If two patients take the exact same medication, why does one recover while the other sees little to no improvement?
Despite the development of advanced biologics and targeted therapies for autoimmune conditions such as rheumatoid arthritis and lupus, a significant number of patients fail to achieve relief—even after trying multiple treatments in sequence.
A South Korean research team has uncovered the biological mechanism behind this frustrating clinical challenge. A team led by Professors Jung Sung-soo, Jeon Chan-hong, and Jung Hye-min from the Division of Rheumatology at Soonchunhyang University Bucheon Hospital announced on the 25th that "treatment nonresponse" in rheumatic diseases is not caused by a failure in a single inflammatory pathway. Instead, it results from the entanglement of multiple internal biological systems—including the immune, nervous, and metabolic networks. The researchers confirmed this system-wide mechanism using a mathematical simulation model.
Their study was published in the August 15 issue of npj Systems Biology and Applications, a peer-reviewed journal published by Springer Nature.
Why Patients with the Same Disease Respond Differently to Medication
Among patients with rheumatoid arthritis, some fail to achieve remission—a state where disease symptoms virtually disappear—despite rotating through various targeted therapies.
In 2021, the European Alliance of Associations for Rheumatology (EULAR) formally defined these refractory cases as "difficult-to-treat rheumatoid arthritis (D2T RA)." Traditionally, treatment development for rheumatic conditions focused on suppressing specific cytokines—signaling molecules released by immune cells—or precisely targeting and blocking a single inflammatory pathway.
However, real-world clinical practice often deviates from this model. Some patients present normal blood inflammatory markers alongside severe pain and fatigue, while others report improved physical symptoms despite persistently high inflammatory levels.
Recognizing these discrepancies, the research team hypothesized that treatment nonresponse is not a localized malfunction of one specific pathway, but rather a state where multiple biological systems interact to lock the body into a persistent diseased state.
The Body’s Interconnected Biological Axes
The researchers compressed the body's intricate network into three primary axes comprising six core variables:
An axis involving the gut mucosa, microbiome, and immune balance that prevents self-attack;
An axis intertwining the nervous system, adipose tissue, and immune cells;
An axis regulating cellular metabolism and stress responses.
They named this integrated mathematical model the "3-axis integrated framework (3-AIF)."
Computer simulations of the model demonstrated that human health does not split neatly into binary categories of "healthy" or "diseased." Instead, the body can settle into several distinct states: complete health, low-symptom partial recovery, active disease, and severe treatment resistance.
Much like a ball coming to rest in different mountain valleys depending on where it falls, patients with the same initial diagnosis settle into different biological "valleys" based on genetic background, environmental factors, and disease progression. Consequently, their response to treatment diverges sharply.
Why the Body Struggles to Shift Back to Health
When pathological triggers surpass a certain threshold, the body shifts from a healthy state to a diseased state; as triggers intensify, it moves deeper into treatment resistance.
Critically, once the body shifts into a diseased state, simply reducing or removing the initial trigger is often insufficient to restore health.
The research team referred to this phenomenon as "hysteresis"—meaning the biological pathway driving disease onset differs from the path required to reverse it. Mathematically, this underscores the vital importance of early diagnosis and intervention before the body crosses this biological threshold.
Notably, among the six model variables, the top three factors influencing disease state were concentrated in the axis connecting the nervous system, adipose tissue, and immune cells. Central to this network was the vagus nerve, which helps suppress inflammation through the parasympathetic nervous system. Vagus nerve activity emerged as the single most critical variable in the entire model, aligning with recent medical paradigms that address the nervous and immune systems simultaneously.
The team validated their model through two distinct approaches:
Ablation simulations: Removing any single axis eliminated the multi-stable states and hysteresis. Stable disease states emerged only when all three axes interacted concurrently.
Real-world patient data: Analyzing six public genomic datasets across five autoimmune conditions—rheumatoid arthritis, lupus, ankylosing spondylitis, dermatomyositis, and systemic sclerosis—revealed that genes predicted by the model exhibited distinct expression patterns in the blood and synovial tissue of rheumatoid arthritis patients.
Treatment Nonresponse Is a Biological State, Not a Lack of Willpower
Professor Jung Sung-soo noted, "This study represents the first attempt to understand the challenges faced by patients with difficult-to-treat rheumatic diseases from a whole-system perspective, moving beyond individual genes or cells. It explains why single-target therapies hit limitations in refractory cases and highlights the necessity of multi-axis therapeutic strategies."
While validated through computer modeling and public genomic datasets, clinical confirmation through long-term patient tracking remains the next phase. Professor Jung plans to validate the framework using multi-omics data—integrating genomic, transcriptomic, proteomic, and metabolomic layers.
"Validating this model with longitudinal multi-omics data from difficult-to-treat patients will allow us to pinpoint which specific biological axis drives resistance in each individual, paving the way for multi-targeted precision medicine," Jung added.
What Patients Can Do Today
While personalized multi-target treatments are still developing, patients can take practical steps now. If symptoms persist after trying two or more biologics or targeted therapies with different mechanisms of action, patients should consult their rheumatologist to evaluate whether their condition falls under difficult-to-treat (D2T) criteria.
Furthermore, disconnects between test results and physical symptoms—such as normalized blood markers despite ongoing pain and fatigue, or reduced pain alongside elevated inflammation—should not be overlooked. These mismatches suggest that rather than simply adding another drug, clinicians should evaluate secondary factors like sleep quality, stress levels, and gut health.
Addressing these mismatches during medical appointments can serve as an important first step toward tailored treatment.
Ultimately, failing to respond to medication is not a reflection of a patient's willpower or effort. The discovery that the body settles into a resistant state through multi-system interactions offers patients who have struggled with treatment frustration both a scientific explanation and new guidance for future therapeutic strategies.
