Early release, published at www.cmaj.ca on November 28, 2016. Subject to revision. CMAJ Commentary

Inflammation, cardiovascular and cancer: moving toward predictive

Paul M. Ridker MD MPH

See also www.cmaj.ca/lookup/doi/10.1503/cmaj.160313

n this issue, Singh-Manoux and colleagues1 ity: AGP, albumin, very-low-density lipoprotein Competing interests: Paul report that measures of low-grade systemic particle size and citrate. However, because con- Ridker is listed as a co-inventor on patents held I inflammation (e.g., C-reactive protein comitant measures of CRP and IL-6 were not by the Brigham and [CRP] and interleukin-6 [IL-6]) in midlife were done in the Estonian and Finnish cohorts, clinical Women’s Hospital that strong independent predictors of all-cause, car- comparisons could not be made. These kind of relate to the use of diovascular and cancer-related mortality in the data are crucial because AGP and albumin both inflammatory biomarkers in cardiovascular disease and Whitehall II study cohort. These findings sup- correlate with systemic low-grade inflammation that have been port evidence from more than a dozen prior pro- and thus in turn with IL-6 and CRP. licensed to Seimens and spective cohort studies with regard to CRP and In this context, Singh-Manoux and colleagues AstraZeneca. He has 2 received research grants all-cause mortality and more than 50 prior report that all three inflammatory biomarkers from the US National cohort studies showing that CRP and IL-6 pre- were associated with all-cause as well as cardio- Heart, Lung, and Blood dict future and stroke.3,4 vascular and cancer-related mortality, both in Institute and from Amgen, Novartis, AstraZeneca, Because CRP and IL-6 are strongly correlated univariate analyses and in analyses controlling Pfizer and Kowa. (both reflect upstream activation from interleu- for traditional risk factors. Overall, these effects kin-1),5 there is no clinical need to measure both weakened over time, with stronger associations This article was solicited and has not been peer factors, and of the two, CRP measured with a in the first five years than in longer follow-up reviewed. high-sensitivity assay (hs-CRP) is less expen- periods. However, when the authors controlled Correspondence to: Paul sive and has regulatory approval for clinical use. for all covariates and biomarkers simultaneously, Ridker, pridker@partners. Following the multi-national JUPITER trial AGP was no longer predictive. By contrast, the org (Justification for the Use of Statins in Preven- magnitude of effects for IL-6 and CRP were CMAJ 2016. DOI:10.1503​ tion: an Intervention Trial Evaluating Rosuvas- largely similar for both cancer-related and car- /cmaj.161033 tatin), which showed large reductions in relative diovascular mortality in the fully adjusted mod- risk for first-ever cardiovascular events (myo- els (with small remaining differences between cardial infarction, stroke and cardiovascular the two likely reflecting intercorrelations ) in the rosuva­statin group among partici- between IL-6 and CRP). pants with low levels of cholesterol but elevated The analysis from Singh-Manoux and col- hs-CRP, the 2009 Canadian Cardiovascular leagues is thus an important reminder that data Society guideline endorsed hs-CRP screening from metabolomics studies need to ensure that ap- for cardiovascular risk prediction, particularly propriate comparisons are made to established among patients at intermediate risk.6 risk markers. Indeed, the metabolomics field has Surprisingly, the third inflammatory bio- suffered from low levels of external validation. In marker in the study by Singh-Manoux and col- this regard, it is worth comparing the earlier Esto- leagues — α1-acid glycoprotein (AGP) — did not nian data with those reported by Cheng fare as well as either CRP or IL-6. This is impor- and colleagues8 from the Offspring Cohort of the tant because the authors’ motivation to perform the new analyses using the Whitehall II study Key points cohort was to confirm or reject data from a recent • Metabolomic and proteomic studies are furthering the understanding metabolomics study that found AGP to be the of “predictive medicine.” strongest predictor of mortality in a nuclear mag- • Inflammation is a biologic determinant of both cardiovascular disease netic resonance spectroscopy evaluation of 106 and cancer and can be identified by measuring high-sensitivity candidate biomarkers.7 That study, which used C-reactive protein or interleukin-6. metabolomic discovery data from the Estonian • Ongoing clinical trials will determine whether reducing inflammation Biobank and validated significant biomarkers in a reduces vascular event rates; if so, the clinical community will need to modify current concepts of residual risk to encompass residual cholesterol population-based cohort from Finland, suggested risk as well as the emerging concept of residual inflammatory risk. that four biomarkers predicted all-cause mortal-

All editorial matter in CMAJ represents the opinions of the authors and not necessarily those of the Canadian­ Medical Association or its subsidiaries.

© 2016 Joule Inc. or its licensors CMAJ 1 Commentary

Framingham Heart Study, where higher concen- ­current concepts of residual risk to encompass trations of isocitrate, an intermediate of the citric not only residual cholesterol risk but also the acid cycle, were associated with lower odds of emerging concept of residual inflammatory longevity. risk.10 This distinction leads to different thera- What are the clinical implications of the cur- pies being used for different patients and repre- rent data? Measure of inflammation as a tool for sents an important move toward personalized cancer screening has found limited clinical util- cardiovascular medicine. ity, although immune-modulating for cancer are a major new form of treatment. By References contrast, since 2000, it has been clear that mea- 1. Singh-Manoux A, Shipley MJ, Bell JA, et al. Association between inflammatory biomarkers and all-cause, cardiovascular and can- sure of inflammation can be effective for cardio- cer-related mortality. CMAJ 2016 Nov. 28 [Epub ahead of print]. vascular screening, and more recently for the doi:10.1503​/cmaj.160313. allocation of statin . In fact, risk predic- 2. Ridker PM. High sensitivity C-reactive protein as a predictor of all-cause mortality: implications for research and patient care. tion algorithms for primary prevention that inte- Clin Chem 2008;54:234-7. grate data on inflammation (e.g., the Reynolds 3. Emerging Risk Factors Collaboration; Kaptoge S, Di Angelanto- nio E, Lowe G, et al. C-reactive protein concentration and risk Risk Score) consistently outperform traditional of coronary heart disease, stroke, and mortality: an individual risk prediction scores such as the Framingham participant meta-analysis. Lancet 2010;375:132-40. 4. Kaptoge S, Seshasai SR, Gao P, et al. Inflammatory cytokines Risk Score and the Pooled Cohort Equations and risk of coronary heart disease: new prospective study and from the American Heart Association and Amer- updated meta-analysis. Eur Heart J 2014;35:578-89. 9 5. Ridker PM. From C-reactive protein to interleukin-6 to interleu- ican College of . kin-1: moving upstream to identify novel targets for atheropro- It is unusual for biomarker development pro- tection. Circ Res 2016;118:145-56. 6. Genest J, McPherson R, Frohlich J, et al. 2009 Canadian Cardiovas- grams to result in a tool useful for risk predic- cular Society/Canadian guidelines for the diagnosis and treatment tion. However, biomarker discovery is crucial of dyslipidemia and prevention of cardiovascular disease in the for thinking about new treatment targets. With adult — 2009 recommendations. Can J Cardiol 2009;25​ :567-79. 7. Fischer K, Kettunen J, Würtz P, et al. Biomarker profiling by regard to AGP, CRP and IL-6, what remains un- nuclear magnetic resonance spectroscopy for the prediction of all- certain is whether reducing inflammation can re- cause mortality: an observational study of 17 345 persons. PLoS Med 2014;11:e1001606. duce cardiovascular event rates. This important 8. Cheng S, Larson MG, McCabe EL, et al. Distinct metabolomic issue has been taken up by several investigative signatures are associated with longevity in humans. Nat Commun 2015;6:6791. groups worldwide, and major hard-outcome tri- 9. DeFilippis AP, Young R, Carrubba CJ, et al. An analysis of cali- als are under way using agents such as low-dose bration and discrimination among multiple cardiovascular risk scores in a modern multiethnic cohort. Ann Intern Med 2015;​ methotrexate and colchicine (which are com- 162:266-75. monly used to treat rheumatoid arthritis and 10. Ridker PM. Residual inflammatory risk: addressing the obverse gout, respectively) as well as novel targeted side of the atherosclerosis prevention coin. Eur Heart J 2016;​ 37:1720-2. agents such as canakinumab (a human mono­ clonal antibody that targets interleukin-1 ). β Affiliation: Center for Cardiovascular Disease Prevention, If the results of these trials are positive, then Brigham and Women’s Hospital, Harvard , the clinical community will need to modify Boston, Mass.

2 CMAJ