The European Society of Cardiology’s Cardiovascular Risk Collaboration- Is it Relevant to Irish Healthcare Professionals?
Written by Ian M Graham, Professor of Cardiovascular Medicine, Trinity College Dublin, Co-Chair, ESC Cardiovascular Risk Collaboration
Take home messages
1. In 2020 the Board of the European Society of Cardiology (ESC) established a multidisciplinary Cardiovascular Risk Collaboration (CRC) in the European Heart health Institute in Brussels.
2. In apparently healthy persons, the risk of a heart attack or stroke relates strongly to the combined effects of smoking, cholesterol level and blood pressure, often combined with overweight, inactivity and consequent diabetes.
3. Clinical estimation of these effects is imperfect, therefore all guidelines on cardiovascular disease (CVD) prevention recommend some form of risk chart or calculator.
4. The previously used European Society of Cardiology’s (ESC) risk estimation system, SCORE, estimated 10 year risk of fatal CVD and there was a strong demand to estimate risk of total (fatal and non-fatal) CVD events.
5. The CRC responded to this by developing total event charts for both middle aged (SCORE2) and older persons (SCORE2-OP). These have been included in the 2021 ESC Guidelines on CVD Prevention in Clinical Practice 1 .
6. All current risk estimation systems are imperfect- they start too late, around age 40, are dominated by age effects, and are much better at estimating risk in populations than in individuals.
7. The CRC believes that the future of risk estimation may lie in applying artificial intelligence to both genetic and environmental factors to allow more personalised risk estimation from early life on, using an exposure time model rather than age per se.
Introduction
This paper outlines the rationale for the ESC’s Cardiovascular Risk Collaboration and its various work packages. We then present its first two completed deliverables, the SCORE2 2 and SCORE2-OP 3 (older persons) risk charts, and finish with some thoughts for the future.
The CRC
The Cardiovascular Risk Collaboration (CRC) was stablished by the Board of the ESC in 2020 to contribute to the ESC’s mission to reduce the burden on cardiovascular disease by researching new aspects of risk estimation and developing their practical application.
Because CVD generally arises from a combination of risk factors, and clinical estimation of these effects is poor, all CVD prevention guidelines recommend some sort of risk estimation system such as the European SCORE 4 or the American Pooled Cohort Equation. However, these systems are relatively crude, attempt to apply population data to individuals, and generally start at age 40- after 40 years of possible exposure to risk factors.
It was therefore decided to develop a multidisciplinary partnership to explore these issues under a steering Committee:
The CRC’s work is arranged through a number of work packages that include, among other activities, the development of new risk estimation systems, exploration of new data science tools such as artificial intelligence and applying them to new data sources, integration of imaging modalities into risk estimation, developing new decision support systems, providing resources for other European Heart Health Institute activities, and providing knowledge to assist developing countries.
We will now focus on the first two major projects delivered by the CRC, SCORE2 and SCORE2-OP (older persons). While the risk charts are presented here, a new interactive risk calculation App is in final test phase
SCORE2 2 :
The ESC’s original SCORE 2 system was a powerful brand but based on old data with a limited number of variables. It was criticised because it estimated only the risk of CVD mortality, and there was a strong demand for total event (fatal +nonfatal CVD) charts. This proved more challenging than one might have expected, due to the heterogeneity and very variable quality of nonfatal event data throughout Europe. It was based on 680,000 individuals from 45 cohorts in 13 countries. A multiplicative approach was used to convert mortality to total events. The process may be summarised from the graphical abstract from the main SCORE2 paper 4 :
Ireland is classified as a moderate risk country and the appropriate chart is as follows. It will be noted that non-HDL cholesterol is used rather than the total cholesterol in SCORE because the former relates somewhat more closely to risk:
SCORE2-OP 3 :
SCORE2-OP was developed in response to demands for a risk estimation system for use in older persons, in view of the ageing populations of Europe. It uses similar methods to SCORE 2 but was more challenging because of more limited and less reliable end point data in older people combined with competing causes of death. The graphical abstract illustrating its development 5 also illustrates the four risk regions of Europe used in both SCORE publications:
The chart appropriate to Ireland is as follows. It will be noted that gender differences are modest, perhaps because the highest risk men have already died.
Risk estimation: can we see the future?
All current risk estimation systems share substantial limitations-
• They are based on cohorts that usually start at around age 40, thus missing the effect of many years of potential exposure to risk.
• The techniques used, such as Cox, essentially derive multipliers and cannot allow for complex biological interactions within different risk factor combinations.
• The results apply to populations and their application to individuals is less certain.
• The effects are dominated by age which is of course essentially a measure of exposure time rather than a risk factor as such.
Based on the work and thoughts of Brian Ference, the CRC is attempting to address some of these issues. This approach regards risk as exposure time, such as units of cholesterol, BP or total risk over time. This reduces the problem of the domination of present systems by age. Future risk is determined by the rate of rise of these variables. Mendelian randomisation studies suggest that the rise in cholesterol and blood pressure from early life on is substantially determined by polymorhisms that may have a small impact on five year risk but a very large impact on lifetime, 50-70 year risk. To this must be added the progressively more important impact of lifestyle and environmental factors. It is planned to apply artificial intelligence techniques to these issues in an attempt to achieve more precise and personalised risk estimates applicable at a much earlier stage, perhaps in the 20s. Further, the gradient (expected rate of rise of risk) may help to determine subjects who could benefit from earlier lifestyle advice and perhaps easrly imaging such as CT calcium scoring, and others in whom this would be expected to have a very low yield. Some of these issues are addressed in an editorial that accompanied the SCORE2 papers. 6
In conclusion, we have presented an outline of the rationale and work plan of the ESC Cardiovascular Risk Collaboration. The new SCORE2 and SCORE2-OP risk estimation systems have been presented and we summarise what may be a radical new approach to risk estimation.
References
1. Visseren FLJ (Chairperson), Francois Mach* (Chairperson). 2021 ESC Guidelines on cardiovascular disease prevention in clinical practice European Heart Journal (2021) 00, 1 111, doi:10.1093/eurheartj/ ehab484.
2. SCORE2 working group and ESC Cardiovascular risk collaboration. SCORE2 risk prediction algorithms:new models to estimate 10-year risk of cardiovascular disease in Europe.
3. SCORE2-OP working group and ESC Cardiovascular risk collaboration. SCORE2-OP risk prediction algorithms: estimating incident cardiovascular event risk in older persons in four geographical risk region
4. Conroy RM, Pyorala K, Fitzgerald AP et al. Estimation of ten-year risk of fatal cardiovascular disease in Europe: the SCORE project. European Heart Journal 2003; 24:987-1003.
5. Graham IM, Di Angelantonio E, Visseren F et al. Systematic Coronary Risk Evaluation (SCORE). JACC Focus seminar: the best of population research studies 4/8. Journal of the American College of Cardiology 2021;77: 3046-57.
6. Tokozoglu L, Torp-Pedersen C. (Editorial). Redefining cardiovascular risk prediction: Is the crystal ball clearer now? European Heart Journal (2021) 00, 1–4, doi:10.1093/ eurheartj/ehab310
Read the full magazine HERE