Author + information
- Masafumi Kitakaze, MD, PhD∗ ()
- ↵∗Reprint requests and correspondence:
Dr. Masafumi Kitakaze, Department of Cardiovascular Medicine, National Cerebral and Cardiovascular Center, 5-7-1 Fujishirodai, Suita 565-8565, Japan.
Personalized medicine will be integral in further driving down the global impact of cardiovascular disease. Basic and clinical research in cardiovascular medicine has rapidly progressed over the last several decades, causing decreases in mortality and morbidity. Longevity is a positive consequence of the battle against the cardiovascular diseases, but the burgeoning, aging population in the forthcoming era may shoulder some unintended consequences. In all developed countries, both aging and an increase in metabolic syndrome will cause serious concerns, because the cardiovascular diseases directly related to these conditions will result in major adverse outcomes. Therefore, we must find novel approaches to prevent and treat cardiovascular diseases. Most importantly, we need to learn how to deliver the beneficial outcomes discovered in basic and clinical science to daily medical practice for each, unique patient with cardiovascular disease, which is commonly known as personalized medicine. In my opinion, refining personalized medicine to the point that it can prevent disease or predict outcomes will require the mathematization of cardiovascular medicine.
First, consider the structure of cardiovascular science or medicine. When we assess the spectrum of cardiovascular diseases, we believe that each cardiologist must apply a 3-dimensional philosophy to decrease the risks of either mortality or morbidity for each patient. As an analogy, the x-, y-, and z-axes can be compared to 3 methods upon which cardiologists may focus their research efforts: 1) types of research; 2) methods of data collection; and 3) mathematization of cardiovascular medical practice.
As for the x-axis, or types of research, our studies are always classified as 1 of the following: 1) epidemiological cohort studies; 2) clinical studies, including observational and randomized studies; 3) translational studies to link the results of basic studies to varied clinical arenas; 4) physiological studies including human, as well as large and small animals; 5) molecular or biological studies; and 6) genetic studies. All types of research are very important for best practices in cardiovascular medicine, and papers in JACC often focus on study types 1 to 3, supported by the results of study types 4 to 6. However, we, as researchers, should not be confined to any single type of study. With a nimble approach, we should evolve with the demands of the field to generate the best answers for the most difficult questions of how to improve cardiovascular outcomes for each patient. This flexibility of an investigator to move from basic to clinical research is truly necessary to define novel management strategies for cardiovascular diseases.
Within each type of research, the method of data collection, or the y-axis, plays a tremendously important role. Data acquisition of genes, basic research, or clinical data can alter the quality and quantity of clinical medicine or research. To enhance our quality, data-mining methods, typically used in the realms of business or economics, can be applied to cardiovascular research to discover novel management strategies or to assess the true efficacy of clinical medicine (1,2).
Finally, we come to the z-axis, which I believe to be the most important, because all of clinical medicine should be able to be mathematically expressed. Clinical medicine should be attributable to natural science, not to the summation of experiences or statistics, and natural science should be mathematically expressed. However, the relationship between each factor of clinical medicine and clinical outcome is qualitative, not quantitative, and each essential factor cannot accurately depict or predict the outcomes of each unique patient. For example, the patients with plasma brain natriuretic peptide levels <170 pg/ml showed a 3.4 times higher survival rate during 2.2 years compared with patients with BNP levels ≥170 pg/ml in patients with chronic heart failure (3). This type of analysis only provides the average tendency of survival for the average patient, but we are obliged to apply this type of knowledge in clinical practice. The use of statistically proven evidence for each patient is a characteristic unique to clinical medicine, because the results and outcomes of basic science in medicine are both qualitatively and quantitatively reproducible. Furthermore, when we evaluate the other basic science fields, such as physics, chemistry, or its applied sciences of technology including material mechanics, thermodynamics, or fluid dynamics, both basic and applied sciences are sufficiently quantitative. Importantly, the observational phenomena in basic or applied science can be depicted by a mathematical equation, such as the law of universal gravitation.
The time it took for an apple to fall from a tree to the ground was assumed to be dependent on the gravity, weight, color, or shape of the apple, as well as the initial speed of the apple. However, Newton found that only 2 factors—gravity and the initial speed of the apple—exclusively determined the time it would take for the apple to reach the ground. Because clinical medicine and practice are classified as natural science, all phenomena, such as the severity of heart failure and patient characteristics before the occurrence of clinical events, may provide the qualitative equation for a clinical outcome. Therefore, I believe that all of clinical medicine or practice can be mathematically expressed. Because the mathematics of a cause-and-effect relationship predict accurate clinical outcomes, we can correct for risk factors of patients before the clinical events might occur. One may claim that there is a disparate personal response to therapy, because humans are not machines. However, this may not be a relevant argument, because if single nucleotide polymorphism or genetic differences impact clinical outcomes, we should consider the action of the drugs or the response to the diseases mathematically normalized by single nucleotide polymorphism or genetic difference. The constitution of a mathematical model or equation is critically important, because this may contribute to personalized medicine and determine whether our clinical medicine is classified in natural science or not (4).
Our group has investigated the possibility of deriving a mathematical formula for the estimation of prognosis of the patients with heart failure. For this purpose, we formulated the equation τ = f(x1,…, xp), where x1,…,xp are clinical features and τ represents the clinical outcome for heart failure, and we attempted to determine the function (f) to mathematically formulate the relationship between the clinical features and outcomes for these patients. The mathematical analysis was performed through a probabilistic modeling of the relational data by assuming a Poisson process for rehospitalization due to heart failure and by linearly approximating the relationship between the clinical factors and the mean elapsed time to rehospitalization. We collected and analyzed 402 clinical parameters and identified 252 factors that substantially influenced the elapsed time until rehospitalization. With the probability model based on the Poisson process, we found that the actual times to rehospitalization tightly correlated to estimated elapsed times to rehospitalization. Therefore, we seemed to establish a mathematical formula that closely predicted the clinical outcomes of patients who were hospitalized with heart failure and discharged after appropriate treatment.
This investigation is only 1 example of the mathematization of medicine. However, we believe it is fascinating not only because we can predict outcomes using clinical parameters beforehand, but also because we can optimize various treatments to extend the days to hospitalization in patients with heart failure. The biggest reason for the importance of this mathematization is the emerging need to personalize medicine for every patient.
It is difficult to prove that a particular mathematical formula is correct, as no finalized answer is ever obtainable in medical science. To help overcome this problem, we are collecting prospective data on patients with heart failure in an effort to predict clinical outcomes. If this equation is proved correct, the application of these risk factors to individual cardiovascular patients may allow us to distinguish patients who are at low risk from those who are at high risk, and the patients may benefit from closer monitoring and aggressive treatment. Furthermore, if this process were successful, we could adopt personalized medicine for each cardiovascular patient.
This equation example demonstrates that clinical medicine or practice is a part of natural science, although to this point, clinical outcomes have been thought to be extracted mainly from medical knowledge and the experience of the physicians.
The question is how JACC contributes to these movements. First, researchers need to forge ahead with sophisticated translational or clinical investigations. This type of dedication to the research process remains important to patch together our fragments of knowledge with regard to cardiovascular diseases for a complete understanding of cardiovascular medicine for each patient. In addition, an expert will contextualize how these data in JACC papers fit into the clinical or translational sciences of cardiovascular medicine through an editorial comment and the perspectives section of each original investigation. Second, JACC will continue to take a global approach, publishing research from varied regions with less or more prevalence of cardiovascular disease to shed light on how to help people with cardiovascular diseases worldwide.
In addition to publishing investigations from the individual researchers, the Journal will seek to extend its own philosophy about clinical medicine and science to decrease the mortality or morbidity of cardiovascular diseases. The philosophies of x-, y-, and z-axes allow us to start to break through current unresolved issues in cardiovascular medicine. Finally, as for the most important z-axis, to consider clinical situations or individual presentations of certain patients through a mathematical lens will be extremely critical to maximize the goal of personalized medicine. To achieve the true rewards of our efforts with x and y, z will be necessary. These paths lead to better care for patients with cardiovascular disease. Now, let's do it.
- American College of Cardiology Foundation
- Kim J.,
- Washio T.,
- Yamagishi M.,
- et al.
- Kim J.,
- Ogai A.,
- Nakatani S.,
- et al.
- Maeda K.,
- Tsutamoto T.,
- Wada A.,
- et al.