The most startling claim in this piece isn't that drug development is broken—it's that the solution has been sitting in our DNA for millennia, waiting for us to stop guessing and start reading. Works in Progress argues that we are finally shifting from a model of expensive, high-stakes human trial-and-error to one where nature has already run the safety tests for us. This isn't just a scientific update; it's a fundamental reorientation of how we approach human health, suggesting that the safest path forward lies in the genetic outliers we've long ignored.
The Cost of Guessing
The article opens with a grim reminder of why the current system is unsustainable. It recounts the 2006 tragedy at Northwick Park Hospital, where a monoclonal antibody called theralizumab turned healthy volunteers into patients with "elephant man" appearances within an hour of infusion. The piece notes that "the drug had activated T cells uncontrollably, flooding the volunteers' bodies with inflammatory proteins called cytokines." This reference to cytokine release syndrome—a topic the publication has explored in depth—highlights a critical flaw: animal models simply cannot predict human immune responses with perfect fidelity.
Works in Progress reports, "Safety and effectiveness in animals often fail to translate to humans." The argument here is that we are burning billions of dollars and risking lives to discover what nature already knows. The piece details how the theralizumab failure led to bankruptcy for the developer and tighter regulations, yet "serious adverse events continue to occur." This persistence of failure despite stricter rules is the piece's strongest motivator for change. It suggests that regulation alone cannot fix a broken methodology.
Critics might argue that relying on natural mutations is too slow or too rare to drive the pace of modern pharmaceutical innovation. However, the article counters this by pointing out that with nearly eight billion people, "any mutation that is compatible with life and reproduction likely exists in at least one person's genome somewhere in the world."
Nature's Laboratory
The core of the argument rests on the concept of "natural experiments." The editors explain that random genetic mutations act as a comprehensive catalog of changes, effectively mimicking the action of a drug. A powerful example provided is the DGAT1 inhibitor trials. While mice suggested the drug would cause weight loss, human trials resulted in severe diarrhea and vomiting. The breakthrough came not from a lab, but from a tragic clinical case: two siblings born with a rare disorder caused by a mutation that inactivated the DGAT1 gene. "The researchers concluded that this lack of functioning DGAT1 had given their intestinal cells an extreme intolerance to food." This real-world data, the piece argues, would have saved years of failed trials if it had been accessible earlier.
The historical context here is vital. Just as the discovery of the CCR5 mutation in the 1990s revealed a natural immunity to HIV and led to the drug Maraviroc, these genetic "knockouts" provide a blueprint for safety. The piece notes that a 2024 analysis found that "the odds of a drug getting regulatory approval doubled if there was supporting genetic evidence," a figure that has since climbed to 2.6 times higher. This statistical leap is the smoking gun for the new approach.
"Nature has already done lots of the trials we want to run, through random genetic mutations and natural selection."
The Power of Scale and Diversity
The commentary shifts to the infrastructure required to unlock this data. The piece argues that early genotyping, which only looked at common variations, was insufficient for finding rare, high-impact mutations. The solution was large-scale sequencing of the exome—the protein-coding instructions of the genome. Works in Progress highlights the role of the UK Biobank and DeCODE genetics in Iceland, noting that these projects have already validated drug targets for Alzheimer's and liver disease.
However, the article makes a crucial pivot toward diversity. It points out that "nearly 90 percent of the UK Biobank participants are of European ancestry," which blinds researchers to variants common in other populations. The piece illustrates this with the discovery of the PCSK9 mutation, which lowers cholesterol. "The initial discovery... was possible because the genetic variants were common in Africans but rare in Europeans." Without a diverse sample, this life-saving insight would have remained hidden.
The editors argue that we must look to populations with high rates of consanguinity, such as South Asians, to find "human knockouts"—individuals with no functional copies of a specific gene. The case of a British Pakistani woman who lacked the HAO1 enzyme but lived a healthy life proved the safety of a new drug for kidney failure. "Her existence showed that a healthy life is possible without glycolate oxidase, and it is safe to switch off HAO1 permanently in humans." This is the ultimate validation of the method: a human volunteer who has effectively taken the drug for a lifetime without side effects.
Bottom Line
The strongest part of this argument is its shift from theoretical possibility to proven track record, using the HAO1 and CCR5 cases to demonstrate that genetic data is not just predictive but prescriptive. Its biggest vulnerability lies in the logistical and ethical complexities of building truly global genetic databases, a hurdle the piece acknowledges but may understate. The reader should watch for how the industry balances the urgent need for diverse data with the historical mistrust of medical research in marginalized communities.
"The odds of a drug getting regulatory approval doubled if there was supporting genetic evidence."
This piece successfully reframes the drug development crisis not as a failure of chemistry, but as a failure of data utilization. It forces us to realize that the most rigorous clinical trial we can ever run has already happened, and the results are written in the DNA of the people we have yet to listen to.