Conference Coverage

Genetic data boost HIV surveillance efforts



– Advances in genetic sequencing are boosting efforts to identify new clusters of HIV infections and guiding public health interventions to address them. The method relies on resistance testing at diagnosis and virologic failure and allows public health researchers to determine the genetic relatedness of viruses responsible for new infections. If the viruses are genetically, geographically, and temporally associated, it indicates a previously unknown transmission cluster.

An illustration of DNA strands copyright Kativ/iStockphoto

“The presence of a cluster indicates gaps in our preventative services, which we must address to improve service delivery and stop transmission,” Alexandra M. Oster, MD, Division of HIV/AIDS Prevention, Surveillance, and Epidemiology at the Centers for Disease Control and Prevention, Atlanta, said during a talk at the Conference on Retroviruses and Opportunistic Infections.

She noted that HIV brings special challenges to outbreak detection. The median delay between infection and diagnosis is 3 years. Individuals are highly mobile, and signals of new outbreaks can be quickly drowned out in high-burden areas. But these challenges aren’t unique. Tuberculosis has a similarly lengthy latency period, yet more than 75% of new TB outbreaks are now identified through the use of genetic data. Sequencing also is used to track food-borne illness. The CDC’s PulseNet is a network of laboratories that examines DNA sequences from bacterial infections in search of previously unrecognized outbreaks.

In the HIV setting, molecular surveillance has great potential in identifying and intervening in evolving networks of HIV transmission, but also carries ethical and other challenges.

Nevertheless, “I hope to make the case that cluster detection and response [using molecular surveillance] can help bring the nation closer to ending the HIV epidemic,” said Dr. Oster.

Molecular surveillance obtains most of its data from drug resistance testing, both at entry to care and after virologic failure, which then gets passed to the U.S. National HIV Surveillance System. The data are then stripped of patient identifying information and submitted to the CDC.

With data from multiple individuals in hand, researchers create a phylogenetic tree, in which closely-related viruses appear as close neighbors on a branch. “By tracing back along the tree, you can see the inferred ancestor of [individual strains], and also the inferred ancestor of all strains on the tree,” said Dr. Oster. Together with geographical data, that information allows researchers to identify clusters of patients connected in a transmission network, and that information can be passed along to federal, state, and local agencies to prevent infections and improve care.

From 1997 through 2012, the CDC’s molecular surveillance program focused on drug resistance patterns, but in 2013 the agency decided to expand to include transmission clusters. It now uses a tool called HIV Trace, which helps public health workers with no background in bioinformatics to visualize the DNA sequences and potential clusters, though Dr. Oster cautioned against overinterpretation of the results. “The links shown can easily be misinterpreted as actual social connections,” she said.

As proof of the approach’s potential, an analysis of the clusters identified showed their potential for HIV spread. On average in the United States, four new HIV infections occur per 100 people living with HIV. In the first 13 clusters that CDC identified, the number of infections was 33 per 100 person-years. The first 60 clusters had an average of 44 transmissions per 100 person-years. “None of these clusters had been found by [standard] epidemiologic methods, demonstrating that rapid transmission can be hard to detect without molecular data,” said Dr. Oster.

In 2018, all health departments began collecting sequencing data, and almost 40% of newly diagnosed patients have had sequencing data reported, more than 340,000 patients in total. Researchers have identified 145 priority clusters.

But use of molecular data is not the only method available. The CDC monitors increases in diagnoses in specific areas and conducts time-space analyses. These more traditional methods are particularly useful in areas with small populations or low HIV burden.

With a cluster identified, public health officials can attempt to identify all of the members of the network and help them to access services, such as testing, preexposure prophylaxis (PrEP), syringe service programs, and linkage to care.

In San Antonio, Tex., an analysis identified a cluster of 24 gay and bisexual men, and further analysis revealed an extended network of 87 sexual or needle-sharing partners. Researchers also identified missed opportunities for diagnosis of acute infection as well as low access to PrEP, so the health department sent out an alert clarifying diagnosis testing guidelines, highlighting the concern over acute infection, and containing PrEP educational material.

Analysis of another network in Michigan found that all identified individuals were virally suppressed, even though the network continued to grow. That suggested that there were unidentified individuals who were contributing to transmission, which prompted efforts by providers to encourage testing, linkage to care, and prevention.

All of these developments are good news for efforts to eradicate HIV, but they come with pitfalls. Local communities have expressed concerned that molecular data could be used to identify direction of transmission and for prosecution, since there are HIV laws that criminalize lack of disclosure and potential exposure to the virus, even when transmission doesn’t occur. “These laws are not aligned with current science and have not been found to help curb HIV,” said Dr. Oster.

She noted that current molecular methods are incapable of identifying direction of transmission. Still, the CDC is reemphasizing efforts to protect public health data from nonpublic health use. “CDC and health departments implement unprecedented policies and procedures to ensure confidentiality and security of the data,” Dr. Oster said.

She reported having no relevant disclosures.

Next Article: