Wistar scientists demonstrate first-ever single-shot HIV vaccine neutralization success

The Wistar Institute

FacebookXLinkedInWeChatBlueskyMessageWhatsAppEmail

 

image: 

The Wistar Institute's Dr. Amelia Escolano

view more 

Credit: The Wistar Institute

PHILADELPHIA — (TUESDAY, FEB. 3, 2026) — Scientists at The Wistar Institute have developed an HIV vaccine candidate that achieves something never before observed in the field: inducing neutralizing antibodies against HIV after a single immunization in nonhuman primates. The innovative approach, published in Nature Immunology, could significantly shorten and simplify HIV vaccination protocols, making them more accessible worldwide.

The research, led by Amelia Escolano, Ph.D., assistant professor in Wistar’s Vaccine and Immunotherapy Center and the senior author of the study, centers on an engineered HIV envelope protein, WIN332, that challenges scientific assumptions about how to design an effective HIV vaccine.

“By going against one commonly held belief in the field, we achieved low neutralization after a single immunization, which was further increased after one additional booster, something that has never been observed before,” said Escolano. “Usually, HIV vaccination protocols require seven, eight, or even ten injections to start seeing any neutralization. For our immunogen, WIN332, we injected once and already saw some neutralization.”

For years, scientists attempting to engineer HIV vaccines have focused on targeting the virus’s envelope protein, a component of the outermost layer of the virus. Dr. Escolano’s team has engineered a specific region of the envelope protein, called the V3-glycan epitope. Conventional wisdom held that antibodies targeting this region required a particular sugar, N332-glycan, to bind effectively. All previous envelope immunogens were designed to preserve this sugar. Escolano’s team took the unprecedented step of removing the N332-glycan completely to create WIN332.

A single injection of WIN332 induced low but detectable neutralization against HIV within just three weeks—an unprecedented timeline. When the researchers gave a second injection using a related immunogen, neutralization levels increased significantly. This represents a potentially marked improvement over current experimental protocols.

“This immunogen could shorten and simplify vaccination protocols,” said Ignacio Relano-Rodriguez, Ph.D., first author of the study. “If this approach proves successful, we could potentially achieve desired immunity with just three injections. This would make vaccination protocols shorter and more affordable.”
By removing the N332-glycan to create their immunogen, the team also revealed the existence of two distinct types of HIV-neutralizing antibodies that target the V3-glycan region. Type I antibodies represent the previously known class that requires the N332 sugar to bind effectively. Type II antibodies are a new class, identified by this research, that doesn’t require the sugar for binding.

“This discovery potentially expands the toolkit available for developing HIV vaccines that provide broader protection against the diverse HIV strains circulating globally,” Escolano said.

The promising results have attracted attention from major global health organizations to advance WIN332 into human clinical trials. Meanwhile, additional preclinical evaluations are underway, along with the design of subsequent immunogens that could be used in a shortened vaccination series to further enhance neutralization efficiency.

Co-authors: Ignacio Relano-Rodriguez, Jianqiu Du, Zi Jie Lin, Margaret Kerwin, Marta Tarquis-Medina, Eduardo Urbano, Jiayan Cui, Rumi Habib, Colby Agostino, Sukanya Ghosh, Joyce Park, Caroline Boroughs, Niharika Shukla, David B. Weiner, Daniel W. Kulp, and Jesper Pallesen from The Wistar Institute; Meagan Watkins and Ronald S. Veazey from Tulane National Primate Research Center; Peng Zhao and Lance Wells from University of Georgia; Michael  S. Seaman from Beth Israel Deaconess Medical Center; Agnes A. Walsh, Mariane B. Melo, and Darrell J. Irvine from Scripps Research Institute; and George M. Shaw and Beatrice H. Hahn from University of Pennsylvania. 

Work supported by: National Institute of Allergy and Infectious Diseases (NIAID) grants R00 AI140770-03, P30 AI045008-23, P30 AI045008-24, and R01 AI172627-01A1 to A.E.; Gates Foundation INV-036995 to A.E.; 5 U19 AI166916-03 to J.P. U19 AI166916, BEAT-HIV UM1AI64570, and the a W.W. Smith Charitable Trust Distinguished Professorship in Cancer Research to D.B.W. National Institutes of Health grants R01GM130915 and R01AI157854 to L.W.; National Science Foundation Biofoundry: Glycoscience Research, Education, and Training Grant 2400220 to L.W.; and The Ching Jer Chern Postdoctoral Fellowship to I.R.R.

Publication information: Rapid Elicitation of Neutralizing Asn332-glycan-independent Antibodies to the V3-glycan epitope of HIV-1 Env in Nonhuman Primates, Nature Immunology, 2026. Online publication.
 
###

The Wistar Institute is an international leader in biomedical research with special expertise in cancer research and vaccine development. Founded in 1892 as the first independent nonprofit biomedical research institute in the United States, Wistar has held the prestigious Cancer Center designation from the National Cancer Institute since 1972. The Institute works actively to ensure that research advances move from the laboratory to the clinic as quickly as possible. wistar.org.
 

---

Journal

Nature Immunology

DOI

10.1038/s41590-025-02408-z 

Method of Research

Experimental study

Subject of Research

Animals

Article Title

Rapid elicitation of neutralizing Asn332-glycan-independent antibodies to the V3-glycan epitope of HIV-1 Env in nonhuman primates'

Article Publication Date

3-Feb-2026

 

Medical AI models need more context to prepare for the clinic

Marinka Zitnik outlines the challenges — and potential solutions

Harvard Medical School

FacebookXLinkedInWeChatBlueskyMessageWhatsAppEmail

Medical artificial intelligence is a hugely appealing concept. In theory, models can analyze vast amounts of information, recognize subtle patterns in data, and are never too tired or busy to provide a response. However, although thousands of these models have been and continue to be developed in academia and industry, very few of them have successfully transitioned into real-world clinical settings.

Marinka Zitnik, associate professor of biomedical informatics in the Blavatnik Institute at Harvard Medical School, and colleagues are exploring why — and how to close the gap between how well medical AI models perform on standardized test cases and how many issues the same models run into when they’re deployed in places like hospitals and doctors’ offices.

In a paper published Feb. 3 in Nature Medicine, the researchers identify a major contributor to this gap: contextual errors.

They explain that medical AI models may produce responses that are useful and correct to an extent but are not necessarily accurate for the specific context in which the models are being used — which include things like medical specialty, geographic location, and socioeconomic factors.

“This is not a minor fluke,” said Zitnik, who is also associate faculty at the Kempner Institute for the Study of Natural and Artificial Intelligence at Harvard University. “It is a broad limitation of all the types of medical AI models that we are developing in the field.”

In a conversation with Harvard Medicine News, Zitnik explains how contextual errors happen in medical AI models, how researchers might overcome this and other challenges, and what else she sees on the horizon for AI in medicine. This interview has been edited for length and clarity.

Harvard Medicine News: Why do contextual errors happen? How can they be fixed?

Marinka Zitnik: We think that they happen because important information for making clinical decisions is not contained in the datasets that are used to train medical AI models. The models then generate recommendations that seem reasonable and sensible but are not actually relevant or actionable for patients.

For medical AI models to perform better, they need to adapt their recommendations in real time based on specific contextual information. We suggest incorporating such information into the datasets used to train models. Additionally, we call for enhanced computational benchmarks (standardized test cases) to test models after training. Finally, we think information about context should be incorporated into the architecture, or structural design, of the models. These three steps will help ensure that models can take different contexts into account and that errors are detected before models are implemented in actual patient-care settings.

HMNews: You give three examples of how a lack of context can lead to errors in medical AI models. Can you expand on them?

Zitnik: Let’s start with medical specialties. Patients may have complex symptoms that span multiple specialties. If a patient comes to the emergency department with neurological symptoms and breathing problems, they might be referred to a neurologist followed by a pulmonologist. Each specialist brings deep expertise shaped by their training and experience, so understandably focuses on their own organ system. An AI model trained mostly on one specialty might do the same, meaning it may provide answers based on data from the wrong specialty or miss that the combination of symptoms points to a multisystem disease.

Instead, we need to develop medical AI models trained in multiple specialties that can switch between contexts in real time to focus on whatever information is most relevant.

HMNews: What about the context of geography?

Zitnik: If a model is presented with the same question in different geographic locations and gives the same answer, that answer is likely to be incorrect because each place will have specific conditions and constraints. If a patient is susceptible to a disease that could lead to organ dysfunction or failure, the clinician would need to figure out the patient’s risk and develop a plan to manage it. However, whether that patient is in South Africa, the United States, or Sweden may make a big difference in terms of how common that disease is and what treatments and procedures are approved and available.

We envision a model that can incorporate geographic information to produce location-specific, and therefore more accurate, responses. We are working on this in our lab, and we think it could have major implications for global health.

HMNews: And the third example, the socioeconomic and cultural factors that affect a patient’s behavior?

Zitnik: Say a patient shows up in the emergency department with severe symptoms after they were previously referred to an oncologist and never made an appointment. A typical response from the ED physician might be to remind the patient to schedule the oncology appointment. However, this overlooks potential barriers such as the patient living far from the oncologist, not having reliable childcare, or not being able to miss work. These types of constraints do not explicitly exist in the patient’s electronic health record, which means they also would not be factored in by an AI model that is helping to manage the patient.

A better model would take these factors into account to offer a more realistic recommendation, perhaps by providing an option for transportation or scheduling an appointment at a time that accommodates childcare or work constraints. Such a model would increase access to care for a broader range of patients rather than reinforcing inequities.

HMNews: What other major challenges are there in medical AI implementation besides contextual errors?

Zitnik: There are many. One relates to how much patients, clinicians, regulatory agencies, and other stakeholders trust medical AI models. We need to identify mechanisms and strategies that both ensure that models are trustworthy and promote trust in these models. We think the answer has to do with building models that provide transparent and easily interpretable recommendations and that say “I don’t know” when they are not confident in their conclusions.

Another challenge relates to human-AI collaboration. Currently, many people think about human-AI interfaces in the context of chatbots, in which you type in a question and get a response. We need interfaces where people can receive responses tailored to their specific backgrounds and levels of expertise — for example, content suitable for a lay audience versus a medical expert. We also need interfaces where clinicians or patients and AI models exchange information in both directions. True collaboration means that there is a question or goal or task that an AI model has to complete, and to do that, it might need to seek more information from the user.

HMNews: What do you see as the promise of medical AI if the challenges can be overcome?

Zitnik: Some models have already had an impact by making every day medical work more efficient. For example, models are helping clinicians draft patient notes and helping researchers quickly find scientific papers that may be relevant to a clinical question.

I am especially excited about the opportunities AI models can create for improving treatment. Models that can switch contexts could adjust their outputs based on the information that is most useful during different parts of the treatment process. For example, a model might shift from analyzing symptoms to suggesting possible causes to providing evidence about treatments that worked in similar patients. A model might then pivot to providing practical information about a patient’s prior medications, potential drug side effects, and what treatments are actually available. If this is done well, it could help clinicians tailor treatment decisions for complex patients with multiple conditions and medications that may fall outside standard treatment guidelines.

HMNews: How do we ensure that medical AI models are doing more good than harm?

Zitnik: I certainly think that AI in health care is here to stay. This technology, while imperfect, is already being used, so everyone in the medical AI community needs to work together to ensure that it is being developed and implemented in a responsible way. This includes considering real-world applications as we design and refine models, doing real-world testing to understand where models succeed and where they fall short, and developing guidelines for how models should be deployed. I feel optimistic that if AI researchers are aligned in the development of these models and we ask the right questions, we can detect any issues early on.

Ultimately, I think there will be many opportunities for medical AI models to improve the efficiency of medical research and clinical work and improve care for patients.

Authorship, Funding, Disclosures

Additional authors on the paper include Michelle M. Li, Ben Y. Reis, Adam Rodman, Tianxi Cai, Noa Dagan, Ran D. Balicer, Joseph Loscalzo, and Isaac S. Kohane.

Support for the research was provided by the Ivan and Francesca Berkowitz Family Living Laboratory Collaboration at HMS and Clalit Research Institute; the National Institutes of Health (grant R01HD108794); the National Science Foundation (CAREER 2339524); the Department of Defense (FA8702-15-D-0001); ARPA-H (BDF program); the Chan Zuckerberg Initiative; the Bill & Melinda Gates Foundation INV-079038; Amazon Faculty Research; the Google Research Scholar Program; AstraZeneca Research; the Roche Alliance with Distinguished Scientists; Sanofi iDEA-TECH; Pfizer Research; the John and Virginia Kaneb Fellowship Award at HMS; a Biswas Family Foundation Computational Biology Grant; a Dean’s Innovation Award for the Use of Artificial Intelligence in Education, Research, and Administration at HMS; the Harvard Data Science Initiative; and the Kempner Institute.

---

Journal

Nature Medicine

DOI

10.1038/s41591-025-04184-7 

Article Title

Scaling medical AI across clinical contexts

Article Publication Date

3-Feb-2026

U$A

Why aren’t more older adults getting flu or COVID-19 shots?

Poll suggests need for more communication about vaccines’ power to reduce risk of severe illness in people over 50

Michigan Medicine - University of Michigan

FacebookXLinkedInWeChatBlueskyMessageWhatsAppEmail

 

image: 

National Poll on Healthy Aging data from adults age 50 and over who replied to a poll in late December 2025 and early January 2026 saying that they had not received a flu vaccination in the last 6 months and/or had not received a COVID-19 vaccination in the last year, showing the main reason why they had not done so.

view more 

Credit: Emily Smith - University of Michigan

This winter’s brutal flu season isn’t over, and COVID-19 cases have risen recently too. But a new poll taken in recent weeks shows that vaccination against both viruses lags among people 50 and over, and the national survey reveals key reasons why.

In all, the University of Michigan National Poll on Healthy Aging shows, 42% of people over 50 haven’t gotten either flu or COVID-19 vaccines in the past six months, though 29% have gotten both and 27% have gotten just the updated flu shot.

The poll also asked about COVID-19 vaccination since it became available in 2021: 49% of people over 50 said it’s been more than a year since their last dose, and 15% said they’ve never received it.

The leading reason people over 50 gave for not getting updated vaccines?

They didn’t think they needed them.

In all, 28% of people over 50 who didn’t get a flu vaccine in the past six months, and 29% of those who didn’t get a COVID-19 vaccine in the past year or ever, gave this as the main reason.

That’s despite clear evidence showing that staying up to date on both vaccines reduces the risk of serious illness and death in older adults, whose immune systems need regular “reminders” with updated vaccines tailored to recent mutations in the viruses.

Coming in second among reasons for not getting vaccinated recently were worries about the vaccines’ side effects (19% for flu and 27% for COVID-19), followed by a belief that the vaccines aren’t effective (18% and 19%, respectively).

Far fewer (10% for flu and 6% for COVID-19, respectively) said they just didn’t think of it. A few (4% and 3%) wanted to wait, and from 1% to 4% cited time, cost, insurance, availability or eligibility concerns.

The findings could help health care providers and public health agencies communicate better with middle-aged and older adults about the health benefits of annual vaccination and address any concerns, misinformation, or barriers.

The poll is based at the U-M Institute for Healthcare Policy and Innovation and supported by Michigan Medicine, U-M's academic medical center.

From late December 2025 to mid-January 2026, the poll team asked 2,964 U.S. adults age 50 and over if they’d gotten a flu vaccine dose in the last six months, and when their most recent COVID-19 vaccination was. Then, the team asked those who hadn’t sought vaccination recently their main reason why.

“The evidence is clear: these viruses can lead to serious illness, hospitalization, and death. That risk increases with age and underlying health conditions, and vaccination reduces that risk,” said Jeffrey Kullgren, M.D., M.P.H., M.S., the poll’s director, an Associate Professor of Internal Medicine at the U-M Medical School and a primary care physician at the VA Ann Arbor Healthcare System.

“These findings suggest that we must do a much better job helping people in their 50s and up understand that they will benefit from getting these updated vaccines each year, that the vaccine side effects are mild and short-lived, and that even if they later get infected and develop symptoms, vaccination means they won’t get as sick,” he added.

Emphasizing individual impacts is important, he noted. Experts and news stories often discuss vaccine effectiveness in percentages based on how well the vaccine reduces the risk of hospitalization or death in a population.

That’s different from what an individual might care most about: whether it will keep them from getting sick at all, or seriously ill, Kullgren said.

Even when a vaccine isn’t a perfect match for the virus strains that are circulating, a recent dose still nudges the immune system to be ready to fight off the virus in general, and can help reduce the severity and duration of symptoms, he notes.

No impact from changed recommendation for COVID-19 shot

Last spring, leaders at the U.S. Food and Drug Administration signaled a change in the agency’s recommendation about which adults should receive the COVID-19 vaccination. This was followed by an official change in the FDA approval and the recommendation from the Centers for Disease Control and Prevention in late summer.

But the poll suggests this change didn’t play a major role in older adults’ decision-making. Less than 1% of respondents who chose not to get vaccinated against COVID-19 in the last year said their main reason was they thought they weren’t eligible.

And in fact, the federal change only affects some of those who were polled: Those who are aged 50 to 64 without any chronic health condition that raises COVID-19 risk.

COVID-19 vaccination is still recommended for most older adults, including two doses a year for everyone over age 65 and anyone with a compromised immune system, and one dose a year for those under 65 who have underlying health conditions that put them at higher risk of severe COVID-19.

Many people ages 50 to 64 have at least one of those qualifying conditions, which include diabetes, asthma, obesity, high blood pressure, current or previous smoking, and physical inactivity.

The CDC didn’t change its recommendation that everyone over the age of 6 months should get an annual flu vaccination. And major medical societies and insurance companies announced that they would continue recommending or covering both vaccines for all children over 6 months and all adults, regardless of health status.

Differences in vaccination uptake

In addition to revealing reasons older adults didn’t get updated vaccines, the new poll data show some key differences in vaccination among different groups of people 50 and over.

The oldest adults (age 75 and over)– those with the highest risk of hospitalization and death from both viruses – had the highest rates of updated vaccination.

In all, 46% of those age 75 and up said they had gotten a COVID-19 vaccine in the last six months, compared with 37% of those age 65 to 74 and 20% of those age 50 to 64.

Flu vaccination was even higher in all age groups, with 76% of those in the oldest age bracket having gotten the latest flu shot, compared with 64% of those age 65 to 74 and 42% of those in their 50s and early 60s.

"This gap between flu and COVID-19 vaccination represents an opportunity to connect the dots for older patients: both of these viruses can put them at risk, both of them mutate rapidly, and both vaccines should be an annual tradition, even if they don’t get them at the same time,” said Kullgren.

He also points to the poll’s positive finding that adults 50 and over who report having at least one chronic health condition were much more likely to have gotten flu and COVID-19 shots recently than those without a chronic condition.

But even among those with chronic health conditions, 39% said they hadn’t had either vaccine in the last six months, though this rate was much lower than the 59% of those without chronic conditions who said so.

The poll also suggests a need for focus on those who have never gotten vaccinated against COVID-19, as they grow older and their risk of severe outcomes rises.

In all, 20% of those age 50 to 64 said they had never gotten a COVID-19 vaccine, along with 12% of those age 65 to 75 and 7% of those age 75 and up.

There’s also an income gap, with 19% of all those with household incomes under $60,000 saying they’ve never had a COVID-19 shot, compared with 12% of those with incomes over $60,000.

Not too late to vaccinate, including second doses of COVID-19 vaccine

Kullgren notes that it’s not too late in the current respiratory virus season for anyone to get a flu or COVID-19 vaccine if they haven’t already done so. It’s also almost time for second doses of the COVID-19 vaccine for those over 65 or with moderately to severely compromised immune systems who got this year’s version soon after it arrived in early September. They are recommended to get another dose starting six months after receiving the first.

So, if someone received their first dose in the first week of September, for instance, they can receive a second dose starting the first week in February. See the detailed adult vaccine schedule and the page about COVID-19 vaccines for immunocompromised people.

About the poll

The poll findings come from a nationally representative survey conducted by NORC at the University of Chicago for IHPI and administered online and via phone from December 29, 2025 to January 13, 2026 among 2,964 U.S. adults ages 50 to 98. The sample was weighted to reflect the national population. 

Read past National Poll on Healthy Aging reports and Michigan findings, and learn about the poll methodology. 

---

Method of Research

Survey

Subject of Research

People