Stefan Kääb outlines the different steps to developing comprehensive personalised medicine strategies in the context of cardiac arrhythmias and reviews what essential elements such strategies require. He spoke about this topic at the European Cardiac Arrhythmia Society’s annual meeting (20–22 April, Munich, Germany).
We all practice personalised medicine on a routine basis daily by applying different stratification algorithms such as age, gender, New York Heart Association functional class, ejection fraction and many others. Genome-based information adds a second level to this group building effort, which-in combination with a multitude of biomarkers including classical serum markers, molecular markers, imaging as well as ECG and more sophisticated mapping markers-will eventually lead to the improved assessment of individual disease risk (third level). A fourth level of personalised care includes differential (biomarker) based therapeutic interventions and may in the long term eventually lead to truly individualised therapy as the fifth level. A prerequisite for all levels of personalised medicine is standardised, detailed, and well documented, individualised patient characterisation (or phenotyping from a genetic point of view). This naturally includes detailed documentation of individual outcomes.
Individualised phenotyping in cardiac arrhythmias currently is crude in many instances, and, generally, it does not consider differential pathophysiological aspects or complex biomarkers. In atrial fibrillation, for example, we rely mostly on a classification by the duration of atrial fibrillation and rely little on related symptoms and hardly at all on the extent of atrial damage or the disease causing mechanisms. However, instead of this approach, as suggested recently, a classification of atrial fibrillation types based on the underlying pathophysiology could help to better select therapies for specific atrial fibrillation patients based on underlying cause and/or degree of atrial damage.1 Subgroups of atrial fibrillation based on underlying pathophysiology could include genetic, focal, complex and postoperative atrial fibrillation (while genetic factors may contribute to all causes to a variable degree).
In a multinational effort, the CHARGE-AF consortium has most recently identified six novel genomic susceptibility regions in a large genome-wide association study (GWAS), identifying transcription factors related to cardiopulmonary development, cardiac-expressed ion channels, and cell signalling molecules associated with atrial fibrillation.2 This information, together with three previously published genetic signals, provides a wealth of information, but detailed and functional characterisation are only at the beginning. Major efforts are currently under way to investigate the potential of these genetic markers to facilitate biomarker guided early treatment and upstream therapy (risk reduction) or more specifically biomarker guided ablation strategy.
Personalisation in atrial fibrillation in advanced structural heart disease, old age and comorbidities will need a more comprehensive approach to biomarkers in general-characterising the different pathophysiological processes involved, such as electrical, structural and contractile remodelling and taking the highly variable degrees of complexity into account. Integrating biomarkers to assess the degree and heterogeneity of atrial fibrosis, assessment of the atrial fibrillation substrate by direct contact mapping or body surface mapping together with imaging markers are essential to meet the need for quantification and specification of the “complexity” to guide safe and (cost) effective therapy in these forms of atrial fibrillation.3
The hype about personalised medicine is generated largely by the enthusiasm about the rapidly evolving technology to generate comprehensive genomic (personalised) data as well as by an urgent need for improved personalised risk stratification and therapeutic decisions-one for all will not be appropriate in the future. It is an important and humbling experience to memorise all the steps necessary to link genomic data to personalised risk assessment, prevention and biomarker-based therapeutic decisions. These steps among others include bioinformatics, fine mapping and functional evaluations in a systems biology approach, pharmacogenomics as well as ethical, social, and legal issues. After all, individualised therapeutic decisions need to demonstrate safety, superiority and cost-effectiveness (evidence-based) compared to established decision making.
With all caution and scepticism alongside improvingly well-characterised patients subgroups based on a multitude of biomarkers, there is a reason for hope for an improved arrhythmia management in the near future.
Stefan Kääb is professor of Medicine and Cardiology at the Department of Medicine I, at Ludwig-Maximilians University Hospital, Munich. Germany.
References
1. Kirchhof P, et al. Thromb Haemost 2011; 106:1012–9
2. Ellinor PT, et al. Nat Genet 2012. doi: 10.1038/ng.2261 Epub
3. Wakili R, et al. J Clin Invest 2011; 121:2955–68
