A pair of antibodies keeps HIV undetectable by ‘bolting the door and masking the key’

A combination of two new antibodies that disrupt the interaction between the virus and the cells maintained viral suppression in most participants for up to six months. Participants replaced their usual daily HIV medication with an infusion of the antibodies once every eight weeks.

Dr Martin Markowitz of Columbia University in New York presented the findings at the International Congress on Drug Therapy in HIV Infection (HIV Glasgow 2026) yesterday. The study, conducted with colleagues at six clinics across the United States, was simultaneously published in The Lancet HIV.

There were no virological failures – persisting increases in viral load above undetectable – during the antibody-only treatment. There were no serious negative health effects during the study, but two participants discontinued due to rash, a known side effect of antibody treatments. Although positive, these observations came from a small group of only 20 participants. 

The authors also note that these results were achieved without first testing to see whether the virus in the participants was susceptible (sensitive) to the antibodies. Such testing helps check whether the antibodies will work against a person’s particular strain of HIV, because an antibody that blocks one strain may be ineffective against another. If testing could be dropped, antibody treatment would be easier to offer in routine care. However, Dr Markowitz said it was too early to draw conclusions.

“This study does not establish that screening is unnecessary,” he told aidsmap. “It was a small, open-label study of only 24 weeks, so we cannot yet define the true rate of virologic failure or know how resistance might emerge over a longer treatment period.”

The antibodies

The two antibodies tackle HIV entry from opposite sides. The first, TMB-365, does not target the virus at all. Instead, it attaches to CD4, a protein on the surface of certain immune cells that HIV uses as its way in. The virus can still attach to CD4, but the antibody blocks the next steps needed to get inside. Think of it as holding the door shut even after the key has entered the lock. This is why TMB-365 is called a post-attachment inhibitor.

Because TMB-356 is related to an older antibody called ibalizumab, a member of the audience asked about the possibility of temporary effects of the antibody on CD4 cell counts, which ibalizumab was associated with. Dr Markowitz reassuringly said the new antibody had no negative effect on CD4 cell counts.

The second antibody, TMB-380, goes after the virus itself. It covers the part of HIV’s surface protein that normally attaches to CD4, effectively covering the virus’ key so it cannot fit into the lock. TMB-380 is a broadly neutralising antibody, meaning it can block many different strains of HIV. 

Together, the antibodies put two obstacles in the virus’s path – one covers the key, while the other keeps the door from opening. Targeting the cell could also make one route to resistance harder. As Dr Markowitz explained, “the virus cannot simply escape by mutating the antibody’s viral target; it must continue to use CD4 to enter susceptible cells.”

Of course, that does not make the combination impossible to develop resistance to, but, on paper, makes it much harder. 

Apart from targeting and inactivating disease-causing agents, antibodies can also tag these pathogens making them more visible to the immune system, thus inducing a stronger and more directed immune response. This action of antibodies is called opsonising and it is part of antibodies’ sometimes vaccine-like effects. When asked whether TMB-380 had a similar effect, Dr Markowitz noted the lack of such observations so far. 

The study

Some of the main criteria for the participants to enter the study were having had an undetectable viral load for the preceding six months, a CD4 count above 350, overall stable health and no history of virological failure (uncontrolled viral load despite treatment). The study lacked a comparison group, so the researchers couldn’t directly compare the antibody treatment to standard treatment.  

Twenty participants were analysed, all of whom were men aged between 24 and 67. Three-quarters were White. Their average CD4 count stood at 653 cells. 

The participants stopped their usual HIV drugs on the day of the first infusion and both antibodies were given together over an hour, with further infusions scheduled at two months and four months. Viral load was checked at least every four weeks.

At six months, 16 of the 17 participants with available results had a viral load below 50 copies, while one participant had a viral load of 59 copies, giving a suppression rate of 94%. At the end of the six-month trial, participants returned to normal oral therapy. 

The researchers concluded that the antibody treatment was safe as there were no serious negative health outcomes during the study. However, some mild to moderate side effects were observed such as rash, fatigue, and headache and – less commonly – flushing, fever, chills and aches. The average CD4 counts also remained stable during treatment.

Three participants had temporary viral load increases between 200 and 300 copies, followed by re-suppression. All three were among the six participants weighing more than 90kg and had lower TMB-365 concentrations. Their TMB-380 concentrations were also relatively low, but remained at levels the researchers expected to provide antiviral coverage. Because nobody met the study’s definition of treatment failure, testing for new resistance to the antibodies was not carried out.

These observations suggest that higher bodyweight may have contributed to lower TMB-365 exposure and possibly leading to the temporary increases in the viral load of the three participants, although it remains unconfirmed. 

The participants also gave the antibodies better ratings for convenience and fit with their lifestyle.

Concluding thoughts  

Antibodies for HIV treatment have been an important research area, yet it has proved difficult to discover or produce antibodies with a broad enough coverage of HIV strains. The present two antibodies may possess a favourable breadth and pharmacological profile, yet they cannot fully overcome the challenges facing antibody treatments in general. One of these is the possibility of our body producing antibodies against the antibodies used for treatment, which was seen in one of the participants, although at low levels and with no effect on antibody concentrations or viral load. However, a more pronounced production of these antibodies targeting the therapeutic antibody may lead to its rapid elimination and loss of its protective concentrations, jeopardising the person’s undetectability. 

Currently, long-acting injections of cabotegravir plus rilpivirine already offer an alternative to daily pills, with dosing every two months. For the antibody combination, the question is what an hour-long infusion could offer beyond the options already available, which Dr Markowitz answered from a very practical angle. 

He highlighted a possible role for people ageing with HIV who take several medicines for other health conditions. Antibodies are processed differently from conventional HIV drugs and do not depend on the liver enzyme system that breaks down many drugs. This could mean fewer concerns about drug interactions.

On a practical level, an infusion every eight weeks requires clinic visits, staff and facilities. However, Dr Markowitz suggested that regular visits could also have value for some people. “An every-eight-week visit may be a reasonable balance for some patients between substantially reducing treatment frequency and maintaining regular clinical contact, particularly for individuals managing other health conditions,” he told aidsmap.

The combination is now being tested in a larger, one-year phase 2b trial, with participants randomly assigned to receive the antibodies or oral HIV treatment. The trial uses bodyweight-based dosing of TMB-365, following the findings of the study presented in Glasgow. These further studies will also need to establish how well the combination works in a more diverse population, including women.

References

Markowitz M et al. A 24-week phase 2a maintenance study of TMB-365/TMB-380 q8w in people with suppressed HIV-1 infection. HIV Glasgow 2026, abstract O31.

Markowitz M et al. Safety, pharmacokinetics, and antiviral activity of a combination of TMB-365 and TMB-380 in people living with HIV suppressed with antiretroviral therapy in the USA: a phase 1b/2a, open-label, dose-escalation study. The Lancet HIV, advance publication, 10 November 2026.

https://doi.org/10.1016/ S2352-3018(26)00148-7