Precision Medicine for Aortic Stenosis

D espite all our success in cardiology, we are still, in some ways, working at an obtuse level. We try and find individual markers for disease pathology (1), risk (2), or for triaging patients for certain therapies (3), and when that is not possible, we create small multivariable statistical models for prediction (4). That strategy has worked to some extent, especially with imaging-related markers (3), but it has also raised questions about generalizability and comparative effectiveness (4,5). In addition, this is a slow trial and error process (needing to find a precise variable or two), and the time line for success is uncertain (needing to test under multiple conditions and together with multiple variables). We also understand success of therapies at a population level, hope that therapy applies to the individual patient, and often end up practicing imprecise medicine by extrapolating therapies to the wider population. This is well beyond the trial enrollment criteria or base therapies on triage variables that are themselves imprecise (e.g., triaging for implantable cardioverter-defibrillators in dilated cardiomyopathy based on ejection fraction [EF] alone). What can hasten the process of discovery and help in precise targeting of therapies to reliably modify outcome? Perhaps a finer understanding of diseases and a major redirection of strategy using large-scale multidisciplinary analytical methods of detecting critical variables or association of variables could immensely speed up this journey of discovery? A better understanding of disease as the basis of