icon-    folder.gif   Conference Reports for NATAP  
  18th CROI
Conference on Retroviruses
and Opportunistic Infections
Boston, MA
February 27 - March 2, 2011
Back grey_arrow_rt.gif
Chronic Inflammation and HIV: CROI 2011
  David H. Shepp, MD
Associate Professor of Medicine
Hofstra North Shore-LIJ School of Medicine
"A large number of abstracts addressing the causes and consequences of chronic immune activation and inflammation in HIV appeared at CROI this year.....inflammation & CVD.....role of innate immunity.....ART interruption/ inflammation.....HIV/endothelial function-effects of coagulation/inflammation.......suggested reduced clotting ability.....immune activation/incomplete immune recovery......causes: microbial translocation, coinfections, residual viremia, ART......Treatment: early HAART, maraviroc, raltegravir, Hydroxychloroquine"
18th CROI Conference on Retroviruses and Opportunistic Infections.....so far well over 100
February 27 - March 2, 2011
Boston, MA reports
Chronic inflammation contributes to the pathogenesis of many chronic medical illnesses, including cardiovascular and cerebrovascular disease (CVD), diabetes, chronic kidney disease, osteoporosis and cancer. These same illnesses now account for most morbidity and mortality in persons with HIV receiving virologically suppressive antiretroviral therapy (ART). Mounting evidence suggests these illnesses occur more often and/or at a younger age in HIV due at least in part to chronic inflammation and accompanying changes in immunologic phenotype and function that suggest accelerated aging. Untreated HIV induces a state of immune activation and resulting chronic inflammation that is only partially corrected by ART. The pathogenesis of HIV-related chronic inflammation is complex and incompletely understood, and no treatment beyond standard ART has been established. Studies of the pathogenesis and treatment of chronic inflammation represent a critical unmet medical need that is now receiving attention in the HIV research community, as indicated by an increased number of abstracts on the topic at CROI 2011.
Inflammation and CVD. Epidemiologic studies have shown HIV infection confers an added risk of CVD. Cohort studies have found an association between serum markers of inflammation, T-cell activation and CVD events or surrogate markers of CVD such as carotid intima-media thickness (cIMT). However, CVD pathogenesis is complex and in these types of analyses, fully controlling for other CVD risk factors, especially smoking, has been difficult. In a case control study of exclusively non-smoking patients, Desvarieux et al. (abstract 803) examined the relation between cIMT and panels of both inflammatory (hsCRP, LI-6, resistin, insulin, IL-18, amyloid A, d-dimer) and anti-inflammatory (adiponectin, IL-27, IL-10) markers among ART naive, ART-treated and HIV-negative controls. After adjustment, a low anti-inflammatory cytokine profile was more strongly associated with cIMT thickness than a high inflammatory profile. Duration of HIV infection but not duration of ART was also independently associated with cIMT thickness. Hunt et al. (abstract 814) used flow-mediated dilatation (FMD) of the brachial artery, an indicator of vascular endothelial dysfunction and a surrogate marker of cardiovascular disease, to examine the relation between CVD and immunologic markers. Higher levels of CCR-5 expression on T-cells significantly correlated with lower FMD, while traditional T-cell activation markers (CD38+/HLA-DR+) did not. Gag-specific CD8+ T-cells also correlated. although the clinical and research implications of these studies is uncertain, they suggest the complexity of chronic inflammation and immune activation in HIV and contribute incrementally to the understanding of which components of inflammation associate with CVD in HIV-infected patients.
Role of Innate Immunity. Much research into the mechanism of chronic inflammation in HIV has focused on activation of the adaptive immune system. However, innate immunity also has an important role. Abnormal function of antigen-presenting cells (APCs) was identified as a potential contributor to chronic inflammation (Nagy, abstract 316). The function in vitro of purified CD11c+ peripheral blood cells from ART naive, ART treated and HIV-negative controls were compared. The proportion of APC isolated and phagocytic function were similar across groups. However, when compared to HIV-negative controls and patients on ART, APCs from the ART naive patients produced greater TNF alpha in response to purified toll-like receptor (TLR) 2 & 4 ligands and to pathogenic strains of Salmonella. Interestingly, ART naive patients also produced more IL-6 and both HIV-infected groups produced more TNF alpha in response to a non-pathogenic strain of lactobacillus. These observations suggest abnormal APC hyperactivity make contribute to chronic inflammation in HIV.
Non-calcified atherosclerotic plaque may be more vulnerable to rupture also more frequent in HIV-infected persons. Activated macrophages are present in plaque and are thought to play a role in rupture. A novel marker of macrophage activation, soluble CD163, was studied in relation to non-calcified coronary plaque, as assessed by multi-detector CT, in HIV-infected patients (most of whom were on ART) and HIV-negative controls (Burdo, abstract 813). None had known CVD. sCD163 was found to be elevated in HIV-infected vs uninfected and in the presence of non-calcified vs. calcified plaque. sCD163 remained significantly associated with plaque after adjustment for demographics and traditional CVD risk factors. Other markers of inflammation such as hsCRP, IL-6 and d-dimer were not significantly associated. The more conventional macrophage activation marker sCD14 (a receptor for LPS) was higher in HIV-infected patients, but levels were not clearly associated with non-calcified plaque.
The SMART study previously demonstrated strong associations between the drug conservation (treatment interruption) strategy, viral replication and elevated plasma levels of certain inflammatory cytokines. SMART investigators used baseline and month 2 specimens to measure levels of an extended panel of cytokines and found ART discontinuation was strongly associated with elevations of IL-6, TNF alpha, IL-10 and CXCL-10 (aka IP-10) all of which are products of macrophage activation (Cozzi-Lepri, abstract 301).
These studies extend knowledge of the important role of the innate immune system in the pathogenesis of chronic inflammation in HIV. The novel marker sCD163 may have a role in identifying patients at high risk for CVD that needs further confirmation.
Effects an coagulation . Complex interactions exist between the immune system and the coagulation cascade. In HIV infection, elevated levels of d-dimer can be attributed to immune activation. D-dimer is a breakdown product of fibrin. Elevated levels are usually interpreted to mean a state of increased coagulation is present. Previously published data suggests HIV may be associated with a pro-thrombotic state that contributes the excess CVD risk. In the presence of endovascular injury, lysis of clot is an important defense against completed vascular occlusion and subsequent infarction. Tissue-type plasminogen activator (t-PA) is a key trigger for fibrinolysis. Diehl et al. (abstract 802) studied t-PA release from vascular endothelium in HIV-infected patients (median age 36) and two groups of HIV-negative controls (median age 39 and 68, respectively) using intra-brachial bradykinin infusion. HIV-infected subjects had significantly lower t-PA release than the younger group of controls, levels that approximated those in the older control group. The difference were seen irrespective of ART treatment status. This study suggests that the phenomenon of premature aging extends to endothelial function and may reduce the ability to lyse clots, predisposing HIV-infected persons to completed vascular thromboses.
Investigators from the INSIGHT/SMART study group examined the effect of initiating or stopping ART on a panel of plasma fibrinolytic, procoagulant and anticoagulant factors (Baker, abstract 811). The results suggest that untreated HIV is associated with an inflammatory, procoagulant state with reduced levels of factors synthesized in the liver. The effect of ART on pro- and anti-coagulant factors is mixed, but increases in hepatic synthesis of anti-thrombin III, protein C, protein S, favor a shift away from the pro-coagulant state.
Despite the suggestion that HIV-infection may be a prothrombotic state, the effects of HIV on coagulation are complex. Investigators from UCSF examined a panel of HIV-infected patients, including ART naives, those on ART with or without virologic suppression, and HIV-negative controls (Hsue, abstract 797). In the HIV-infected groups, the finding of elevated levels of anti-thrombin III (in the ART groups) and lower endogenous thrombin potential (all HIV groups) compared to HIV-negatives suggest reduced clotting ability.
Suboptimal immune recovery. The immunologic non-responder (INR) - a patient with good virologic suppression on ART but poor CD4 recovery - is commonly encountered in clinical practice. When compared to patients with similar CD4 counts but without viral suppression, INRs have a greatly reduced risk of AIDS-related illness, but when compared to immunologic responders they are at increased risk for CVD and other non-AIDS illnesses. Previous studies have found higher levels of immune activation in INRs compared to those with a greater CD4 response. Several abstracts presented data further exploring immune system abnormalities that accompany immunologic non-response. Shive et al. studied patients on virologically suppressive ART for ≥ 2 years and compared INRs (absolute CD4+ <350) to responders (CD4+ >500) and to HIV-negative controls (abstract 320). CD4+ and CD8+ T cell activation markers, IL-6 and sCD14, but not endotoxin, were all significantly higher in INRs compared to both responders and negative controls. D-dimer did not differ between responders and non-responders, but both groups had elevated values compared to controls.
Because of a much higher prevalence of co-infections such as tuberculosis, malaria and intestinal helminths, the relation between immune activation and CD4+ recovery might be different in Africans receiving ART. Nakanjako, et al. (abstract 572 ) studied Ugandan patients, all with advanced HIV disease at baseline, who achieved virologic suppression on ART for 4 years. They found levels of activation and apoptosis markers in both CD4+ and CD8+ T-cells that correlated inversely with the magnitude CD4+ response. Large differences in levels of these markers were found when the lowest quartile of CD4+ response was compared to the middle two quartiles. Smaller but still statistically significant differences also existed between the middle two quartiles and the highest CD4 response quartile.
Levels of T-cell activation markers were found to distinguish responders from INRs in patients from the SCOPE cohort at UCSF (Hatano, abstract 540). Patients on long-term suppressive ART were divided into those achieving or not achieving a CD4 count >350. Activation (CD38+/HLA-DR+) and apoptosis (PD-1) markers were significantly higher in INRs. These differences were especially marked in the central memory (CD45RA-/CCR7+) CD4 T-cell compartment. Disruption of normal lymphoid tissue architecture is another potential cause of poor immunologic recovery. A small subset of patients from the SCOPE cohort study (Hunt, abstract 319) also underwent rectal biopsies to assess the extent of fibrosis in gut-associated lymphoid tissue (GALT). The percent of fibrosis observed in lymphoid aggregates correlated inversely with CD4+ count and directly with CD4+ activation markers. Correlation of fibrosis with total CD8+ or activated CD8+ was poor but gag-specific CD8+ cells correlated inversely with fibrosis. Although this study was very small (n=9) and needs confirmation, it is consistent with the hypothesis that lymphoid tissue fibrosis disrupts the immune reconstitution process and is associated with abnormal T-cell activation.
There are multiple potential drivers of the chronic immune activation in HIV infection. Antigenic stimulation by HIV itself, translocation of microbial products from the gut, persistent or reactivated co-infections, or effects of antiretrovirals may all be contributing factors.
Microbial Translocation. Translocation of microbial products from the gut lumen into the portal and systemic circulation has been identified as an important component of HIV pathogenesis in untreated patients and a cause of persistent chronic inflammation after virologically successful ART. Defects in immune reconstitution of GALT are thought to be permissive for microbial translocation (MT). Markers of MT include elevated plasma levels of endotoxin, LPS, and 16S rDNA. Two abstracts identified novel markers of MT. Anti-flagellin antibody levels were markedly elevated in untreated HIV-infected patients from Vietnam, Ethiopia and Sweden (Abdurahman, abstract 313). Levels were significantly higher in patients from the developing countries, but patients from Sweden still had elevated levels compared to HIV-negative controls. Levels declined but did not normalize after 2 years of ART. Antibodies to a panel of microbial products often used in the diagnosis of inflammatory bowel disease (IBD) were assessed in patients with advanced HIV (CD4 <300) and elevated levels of LPS (Kamat, abstract 315). Positive results for at least one antibody were found in 65%, and 46% had a pattern of results considered consistent with IBD. The effects of MT in HIV may depend not only on immunologic defects in GALT but also on altered composition of gut flora. Ellis et al. used molecular techniques to classify bacteria in stool of HIV-infected patients beginning ART and HIV-negative controls (abstract 317). Preliminary results suggest that compared to controls, there are substantial differences in the composition of gut flora in HIV-infected patients, with an increased prevalence of certain bacterial orders with high LPS content and greater TLR-stimulating potential.
Residual viremia. Although virologic suppression below the level of detection of standard clinical viral load assays is associated with good long-term ART efficacy, ongoing viral production detectable by more sensitive assays is common and persists for years. This low level viremia is a potential source of immunologic stimulation. Controversy exists as to whether low level viremia represents ongoing cycles of replication that might be amenable to antiretroviral intensification or persistent release of virus from long-lived cells infected prior to initiation of ART. Because of its potency, safety, tolerability and lack of cross resistance with other antiretrovirals, raltegravir (RAL) is an attractive candidate to test intensification. Previous studies of RAL intensification have produced unclear results. ACTG investigators reported results of a randomized, placebo controlled 12-week cross-over design study of RAL intensification of standard ART (Gandhi, abstract 51). They were unable to demonstrate any effects on viremia as detected by a single copy assay, proviral DNA, 2-LTR circles (a sensitive marker of ongoing viral replication cycles) or CD4+ and CD8+ activation markers. The strength of this study is the controlled, cross-over design, but a limitation is the relatively short duration of treatment. Two other abstracts with a less rigorous design left open the possibility of improvement in immune activation with longer periods of intensification. Lichtenstein et al. added RAL to standard ART in a group of INRs and after one year, saw declines from baseline in CD4+ activation markers and CD4+ percent (but not absolute CD4) as well as improvement in a variety of other cytokines (abstract 276). The lack of a control group in this study makes it difficult to be certain the observed effects are due to additional of RAL. Massenella et al. (abstract 281) also conducted a single-arm study of RAL intensification for 48 weeks. Among patients with detectable 2 LTR circles at baseline, they found higher levels of activated CD8+ memory cells (CD45RO+/CD38+ with or without HLA-DR+) at baseline, significant declines in these markers at 48 weeks, and a rebound when RAL was discontinued. These results suggest that at least in the subset with detectable 2-LTR circles at baseline, the observed changes may have been attributable to the virologic activity of RAL.
Co-infections. CMV, HCV and HBV infections are common and persistent in HIV infected persons, even after effective ART, and may contribute to chronic inflammation. Previous data has shown that CMV-specific CD8+ cells constitute a substantial portion of activated CD8+ in HIV-infected patients. Chronic hepatitis B or C are other potential drivers of ongoing immune activation, either by direct viral stimulation of immune responses or through higher levels of MT due to liver dysfunction. Marchetti et al. (abstract 936) compared markers of MT, immune activation and inflammation in patients with and without hepatitis co-infection (93% HCV). Patients with CD4+ counts <200 or acute flares of liver disease were excluded. LPS levels trended higher but did not reach statistical significance in co-infected vs mono-infected patients. Other markers (IL-6, sCD14, TNF alpha, and CD8+ activation) did not differ. During longitudinal follow-up among co-infected patient, elevated levels of TNF alpha identified patients at greater risk of developing either an ALT flare or progression of fibrosis as assessed by FIB-4 score. This study suggests hepatitis co-infection (mainly HCV) may not be an important driver of chronic immune activation in HIV, but that levels of inflammation may contribute to risk of liver disease progression.
Effects of ART. Data from the DAD cohort study, first presented at CROI 2008, suggested abacavir use may be a trigger for myocardial infarction. These findings have been controversial. Attempts to reproduce the association have produced mixed results and no clear mechanism by which abacavir might contribute to CVD risk has been established. Three abstracts at CROI 2011 add to the controversy. Leukocyte adhesion to vascular endothelium is a potential mechanism for endovascular inflammation that may predispose to MI. Using both an in vivo rat model and human cells in vitro, abacavir was found to increase leukocyte adhesion to vascular endothelium (dePablo, abstract 815). Levels of expression of mRNA for a panel of cytokine genes were compared in pregnant women randomized to ART containing either abacavir or lopinavir/ritonavir, both given with ZDV/3TC (MacLeod, abstract 816). Three genes (CCL-5, CD40LG, LTA) were found to be significantly altered in abacavir recipients, creating a pro-inflammatory pattern. A similar pattern was noted when abacavir was used to stimulate TLR 8-expressing cells in vitro. This study did not determine if the cytokines themselves were elevated in these patients or if an increased state of inflammation resulted. Finally, the FDA carried out a meta-analysis of 26 clinical trials in which a total of 9868 patients were randomized to ART employing abacavir or a comparator antiretroviral (Ding, abstract 808). This analysis did not find any relationship between abacavir use and MI.
Treatment. Activation of innate immunity plays an important role in the chronic inflammation in HIV. It may be amenable to treatment. Hydroxychloroquine (HCQ) is a licensed drug with immunomodulatory properties as a consequence of down regulation of TLR expression and interference with TLR signaling. Piconi et al. treated 20 INRs (CD4 <200 despite ≥12 months of virologically suppressive ART) with 400 mg of HCQ for 6 months in an uncontrolled single arm pilot study (abstract 382). Compared with baseline, they observed improvement in percent CD4+, decreases in the expression of proliferation marker Ki67 in CD4+ cells, decreased levels of LPS, down regulation of express of TLRs on macrophages and increases in regulatory T-cells and plasmacytoid dendritic cells. Safety was not reported, but the drug has a known safety profile in other disease states. Although absolute CD4+ T-cells were not significantly increased, the other favorable changes support further study of hydroxychloroquine in a randomized controlled trial.
Maraviroc (MVC), an anti-retroviral which inhibits HIV entry by altering the interaction of HIV envelope with the cytokine receptor CCR-5, also has immunomodulatory properties that result from its inhibition of the interaction with the receptors natural ligand, the c-c chemokines. There has been much interest in exploiting the immunomodulatory effects of MVC to impact residual, chronic inflammation in patients on ART. Wilkins et al. reported 48 week follow-up from a single-arm study of 24 weeks of MVC intensification in INRs (abstract 574). The previously reported declines in CD4+ and CD8+ activation and cell cycling markers were seen to partially reverse 24 weeks after MVC discontinuation. Absolute CD4 counts were not benefited in this trial. The absence of a control group makes it difficult to attribute with certainty the observed improvements in activation to MVC intensification, but their partial reversal after discontinuation seems to support the notion.
Another trial addressed some of the limitations of single arm intensification study design. Hunt et al. reported results of a randomized, placebo-controlled trial of 24 weeks of MVC intensification in INRs (abstract 153LB). Participants underwent virologic and immunologic monitoring that included multiple markers of immune activation in blood and rectal GALT. HIV RNA as measured by a single copy assay declined significantly but similarly in both arms, suggesting a possible improvement in adherence. Absolute CD4+ counts improved modestly with no difference between arms. In the MVC arm, CD8+ counts rose in peripheral blood but percents fell in rectal mucosa, suggesting a redistribution effect, while no changes were seen with placebo. With MVC, activated CD8+ and CD4+ percents rose and remained unchanged, respectively, in the periphery but both declined modestly with placebo. In GALT, both CD4+ and CD8+ activation percents increased about 2-fold with MVC but were unchanged with placebo. LPS levels declined and sCD14 increased in the plasma of MVC recipients, while no changes occurred with placebo. This mixture of seemingly positive and negative effects was unanticipated and the finding of increased T-cell activation contradicts previous single arm intensification studies. The investigators hypothesize that CCR-5 blockade by MVC upregulates expression of CCR-5 on cells and increases levels of CCR-5 ligands, such as MIP 1-beta. These cytokines in turn bind to alternate targets, increasing T-cell and macrophage activation and clearing LPS. Data from study participants was presented to support this hypothesis. The clinical implications of these finding are unclear and further study will be need to determine if the alterations in immunologic function seen with MVC intensification are beneficial or deleterious.
At ICAAC 2009, a provocative trial with a memorable name reported delayed progression of HIV in patients not on ART who received a multicomponent nutritional regimen of prebiotics, bovine colostrum, vitamins and antioxidants, presumably by reducing translocation of gut microbes and/or its systemic inflammatory effects. The effect of the BITE trial regimen on immune activation were presented at CROI 2011 (Clerici, abstract 383). Compared to patients receiving control nutrition, patients on the BITE regimen had lower levels of CD4+/CD25+ cells. CD8+/CD38+ cells did not differ significantly between treatment groups. Shifts in gut flora were also seen with the treatment regimen that correlated with changes in CD4+/CD25+ cells. The findings of this trial are novel and thought provoking with regard to the pathogenesis of HIV, but their clinical significance is unclear, since the regimen did not reduce viral load and adherence to a nutritional regimen is probably for more challenging than adherence to contemporary ART. It is also uncertain if the described decline in CD25+ cells is desirable since some of these cells may have regulatory function which is usually deficient in chronic HIV.
An older ART approach received renewed attention at CROI 2011 for its potential to reduce T-cell activation as well as viral load. Baev et al. (abstract 381) reported a 28 day, phase II pilot study of various doses of hydroxycarbamide (HC), also known as hydroxyurea, together with didanosine (ddI) without any additional antiretrovirals. Viral load reductions ranged from 1.2 to 1.8 logs while absolute CD4+ counts increases ranged from almost none to 135 cells, depending on the doses used. T-cell activation, proliferation and apoptosis markers were significantly reduced. These effects may be attributable to both viral load reduction and the cytostatic/antiproliferative effects of HC. No comparison to standard ART or to ddI without HC was made. HC has been previously shown to enhance both the virologic activity and the toxicity of ddI, but no safety data was presented. New strategies are needed to address the state of immune activation that persists after viral suppression is achieved, but the relevance of this strategy in the current era of safer and more potential ART is unclear.
Early initiation of ART is another potential strategy to limit the duration and magnitude of exposure to chronic inflammation. Immune activation and inflammation improve but usually do not normalize when ART is initiated in the chronic phase of HIV. Jain et al. compared residual levels of T-cell activation in patients who initiated ART within 6 months and those initiating 2 or more years after primary HIV infection (abstract 517). After a median treatment duration of a little over 2 years, both CD4+ and CD8+ activation marker levels were significantly lower in early initiators. However, CD8+ activation levels remained significantly higher and CD4+ activation levels trended toward remaining higher than HIV-negative controls. This study suggests a more beneficial effect with early ART initiation, but two years of treatment is not adequate for full normalization. Longer follow-up should determine if normalization can be achieved, but adjunctive strategies may be needed, even when ART is began very early in infection.
Conclusions. A large number of abstracts addressing the causes and consequences of chronic immune activation and inflammation in HIV appeared at CROI this year. The immunologic mechanisms are complex, incompletely understood and provide fertile ground for future research. Both innate and adaptive immunity play a role in HIV-related inflammation. Coagulation is also disturbed. Poor immunologic response to ART is associated with greater levels of residual immune activation, but no treatment strategy has proven yet proven effect for improving CD4+ counts, despite some effects on immune activation. Maraviroc has demonstrable effects on immunologic function but the clinical significance needs further exploration. Consistent effects of raltegravir intensification have not been proven. Novel therapies for immune activation such as hydroxychloroquine have begun clinical assessment. Early initiation of ART appears to provide a readily available, if not always easily implemented strategy to reduce exposure to chronic inflammation in HIV.