Report of a Functional Cure in an HIV-Infected Infant
On Monday March 4th at the Conference on Retroviruses and Opportunistic Infections (CROI) in Atlanta, details were presented on a case of a potential "functional cure" of HIV infection in an infant. The story has already broken widely in the media due to a CROI press conference held on Sunday afternoon, during which the researcher describing the case, Deborah Persaud, gave a preview of the data. The webcast of Persaud's talk is now available for viewing online (it is the 7th presentation in the session entitled "Is There Hope for HIV Eradication?").
The context of the case report is unusual because the mother was not diagnosed with HIV infection until in labor, and there was no opportunity to give her antiretroviral therapy (ART) to reduce the risk of mother-to-child transmission. The infant was therefore considered at high risk of acquiring infection and HIV DNA and RNA tests were performed at 30 and 31 hours after birth, respectively. The HIV DNA test was positive, and the RNA test showed a viral load of 19,812 copies. Although current US guidelines recommend initiating a prophylactic regimen of AZT and nevirapine pending the results of virologic testing, the treating pediatrician, Hannah Gay, chose to initiate combination ART including AZT, 3TC, and nevirapine (the latter at the therapeutic rather than prophylactic dose). Nevirapine was switched in favor of Kaletra after seven days. Subsequent sequential viral-load testing showed the typical decline in response to treatment, with measurements of 2617 copies, 516 copies, 265 copies and then less than 48 copies (beneath the limit of detection).
ART was maintained for around 18 months until the mother and infant were lost to follow-up. When the infant was returned to care around five months later, Gay learned that treatment had been stopped. The reasons for the disengagement from care and cessation of treatment have not been made clear, with Gay stating in one article only that the mother was battling "some life changes."
Testing of the infant after the return to care revealed a highly unusual outcome: viral load remained undetectable despite the five-month period off ART (there have been many studies of pediatric ART interruptions, all demonstrating that viral-load rebound occurs rapidly, typically to very high levels, almost without exception). This prompted Gay to consult with external experts in pediatric HIV infection Deborah Persaud at Johns Hopkins University and Katherine Luzuriaga at Massachussetts General Hospital. The results of further investigations led to the findings presented by Persaud at CROI: several independent laboratories with expertise in searching for trace amounts of HIV evaluated samples from the infant but could only rarely detect extremely low levels of viral genetic material (there was no detectable replication-competent HIV).
These results echo those from follow-up studies of the single adult considered cured of HIV infection, Timothy Brown. Methods employed included digital droplet PCR, single-copy assays for HIV RNA, and a test for viral outgrowth from 22 million resting CD4 T cells. Immune responses to HIV, including antibody and CD4 and CD8 T-cell responses, were not detectable. Genetic studies showed that neither mother nor infant possessed the CCR5-Delta32 mutation or any HLA genes known to be associated with control of HIV replication. The totality of the findings has led the researchers to conclude that the infant represents a case of a functional cure of HIV infection (essentially an infant equivalent of Timothy Brown); the virus may not have been completely cleared, but at the current time -- after over 10 months off ART -- no viral activity is detectable.
Not too surprisingly, the media interest in the story has been intense. Several stories, including a piece in The New York Times, have included quotes from scientists suggesting the possibility that the infant was not infected in the first place. This scenario is difficult to reconcile with the stepwise decline in HIV RNA levels after ART initiation (this would have to involve not only four sequential false positive RNA results, but false positive results that also coincidentally mirrored the expected viral-load decline in response to ART -- a possibility that seems vanishingly unlikely). Another skeptical viewpoint cited by some articles is that HIV might have been cleared even in the absence of ART. This suggestion is based on occasional reports in the scientific literature of apparent transient HIV infection in exposed infants; however, a study published in 1998 that analyzed many of these cases in detail showed that most were explained by PCR testing contamination or sample mislabeling. Even the few reports that could not be fully explained by these problems were considered to be lacking the evidence necessary to formally prove that transient infection had occurred.
Some other aspects of the case deserve comment: although the baseline viral load of 19,812 copies might seem relatively low in the context of adult values, it should be noted that data on early viral loads among infected infants indicate this level is not atypical. A study published in 1998 reported that "all 18 infants defined as in utero infected by DNA PCR or coculture (or both) were also positive with a wide range of plasma HIV-1 RNA values within 48 h of birth and a median level of 26,940 HIV-1 RNA copies/mL (25th and 75th percentiles, 1556 and 468,390)."
In terms of other cases of pediatric ART interruptions in which viral-load rebound did not occur, a search of the literature did not turn up many examples. In 2006, researchers from Massachusetts General Hospital published a case report regarding a perinatally infected child who stopped ART as a teenager and maintained undetectable viral-load levels for five years of follow-up; the individual was heterozygous for the CCR5-Delta32 mutation and, unlike the infant described by Persaud, displayed strong HIV-specific T-cell responses. A 2008 survey of outcomes among children and adolescents undergoing unstructured ART interruptions notes the following: "In only 1 patient (6% [1 of 17]) did HIV-1 RNA remain undetectable after 12 months off therapy." However, no additional information is provided, and the cut-off for the viral-load assay used was <400 copies.
Based on the data presented, the Persaud case report appears unique, and the evidence that HIV infection occurred and has subsequently been controlled or even cleared after short-term ART is more compelling than some of the skeptical comments in the press have suggested. However, there is a lack of stored samples that might be able to bolster the evidence by confirming the genetic relationships between the separate virus samples from the infant and the linkage to the virus in the mother. It also remains possible that viral load could rebound at some point in the future.
The implications for pediatric HIV care and cure research are now the subject of ongoing discussions:
- Most basically, the case raises questions about the accessibility and quality of the health care system in Mississippi and the U.S. generally (as eloquently described in a posting by Jim Merrell of the Prevention Justice Alliance).
- Deborah Persaud has cited the possibility of studying triple-combination ART instead of the normally recommended dual-prophylactic regimen of AZT and nevirapine in infants with HIV-positive mothers who have not received prophylaxis, in order to assess whether similar outcomes can be obtained (the British HIV Association guidelines already recommend triple-drug therapy as postexposure prophlylaxis for infants whose mothers have detectable viral loads at delivery).
- For other perinatally infected children treated very early with ART, it has been suggested that there is now reason to carefully study whether control of HIV might be maintained after ART interruption (the reason that many ART interruption studies have already been conducted in the pediatric population is the recognition that they face the greatest burden of lifelong treatment). A related proposal is that perinatally infected children on ART should be candidates for studies of approaches that might increase the possibility of HIV control after ART withdrawal (such as therapeutic vaccines). A poster presentation at CROI by Katherine Luzuriaga describes five early-treated children in whom no replication-competent HIV can be detected, and makes the point that these individuals are "prime candidates for interventions to achieve functional cure or eradication."
- More broadly for the cure research field, the case suggests that early ART may have prevented the establishment of a long-lived latent reservoir of HIV-infected memory CD4 T cells. This suggestion fits with current efforts to deplete the latent HIV reservoir as a means to a cure.
- The question whether very early ART could lead to a similar outcome in adults might potentially be addressed by another ongoing study presented at CROI by Jintanat Ananworanich from Thailand, which involves a sizable population of individuals initiating ART at the very earliest stage of acute HIV infection (Feibig I). Ananworanich reported that the majority of individuals treated at this stage showed undetectable levels of integrated HIV DNA in both blood and colon after 24 weeks of ART; future plans include analytical treatment interruptions to assess whether HIV viral load returns.
Update 3/14/2013: On Monday March 11th in the Wall Street Journal, Dr. Mark J. Seidner from Massachussetts General Hospital and Harvard Medical School published an opinion piece arguing that the case represents an example of successful infant post-exposure prophlaxis (PEP) against HIV infection, rather than a cure of HIV infection that had occurred in utero as Persaud and colleagues have suggested. There is a published case report of an adult recipient of an HIV-infected blood transfusion who showed a single viral load value of 3 copies/mL (the lower limit of detection of the assay) but proved to be uninfected after receipt of PEP. However, there are also published studies that include assessments of viral load values in infants receiving PEP to prevent mother-to-child transmission and none appear to offer examples consistent with the idea that viral load values are detectable in this situation e.g. see this large French study from 2012, and a US analysis from 2003; the latter reports two examples of low values in infants receiving AZT prophlyaxis, but they are considered false positives because DNA PCR tests taken the same day were negative (unlike the case reported at CROI).
Richard Jefferys is the coordinator of the Michael Palm HIV Basic Science, Vaccines & Prevention Project Weblog at the Treatment Action Group (TAG). The original blog post may be viewed here.