By Matthew Herper
November 8, 2026
The next time a venture capitalist or a tech visionary asserts that artificial intelligence, machine learning, or a breakthrough in genomic data will inevitably "solve" the crisis of chronic disease, point them toward the current state of cardiovascular medicine. Heart attacks and strokes remain the leading cause of death in the developed world, a stubborn, biological monolith that has defied decades of multi-billion-dollar R&D efforts.
This past weekend, at the annual meeting of the American Heart Association (AHA), the industry received a sobering reality check. Despite the promise of "precision medicine" and the deployment of advanced genetic insights, two high-profile, late-stage clinical trials ended in failure. These results serve as a stark reminder: in the complex ecosystem of the human body, a clear genetic target does not always translate into a clinical cure.
Main Facts: The Failure of Two "Sure Bets"
The cardiovascular community arrived at the AHA conference with high expectations for two distinct therapeutic approaches, both underpinned by what appeared to be ironclad biological rationale.
First, Novartis presented findings from its trial of pelacarsen, a drug designed to lower lipoprotein(a)—or lp(a)—a cholesterol-like particle that is heavily influenced by genetics. For years, epidemiological data had suggested that elevated levels of lp(a) were a primary, independent risk factor for cardiovascular disease. The logic was elegant: lower the lp(a), reduce the risk of heart attacks and strokes. Yet, despite successfully lowering the substance in patients’ blood, the drug failed to produce a statistically significant reduction in cardiovascular events.
Second, Novo Nordisk delivered equally disappointing results regarding their anti-inflammatory candidate, ziltivekimab. This drug targets the IL-6 protein, a key driver of inflammation. Given the medical community’s growing consensus that chronic inflammation is a silent engine behind atherosclerosis and plaque rupture, researchers were optimistic that blocking this pathway would yield significant protective benefits. Once again, the trial data failed to move the needle on patient outcomes.
These failures are not merely minor setbacks; they represent a fundamental disconnect between our ability to manipulate biomarkers and our ability to improve clinical reality.
A Chronology of Expectation vs. Reality
To understand the weight of these failures, one must look at the timeline of the "Genetic Revolution" in cardiology.
- 2015–2018: The rise of Large-Scale Genomic Studies. Researchers identified specific gene variants that predisposed individuals to high lp(a) levels and chronic, low-grade inflammation. The pharmaceutical industry pivoted, pouring billions into "validated" targets.
- 2020–2023: The "AI Boom." As Big Tech entered the drug discovery space, firms began utilizing machine learning to predict how molecules would interact with these specific protein targets. Both pelacarsen and ziltivekimab were hailed as the "low-hanging fruit" of this new era.
- 2024: Enrollment in Phase 3 trials reached its peak. Investors and analysts pegged these drugs as potential blockbusters, assuming that the biological data was so robust that failure was statistically unlikely.
- November 2026: The AHA conference. Both trials reached their primary endpoints (or lack thereof), signaling that the biological hypothesis—while accurate in theory—was clinically insufficient.
Supporting Data: Why the Logic Failed
The failure of these trials reveals a profound truth about human physiology: the body is not a simple machine with a single "off" switch.
The Lp(a) Conundrum
Novartis’s pelacarsen was designed to inhibit the production of apolipoprotein(a), the main component of lp(a). The data confirmed the drug did exactly what it was designed to do: it significantly reduced plasma lp(a) levels. However, the trial showed that this reduction did not equate to a decrease in major adverse cardiovascular events (MACE). Scientists are now scrambling to determine if the drug was started too late in the disease process, or if lp(a) is a biomarker of risk rather than a direct driver of the disease itself.
The Inflammation Barrier
Novo Nordisk’s Ziltivekimab targeted the IL-6 pathway, a centerpiece of the "inflammation hypothesis" of heart disease. While inflammation is undeniably linked to plaque instability, the immune system is highly redundant. By blocking one pathway (IL-6), the body may have compensated through other inflammatory markers, leaving the underlying risk of heart attack largely untouched. This suggests that the "precision medicine" approach of targeting a single protein may be fundamentally incompatible with the systemic nature of cardiovascular inflammation.
Official Responses and Industry Sentiment
The atmosphere at the AHA conference was one of measured caution. Industry leaders and academic researchers are now engaged in a period of intense soul-searching.
"We have been operating under the assumption that if we can fix the upstream biomarker, the clinical outcome will inevitably follow," said one lead investigator during a panel discussion. "These results force us to admit that we don’t fully understand the ‘downstream’ consequences of these interventions. Biology is not a linear equation."
Novartis and Novo Nordisk have both issued statements pledging to conduct "deep dives" into the data. Novartis noted that while the overall primary endpoint was not met, they are investigating whether specific subsets of patients—perhaps those with extremely high baseline lp(a) levels—might still derive some benefit. Novo Nordisk, meanwhile, emphasized that the safety profile of ziltivekimab remained solid, even if the efficacy was absent.
The Implications: A Shift in R&D Strategy
The failure of these two high-profile candidates carries significant implications for the future of pharmaceutical investment and drug discovery.
1. The End of "Target-First" Naivety
For years, the industry has favored a "target-first" approach: find a gene or protein, validate it in a lab, and build a drug around it. These latest failures suggest that this method is increasingly reaching a point of diminishing returns. The focus is shifting toward "phenotype-first" approaches—looking at the patient’s clinical manifestation of disease rather than just their molecular profile.
2. The AI Reality Check
Artificial intelligence has shown immense promise in folding proteins and predicting molecular binding, but these trials prove that AI cannot "solve" a disease if the underlying biological hypothesis is flawed. If we feed algorithms incomplete data about human physiology, the algorithms will only return more precise errors. Innovation in AI must be paired with a more holistic understanding of human biology, not just high-throughput data processing.
3. The Re-evaluation of Chronic Disease
Cardiovascular disease is not a single entity; it is a multifactorial process involving genetics, lifestyle, environment, and aging. The "silver bullet" mentality that has dominated cardiovascular R&D for two decades is clearly failing. Moving forward, the industry may need to pivot toward combination therapies—targeting multiple pathways simultaneously—rather than relying on single-agent blockbusters.
4. Economic Consequences for Pharma
The loss of these programs represents billions of dollars in sunk costs and a significant hit to the pipelines of two of the world’s largest pharmaceutical companies. This will likely lead to a cooling-off period for investment in cardiovascular R&D, as companies pivot toward more "predictable" areas like oncology or rare genetic diseases, where the link between target and outcome is often more direct.
Conclusion: The Long Road Ahead
As the medical community departs the 2026 AHA conference, the mood is one of humble reflection. The dream of "curing" heart disease through genomic precision has proven to be far more elusive than we imagined.
We are entering a new era of biological complexity. The lesson of this year’s conference is not that we should abandon our search for new medicines, but that we must change our definition of what "innovation" looks like. It is no longer enough to build a better molecule; we must build a better understanding of the system in which that molecule operates. The technology of the future will not be defined by the speed of our computers, but by the depth of our biological humility.
The battle against the world’s largest killer continues, but it will be won through the slow, methodical process of understanding human biology—not by chasing the mirage of the next technological quick fix.
