Yes, HIV can stay dormant, but only inside a small population of latently infected CD4 T cells where the viral genome is stitched into human DNA and produces no new virus. This latent reservoir, estimated at roughly 1 to 1 million cells per patient, is invisible to antiretroviral drugs and mostly invisible to the immune system. It is why modern therapy suppresses HIV to undetectable levels yet almost never cures it. Understanding latency means understanding integration, transcriptional silencing, and the rare events that wake the virus up.
What Dormancy Actually Means at the Molecular Level
Dormancy in HIV is not a sleeping virus particle. It is a silent DNA blueprint. After HIV enters a CD4 T cell, reverse transcriptase converts its RNA genome into DNA, and integrase splices that DNA into the host chromosome. In an activated T cell, this provirus is transcribed aggressively and makes new virions. In a resting memory T cell, the same provirus sits in chromatin without producing viral RNA.
The distinction matters because "dormant HIV" can describe three different states. First is latent provirus in a resting cell, with no detectable viral protein. Second is low-level ongoing replication in tissue sanctuaries, which is a failure of drug penetration rather than true latency. Third is a defective provirus that can never reactivate, which makes up the majority of integrated sequences. Roughly 90 to 95 percent of integrated HIV DNA in patients on suppressive therapy is defective and cannot produce infectious virus, per landmark sequencing work from the Siliciano laboratory. Only a small fraction is intact and replication-competent.
The mechanism of silencing is epigenetic. The host protein CTIP2 and histone deacetylases compact chromatin around the provirus. DNA methylation at CpG islands near the viral promoter adds a second lock. The viral Tat protein, which normally amplifies transcription in a positive feedback loop, is absent in resting cells, so the loop never starts. In plain terms: latency is a chromatin problem, not a drug problem.
How the Reservoir Forms and Why It Persists
Latency is established in the first days to weeks of infection, before most people know they are positive. A fraction of infected CD4 T cells naturally transition into long-lived memory subsets. When those cells return to a resting state, the integrated provirus goes quiet with them. The reservoir is therefore seeded early and becomes essentially permanent.
Three memory subsets dominate the reservoir. Central memory T cells (Tcm) carry the largest share of intact provirus. Transitional memory cells (Ttm) are the most transcriptionally active. Effector memory cells (Tem) are more differentiated and shorter-lived. The relative proportions shift with time on therapy, and this matters for cure research because each subset responds differently to latency-reversing agents.
Cell proliferation keeps the reservoir alive without new infection. Homeostatic proliferation driven by interleukin-7 and antigen-driven clonal expansion copy the provirus into daughter cells. Recent estimates suggest that clonal expansion accounts for the majority of reservoir persistence in patients suppressed for years, not residual replication. This is a non-obvious correction to the older assumption that ongoing viral replication maintained the reservoir. The practical consequence: blocking new infection does not shrink the reservoir, because the cells are copying themselves and their silent viral cargo.
Quantified context helps. Total body HIV DNA in a suppressed patient typically ranges from 10 to 1,000 copies per million CD4 T cells. The replication-competent reservoir measured by the quantitative viral outgrowth assay (QVOA), the gold standard, sits at roughly 0.1 to 1 infectious unit per million resting CD4 T cells. That is a measurement floor issue as much as a biology issue: QVOA underestimates the true reservoir by a factor of 10 to 100 because most intact proviruses do not reactivate in a single round of stimulation.
Timeline: From Exposure to a Silent Reservoir
The dormancy story unfolds on a predictable schedule in untreated infection, and each row below maps to a measurable window.
| Time after exposure | What happens | Reservoir relevance |
|---|---|---|
| 0 to 72 hours | Virus crosses mucosa, infects local CD4 cells and dendritic cells | No reservoir yet; transmission risk is highest here |
| 3 to 10 days | Virus reaches draining lymph nodes, systemic spread begins | First latently infected cells appear |
| 2 to 4 weeks | Peak viremia, acute retroviral syndrome in 50 to 70 percent | Reservoir seeded broadly across memory subsets |
| 1 to 3 months | Immune control begins, viral set point established | Reservoir size stabilizes at roughly 10 to 1,000 copies per million CD4 cells |
| Months to years untreated | Progressive CD4 decline, 50 to 100 cells per microliter untreated | Reservoir persists and may contract slightly |
| First weeks on ART | Viremia falls 1 to 2 log10, then slower second phase | Reservoir half-life lengthens to months or years |
| Years on suppressive ART | Undetectable plasma RNA, stable reservoir | Clonal expansion dominates persistence |
The critical implication: starting antiretroviral therapy within days of infection, rather than weeks, shrinks the reservoir but does not eliminate it. Even the most aggressive early-treatment cohorts, followed for years, show measurable intact provirus.
The Numbers: How Small and How Stable Is the Reservoir
Reservoir size is the central variable in every cure strategy, and the numbers are sobering. In patients on suppressive antiretroviral therapy with undetectable plasma viral loads, the intact proviral DNA assay (IPDA) finds roughly 50 to 100 intact proviruses per million CD4 T cells. That is a 10-fold to 100-fold lower figure than older total-DNA measurements suggested, because IPDA excludes defective sequences. The revision is important: it means the true target for cure is smaller than once feared, but also that measuring it requires precise assays.
The reservoir's decay is painfully slow. On suppressive therapy, the latent reservoir's half-life is estimated at roughly 44 months in some cohorts, meaning complete natural eradication would take more than 60 years of continuous therapy. Some studies put the half-life closer to 6 to 12 months early on and much longer later, as the most stable clones dominate. Either way, the arithmetic does not deliver a cure within a normal lifespan.
Reactivation frequency is the other key number. Even in suppressed patients, rare latently infected cells reactivate and produce transient viral blips. These blips usually stay below 50 copies per milliliter and are controlled by residual immune pressure, but they demonstrate that the reservoir is not truly frozen. It is a slow leak, not a sealed vault.
Variations and Edge Cases That Change the Picture
Not every patient's reservoir behaves the same way. Elite controllers, roughly 0.5 percent of people with HIV, maintain undetectable viral loads without therapy and often carry extremely small reservoirs. Their CD4 cells show enhanced restriction of viral transcription, which is a partial natural latency that cure researchers study closely.
Post-treatment controllers are rarer and more surprising. Some individuals who started therapy very early and then stopped maintain viral suppression for months or years. The VISCONTI cohort documented this in a small number of patients, but most eventually lose control. These cases confirm that the reservoir can be functionally contained, not eliminated.
Tissue sanctuaries add another layer. Lymphoid tissue in the gut and lymph nodes harbors provirus that blood sampling misses. Central nervous system and genital tract compartments can hold distinct viral populations with different reactivation thresholds. A blood test showing near-zero reservoir does not guarantee the same in tissue, a caveat that complicates every "functional cure" claim.
Finally, defective proviruses are not harmless bystanders. They can produce aberrant viral proteins that drive immune activation and inflammation, contributing to the elevated cardiovascular and metabolic risk seen in long-term survivors. Elimination strategies must account for these sequences, not just intact ones.
Common Misconceptions About Dormant HIV
The most dangerous misconception is that "undetectable means the virus is gone." Undetectable refers to plasma viral load below roughly 50 copies per milliliter, which is a treatment success, not a cure. The reservoir persists at levels far below that detection threshold.
A second misconception is that latency-reversing agents can flush the virus out and cure patients. Clinical trials of HDAC inhibitors such as vorinostat and panobinostat showed measurable increases in viral transcription, but no reduction in reservoir size. The "shock and kill" strategy works on the shock side far better than the kill side, an edge case that reshaped cure research toward immune enhancement.
A third error is conflating low-level viremia with true latency. Detectable virus in a suppressed patient often reflects drug penetration gaps or adherence issues, not waking of a dormant cell. Per the 2021 IAS-USA treatment guidelines, persistent low-level viremia between 50 and 200 copies warrants resistance testing and regimen review rather than a diagnosis of reservoir reactivation. This distinction has direct clinical consequences.
What to Actually Do If You Live With HIV
- Stay on antiretroviral therapy without interruption; adherence above 95 percent keeps plasma RNA undetectable and limits reservoir reseeding.
- Get viral load monitored every 3 to 6 months once stable, and sooner after any change in regimen.
- Discuss any viral blip with your clinician rather than assuming the dormant virus has "woken up."
- Enroll in cure research only through accredited trials; understand that latency-reversing agents are experimental, not therapeutic.
- Report new infections, fevers, or inflammation symptoms; residual immune activation from the reservoir does have long-term health effects.
Limits of What We Know
Latency is established science. The mechanism of transcriptional silencing, the role of chromatin, and the persistence of intact provirus are all well documented. What remains hypothesis is whether any intervention can eliminate the reservoir safely in a living human. The "block and lock" approach, which deepens latency rather than reversing it, is promising in animal models but unproven in people. The Berlin and London patients who achieved sustained remission after stem cell transplants are proof of principle, not a scalable cure; both received treatment for cancer, and the procedure carries mortality risk above 10 percent. Gene editing strategies targeting CCR5 or the provirus itself are in early trials with no long-term human data.
The honest answer to whether HIV can stay dormant is yes, and that dormancy is the single biggest obstacle between current therapy and a cure. The virus hides not by outrunning drugs but by becoming part of the cell it infects. Until science finds a way to identify and eliminate every intact provirus, suppression will remain lifelong, and the reservoir will remain the target