Discrepancies in assignment of subtype/recombinant forms by genotyping programs for HIV type 1 drug resistance testing may falsely predict superinfection

Development of an HIV-1 Subtype Panel in China: Isolation and Characterization of 30 HIV-1 Primary Strains Circulating in China

Background

The complex epidemic and significant diversity of HIV-1 strains in China pose serious challenges for surveillance and diagnostic assays, vaccine development and clinical management. There is a lack of HIV-1 isolates in current canonical HIV-1 subtype panels that can represent HIV-1 diversity in China; an HIV-1 subtype panel for China is urgently needed.

Methods

Blood samples were collected from HIV-1 infected patients participating in the drug-resistance surveillance program in China. The samples were isolated, cultured and stored as neat culture supernatant. The HIV-1 isolates were fully characterized. The panel was used to compare 2 viral load assays and 2 p24 assays as the examples of how this panel could be used.

Results

An HIV-1 subtype panel for China composed of 30 HIV-1 primary strains of four subtypes (B [including Thai-B], CRF01_AE, CRF07_BC and G) was established. The samples were isolated and cultured to a high-titer (10⁶-10⁹ copies/ml)/high-volume (40ml). The HIV-1 isolates were fully characterized by the final viral load, p24 concentration, gag-pol and envC2V3 sequencing, co-receptor prediction, determination of the four amino acids at the tip of the env V3-loop, glycosylation sites in the V3 loop and the drug-resistance mutations. The comparison of two p24 assays and two viral load assays on the isolates illustrated how this panel may be used for the evaluation of diagnostic assay performance. The Pearson value between p24 assays were 0.938. The viral load results showed excellent concordance and agreement for samples of Thai-B, but lower correlations for samples of CRF01_AE.

Conclusion

The current panel of 30 HIV-1 isolates served as a basis for the development of a comprehensive panel of fully characterized viral isolates, which could reflect the current dynamic and complex HIV-1 epidemic in China. This panel will be available to support HIV-1 research, assay evaluation, vaccine and drug development.

HIV diversity and drug resistance from plasma and non-plasma analytes in a large treatment programme in western Kenya

INTRODUCTION

Antiretroviral resistance leads to treatment failure and resistance transmission. Resistance data in western Kenya are limited. Collection of non-plasma analytes may provide additional resistance information.

METHODS

We assessed HIV diversity using the REGA tool, transmitted resistance by the WHO mutation list and acquired resistance upon first-line failure by the IAS-USA mutation list, at the Academic Model Providing Access to Healthcare (AMPATH), a major treatment programme in western Kenya. Plasma and four non-plasma analytes, dried blood-spots (DBS), dried plasma-spots (DPS), ViveST(TM)-plasma (STP) and ViveST-blood (STB), were compared to identify diversity and evaluate sequence concordance.

RESULTS

Among 122 patients, 62 were treatment-naïve and 60 treatment-experienced; 61% were female, median age 35 years, median CD4 182 cells/µL, median viral-load 4.6 log10 copies/mL. One hundred and ninety-six sequences were available for 107/122 (88%) patients, 58/62 (94%) treatment-naïve and 49/60 (82%) treated; 100/122 (82%) plasma, 37/78 (47%) attempted DBS, 16/45 (36%) attempted DPS, 14/44 (32%) attempted STP from fresh plasma and 23/34 (68%) from frozen plasma, and 5/42 (12%) attempted STB. Plasma and DBS genotyping success increased at higher VL and shorter shipment-to-genotyping time. Main subtypes were A (62%), D (15%) and C (6%). Transmitted resistance was found in 1.8% of plasma sequences, and 7% combining analytes. Plasma resistance mutations were identified in 91% of treated patients, 76% NRTI, 91% NNRTI; 76% dual-class; 60% with intermediate-high predicted resistance to future treatment options; with novel mutation co-occurrence patterns. Nearly 88% of plasma mutations were identified in DBS, 89% in DPS and 94% in STP. Of 23 discordant mutations, 92% in plasma and 60% in non-plasma analytes were mixtures. Mean whole-sequence discordance from frozen plasma reference was 1.1% for plasma-DBS, 1.2% plasma-DPS, 2.0% plasma-STP and 2.3% plasma-STB. Of 23 plasma-STP discordances, one mutation was identified in plasma and 22 in STP (p<0.05). Discordance was inversely significantly related to VL for DBS.

CONCLUSIONS

In a large treatment programme in western Kenya, we report high HIV-1 subtype diversity; low plasma transmitted resistance, increasing when multiple analytes were combined; and high-acquired resistance with unique mutation patterns. Resistance surveillance may be augmented by using non-plasma analytes for lower-cost genotyping in resource-limited settings.

HIV-1 subtypes B and C unique recombinant forms (URFs) and transmitted drug resistance identified in the Western Cape Province, South Africa

South Africa has the largest worldwide HIV/AIDS population with 5.6 million people infected and at least 2 million people on antiretroviral therapy. The majority of these infections are caused by HIV-1 subtype C. Using genotyping methods we characterized HIV-1 subtypes of the gag p24 and pol PR and RT fragments, from a cohort of female participants in the Western Cape Province, South Africa. These participants were recruited as part of a study to assess the combined brain and behavioural effects of HIV and early childhood trauma. The partial HIV-1 gag and pol fragments of 84 participants were amplified by PCR and sequenced. Different online tools and manual phylogenetic analysis were used for HIV-1 subtyping. Online tools included: REGA HIV Subtyping tool version 3; Recombinant Identification Program (RIP); Context-based Modeling for Expeditious Typing (COMET); jumping profile Hidden Markov Models (jpHMM) webserver; and subtype classification using evolutionary algorithms (SCUEAL). HIV-1 subtype C predominates within the cohort with a prevalence of 93.8%. We also show, for the first time, the presence of circulating BC strains in at least 4.6% of our study cohort. In addition, we detected transmitted resistance associated mutations in 4.6% of analysed sequences. With tourism and migration rates to South Africa currently very high, we are detecting more and more HIV-1 URFs within our study populations. It is stil unclear what role these unique strains will play in terms of long term antiretroviral treatment and what challenges they will pose to vaccine development. Nevertheless, it remains vitally important to monitor the HIV-1 diversity in South Africa and worldwide as the face of the epidemic is continually changing.

Automated subtyping of HIV-1 genetic sequences for clinical and surveillance purposes: performance evaluation of the new REGA version 3 and seven other tools

BACKGROUND

To investigate differences in pathogenesis, diagnosis and resistance pathways between HIV-1 subtypes, an accurate subtyping tool for large datasets is needed. We aimed to evaluate the performance of automated subtyping tools to classify the different subtypes and circulating recombinant forms using pol, the most sequenced region in clinical practice. We also present the upgraded version 3 of the Rega HIV subtyping tool (REGAv3).

METHODOLOGY

HIV-1 pol sequences (PR+RT) for 4674 patients retrieved from the Portuguese HIV Drug Resistance Database, and 1872 pol sequences trimmed from full-length genomes retrieved from the Los Alamos database were classified with statistical-based tools such as COMET, jpHMM and STAR; similarity-based tools such as NCBI and Stanford; and phylogenetic-based tools such as REGA version 2 (REGAv2), REGAv3, and SCUEAL. The performance of these tools, for pol, and for PR and RT separately, was compared in terms of reproducibility, sensitivity and specificity with respect to the gold standard which was manual phylogenetic analysis of the pol region.

RESULTS

The sensitivity and specificity for subtypes B and C was more than 96% for seven tools, but was variable for other subtypes such as A, D, F and G. With regard to the most common circulating recombinant forms (CRFs), the sensitivity and specificity for CRF01_AE was ~99% with statistical-based tools, with phylogenetic-based tools and with Stanford, one of the similarity based tools. CRF02_AG was correctly identified for more than 96% by COMET, REGAv3, Stanford and STAR. All the tools reached a specificity of more than 97% for most of the subtypes and the two main CRFs (CRF01_AE and CRF02_AG). Other CRFs were identified only by COMET, REGAv2, REGAv3, and SCUEAL and with variable sensitivity. When analyzing sequences for PR and RT separately, the performance for PR was generally lower and variable between the tools. Similarity and statistical-based tools were 100% reproducible, but this was lower for phylogenetic-based tools such as REGA (~99%) and SCUEAL (~96%).

CONCLUSIONS

REGAv3 had an improved performance for subtype B and CRF02_AG compared to REGAv2 and is now able to also identify all epidemiologically relevant CRFs. In general the best performing tools, in alphabetical order, were COMET, jpHMM, REGAv3, and SCUEAL when analyzing pure subtypes in the pol region, and COMET and REGAv3 when analyzing most of the CRFs. Based on this study, we recommend to confirm subtyping with 2 well performing tools, and be cautious with the interpretation of short sequences.

Pilot studies for development of an HIV subtype panel for surveillance of global diversity

The continued global spread and evolution of HIV diversity pose significant challenges to diagnostics and vaccine strategies. NIAID partnered with the FDA, WRAIR, academia, and industry to form a Viral Panel Working Group to design and prepare a panel of well-characterized current and diverse HIV isolates. Plasma samples that had screened positive for HIV infection and had evidence of recently acquired infection were donated by blood centers in North and South America, Europe, and Africa. A total of 80 plasma samples were tested by quantitative nucleic acid tests, p24 antigen, EIA, and Western blot to assign a Fiebig stage indicative of approximate time from initial infection. Evaluation of viral load using FDA-cleared assays showed excellent concordance when subtype B virus was tested, but lower correlations for subtype C. Plasma samples were cocultivated with phytohemagglutinin (PHA)-stimulated peripheral blood mononuclear cells (PBMCs) from normal donors to generate 30 viral isolates (50-80% success rate for samples with viral load >10,000 copies/ml), which were then expanded to 10(7)-10(9) virus copies per ml. Analysis of env sequences showed that sequences derived from cultured PBMCs were not distinguishable from those obtained from the original plasma. The pilot collection includes 30 isolates representing subtypes B, C, B/F, CRF04_cpx, and CRF02_AG. These studies will serve as a basis for the development of a comprehensive panel of highly characterized viral isolates that reflects the current dynamic and complex HIV epidemic, and will be made available through the External Quality Assurance Program Oversight Laboratory (EQAPOL).

More accurate recombination prediction in HIV-1 using a robust decoding algorithm for HMMs

Background

Identifying recombinations in HIV is important for studying the epidemiology of the virus and aids in the design of potential vaccines and treatments. The previous widely-used tool for this task uses the Viterbi algorithm in a hidden Markov model to model recombinant sequences.

Results

We apply a new decoding algorithm for this HMM that improves prediction accuracy. Exactly locating breakpoints is usually impossible, since different subtypes are highly conserved in some sequence regions. Our algorithm identifies these sites up to a certain error tolerance. Our new algorithm is more accurate in predicting the location of recombination breakpoints. Our implementation of the algorithm is available at http://www.cs.uwaterloo.ca/~jmtruszk/jphmm_balls.tar.gz.

Conclusions

By explicitly accounting for uncertainty in breakpoint positions, our algorithm offers more reliable predictions of recombination breakpoints in HIV-1. We also document a new domain of use for our new decoding approach in HMMs.

An evolutionary model-based algorithm for accurate phylogenetic breakpoint mapping and subtype prediction in HIV-1

Genetically diverse pathogens (such as Human Immunodeficiency virus type 1, HIV-1) are frequently stratified into phylogenetically or immunologically defined subtypes for classification purposes. Computational identification of such subtypes is helpful in surveillance, epidemiological analysis and detection of novel variants, e.g., circulating recombinant forms in HIV-1. A number of conceptually and technically different techniques have been proposed for determining the subtype of a query sequence, but there is not a universally optimal approach. We present a model-based phylogenetic method for automatically subtyping an HIV-1 (or other viral or bacterial) sequence, mapping the location of breakpoints and assigning parental sequences in recombinant strains as well as computing confidence levels for the inferred quantities. Our Subtype Classification Using Evolutionary ALgorithms (SCUEAL) procedure is shown to perform very well in a variety of simulation scenarios, runs in parallel when multiple sequences are being screened, and matches or exceeds the performance of existing approaches on typical empirical cases. We applied SCUEAL to all available polymerase (pol) sequences from two large databases, the Stanford Drug Resistance database and the UK HIV Drug Resistance Database. Comparing with subtypes which had previously been assigned revealed that a minor but substantial (≈5%) fraction of pure subtype sequences may in fact be within- or inter-subtype recombinants. A free implementation of SCUEAL is provided as a module for the HyPhy package and the Datamonkey web server. Our method is especially useful when an accurate automatic classification of an unknown strain is desired, and is positioned to complement and extend faster but less accurate methods. Given the increasingly frequent use of HIV subtype information in studies focusing on the effect of subtype on treatment, clinical outcome, pathogenicity and vaccine design, the importance of accurate, robust and extensible subtyping procedures is clear.

There are nine different subtypes of the main group of HIV-1, each originating as a distinct subepidemic of HIV-1. The distribution of subtypes is often unique to a given geographic region of the world and constitutes a useful epidemiological and surveillance resource. The effects of viral subtype on disease progression, treatment outcome and vaccine design are being actively researched, and the importance of accurate subtyping procedures is clear. In HIV-1, subtype assignment is complicated by frequent recombination among co-circulating strains, creating new genetic mosaics or recombinant forms: 43 have been characterized to date, and many more likely exist. We present an automated phylogenetic method (SCUEAL) to accurately characterize both simple and complex HIV-1 mosaics. Using computer simulations and biological data we demonstrate that SCUEAL performs very well under various conditions, especially when some of the existing classification procedures fail. Furthermore, we show that a small, but noticeable proportion of subtype characterization stored in public databases may be incomplete or incorrect. The computational technique introduced here should provide a much more accurate characterization of HIV-1 strains, especially novel recombinants, and lead to new insights into molecular history, epidemiology and geographical distribution of the virus.