Abstract
In resource-constrained settings, antiretroviral treatment (ART) is often continued based on clinical and CD4 responses, without virologic monitoring. ART with incomplete viral suppression was assessed in 27 subjects with subtype C HIV-1 by measuring plasma HIV-1 RNA, drug resistance, viral tropism, and evolution in polymerase (pol) and envelope (env) genes. The association between these viral parameters and CD4 cell change over time was analyzed using linear regression models. Increased area under the curve of HIV-1 RNA replication was a predictor of lower CD4 cell gains (p <0.007), while less drug resistance measured as a genotypic susceptibility score (GSS) (p=0.065), and lower rates of evolution in pol and env genes (p= 0.08 and 0.097, respectively) measured as genetic distance were modestly associated with increasing CD4 cell counts. Evolution of pol and env were correlated (R2 = 0.48, p=0.005), however, greater evolution was identified in env vs. pol (p <0.05). CXCR4-usage (X4) was detected in 14/27 (52%) but no differences in CD4 cell change or plasma viremia were associated with X4-usage. Among subtype C HIV-1 infected patients in Zimbabwe receiving incompletely suppressive ART, higher virus replication and lower CD4 cell gains were associated with drug resistance and evolution of polymerase and envelope.
Keywords: Viral evolution, tropism, CD4 response, subtype C HIV, antiretroviral therapy
Introduction
In resource-constrained settings, clinical management of antiretroviral treatment (ART) is often based on drug tolerance, clinical stability and increasing CD4 counts, without virologic monitoring. However, continued ART with detectable viremia and active viral replication may lead to the selection of drug resistance and promote viral evolution, with opportunity for a switch in co-receptor usage independent of CD4 response.
Viral evolution can be estimated by evaluating sequential viral sequences from the same individual. A change in viral tropism can also indicate ongoing viral evolution. R5 viruses are frequently identified in genital secretions and in recent infection 1, 2 while X4 viruses may appear later, associated with disease progression.3 However, up to 17% of recently infected patients may have X4 virus.4 Switching from R5 to X4 or D/M virus is observed in untreated individuals, associated with declining CD4 count, possibly as a result of diminished immune control or increased pathogenicity of X4 viruses.5-9 Rapid progression to AIDS in cases of primary infection with X4 virus support the latter hypothesis.2, 3, 10 Various techniques can be used to distinguish co-receptor usage. X4 viruses are associated with syncytia induction (SI) in culture, and may be estimated by the presence of basic amino acids at position 11 and 25 of the V3 loop,11-13 although V-3 loop net charge, N-linked V1 and V2 glycosylation sites, and env length polymorphisms have also been implicated. 14, 15
The relationship between a CD4 response to treatment, selection of drug resistance and changes in env tropism is not clear.16 In the setting of ART, X4 virus is more prevalent among patients with lower CD4 cells and drug resistance.17, 18 Drug resistance mutations may modulate HIV replication and the rate of CD4 decline. In primary infection, individuals with HIV RNA levels > 100,000 copies/ mL are more likely to have wild type viruses.19 Treatment experienced individuals with drug resistance mutations have lower HIV RNA compared to antiretroviral naïve individuals.20 This may be explained by decreased replication and fitness of viruses with drug resistance mutations.21, 22
Subtype C HIV-1 accounts for the majority of global infections.23, 24 However, drug resistance, tropism, and ART have not been extensively studied in subtype C viruses. We previously described a high frequency of X4 viruses among a cohort of treatment experienced subtype C HIV-1 infected individuals in Zimbabwe, with multi-drug resistance and viremia despite ART.25 Among these individuals followed for 1-3 years we now explored the relationships between persistent viremia, drug resistance, co-receptor usage, evolution of polymerase (pol) and envelope (env) genes, and the effect of these viral parameters on CD4 recovery and evolution of env and pol.
Methods
Study Population, CD4, and HIV-1 RNA testing
Study participants were recruited from The Centre, a community-based outpatient treatment program in Harare, Zimbabwe providing ART under the supervision of local physicians. Clinical and ART histories were obtained from 30 patients who provided blood samples for CD4, HIV-1 RNA, genotypic resistance testing, and culture in 2001, 2003, and 2004. Ethical approval for this study was obtained from the Human Subjects panel at Stanford University as an exempt protocol using de-identified samples obtained for clinical management. Results were returned to the patients and their physician.
CD4 testing was performed on fresh blood samples. Plasma was stored within 4 hours of collection, and maintained at -70°C. Plasma HIV-1 RNA was quantified using the Roche AMPLICOR™ Monitor test, version 1.5 (Roche Molecular Diagnostics, Branchberg NJ), with a lower limit of detection at 400 copies/ mL (2.6 log10 copies/ mL). An area under the curve (AUC) of plasma virus was calculated based on sequential HIV-1 RNA levels to provide a continuous representation of virus replication. The time points with HIV-1 RNA levels < 2.6 log10 copies/ mL were assigned a value of 2.59 log10 copies/ mL.
Genotype Resistance Testing
HIV RNA was isolated from plasma using the Roche amplicor extraction. RNA was reverse transcribed to cDNA using random hexamer primers, and reverse transcriptase (Superscript III, Invitrogen). PCR amplification of cDNA generated an approximately 1,100 base pair amplicon of the pol gene for genotype resistance testing using primers MAW26 (forward-5′TTGGAAATGTGGAAAGGAAGGAC) and RT21 (reverse-5′CTGTATTTCTGCTATTAAGTCTTTTGATGGG) for first round amplification, and Pro1 (forward-5′CAGAGCCAACAGCCCCACCA) and RT20 (reverse-5′CTTCGTCTCGATCTTGACCGTC) for second round amplification. RNA from the 2001 sample collection was extracted with the TRUPREP™ Viral RNA Kit (Visible Genetics Inc/ Bayer, Inc. Emeryville, CA), and reverse transcribed and amplified using enzymes and primers included in the TRUGENE™ Genotyping System kit, version 1.5 RT-PCR Research Use Only primers, optimized for non-subtype B HIV (Visible Genetics/Bayer, Inc.). Second round amplicon was purified using the QiaQuick PCR purification kit (Qiagen Inc.), and sequenced using Big Dye chemistry on an ABI 377 DNA sequencer. For population sequencing, chromatograms were manually inspected to verify mixtures at greater than 20% identified by the ABI or TRUGENE™ electropherogram interpretation algorithms. The presence of drug resistance and genotypic susceptibility score (GSS) was determined using the Stanford HIV Drug Resistance database, HIVDB.stanford.edu (accession date 12 December, 2006). The current antiretroviral drug regimen and resistance mutations for each individual were used to derive a GSS: a GSS of 3 indicates full susceptibility to 3 drugs, while a GSS of 0 indicates complete lack of activity of the treatment regimen against the patient's consensus virus. Genotypic susceptibility testing was not performed in real time, and treatment changes were undertaken by local physicians based on history of antiretroviral drug exposure and local availability of antiretroviral agents.
Tropism determination
Tropism was determined by passage of virus isolates in MT-2 cells and observation of SI/ NSI phenotype with env consensus sequence analysis. In some cases, a clonal env analysis was performed as previously described when accurate interpretation of the population sequence was not possible.25 Among samples propagated in PBMC cultures, phenotype was determined in low passage virus isolates in GHOST cells, Human Osteosarcoma (HOS) cells expressing Green Fluorescent Protein (GFP) under the control of an HIV promoter. GHOST cells have been transfected with either CCR5 or CXCR4 to determine tropism. The web-based C-PSSM program (http://ubik.microbiol.washington.edu/computing/pssm/) for subtype C HIV-1 was used with V3 loop sequences to estimate tropism from plasma samples without a corresponding virus isolate.26 MT-2 cells and GHOST cells were obtained through the AIDS Research and Reference Reagent Program, Division of AIDS, NIAID, NIH.
Phylogenetic Analysis
Sequences of nucleotides and amino acids were aligned using ClustalW, and manually edited in DAMBE. ModelTest and PAUP*4.0b10 were used to identify the optimal evolutionary model to estimate genetic distances in the C2-V5 of the env gene (approximately 650 base pairs), and pol gene sequences being analyzed.27 The general time reversible model option in PAUP which accommodates six substitution types and selected gamma distribution was used to estimate genetic distance using the neighbor-joining method, and phylogenetic trees were generated in MEGA 3.1. 1000 replicate bootstrap values were calculated in PAUP to ascertain the relationship between viral sequences. Synonymous (dS) to nonsynonymous (dN) changes indicating positive or neutral selection was determined using SynScan (http://hivdb.stanford.edu/pages/synscan.html), with a dS/dN ratio greater than 1 indicating neutral selection, and less than 1 a change in the encoded amino acid interpreted as positive selection. The REGA Subtyping Tool (version 2.0) on HIVDB.stanford.edu was used to verify HIV-1 subtype of pol sequences. The Recombination Identification Program, RIP 3.0 (http://www.hiv.lanl.gov/content/sequence/RIP/RIP.html) was used to assess recombination within sequences.
Statistical Analysis
Cross-tabulations were performed to describe the cohort. Differences between categorical variables were explored using the chi-square test or Fisher's Exact test where appropriate. Relationships between continuous non-parametric variables were examined using the Wilcoxon rank-sum test, and between paired non-parametric data using the Wilcoxon signed rank sum test and a two-sided p-value of < 0.05 considered statistically significant. Univariate linear regression models (PROC GLM; SAS, version 9.1) were used to model the rate of CD4 change over time. The relationship between genetic distance in sequential pol and env sequences, age, gender, CD4 change, viremia (AUC of HIV RNA), changing genotypic susceptibility to ART (GSS), and viral tropism was explored with univariate linear regression models. Multivariate analyses and mixed effects models were performed but are not reported due to indication of overfitting of variables in this small data set.
Sequence Accession Numbers
Nucleotide and amino acid sequences for the pol and env sequences have been deposited in Genbank. The pol sequences have the following accession numbers: AY090840-41, AY090843-49, AY090851, AY090853, AY090855-56, AY090859, XXXX-XXXX. The env sequences have the following accession numbers: AY265929, AY265934, AY265936-37, AY265941-43, AY265948-49, AY265951, AY265954, AY265957, XXXX-XXXX. Participants are identified as TC or ZI followed by a unique number in the submitted sequences.
Results
Baseline HIV RNA, CD4, and CD4 Response
Thirty individuals were repeatedly evaluated in 2001, 2003, and/or 2004, with two sequential CD4 tests performed on all participants, and three CD4 tests performed on 8 of the 30 participants. Patient characteristics in 2001 have been previously published,28 and is updated in table 1 to include subjects recruited in 2003 with follow up in 2004. Among 12 men and 18 women, the median CD4 at enrollment was 122.5 per cu mm (range: 3-441) and HIV-1 RNA level was 4.46 (range: < 2.6 to 5.84 log10 copies/ mL). Median CD4+ T-lymphocyte count increased by 21.7/year (range:-111 to 357) and median percent CD4 by 20% (range: 95% to 5533%). Table 2 shows the CD4 and HIV RNA levels at initial enrolment and at the last study visit, and viral co-receptor usage.
Table 1.
Characteristics at enrollment among subtype C HIV-1 infected Zimbabweans receiving antiretroviral therapy
| Variable | Overall | Male (n=12) | Female (n=18) | * P-value |
|---|---|---|---|---|
| Age | 42.5 (39-49) | 48 (39.5-51.5) | 40.5 (39-45) | 0.06 |
| Initial CD4 per cu mm | 122.5 (42-264) | 128.5 (48-260) | 103 (42- 272) | 0.93 |
| Initial HIV-1 RNA log10 copies/ ml | 4.46 (3.14-5.07) | 3.87 (3.16-5.17) | 4.46 (2.88- 5.07) | 0.76 |
| Change in CD4 per year | 21.7 (-32 - 81) | 14.9 (-39 – 81) | 23.66 (-17 -104) | 0.66 |
| R5 virus at final determination – proportion (%) | 16/27 (59) | 8/11 (73) | 8/16 (50) | 0.43 |
Note: Data are median (interquartile range) unless otherwise indicated
P-values were determined using the Wilcoxon rank-sum test for continuous, non-parametric data, and using the Fisher Exact test for categorical data
Table 2.
Cohort immunologic and viral characteristics at study enrolment and completion
| Subject | Baseline CD4 (cells per cu mm) | Baseline HIV RNA (log10 copies/ mL) | Final CD4 (cells per cu mm) | Final HIV RNA (log10 copies/ mL) | Coreceptor Usage |
|---|---|---|---|---|---|
| TC23 | 255 | 5.02 | 25 | 5.47 | X4 |
| TC67 | 132 | 3.14 | 160 | 5.5 | R5a |
| TC41 | 75 | 2.88 | 29 | 4.88 | R5a |
| TC31 | 13 | 4.48 | 15 | 5.07 | R5 |
| TC52 | 33 | 5.43 | 222 | 4.61 | R5 --> X4 |
| TC28* | 71 | 5.17 | 16 | 4.92 | Dual/Mixed --> R5 |
| TC05 | 321 | 3.26 | 133 | 4.7 | R5a |
| TC06* | 441 | 2.61 | 425 | 2.59 | X4 a --> R5 |
| TC68 | 86 | 2.59 | 249 | 5.35 | R5 a --> X4 a |
| TC13 | 3 | 2.59 | 169 | 3.89 | X4 |
| TC19 | 41 | 5.53 | 149 | 4.92 | R5 |
| TC02 | 9 | 5.07 | 81 | 3.94 | R5 --> Dual/Mixed |
| TC30 | 42 | 5.04 | 14 | 5.68 | X4 |
| TC08* | 34 | 3.15 | 167 | 4.04 | X4 |
| TC22 | 214 | 4.23 | 11 | 5.29 | X4 --> R5 |
| TC35* | 158 | 3.26 | 54 | 5.5 | R5 |
| TC47 | 337 | 4.95 | 459 | 4.45 | R5 a |
| TC114 | 264 | 5.32 | 249 | 5.99 | R5 a |
| TC12* | 244 | 5.42 | 290 | 5.39 | R5 |
| TC113 | 323 | 3.12 | 507 | 4.23 | R5 a |
| TC117 | 22 | 5.21 | 245 | 4.01 | X4 a |
| TC01 | 272 | 4.46 | 46 | 2.69 | R5 a |
| TC03* | 81 | 5.02 | 476 | 3.04 | X4 |
| TC10* | 286 | 113 | 3.98 | X4 | |
| TC26* | 54 | 5.84 | 223 | 2.59 | R5 |
| TC53 | 120 | 2.59 | 196 | 4.9 | R5 a --> X4 |
| TC54 | 222 | 2.59 | 303 | ||
| TC63 | 81 | 4.99 | 86 | 5.79 | R5 a |
| TC101 | 314 | 3.52 | 661 | 3.51 | |
| TC107 | 125 | 3.18 | 81 | 4.4 |
Subjects in whom 3 measurements were available. Provided are the baseline and final CD4 and HIV RNA levels
Coreceptor usage – R5 indicates CCR5-using virus; X4 indicates CXCR4-using virus; and Dual/mixed indicates the presence of both R5 and X4 using virus; arrows are used to indicate a change in coreceptor usage in sequential samples
Coreceptor usage determined from sequence using C-PSSM program
Antiretroviral Regimens. HIV RNA Response, and Drug Resistance
Individuals had initiated ART between 1997 and 2001. Ten of the twenty-seven (37%) had been treated with mono or dual-therapy, and 2 were continuing mono or dual therapy in 2001. Most individuals had received multiple drug regimens; 5 were on their first treatment regimen, 8 individuals on their second, 8 on their third, 5 on their fourth, and one on their fifth regimen. Treatment interruptions were due to intermittent antiretroviral availability and toxicities. The last ART regimen included a protease inhibitor (PI) in 12 individuals, non-nucleoside reverse-transcriptase inhibitor (NNRTI) in 11 individuals, and 2 were receiving PI and NNRTI-containing regimens. One person was maintained on DDI and hydroxyurea, and subsequently on nevirapine monotherapy, with an HIV-1 RNA level <500 copies/ mL for three years. No treatment history was available for three subjects.
Three individuals achieved virologic suppression to <2.6 log10 copies/ mL, while 27 individuals remained viremic. Among individuals with viremia, 16/27 (59%) had a CD4 increase over time, and 11/27 (41%) had a fall in CD4 count during the study period. The median genotypic sensitivity score (GSS) among 20 viremic individuals in whom pol sequence data and treatment history was available was 1.97 (predicted sensitivity to 2 of their drugs) compared with a median GSS of 2.89 among the three individuals who achieved virologic suppression. Treatment history was not available for three individuals, and pol sequence was not amplifiable in four individuals. Table 3 shows the treatment regimens at the time of blood sample collection for this study and genotypic drug resistance for the 23 individuals in whom pol sequence and treatment information was available. Among twenty individuals with viremia and sequential pol sequences, there was accumulation of drug resistance mutations in fourteen individuals. Figure 1a shows the phylogenetic tree of pol sequences, with clustering of sequences from the same individual.
Table 3.
Treatment history and drug resistance associated mutations among subtype C HIV-1 infected Zimbabweans§
| Subject | Treatment Time 1* | Treatment Time 2 and 3* | Protease Mutations (Time 1) | Protease Mutations (Time2) | Protease Mutations (Time 3) | Reverse Transcriptase Mutations (Time 1) | Reverse Transcriptase Mutations (Time 2) | Reverse Transcriptase Mutations (Time 3) |
|---|---|---|---|---|---|---|---|---|
| TC23 | AZT, DDC, SQV | L90M (2001) | L90M (2003) | D67G, K70R, K219Q (2001) | D67G, K70R, K103KR, V118I, K219Q (2003) | |||
| TC67 | AZT, 3TC, SQV | None (2003) | None (2004) | None (2003) | M184MV (2004) | |||
| TC41 | CBV, NVP | D4T,3TC,NVP | None (2003) | None (2004) | A62AV, K65R, K103R, Y181CY, K219KQ (2003) | A62V, K65R, K70R, V75I, F77L, Y115F, F116Y, V118I, Q151M, Y181C, M184I, K219E (2004) | ||
| TC31 | DDI, APV, NVP | EFV, LPV, IDV | I54FL (2001) | V32VG, M46MI, I47IM, L76V (2003) | M41L, D67N, L74V, Y181I, T215Y (2001) | D41L, D67N, L74V, K103N, Y181C, G190A, T215Y (2003) | ||
| TC52 | D4T, 3TC, NVP | D4T, DDI, IDV | None (2003) | None (2004) | M184V, G190A (2003) | D67N, K70R, M184V, K219E (2004) | ||
| TC28 | CBV, SQV | NVP, D4T, 3TC(2003) TDF,LPV,RTV (2004) | L90M (2001) | L90M (2003) | M46I, I84V, L90M (2004) | D67N, K70R, M184V, T215IT, K219Q (2001) | D67N, T69N, K70R, Y181C, M184V, T215V, K219Q (2003) | D67N, T69N, K70R, K101E, Y181C, G190A, T215F, K219Q (2004) |
| TC05 | D4T, 3TC, SQV | L90M (2001) | L90M (2003) | M41L, M184V (2001) | M184V (2003) | |||
| TC06 | DDI, HYD | DDI, HYD (2003) NVP (2004) | None (2001) | None (2003) | None (2004) | None (2001) | None (2004) | None (2004) |
| TC68 | D4T, 3TC, NVP | AZT, 3TC, NFV | None (2003) | I54L, L90LM (2004) | T69A, V106MV, K238KN (2003) | T69A, M184V (2004) | ||
| TC13 | AZT, 3TC, SQV | CBV, NFV | None (2001) | None (2003) | None (2001) | M184V (2003) | ||
| TC19 | CBV | DDI, D4T, EFV | None (2001) | None (2003) | M41LM, D67DN, K70KR, M184V, T215FIST (2001) | M41L, L74V, K103N, V106M (2003) | ||
| TC02 | 3TC, D4T, SQV | None (2001) | L90M (2003) | None (2001) | M184V (2003) | |||
| TC30 | CBV, NFV | D4T, 3TC, NVP | None (2001) | I54V, I84V (2003) | M184V (2001) | M41L, D67N, K70R,Y181C, T215F, K219Q (2003) | ||
| TC08 | D4T, 3TC, SQV | 3TC,D4T,SQV(2003) 3TC,D4T,EFV (2004) | None (2001) | None (2003) | None (2004) | M184V (2001) | L74LV, K103KR, M184V, M230LM, K238EGKR (2003) | K103R, V179D, M184V, T215F, M230L (2004) |
| TC22 | IDV, NVP, SQV | None (2001) | L90M (2003) | M41L,K103N,M184V,T215Y (2001) | M41L, V75I, L100I, K103N, T215Y (2003) | |||
| TC35 | AZT, 3TC, NFV | D4T, 3TC, NVP | None (2001) | None (2003) | I54IF (2004) | M41L, D67N, K70R, Q151M, Y181C, M184V, G190A, T215Y, K219E (2001) | M41L, D67N, K70R, Q151M, Y181C, G190A, T215Y, K219E (2003) | M41L, A62V, K65R, K70R, V75T, Q151M, Y181C, G190A, T215Y, K219E (2004) |
| TC47 | D4T, DDI, NVP | D4T, 3TC, NVP | None (2003) | None (2004) | K103KN, V106MV, Q151KLMQ, Y181CY (2003) | K103N, F116Y, Q151M, Y181C, M184V (2004) | ||
| TC12 | 3TC, DDI, HYD | D4T, 3TC, NVP | None (2001) | None (2003) | None (2004) | None (2001) | M41L, K103N, M184V (2003) | M41L, K103N, M184V, T215Y (2004) |
| TC03 | DDI, HYD, AZT | IDV, D4T, 3TC | None (2001) | D67N,T69N,K70R,F77FL, Q151M,T215IT, K219EQ (2001) | ||||
| TC26 | IDV, RTV, SQV | IDV, CBV | None (2001) | None (2001) | ||||
| TC 53 | AZT, 3TC, NFV | D4T, 3TC, NVP | None (2004) | None (2004) | ||||
| TC54 | AZT, 3TC, NVP | D4T, 3TC, NVP | None (2003) | V118I (2003) | ||||
| TC63 | CBV, EFV | None (2003) | K103N (2003) |
Individuals for whom treatment histories or pol sequence are unavailable are not included in the table.
Treatment histories indicate the antiretroviral regimen the individual was receiving prior to the collection of the blood sample.
Abbreviations: 3TC, lamivudine; D4T, stavudine; AZT, zidovudine; CBV, combivir; DDI, didanosine; EFV, efavirenz; HYD, hydroxyurea; IDV, indinavir; LPV, lopinavir; NFV, nelfinavir; NVP, nevirapine; RTV, ritonavir; SQV, saquinavir; TDF, tenofovir;
Figure 1.
Phylogenetic analysis of the pol (Figure 1a) and env genes (Figure 1b) of HIV among individuals with viremia while receiving combination antiretroviral therapy. Tree topology was inferred using the general time reversible evolutionary model based on parameters estimated using ModelTest. Optimal evolutionary model parameter estimates identified for env sequences based on the hierarchical likelihood ratio tests were as follows: Model selected: TVM+I+G; fA = 0.4309, fC = 0.1744, fG = 0.1885, and fT = 0.2062; R matrix values, RA->C = 1.5971, RA->G = 3.7064, RA->T = 0.8971, RC->G = 1.0596, RC->T = 3.7064, and RG->T = 1.0000; the proportion of invariable sites was 0.1994, and gamma distribution shape parameter=0.4728. Evolutionary model parameter estimates for pol sequences were as follows: Model selected: TVM+I+G fA = 0.3763, fC = 0.1691, fG = 0.2015, and fT = 0.2531; R matrix values, RA->C = 3.1596, RA->G = 10.8420, RA->T = 1.6226, RC->G = 0.9179, RC->T = 10.8420, and RG->T = 1.0000; the proportion of invariable sites was 0.4501, and gamma distribution shape parameter=0.7206. The subtype B HXB2 isolate was used as the outgroup to root the tree. Bootstrap values shown are based on 1000 replicates calculated in PAUP to ascertain the relationship between viral sequences. Samples from which tropism was determined by sequence only are indicated with an asterix, and the tropism determination of the env sequences and/or isolates is indicated using solid blue diamonds for R5 virus, and solid red squares for X4 or dual/mixed virus. Sequential sequences from individuals in whom a tropism switch was detected are indicated with an arrow.
Tropism and CD4 Response
Tropism was determined in 27 individuals, of whom 13 had exclusively R5 and 14 had X4 or D/M virus. Tropism determination was unsuccessful in 3 individuals, 1 of whom had achieved virologic suppression. Culture-based methods were unsuccessful for 9/30 samples, and tropism determination was performed using sequence only. Both culture-based assays and sequence analysis were performed on the remaining 21 samples, with agreement in tropism prediction in all but 1 subject. TC52 had two samples drawn in 2003: the first sample was NSI by culture but the PSSM interpretation of sequence was X4, while the second isolate was SI by culture with a PSSM sequence interpretation indicating R5 virus.
Paired env sequences are shown in the phylogenetic tree in figure 1b. Of the 27 individuals with detectable HIV-1 RNA levels despite ART, 12 had R5 virus, and 13 had X4 virus. Paired sequential env sequences were available from 14 individuals, and these are shown in the phylogenetic tree in figure 1b. 7 individuals had a tropism switch during the study period: 3 from R5 to X4, 2 from X4 to R5, and 1 each from R5 to D/M and D/M to R5; the five subjects in who tropism switch was detected, and for whom paired sequence data are available are depicted with arrows in figure 1b. Based on the tropism at the final sampling point, 8/16 women had X4 or D/M virus, and 3/11 men had X4 virus (p=0.43, Fisher's Exact Test). The median initial CD4 count per cu mm among individuals with exclusively R5 virus was 158 (range: 13-337; 25-75% IQR 75-272) compared to 76 (range: 3-441; 25-75% IQR 33-214) among individuals with evidence for X4 virus at some point during the study period (p=0.16, Wilcoxon rank-sum test). The median initial HIV RNA level among individuals with exclusive R5 virus was 4.48 (range: 2.88 – 5.84; 25-75% IQR 3.26 – 5.32) and 5.02 (range: 2.59 – 5.43; 25-75% IQR 2.61 – 5.07) among those with evidence for X4 virus (p=0.58, Wilcoxon rank-sum test).
Envelope and Polymerase Phylogenetic Analysis
Neighbor-joining phylogenetic trees for consensus pol and env sequences are shown in figure 1a and 1b, respectively. Pol sequences were subtype C HIV-1 with bootstrap values >97%. Sequential sequences from the same individuals were closely related to one another, with bootstrap values ≥ 88%. There was no evidence for viral recombination by sequence analysis. One consensus sequence from the plasma of TC28 in 2001 was not interpretable with respect to tropism, however, six molecular clones generated from second round amplicon from the 2001 sample showed two discrete viral populations by sequence, one X4, the other R5. The viral population consensus sequence in 2003 demonstrated only a pure R5 viral population phylogenetically closer to the R5 clonal population identified in the 2001 patient sample, suggesting a fitness advantage associated with R5 usage,.
Sequential pol or env sequences were used to determine an intra-individual median genetic distance using the optimal evolutionary model identified by ModelTest. On treatment distribution of genetic distance between sequential time points in the pol gene among 21 individuals and the env gene among 14 individuals is shown in Figure 2. The genetic distance in env exceeded that in pol per unit time, with a median genetic distance in env of 0.013 compared with 0.0085 in pol (p=0.0494, Wilcoxon signed rank sum test). There was a significant correlation between genetic distance in pol and env (R2 = 0.48, p=0.005). Genetic distance in pol was similar among individuals infected with R5 and X4 or D/M virus, with an arithmetic mean genetic distance of 0.014 among 10 individuals with R5 virus and 11 individuals with X4 virus. Positive selection was seen in the sequential consensus env sequence of 3/14 individuals and 1/19 consensus pol sequences.
Figure 2.
Figure 2 shows the genetic distance per year in pol and env genes among individuals with sequential sequences available on treatment. The bold line represents the median genetic distance for the pol gene genetic distances (represented as blue triangles), and for env gene genetic distances (represented as red triangles). The Wilcoxon signed rank sum test was used to perform the equivalent of a paired t-test for non-parametric data, and showed genetic distance per year in env than in pol, p=0.0494.
Predictors of CD4 Response
Results of the linear regression models to evaluate predictors of CD4 cell numbers over time among individuals receiving non-suppressive ART are shown in table 4. Positive parameter estimates indicate an increase in CD4 cell numbers, while negative parameter estimates indicate a decline in CD4 in association with the variable being analyzed. By univariate analysis there was a significant negative correlation between rate of CD4 gain and increased viremia. Those with higher HIV RNA levels over time (higher AUC viremia) achieved less CD4 increase than those with lower levels of viremia (p=0.04). An increase in genotypic susceptibility score (GSS) to the drugs in a treatment regimen, indicating more potent drug activity, demonstrated a trend towards an increased CD4 gain, p=0.065. Greater viral evolution in pol and env, as measured by genetic distance, were associated with a trend towards lower rates of CD4 gain, p=0.08 and p=0.097 respectively (Figure 3). Age, gender, and tropism (measured as R5 only at all time points, or X4/ D/M on at least one time point) were not associated with the rate of CD4 change on treatment.
Table 4.
Predictors of rate of CD4 cell change per year among subtype C HIV-1 infected individuals receiving non-suppressive combination antiretroviral therapy, linear regression model (Proc GLM, SAS version 9.1).
| Variable | Univariate | |
|---|---|---|
| Parameter Estimate | p-value | |
| Age | 1.92 | 0.39 |
| Gender | 28.7 | 0.48 |
| * AUC HIV RNA | -0.93 | 0.04 |
| R5 Tropism | -28.6 | 0.43 |
| Increase in GSS | 65.63 | 0.065 |
| Genetic Distance - env | -20.47 | 0.097 |
| Genetic Distance – pol | -43.32 | 0.08 |
AUC HIV-1 RNA level represents the area under the curve of viremia over time
Figure 3.
Figure 3 shows the relationship between the rate of CD4 change and evolution in the envelope (figure 3a) and polymerase (figure 3b) genes as measured by genetic distance between sequential sequences of the respective gene. Each data point is indicated with a red triangle for envelope (env) and blue triangle for polymerase (pol), with the regression slope shown for each data set. An inverse relationship is seen between CD4 response and evolution, where greater evolution in the envelope and polymerase gene is associated with lower CD4 gain.
Predictors of Viral Evolution
The relationship between viral evolution measured as genetic distance between sequential plasma samples, sequential CD4 measurements, HIV-1 RNA levels, and viral tropism was explored using linear regression models, and the results shown in table 5. Evolution in env, represented by genetic distance, showed a trend towards association with level of HIV viremia (p=0.069), with increasing levels of viremia associated with greater evolution between sequential samples. CD4 change also showed a trend towards association with env evolution (p=0.097) with greater env evolution seen with among individuals with lower CD4 recovery. Evolution in env was not associated with age, gender, change in GSS, or tropism. There was a strong association between evolution in pol and level of viremia over time (p=0.007), with higher levels of viremia associated with increased pol evolution as measured by genetic distance between sequential samples. There was a trend towards an association between pol evolution and rate of CD4 change, p=0.08, with greater pol evolution seen among individuals with poor CD4 responses to treatment.
Table 5. Predictors of Env and Pol Gene Evolution by univariate linear regression analysis.
| Variable | Env Evolution Univariate | Pol Evolution Univariate | ||
|---|---|---|---|---|
| Parameter Estimate | p-value | Parameter Estimate | p-value | |
| Age | -0.059 | 0.3687 | -0.033 | 0.1178 |
| Gender | 1.462 | 0.1228 | 0.61 | 0.1053 |
| CD4 Change | -0.01 | 0.097 | -0.004 | 0.0805 |
| *AUC HIV RNA | 0.02079 | 0.0693 | 0.01 | 0.0071 |
| Increase in GSS | -1.308 | 0.1876 | -0.363 | 0.3148 |
| R5 Tropism | -1.226 | 0.2187 | -0.077 | 0.842 |
Discussion
Virologic predictors of clinical benefit, measured by changes in CD4 cell count, were explored among ambulatory subtype C HIV-1 infected individuals on non-suppressive ART being monitored in an outpatient community-based clinical setting. Linear regression modeling identified the AUC of HIV-1 RNA as the most significant predictor of CD4 cell change among ART-treated individuals with multi-drug resistance. Similarly, the area under the curve of viremia is the primary predictor of evolution in env and pol. Infrequent sampling for HIV RNA quantitation may affect the AUC of HIV RNA levels that was used in this analysis. However, individuals continued treatment on clinical grounds, without the provision of viral or immunologic monitoring in real-time, and with very limited access to effective second-line or alternative drug regimens. While more frequent HIV RNA measurements, especially timed to reflect the impact of treatment changes, would have been ideal in the context of this study, we believe that the AUC HIV RNA levels that we used in this study reflect the virologic status of these participants, most of whose treatment changes provided less than one effective new antiretroviral agent. We hypothesize that reduced virus replication, and decreased area under the curve of viremia, may limit viral evolution. Less evolution is associated with higher CD4 gains as shown in Figure 3. In contrast, chemokine co-receptor use was not associated with area under the curve of viremia, nor evolution in env or pol. This may explain why X4-usage was not associated with decreased CD4 recovery. Despite the association of X4 virus with more rapid progression of disease in the absence of ART,5-10 among these subtype C HIV-1 infected, ART treated individuals, there was no evidence of an association between X4 virus and either increasing HIV RNA or decreasing CD4 cell counts. X4 usage may not mark increased pathogenicity and effective modulation of viral immune escape in the presence of drug resistance associated mutations.
A high frequency of X4 or D/M virus has been identified in subtype B HIV, associated with failing therapy and nadir CD4 count.18, 29 In ACTG 175, a pre-HAART era study of nucleoside RTI drugs among subtype B HIV-infected patients X4 virus (measured as SI) was independently associated with CD4 cell decline, AIDS and death 30, 31. More recent studies of treatment and tropism in subtype B viruses suggest that X4 virus is associated with lower CD4 counts on ART,17 and that X4 viruses have increased replicative capacity and reduced CD4 response to treatment associated with drug resistance and persistent viremia.16 X4 viruses have been infrequently identified in prior studies of treatment naïve subtype C HIV-1 infected individuals in the absence of ART exposure.23, 24 In the present study, about half of subtype C HIV-1 infected Zimbabweans receiving ART carried an X4 virus, providing additional evidence that incomplete viral suppression may select or allow X4 usage. Higher on-treatment HIV-1 RNA levels were associated with greater evolution in env than pol, and persistent viremia on ART was associated with accumulation of drug resistance associated mutations in pol. This is consistent with studies of subtype B HIV-1 in which infected individuals with viremia, despite ART develop new drug resistance mutations.29, 32
There are several limitations to this observational study. The patients who returned for repeated measures to this community clinic reflect a selected group with the potential for survival bias. Misclassification of reported drug regimens, treatment interruptions, poor adherence and limited access to medications may have contributed to continued viremia and drug resistance. Individuals continued treatment on clinical grounds, without viral or immunologic monitoring, and had very limited access to second-line or alternative drug regimens. Therapy with virologic non-suppressive (sub-optimal) regimens, as occurred here, is not recommended. However, in the absence of HIV RNA monitoring and with limited access to second line treatments, clinical monitoring alone and continued ART are the only option for many HIV infected individuals in resource-constrained settings.
Continued viremia with switches in tropism and the development of drug resistance on ART demonstrates that increasing ART access in resource-constrained settings may drive the evolution of HIV-1. The availability of a continuous supply of antiretroviral drugs is therefore critical to minimize treatment interruption, breakthrough viremia, and the potential for development of drug resistance and viral evolution. The virologic outcome of this treatment mono- and dual-therapy exposed cohort does not necessarily reflect the typical characteristics of current antiretroviral naïve populations receiving effective first-line combination ART. The relationship that we have explored in this treatment experienced group with multi-drug resistance should be further evaluated in a prospectively enrolled cohort more representative of the current recipients of combination ART globally. However, serial samples from these individuals in a community-based treatment program demonstrate an association between the level of viremia despite ART and increasing viral evolution. CD4 benefit was seen despite persistent viremia in the majority of patients, a finding which highlights the limitations of CD4 monitoring to gauge virologic response. A gain in CD4 was seen among 8/14 subjects with increasing drug resistance, suggesting that in some cases, the increase in drug resistance may be accompanied with a fitness cost to the viral population. In contrast, virologic outcomes, including greater HIV RNA levels, drug resistance and evolution are associated with lower CD4 cell responses. Where possible, access to virologic monitoring in resource-constrained settings will enhance timely access to effective second-line treatment regimens to reduce accumulation of drug resistance, diminish viremia and viral evolution, while sustaining clinical benefit.
Acknowledgments
DDCF Distinguished Clinical Scientist Award, 2000-2005, and NIH, NIAID R01 AI060399, Drug Resistance and Pathogenesis in Subtype C-HIV-1 to David Katzenstein
Footnotes
Presented in part: 13th Conference on Retroviruses and Opportunistic Infections, Los Angeles, CA, 2007, abstract no.171; and Conference on Retroviruses and Opportunistic Infections,Denver, CO, February 2006, abstract no. 279.
Conflict of Interest: None
References
- 1.Pillai SK, Good B, Pond SK, et al. Semen-specific genetic characteristics of human immunodeficiency virus type 1 env. J Virol. 2005 Feb;79(3):1734–1742. doi: 10.1128/JVI.79.3.1734-1742.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Fiore J, Zhang Y, Björndal A, et al. Biological correlates of HIV-1 heterosexual transmission. AIDS. 1997;11(9):1089–1094. doi: 10.1097/00002030-199709000-00002. [DOI] [PubMed] [Google Scholar]
- 3.Shankarappa R, Gupta P, Learn GH, Jr, et al. Evolution of human immunodeficiency virus type 1 envelope sequences in infected individuals with differing disease progression profiles. Virology. 1998 Feb 15;241(2):251–259. doi: 10.1006/viro.1997.8996. [DOI] [PubMed] [Google Scholar]
- 4.de Mendoza C, Rodriguez C, Garcia F, et al. Prevalence of X4 tropic viruses in patients recently infected with HIV-1 and lack of association with transmission of drug resistance. J Antimicrob Chemother. 2007 Apr;59(4):698–704. doi: 10.1093/jac/dkm012. [DOI] [PubMed] [Google Scholar]
- 5.Connor RI, Ho DD. Human immunodeficiency virus type 1 variants with increased replicative capacity develop during the asymptomatic stage before disease progression. J Virol. 1994 July 1;68(7):4400–4408. doi: 10.1128/jvi.68.7.4400-4408.1994. 1994. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Arien KK, Gali Y, El-Abdellati A, Heyndrickx L, Janssens W, Vanham G. Replicative fitness of CCR5-using and CXCR4-using human immunodeficiency virus type 1 biological clones. Virology. 2006 Mar 30;347(1):65–74. doi: 10.1016/j.virol.2005.11.045. [DOI] [PubMed] [Google Scholar]
- 7.Koot M, van 't Wout AB, Kootstra NA, de Goede RE, Tersmette M, Schuitemaker H. Relation between changes in cellular load, evolution of viral phenotype, and the clonal composition of virus populations in the course of human immunodeficiency virus type 1 infection. J Infect Dis. 1996 Feb;173(2):349–354. doi: 10.1093/infdis/173.2.349. [DOI] [PubMed] [Google Scholar]
- 8.Moyle GJ, Wildfire A, Mandalia S, et al. Epidemiology and predictive factors for chemokine receptor use in HIV-1 infection. J Infect Dis. 2005 Mar 15;191(6):866–872. doi: 10.1086/428096. [DOI] [PubMed] [Google Scholar]
- 9.Koot M, Keet IP, Vos AH, et al. Prognostic value of HIV-1 syncytium-inducing phenotype for rate of CD4+ cell depletion and progression to AIDS. Ann Intern Med. 1993 May 1;118(9):681–688. doi: 10.7326/0003-4819-118-9-199305010-00004. [DOI] [PubMed] [Google Scholar]
- 10.Markowitz M, Mohri H, Mehandru S, et al. Infection with multidrug resistant, dual-tropic HIV-1 and rapid progression to AIDS: a case report. Lancet. 2005 Mar 19-25;365(9464):1031–1038. doi: 10.1016/S0140-6736(05)71139-9. [DOI] [PubMed] [Google Scholar]
- 11.Jensen MA, Li FS, van 't Wout AB, et al. Improved Coreceptor Usage Prediction and Genotypic Monitoring of R5-to-X4 Transition by Motif Analysis of Human Immunodeficiency Virus Type 1 env V3 Loop Sequences. J Virol. 2003 December 15;77(24):13376–13388. doi: 10.1128/JVI.77.24.13376-13388.2003. 2003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Fouchier RA, Brouwer M, Broersen SM, Schuitemaker H. Simple determination of human immunodeficiency virus type 1 syncytium-inducing V3 genotype by PCR. J Clin Microbiol. 1995 Apr;33(4):906–911. doi: 10.1128/jcm.33.4.906-911.1995. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Fouchier RA, Groenink M, Kootstra NA, et al. Phenotype-associated sequence variation in the third variable domain of the human immunodeficiency virus type 1 gp120 molecule. J Virol. 1992 May 1;66(5):3183–3187. doi: 10.1128/jvi.66.5.3183-3187.1992. 1992. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Pollakis G, Abebe A, Kliphuis A, et al. Phenotypic and genotypic comparisons of CCR5- and CXCR4-tropic human immunodeficiency virus type 1 biological clones isolated from subtype C-infected individuals. J Virol. 2004 Mar;78(6):2841–2852. doi: 10.1128/JVI.78.6.2841-2852.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Cardozo T, Kimura T, Philpott S, Weiser B, Burger H, Zolla-Pazner S. Structural Basis for Coreceptor Selectivity by The HIV Type 1 V3 Loop. AIDS Res Hum Retroviruses. 2007 Mar;23(3):415–426. doi: 10.1089/aid.2006.0130. [DOI] [PubMed] [Google Scholar]
- 16.Solomon A, Lane N, Wightman F, Gorry PR, Lewin SR. Enhanced replicative capacity and pathogenicity of HIV-1 isolated from individuals infected with drug-resistant virus and declining CD4+ T-cell counts. J Acquir Immune Defic Syndr. 2005 Oct 1;40(2):140–148. doi: 10.1097/01.qai.0000173460.75322.93. [DOI] [PubMed] [Google Scholar]
- 17.Delobel P, Sandres-Saune K, Cazabat M, et al. R5 to X4 switch of the predominant HIV-1 population in cellular reservoirs during effective highly active antiretroviral therapy. J Acquir Immune Defic Syndr. 2005 Apr 1;38(4):382–392. doi: 10.1097/01.qai.0000152835.17747.47. [DOI] [PubMed] [Google Scholar]
- 18.Hunt PW, Harrigan PR, Huang W, et al. Prevalence of CXCR4 tropism among antiretroviral-treated HIV-1-infected patients with detectable viremia. J Infect Dis. 2006 Oct 1;194(7):926–930. doi: 10.1086/507312. [DOI] [PubMed] [Google Scholar]
- 19.Chin BS, Choi J, Nam JG, et al. Inverse relationship between viral load and genotypic resistance mutations in Korean patients with primary HIV type 1 infections. AIDS Res Hum Retroviruses. 2006 Nov;22(11):1142–1147. doi: 10.1089/aid.2006.22.1142. [DOI] [PubMed] [Google Scholar]
- 20.de Mendoza C, Martin-Carbonero L, Gallego O, Corral A, Gonzalez-Lahoz J, Soriano V. Relationship between drug resistance mutations, plasma viremia, and CD4+ T-cell counts in patients with chronic HIV infection. J Med Virol. 2005 May;76(1):1–6. doi: 10.1002/jmv.20316. [DOI] [PubMed] [Google Scholar]
- 21.Kaplan SS, Ferrari G, Wrin T, et al. Longitudinal assessment of immune response and viral characteristics in HIV-infected patients with prolonged CD4(+)/viral load discordance. AIDS Res Hum Retroviruses. 2005 Jan;21(1):13–16. doi: 10.1089/aid.2005.21.13. [DOI] [PubMed] [Google Scholar]
- 22.Bangsberg DR, Acosta EP, Gupta R, et al. Adherence-resistance relationships for protease and non-nucleoside reverse transcriptase inhibitors explained by virological fitness. AIDS. 2006 Jan 9;20(2):223–231. doi: 10.1097/01.aids.0000199825.34241.49. [DOI] [PubMed] [Google Scholar]
- 23.Peeters M, Vincent R, Perret JL, et al. Evidence for differences in MT2 cell tropism according to genetic subtypes of HIV-1: syncytium-inducing variants seem rare among subtype C HIV-1 viruses. J Acquir Immune Defic Syndr Hum Retrovirol. 1999 Feb 1;20(2):115–121. doi: 10.1097/00042560-199902010-00002. [DOI] [PubMed] [Google Scholar]
- 24.Abebe A, Demissie D, Goudsmit J, et al. HIV-1 subtype C syncytium- and non-syncytium-inducing phenotypes and coreceptor usage among Ethiopian patients with AIDS. AIDS. 1999 Jul 30;13(11):1305–1311. doi: 10.1097/00002030-199907300-00006. [DOI] [PubMed] [Google Scholar]
- 25.Johnston ER, Zijenah LS, Mutetwa S, Kantor R, Kittinunvorakoon C, Katzenstein DA. High frequency of syncytium-inducing and CXCR4-tropic viruses among human immunodeficiency virus type 1 subtype C-infected patients receiving antiretroviral treatment. J Virol. 2003 Jul;77(13):7682–7688. doi: 10.1128/JVI.77.13.7682-7688.2003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Jensen MA, Coetzer M, van 't Wout AB, Morris L, Mullins JI. A reliable phenotype predictor for human immunodeficiency virus type 1 subtype C based on envelope V3 sequences. J Virol. 2006 May;80(10):4698–4704. doi: 10.1128/JVI.80.10.4698-4704.2006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Posada D, Crandall KA. MODELTEST: testing the model of DNA substitution. Bioinformatics. 1998;14(9):817–818. doi: 10.1093/bioinformatics/14.9.817. [DOI] [PubMed] [Google Scholar]
- 28.Kantor R, Zijenah LS, Shafer RW, et al. HIV-1 subtype C reverse transcriptase and protease genotypes in Zimbabwean patients failing antiretroviral therapy. AIDS Res Hum Retroviruses. 2002 Dec 10;18(18):1407–1413. doi: 10.1089/088922202320935483. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Hatano H, Hunt P, Weidler J, et al. Rate of viral evolution and risk of losing future drug options in heavily pretreated, HIV-infected patients who continue to receive a stable, partially suppressive treatment regimen. Clin Infect Dis. 2006 Nov 15;43(10):1329–1336. doi: 10.1086/508655. [DOI] [PubMed] [Google Scholar]
- 30.Katzenstein DA, Hammer SM, Hughes MD, et al. The relation of virologic and immunologic markers to clinical outcomes after nucleoside therapy in HIV-infected adults with 200 to 500 CD4 cells per cubic millimeter. AIDS Clinical Trials Group Study 175 Virology Study Team. N Engl J Med. 1996 Oct 10;335(15):1091–1098. doi: 10.1056/NEJM199610103351502. [DOI] [PubMed] [Google Scholar]
- 31.Tierney C, Lathey JL, Christopherson C, et al. Prognostic value of baseline human immunodeficiency virus type 1 DNA measurement for disease progression in patients receiving nucleoside therapy. J Infect Dis. 2003 Jan 1;187(1):144–148. doi: 10.1086/345870. [DOI] [PubMed] [Google Scholar]
- 32.Kantor R, Shafer RW, Follansbee S, et al. Evolution of resistance to drugs in HIV-1-infected patients failing antiretroviral therapy. AIDS. 2004 Jul 23;18(11):1503–1511. doi: 10.1097/01.aids.0000131358.29586.6b. [DOI] [PMC free article] [PubMed] [Google Scholar]