Despite effective antiretroviral therapy (ART), patients with HIV infection have higher morbidity and mortality rates than the general population, with a greater incidence of non–AIDS-defining cancer, cardiovascular disease, kidney failure, osteoporosis, neurocognitive impairment, and more rapid progression of chronic hepatitis C.1–3 Comorbidities associated with HIV infection have been attributed to chronic inflammation and immune system dysfunction, secondary, among other causes, to microbial translocation (MT) or low-level HIV replication in cell reservoirs or plasma.4–10
The goals of ART are to achieve a HIV viral load (VL) <50 copies per milliliter within 6 months after starting treatment and to maintain it as long as possible. Viremia control is accompanied by decrease in MT, chronic inflammation, and immune system activation parameters, although certain biomarkers remain altered for years.2,8,11,12 MT and inflammation markers behavior in patients with negative VL according to the current third-generation techniques (cutoff of 20 RNA copies/mL) is not known, and clinical benefit of a target of <20 copies per milliliter compared with other less stringent targets has not been established.13
Lipopolysaccharide (LPS) is the most widely studied surrogated marker of MT. Patients with HIV infection and high levels of LPS have an increased risk of progression to AIDS or death, irrespective of their CD4+ cell counts and HIV-VL.6 Soluble CD14 (sCD14) is produced in response to stimulation by LPS. This marker has been demonstrated to be a predictor of mortality, independent of CD4+ cell count and VL.14,15 However, because LPS is only present in the wall of gram-negative bacteria, 16S ribosomal DNA (rDNA), which is common to most bacteria, is considered a more useful marker.5 Our investigations have previously shown that detection of bacterial DNA by means of automated nucleoside sequencing is feasible in patients with decompensate cirrhosis.16 Subsequent investigations showed that the presence of bacterial DNA in biological fluids is invariably associated with its presence in mesenteric lymph nodes in an animal model of cirrhosis, confirming its usefulness as surrogated marker of MT.17
rDNA can be demonstrated in blood in most naive HIV patients. With effective ART and VL <400 copies per milliliter, there is a decrease in rDNA levels.8 However, rDNA behavior has not been studied in the setting of a strict control of HIV viremia (<20 copies/mL). Of the chronic inflammation markers, ultrasensitive C-reactive protein, tumor necrosis factor α (TNF-α), and interleukin 6 (IL-6) are those most clearly associated with mortality and development of opportunistic infections in patients with HIV.5
The aim of our study was to determine whether strict control of HIV-VL (<20 copies/mL) is associated with lower prevalence of MT, as shown by the detection of bacterial DNA, LPS, and sCD14 in blood, and lower inflammatory biomarker levels (TNF-α and IL-6) than low-level viremias (20–200 copies/mL).
Observational cross-sectional study conducted in the Infectious Disease Unit of the Hospital General Universitario de Alicante. The study was approved by the local ethics committee.
Adults with HIV infection on antiretroviral treatment and HIV-VL <200 copies per milliliter for at least 6 months signed the informed consent.
Poor adherence to treatment (self-reported intake of <95% of the scheduled dose in the past 15 days), active infection or antibiotic intake in the previous month (including prophylactic cotrimoxazole), upper gastrointestinal bleeding in the preceding month, chronic liver disease irrespective of cause, and active alcoholism.
Number of Subjects
Among 1170 patients with HIV currently being followed in the outpatient clinic, roughly 500 receive ART, consistently have a HIV-VL of <200 copies per milliliter, and do not have chronic hepatitis. Assuming an expected prevalence of translocation (rDNA) in these patients between 15% and 35% and an alpha error of 5%, the number of patients to be included was calculated as 63.
Patients who consecutively visited the outpatient clinic in November 2011 were recruited. After patients signed the informed consent, information was collected along patient interviews and from medical records. A blood sample was taken to determine MT and inflammation markers, lymphocyte counts, HIV-VL, kidney function, and lipid levels.
MT and Chronic Inflammation Study
The primary endpoint was the presence of molecular MT as determined by the presence in plasma of 16S rDNA. Secondary endpoints were plasma levels of LPS and sCD14 (markers of MT) and TNF-α and IL-6 (markers of inflammation). These markers were determined blindly (regarding patient characteristics and HIV-VL) by investigators of the Centro de Investigación Biomédica en Red (CIBERehd) of the Institution.
To detect and identify the origin of bacterial DNA fragments in blood, a broad-range polymerase chain reaction (PCR) and nucleotide sequencing analysis were performed according to the methodology previously described.16 Briefly, DNA was isolated using a QIAmp DNA Blood Mini Kit (Qiagen, Hilden, Germany), and a broad-range PCR amplification of the bacterial 16S rRNA gene conserved region was performed using the following primers: 5′-TTCCGGTTGATCCTGCCGGA-3′ as forward and 5′-GGTTACCTTGTTACGACTT-3′ as reverse. BactDNA fragments were purified using a QIAquick purification kit (Qiagen), and purified amplicons were used for the sequencing reactions with a Big Dye Terminator version 3.1 Cycle Sequencing kit (Applied Biosystems, Foster City, CA). The same reverse oligonucleotide used for PCR amplification was used as a sequencing primer. The final product was purified through precipitation with ethanol–acetate and analyzed in the ABI-Prism 310 automated sequencer (Applied Biosystems). Sequences obtained were compared with the database from the National Center for Biotechnology Information (www.ncbi.nih.gov) using the advanced BLAST search tool.
LPS levels were measured by Limulus amebocyte lysate assay (Cambrex, NJ) as previously described.18 Serum levels of cytokines and sCD14 were measured by enzyme-linked immunosorbent assay according to the manufacturer's instructions (R&D Systems Minneapolis, MN).18 HIV-VL was determined by real-time reverse transcriptase–PCR using COBAS AmpliPrep/COBAS TaqMan HIV-1 test, version 2.0 (Roche Diagnostics, NJ) according to the manufacturer's instructions. The HIV-1 RNA concentration that could be detected with a positivity rate >95% was 20 copies per milliliter.
Patients were classified into 2 groups according to whether they had low-level (20–200 copies/mL) or negative HIV-VL (<20 copies/mL). We investigated the association between HIV-VL (the primary explanatory variable) and MT and inflammation markers, possible associations between other factors and rDNA (the primary endpoint), and the differences between the 2 groups of patients. We also investigated the correlation between the different inflammation and translocation markers and between these and HIV-VL. Categorical variables were expressed as numbers and percentages, and quantitative variables were expressed as medians with 25th–75th percentiles (given the nonparametric distribution).
The χ2 test and the Mann–Whitney test were used to study the association between qualitative and quantitative variables, respectively. The Spearman coefficient was used to test the correlation between variables. To study the independent effect of HIV-VL and MT on inflammation markers, analysis of variance for 2 factors was used. Multivariate logistic regression analysis was performed for variables associated simultaneously with the main explanatory variable and the endpoint. Significance was set to 0.05. Analyses were performed using SPSS version 19.0 (Chicago, IL).
Sixty-three patients entered the study, but 4 were excluded because of HIV-VL >200 copies per milliliter and 5 because HIV-VL and lymphocyte counts were not simultaneously determined with biomarkers. The remaining 52 patients are the subjects of the study. Sixty-five percent were men, with a median age of 45 years (range, 20–70 years) and had acquired HIV for a median of 12 years earlier (range, 2–25 years), predominantly through sexual intercourse (75%). Forty percent were in stage C of the Centers for Disease Control and Prevention and had a median baseline CD4 count of 552 cells per cubic millimeter (range, 126–1640 cells/mm3). Only 3 patients had CD4 cells <200 cells per cubic millimeter. Thirteen had low-level HIV-VL (>20 copies/mL) at the time of the study, with a median of 49 copies per milliliter (25th–75th percentile, 28–93; range, 20–97 copies/mL).
Association Between HIV-VL and MT and Inflammation Biomarkers
Table 1 shows MT and inflammation markers according to HIV-VL. MT (rDNA) was observed in 46% of patients with low-level HIV-VL and in 18% of patients with negative HIV-VL (P < 0.05). Higher levels of IL-6 were also observed in patients with low level as opposed to undetectable HIV-VL (median, 90 vs 79 pg/mL; P < 0.05). Values for the remaining markers (LPS, TNF-α, and sCD14) were also higher in patients with low-level HIV viremia, but the differences did not reach significance. The correlation coefficient for HIV-VL with IL-6 was 0.3 (P < 0.05) and for HIV-VL with TNF, sCD14, and LPS were 0.2 (P = 0.2, 0.2, and 0.1, respectively). The HIV-VL values for patients with and without MT are shown in Figure 1.
In patients with negative HIV-VL, we studied if the duration of undetectability showed any correlation with the presence of MT. MT was associated with a shorter period of negative VL than in patients without MT, although values did not reach significance [9 months (range, 1–72 months) vs 20 months (range, 1–120 months)].
Table 2 shows the values of biomarkers depending on HIV-VL and rDNA. LPS, sCD14, TNF-α, and IL-6 values were significantly higher in patients with rDNA compared with patients with negative rDNA (P < 0.01). For IL-6, a cutoff of 109 pg/mL fully discriminated between patients with and without MT (rDNA).
Patients with rDNA had the higher levels of inflammation biomarkers (TNF and IL-6), without differences between low-level or negative HIV-VL. Patients with low-level HIV viremia but negative rDNA had low levels of inflammation biomarkers, suggesting that inflammation is caused by MT and not by HIV replication.
The coefficients of correlation between LPS and sCD14, TNF-α, and IL-6 were 0.5 (P < 0.01), 0.4 (P < 0.01), and 0.4 (P < 0.01), respectively, and between IL-6 and sCD14 and TNF-α were 0.4 (P < 0.01) and 0.6 (P < 0.01). The coefficient of correlation between sCD14 and TNF-α was 0.5 (P < 0.01).
Factors Associated With MT as Determined by rDNA
Table 3 shows the association between risk factors and MT as determined by rDNA. Patients with MT had a lower baseline CD4 cell count (422 vs 578 cells/mm3) and a lower nadir count (83 vs 199 cells/mm3) than patients without evidence of MT, although the values did not reach significance. Therefore, the only factor statistically associated with MT was HIV-VL.
Factors Associated With the Presence of Low-Level HIV-VL
Table 4 compares the characteristics of patients with low-level and negative HIV-VL. The groups were similar in terms of age, sex, risk factors for HIV acquisition, smoking, Centers for Disease Control and Prevention stage, CD4 count (baseline and nadir), CD8 count, and comorbidity. However, patients with low level as opposed to negative HIV-VL had a longer history of HIV infection (18 vs 10 years; P < 0.01) and were receiving nucleoside analogs less frequently (69% vs 92%; P < 0.05). The most commonly used regimen for suppressed patients was a combination of nucleoside and nonnucleoside analogs reverse transcriptase inhibitors.
Multivariate analysis was not performed because variables significantly associated with HIV-VL were not associated with MT.
In our study, patients with more rigorous control of HIV viremia (<20 copies/mL) had a significantly lower rate of MT than patients with low-level replication (20–200 copies/mL), as determined by microbial DNA (18% vs 46%). The patients of the study were very homogeneous and had good adherence to ART, persistent HIV-VL <200 copies per milliliter, and good immune status (median CD4 count, 552 cells/mm3); furthermore, they were not featured by any confounding factors, such as chronic hepatitis, intercurrent infections, or antibiotic treatment.
Regarding the inflammatory biomarkers, higher levels of TNF-α and IL-6 were observed in patients with MT (rDNA). Interestingly, low-level HIV viremia was not accompanied by increased inflammatory biomarkers in the absence of MT, suggesting that inflammation is produced by MT and not directly by HIV viremia.
It would have been interesting to determine which factors could predict the appearance of MT in patients with suppressed HIV-VL, but the study design and the number of patients did not permit this. It makes sense to think that the duration of HIV-VL undetectability affects MT and inflammatory biomarkers, although this is a hypothesis that remains to be definitively settled.
MT can be demonstrated in most HIV-naive patients and is considered the most important trigger of the chronic inflammation observed in HIV infection.14 MT is caused by depletion of CD4 cells in the gut mucosa and the gut’s increased permeability19; it is also observed in idiopathic CD4 lymphocytopenia.20 ART induces a progressive decrease in plasma levels of microbial DNA, which stabilize after several weeks but do not normalize.8 Reductions in MT and inflammatory markers are broadly related to a decrease in HIV-VL. ART itself is probably beneficial even if it does not suppress HIV viremia,8 but this relationship cannot always be proved.15 In contrast, VL rebound after interruption of ART is followed by immune activation and inflammation, with increased risk for HIV progression and cardiovascular disease.13,15
We observed that patients with low-level HIV-VL were more likely to be treated with protease inhibitors, although not to a significant degree. This fact has been repeatedly reported in the literature13; that is, patients treated with protease inhibitors more frequently have detectable HIV-VL than patients treated with nonnucleoside analogs. It is not known whether this is due to different reservoir penetration, confounding factors (such as adherence), or, more likely, selection bias (patients treated with protease inhibitors may have longer evolution of HIV infection or more previous treatment failures).
It is not fully established which is the ideal HIV-VL target to be achieved with ART. Undetectability (VL <50 copies/mL) is the recommended target in guidelines, but this goal has been defined according to the detection limits of commercial assays rather than according to clinical considerations.13 Second-generation tests have a lower quantification threshold of 50 copies per milliliter, the common cutoff for current clinical trials. Third-generation VL assays (as used in our study) have a lower quantification threshold of 20 copies per milliliter. The clinical significance of low-level viremia in terms of HIV progression and persistent inflammation is not known nor is the source of the virus (possibly sanctuary sites).21,22 In most studies, transient rebounds in viremia (blips) do not predict poorer clinical outcomes, but persistent viremia of 50–400 copies per milliliter does predict virological failure.23 The relevance of detectable HIV-VL under the detection threshold is not known, but some data suggest an increased risk of VL rebound in the following year.24 For this reason, the current ART target of 50 copies per milliliter may need to be revised downward.
Clinical trials with HIV progression or mortality as endpoints, which aim to demonstrate that tight control over viremia (<20 copies/mL) is beneficial compared with less rigorous control (20–50 or 20–200 copies/mL), are not practical, as they would require a huge number of patients with very long follow-up.
MT and inflammation markers are repeatedly implicated in HIV progression and mortality. In naive patients, LPS levels >110 pg/mL are associated with rapid progression of HIV infection, independent of HIV-VL and CD4 cell count.6 Patients in the SMART study who received episodic rather than continuous ART had higher mortality and morbidity rates (AIDS-related and cardiovascular diseases), largely associated with levels of sCD14.14 In other studies, IL-6, ultrasensitive C-reactive protein, and D-dimer have also been associated with disease progression and mortality.5 MT and inflammation markers have been implicated in the poor CD4 cells recovery in response to treatment and also in much of the morbidity associated with HIV, such as dyslipidemia, cardiovascular disease, neurocognitive impairment, osteoporosis, non–HIV-related cancers, or kidney failure.25
Our study shows that patients with HIV-VL <20 copies per milliliter have MT more infrequently than patients with low-level HIV-VL and their inflammation biomarkers are lower. Currently, there are no therapies that specifically work against MT and inflammation markers2; hence, the fact that strict control of HIV-VL reduces these biomarkers supports that we should not focus just on low viremia but also on undetectability by the current third-generation techniques. In the absence of conclusive data demonstrating the benefits of reducing MT and inflammation markers, the presence of a state of chronic inflammation is logically not beneficial to the patient, as is evident from other chronic inflammatory diseases.
Our study has some limitations. Its cross-sectional design does not allow us to establish causal relationships, merely to detect associations. If viremia was not the reason for the increased MT, improving viremic control would have no impact. However, studies on raltegravir intensification support that CD8+ T-cell activation decreases.26,27 Literature repeatedly indicates that the beginning of effective ART and a decrease in HIV-VL is followed by an evident decrease in biomarkers, so there is no reason to believe that this relationship will not be maintained even if HIV-VL is very low, as in our study.
Another limitation is the fact that the accuracy of HIV-VL quantification techniques decreases for low VL and may have led to misclassification of patients by groups according to VL. However, in a comparison of our technique with another widely used third-generation technique (Abbott RealTime HIV-1, with a quantification limit of 40 copies/mL), concordance was 90% for the cutoff of 50 copies per milliliter.28 Quantification of HIV-VL <20 copies per milliliter is possible but only for research purposes.13
Finally, our study does not distinguish between blips and persistently low viremia. Although both situations are probably different in terms of risk of virological failure, persistent viremia and blips are likely to lead to raised inflammation markers and may not be so harmless on the long run as previously thought; this aspect has not, however, been addressed in our study.
In conclusion, HIV-treated patients with negative HIV-VL (<20 copies/mL) present less frequently MT and have lower levels of inflammation markers than patients with low-level HIV viremias. Inflammation seems to be induced by MT and not by HIV viremia itself.
1. Dandekar S, George MD, Bamler AJ. Th17 cells, HIV and the gut mucosal barrier. Curr Opin HIV AIDS. 2010;5:173–178.
2. Hunt PW. Th17, gut and HIV: therapeutic implications. Curr Opin HIV AIDS. 2010;5:189–193.
3. Balagopal A, Philp FH, Astemborski J, et al.. Human immunodeficiency virus-related microbial translocation and progression of hepatitis C. Gastroenterol. 2008;135:226–233.
4. Brenchley JM, Price DA, Schacker TW, et al.. Microbial translocation is a cause of systemic immune activation in chronic HIV infection. Nat Med. 2006;12:1365–1371.
5. Nixon DE, Landay AL. Biomarkers of immune dysfunction in HIV. Curr Opin HIV AIDS. 2010;5:498–503.
6. Marchetti G, Cozzi-Lepri A, Merlini E, et al.. Microbial translocation predicts disease progression of HIV-infected antiretroviral-naïve patients with high CD4 cell count. AIDS. 2011;25:1385–1394.
7. Bentwich Z. Bacterial translocation: a useful biomarker for immune activation and disease progression. AIDS. 2011;25:1439–1441.
8. Jiang W, Lederman MM, Hunt P, et al.. Plasma levels of bacterial DNA correlate with immune activation and the magnitude of immune restoration in persons with antiretroviral-treated HIV infection. J Infect Dis. 2009;199:1177–1185.
9. Hofer U, Speck RF. Disturbance of the gut-associated lymphoid tissue is associated with disease progression in chronic HIV infection. Semin Immunopathol. 2009;31:257–266.
10. Nowroozalizadeh S, Mansson F, da Silva Z, et al.. Microbial translocation correlates with the severity of both HIV-1 and HIV-2 infections. J Infect Dis. 2010;201:1150–1154.
11. Troseid M, Nowak P, Nystrom J, et al.. Elevated plasma levels of lipopolysaccharide and high mobility group box-1 protein are associated with high viral load in HIV-1 infection: reduction by 2-year antiretroviral therapy. AIDS. 2010;24:1733–1737.
12. Wallet MA, Rodríguez CA, Yin L, et al.. Microbial translocation induces persistent macrophage activation unrelated to HIV-1 levels or T-cell activation following therapy. AIDS. 2010;24:1281–1290.
13. Doyle T, Geretti AM. Low level viraemia on HAART: significance and management. Curr Opin Infect Dis. 2012;25:17–25.
14. Sandler NG, Wand H, Roque A, et al.. Plasma levels of soluble CD14 independently predict mortality in HIV infection. J Infect Dis. 2011;203:780–790.
15. El Sadr WM, Lundgren JD, Neaton JD, et al.. CD4 count-guided interruption of antiretroviral treatment. N Engl J Med. 2006;355:2283–2296.
16. Such J, Frances R, Munoz C, et al.. Detection and identification of bacterial DNA in patients with cirrhosis and culture-negative, nonneutrocytic ascites. Hepatology. 2002;36:135–141.
17. Guarner C, González-Navajas JM, Sánchez E, et al.. The detection of bacterial DNA in blood of rats with CCl4-induced cirrhosis with ascites represents episodes of bacterial translocation. Hepatology. 2006;44:633–639.
18. Frances R, González-Navajas JM, Zapater P, et al.. Translocation of bacterial DNA from gram-positive microorganism is associated with a species-specific inflammatory response in serum and ascitic fluid of patients with cirrhosis. Clin Exp Immunol. 2007;150:230–237.
19. Estébanez-Muñoz M, Stella N, Bernardino JI. Infección por el virus de la inmunodeficiencia humana tipo 1 e inflamación persistente: ¿es la traslocación bacteriana la causa oculta? Med Clin (Barc). 2012;138:673–677.
20. Lee PI, Ciccone EJ, Read SW, et al.. Evidence for translocation of bacterial products in patients with idiopathic CD4 lymphocytopenia. J Infect Dis. 2009;199:1664–1970.
21. Yukl SA, Gianella S, Sinclair E, et al.. Differences in HIV burden and immune activation within the gut of HIV-positive patients receiving suppressive antiretroviral therapy. J Infect Dis. 2010;202:1553–1561.
22. North TW, Higgins J, Deere JD, et al.. Viral sanctuaries during highly active antiretroviral therapy in a nonhuman primate model for AIDS. J Virol. 2010;84:2913–2922.
23. Geretti AM, Smith C, Haberl A, et al.. Determinant of virological failure after successful viral load suppression in first-line highly active antiretroviral therapy. Antivir Ther. 2008;13:927–936.
24. Doyle T, Smith C, Vitiello P, et al.. Plasma HIV-1 RNA detection below 50 copies/ml and risk of virologic rebound in patients receiving highly active antiretroviral therapy. Clin Infect Dis. 2012;54:724–732.
25. Ross J, Fan C, Ross MD, et al.. HIV-1 infection initiates an inflammatory cascade in human renal tubular epithelial cells. J Acquir Immune Defic Syndr. 2006;42:1–11.
26. McMahon D, Jones J, Wiegand A, et al.. Short-course raltegravir intensification does not reduce persistent low-level viremia in patients with HIV-1supression during receipt of combination antiretroviral therapy. Clin Infect Dis. 2010;50:912–919.
27. Llibre JM, Buzón MJ, Massanella M, et al.. Treatment intensification with raltegravir in subjects with sustained HIV-1 viraemia suppression: a randomized 48-week study. Antivir Ther. 2012;17:355–364.
28. Karasi JC, Dziezuk F, Quennery L, et al.. High correlation between the Roche Cobas (AmpliPrep/COBAS) TaqMan HIV-1 v2.0 and the Abbott m2000 RealTime HIV-1 assays for quantification of viral load in HIV-1 B and non-B subtypes. Clin Virol. 2011;52:181–186.
© 2013 Lippincott Williams & Wilkins, Inc.