Systems Biology and network medicine can give us new insights into disease causation, but they have their limitations.
By Paul Cerrato, MA, senior research analyst and communications specialist and John Halamka, M.D., Diercks President, Mayo Clinic Platform
Scientists have literally spent centuries using a reductionistic approach to determine what causes disease. And much of this research has been very successful. By one definition, reductionism is “a philosophical and scientific approach that explains complex systems, ideas, or phenomena by breaking them down into their simpler, more fundamental parts.” Medicine has used this approach with countess diseases states, linking collections of signs and symptoms to single organs.
This approach is similar to how an auto mechanic repairs a car. As a diagnostician, a mechanic isolates a broken part and repairs or replaces it. Similarly, when a clinician detects an infection, they isolate a specific pathogen and attack it with antibiotics or antiviral agents. If cancer is diagnosed, the tumor becomes the target to be destroyed. If gastrointestinal bleeding is detected, we try to locate the source and stop the bleeding.
Although this strategy has saved many lives and reduced pain and suffering, it nevertheless treats the disease and not the patient. Patients have unique habits, lifestyle mistakes, environmental exposures, psychosocial interactions, and genetic predispositions.
The traditional approach to many physiological or biochemical imbalances based on this reductionistic model has its shortcomings. Correcting a low serum potassium, for instance, by administering the electrolyte does not necessarily address the underlying pathology causing the drop in potassium levels. The management of hypertension poses a similar dilemma for the one-size-fits-all mentality behind reductionism. National guidelines recommend lower systolic and diastolic blood pressure targets for the general public, leading many patients to be advised to lower their blood pressure even though they do not have heart disease and may never develop it.
A more wholistic systems approach would dictate a multidimensional risk assessment that individualizes recommendations. As Andrew Ahn and colleagues explain, “In reductionism, multiple problems in a system are typically tackled piecemeal. Each problem is partitioned and addressed individually. In coronary artery disease, for example, each known risk factor is addressed individually, whether it be hyperlipidemia or hypertension. The strategy is also extended to coexisting diseases, such as hypothyroidism, diabetes, and coronary artery disease. Each disease is treated individually, as if the treatment of one disorder (such as coronary artery disease) has minimal effects on the treatment of another (such as hypothyroidism). While this approach is easily executable in clinical practice, it neglects the complex interplay between disease and treatment. The assumption is that the results of treatments are additive rather than nonlinear.”
The systems biology paradigm is better suited to the management of chronic, complex diseases, as opposed to acute, simple disorders. Acute appendicitis, for instance, has an immediately curable pathology. Removal of the organ usually solves the problem, with no further need for long-term therapy.
On the other hand, disorders such as diabetes, coronary artery disease, and asthma will likely respond better to the in-depth analysis and comprehensive data gathering that are characteristic of systems based clinical care, sometimes referred to as network medicine. Diabetes, for instance, is a multidimensional disease involving genetic mutations, inflammation, leptin, cortisol, diet, body mass index, and a host of other contributing factors that all need attention.
The latest research is now merging systems biology with AI and multi-omics data to redefine specific diseases as molecular clusters or “modules” that take advantage of knowledge graphs and combine protein/protein interactions, genome-wide association studies (GWAS), and other omics data. The result is the emergence of previously hidden pathological mechanisms that look beyond the one cause-one disease perspective.
The practical benefits of this new approach include the development of repurposed drugs, the effects of which have been validated in randomized clinical trials. For example, with the help of AI-driven protocols, baricitinib, which was originally approved by FDA for rheumatoid arthritis, has been approved for the treatment of Covid infection in hospitalized patients.
Taking us to the next step, there is evidence to suggest that network medicine needs to be married to frontier AI models. The mechanistic models generated by network medicine have their shortcomings, the least of which is the fact that they have rarely been tested in prospective clinical trials to see if they actually improve patient outcomes. Right now, a hybrid approach that uses both methodologies looks like the best way forward.
Share this:
