Education in Heart Imaging as a surrogate marker of drug efficacy in Heart: first published as 10.1136/heartjnl-2017-311213 on 31 October 2018. Downloaded from cardiovascular disease Jason M Tarkin,1,2 Marc R Dweck,3 James H F Rudd1 1Division of Cardiovascular INTRODUCTION Learning objectives Medicine, University of Many cardiovascular drugs in the pipeline will fail to Cambridge, Addenbrooke’s demonstrate a clear clinical benefit when evaluated in Hospital, Cambridge, UK ► Learn how imaging can be applied to gain 2 large-scale clinical outcome trials, which are costly, National Heart & Lung Institute, early insights into drug efficacy and inform the Hammersmith Hospital, Imperial require lengthy follow-up and can potentially expose design of phase III clinical outcome trials in College London, London, UK patients to unforeseen risks. There exists an enor- 3 cardiovascular disease. Centre for Cardiovascular mous gap between early mechanistic studies demon- Science, University of Edinburgh, ► Understand which markers of atherosclerotic strating proof-of-principle drug efficacy in preclinical Little France Crescent, disease severity are most useful as imaging models and successful translation of these therapies Edinburgh, UK endpoints in drug intervention studies. into everyday clinical practice. To help overcome this ► Learn about the emerging role for molecular Correspondence to challenge, cardiovascular imaging techniques can be imaging of inflammation and disease activity in Dr Jason M Tarkin, Division applied to quantify early changes in disease severity cardiovascular drug development. of Cardiovascular Medicine, owing to drug intervention, or lack thereof, with the University of Cambridge, Addenbrooke’s Hospital, aim of informing subsequent clinical outcome trials. Cambridge CB2 2QQ, UK; This approach can be used to direct valuable resources jt545@ cam. ac. uk towards development of drugs most likely to provide and strong prognostic link to important clinical real clinical impact. The rationale here is that ‘surro- outcomes. Importantly, an imaging biomarker should gate’ imaging outcomes can be powered using far also reflect and track the mechanism of the tested less subjects than clinical outcomes in drug trials, as drug. For example, measurement of plaque lipid each participant will contribute an imaging endpoint content would be an appropriate biomarker for regardless of whether they then go on to develop a studying lipid-lowering therapies. The relative clinical event. In addition, drug efficacy can be more change in response to drug intervention for an ideal rapidly tested using imaging markers as there is no imaging endpoint should also be detected within a need to wait long periods of time for clinical outcomes relatively short time-frame. Methods for quantifi- to occur. cation of atherosclerotic disease severity have been comprehensively reviewed elsewhere1 2; here we Imaging endpoints in clinical trials might also be http://heart.bmj.com/ used in the future to identify specific subgroups of focus on their ability to predict clinical events. patients who are more likely than others to respond to targeted pharmacotherapies in cardiovascular Validation of imaging for risk prediction disease—the so-called precision medicine. Indeed, Among the most widely used imaging endpoints better methods are needed to identify those patients for cardiovascular drug trials are arterial inflam- with cardiovascular disease who are most at risk of mation, vascular intima media thickness (IMT), future or recurrent, clinical events despite secondary plaque burden (or atheroma volume) and plaque prevention. Many of these patients will have ‘residual’ morphology. on September 24, 2021 by guest. Protected copyright. on-treatment risk and could benefit from higher inten- Vascular inflammation is the earliest modifiable sity lipid lowering or anti-inflammatory therapies link between clinical cardiovascular risk factors and currently under evaluation in atherosclerosis. Mecha- disease activity that can be detected using imaging. 18 nisms underlying on-treatment residual risk are widely When imaged using F-fluorodeoxyglucose (FDG) heterogeneous and patient specific, with different positron emission tomography (PET), arterial disease substrates (ie, thrombotic tendency, lipid accu- inflammation can offer prognostic information mulation and inflammation) contributing in varying beyond clinical risk factors, including Framingham degrees to an individual’s future cardiovascular risk. risk score, with an increased HR of 2.9–4.7 for the 3 4 Consequently, applying targeted antiatherosclerotic highest risk groups in large retrospective analyses. drugs on top of standard therapies broadly in unse- IMT provides a measure of local atherosclerotic lected patient populations is likely to produce at most burden, including early subclinical disease, which a modest impact on clinical outcomes. This article will has also been correlated with risk of future myocar- © Author(s) (or their dial infarction (MI) and stroke.5 6 However, the link employer(s)) 2018. Re-use discuss the potential scope of imaging to improve drug permitted under CC BY. efficacy testing of current and emerging disease-modi- between vascular IMT and future cardiovascular 7 Published by BMJ. fying therapies in atherosclerosis. risk remains unproven, and carotid IMT might represent vascular changes arising from arterial To cite: Tarkin JM, hypertension rather than a direct marker of athero- Dweck MR, Rudd JHF. 8 Heart Epub ahead of IMAGING ENDPOINTS sclerosis per se. print: [please include Day The ideal imaging endpoint for use in any cardio- Total plaque burden is in fact the strongest prog- Month Year]. doi:10.1136/ vascular drug trial should be easily measurable and nostic indicator that has been identified in large heartjnl-2017-311213 highly reproducible, with sound biological rationale prospective imaging trials.9–11 While the presence Tarkin JM, et al. Heart 2018;0:1–12. doi:10.1136/heartjnl-2017-311213 1 Education in Heart of high-risk plaque features associated with the Exposure to ionising radiation, additional risks Heart: first published as 10.1136/heartjnl-2017-311213 on 31 October 2018. Downloaded from histological appearance of ‘vulnerable’ rupture- associated with invasive imaging procedures and prone thin-cap fibroatheromas (TCFAs) are also local accessibility to imaging technology are other predictors of major adverse cardiovascular events factors to consider when choosing between surro- (MACE),10–12 it remains unclear whether identifying gate imaging endpoints for cardiovascular drug individual plaque characteristics is of incremental trials. While plaque volume and composition can value to simpler assessments of plaque burden, be more precisely quantified using invasive versus as at the plaque level, this approach is limited by non-invasive coronary imaging, it is worth bearing poor positive predictive value.13 Indeed, the vast in mind that there is also a high ~25% partici- majority of coronary artery TCFAs identified using pant dropout in contemporary invasive imaging virtual histology (VH)-intravascular ultrasound studies.21 In addition, the concept of an overall (IVUS) do not go on to cause clinical events because ‘barometer’ of disease severity that might be modi- they either heal or rupture silently without clinical fiable with drug intervention can be more readily sequelae.14 attained using non-invasive than invasive imaging, New imaging techniques may offer opportunities for example, with PET or MRI, where the entire to measure plaque lipid content, both invasively vascular bed can be imaged simultaneously. using near infrared spectroscopy and non-invasively using carotid MRI with T2 mapping.15 Quantifica- USE OF IMAGING FOR TESTING DRUG EFFICACY tion of pericoronary adipocyte content and inflam- IN ATHEROSCLEROSIS mation could provide additional surrogate markers Here we discuss how various imaging biomarkers of cardiovascular risk for use in clinical drugs trials have been applied in clinical drug trials to study the 16 in the future. efficacy of disease modifying therapies in athero- Degree of intraluminal stenosis, ischaemic sclerosis, including both long-established and newly burden and coronary artery calcification (CAC) tested lipid-lowering and anti-inflammatory agents. are additional imaging markers that have been tested as surrogate endpoints in drug trials. While angiographic stenosis severity and functional isch- Lipid-lowering and other drugs affecting aemia are among the most important factors used cholesterol to guide everyday clinical management decisions, Statins particularly when contemplating coronary revascu- Statins reduce low-density lipoprotein cholesterol larisation, they represent a late stage in the disease (LDL-c) through inhibition of β-Hydroxy β-meth- process that is not easily modifiable by drug inter- ylglutaryl-CoA (HMG-CoA (HMG CoA)reductase vention. Moreover, although there is a well-estab- and have been proven in landmark clinical trials lished association between ‘flow-limiting’ coronary to dramatically reduce the incidence of cardio- vascular events in a range of individuals,22 with disease and hard clinical outcomes, this relationship http://heart.bmj.com/ might not be causal. Indeed, reversal of coronary greater benefit seen for intensive versus moderate or low-dose therapy in patients with stable angina ischaemia with drug intervention and percutaneous 23 24 coronary stenting in patients with stable angina or previous