A $4 Drug Could Lock HIV in Permanent Remission

For 15 years, the dominant strategy for curing HIV has been shock-and-kill: flush the virus out of the cells where it hides, then destroy it. The approach made elegant sense. It has also, by any honest accounting, underdelivered. Billions in funding, dozens of clinical trials, and the virus still rebounds in most patients within weeks of stopping antiretroviral therapy.

Now a team at Gladstone Institutes, the San Francisco-based biomedical research organization, is publishing the strongest evidence yet for the opposite idea: instead of waking the virus up, lock it down permanently. And the drug that might do it costs about four dollars a month.

The study00049-X), published March 20 in the peer-reviewed journal Immunity, analyzed blood from 75 people across four analytical treatment interruption trials—experiments where patients stop taking antiretroviral drugs under close medical supervision so researchers can study what happens next. The team, led by immunologist Nadia Roan and co-first authors Tongcui Ma and Ashley George, identified two genes in CD4+ T cells—the reservoir cells where HIV hides—that appear to act as molecular locks, keeping the virus silent.

The genes are DDIT4 and ZNF254. Both were expressed at significantly higher levels in people whose virus stayed suppressed longer after stopping treatment. ZNF254, in particular, showed dramatically elevated expression in elite controllers, the fewer than 1 percent of HIV-positive individuals who suppress the virus without any medication at all. George, in the Gladstone press release, floated the idea of delivering ZNF254 via gene therapy to effectively turn patients into controllers—ambitious, but the biology is now pointing there.

The metformin finding is what will grab headlines, and it should. Metformin, the ubiquitous diabetes drug prescribed to tens of millions of people worldwide, upregulates DDIT4. When the researchers treated patient cells with metformin ex vivo—in the lab, not in a person—it blocked HIV from reactivating. A cheap, well-characterized generic with decades of safety data, potentially repurposed to silence the most persistent virus in modern medicine.

But here is where the story gets genuinely interesting, and where every other outlet covering this paper has stopped short.

A 2024 study published in iScience01895-9) by Banga et al. found that metformin does something closer to the opposite: it reactivates latent HIV reservoirs and facilitates antibody recognition of infected cells. That is not block-and-lock. That is shock-and-kill—the very strategy the Gladstone paper implicitly argues against.

Both findings may be correct. The Roan team is looking at DDIT4-mediated silencing through mTOR pathway inhibition. The Banga paper identified a different mechanism involving BST2, a protein that helps the immune system flag infected cells. Metformin is a famously promiscuous drug—it touches multiple cellular pathways, and its effects can vary dramatically depending on the cell type, the dose, and the context. The idea that the same four-dollar pill could both silence the virus and wake it up is not a contradiction so much as a reminder that biology does not read press releases.

This tension matters for anyone watching HIV cure research. If metformin suppresses in some contexts and activates in others, the clinical path becomes considerably more complicated than "give everyone metformin." The only prior human trial testing this idea, the LILAC pilot study published in eBioMedicine by Yek et al., treated 22 non-diabetic people living with HIV and found only minor variations in blood reservoir levels—though it did show decreased mTOR signaling in gut tissue, which is where a large fraction of the HIV reservoir actually lives.

The author list on the Immunity paper tells its own story about where the field is heading. Robert Siliciano of Johns Hopkins University, who discovered the latent HIV reservoir in the 1990s—the finding that made a cure seem nearly impossible—is a co-author. So is Steven Deeks of the University of California, San Francisco, a leading figure in treatment interruption research, and Warner Greene and Melanie Ott, both of Gladstone. The study emerged from the HOPE Collaboratory, a $26.5 million NIH-funded consortium of 16 institutions, with industry partners including Amgen, the biotech giant, and Sangamo Therapeutics, a gene therapy company. The collaboratory's stated mission is "block, lock, and excise"—silence the virus, cement the silence, then cut the proviral DNA out of the genome entirely.

That framing—block-lock-excise rather than shock-and-kill—represents a strategic pivot in HIV cure research that has been building for several years but now has its most compelling dataset. The A5345 trial, one of the four analyzed in the Immunity paper, previously showed that among its 45 participants, the median time to viral rebound was just 22 days. That is the baseline the field is working against. The new findings suggest that the difference between 22 days and indefinite suppression may come down to which genes are active in a patient's reservoir cells—and that those genes might be pharmacologically targetable.

The study also identified immune signatures beyond the reservoir itself. Higher levels of stem cell memory CD8+ T cells and atypical natural killer cells correlated with delayed rebound, according to the EurekAlert release summarizing the paper. These are not the usual suspects in HIV immunology, and they suggest the immune system's role in post-treatment control is more layered than previously appreciated.

None of this is a cure. The ex vivo metformin result needs to be replicated in people. The LILAC trial's modest blood findings are a cautionary note. The ZNF254 gene therapy concept is years away from even a Phase 1 trial. And the Banga contradiction needs to be resolved before anyone designs a large clinical study around metformin for HIV.

But the shift in thinking is real. For a decade and a half, cure research bet on flushing the virus out. The Gladstone team is arguing, with data from 75 patients and some of the biggest names in HIV biology, that the better bet may be to make the locks stronger. If they are right, the tools might already be in the medicine cabinet.

The GEN News wire report that flagged this story covered the basics but missed the metformin tension entirely—a gap worth noting for anyone tracking how HIV cure research gets communicated to the public.