What Is a Viral Reservoir, and Why Does It Make HIV So Hard to Cure?

Short answer: an HIV reservoir is a population of infected cells that carries persistent viral genetic material while producing little or no virus. Antiretroviral therapy can stop active replication, but it cannot reliably identify and remove every silent infected cell. If treatment stops, some reservoir cells can reactivate and restart infection.

How HIV creates a long-lived archive

HIV is a retrovirus. After entering a susceptible cell, it copies its RNA into DNA and inserts that DNA into the cell's chromosomes. The integrated copy is called a provirus. An active provirus directs the cell to make new virus; a latent one remains largely silent.

This integration is central to the viral reservoir. A resting memory CD4 T cell can preserve the provirus for years and can divide, copying the viral DNA along with its own genome. The reservoir can therefore persist without continuous rounds of new infection.

Why antiretroviral therapy does not remove it

Combination antiretroviral therapy blocks stages such as reverse transcription, integration, and viral maturation. It is extraordinarily effective at preventing new productive infections and can reduce virus in blood below routine detection. But the drugs act on active viral processes. A latent provirus performing almost none of those steps presents little target.

The immune system faces a similar problem. Cytotoxic T cells can recognize infected cells displaying viral fragments, but a silent cell may display too little to reveal itself. HIV can also accumulate escape mutations and persist in anatomical sites where immune surveillance and drug penetration differ.

What the reservoir contains

Most proviral DNA detected in treated people is defective. Cure research must distinguish total HIV DNA from the much smaller fraction capable of reigniting infection.

When is the reservoir established?

HIV begins seeding long-lived cells very early after infection. Starting therapy quickly can limit the reservoir's size, preserve immune function, and reduce viral diversity. It does not guarantee eradication because integration can occur before diagnosis or treatment.

Infants and adults also differ in immune development, cell populations, exposure timing, and drug delivery. Results in newborn animal models can reveal mechanisms, but they do not automatically predict treatment outcomes in human infants or adults.

Undetectable is not absent

A standard viral-load test samples blood and has a detection threshold. “Undetectable” means the assay found less virus than that limit in that sample. It does not prove that every tissue is virus-free or that no replication-competent provirus remains.

For people maintaining therapy, sustained viral suppression prevents sexual transmission, summarized as U=U, undetectable equals untransmittable. That public-health fact is compatible with persistence of a latent reservoir: treatment suppresses the virus so effectively that it is not sexually transmitted, while stopping therapy may still allow rebound.

How scientists test for rebound

The most direct test of treatment-free control is an analytical treatment interruption, in which therapy is paused under a strict protocol with frequent monitoring and predefined restart criteria. Because interruption can harm the participant and permit transmission, it is used cautiously and cannot be replaced by a single negative blood test.

Researchers also use quantitative viral outgrowth assays, intact-proviral-DNA measurements, tissue sampling, and sequencing. Each captures a different part of the reservoir; none measures every relevant cell perfectly.

Strategies under investigation

Remission, functional cure, and sterilizing cure

• Durable remission: virus remains controlled for a prolonged period without therapy, although reservoir material may persist.
• Functional cure: often used for long-term treatment-free control that prevents disease and transmission, without proving every provirus is gone.
• Sterilizing cure: no replication-competent HIV remains anywhere in the body. This is the hardest claim to demonstrate.

Terminology varies among researchers, so reports should state the actual observation: how long treatment stopped, which samples were tested, and what assays found.

What an infant-macaque study showed

A 2026 study treated infant macaques within three days of infection using antiretroviral drugs, broadly neutralizing antibodies, and an experimental therapy that blocks CCR5. Eight animals had no detectable virus a year after treatment ended. The result suggests that striking the virus before the reservoir is fully established, through several mechanisms at once, may enable durable control. Read A Three-Drug Combination Cleared HIV in Baby Monkeys for a Year.

It did not yet establish a human cure. The sample was small, the subjects were nonhuman primates, treatment began exceptionally early, and follow-up cannot prove that no rare replication-competent cell remains. Human trials must establish safety, feasibility, and whether the effect generalizes.

Why the reservoir remains the central obstacle

HIV treatment transformed infection into a manageable chronic condition, but control and eradication are different tasks. A cure must reach rare cells distributed across tissues, distinguish dangerous proviruses from defective remnants, prevent new infection, and avoid damaging essential immune cells. The reservoir is difficult not because it is always active, but because it can remain biologically quiet, scattered, and capable of returning.