A new pattern-based method for identifying recent HIV-1 infections from the viral env sequence

Statistical considerations for cross-sectional HIV incidence estimation based on recency test

Longitudinal cohorts to determine the incidence of HIV infection are logistically challenging, so researchers have sought alternative strategies. Recency test methods use biomarker profiles of HIV-infected subjects in a cross-sectional sample to infer whether they are "recently" infected and to estimate incidence in the population. Two main estimators have been used in practice: one that assumes a recency test is perfectly specific, and another that allows for false-recent results. To date, these commonly used estimators have not been rigorously studied with respect to their assumptions and statistical properties. In this article, we present a theoretical framework with which to understand these estimators and interrogate their assumptions, and perform a simulation study and data analysis to assess the performance of these estimators under realistic HIV epidemiological dynamics. We find that the snapshot estimator and the adjusted estimator perform well when their corresponding assumptions hold. When assumptions on constant incidence and recency test characteristics fail to hold, the adjusted estimator is more robust than the snapshot estimator. We conclude with recommendations for the use of these estimators in practice and a discussion of future methodological developments to improve HIV incidence estimation via recency test.

Modeling methods for estimating HIV incidence: a mathematical review

Estimating HIV incidence is crucial for monitoring the epidemiology of this infection, planning screening and intervention campaigns, and evaluating the effectiveness of control measures. However, owing to the long and variable period from HIV infection to the development of AIDS and the introduction of highly active antiretroviral therapy, accurate incidence estimation remains a major challenge. Numerous estimation methods have been proposed in epidemiological modeling studies, and here we review commonly-used methods for estimation of HIV incidence. We review the essential data required for estimation along with the advantages and disadvantages, mathematical structures and likelihood derivations of these methods. The methods include the classical back-calculation method, the method based on CD4+ T-cell depletion, the use of HIV case reporting data, the use of cohort study data, the use of serial or cross-sectional prevalence data, and biomarker approach. By outlining the mechanistic features of each method, we provide guidance for planning incidence estimation efforts, which may depend on national or regional factors as well as the availability of epidemiological or laboratory datasets.

HIV-1 envelope sequence-based diversity measures for identifying recent infections

Identifying recent HIV-1 infections is crucial for monitoring HIV-1 incidence and optimizing public health prevention efforts. To identify recent HIV-1 infections, we evaluated and compared the performance of 4 sequence-based diversity measures including percent diversity, percent complexity, Shannon entropy and number of haplotypes targeting 13 genetic segments within the env gene of HIV-1. A total of 597 diagnostic samples obtained in 2013 and 2015 from recently and chronically HIV-1 infected individuals were selected. From the selected samples, 249 (134 from recent versus 115 from chronic infections) env coding regions, including V1-C5 of gp120 and the gp41 ectodomain of HIV-1, were successfully amplified and sequenced by next generation sequencing (NGS) using the Illumina MiSeq platform. The ability of the four sequence-based diversity measures to correctly identify recent HIV infections was evaluated using the frequency distribution curves, median and interquartile range and area under the curve (AUC) of the receiver operating characteristic (ROC). Comparing the median and interquartile range and evaluating the frequency distribution curves associated with the 4 sequence-based diversity measures, we observed that the percent diversity, number of haplotypes and Shannon entropy demonstrated significant potential to discriminate recent from chronic infections (p<0.0001). Using the AUC of ROC analysis, only the Shannon entropy measure within three HIV-1 env segments could accurately identify recent infections at a satisfactory level. The env segments were gp120 C2_1 (AUC = 0.806), gp120 C2_3 (AUC = 0.805) and gp120 V3 (AUC = 0.812). Our results clearly indicate that the Shannon entropy measure represents a useful tool for predicting HIV-1 infection recency.

Strengthening HIV surveillance: measurements to track the epidemic in real time

Surveillance for HIV as a public health initiative requires timely, detailed and robust data to systematically understand burden of infection, transmission patterns, direct prevention efforts, guide funding, identify new infections and predict future trends in the epidemic. The methods for HIV surveillance have evolved to reliably track the epidemic and identify new infections in real time. Initially HIV surveillance relied primarily on the reporting of AIDS cases followed by measuring antibodies to HIV to determine prevalence in key populations. With the roll-out of antiretroviral therapy (ART) resulting in better survival and the corresponding increase in HIV prevalence, the landscape of surveillance shifted further to track HIV prevalence and incidence within the context of programmes. Recent developments in laboratory assays that potentially measure and differentiate recent versus established HIV infection offer a cost-effective method for the rapid estimation of HIV incidence. These tests continue to be validated and are increasingly useful in informing the status of the epidemic in real time. Surveillance of heterogeneity of infections contributing to sub-epidemics requires methods to identify affected populations, density, key geographical locations and phylogenetically linked or clustered infections. Such methods could provide a nuanced understanding of the epidemic and prioritise prevention efforts to those most vulnerable. This paper brings together recent developments and challenges facing HIV surveillance, together with the application of newer assays and methods to fast-track the HIV prevention and treatment response.

Identifying Recent HIV Infections: From Serological Assays to Genomics

In this paper, we review serological and molecular based methods to identify HIV infection recency. The accurate identification of recent HIV infection continues to be an important research area and has implications for HIV prevention and treatment interventions. Longitudinal cohorts that follow HIV negative individuals over time are the current gold standard approach, but they are logistically challenging, time consuming and an expensive enterprise. Methods that utilize cross-sectional testing and biomarker information have become an affordable alternative to the longitudinal approach. These methods use well-characterized biological makers to differentiate between recent and established HIV infections. However, recent results have identified a number of limitations in serological based assays that are sensitive to the variability in immune responses modulated by HIV subtypes, viral load and antiretroviral therapy. Molecular methods that explore the dynamics between the timing of infection and viral evolution are now emerging as a promising approach. The combination of serological and molecular methods may provide a good solution to identify recent HIV infection in cross-sectional data. As part of this review, we present the advantages and limitations of serological and molecular based methods and their potential complementary role for the identification of HIV infection recency.

Assessing Biases in the Evaluation of Classification Assays for HIV Infection Recency

Identifying recent HIV infection cases has important public health and clinical implications. It is essential for estimating incidence rates to monitor epidemic trends and evaluate the effectiveness of interventions. Detecting recent cases is also important for HIV prevention given the crucial role that recently infected individuals play in disease transmission, and because early treatment onset can improve the clinical outlook of patients while reducing transmission risk. Critical to this enterprise is the development and proper assessment of accurate classification assays that, based on cross-sectional samples of viral sequences, help determine infection recency status. In this work we assess some of the biases present in the evaluation of HIV recency classification algorithms that rely on measures of within-host viral diversity. Particularly, we examine how the time since infection (TSI) distribution of the infected subjects from which viral samples are drawn affect performance metrics (e.g., area under the ROC curve, sensitivity, specificity, accuracy and precision), potentially leading to misguided conclusions about the efficacy of classification assays. By comparing the performance of a given HIV recency assay using six different TSI distributions (four simulated TSI distributions representing different epidemic scenarios, and two empirical TSI distributions), we show that conclusions about the overall efficacy of the assay depend critically on properties of the TSI distribution. Moreover, we demonstrate that an assay with high overall classification accuracy, mainly due to properly sorting members of the well-represented groups in the validation dataset, can still perform notoriously poorly when sorting members of the less represented groups. This is an inherent issue of classification and diagnostics procedures that is often underappreciated. Thus, this work underscores the importance of acknowledging and properly addressing evaluation biases when proposing new HIV recency assays.

A comparison of two measures of HIV diversity in multi-assay algorithms for HIV incidence estimation

BACKGROUND

Multi-assay algorithms (MAAs) can be used to estimate HIV incidence in cross-sectional surveys. We compared the performance of two MAAs that use HIV diversity as one of four biomarkers for analysis of HIV incidence.

METHODS

Both MAAs included two serologic assays (LAg-Avidity assay and BioRad-Avidity assay), HIV viral load, and an HIV diversity assay. HIV diversity was quantified using either a high resolution melting (HRM) diversity assay that does not require HIV sequencing (HRM score for a 239 base pair env region) or sequence ambiguity (the percentage of ambiguous bases in a 1,302 base pair pol region). Samples were classified as MAA positive (likely from individuals with recent HIV infection) if they met the criteria for all of the assays in the MAA. The following performance characteristics were assessed: (1) the proportion of samples classified as MAA positive as a function of duration of infection, (2) the mean window period, (3) the shadow (the time period before sample collection that is being assessed by the MAA), and (4) the accuracy of cross-sectional incidence estimates for three cohort studies.

RESULTS

The proportion of samples classified as MAA positive as a function of duration of infection was nearly identical for the two MAAs. The mean window period was 141 days for the HRM-based MAA and 131 days for the sequence ambiguity-based MAA. The shadows for both MAAs were <1 year. Both MAAs provided cross-sectional HIV incidence estimates that were very similar to longitudinal incidence estimates based on HIV seroconversion.

CONCLUSIONS

MAAs that include the LAg-Avidity assay, the BioRad-Avidity assay, HIV viral load, and HIV diversity can provide accurate HIV incidence estimates. Sequence ambiguity measures obtained using a commercially-available HIV genotyping system can be used as an alternative to HRM scores in MAAs for cross-sectional HIV incidence estimation.

High-accuracy identification of incident HIV-1 infections using a sequence clustering based diversity measure

Accurate estimates of HIV-1 incidence are essential for monitoring epidemic trends and evaluating intervention efforts. However, the long asymptomatic stage of HIV-1 infection makes it difficult to effectively distinguish incident infections from chronic ones. Current incidence assays based on serology or viral sequence diversity are both still lacking in accuracy. In the present work, a sequence clustering based diversity (SCBD) assay was devised by utilizing the fact that viral sequences derived from each transmitted/founder (T/F) strain tend to cluster together at early stage, and that only the intra-cluster diversity is correlated with the time since HIV-1 infection. The dot-matrix pairwise alignment was used to eliminate the disproportional impact of insertion/deletions (indels) and recombination events, and so was the proportion of clusterable sequences (P_c) as an index to identify late chronic infections with declined viral genetic diversity. Tested on a dataset containing 398 incident and 163 chronic infection cases collected from the Los Alamos HIV database (last modified 2/8/2012), our SCBD method achieved 99.5% sensitivity and 98.8% specificity, with an overall accuracy of 99.3%. Further analysis and evaluation also suggested its performance was not affected by host factors such as the viral subtypes and transmission routes. The SCBD method demonstrated the potential of sequencing based techniques to become useful for identifying incident infections. Its use may be most advantageous for settings with low to moderate incidence relative to available resources. The online service is available at http://www.bioinfo.tsinghua.edu.cn:8080/SCBD/index.jsp.

HIV diversity as a biomarker for HIV incidence estimation: including a high-resolution melting diversity assay in a multiassay algorithm

Multiassay algorithms (MAAs) can be used to estimate cross-sectional HIV incidence. We previously identified a robust MAA that includes the BED capture enzyme immunoassay (BED-CEIA), the Bio-Rad Avidity assay, viral load, and CD4 cell count. In this report, we evaluated MAAs that include a high-resolution melting (HRM) diversity assay that does not require sequencing. HRM scores were determined for eight regions of the HIV genome (2 in gag, 1 in pol, and 5 in env). The MAAs that were evaluated included the BED-CEIA, the Bio-Rad Avidity assay, viral load, and the HRM diversity assay, using HRM scores from different regions and a range of region-specific HRM diversity assay cutoffs. The performance characteristics based on the proportion of samples that were classified as MAA positive by duration of infection were determined for each MAA, including the mean window period. The cross-sectional incidence estimates obtained using optimized MAAs were compared to longitudinal incidence estimates for three cohorts in the United States. The performance of the HRM-based MAA was nearly identical to that of the MAA that included CD4 cell count. The HRM-based MAA had a mean window period of 154 days and provided cross-sectional incidence estimates that were similar to those based on cohort follow-up. HIV diversity is a useful biomarker for estimating HIV incidence. MAAs that include the HRM diversity assay can provide accurate HIV incidence estimates using stored blood plasma or serum samples without a requirement for CD4 cell count data.

Future of phylogeny in HIV prevention

The success of the HIV Prevention Trials Network 052 trial has led to revisions in HIV-1 treatment guidelines. Antiretroviral therapy may reduce the risk of HIV-1 transmissions at the population level. The design of successful treatment as prevention interventions will be predicated on a comprehensive understanding of the spatial, temporal, and biological dynamics of heterosexual men who have sex with men and intravenous drug user epidemics. Viral phylogenetics can capture the underlying structure of transmission networks based on the genetic interrelatedness of viral sequences and cluster networks that could not be otherwise identified. This article describes the phylogenetic expansion of the Montreal men who have sex with men epidemic over the last decade. High rates of coclustering of primary infections are associated with 1 infection leading to 13 onward transmissions. Phylogeny substantiates the role of primary and recent stage infection in transmission dynamics, underlying the importance of timely diagnosis and immediate antiretroviral therapy initiation to avert transmission cascades.