iconstar paper   HIV Articles  
Back grey arrow rt.gif
Rekindled HIV infection (latent reservoirs) - Perspective
  Download the PDF here
Science 29 August 2014
Janet D. Siliciano and Robert F. Siliciano
Palpable disappointment accompanied the news last month of a child (the "Mississippi baby") who was thought to be cured of HIV-1 infection by early antiretroviral treatment, but suddenly had detectable virus in the blood 2 years after the treatment was stopped. This setback followed the report of similar delayed viral rebounds 3 to 8 months after antiretroviral treatment was ended in two HIV-1-infected adults who had received bone marrow transplants (1). These findings have been viewed as a blow to HIV-1 eradication efforts. Although disappointing, the late rebounds are of enormous scientific importance. They reaffirm the concept that HIV-1 can persist by establishing latent infection and point to this challenge as the major hurdle in achieving a cure for HIV-1 infection.
The Mississippi baby was born to an HIV-1-infected mother who had no prenatal antiviral treatment. The infant tested positive for virus in the blood and was started on antiretroviral therapy 30 hours after delivery (2). The amount of virus in the blood (viremia) fell rapidly to below the limit of detection. At 18 months, treatment was stopped (against medical advice), but subsequent testing failed to reveal a rebound in viremia, which normally occurs 2 weeks after treatment interruption. There was widespread excitement that early therapy prevented the establishment of stable viral reservoirs in this infant. However, viremia suddenly rebounded 27 months after treatment interruption.


Silent reservoir.
HIV-1 may establish a latent infection in T cells that is unaffected by antiretroviral therapy. The virus may rebound later.
The two HIV-1-infected adults were given bone marrow transplants for other conditions while on antiretroviral therapy. The antiretroviral drugs protected donor cells from infection while host immune cells, including infected cells, were largely eliminated by chemotherapy and graft-versus-host disease (in which immune cells in the grafted tissue attack the host's immune cells) (1).
Antiretroviral therapy was eventually stopped, and initially there was no viral rebound, raising hopes for cure. However, both patients later experienced sudden and dramatic rebounds in viremia.
The major barrier to cure is a small but extremely stable pool of resting memory CD4+ T cells that carry an integrated but transcriptionally silent viral genome (3-6). Upon cellular activation, these latent proviruses are transcribed and replication-competent virus is released. Strategies for curing HIV-1 infection involve early antiretroviral therapy to prevent the establishment of a reservoir of latently infected cells, and a "shock and kill" approach (in which latency is reversed so the infected cells can be eliminated) (7). Due to the associated risks, bone marrow transplantation is only a curative strategy in patients requiring a transplant for other conditions.
The recent cases of delayed rebound suggest that even a small number of latently infected cells established in early infection or persisting after transplantation may be sufficient to rekindle viral replication. But there has been some promising news too. Established simian immunodeficiency virus infection in nonhuman primates was cleared by vaccine-induced T cell responses (8). There is also evidence that a histone deacetylase inhibitor can perturb the latent HIV-1 reservoir in humans, potentially making the infected cells vulnerable to immune clearance (9). In addition, the well-known "Berlin patient"-an HIV-1-infected adult who developed leukemia-remains virus-free after being off of antiretroviral therapy for 5 years. This patient received a bone marrow transplant from a donor who did not express C-C chemokine receptor type 5, a co-receptor for HIV-1 (and was therefore resistant to commonly transmitted forms of HIV-1) (10).
There has been lingering concern that HIV-1 persistence in patients who are on antiretroviral therapy might be due in part to low amounts of viral replication, occurring perhaps in an anatomical site with low drug concentrations (11). If persistent replication were occurring in these patients during therapy, rebound should have occurred rapidly after treatment interruption, particularly because HIV-1-specific immune responses were absent or diminished (1, 2). The finding that viremia can remain undetected with the most sensitive assays for months to years in patients off therapy before a dramatic and sudden rebound to high concentrations (>2,000,000 copies/ml in the transplant patients) can only be explained by the persistence of a latent form of HIV-1. It is not clear where the rebound virus came from in the Mississippi baby or in the two adults, but the latent reservoir in resting memory CD4+ T cells is present in every patient and persists indefinitely even in patients on suppressive antiretroviral therapy (12).
The recent cases illustrate a major challenge: measuring the latent reservoir. This reservoir was initially defined using a quantitative viral outgrowth assay (QVOA), which measures the frequency of resting CD4+ T cells that produce replication-competent virus after the T cells are activated. Typical values are ~1 per 106 resting CD4+ T cells. Measurement of HIV-1 DNA by polymerase chain reaction (which amplifies a targeted DNA sequence) gives higher frequencies for infected cells because both replication-competent and defective viral genomes are detected (13, 14). Before rebound, the QVOA was negative in the recent cases, and HIV-1 DNA was at or below the limit of detection. Thus, viral rebound occurred even though the best available assays for persistent HIV-1 were negative. The problem is not assay sensitivity. Both the QVOA and polymerase chain reaction can detect a single latently infected cell. The limitation is the size of the tissue sample tested. Even large blood or tissue samples may be devoid of such cells although they persist elsewhere in the patient.
The rebound cases also provide insight into the likely outcome of cure strategies. For example, if treatments that produce multilog reductions in the latent reservoir can be found, patients may interrupt therapy and remain aviremic for months to years, during which time assays for viral persistence would be negative (because of sample size). However, rebound could occur in a sudden and unpredictable fashion at late time points after interruption (15). Although the time to rebound will increase with reservoir reduction, enormous patient-to-patient variation in time to rebound can be expected because of the stochastic nature of the reactivation of latently infected cells. Thus, frequent monitoring of viremia over long periods of time would be needed. However, patients can only be considered safe from rebound if all latently infected cells are eliminated. It is much more likely that we will be able to achieve partial reductions, allowing a prolonged but uncertain interval off therapy-something that many patients would welcome. It is not too soon to begin planning for how to monitor this type of cure scenario.
  iconpaperstack View Older Articles   Back to Top   www.natap.org