Integration of Contact Tracing and Phylogenetics in an Investigation of Acute HIV Infection

Abstract

Background: The integration of traditional contact tracing with HIV sequence analyses offers opportunities to mitigate some of the barriers to effective network construction. We used combined analyses during an outbreak investigation of spatiotemporally clustered acute HIV infections to evaluate if the observed clustering was the product of a single outbreak.

Methods: We investigated acute and recent HIV index cases reported in North Carolina from 2013 to 2014 and their reported contacts. Contact tracing networks were constructed with surveillance data and compared with phylogenetic transmission clusters involving an index case using available HIV-1 pol sequences including 1672 references. Clusters were defined as clades of 2 or more sequences with a less than 1.5% genetic distance and a bootstrap of at least 98% on maximum-likelihood phylogenies.

Results: In total, 68 index cases and 210 contacts (71 HIV infected) were reported. The contact tracing network involved 58 components with low overall density (1.2% statewide); 33% of first-degree contacts could not be located. Among 38 (56%) of 68 index cases and 34 (48%) of 71 contacts with sequences, 13 phylogenetic clusters were identified (size 2-4 members). Four clusters connected network components that were not linked in contact tracing. The largest component (n = 28 cases) included 2 distinct phylogenetic clusters and spanned 2 regions.

Conclusions: We identified the concurrent expansion of multiple small transmission clusters rather than a single outbreak in a largely disconnected contact tracing network. Integration of phylogenetic analyses provided timely information on transmission networks during the investigation. Our findings highlight the potential of combined methods to better identify high-risk networks for intervention.

Figure 1

A. Trends in acute HIV diagnoses in North Carolina reported in Regions 1 (Black Mountain) and 2 (Charlotte) from 2012–2014. An outbreak investigation was triggered after the observed spatiotemporal spike in acute HIV diagnoses from December 2013 – February 2014. B. Map of North Carolina regions for HIV field services. Locations of acute HIV infections by county are indicated by red circles. As reference, the estimated 2012 newly diagnosed HIV infection rates per 100,000 population for North Carolina Regional Networks of Care and Prevention are provided in parentheses after each HIV Field Service Region. Regions with multiple values include several Regional Care Networks

Figure 2

Contact tracing network components stemming from index cases diagnosed during acute HIV infection (n=32) and their reported partners (n=118 non-index cases) investigated in two regions in North Carolina, 2012–2014. Components are labeled with black numbers and index cases nodes are large while contacts nodes are small. HIV-infected cases and contacts have red labels and those with available sequences are indicated with red outlines. Dashed lines depict cases linked in phylogenetic clusters and are labeled with cluster names (A–G). Panel A. Region 1 (Black Mountain) involves 4 components, including components 3 and 2 which are linked by phylogenetic cluster (B) but not by contact tracing. Panel B. Region 2 (Charlotte), where most index cases were found in small components not linked in clusters. One large component (8) involves two distinct phylogenetic clusters (C & D) and multiple index cases not in clusters. Clusters F and G phylogenetically link cases that are not connected in contact trancing components.