Strategies to Identify Cardiovascular Risk and Improve Risk Prediction in Older Populations

Cardiovascular disease (CVD) is a leading cause of death and disability in the elderly. However, the pathophysiological processes causing atherothrombosis are less understood in the elderly. Whether platelets behave in the same way as in younger people and whether conventional CVD risk factors are also predictive in older people are unclear. There is a paucity of evidence to guide CVD risk prediction and management in older individuals, which is further complicated by a higher comorbidity burden, increasing frailty, and functional decline. These factors are likely to underlay the significant disparities in care and delays to treatment experienced by older populations with CVD. As western populations continue to age, the health and economic repercussions of CVD will only be further magnified in the future. The ability to accurately and efficiently predict CVD risk in the elderly is therefore imperative. The primary aims of this thesis were: (i) to explore how factors that predispose to atherosclerotic plaque formation such as platelet activation and traditional CVD risk factors change with ageing; (ii) to explore how our ability to predict cardiovascular risk subsequently changes with ageing; and, (iii) to develop novel strategies to assess CVD risk in elderly community-dwelling populations. In a prospectively designed age-controlled study, we investigated how platelet activation in the resting circulation and in response to platelet agonists differed in the young, middle aged and relatively healthy, community-dwelling elderly. Key age-related changes in platelet activation include elevated basal activity, hyperreactivity to adenosine diphosphate (ADP) and hypo-responsiveness to thrombin in the elderly. The last of these was tracked to changes in the basal and inducible expression of the PAR1 and PAR4 thrombin receptors, and the combination of hyperresponsiveness to ADP and hypo-responsiveness to thrombin appears to be a novel phenotype. ADP and thrombin are important targets of antiplatelet agents used in the treatment of CVD, and changes in these activation pathways may have implications for drug efficacy. There was a correlation between markers of inflammation and increased basal platelet activation and reduced thrombin reactivity in the elderly. Increased chronic inflammation may predispose to a prothrombotic platelet phenotype and provide mechanistic explanation for how platelet activation and the pathogenesis of atherothrombosis change with ageing. The age-related hyporeactivity to thrombin appears to be related to multiple factors: lower numbers of thrombin binding receptor GPIbα, impaired downstream ADP feed-forward loop, as well as thrombin receptor desensitisation evidenced by decreased receptor (PAR-1 and PAR-4) cleavage. This relationship between age, inflammation and platelet receptor down-regulation is novel. It is unclear whether the thrombin receptor hypo-reactivity identified in a healthy elderly population is adaptive or maladaptive. It is possible that this represents a protective response in healthy ageing that is lost in patients with cardiovascular disease. Understanding these age-related changes may pave the way for novel and age-specific antiplatelet strategies that selectively target these pathways in order to effectively reduce the burden of atherothrombotic disease in the elderly. As well as demonstrating important changes in age-related platelet activation, this thesis confirmed that the ability of traditional CVD risk factors to predict risk declines with ageing. Age and smoking appeared to be the most robust traditional CVD risk factors in people aged ≥70 years. However, traditional risk factors including hypercholesterolaemia were not predictive of CVD in the cohorts examined in this thesis, even when stratified by statin use. Rather, non-traditional risk factors such as frailty, chronic kidney disease, anaemia, iron deficiency and low albumin appear to have greater prognostic value in older populations. The well documented association between these novel risk factors and inflammation has been evidenced in prior literature, providing further support for the role of inflammation in the development of CVD in the elderly. As traditional risk factors become less predictive of CVD, the need for a novel age-appropriate risk calculator increases. Contemporary risk models fail to account for increasing frailty, and the changing distribution of cardiac risk factors and metabolic profiles in older people. We therefore developed the first age-specific risk calculator specifically for populations aged ≥70 years to predict major adverse cardiovascular outcomes (MACE) incorporating smoking, age, frailty, albumin, anaemia, renal failure and iron (SAFAARI). Our SAFAARI risk model estimated MACE with good discrimination and calibration in men aged ≥70 years, appropriately classifying men across a broad range of risks. The C-index was 0.71 (95%CI:0.68-0.74), with good agreement between predicted and observed MACE incidence (Hosmer-Lemeshow χ2(9)=10.33, p=0.32). Discrimination analyses showed the SAFAARI model performed better than the original and refitted Framingham equations (p<0.001). The advantage of the SAFAARI model is its incorporation of routine, inexpensive and easily obtainable clinical biomarkers, meaning that our model has the potential to be a cost-effective and practical risk assessment tool. Our calculator provides specific risk modelling that predicts cardiovascular events in older populations within a temporally clinically relevant 5-year time-period and includes multimorbid participants aged 70-96 years. The dominance of non-traditional risk factors in the SAFAARI model differentiates the model from existing CVD risk models, but also accounts for its robustness. We investigated other strategies to predict CVD risk in older populations given the declining utility of traditional CVD risk factors. We validated for the first time the role of CT coronary angiography (CTCA)-identified non-obstructive and obstructive coronary artery disease (CAD) to predict long term major adverse cardiac events (MACE) and all-cause mortality. We found that the presence and extent of CAD was a meaningful stratifier of both long-term mortality and MACE in patients aged