagnosis of LC poses a significant challenge, resulting in a short median survival of about 18 months and an overall 5-year survival rate ranging from 15% to 21% depending on gender [6,7]. Sadly, as many as 75% of individuals diagnosed with NSCLC have already reached an advanced stage or developed metastases at the time of diagnosis, leading to a dismal 5-year survival rate below 5% [1]. Despite extensive research conducted to investigate LC, including studying the molecular mechanism, cytological behavior, and biological process, our available treatments for LC remain inadequate. Previous researches have shown that metabolites are crucial factor affecting LC. In LC, researchers have conducted metabolomic studies in blood and urine samples to discover metabolite biomarkers for diagnosing lung cancer [8–14], characterizing tumors [15,16], and monitoring disease progression [8,11]. However, there is a lack of a conclusive and all-encompassing summary regarding the causal relationship between metabolites and LC, and this research aims to address this gap.

Metabolites refer to the substances produced and utilized in the biological metabolism of both host organisms and microorganisms and xenobiotics [17]. Metabolomics is a widely employed technique in which the analysis of metabolites present in biofluids, cells, and tissues is carried out. This approach is commonly utilized for the identification of biomarkers [18]. By employing metabolites obtained from human studies, it becomes possible to establish links between diseases and metabolic pathways [19]. Due to the high sensitivity of metabolomics, even small changes in biological behavior can be identified and used to gain understanding of the mechanisms behind different physiological conditions, abnormal processes, and diseases [17]. Despite the long history of metabolite identification and validation, a major obstacle in biomarker validation remains overcoming the variations in metabolite levels among individuals, caused by differences in genetic factors and environmental exposures [20,21]. Identifying the metabolites and understanding their biological roles is a crucial step in cancer research. Numerous studies have demonstrated the immense potential of metabolomics in this field [22,23].

MR is a technique that utilizes genetic variants as instrumental variables (IVs) to assess the causal effects between exposure and outcome. It is a widely applicable method built upon Genome-Wide Association Studies (GWAS) summary data [24]. To put it differently, the genetic variants found in GWAS can serve as a substitute for a randomized controlled trial to establish a cause-and-effect relationship between different factors [22,23]. Additionally, GWAS has been expanded to include metabolic profiles. In this research, an extensive TSMR analysis was conducted to ascertain the causal link between metabolites and NSCLC.

Materials and methods

Study design

Using a TSMR study, we conducted a comprehensive analysis to establish the causal impacts of 1400 metabolites on the susceptibility to NSCLC. In the MR study, three basic assumptions need to be adhered to [25]. Firstly, the genetic variants must display a strong correlation with the exposure factor. Secondly, they must be independent of any potential factors that may confound the results. Lastly, it is essential that these genetic instruments impact the outcome solely through the exposure.

GWAS data sources for NSCLC and metabolites

We performed a search using the GWAS summary data(http://gwas.mrcieu.ac.uk), which includes a comprehensive collection of summary statistics from multiple GWAS. For our analysis, we utilized the openly available summary statistic datasets obtained from a GWAS conducted on individuals with European ancestry diagnosed with NSCLC (total n = 217165; ncase = 1627, ncontrol = 217,165).Various metabolites GWAS summary statistics were deposited to GWAS Catalog (https://www.ebi.ac.uk/gwas/). Accession numbers for European GWASs: GCST90199621-90201020. These statistics cover a total of 1400 metabolites [26].

Selection IVs

Based on our study, we established a significance level of 1×10−5 for the independent variables each metabolite. In order to obtain independent IVs, we employed clumping technique by using a linkage disequilibrium (LD) reference panel sourced from the 1000 Genomes Project. Clumping was performed based on a criterion of R2 < 0.001 within a 1,000-kb distance [27]. For NSCLC, when we adjusted the significance level to 5×10−8,there is only one significant instrumental variable. therefore, we adjusted the significance level to 1×10−5 for NSCLC.Furthermore, in order to mitigate any potential bias caused by weak instruments, we only included IVs with F statistics greater than 10, which we identified as strong instruments, for the subsequent analysis.We harmonized tthe exposure and outcome single nucleotide polymorphism(SNPs) so that the effect estimates were consistent for the same effect allele. We removed palindromic SNPs with intermediate effect allele frequencies (EAFs > 0.42) or SNPs with incompatible alleles [28].

Statistical analysis

The data analysis in the study was carried out using R software (version 4.2.1). The TwoSampleMR package (version 0.5.6) and MRPRESSO package (version 1.0) were utilized for the analysis [29]. As our primary analysis, we used the random-effects inverse variance weighted (IVW), weighted median (WM), MREgger(ME),Simple mode(SM) and Weighted mode(WM)methods,providing an estimate of the effect of the exposure on the outcome when MR assumptions are valid.We applied Cochran’s Q test (p<0.05) to estimate residual heterogeneity for the IVW model and the MR-Egger intercept test (p < 0.05) to indicate potential pleiotropy on causal estimates. In order to eliminate the influence of horizontal pleiotropy, the researchers employed a widely-used approach known as MR-Egger. If the intercept term of MR-Egger is found to be statistically significant, it suggests the existence of horizontal multiplicity [30]. Furthermore, the MR pleiotropy residual sum and outlier (MR-PRESSO) method, which is a robust approach, was employed to eliminate potential horizontal pleiotropic outliers that could have had a significant impact on the estimation outcomes within the MR-PRESSO package [31]. In addition, we correction for P-values (FDR<0.05) with BH method make the results more rigorous and reliable.

Results

Causal estimates between metabolites and NSCLC

To explore the causal effects of 1400 metabolites on NSCLC, TSMR analysis was performed, and the IVW method was used as the main analysis (S1 Table). For sensitivity analysis, we utilized the Cochranwith a significance level of p < 0.05) to assess heterogeneity in the IVW model. Additionally, we performed the MR-Egger intercept test (with a significance level of p < 0.05) to evaluate the potential impact of pleiotropy on the causal estimates (S2 and S3 Tables). Finally, a leave-one-out (LOO) analysis was conducted to determine if a single nucleotide polymorphism (SNP) could influence the accuracy of the causal estimate. 61 metabolites were found to be potentially significant, with a significance level of P<0.05.After correction for P-values (FDR<0.05) with BH method, we detected the significant effect of two metabolites(Oleate/vaccenate (18:1) levels and Caffeine to paraxanthine ratio) on NSCLC (S4 Table). Oleate/vaccenate (18:1) levels and Caffeine to paraxanthine ratio is a risk factor for NSCLC (Fig 1). Our study displays the relationship between the two metabolites and the risk of NSCLC, indicating no heterogeneity among the IVs with a P-value of 0.679 and 0.617 (S2 and S4 Tables). The Forest plots in Fig 2 depict the impact of SNP effect sizes on Oleate/vaccenate (18:1) levels and Caffeine to paraxanthine ratio. The leave-one-out analysis does not uncover any statistically significant outliers or influential data points (Fig 3). Our results support genetic relationship between the two metabolites and NSCLC (Fig 4). Notably, Significantly, our findings have revealed a robust cause-and-effect relationship between a heightened susceptibility to NSCLC and elevated levels of the two metabolites, as evidenced by rigorous IVW analysis[Oleate/vaccenate (18:1) levels:odds ratio (OR) = 1.172, 95% confifidence interval(CI) = 1.085–1.265, P<0.001; Caffeine to paraxanthine ratio:OR = 1.386, CI = 1.192–1.612, P<0.001], Similar positive trends with weighted median and weight mode methods were observed (Fig 4, S1 Table).

thumbnail
Download:
Fig 1. Scatterplot of genetic association Oleate/vaccenate (18:1) levels and Caffeine to paraxanthine ratio With NSCLC.

(a) Oleate/vaccenate (18:1) levels with NSCLC and (b) Caffeine to paraxanthine ratio With NSCLC.

https://doi.org/10.1371/journal.pone.0300904.g001

thumbnail
Download:
Fig 2.

Forest plots of causal efects of Oleate/vaccenate (18:1) levels-associated SNPs on NSCLC (a) and Caffeine to paraxanthine ratio-associated SNPs on NSCLC (b).

https://doi.org/10.1371/journal.pone.0300904.g002

thumbnail
Download:
Fig 3.

Leave-one-out analysis of SNPs associated with Oleate/vaccenate (18:1) levels and NSCLC(a) and Caffeine to paraxanthine ratio and NSCLC(b).

https://doi.org/10.1371/journal.pone.0300904.g003

thumbnail
Download:
Fig 4. MR estimates for the relationship Oleate/vaccenate (18:1) levels and Caffeine to paraxanthine ratio with NSCLC.

https://doi.org/10.1371/journal.pone.0300904.g004

Then we perform reverse MR analysis, we investigated the relationship between NSCLC as the exposure and the two metabolites(Oleate/vaccenate (18:1) levels and Caffeine to paraxanthine ratio) as the outcome. We identified 16 SNPs that were found to be associated with these variables. The results of our analysis revealed no significant correlations between NSCLC and the two metabolites.as shown by the IVW (P = 0.618 and P = 0.904),as depicted in Fig 5.

thumbnail
Download:
Fig 5. Reverse MR estimates for the relationship NSCLC with Oleate/vaccenate (18:1) levels and Caffeine to paraxanthine ratio.

https://doi.org/10.1371/journal.pone.0300904.g005

Discussion

Using extensive genetic data that is publicly accessible, we conducted an analysis to investigate the causal connections between 1400 metabolites and NSCLC. As far as we know, this is the initial MR study to examine the causal relationship between numerous metabolites and NSCLC. In this study, 61 metabolites was found to have causal effects on NSCLC(P<0.05), and two metabolites(Oleate/vaccenate (18:1) levels and Caffeine to paraxanthine ratio) had significant causal effects on NSCLC(FDR<0.05). However, the reverse MR analysis representation NSCLC had no significant causal effects on the two metabolites.

Our study found that the risk of NSCLC increase with an increase in the levels of two metabolites(Oleate/vaccenate (18:1) levels and Caffeine to paraxanthine ratio). These two metabolites have not been linked to lung cancer in previous studies. Oleate/vaccenate (18:1) levels are monounsaturated fatty acids(MUFA). It is an important component of lipid synthesis. Lipids play a crucial role in cellular growth and division, as well as in the formation of new cell membranes, energy production, and signaling processes. Therefore, it is not surprising that cancer cells necessitate greater amounts of lipids to sustain their unregulated proliferation [32,33]. To meet this condition, cancer cells boost their lipid production through de novo lipogenesis or increase their uptake of lipids, in order to make use of lipids obtained from their surrounding microenvironment. Complex lipids are synthesized using fatty acids as their building blocks. The properties of these fatty acids, such as their length and degree of unsaturation, play a significant role in determining the properties of the resulting lipids. In general, fatty acids can be categorized into three primary groups: Saturated fatty acids(SFA), MUFA, and polyunsaturated fatty acids(PUFA) [34]. Cancer has been associated with oleic acid (OA), the most prevalent MUFA found in the bloodstream, according to several research studies. OA has been proven to be consumed and utilized by metastatic breast cancer and prostate cancer cells in culture, indicating its association with the development and progression of cancer [35–37]. These studies emphasize the significance of MUFAs and OA in relation to cancer, as cancer cells specifically amplify MUFA levels and utilization. Stearoyl CoA desaturase (SCD1) is a vital enzyme in the production of fresh lipids. It is found to be excessively expressed in melanoma, colorectal cancer, and clear-cell renal carcinoma [38–40]. Moreover, elevated levels of SCD1 have been linked to the progression of lung cancer and unfavorable prognoses in patients [41]. The previously unexplored n-7 product of SCD1, known as cis-vaccenic acid, has been revealed to play a significant role in regulating the growth of prostate cancer cells and oxidative stress [42]. This is in spite of numerous studies that have established a connection between Oleate/vaccenate (18:1) levels and cancer, but they are rarely associated with lung cancer. Our study provides new evidence for future research.

Caffeine is completely absorbed, rapidly distributed and extensively metabolized in the liver to three major metabolites: paraxanthine (1,7 dimethylxanthine), theophylline (1,3 dimethylxanthine), and theobromine (3,7 dimethylxanthine) which undergo further metabolism to several xanthines and uric acids [43]. The metabolism of caffeine is primarily carried out by CYP1A2, with more than 95% of the substance being metabolized through this enzyme. The variability in the activity of CYP1A2 among individuals is believed to strongly correlate with the differences observed in caffeine metabolism [44–46]. The paraxanthine to caffeine ratio provides valuable information as paraxanthine is formed solely through the catalytic action of CYP1A2 [47]. This step accounts for approximately 80% of caffeine metabolism. There is a scarcity of reports exploring the association between caffeine and LC. It is possible that caffeine could hinder the repair of DNA, thereby increasing the vulnerability to lung adenocarcinoma in women with rapid CYP1A2 activity and slow N-acetyltransferase2(NAT2) [48,49]. No link between Caffeine to paraxanthine ratio and lung cancer risk has been reported.Our research may provide clues for future exploration.

This study conducted two-sample MR analysis based on the published results of large GWAS cohorts,with a large sample size of 1400 metabolites,so it has high statistical efficiency.The conclusions of this study are based on genetic instrumental variables, and causal inference is made using a variety of MR analysis methods. The results are robust and were not confounded by horizontal pleiotropy and other factors. Our study has limitations. First,the study was based on a European database, so the conclusion cannot be extended to other ethnic groups. Second, we used strict threshold to evaluate the results, which may decrease some true positives.Finally, our study was limited to non-small cell lung cancer, and the relationship between metabolites and risk of small cell lung cancer needs to be explored in future studies.

Conclusions

In summary, our comprehensive bidirectional MR analysis has established the causal associations between certain metabolites and NSCLC. This sheds light on the intricate interactions between metabolites and NSCLC.

Additionally, our research offers researchers a fresh opportunity to delve into the biological mechanisms of NSCLC, potentially paving the way for earlier intervention and treatment options. By building upon the existing knowledge in metabolomics of NSCLC, our findings present crucial hints for preventing the onset of NSCLC.

Supporting information

S1 Table. Results of the causal effect of metabolites on NSCLC.

https://doi.org/10.1371/journal.pone.0300904.s001

(XLSX)

S2 Table. Sensitivity analysis results of causal effects of metabolites on NSCLC.

https://doi.org/10.1371/journal.pone.0300904.s002

(XLSX)

S3 Table. Pleiotropy analysis results of causal effects of metabolites on NSCLC.

https://doi.org/10.1371/journal.pone.0300904.s003

(XLSX)

S4 Table. 61 metabolites associated with NSCLC.

https://doi.org/10.1371/journal.pone.0300904.s004

(XLSX)

References

  1. 1. Arrieta O.; Cruz-Rico G.; Soto-Perez-de-Celis E.; Ramírez-Tirado L.A.; Caballe-Perez E.; Martínez-Hernández J.N.; et al. Reduction in Hepatocyte Growth Factor Serum Levels is Associated with Improved Prognosis in Advanced Lung Adenocarcinoma Patients Treated with Afatinib: A Phase II Trial. Target Oncol. 2016, 11, 619–629. pmid:27033062
  2. 2. Neal J.W.; Dahlberg S.E.; Wakelee H.A.; Aisner S.C.; Bowden M.; Huang Y.; et al. Erlotinib, cabozantinib, or erlotinib plus cabozantinib as second-line or third-line treatment of patients with EGFR wild-type advanced non-small-cell lung cancer (ECOG-ACRIN 1512): A randomised, controlled, open-label, multicentre, phase 2 trial. Lancet Oncol. 2016, 17, 1661–1671. pmid:27825638
  3. 3. Okamoto I.; Sakai K.; Morita S.; Yoshioka H.; Kaneda H.; Takeda K.; et al. Multiplex genomic profiling of non-small cell lung cancers from the LETS phase III trial of first-line S-1/carboplatin versus paclitaxel/carboplatin: Results of a West Japan Oncology Group study. Oncotarget 2014, 5, 2293–2304. pmid:24810493
  4. 4. Yoshioka H.; Azuma K.; Yamamoto N.; Takahashi T.; Nishio M.; Katakami N.; et al. A randomized, double-blind, placebo-controlled, phase III trial of erlotinib with or without a c-Met inhibitor tivantinib (ARQ 197) in Asian patients with previously treated stage IIIB/IV nonsquamous nonsmall-cell lung cancer harboring wild-type epidermal growth factor receptor (ATTENTION study). Ann. Oncol. 2015, 26, 2066–2072. pmid:26153496
  5. 5. Spigel D.R.; Ervin T.J.; Ramlau R.A.; Daniel D.B.; Goldschmidt J.H. Jr.; Blumenschein G.R. Jr.; et al. Randomized phase II trial of Onartuzumab in combination with erlotinib in patients with advanced non-small-cell lung cancer. J. Clin. Oncol. 2013, 31, 4105–4114. pmid:24101053
  6. 6. Urvay S.E.; Yucel B.; Erdis E.; Turan N. Prognostic Factors in Stage III Non-Small-Cell Lung Cancer Patients. Asian. Pac. J. Cancer Prev. 2016, 17, 4693–4697. pmid:27893199
  7. 7. Önal Ö.; Koçer M.; Eroğlu H.N.; Yilmaz S.D.; Eroğlu I.; Karadoğan D. Survival analysis and factors affecting survival in patients who presented to the medical oncology unit with non-small cell lung cancer. Turk. J. Med. Sci. 2020, 50, 1838–1850. pmid:32512671
  8. 8. Puchades-Carrasco L., Jantus-Lewintre E., Pérez-Rambla C., García-García F., Lucas R., Calabuig S., Serum metabolomic profifiling facilitates the non-invasive identifification of metabolic biomarkers associated with the onset and progression of non-small cell lung cancer, Oncotarget 7 (2016) 12904–12916.
  9. 9. Wen C.P., Zhang F., Liang D., Wen C., Gu J., Ye Y., et al., The ability of bilirubin in identifying smokers with higher risk of lung cancer: a large cohort study in conjunction with global metabolomic profifiling, Clin. Cancer Res. 21 (2015) 193–200.
  10. 10. Maeda J., Higashiyama M., Imaizumi A., Nakayama T., Yamamoto H., Daimon T., Possibility of multivariate function composed of plasma amino acid profifiles as a novel screening index for non-small cell lung cancer: a case control study, BMC Cancer 10 (2010) 690.
  11. 11. Hori S., Nishiumi S., Kobayashi K., Shinohara M., Hatakeyama Y., Kotani Y., A metabolomic approach to lung cancer, Lung Cancer 74 (2011) 284–292. pmid:21411176
  12. 12. Deja S., Porebska I., Kowal A., Zabek A., Barg W., Pawelczyk K., Metabolomics provide new insights on lung cancer staging and discrimination from chronic obstructive pulmonary disease, J. Pharm. Biomed. Anal. 100 (2014) 369–380. pmid:25213261
  13. 13. Chen Y., Ma Z., Min L., Li H., Wang B., Zhong J., Biomarker identifification and pathway analysis by serum metabolomics of lung cancer, BioMed Res. Int. 2015 (2015) 183624.
  14. 14. Chen Y., Ma Z., Li A., Li H., Wang B., Zhong J., Metabolomic profifiling of human serum in lung cancer patients using liquid chromatography/hybrid quadrupole time-of-flflight mass spectrometry and gas chromatography/mass spectrometry, J. Cancer Res. Clin. Oncol. 141 (2015) 705–718.
  15. 15. Jordan K.W., Adkins C.B., Su L., Halpern E.F., Mark E.J., Christiani D.C., Comparison of squamous cell carcinoma and adenocarcinoma of the lung by metabolomic analysis of tissue-serum pairs, Lung Cancer 68 (2010) 44–50. pmid:19559498
  16. 16. Lokhov P.G., Balashova E.E., Voskresenskaya A.A., Trifonova O.P., Maslov D.L., Archakov A.I., Mass spectrometric signatures of the blood plasma metabolome for disease diagnostics, Biomed Rep. 4 (2016) 122–126. pmid:26870348
  17. 17. Johnson CH, Patterson AD, Idle JR, et al. Xenobiotic metabolomics: major impact on the metabolome. Annu Rev Pharmacol Toxicol 2012;52:37–56. pmid:21819238
  18. 18. Koeth RA, Wang Z, Levison BS, et al. Intestinal microbiota metabolism of L-carnitine, a nutrient in red meat, promotes atherosclerosis. Nat Med 2013;19:576–85. pmid:23563705
  19. 19. Ridder L, van der Hooft JJ, Verhoeven S, et al. Automatic chemical structure annotation of an LCMS(n) based metabolic profile from green tea. Anal Chem 2013;85:6033–40. pmid:23662787
  20. 20. Hamdalla MA, Ammar RA, Rajasekaran S. A molecular structure matching approach to efficient identification of endogenous mammalian biochemical structures. BMC Bioinformatics 2015;16 Suppl 5:S11. pmid:25859612
  21. 21. Shin SY, Fauman EB, Petersen AK, et al. An atlas of genetic influences on human blood metabolites. Nat Genet 2014;46:543–50. pmid:24816252
  22. 22. Telomeres Mendelian Randomization Collaboration; Haycock PC, Burgess S, et al. Association Between Telomere Length and Risk of Cancer and Non-Neoplastic Diseases: A Mendelian Randomization Study. JAMA Oncol 2017;3:636–51. pmid:28241208
  23. 23. Day FR, Thompson DJ, Helgason H, et al. Genomic analyses identify hundreds of variants associated with age at menarche and support a role for puberty timing in cancer risk. Nat Genet 2017;49:834–41. pmid:28436984
  24. 24. Krumsiek J, Suhre K, Evans AM, et al. Mining the unknown: a systems approach to metabolite identification combining genetic and metabolic information. PLoS Genet 2012;8:e1003005. pmid:23093944
  25. 25. Bowden J, Holmes MV. Meta-analysis and mendelian randomization: a review. Res Synth Methods (2019) 10(4):486–96. pmid:30861319
  26. 26. Chen Y, Lu T, Pettersson-Kymmer U, Stewart ID, Butler-Laporte G, Nakanishi T, et al. Genomic atlas of the plasma metabolome prioritizes metabolites implicated in human diseases. Nat Genet. 2023 Jan;55(1):44–53. pmid:36635386
  27. 27. Genomes Project C, et al. A global reference for human genetic variation. Nature. 2015;526(7571):68–74. pmid:26432245
  28. 28. Cai J, Li X, Wu S, Tian Y, Zhang Y, Wei Z, et al. Assessing the causal association between human blood metabolites and the risk of epilepsy. J Transl Med (2022) 20 (1):437. pmid:36180952
  29. 29. Rasooly D, Patel CJ (2019) Conducting a reproducible Mendelian randomization analysis using the R analytic statistical environment. Curr Protoc Hum Genet 101(1):e82. pmid:30645041
  30. 30. Burgess S, Thompson SG. Interpreting findings from mendelian randomization using the MR-Egger method. Eur J Epidemiol. 2017;32(5):377–89. pmid:28527048
  31. 31. Verbanck M, et al. Detection of widespread horizontal pleiotropy in causal relationships inferred from mendelian randomization between complex traits and diseases. Nat Genet. 2018;50(5):693. pmid:29686387
  32. 32. Butler LM, Perone Y, Dehairs J, Lupien LE, De Laat V, Talebi A, et al. Lipids and cancer: Emerging roles in pathogenesis, diagnosis and therapeutic intervention. Adv Drug Deliv Rev 2020;159:245–93. pmid:32711004
  33. 33. Zaidi N, Lupien L, Kuemmerle NB, Kinlaw WB, Swinnen JV, Smans K. Lipogenesis and lipolysis: the pathways exploited by the cancer cells to acquire fatty acids. Prog Lipid Res 2013;52:585–9. pmid:24001676
  34. 34. Scott JS, Nassar ZD, Swinnen JV, Butler LM. Monounsaturated Fatty Acids: Key Regulators of Cell Viability and Intracellular Signaling in Cancer. Mol Cancer Res. 2022 Sep 2;20(9):1354–1364. pmid:35675039.
  35. 35. Li S, Zhou T, Li C, Dai Z, Che D, Yao Y, et al. High metastatic gastric and breast cancer cells consume oleic acid in an AMPK dependent manner. PLoS One 2014; 9:e97330. pmid:24823908
  36. 36. Liotti A, Cosimato V, Mirra P, Cal G, Conza D, Secondo A, et al. Oleic acid promotes prostate cancer malignant phenotype via the G protein-coupled receptor FFA1/GPR40. J Cell Physiol 2018;233:7367–78. pmid:29663374
  37. 37. Chen M, Zhang J, Sampieri K, Clohessy JG, Mendez L, Gonzalez-Billalabeitia E, et al. An aberrant SREBP-dependent lipogenic program promotes metastatic prostate cancer. Nat Genet 2018;50:206–18. pmid:29335545
  38. 38. Pisanu ME, Maugeri-Sacca M, Fattore L, Bruschini S, De Vitis C, Tabb E, et al. Inhibition of Stearoyl-CoA desaturase 1 reverts BRAF and MEK inhibitioninduced selection of cancer stem cells in BRAF-mutated melanoma. J Exp Clin Cancer Res 2018;37:318. pmid:30558661
  39. 39. Chen L, Ren J, Yang L, Li Y, Fu J, Li Y, et al. Stearoyl-CoA desaturase-1 mediated cell apoptosis in colorectal cancer by promoting ceramide synthesis. Sci Rep 2016;6:19665. pmid:26813308
  40. 40. Von Roemeling CA, Marlow LA, Wei JJ, Cooper SJ, Caulfield TR, Wu K, et al. Stearoyl-CoA desaturase 1 is a novel molecular therapeutic target for clear cell renal cell carcinoma. Clin Cancer Res 2013;19:2368–80. pmid:23633458
  41. 41. She K, Fang S, Du W, Fan X, He J, Pan H, et al. SCD1 is required for EGFRtargeting cancer therapy of lung cancer via re-activation of EGFR/PI3K/AKT signals. Cancer Cell Int 2019;19:103 pmid:31019378
  42. 42. Centenera MM, Scott JS, Machiels J, Nassar ZD, Miller DC, Zinonos I, et al. ELOVL5 is a critical and targetable fatty acid elongase in prostate cancer. Cancer Res 2021;81:1704–18. pmid:33547161
  43. 43. Cornelis MC, Kacprowski T, Menni C, Gustafsson S, Pivin E, Adamski J, et al. Genome-wide association study of caffeine metabolites provides new insights to caffeine metabolism and dietary caffeine-consumption behavior. Hum Mol Genet. 2016 Dec 15;25(24):5472–5482. pmid:27702941.
  44. 44. Thorn C.F., Aklillu E., McDonagh E.M., Klein T.E. and Altman R.B. (2012) PharmGKB summary: caffeine pathway. Pharmacogenet. Genomics, 22, 389. pmid:22293536
  45. 45. Gunes A. and Dahl M.L. (2008) Variation in CYP1A2 activity and its clinical implications: influence of environmental factors and genetic polymorphisms. Pharmacogenomics, 9, 625–637. pmid:18466106
  46. 46. Thorn C.F., Aklillu E., Klein T.E. and Altman R.B. (2012) PharmGKB summary: very important pharmacogene information for CYP1A2. Pharmacogenet. Genomics, 22, 73–77. pmid:21989077
  47. 47. Perera V., Gross A.S. and McLachlan A.J. (2012) Measurement of CYP1A2 activity: a focus on caffeine as a probe. Curr. Drug Metab., 13, 667–678. pmid:22554278
  48. 48. Ogasawara H, Nishio K, Ishida T, Arioka H, Fukuoka K, Saijo N. In vitro enhancement of antitumor activity of a water-soluble duocarmycin derivative, KW-2189, by caffeine-mediated DNA-repair inhibition in human lung cancer cells. Jpn J Cancer Res. 1997; 88:1033–7. pmid:9439677
  49. 49. Seow A, Zhao B, Lee EJ, Poh WT, Teh M, Eng P. Cytochrome P4501A2 (CYP1A2) activity and lung cancer risk: a preliminary study among Chinese women in Singapore. Carcinogenesis. 2001; 22:673–7. pmid:11285205