Ultra-Short Radiotherapy Schedule for Prostate Cancer Supported by HERMES Trial

U.K. research has shown that condensing prostate cancer radiotherapy into two sessions, but with a higher dose per session, rather than the traditional five sessions does not lead to increased side effects when treatment is delivered using state-of-the-art magnetic resonance imaging (MRI)-guided technology.

Lead researcher Sian Cooper, clinical research fellow at The Royal Marsden NHS Foundation Trust and the Institute of Cancer Research in London, told Inside Precision Medicine that the results of the HERMES trial “point toward a future where prostate radiotherapy is tailored to the disease, to the individual’s anatomy and symptom profile, whilst being mindful of the impact of treatment time on patients’ lives.”

Presenting at the 2026 Congress of the European Society for Radiotherapy and Oncology (ESTRO), Cooper said that the number of people diagnosed with prostate cancer is projected to double by 2040, meaning the demand for effective and efficient treatment has never been greater.

At present, localized prostate cancer is typically treated with stereotactic body radiotherapy over five sessions, but there has been increasing interest delivering the treatment in fewer sessions, with a larger dose each session.

“For patients, a two-session treatment course would be far less disruptive than the weeks of daily hospital visits that radiotherapy has traditionally required. This convenience comes with clear benefits for work, leisure, family life and travel. For clinicians and health systems, fewer fractions mean faster workflow throughput, and getting patients the right treatment, quicker,” said Cooper.

“We wanted to find out whether giving the equivalent dose in just two treatment sessions could be feasible and safe for patients, and to understand how it might affect the potential side effects patients can experience, such as problems with urinary and bowel function.”

The move toward giving higher radiotherapy doses in fewer sessions has been made possible by improvements in radiotherapy delivery technology over recent years.

“It has allowed us to harness the power of modern computing and discoveries in clinical physics, to create radiotherapy doses which conform very tightly to the edge of the prostate,” Cooper explained. “This results in vastly less dose to the normal, healthy tissues around the cancer.”

The HERMES trial used a Unity MR-Linac (Elekta AB, Sweden) machine, which Cooper describes as “the ultimate evolution of this progress,” to deliver radiotherapy to participants.

The device combines real-time MR scanning with a linear accelerator and is known as MRI-guided adaptive radiotherapy. It allows adaptation of the radiotherapy beam design to changes in patient anatomy on the treatment day as well as moment by moment motion management, meaning that if there are any changes in the target or healthy bystander organs, the radiation beam can be switched off.

“This level of precision was needed to safely deliver the high dose of radiation necessary to maintain the biologically equivalent dose in just two fractions,” Cooper noted.

In all, 46 participants (median age 74 years) with intermediate or lower high-risk prostate cancer were randomly assigned to receive radiotherapy at a dose of 24 Gy in two fractions over 8 days, with a 27 Gy integrated boost to the MRI defined tumor (n=22), or 36.25 Gy in five fractions over two weeks to the planning target volume, with 40 Gy to the prostate and proximal one cm seminal vesicles (n=24). All participants had androgen deprivation treatment for at least six months.

Cooper reported that, at two years, four (18%) participants in the two-fraction arm and six (25%) in the five-fraction arm experienced moderate (grade 2) urinary adverse events (AEs) such as increased frequency or urgency. Grade 2 gastrointestinal AEs occurred in one participant in each arm (5% and 4%, respectively).

There were no grade 3 or worse genitourinary or gastrointestinal events in either group.

The team also showed that quality of life, measured by the International Prostate Symptom Score and the International Index of Erectile Function, showed minimal change up to two years but will continue to be monitored up to five years.

Cooper said that the investigators will present efficacy data, a secondary endpoint, when it matures but she pointed out that as “the cancer control rate for prostate cancer is so high, it often takes many years for failures to appear, whereas trial data for ultra hypofractionation show that the genitourinary adverse event rate is the primary concern after treatment.”

ESTRO president, Matthias Guckenberger, MD, from University Hospital Zurich, Switzerland, who was not involved in the research said: “While the technology used in this trial is currently available in only a limited number of specialist centers worldwide, they are growing rapidly. These results can help guide how they are used and help us understand whether two-session radiotherapy should become a new standard of care.”

The post Ultra-Short Radiotherapy Schedule for Prostate Cancer Supported by HERMES Trial appeared first on Inside Precision Medicine.

Mutating Antibodies for Easier Drug-Conjugate Manufacturing

Scientists in the United States have developed a general-purpose antibody that they hope will help revolutionize antibody-drug conjugate (ADC) manufacturing. The team, from Johns Hopkins University, says they mutated the fragment crystallizable (FC) region, the part of an antibody that modulates immune response. The aim was to create new sites to attach molecules, including nanoparticle drugs or fluorescent markers for quality assurance.

According to Jamie Spangler, PhD, associate professor of biomedical engineering and chemical & biomolecular engineering, the new antibodies could—in the future—lead to more effective and easier-to-manufacture drug conjugates.

“The chemistry of antibody drug conjugates is so heterogeneous. It can be hard to characterize the drug-to-antibody ratio and to [do things like] maintain consistency in formulations.”

To get around this problem, Spangler’s team installed six mutations on the FC region of an antibody that can act as attachment sites for a variety of molecules. The team was able to attach a dye to quantify how many sites were available and discovered the best productivity was found when using up to four sites.

They emphasize that the sites can be used for many purposes.

“You can attach whatever you want,” Spangler explains. “You could use [them] to make an antibody-dye conjugate or even a drug conjugate.”

According to Spangler, the team has already shown that the mutations can be used to conjugate with nanoparticles. “We encapsulate the protein we want to deliver within the nanoparticle, and then we coat the surface with an antibody. The nanoparticle we’re carrying, in this case, contains some GFP [green fluorescent protein], which is a fluorescent readout, but we can attach that to an antibody.”

After the antibody binds to a cell expressing the target, it’s internalized, and the nanoparticle can release its cargo, she explains. This system can be used for any number of purposes.

“The sky’s the limit for how people want to use this in their own research and their own work,” she says. “It’s a fully tuneable and generalisable system, and we’d encourage people to think broadly and creatively about the different attachments they can use.”

The post Mutating Antibodies for Easier Drug-Conjugate Manufacturing appeared first on GEN – Genetic Engineering and Biotechnology News.

The many manifestations of magical thinking: a systematic review

Magical thinking (MT) involves beliefs that thoughts or actions can influence events in unrealistic ways. While MT is integral to obsessive compulsive disorder, and reflected in the cognitive features of schizophrenia, it is observable across the general population in various forms. Given its prevalence and potential relevance to a range of psychiatric conditions, understanding more about what may predispose an individual to MT, and how it may in some cases culminate in psychological distress or dysfunction would be helpful. This paper reports a systematic review of studies investigating MT, encompassing both magical ideation and thought-action fusion specifically, across the disciplines of psychiatry and psychology, to shed further light on the likely predisposing factors and behavioural consequences of MT, its potential neurobiological underpinnings, and role in psychiatric symptomatology. After exclusions, 191 studies were identified that explored MT in association with a diverse array of secondary topics, from gambling compulsions to childhood trauma, within both clinical and non-clinical samples, across a range of cultural contexts. On an intra-individual level, MT demonstrates numerous cognitive and emotional correlates, and on a societal level it may influence both social custom and religious tradition. A synthesis of the available evidence uncovers unexplored relationships with social cognition and mental health, and future research investigating its emerging relationships with stress, mood and social connection, may uncover functions beyond those exhibited by a simple marker of psychopathology.

Undergraduate Nursing Students’ Experiences of Individualized Digital Reminiscence Using the InspireD App in Care Home Placements: Qualitative Focus Group Study

<strong>Background:</strong> Care home placements offer important opportunities for student nurses to develop relational and person-centered approaches to dementia care. Digital reminiscence platforms are increasingly used to support the well-being of people living with dementia; however, little is known about how such platforms may shape student learning within practice settings. There is limited qualitative evidence examining how digital reminiscence is experienced by students and how it influences their understanding of personhood, relationships, and care practices. <strong>Objective:</strong> The aim of the study is to explore undergraduate nursing students’ experiences of engaging with individualized digital reminiscence using the InspireD reminiscence app during care home placements. <strong>Methods:</strong> Following a pilot implementation of the intervention, a qualitative exploratory study was conducted, in which 13 undergraduate nursing students participated in 4 focus groups. Data were analyzed using reflexive thematic analysis. <strong>Results:</strong> Three themes were developed to capture how participants made sense of their learning and practice experiences when engaging with individual reminiscence using the InspireD reminiscence app. The first theme, “deepening empathy and understanding through reminiscence,” describes how participants developed a greater appreciation of residents’ life histories and personhood. The second theme, “learning through connection,” reflects how relationships with residents and families shaped communication, confidence, and emotional engagement. The third theme, “growing as person-centered practitioners within the realities of care home practice,” highlights how participants reflected on translating this learning into practice while navigating organizational constraints and everyday care demands. <strong>Conclusions:</strong> Findings suggest that the InspireD reminiscence app can support the development of person-centered learning within care home placements, although successful implementation is contingent on supportive organizational cultures. These findings contribute to wider discussions in health professions education by illustrating how digital platforms can mediate experiential learning in practice settings and support the preparation of future health professionals to use digital tools in relational and values-based ways. Future research should examine longer-term learning outcomes and implementation across diverse placement contexts.

Google DeepMind and Edison Are Building the AI Scientist

Google DeepMind and Edison Scientific are on an ambitious mission to build the AI scientistThese platforms propose to automate the scientific method using reasoning systems that connect hypothesis generation, experimental design, and data interpretation in one platform. In drug discovery, where traditional development timelines can stretch beyond a decade, such systems promise to dramatically accelerate the pace of biomedical research.

The AlphaFold developer and the nonprofit home organization behind Edison, FutureHouse, originally introduced their respective systems, Co-Scientist and Robin, as bioRxiv preprints in early 2025. Those studies have now been published in Nature, marking another step toward a growing ecosystem of specialized AI agents for life science research. 

Led by Demis Hassabis, PhD, CEO, and 2024 Nobel laureate in Chemistry, DeepMind is no stranger to expanding biomedicine. The team published a January Nature paper describing AlphaGenome, a unifying DNA sequence model for regulatory variant-effect prediction to support understanding of genome function and disease biology. 

Additionally, DeepMind drug discovery spinout, Isomorphic Labs, recently made waves after securing a whopping $2.1 billion Series B led by Thrive Capital, signaling the industry’s growing investment in AI-driven therapeutics. 

I’ve always believed the No.1 application of AI should be to improve human health,” wrote Hassabis on LinkedIn when announcing Isomorphic’s blockbuster raise. 

DeepMind’s newly published AI assistant, Co-Scientist, is a general-purpose multi-agent system built with Google’s Gemini and driven by natural language prompts. The platform demonstrated initial validation across three biomedical applications: drug repurposing for acute myeloid leukemia, novel target discovery for liver fibrosis, and explaining mechanisms of anti-microbial resistance. 

Co-Scientist’s design scales test-time compute to iteratively reason, evolve, and improve the output as it gathers more knowledge. Researchers can also actively steer the system by refining generated ideas or providing feedback through the natural language chat.

Vivek Natarajan, research scientist at DeepMind, emphasizes that time is a valuable commodity when tackling disease. Co-Scientist aims to support humans scientists in reaching answers to their problems much faster than before, from “months and years to minutes and hours.”

“To realize this vision, we need to build in reliability, trustworthiness and ensure a collaborative human-AI interaction paradigm. We have done a lot of research on these aspects and we are continuing to improve,” Natarajan told GEN Edge.

Closing the loop 

Edison is the commercial spinout of FutureHouse, an AI scientist non-profit backed by former Google CEO Eric Schmidt and co-founded by Sam Rodriques, PhD, former group leader at The Francis Crick Institute and Edison’s CEO. The team’s newly published platform, Robin, leverages both OpenAI o4-mini and Anthropic Claude 3.7 to aid biological discovery.  

In research tasks, Robin proposed repurposing Ripasudil, an existing drug for treatment of glaucoma, to address dry age-related macular degeneration (dAMD) via a novel mechanism that enhanced retinal pigment epithelial cell phagocytosis. The platform also suggested a circadian clock modulator, KL001, as an unexpected treatment for dAMD, illustrating the ability to make new connections not found in existing literature. Both insights were experimentally validated in patient-derived retinal pigment epithelium (RPE) cells. 

Since Robin’s May 2025 preprint release, Edison unveiled an updated AI scientist, Kosmos, last November. Kosmos can reason over 175 million full-text papers, clinical trials and patents, and operate interactively as a colleague that can sends updates mid-run. The system is reported to perform hundreds of research tasks in parallel to compress months of work into a single day.  

Today, Edison announced a collaboration with Incyte to employ Kosmos across the global pharma’s discovery and development pipeline. The partnership will focus on enabling continuous learning from translational and clinical data, real-time synthesis of evidence, and predictive models of therapeutic performance.

Michaela Hinks, founding member of technical staff at Edison, says the main bottlenecks for AI scientist adoption are trust, validation, and the gap in end-to-end solutions.  

“Most AI tools accelerate the cheaper and easier upstream work, but not the expensive and regulated downstream stages of scientific research,” Hinks told GEN Edge. 

She also highlights Robin as the first demonstration of an agentic AI scientist generating a hypothesis that is tested and validated in patient-derived cells, not an immortalized cell line, supporting clinically actionable insights for patients in need. 

Whether AI scientists will truly revolutionize discovery remains to be seen, but researchers are already beginning the experiment.  

The post Google DeepMind and Edison Are Building the AI Scientist appeared first on GEN – Genetic Engineering and Biotechnology News.

Understanding the modern cybercrime landscape

Throughout 2025, HPE observed significant changes in how cybercriminals operate. Analyzing real-world threats, our HPE Threat Labs highlighted an industrialization of the cyber criminals’ methods in its new In the Wild Report, enabling greater scale, speed and structure in their campaigns. They typically use automation and AI to exploit longstanding vulnerabilities, and many have adopted a professional, corporate hierarchy to optimize their efficiency.

Cybersecurity threats today are as menacing as ever for enterprises, as any CISO or CIO can probably confirm. But, digging behind that straightforward statement, there is a much more nuanced, complex cybersecurity landscape at play. This can make it significantly harder to plan, execute, and sustain effective strategies and solutions to protect the network—plus the often valuable—sometimes priceless—data, apps, and assets it transports and stores.

But it can be done, with the right philosophy and strategy, and the right tools and insights.

We must first understand the contemporary cybersecurity landscape. This understanding can unlock the right strategy and then onward to identify the tools and insights necessary to protect an enterprise’s network effectively.

There are five primary factors influencing the landscape, some old, some new, all dynamic. These factors are distinct but often interdependent, both within themselves and with one or more of the others. Another meaningful way of looking at them is “internal” and “external”; as ever, understanding and dealing with what is in your control can also help to navigate and mitigate what is beyond your control.

Five key factors influencing today’s dynamic cybersecurity landscape

1. Expectations

The first factor is predicated on the fundamental reality of an enterprise’s reliance on its network. Most enterprises have already undergone some form of digital transformation and are reaping the day-to-day benefits. This means that the number of people, devices, and things using the network continues to grow; it also means that people’s expectations of the network are higher than ever before – they demand that it does exactly what they need it to do, typically across a proliferation of devices and from multiple locations. Conversely, many employees might not be fully aware of cyber threats and infiltration methods, so their skillsets can easily be the weak point that admits bad actors into the network.

Equally, senior management and board members have high expectations at a meta level. Embracing digital transformation and network reliance means the enterprise’s function and reputation are inextricably tied to that. Loss of reputation due to a security breach is a chilling prospect, as is the threat of financial penalty and revenue loss. So, in the minds of leadership, the network has to be safe from cyber threats and be compliant.

2. Financial pressures

The first factor arguably contradicts its neighbor in the landscape: general financial constraints and the pressure on CISOs and CIOs to achieve more with less. Despite the strategic reliance on the network and the expectation that it will be protected from cyber threats regardless, the appropriate latticework of defenses (e.g., skilled and right-sized IT teams using progressive tools and meaningful data insights, plus constant workforce education) is not always properly funded and sustained, particularly in the current tough economic climate.

3. Complex infrastructure operations

The ongoing pursuit of digital transformation and consequent network reliance also drives the third factor. Ironically, there is another facet of enterprise protection and financial control wrapped up in this. The widespread move from one-stop shops (avoiding IT vendor lock-in in favor of more competitive pricing and autonomy) has created a more complex, multivendor environment. This is coupled with multiple IT domains required to handle many diverse functions and layers of IT infrastructure (e.g., cloud, on-prem), all connected to the network. Complex, mission-critical IT operations now need to be monitored and protected from increasingly sophisticated cyber breaches.

4. Unpredictable geopolitics and economics

Shifting from the first three factors—all internal to an enterprise—the fourth is unquestionably external and without doubt the most intractable risk for any enterprise, individual, or industry group. Global uncertainty and tension are unavoidably putting even greater pressure on already-tight IT budgets, component supply chains and power costs. This can easily exacerbate existing constraints on cybersecurity budgets when vigilance and protection are more needed than ever. Unfortunately, in cyberspace one cannot always point a finger in one direction to identify an adversary. Geopolitical alliances in cyberspace are much more difficult to track, and defending against an escalating tension becomes an all-out fight to secure the network.

5. Evolving cyber threats

The fifth factor is obviously the epicenter of today’s cyber security landscape. According to the HPE Threat Labs’ report, governments were the most frequently targeted sector globally in 2025, followed by finance, technology, defense, and manufacturing. The prevailing global geopolitical and economic situation may further accelerate the twin motivations of nation state-linked espionage and organized crime for extortion and theft.

Use the network to protect the network… and beyond

The current cybersecurity landscape calls for a re-think of the network’s pivotal role and how it can manage an enterprise’s digital defenses effectively, dynamically, and comprehensively. Overall, the network can be an excellent security sensor and enforcement point, using built-in security capabilities rather than being a collection of devices with an inflexible, bolted-on security layer.

Much as cybercriminals use agentic and generative AI to intensify their campaigns, CISOs can stay ahead more easily by leveraging AI-driven network platforms for 24×7 automated management of security policy enforcement (e.g., zero trust), threat monitoring, and mitigation, encompassing devices, things, and users. Meaningful data insights can be harvested, analyzed, and recycled back into secure networking management tools for dynamic protection.

This approach helps the progressive enterprise to overcome increasingly sophisticated, multi-step, and prolific attacks, while better managing IT costs and simplifying oversight of IT operations. It can also significantly improve the user experience, going a long way to meet and even exceed those rising expectations consistently. 

As a strategy in today’s uncertain world, embracing this self-driving network paradigm enables flexibility, visibility, and consistency in an enterprise’s frontline digital defenses.

For more, read the “In the Wild” report.

This content was produced by HPE. It was not written by MIT Technology Review’s editorial staff.

Here’s why Elon Musk lost his suit against OpenAI

On Monday, the jury in Musk v. Altman dealt Elon Musk a major blow—reaching a unanimous advisory verdict that he had sued OpenAI too late and, as a result, his claims are barred by the applicable statutes of limitations. US District Judge Yvonne Gonzalez Rogers immediately accepted it. 

Musk announced on X that he will be appealing the decision. “The judge & jury never actually ruled on the merits of the case, just on a calendar technicality,” he wrote.

OpenAI was cofounded by Musk and a group of researchers in 2015 as a nonprofit with a mission to develop AI for the benefit of humanity, unconstrained by a need to generate financial returns. Musk donated $38 million to the company during its early days, allegedly on the basis that OpenAI CEO Sam Altman and president Greg Brockman had promised to keep the company a nonprofit committed to the mission.   

Musk brought two claims against OpenAI. First, he argued that Altman and Brockman breached the charitable trust he created through his donations by breaking their promise to keep the company a nonprofit and creating a for-profit subsidiary that ballooned over the years. Second, he argued that Altman and Brockman unjustly enriched themselves at Musk’s expense. He sued OpenAI in 2024. 

Musk asked the court to unwind a 2025 restructuring that converted OpenAI’s for-profit subsidiary into a public benefit corporation and to remove Altman and Brockman from their roles.

OpenAI argued that the time for Musk to sue the company had run out before he brought the case. The statute of limitations on the breach of charitable trust claim is three years, while the statute of limitations on the unjust enrichment claim is two years. This means that Musk should have discovered, or had reason to discover, Altman and Brockman’s alleged breach of charitable trust no earlier than 2021 and their alleged unjust enrichment no earlier than 2022. 

While Musk argued he discovered that Altman and Brockman had broken their promise only in 2022, OpenAI claimed that Musk had reason to think this well before 2021. 

Musk told the jury that he has gone through “three phases” in his beliefs about OpenAI: In phase one, he was “enthusiastically supportive” of the company. In phase two, “I started to lose confidence that they were telling me the truth,” he said. In phase three, “I’m sure they’re looting the nonprofit.” 

Here’s a deeper dive into a timeline of the events as testified in the trial. You can read my dispatches from all three weeks of the trial here and here and here

2017: Musk proposes creating a for-profit subsidiary

In 2017, two years after OpenAI was founded, Musk and the other cofounders tried to create a for-profit subsidiary to raise enough capital to build artificial general intelligence—powerful AI that can compete with humans on most cognitive tasks. They fought a bitter power battle over who would get to control the entity. Musk also proposed merging OpenAI with his electric-car company, Tesla. 

During the trial, OpenAI’s lawyers pressed Musk on these discussions, suggesting that Musk knew in 2017 about Altman and Brockman’s plans to pivot the company—even participating in such plans—and had reason to sue then.

“I was not opposed to there being a small for-profit that provides funding to the nonprofit,” Musk told the jury, “as long as the tail didn’t wag the dog.” 

2019: OpenAI creates a for-profit subsidiary with capped profits

In 2019, OpenAI created a for-profit subsidiary, under which employees and investors would receive a capped return on their investment. At the same time, the company secured a $1 billion investment from Microsoft. OpenAI argued that Musk again had reason to sue the company then. 

But Musk testified that he didn’t think the move was violating the nonprofit’s mission. “If you’ve got a capped-profit situation, it hasn’t violated the nonprofit’s goal,” Musk told the jury earlier in the trial. “There was no basis for me to file a lawsuit at that time.”

2020: Microsoft snags an exclusive license 

In 2020, when Microsoft secured an exclusive license to OpenAI’s GPT-3 model, Musk posted on X: “This does seem like the opposite of open. OpenAI is essentially captured by Microsoft.” OpenAI once again argued that Musk had reason to sue then. 

But Musk testified that after reading the post, Altman reassured him that “OpenAI was staying on the mission as a nonprofit.” Musk said although he was skeptical, he still had no reason to sue the company at that point.

2022: Microsoft prepares to invest $10 billion in OpenAI

It was only in 2022, Musk testified, that he discovered OpenAI had abandoned its nonprofit mission. At that time, Microsoft was preparing to invest $10 billion in OpenAI—a deal that closed in 2023. 

“I was disturbed to see OpenAI with a $20B valuation,” Musk texted Altman after reading the news. “This is a bait and switch.”

Musk told the jury this was the moment that made him realize “the for-profit is the tail wagging the dog.” He thought Microsoft would give $10 billion only if it expected “a very big financial return.” He argued that this was the point he realized “OpenAI had become, for all intents and purposes, a for-profit company with a $20 billion valuation.” 

“The 2023 deal was different,” Steven Molo, one of Musk’s lawyers, hammered home during his closing argument.

The jury sides with OpenAI

It was up to the jury to decide whether the evidence supported Musk’s claim that he first realized in 2023 that OpenAI was no longer a nonprofit committed to its mission. In the verdict announced today, they found Musk did in fact have reason to think that he was being misled by Altman and Brockman before 2021. They did not address whether he was in fact misled. 

Courts often decide cases on procedural grounds like statutes of limitations when they can, because it can be the cleaner way to resolve a case than to grapple with its merits.

Musk has said he will appeal the decision to the Ninth Circuit Court of Appeals, a federal appellate court that reviews decisions from district courts in California and other states.

Reporter’s Notebook: The Day the Scientific Debate Died

When the news broke on May 5 that U.S. Food and Drug Administration (FDA) officials had blocked the publication of two major COVID-19 vaccine safety studies in 2025 after being accepted for publication in medical journals, many researchers saw more than a scientific dispute. They saw it as further evidence that America’s most powerful public health agencies were devolving into ideological warfare, institutional instability, and political distrust.

By the time FDA commissioner Martin Makary, MD, resigned a week later on May 12, 2026, there was a growing conflict among scientists, political appointees, public health officials, and outside activists not only over vaccines and public health policy but also over something even more fundamental: who gets to determine what constitutes legitimate science and whether scientific disagreement itself would still be permitted to occur in public.  

A decade in the making 

Makary’s departure came amid a broader transformation of the FDA and HHS from relatively stable, technocratic agencies into politicized institutions shaped by pandemic-era conflicts. Under Obama-era leaders Robert Califf, MD, and Margaret Hamburg, MD, the FDA emphasized regulatory continuity and evidence-based policymaking, while the Department of Health and Human Services (HHS) focused largely on healthcare administration and implementing the Affordable Care Act. 

That changed during the first Trump administration and accelerated during COVID-19, when disputes over vaccines, masking, emergency authorizations, and therapeutics turned the FDA into a political flashpoint. Leadership turnover has increased, tensions between political appointees and career scientists have deepened, and public trust has fractured along ideological lines.  

The Biden administration attempted to restore institutional stability by returning Califf to the FDA and appointing lawyer and politician Xavier Becerra, JD, to HHS, but the agencies remained mired in conflicts over pandemic policy and public health authority. Under HHS Secretary Robert F. Kennedy Jr. in the second Trump administration, those tensions intensified further through staffing and funding cuts, ideological battles over vaccines and food policy, and growing distrust inside federal health agencies.  

Makary initially aligned with parts of the administration’s “Make America Healthy Again” (MAHA) agenda, particularly on food reform and criticism of segments of the pharmaceutical industry. But reports suggest he became caught between competing pressures from the White House, HHS leadership, industry groups, conservative activists, and public-health officials. He ultimately resigned amid disputes over vaping regulation, drug approvals, and broader public health policy during a sweeping restructuring of federal health agencies. Reports also indicated he was already at risk of removal and that his departure was not directly tied to controversy over the blocked COVID publication.  

Instead, the move signals the White House’s continued support for Robert F. Kennedy Jr., “MAHA,” and a shift toward more centralized control over food-safety strategy, inspections, and outbreak response—changes that could affect how aggressively the FDA enforces nutrition standards and responds to contamination events. More broadly, the episode has done little to ease concerns among scientists and public-health experts that political considerations are increasingly shaping regulatory decisions and narrowing the space for independent scientific debate within federal health agencies. 

Seeing double(think) 

One of the two was posted online as a medRxiv preprint in 2025 by lead author Joann F. Gruber, PhD, and senior author Steven A. Anderson, PhD, and examined updated Covid-19 vaccines in adults over 65, the population most vulnerable to severe disease and death from the virus. The analysis drew on data from 7.6 million Medicare FFS beneficiaries who received a COVID-19 vaccination in 2023–2024—either the Pfizer-BioNTech (3.68 million) or Moderna (3.84 million) mRNA vaccine or the Novavax protein-based vaccine (30,000)—and found no new vaccine safety signals. 

But before the study could move through peer review, publication was halted by the FDA, according to a spokesperson for the HHS. As reported by the New York Times, an HHS spokesperson said the studies were withdrawn “because the authors drew broad conclusions that were not supported by the underlying data. The FDA acted to protect the integrity of its scientific process and ensure that any work associated with the agency meets its high standards.”  

Both Gruber, whose work with the FDA’s Center for Biologics Evaluation and Research (CBER) began in 2017, and Anderson, a veteran of CBER having joined in 2001, left the FDA at some point in 2025. It’s worth noting that Anderson’s team posted a second preprint on influenza vaccines that mirrored the COVID-19 vaccines study, using the same patient population, on the same day, which was also posted online as a medRxiv preprint on January 5, 2025, and was accepted in the peer-reviewed journal Vaccine on March 25, 2025, and made available online April 8, 2025. 

On June 25, 2025, Makary and former CBER director Vinay Prasad, MD, PhD, in conjunction with manufacturers, added class safety warnings for myocarditis and pericarditis to COVID-19 mRNA vaccines’ prescribing information. The exact timing of when the accepted Gruber and Anderson study was pulled from publication has yet to be reported. That it occurred before June 2025 to prevent contradiction with Makary’s and Prasad’s safety update to the COVID-19 mRNA vaccine is entirely possible. Anderson’s team posted a second preprint on influenza vaccines that mirrored the COVID-19 vaccines study, using the same patient population, on the same day, which was also posted online as a medRxiv preprint on January 5, 2025, and was accepted in the peer-reviewed journal Vaccine on March 25, 2025, and made available online April 8, 2025. 

Both Gruber, whose work with the FDA’s Center for Biologics Evaluation and Research (CBER) began in 2017, and Anderson, a veteran of CBER having joined in 2001, no longer work at the FDA. According to their LinkedIn profiles, Gruber left in June 2025 and Anderson in December 2024. 

Gold-standard science 

I spoke with several leading epidemiologists to assess whether there was substance to the HHS statement. All immediately noted the lack of specificity in the agency’s criticism, particularly the vague references to “gold-standard science.” 

An epidemiologist with experience conducting vaccine safety studies, requesting anonymity, told Inside Precision Medicine, “If someone wants to criticize the study, there should either be a very clear articulation of what exactly they mean when they invoke terms like ‘gold-standard science’—specifically, which methods are acceptable, which are not, and why—or they should bring to the table the scientific credibility that would justify dismissing this kind of work outright. Frankly, neither of those things has happened. Broad, nonspecific attacks like these actually undermine the critique itself.”  

The epidemiologists I spoke to emphasized that the study’s methods were not novel but reflected established approaches used in vaccine surveillance and prior scientific work, including self-controlled case series and cohort studies. A Harvard University researcher familiar with the study said the framework was specifically designed for this purpose. “The system/infrastructure (FDA BEST), data source, study design, and analytic approach are all fit-for-purpose for the study question,” the Harvard researcher told me. “Self-controlled designs are robust and control for non-time-varying factors, especially in adults.” 

Céline Gounder, MD, an infectious disease specialist, epidemiologist, and editor-at-large for public health at KFF Health News, said the report used one of the strongest available methods for post-market vaccine surveillance. “This study used one of the best methods we have to check if vaccines cause side effects, and it found that the updated COVID vaccines are safe,” Gounder told me. “Pulling this study from publication doesn’t protect good science. It’s not radical transparency, and it’s not gold-standard science.” 

Gounder also noted that the study analyzed data from more than seven million people and found no new safety concerns. “That’s a careful conclusion backed by solid data,” she said. “Blocking a study because you don’t like the answer is censorship.” 

Indeed, the consensus among interviewees was that the study appeared adequately powered and appropriately cautious in its conclusions. “This study includes a large number of people, and from what I can see, it appears adequately powered for the conclusions they’re making,” said the epidemiologist with vaccine expertise. “Importantly, the authors are framing the findings appropriately. They are not claiming more than the data support. Their framing is essentially ‘No new safety signals identified.’ That’s a careful and reasonable way to present findings like this.” 

The study also openly acknowledged limitations, including possible outcome misclassification and residual uncertainty, while describing how these issues were addressed in the analysis. “Seasonality can be a concern with the study design, but it was adjusted for in the study,” said the Harvard researcher. “Claims data are well equipped for studying the exposure and outcomes of interest… In this study, misclassification was accounted for.”

Steven Goodman, MD, PhD, associate dean of clinical and translational research and professor of epidemiology and population health and medicine at Stanford University, told Inside Precision Medicine the study was informative and aligned with broader evidence supporting the low-risk profile of COVID vaccines in adults over 65. Goodman also highlighted the restraint of the authors’ interpretations. “They do not make a statement about the risk-benefit balance, which they can’t because they didn’t study the benefit, but they note that the FDA felt that the balance was positive,” he said. “Their main conclusion was, ‘Our study contributes to growing evidence on the safety of COVID-19 vaccines.’ It is hard to argue with that.” 

Goodman added, “All studies have strengths and limitations, i.e., none establish a scientific truth all by themselves. But this is fundamentally good science that adds valuable information to the COVID vaccine safety picture in adults >65.” 

The epidemiologists stressed that vaccine safety science depends on cumulative evidence across multiple studies, methods, and datasets. “Public health surveillance has always operated this way,” the Harvard University researcher said. “No single study claims to be the final word on a topic. You accumulate evidence across multiple studies, multiple methods, and multiple datasets and then interpret the totality of evidence together.” 

The Harvard researcher added, “The results are consistent with what has been reported by others, including in other countries. There is no clear scientific reason for this work being pulled from publication.” 

Truth welcomes questions 

Further, many of the epidemiologists I interviewed stressed that publication does not imply unquestioned acceptance. Instead, they argued that publication serves as the mechanism through which scientific claims challenge, refine, or overturn one another.   

Goodman emphasized that the unpublished manuscript was intended for scientific scrutiny and peer review and said imperfections in such work are neither unusual nor disqualifying. “Is it perfect? No, but this is a preprint, and usually the peer review and editing process improves the analyses, exploring robustness to various assumptions, the reporting, and the interpretation,” he said. “I would presume that the final version would have come out with some more qualifications, limitations, sensitivity analyses and caveats.” 

The experts I contacted emphasized how suppressing publication interrupts the ordinary process through which scientific consensus develops. “This study should be out there, clearly labeled as one piece of evidence among many, with all the necessary caveats attached,” said the epidemiologist. “Then additional studies come in, more data accumulate, and eventually the field interprets the evidence in totality.” 

The unnamed epidemiologist added, “For 250 years, this country has benefited from exactly that: reasonable people openly disagreeing about difficult issues. So why not say, ‘Fine, publish the study,’ and then publish an editorial alongside it explaining the caveats, limitations, and alternative interpretations? That’s how science is supposed to work. You respond to speech you disagree with by adding more speech, not by suppressing speech and certainly not by suppressing scientific speech.” 

Several of the epidemiologists argued that blocking the manuscript conflicts with repeated public calls for open scientific debate from directors at agencies under the purview of HHS, notably Jay Bhattacharya, PhD, Director of the National Institutes of Health (NIH). 

Goodman said that how the FDA handled this study “contrasted with Dr. Bhattacharya’s many public remarks stressing the criticality of open discussion of scientific results and his objections to suppressing science whose results one doesn’t like. The forced withdrawal of this manuscript prevented that process from occurring, shutting down the open discussion Dr. Bhattacharya has called for in innumerable forums.” 

Goodman added that if officials believe the study contains fatal flaws, they should articulate those concerns publicly and subject them to scientific scrutiny, “letting the authors respond and the scientific community decide… Their own critique should be subjected to peer review.” 

The fundamental process of science encourages that disputes over evidence should unfold transparently in scientific journals and public debates. “If someone has objections, they should make those objections publicly and specifically in the scientific literature where others can critique them, evaluate them, or even prove them right,” said the epidemiologist. “That’s how science advances. That’s what real science looks like: gold-, platinum-, titanium-, or whatever rare metal metaphor people want to use for standards. The scientific enterprise in this country has succeeded because ideas are tested openly, criticized openly, and refined openly. This kind of amateur hour behavior at regulatory agencies doesn’t help anybody.” 

Nostrums, not normalcy 

The culling of FDA scientists, be it via resignations or firings in 2025–2026, has continued since Makary’s resignation. Tracy Beth Hoeg, MD, PhD, the head of the FDA’s Center for Drug Evaluation and Research (CDER), was fired Friday (according to a social media post reported by Reuters on Sunday) and replaced by Michael Davis, MD, PhD, who had served as deputy director of CDER for about a year. 

There is no concrete evidence connecting the FDA’s blocking publication of two studies accepted into medical journals to Makary’s departure. But the contradictory messaging within the agency on COVID-19 mRNA vaccinesthe product of Operation Warp Speed, considered a signature accomplishment of the first Trump administrationis obvious. The collapse of confidence within institutions that once relied on scientific independence as their organizing principle. Increasingly, senior scientists and regulators appear unwilling to publicly defend decisions, studies, or processes they privately regarded as scientifically sound. That shift matters more than any single resignation. 

The appointment of Kyle Diamantas as acting head of the FDA, however, is far from a course correction, marking another sharp turn away from independent scientific leadership at America’s top health regulator. A former corporate lawyer for Abbott Laboratories with no medical or research background, Diamantas rose through the agency by advancing the “MAHA” food agenda and cultivating ties to politically aligned health influencers rather than the scientific establishment. A close friend of Donald Trump Jr., the appointment of the 38-year-old Diamantas only reinforces concerns that ideological loyalty is increasingly outweighing scientific expertise inside the FDA. 

For decades, FDA and HHS leadership operated with relative continuity, assuming that disputes would be resolved through open scientific debate. That assumption now appears badly weakened. The blocked vaccine safety study became symbolic not merely because of the substance of the research but also because even many scientists who believed the work was rigorous hesitated to say so publicly. When experts become reluctant to attach their names to conclusions that they consider obvious or well-supported, the problem extends beyond politics or personnel. It reflects a deeper institutional fear inside the scientific establishment itself. 

Makary’s resignation earlier this month therefore represented more than another leadership change in Washington. It exposed how federal health agencies have been pulled into a culture where scientific judgments are increasingly filtered through ideological loyalty, political risk, and reputational self-preservation. The larger danger is not only instability at the FDA or HHS, but the emergence of a scientific culture in which silence becomes safer than candor. Institutions built to evaluate evidence cannot function for long under those conditions.  

The post Reporter’s Notebook: The Day the Scientific Debate Died appeared first on Inside Precision Medicine.