Home > News & Stories > Article
  • Article - Pinpointing HIV spread in Africa poses risks
  • Blue Ocean l 17-06-12 00:00:00 l View : 529
  • Science Magazine May 2017

    Pinpointing HIV spread in Africa poses risks

    Researchers hash out ethical, legal issues raised by large sequence database of AIDS virus

    By Jon Cohen


     In 1999, prosecutors in Scotland served evolutionary geneticist Andrew Leigh Brown with a warrant. Brown, who is based at the University of Edinburgh, was documenting a cluster of people who had been infected with HIV through heterosexual contact. The prosecutors wanted one man’s viral DNA sequence. “This was confidential information,” says Brown, who had no choice but to comply. The sequence turned out to be a close match to HIV from a woman the man had had sex with. The genetic match, together with the man’s own testimony that he had not told the woman he was infected, led to a 5-year prison sentence for “reckless conduct.”

     Increasingly, researchers hoping to slow HIV’s spread are collecting sequences of the highly mutable virus from multiple people to obtain a finer view of transmission patterns between individuals. But that raises complex legal and ethical issues (Science, 12 June 2015, p. 1188). On 20–21 May, scientists, bioethicists, and lawyers will gather in London to discuss how those quandaries might play out in sub-Saharan Africa, where a consortium called Phylogenetics and Networks for Generalised HIV Epidemics in Africa (PANGEA HIV) is sequencing viral DNA from thousands of people. “It raises a whole range of questions that hitherto have been limited to the resource-rich world,” says virologist Deenan Pillay, principal investigator of PANGEA HIV, which is sponsoring the meeting.

     PANGEA HIV aims to understand how the virus is moving through clusters of people so that health workers can better target prevention and treatment resources. Since 2013, it has obtained blood samples from 20 different clinical studies in Africa and sequenced viral DNA from more than 12,500 people. By organizing the sequences into family trees, or phylogenies, researchers can identify individuals who likely have infected several others, geographic areas where the virus is spreading, or people failing on treatment who are still highly infectious.

     But Pillay, who is head of the Africa Health Research Institute in Durban and Somkhele, South Africa, worries that what happened to Brown in Scotland could occur in sub-Saharan Africa. In addition to linking transmission between individuals, phylogenies can reveal other sensitive information such as groups of men who are having sex with each other, or drug users sharing needles. Homosexuality and drug use both can lead to prison sentences or blackmail in many sub-Saharan African countries. Pillay is especially concerned that fears about improper use of their data could scare people away from being tested for HIV.

     Anne-Mieke Vandamme, a virologist at the University of Leuven in Belgium and an expert witness for both the prosecution and defense in two dozen court cases, worries about wrongful convictions. “There’s this false confidence that you can actually identify the person who infected another person” from sequence matches, she says. When two people have closely matched HIVs, she notes, a third person could have infected either or both. In court cases phylogenetics can exonerate someone, but to convict, “everything depends on testimonies,” she says.

     Advances in genetic sequencing and statistical modeling could make phylogenetics more revealing, however. “The entire ethical discussion on phylogenetics and HIV has gone from zero to 60 in a short amount of time,” says epidemiologist Oliver Laeyendecker of Johns Hopkins Bloomberg School of Public Health in Baltimore, Maryland. Researchers traditionally have sequenced maybe 10% of one HIV isolated from an individual. (The virus makes mistakes when it copies itself, creating a “swarm” of mutants in each infected person.) Because costs have fallen and the technology has improved, they now can produce more detailed family trees that include the complete viral genome of several HIVs from each person.

     Thomas Leitner, a computational biologist at the Los Alamos National Laboratory in New Mexico, and his colleagues say their new statistical model can use the improved genetic data to determine who infected whom with high confidence. The model, which they described in the 8 March 2016 issue of the Proceedings of the National Academy of Sciences, analyzes at least 20 HIVs from each infected person. Skeptics point out that it has yet to be independently verified. But Leitner, who wants to use the information for public health aims, is confident. “People have been saying it’s impossible to infer directionality and here I show you can do that and that’s shocking to them,” Leitner says. 

     The PANGEA HIV consortium has been “extremely careful” to protect anonymity, Brown says, and employs a far more sophisticated scheme than the one Scottish prosecutors cracked to identify his trial participants. PANGEA HIV samples arrive with an ID code that is repeatedly changed before the sequence appears in the project’s database. What’s more, that database is housed at the Farr Institute of Health Informatics Research in London, a secure site that also contains patient records for the United Kingdom’s National Health Service.

     At the London meeting, researchers plan to discuss whether to include clinical information along with the samples in the database. Doing so could, in principle, make it easier to identify an infected individual if someone had access to original patient records. “My main concern is deductive disclosure,” says Anne Hoppe, a virologist at University College London who serves as PANGEA HIV’s project manager.

     The power of viral sequencing to reveal private details also raises questions about informed consent. “It’s a complete ethical minefield,” says Janet Seeley, a social anthropologist at the London School of Hygiene & Tropical Medicine. All PANGEA HIV participants signed consent forms that allow future research to be done with their blood samples. But Seeley, who studies social aspects of HIV/AIDS in fishing villages in Uganda, questions how much participants understand. “If one is thinking about anonymized phylogenetic trees, what does that mean to people?” she asks

     Edwin Bernard, head of the HIV Justice Network in Brighton, U.K., says however beneficial phylogenetic studies may be, researchers should keep in mind the potential cost to subjects in sub-Saharan Africa, where legal systems are often shaky. “I want people working in public health to understand that they are working in a punitive legal environment where stigma continues to exist,” Bernard says. “We all want to end the epidemic, but they need to be extra careful about how the information is used.”