Impact of Multi-Targeted Antiretroviral Treatment on Gut T Cell Depletion and HIV Reservoir Seeding during Acute HIV Infection : functional cure
Download the PDF here|
"There may be a window of opportunity during early AHI to intervene and limit CD4 destruction and HIV reservoir formation with the ultimate goal of achieving drug-free remission of HIV. In addition to ART during early Fiebig stages, strategies such as therapeutic HIV vaccines or drugs that target the long-lived cellular reservoir may be necessary to achieve this goal ,  - and persons aggressively treated in these early Fiebig stage infections may be ideal candidates for these interventions.....In order to test for HIV functional cure, treatment interruption will be necessary........Our study provided evidence that identifying AHI subjects by NAT and sequential EIA, and enrolling them in a study is feasible but technically and logistically challenging and costly, . Strengthening the awareness of symptomatic AHI is a less challenging way to identify acute and recent HIV infection making early treatment far more possible......
These findings favor early intervention during AHI to limit immune destruction and HIV reservoir size, and also highlight that immune destruction begins in the earliest days after infection. The latter finding raises concerns regarding the interpretation of these data in the context of protection when considering treatment in later stages of primary infection .
It is thought that the early depletion of the GALT, the largest reservoir of CD4+ T cells in the body, is a blow from which the host may not recover even after prolonged ART in the chronic phase of infection , , , . We employed a strategy that blocked HIV at three steps in the viral life cycle -at entry (CCR5 inhibitor), reverse transcription [nucleoside reverse transcriptase inhibitors (NRTIs) and non-NRTI) and integration (integrase inhibitor)-and found a marked reduction in viral burden in both gut and plasma HIV RNA and DNA. The extent of HIV DNA reduction after 6 months of therapy exceeded that achieved in chronically infected patients following almost 5 years of conventional three-drug ART. Importantly, in persons whose gut CD4+CCR5+ T cells were depleted, megaHAART was associated with reconstitution of gut CD4+CCR5+ T cells to the normal range.
Limited knowledge exists on early HIV events that may inform preventive and therapeutic strategies. This study aims to characterize the earliest immunologic and virologic HIV events following infection and investigates the usage of a novel therapeutic strategy.
Methods and Findings
We prospectively screened 24,430 subjects in Bangkok and identified 40 AHI individuals. Thirty Thais were enrolled (8 Fiebig I, 5 Fiebig II, 15 Fiebig III, 2 Fiebig IV) of whom 15 completed 24 weeks of megaHAART (tenofovir/emtricitabine/efavirenz/ralte gravir/maraviroc).Sigmoid biopsies were completed in 24/30 at baseline and 13/15 at week 24.
At baseline, the median age was 29 years and 83% were MSM. Most were symptomatic (87%), and were infected with R5-tropic (77%) CRF01_AE (70%). Median CD4 was 406 cells/mm3. HIV RNA was 5.5 log10 copies/ml. Median total blood HIV DNA was higher in Fiebig III (550 copy/106 PBMC) vs. Fiebig I (8 copy/106 PBMC) (p = 0.01) while the median %CD4+CCR5+ gut T cells was lower in Fiebig III (19%) vs. Fiebig I (59%) (p = 0.0008).
After 24 weeks of megaHAART, HIV RNA levels of <50 copies were achieved in 14/15 in blood and 13/13 in gut. Total blood HIV DNA at week 0 predicted reservoir size at week 24 (p<0.001). Total HIV DNA declined significantly and was undetectable in 3 of 15 in blood and 3 of 7 in gut. Frequency of CD4+CCR5+ gut T cells increased from 41% at baseline to 64% at week 24 (p>0.050); subjects with less than 40% at baseline had a significant increase in CD4+CCR5+ T cells from baseline to week 24 (14% vs. 71%, p = 0.02).
Gut T cell depletion and HIV reservoir seeding increases with progression of AHI. MegaHAART was associated with immune restoration and reduced reservoir size. Our findings could inform research on strategies to achieve HIV drug-free remission.
Three decades after the discovery of antiretroviral therapy (ART), complete eradication of HIV infection has not been achieved except under unique circumstances . A slightly less difficult target may be the long-term, drug-free remission of HIV or functional cure through modulation of immune responses; the basis for therapeutic vaccination approaches that, to date, have not provided evidence of control . Preservation of immune function by preventing the CD4 depletion in acute HIV infection (AHI), which occurs prominently in the gut, may be a prerequisite to achieving functional cure . Gut CD4 T cell destruction and mucosal breakdown are linked to immune activation in chronic HIV - a key driver of chronic CD4 decline , . Data on immunologic and virologic events as early as Fiebig stages I to III in humans infected with HIV are lacking; yet, such knowledge could inform the design of preventive HIV vaccines and therapeutics , .
Definitions of AHI vary by study but generally include persons with HIV viremia in the absence of IgG antibody to HIV proteins . Because of high HIV viremia and infectiousness of acute transmitted-founder viruses , AHI subjects are more likely to transmit HIV . Symptomatic AHI, sometimes referred to as acute retroviral syndrome (ARS), is characterized by a flu-like syndrome that coincides with peak HIV viremia, and occurs around three weeks after infection .
The best practice for clinical management of AHI is currently unknown. Highly active ART (HAART) instituted during early infection could alleviate CD4 loss, suppress viremia, and limit the size of the latent reservoir. However, there is no conclusive evidence for improved long term clinical outcome, and treatment is not always benign , . Therefore, treatment of AHI is optional in guidelines . Theoretically, usage of a CCR5 inhibitor and/or an integrase inhibitor in addition to standard ART in AHI may reduce HIV spread and limit immune damage .
HIV circulating recombinant form (CRF) 01_AE, and to a much lesser extent, subtype B are prevalent in Thailand . We investigated the clinical, immunologic and virologic characteristics of AHI Thai subjects as well as the short-term outcomes of using multi-targeted HAART. We hypothesized that 1) Real-time pooled nucleic acid testing (NAT) and sequential enzyme immunoassay (EIA) of high prevalence HIV-seronegative subjects from HIV voluntary counseling and testing centers (VCT) in Bangkok would yield volunteers with AHI that is predominantly non-subtype B, and 2) The use of 5-drug ART will have a significant impact on immunity and HIV viral burden.
This study informs the usage of a novel therapeutic strategy with a CCR5 inhibitor in addition to an integrase inhibitor and reverse transcriptase inhibitors in the earliest clinical stage of HIV infection.
Diagnosis of AHI and characteristics of AHI subjects
Between 20 April 2009 to 31 December 2010, 24,430 samples were prospectively screened to identify 40 subjects with AHI (Figure 1). Twenty-six were identified by pooled NAT (negative HIV IgM antibody - Fiebig I/II) and 14 were identified by sequential EIA (positive HIV IgM antibody - Fiebig III/IV) with an overall AHI incidence of 1.7/1000 screened; 95% confidence interval 1-2.7/1000. Seven did not enroll in the study -3 were not able to be contacted and 4 were non-Thais. Thirty-three subjects enrolled in the study but 3 were excluded from the baseline analysis as they had progressed to Fiebig V at enrollment.
Subjects were mostly young MSM with an estimated time of HIV exposure of about two weeks (Table 1). The median CD4 count was 406 cells/mm3 and the plasma HIV RNA was 5.5 log10copies/ml. The median gut HIV RNA was 596 copies/mg tissue and 6 had levels <50 copies/mg tissue. Most (70%) were infected with HIV-1 CRF01_AE. The seven non-typable samples by MHAbce contained CRF01_AE (n = 3), B (n = 3) and CRF01_AE/B (n = 1) genetic material. Almost 80% had R5 HIV virus by the Trofile assay, and 10% had primary NNRTI resistance. Most (n = 26) were symptomatic with fever, myalgia and fatigue being the most common manifestations, occurring around 11 days after estimated time of exposure (Table 2). Notably absent were respiratory symptoms. The mean duration (SD) from onset of HIV exposure by self-reporting history to Fiebig stages was 12 (9.6) days for I, 16 (5.6) days for II, 18 (7.8) days for III, and 29 (3.6) days for IV.
The first 15 subjects on megaHAART had treatment outcome data up to week 24 (Figure 1). They had similar characteristics as the whole cohort (Table 1). Figure 2 shows a rise in HIV RNA between the screening and baseline visits [mean (SD) of 3 (1.6) days] suggesting that our subjects are captured early before reaching peak viremia. After treatment, plasma HIV RNA declined rapidly with 6/15 achieving HIV RNA<50 copies/ml by week 4 and 14/15 having undetectable HIV RNA at week 24 (Figure 3). At week 24, HIV RNA in gut tissue was below 50 copies/mg tissue in all 13 subjects tested. Median peripheral blood CD4 count rose by 188 cells/mm3 at week 2 followed by a sustained rise to 591 cells/mm3 at week 24. Four discontinued EFV before week 12 due to CNS symptoms (1), rash (1) and primary NNRTI resistance (2).
Gut T cell depletion during AHI and its restoration following megaHAART
For the sigmoid biopsy flow analysis, data for 22 subjects at baseline and 11 subjects at week 24 were available for analysis. At baseline, a significant decrease in the median frequency of CD4+CCR5+ T cells with progression of Fiebig stages was observed with 53% at Fiebig I and 19.3% at Fiebig III (p = 0.001) (Figure 4A). The loss of CD4+CCR5+ T cells between Fiebig I and Fiebig III mainly occurred within the effector memory (EM: CD27-CD45RO+) and central memory (CM: CD27+CD45RO+) CD4+ T cell sub-sets. The median frequency of CD4+CCR5+ EM T cells dropped from 56% at Fiebig I to 18% at Fiebig III (p = 0.008) while CM T cells dropped from 80% to 46% from Fiebig I to Fiebig III, respectively (p = 0.001).
This association was not seen for PBMCs, where the frequency of CD4+CCR5+ T cells at Fiebig I was 9.1% (n = 6) compared to 9.6% at Fiebig III (n = 8) (p>0.05). However there was no change in the frequency of CD8+CCR5+ T cells in the gut mucosa (Fiebig I: 91.6%; Fiebig III: frequency 94.6%, p>0.05) that could contribute to the changes observed in the frequency of CD4+CCR5+ T cells. Additionally we observed that in the gut mucosa CD4+CCR5+ T cells depletion increased with higher total HIV DNA, p = 0.05 and r = 0.46.
After 24 weeks of megaHAART, the median frequency of CD4+CCR5+ T cells in the gut mucosa showed an increased trend compared to baseline (baseline: 41%, 24 weeks: 64%, p>0.05). However, in five subjects with a frequency of CD4+CCR5+ T cells below the median baseline value of 40%, a significant increase in the frequency of CD4+CCR5+ T cells from baseline to 24 weeks of treatment of 14% to 71% (p = 0.02), respectively, was observed (Figure 4B). The increase in CD4+CCR5+ T cells was mainly seen in the EM (baseline: 36%, 24 weeks: 82.2%, p = 0.02) and CM subsets (baseline: 18%, 24 weeks: 62.6%, p = 0.003). In contrast, a variable response to treatment was observed in patients with baseline CD4+CCR5+ frequency above the median of 40% (Figure 4C). The same trend (although not statistically significant) could be observed for the actual count of CD4+CCR5+ T cells with those patients having a median baseline value below 40% showing an increase in their CD4+CCR5+ T cell count after treatment (baseline: median 696, 24 weeks: median 7165, p<0.05). Additionally there was no change observed in the frequency and actual count of CD8+CCR5+ T cells in the sigmoid colon (baseline: 93% and median: 16812, 24 weeks: 93% and median: 18320, p>0.05). In the peripheral blood, no significant changes in the frequency of CD4+CCR5+ T cells were observed as a response to 24 weeks of megaHAART (baseline: 9%, 24 weeks: 8.3%, p>0.05).
The size of the HIV reservoir was determined by quantifying the total number of HIV DNA copies per 106 cells in PBMCs and in gut mucosal tissue. Data on HIV DNA in PBMCs were available for 29/30 subjects at baseline and 15/15 at week 24. The total HIV DNA in PBMCs was significantly higher in subjects at Fiebig III (median 550 copies/106 PBMCs, n = 15) and Fiebig II (median 96 copies/106 PBMCs, n = 5) compared to those at Fiebig I (median 8 copies/106 PBMCs, n = 7) AHI, p = 0.01 (Figure 5A). As more than 95% of the reservoir is harbored by CD4+ T cells (Chomont N, unpublished data), we estimated the frequency of infection of CD4+ T cells in the blood based on their percentage among PBMCs (measured by flow cytometry). We saw similar results when infection frequencies in CD4+ T cells were used (Figure 5B) rather than infection frequencies in PBMCs. The median total HIV DNA were significantly higher in Fiebig III (2308.5 copies/106 CD4) and Fiebig II (319.7 copies/106 CD4) compared to Fiebig I (20.5 copies/106 CD4) subjects (p = 0.01 and 0.02 respectively). The total HIV DNA in PBMCs (Figure 5C) and in CD4+ T cells (Figure 5D) at baseline predicted reservoir size at week 24, (ρ = 0.8, p = 0.0002). After megaHAART, the median total HIV DNA declined from 1513 copies/106 CD4 at baseline to 106 copies/106 CD4 at week 24 (p = 0.002).
By week 24, acutely treated subjects achieved total DNA levels (median 40 copies/106 PBMCs) lower than those in virologically suppressed, chronically HIV-infected patients on long-term standard HAART (median 109 copies/106 PBMCs, n = 14), and 9 patients had levels as low as those seen in elite controllers who maintain undetectable viremia without ART (median 4.5 copies/106 PBMCs, n = 13) (Figure 5E). Three of 15 acutely treated subjects had undetectable total HIV DNA in PBMCs.
HIV DNA data in gut tissue were available in 19/30 subjects at baseline and 7/15 at week 24. The median total gut HIV DNA was 336 copies/106 cells in Fiebig III (n = 11) and 10 copies/106 cells in Fiebig I (n = 5) (p>0.05). The total HIV DNA in gut among the 7 subjects with week 24 results decreased from 319 copies/106 cells at baseline to 39 copies/106 cells at week 24 (p = 0.047).
Plasma inflammatory biomarkers.
Significantly lower levels of plasma IFNα (Figure 6A), IL-17 (Figure 6B) and interferon gamma-induced protein (IP-10) (Figure 6C) were seen following treatment. Of the inflammatory markers measured, only D-dimer showed a significant reduction following 24 weeks of megaHAART whereas plasma LPS, CRP, sCD14 and HA concentrations were not different from baseline (Figures 7A to 7E).
At the incipient stages of HIV infection, there is a massive loss of GALT-associated CD4+ T cells in animal models of simian immunodeficiency virus (SIV) and HIV in humans as a consequence of direct infection , , . Increased gut permeability and chronic activation of CD4+ T cells by bacterial products further potentiates T cell loss , , . The clinical course of acute CRF01_AE infection has not been described before. In our study, we identified subjects in the earliest stage of infection (90% were Fiebig I-III) and evaluated the immunologic and virologic changes in both peripheral blood and gut before and after aggressive ART. We demonstrated that gut T cell depletion and HIV DNA reservoir size increased as Fiebig stage progressed, and gut total HIV DNA correlated with the depletion of gut CD4+ T cells. Importantly, the amount of total HIV DNA at entry into the study predicted reservoir size after treatment. These findings favor early intervention during AHI to limit immune destruction and HIV reservoir size, and also highlight that immune destruction begins in the earliest days after infection. The latter finding raises concerns regarding the interpretation of these data in the context of protection when considering treatment in later stages of primary infection .
It is thought that the early depletion of the GALT, the largest reservoir of CD4+ T cells in the body, is a blow from which the host may not recover even after prolonged ART in the chronic phase of infection , , , . We employed a strategy that blocked HIV at three steps in the viral life cycle -at entry (CCR5 inhibitor), reverse transcription [nucleoside reverse transcriptase inhibitors (NRTIs) and non-NRTI) and integration (integrase inhibitor)-and found a marked reduction in viral burden in both gut and plasma HIV RNA and DNA. The extent of HIV DNA reduction after 6 months of therapy exceeded that achieved in chronically infected patients following almost 5 years of conventional three-drug ART. Importantly, in persons whose gut CD4+CCR5+ T cells were depleted, megaHAART was associated with reconstitution of gut CD4+CCR5+ T cells to the normal range. This has not been described in published data of macaques or humans treated with standard HAART which may reflect ART timing and/or regimen , , , . Lower CCR5 may reduce the replication of HIV by the fusion inhibitor enfurvitide, and a similar mechanism may be at work for MVC . CD4+ CM T cells are preserved in elite controllers and natural SIV hosts suggesting that maintaining CD4+ CM T cells and limiting HIV integration may be crucial to achieving drug-free control of HIV replication , , . Our data showed that both CM and EM CD4+CCR5+ T cells are predominantly depleted in the gut during AHI, and institution of ART during Fiebig I to III reconstituted these subsets. Decreased IFNα and IL-17 levels following treatment may be a consequence of the diminished viral reservoir leading to reduced immune activation, but further studies are needed . Damage to the gut mucosal barrier leading to microbial translocation has been proposed as a cause of ongoing immune activation despite successful ART in chronic HIV infection reflected by elevated inflammatory biomarkers such as those investigated in our study , . These markers have been linked to HIV progression and complications in chronic HIV infection , . In contrast to the cytokine rise seen prior to peak viremia , elevated levels of markers of gut microbial translocation have not been consistently reported in the acute HIV infection period . Our patients were enrolled in early acute HIV infection when significant gut microbial translocation may have not yet occurred. Supporting evidence were the lower levels of inflammatory biomarkers in our patients compared to those reported in chronically HIV-infected Thais , and the lack of significant reductions, except for D-dimer, following treatment despite a profound decline in plasma viremia .
Results have been mixed with respect to the benefit of initiating treatment during AHI , possibly owing to varying definitions of AHI and timing of ART initiation. In our study, patients initiated treatment in the earliest stage of infection, most at Fiebig stages I-III. Further, we used an aggressive approach with five antiretrovirals, affecting both HIV entry and integration. The integrase inhibitor accelerates HIV viremic decay , and the CCR5 inhibitor could be beneficial in AHI since transmitted/founder HIV virurses are almost exclusively CCR5-tropic . A randomized comparison of megaHAART versus standard HAART to discern the benefit of using these additional drugs is now being implemented in our study, and is the subject of ongoing investigation by others .
AHI subjects are 20-30 times more likely than their chronically-infected counterparts to transmit infection , , and early treatment could avert new infections . Our study provided evidence that identifying AHI subjects by NAT and sequential EIA, and enrolling them in a study is feasible but technically and logistically challenging and costly, . Strengthening the awareness of symptomatic AHI is a less challenging way to identify acute and recent HIV infection making early treatment far more possible.
Our study has limitations such as the relatively small number of subjects with AHI despite screening a large number of VCT clients and the lack of comparator groups, namely, a group in which treatment was not instituted and a group that received standard HAART. This limits our interpretation of the effect of megaHAART on HIV reservoir size and immune restoration. Additionally the gut biopsy was done at the sigmoid site only. Yukl et al reported differences in cells harboring HIV DNA across gut sites; the amount of HIV DNA appeared to increase from duodenum to terminal ileum, right colon and rectum. In that study, sigmoid colon was not studied but the rectum was described as a major site for the persistence of cells harboring HIV DNA . Yukl et al normalized viral loads to 106 CD4+ T cells, and our data do not directly address the differences in assays. Although we cannot exclude that other sites of the GI tract might contribute to HIV persistence, the serial increase in HIV DNA from duodenum to rectum would suggest that sigmoid might be more akin to the latter, and we believe that the measurement of viral DNA in longitudinal sigmoid biopsy provides novel information by indicating that early ART greatly reduces the reservoir size in this compartment. Furthermore, we did not include information on integrated DNA as the values were extremely low and difficult to interpret for the small number of patients. The assay for 2LTR (long terminal repeat) circles is currently being optimized for CRF01_AE. Transient increases in 2LTR circles could occur as a result of raltegravir intensification . Therefore, the total HIV DNA reported in our study may reflect non-integrated HIV DNA. In addition, we did not measure cell-associated HIV RNA (unspliced and multiply-spliced). Decreases in unspliced HIV RNA of ileal CD4+ T cells were reported in some patients who received raltegravir intensification . Taken with these observations, the reduction of total DNA in sigmoid MMC at 24 weeks could be a lower bound estimate for integrated DNA.
Studying the immunologic and virologic changes during AHI could provide important insight that will inform the design of immunogens for preventive vaccines . There may be a window of opportunity during early AHI to intervene and limit CD4 destruction and HIV reservoir formation with the ultimate goal of achieving drug-free remission of HIV. In addition to ART during early Fiebig stages, strategies such as therapeutic HIV vaccines or drugs that target the long-lived cellular reservoir may be necessary to achieve this goal ,  - and persons aggressively treated in these early Fiebig stage infections may be ideal candidates for these interventions. In order to test for HIV functional cure, treatment interruption will be necessary. The risk for serious non-AIDS events due to treatment interruption seen in the SMART study is a concern ; however, this may be mitigated in early-treated AHI subjects who have preserved immunity and limited HIV reservoir , provided it is done in the setting of a closely monitored clinical trial.