Long-acting, bNAbs, and cure — where the science actually is in 2026.

If you live with HIV, you have probably had the experience of someone forwarding you a headline that says HIV has been cured,. Usually the headline is true and the framing is wrong. Something real happened — a person genuinely no longer has HIV in their body — but the thing that happened is not something you can go ask for.

The confusion is not the reader's fault. Three separate scientific stories are running at the same time, and press coverage blends them constantly.

This page keeps those three apart on purpose. It then lays out, case by case, every documented cure as of the 26th International AIDS Conference in Rio de Janeiro, July 2026.2

Three worlds, one word: "long-acting"

Before anything else, the sorting. If you take one thing from this page, take this.

World one — long-acting treatment

This is antiretroviral therapy delivered by injection instead of by mouth, for people who already live with HIV. It does not remove HIV from your body. It does exactly what your pills do — hold the virus at undetectable levels — on a different schedule. To start most long-acting treatment you have to already be virally suppressed, because the regimen is designed to maintain suppression rather than to bring a high viral load down.11

World two — long-acting prevention

This is pre-exposure prophylaxis — PrEP — for people who are HIV negative. It has nothing to do with treating HIV and it does nothing for someone who already has it. Long-acting PrEP is where the most spectacular clinical trial results of the past three years have landed, with a twice-yearly injection that came close to eliminating new infections in its trials.15

World three — cure research

This is the attempt to remove HIV from the body, or to make the body permanently control it without medication. It is a genuinely different scientific problem from suppressing the virus, because HIV writes itself into the DNA of long-lived immune cells and sits there silently, untouched by antiretrovirals. That hidden population is called the reservoir, and it is the wall. The cure cases we have are real proof the wall can come down. They are not a treatment.

Long-acting treatment — Cabenuva and Sunlenca

For most of the history of effective HIV treatment, "taking your meds" meant swallowing something every single day for the rest of your life. That is a remarkable achievement and also, for many people, a daily reminder they did not ask for. Long-acting injectable treatment is the first serious alternative.

Cabenuva — cabotegravir plus rilpivirine

Cabenuva is two long-acting drugs given as intramuscular injections, either monthly or every two months, for people living with HIV who are already virally suppressed on a stable regimen. The every-two-month schedule rests on ATLAS-2M, a phase 3b randomized, open-label, non-inferiority trial that enrolled 1,045 adults across 13 countries and compared cabotegravir 600 mg plus rilpivirine 900 mg every eight weeks against cabotegravir 400 mg plus rilpivirine 600 mg every four weeks.11

The trial ran long. Through week 152 — nearly three years — 87% of the every-two-month group and 86% of the monthly group were still under 50 copies/mL, with confirmed virologic failure in 12 people (2.3%) on the eight-week schedule versus 2 (0.4%) on the four-week schedule.11

Read honestly, that is a durable regimen with a small but real difference in failure rate favoring monthly dosing.

That result deserves emphasis because it inverts an old assumption. Long-acting injectables were often positioned as a lifestyle upgrade for people already doing well. LATITUDE suggests they may matter most for people whose lives make a daily pill genuinely hard — housing instability, incarceration, mental health crises, substance use, work that does not follow a schedule. Sticking with treatment is a logistics problem far more often than a willpower problem, and this is a logistics solution.

Sunlenca — lenacapavir for multidrug-resistant HIV

Sunlenca is lenacapavir, the first approved HIV capsid inhibitor, given as a subcutaneous injection every six months alongside oral loading doses. Its treatment approval is narrow and specific: it is indicated in combination with other antiretrovirals for adults with multidrug-resistant HIV-1 whose current regimen is failing because of resistance, intolerance, or safety considerations.13

That approval came from CAPELLA, a trial in 72 people with multiclass resistance — participants had to have a viral load of at least 400 copies/mL, resistance to at least two antiretrovirals from at least three of the four main drug classes, and no more than two fully active agents left. In the randomized cohort, 30 of 36 people (83%) reached a viral load under 50 copies/mL by week 52, with a mean CD4 increase of 82 cells/µL.13 For people who had run out of options, that is a genuinely different life.

The label is also honest about the costs. Injection-site reactions were common — 65% of participants had one by week 52 — along with nausea, though only 1% discontinued because of an adverse event. Capsid resistance substitutions emerged in 9 of 22 people (41%) with confirmed virologic failure, most often M66I.13 Lenacapavir is powerful, and it is not indestructible.

Long-acting prevention — Apretude and Yeztugo

Now switch worlds completely. Everything in this section is for people who do not have HIV.

Apretude — cabotegravir every two months

Long-acting injectable cabotegravir for prevention was tested in two large trials. HPTN 083 enrolled cisgender men who have sex with men and transgender women and compared injections every two months against daily oral tenofovir/emtricitabine: 13 HIV infections in the cabotegravir group (0.40 per 100 person-years) versus 39 in the oral group (1.22 per 100 person-years) — a hazard ratio of 0.34, roughly a 66% reduction in new infections compared with an already highly effective daily pill.14

HPTN 084, in cisgender women across sub-Saharan Africa, was even more striking: 4 infections on injectable cabotegravir (0.20 per 100 person-years) against 36 on the daily oral comparator — about a 90% reduction, with superiority declared at a planned interim review in November 2020.14

Yeztugo — lenacapavir twice a year

Then came the results that reset expectations. The same molecule used in Sunlenca, given subcutaneously every six months, was tested as PrEP in the PURPOSE program and approved by the FDA on June 18, 2025 for adults and adolescents weighing at least 35 kg, with a required negative HIV-1 test before starting and before each injection.16

It is worth sitting with what those numbers mean. A twice-yearly injection produced essentially zero infections in a population of young women where oral PrEP had struggled for a decade — not because anyone lacked motivation, but because a daily pill you must hide, store, and remember is a different product than two appointments a year. CDC's 2025 recommendation places lenacapavir alongside oral PrEP and injectable cabotegravir as an option to be discussed with anyone who could benefit.15

None of this touches HIV that is already established in the body. Lenacapavir prevents HIV from taking hold; it does not clear a reservoir that exists. For the prevention side in full, see our PrEP page.

Broadly neutralizing antibodies — promising, and not a product

Broadly neutralizing antibodies, or bNAbs, are antibodies that a small minority of people living with HIV naturally develop after years of infection — antibodies potent enough to block many different HIV strains at once. Researchers isolate them, manufacture them, and infuse them. They are not antiretrovirals; they work by binding the virus directly.

The landmark proof that this can work in humans came from the Antibody Mediated Prevention (AMP) trials, two parallel efficacy studies of a single bNAb called VRC01. The headline result was a disappointment: neither trial showed significant overall protection. The pre-specified secondary analysis was the real finding — against HIV strains that were actually susceptible to VRC01 in the laboratory, prevention efficacy was 75%, irrespective of gender, route, or viral subtype. The problem was that only about 30% of circulating strains were susceptible.17

That reframed the whole field. A single antibody is not enough; you need combinations covering different targets on the viral envelope, or engineered antibodies with a longer half-life, or both. The most-studied combination is 3BNC117 and 10-1074, two bNAbs hitting independent sites on the HIV envelope spike. In their long-acting engineered forms — 3BNC117-LS and 10-1074-LS — given to people who then interrupted antiretroviral therapy, results presented at CROI 2025 showed that at week 20, 75% of bNAb recipients had not experienced viral rebound, compared with 11% of placebo recipients.17

Therapeutic vaccines — the hardest problem in immunology

A preventive HIV vaccine would teach an HIV-negative immune system to block the virus. A therapeutic vaccine would teach the immune system of someone already living with HIV to control the virus without daily medication. Neither exists. HIV has defeated more vaccine candidates than any pathogen in modern medicine, for structural reasons: it mutates constantly, integrates into host DNA within days, hides the sites antibodies need, and destroys the CD4 cells that coordinate immune responses.

The most encouraging recent movement is in mRNA-based immunogen design. Rather than trying to produce broadly neutralizing antibodies in one shot, this generation of candidates tries to prime the rare naive B cells that could eventually mature into bNAb producers — a deliberate, multi-step immune-education strategy. In trials published in 2025, mRNA-encoded HIV envelope nanoparticle vaccines drove early maturation of bNAb precursors in humans, and mRNA-encoded membrane-anchored envelope trimers elicited neutralizing antibody responses in 80% of recipients across 108 HIV-negative adults at ten United States sites, with responses appearing after the second dose and strengthening after the third.18

Those trials were also candid about tolerability: about 18% of participants in one study had skin-related adverse events, including urticaria in roughly 12%.18 And "neutralizing antibodies against the vaccine's own strain" is several steps short of "protection against the strains circulating in the world." This is real scientific progress on a problem that has resisted forty years of effort. It is not a vaccine.

On the therapeutic side specifically, earlier attempts illustrate the pattern. An intranodally administered mRNA therapeutic vaccine was safe and increased the proportion of participants mounting HIV-specific T-cell responses from 31% to 80%, and transiently increased cell-associated HIV RNA — but had no impact on the amount of HIV DNA persisting in cells.20 The gap between "the immune system responded" and "the reservoir shrank" is the recurring story of therapeutic vaccination.

What "cure" actually means — three different words

Cure language in HIV is precise, and the precision is not pedantry. Three distinct outcomes get called a cure in the press.

• Sterilizing cure (eradication). Every replication-competent copy of HIV is gone from the body. This is the strongest claim and the hardest to prove, because proving absence requires searching tissue you cannot fully sample. It is what the stem cell transplant cases appear to represent.
• Functional cure (remission). HIV is still present somewhere, but the body controls it indefinitely without medication and without disease progression. This is what most cure research is actually aiming for, because it may be achievable at scale.
• Post-treatment control. A person stops antiretroviral therapy and stays undetectable for an extended period without any intervention designed to cure them. Cohort studies have documented this in a small number of people, some for more than five years.1

Why the reservoir is the wall

Within days of acquisition, HIV integrates its genetic material into the DNA of long-lived CD4 cells. Some of those cells then go quiet. Stop treatment and, in almost everyone, some of those cells wake up and viral load rebounds within weeks.

This is why "undetectable" and "cured" are different states, and why the honest framing of the entire cure field is the one the Treatment Action Group uses: across many therapeutic approaches under study, no broadly usable intervention has yet produced clear evidence of cures or remissions — the best results so far involve small reductions in the reservoir and some cases of extended viral control after stopping treatment.22

The people who have been cured of HIV

Here is the roster. As of AIDS 2026 in Rio de Janeiro, 13 people have been cured of HIV following allogeneic stem cell transplantation performed to treat a life-threatening blood cancer, according to amfAR's running summary of the field.1 Separately, two people appear to have cleared HIV without any transplant at all — described as exceptional elite controllers.1 Different sources count to different totals depending on whether they include the elite controllers, cases still under observation, or remissions not yet formally published. Names are used only where the person made them public.

The two people who were never transplanted

These two cases belong in a category of their own, and they may ultimately matter more to the rest of us than any of the thirteen.

Not a practical cure — read this part twice

Researchers say this themselves, consistently and without hedging. Reporting from AIDS 2026 noted that scientists presenting the newest cases stressed the procedure is too dangerous and costly to be a cure for anyone without a life-threatening cancer, and that its value lies in guiding safer approaches — such as gene editing to disable CCR5 — that could one day reach far more people.2 Professor Sharon Lewin, then president-elect of the International AIDS Society, put it plainly when the New York case was announced: "A bone marrow transplant is not a viable large-scale strategy for curing HIV, but it does present a proof of concept that HIV can be cured."6

Gene editing — the attempt to make it scalable

If the transplant cases prove that removing CCR5 or removing HIV DNA can cure HIV, the obvious question is whether you can do that without the transplant. That is gene editing, and it is very early.

The furthest-along attempt in humans was EBT-101, from Excision BioTherapeutics — a CRISPR-Cas9 therapy delivered by an AAV9 viral vector with two guide RNAs designed to cut the HIV provirus at three sites simultaneously, snipping out the integrated virus rather than editing the host. It was tested in a phase 1/2 trial with long-term follow-up planned out to 2037.

The safety results were genuinely encouraging. The trial met its primary safety endpoint: no serious adverse events, only grade 1 events attributed to the therapy, and no detectable off-target DNA damage.19 The efficacy results were not. Of the participants who stopped antiretroviral therapy after dosing, all rebounded and restarted treatment, though one stayed suppressed for 16 weeks — longer than the typical rebound interval. The guide RNAs had been designed against HIV subtype B and did not perfectly match some participants' viral sequences.19

Dr Rachel Presti of Washington University in St. Louis, presenting the data, framed it the way early-phase investigators should: the trial provides important clinical evidence that a gene editing modality can be safely delivered to target HIV DNA reservoirs in human cells.19

The reservoir — shock and kill, block and lock

Two opposite strategies dominate reservoir research, and it helps to know both names because you will see them in coverage.

Shock and kill

The idea: use a latency-reversing agent to wake up silent infected cells, so they start producing viral proteins and become visible to the immune system, which then clears them. The best-studied latency-reversing agents are histone deacetylase inhibitors — vorinostat, romidepsin, panobinostat — drugs borrowed from oncology.

They do the first half. In a carefully designed study of 16 people, serial doses of vorinostat spaced 72 hours apart reliably increased HIV RNA inside resting CD4 cells, and vorinostat was well tolerated. But despite a month of repeatedly reversing latency, investigators observed no measurable decrease in the frequency of latent infection.20 A 2026 review of the whole class reached the same conclusion: no clinical trial to date has demonstrated that any latency-reversing agent achieves a statistically significant reduction in the size of the latent reservoir.20

You can wake the virus up. The immune system does not reliably kill what wakes up. That is the gap.

Block and lock

Block and lock is still preclinical. The bet runs the opposite way: instead of waking HIV, force it into a permanent, unbreakable sleep. Latency-promoting agents aim to induce epigenetic changes at the HIV promoter that keep the provirus silent even after the drug is withdrawn — a functional cure by silence rather than by clearance. The most studied candidate is didehydro-cortistatin A, which blocks the HIV Tat protein and appears to drive durable silencing in laboratory and animal models.20

The pipeline map — realistic timelines

A rough honest picture of where each approach sits. Approvals are facts; timelines for unapproved approaches are estimates, and the history of this field is that estimates slip.

What to watch, and how to join a trial safely

A short list of the things most likely to generate real news in the next few years: continued readouts from the PURPOSE prevention program, including data in pregnant and lactating people and adolescents; the next generation of bNAb combinations that combine potency with long half-lives; the next-generation gene editing vectors being developed after EBT-101; and the annual science at CROI, IAS, and the International AIDS Conference, where cure cases are typically presented before they are published.

The other thing worth watching is the funding and delivery environment, which is not neutral scientific ground. Reporting from AIDS 2026 described the replacement of PEPFAR programs with bilateral memoranda of understanding — 34 signed since Kenya's in December 2025 — at roughly a third lower funding.2 A twice-yearly injection that prevents nearly every infection is only as good as the system that delivers it.

Who is doing the work — and who is holding it accountable

Cure research is unusually shaped by community advocacy, in part because the people asked to take risks in cure trials are people living with HIV.

The Martin Delaney Collaboratories

Named for the late treatment activist Martin Delaney, the Collaboratories are the flagship United States public investment in HIV cure research, funded by NIAID. The program launched in July 2011 with three consortia, expanded to six in 2016, and now supports ten Collaboratories with roughly $53 million per year over five years — an increase of about 75% over the prior cycle.21

amfAR

amfAR funded the original research identifying CCR5 as HIV's entry co-receptor — the discovery underneath every case in this article — and maintains the running public summary of cure cases this page relies on.1 Its ICISTEM collaboration systematically studies people living with HIV who need stem cell transplants, which is how cases like the Düsseldorf patient's were confirmed rather than merely noticed.1

Treatment Action Group

TAG's periodically updated HIV Cure Research information sheet and its annual Pipeline Report are the most useful plain-language accountability documents in the field, precisely because they refuse to overstate. TAG's summary judgment — that no broadly usable intervention has yet produced clear evidence of cures or remissions, and that the best results are small reservoir reductions and some extended viral control after stopping treatment — is the sentence to hold onto when a headline gets ahead of the data.22

Alongside these, the International AIDS Society coordinates global cure strategy work, and community publications including POZ, Positively Aware, and TheBody carry the voices of people who have actually sat in cure trials — worth reading for what the experience is like, alongside the primary sources for what the science shows.

What to do with all of this

Concrete next steps, in rough order of how much they will change your life.

1. Stay on treatment. Nothing on this page changes that, and one thing on this page reinforces it hard: consistent treatment started early is associated with a life expectancy approaching the general population.9 If your regimen is a burden, that is a reason to change the regimen, not to stop.
2. Ask whether a long-acting regimen fits you. If daily pills are hard — for any reason at all, including reasons you would rather not explain — the LATITUDE result is your evidence base to bring to the appointment.12 Ask about eligibility, dosing windows, injection-site reactions, coverage, and what happens if you miss a visit.
3. Get the PrEP conversation to the people who need it. If you are in a mixed-status relationship, your partner now has options that did not exist three years ago, including a twice-yearly injection that CDC strongly recommends.15 That is one of the most useful things you can hand someone you love.
4. Learn how to read a cure headline. Ask: treatment, prevention, or cure? Humans or cells? Did anyone stop antiretroviral therapy? Did the virus stay away? Those four questions will correctly deflate most of the coverage.
5. If you want to contribute, contribute deliberately. Cure research needs participants, and it needs them informed. Read our clinical trials page, talk to your own clinician, and never enter a study that asks you to stop treatment without a documented monitoring and restart plan.
6. Push on access, not just discovery. The gap between what science can do and what people can get is where the epidemic actually lives — see the 2026 state of the epidemic and UNAIDS and WHO for how that gap is currently trending.

Thirteen people have been cured of HIV, and two more appear to have cured themselves. That is real, and it is worth celebrating without pretending it is available. Timothy Ray Brown's own hope was that he would not be the last — that other people would, in his words, join his club.5 Twelve people have. In the meantime, the thing that keeps you alive and keeps your partners safe is the least dramatic tool in the article: treatment that works, taken on whatever schedule fits your life.