The introduction of highly active antiretroviral therapy (HAART) in 1996 led to a reduction in opportunistic infections (OIs), decreased mortality, and improved quality of life for those infected with HIV.1 The restoration of the immune system may also induce a subset of HIV-infected patients on HAART to develop immune reconstitution inflammatory syndrome (IRIS), however. IRIS, which has also been called immune restoration disease, is a pathologic immune recognition of antigens associated with a known replicating infection or persistent nonreplicating antigens from a prior infection. It is clinically manifested by worsening inflammatory symptoms. It occurs in HIV-positive patients who have a decrease in the HIV-1 RNA level that occurs after antiretroviral treatment2-5 and in whom symptoms develop that cannot be explained by a newly acquired infection or inflammatory condition.6 IRIS develops in 15% to 25% of patients receiving HAART.4,5,7-9 Higher incidence rates, up to 45%, have been reported when patients with known OIs are studied.10-13 Most cases develop within the first 3 months after treatment,3,7-9,14,15 although patients presenting up to 2 years after the initiation of HAART have been reported.16,17
Several studies have shown that lower CD4+ T-cell counts at the time of HAART initiation increase the risk of developing IRIS.7,8,15,18 There is little consensus among the studies on other risk factors for IRIS and optimal treatment modalities. Data are lacking on the optimal timing of initiation of HAART in the setting of concomitant treatment of OIs, because studies have shown both an increased risk of IRIS if HAART is initiated close to the time of the OI13 and no association.10 The strongest data have accumulated for tuberculosis treatment. For patients with the lowest CD4+ T-cell counts, the risk of IRIS correlates well with early initiation of HAART after specific antituberculous therapy,12,19 with decreasing risk with increasing interval.18 As HAART becomes more widely available worldwide, the incidence of IRIS is likely to increase. Further information on risk factors and ideal treatment strategies in patients with the highest acuity of illness with IRIS can help clinicians to predict who is most likely to develop severe IRIS, monitor patients adequately, and possibly prevent its occurrence. To define these factors better, we conducted a case-control study of HIV-positive patients who developed an inflammatory condition consistent with IRIS within 6 months after the initiation of HAART so as to identify clinical, demographic, and laboratory predictors of IRIS.
The Johns Hopkins HIV Clinic provides longitudinal care to more than 3000 HIV-infected patients at the Johns Hopkins Hospital in Baltimore, Maryland. As part of this clinic, a longitudinal cohort database is maintained of demographic, clinical, and therapeutic data on serologically confirmed HIV-infected patients.20 Data are abstracted from patients' medical records by trained monitors using structured instruments and are also obtained from institutional electronic databases. Using this database, we searched for patients who had a new or recurrent OI diagnosed within 6 months of starting a HAART regimen (defined as a regimen that contained at least 1 nucleoside reverse transcriptase inhibitor [NRTI] agent and a protease inhibitor [PI], boosted protease inhibitor [BPI], nonnucleoside reverse transcriptase inhibitor [NNRTI], or fusion inhibitor) or who had been prescribed prednisone or another corticosteroid. In addition, all patients who had a viral load response to HAART (of any magnitude) and who were admitted within 6 months of HAART initiation were reviewed using the electronic and paper patient records. We supplemented these search criteria with the addition of any HIV-infected patients diagnosed with IRIS who were identified by care providers in the Johns Hopkins HIV Clinic. Next, we determined through medical record review by the authors (YCM, ES) which of these patients met the case definition for IRIS. IRIS was defined as symptoms or signs consistent with an inflammatory and/or atypical presentation of OIs or tumors that was not a side effect of antiretroviral therapy, which occurred after initiation, reintroduction, or change in antiretroviral therapy in a patient who had evidence of HIV viral RNA suppression as previously defined (AIDS Clinical Trials Group [ACTG] IRIS criteria revised May 24, 2005).4 The pathogen or the process had to be identified microbiologically or histopathologically to be included in this cohort. Therefore, only definite cases of IRIS were included for study to increase the specificity of the analysis.
For cases in which the pathogen or process was identified at the time of IRIS, we used definitions of “unmasking” as follows: for Mycobacterium tuberculosis, as a focal inflammatory process that was rapidly enlarging or necrotic;21 for Mycobacterium avium disease, as a site-limited inflammatory process or rapidly enlarging lymphadenopathy;13,22 for cryptococcal disease, as atypical inflammatory disease, including inflammatory nodules in the lungs;23 for progressive multifocal leukoencephalopathy (PML), as acute worsening of inflammatory complications;13 for Kaposi sarcoma, as worsening or new presentation of visceral disease;24,25 for Pneumocystis jirovecii pneumonia (PCP) as severe pulmonary inflammation; for varicella-zoster virus (VZV), as previously defined;26 and for cytomegalovirus (CMV), as site localized and inflammatory disease.
For each case of IRIS identified, 4 controls were randomly selected from within the Johns Hopkins HIV Clinical Cohort database. Controls were matched to cases solely on the basis of initiation of HAART within 6 months of the case.
Bivariate associations between candidate variables and IRIS were determined by the Yates-corrected χ2 test (for categoric variables), the Student t test (for normally distributed continuous variables), or the Wilcoxon rank sum test for nonparametric variables. Variables examined included age, gender, race (white vs. African American), HIV transmission risk group (injection drug user [IVDU] vs. nonusers), CD4+ T-cell counts (nadir, at time of initiation of HAART, and at time of diagnosis of IRIS), HIV-1 RNA concentrations (at initiation of HAART and at the time of diagnosis of IRIS), HAART regimen (NNRTI-based, PI-based, or BPI-based), and whether patients were naive to antiretrovirals at initiation of HAART. For controls, the CD4+ T-cell count and the HIV RNA levels at the time of IRIS were imputed from values obtained at a time interval closest to the number of days after the initiation of HAART that IRIS occurred in the case patient. Conditional multivariate logistic regression analysis was used to identify independent predictors of IRIS among cases. Statistical analysis was accomplished using STATA SE version 8 (Stata Corporation, College Station, TX).
Clinical Characteristics of IRIS Cases and Controls
A total of 49 cases of IRIS were identified (35 cases from database screen and 14 cases of IRIS by a care provider within the clinic), which had occurred from November 1998 through June 2006. Clinical characteristics of the 49 cases and the 196 controls are presented in Table 1. Patients presented a median of 29 days from the initiation of HAART (range: 9 to 102 days); 63% were men and 75% were African American (see Table 1). The median nadir CD4+ T-cell count was 20 cells/mm3 (range: 2 to 351 cells/mm3). All patients were treated with regimens containing NRTIs in combination with an NNRTI, PI, or ritonavir-BPI. There were no significant demographic differences between patients from whom a pathogen was recovered and treated before IRIS compared with those from whom a pathogen or process was diagnosed coincidentally with IRIS. Only the number of days to IRIS from the initiation of HAART was significantly different between the 2 types of IRIS cases (43 vs. 21 days; P = 0.03). This may have been attributable to a difference in the number of organism-specific antigens present, because none of the coincidentally diagnosed patients had been treated for their OI.
Twenty-four of the cases were mycobacterial: 15 cases of M. avium complex (MAC) disease characterized by fever and/or symptomatic lymphadenopathy, 5 cases of M. tuberculosis (3 with cavitary pulmonary and/or disseminated disease and 2 with tuberculous meningitis), and 4 cases with other mycobacteria (Mycobacterium simiae and Mycobacterium kansasii) (Table 2). Cryptococcus-associated IRIS was also common in our study, accounting for 9 cases. Other OIs included VZV (causing herpes zoster), CMV, and PCP. Other processes mediated by viral infection, including Epstein-Barr virus (EBV)-related lymphoma, PML (JC virus), and Kaposi sarcoma (human herpesvirus-8), were also included in our case series.
The cases had a high acuity of illness; 92% of the patients were admitted to the hospital at the time of IRIS diagnosis consistent with the search strategy used to identify patients (part of the original case-finding strategy was to review records of patients who had been admitted to the hospital within 6 months of initiating HAART). Sixteen (32%) of the patients received steroids as treatment for IRIS. Five (10%) patients died as a complication of IRIS or the underlying opportunistic disease. Two patients died of brain lymphoma. One patient had culture-negative aseptic meningitis diagnosed before initiation of HAART. He was empirically treated for tuberculosis for 4 weeks, and HAART was then initiated. Twenty-two days after the initiation of HAART, he had fulminant worsening with new and rapidly enlarging inflammatory brain masses, cranial nerve palsies, and progressive obtundation. His condition was recognized as Candida meningitis by brain autopsy and culture after death. Another patient presented with colonic Kaposi sarcoma with gastrointestinal hemorrhage and hepatitis B-associated liver inflammation and jaundice despite ongoing treatment with lamivudine and tenofovir. Finally, the last patient had multiple inflammatory pulmonary nodules with Cryptococcus and acute respiratory distress syndrome, and that patient died in the intensive care unit.
Bivariate Analysis of IRIS Cases and Controls
In our bivariate analysis, cases and controls were not statistically different in the proportion of patients of male gender, black race, intravenous drug use, or prior antiretroviral use. Forty-eight of 49 patients as compared with 150 of 196 controls (odds ratio [OR] = 14.72; P < 0.001) were taking a HAART regimen anchored by a BPI, an NNRTI, or both. Cases also had significantly lower CD4+ T-cell counts (mean: 50 vs. 153 cells/mm3; P < 0.001) and significantly higher HIV-1 RNA levels (mean: 390,259 vs. 199,446 copies/mL; P < 0.001) at the time of HAART initiation. Cases were also more likely to have a >2.5-log decrease in viral HIV RNA from the initiation of HAART to the time of IRIS (OR = 3.06; P = 0.001) (Table 3). There was no significant difference in the proportion of patients versus controls who had achieved a 50-cell/mm3 increase in CD4 T cells in the same time interval. Values for controls “at the time of IRIS” were imputed as described in the Methods section.
Multivariate Logistic Regression Analysis
A conditional multivariate logistic regression analysis confirmed that the development of IRIS was independently associated with using a ritonavir-BPI (OR = 7.41, 95% confidence interval [CI]: 1.76 to 31.29), nadir CD4 count <100 cells/mm3 (OR = 5.97, 95% CI: 2.33 to 15.31), and the magnitude of the HIV RNA level decrease after the initiation of HAART (OR = 2.43 for at least a 2.5-log reduction in viral load, P = 0.05) (Table 3). Interestingly, the strength of the association increased with each quartile increase in the magnitude of viral suppression (Fig. 1). The same conditional logistic regression model was applied independently to the subset of patients with IRIS (and their controls) in whom an OI was recognized before the initiation of HAART and to the subset of patients with IRIS (and their controls) in whom the OI or process was noted coincidently. The same risk factors were identified in both groups (data not shown).
Clinical Characteristics of IRIS Cases Treated With Corticosteroids
One-third of our cases with IRIS received steroids as part of their treatment. Because of the ongoing debate regarding the efficacy and utility of corticosteroids in IRIS, we describe the course and outcome of patients given prednisone. Table 4 describes their clinical characteristics. Among the patients who received steroids, 3 had PCP (1 coinfected with MAC), 1 had large B-cell lymphoma, 1 had VZV encephalitis, and the remaining 11 patients had mycobacterial disease only (see Table 4). All these patients were treated with oral prednisone. The median time of prednisone treatment was 138 days (range: 21 to 551 days). The shortest duration of treatment was in the patient with PCP only. The longest duration of prednisone treatment was in a patient with tuberculous meningitis, who had intractable nausea, vomiting, and headache with prednisone taper and remained on prednisone until the conclusion of his antituberculous therapy, which was extended an additional month because of recurrent fever after discontinuation of antituberculous medication after 17 months of therapy. The patients with the longest durations of prednisone treatment were infected with atypical mycobacteria (M. simiae, M. kansasii, or M. avium) and had disseminated mycobacterial disease or had extrapulmonary tuberculosis that was in relatively sequestered sites (gastrointestinal and central nervous system), although repeated biopsies of these sites were not culture-positive, suggesting that the process may have been driven by antigen from dead organisms.
IRIS is a clinical entity that occurs after the initiation of HAART and measurable viral suppression. In this case-control study, we found that nadir CD4+ T-cell counts <100 cells/mL were independently predictive of the development of IRIS. This finding corroborates those of other published studies,8,13,15,27 with low nadir CD4+ T-cell count as a consistent finding in many of the larger retrospective case-control studies.
The magnitude of the viral load decline was also associated with increased risk of developing IRIS in a multivariate analysis; in general, this viral burden decrease occurred rapidly with a mean time to IRIS of 29 days after initiation of HAART. Rapid viral load decline has been identified in previous observational11,12 and retrospective studies as a risk factor for IRIS.13,15,28 The absolute drop in viremia positively correlated with increasing risk for IRIS; the greatest point estimate of risk for IRIS was found in the quartile of patients with the highest viral load decline.
Interestingly, absolute CD4 T-cell increase did not correlate with increasing risk for IRIS. This concurs with the observation of others that CD4+ T-cell increases are not found in all patients with IRIS.8,29 In 1 series, approximately 10% of MAC infections complicated by IRIS occurred in the absence of an increase blood CD4+ T-cell count.30 In contrast, other series have shown an increase in CD4+ T-cell absolute count13 and in the percentage of all T cells,10 although the acuity of illness in these studies differed from ours. It is likely that CD4+ T cells are not the only mediators of IRIS; therefore, increases in CD4+ T-cell counts with HAART do not consistently correlate with the development of IRIS. Published data suggest that with high-level viremia, HIV-1-specific CD4+ T-cell proliferation is impaired because of decreased levels of interleukin (IL)-2.31,32 Further data from the same laboratory suggest that during viremia, increased levels of monocyte interferon-α (IFNα) lead to suppression of monocyte cytokines, including IL-6. Levels of IL-6 rose with suppression of viremia.33 Independent data have shown that monocyte IL-6 levels are increased in patients with IRIS.34,35 In tuberculosis IRIS, restoration of the delayed-type hypersensitivity responses (which require T-cell and monocyte responses) and lymphoproliferation to mycobacterial antigens have been demonstrated,11,34 although a correlation with viremia has not been made.
Our report is the first to suggest that the use of the most potent HIV regimens (BPIs and/or NNRTIs) is an independent risk factor for the development of IRIS. In particular, the use of BPIs was strongly associated with IRIS. This association may be related to potent and rapid suppression of the HIV-1 RNA with this class of antiretroviral therapy. Clearly, a decrease in the level of viremia is important for the development of IRIS. In our multivariate model, however, the association with BPI use was independent of a decline in viral load. This is particularly curious, given recent data from ACTG 5160s; BPIs are not as potent as NNRTIs or the combination of an NNRTI and BPI in decreasing HIV RNA levels in the first week of therapy.36 A meta-analysis of all trials of HAART has shown that significantly higher CD4+ T-cell count increases had occurred with BPI-based regimens compared with NNRTI-based regimens 48 weeks after treatment initiation.37 Taken together with published evidence that HIV PIs have immunomodulatory effects, including antiapoptotic effects,38,39 positive effects on lymphoproliferation,40 and increases in macrophage proinflammatory cytokines such as IL-6 and TNFα,41 there may be an independent effect of BPI in the development of IRIS separate from potent HIV RNA suppression. Further research into the effect of BPI on the regulation and balance of the proinflammatory response may clarify the association with IRIS.
Finally, our report describes the problematic nature of steroid treatment in IRIS. In our cohort, patients with mycobacterial disease were most likely to receive steroids (12 of 16 patients). Three of the patients treated with steroids had PCP: the first was coinfected with MAC, the second patient had only PCP and received steroids as part of the treatment algorithm for hypoxia, and the third patient developed bronchiolitis obliterans as a consequence of PCP infection. In general, patients with the highest mycobacterial burden (eg, disseminated, often extrapulmonary, disease) had the longest courses of steroids, with recrudescence in some patients when steroids were tapered. These data point to the relative importance of the burden of antigen attributable to the opportunistic pathogen in the pathogenesis of IRIS. Furthermore, patients who had not yet initiated treatment for their opportunistic pathogen (“unmasked”) presented earlier after the initiation of HAART than patients who were already on treatment (paradoxical worsening), suggesting that untreated organism burden contributes to the immunopathogenesis of immune reconstitution inflammatory disease.
Our study has several potential limitations. First, because we included only a single urban cohort, our findings may not be more widely applicable and may reflect local prevalence of OI and HAART prescription practices. We did not systematically review all patients who had initiated HAART in our cohort, only those who were admitted within 6 months; thus, less severe or undiagnosed cases of IRIS were excluded from our study. In our study, all but 4 of the patients were admitted as a result of IRIS, and this emphasizes the high acuity of illness in our cohort. Although these cases may represent only a subset of patients who developed IRIS in our clinic, these patients are diagnostically difficult and usurp resources. This population is particularly important to study, given the high morbidity and occasional mortality. Additionally, use of steroids was a criterion for case screening; this strategy may have increased the proportion of ascertained cases that received this therapy. In focusing on these patients, however, we identified those patients who represent the most challenging clinical cases.
In summary, the development of IRIS is influenced by at least 3 major factors: rapid and significant viral load decline, burden of antigen (from dead or live organisms), and host immunity. In our study, we found that patients with the most severe immunosuppression (CD4+ T-cell nadirs <100 cells/mL) who were taking potent antiretroviral regimens (especially BPI-containing regimens) with a rapid decrease in the viral load may be at the highest risk for the development of IRIS after the initiation of HAART. Steroid treatment was difficult to taper in the patients with disseminated mycobacterial disease and/or those patients with the highest bacillary burdens. Further prospective study of IRIS and its treatment is important in defining patients at greatest risk and to optimize immunomodulatory therapies that are currently being used empirically.
The authors thank Paul Pham, PharmD, and Joel Gallant, MD, for helpful discussions regarding antiretroviral therapy and Kelly Gebo, MD, for invaluable statistical input and critical review of the manuscript. In addition, the authors gratefully acknowledge the patients of the Moore Clinic. No author has a conflict of interest to report.
1. Mocroft A, Ledergerber B, Katlama C, et al. Decline in the AIDS and death rates in the EuroSIDA study: an observational study. Lancet
2. DeSimone JA, Pomerantz RJ, Babinchak TJ. Inflammatory reactions in HIV-1-infected persons after initiation of highly active antiretroviral therapy. Ann Intern Med
3. Hirsch HH, Kaufmann G, Sendi P, et al. Immune reconstitution in HIV-infected patients. Clin Infect Dis
4. French MA, Price P, Stone SF. Immune restoration disease after antiretroviral therapy. AIDS
5. Shelburne SA, Montes M, Hamill RJ. Immune reconstitution inflammatory syndrome: more answers, more questions. J Antimicrob Chemother
6. Robertson J, Meier M, Wall J, et al. Immune reconstitution syndrome in HIV: validating a case definition and identifying clinical predictors in persons initiating antiretroviral therapy. Clin Infect Dis
7. Jevtovic DJ, Salemovic D, Ranin J, et al. The prevalence and risk of immune restoration disease in HIV-infected patients treated with highly active antiretroviral therapy. HIV Med
8. French MA, Lenzo N, John M, et al. Immune restoration disease after the treatment of immunodeficient HIV-infected patients with highly active antiretroviral therapy. HIV Med
9. de Boer MG, Kroon FP, Kauffmann RH, et al. Immune restoration disease in HIV-infected individuals receiving highly active antiretroviral therapy: clinical and immunological characteristics. Neth J Med
10. Breton G, Duval X, Estellat C, et al. Determinants of immune reconstitution inflammatory syndrome in HIV type 1-infected patients with tuberculosis after initiation of antiretroviral therapy. Clin Infect Dis
11. Narita M, Ashkin D, Hollender ES, et al. Paradoxical worsening of tuberculosis following antiretroviral therapy in patients with AIDS. Am J Respir Crit Care Med
12. Navas E, Martin-Davila P, Moreno L, et al. Paradoxical reactions of tuberculosis in patients with the acquired immunodeficiency syndrome who are treated with highly active antiretroviral therapy. Arch Intern Med
13. Shelburne SA, Visnegarwala F, Darcourt J, et al. Incidence and risk factors for immune reconstitution inflammatory syndrome during highly active antiretroviral therapy. AIDS
14. Battegay M, Nuesch R, Hirschel B, et al. Immunological recovery and antiretroviral therapy in HIV-1 infection. Lancet Infect Dis
15. Ratnam I, Chiu C, Kandala NB, et al. Incidence and risk factors for immune reconstitution inflammatory syndrome in an ethnically diverse HIV type 1-infected cohort. Clin Infect Dis
16. Buckingham SJ, Haddow LJ, Shaw PJ, et al. Immune reconstitution inflammatory syndrome in HIV-infected patients with mycobacterial infections starting highly active anti-retroviral therapy. Clin Radiol
17. Shelburne SA 3rd, Hamill RJ, Rodriguez-Barradas MC, et al. Immune reconstitution inflammatory syndrome: emergence of a unique syndrome during highly active antiretroviral therapy. Medicine (Baltimore)
18. Lawn SD, Myer L, Bekker LG, et al. Tuberculosis-associated immune reconstitution disease: incidence, risk factors and impact in an antiretroviral treatment service in South Africa. AIDS
19. Breen RA, Smith CJ, Bettinson H, et al. Paradoxical reactions during tuberculosis treatment in patients with and without HIV co-infection. Thorax
20. Moore RD. Understanding the clinical and economic outcomes of HIV therapy: the Johns Hopkins HIV Clinical Practice Cohort. J Acquir Immune Defic Syndr Hum Retrovirol
. 1998;17(Suppl 1):S38-S41.
21. Breen RA, Smith CJ, Cropley I, et al. Does immune reconstitution syndrome promote active tuberculosis in patients receiving highly active antiretroviral therapy? AIDS
22. Lawn SD, Bekker LG, Miller RF. Immune reconstitution disease associated with mycobacterial infections in HIV-infected individuals receiving antiretrovirals. Lancet Infect Dis
23. Skiest DJ, Hester LJ, Hardy RD. Cryptococcal immune reconstitution inflammatory syndrome: report of four cases in three patients and review of the literature. J Infect
24. Crane HM, Deubner H, Huang JC, et al. Fatal Kaposi's sarcoma-associated immune reconstitution following HAART initiation. Int J STD AIDS
25. Leidner RS, Aboulafia DM. Recrudescent Kaposi's sarcoma after initiation of HAART: a manifestation of immune reconstitution syndrome. AIDS Patient Care STDS
26. Domingo P, Torres OH, Ris J, et al. Herpes zoster as an immune reconstitution disease after initiation of combination antiretroviral therapy in patients with human immunodeficiency virus type-1 infection. Am J Med
27. Michailidis C, Pozniak AL, Mandalia S, et al. Clinical characteristics of IRIS syndrome in patients with HIV and tuberculosis. Antivir Ther
28. Olalla J, Pulido F, Rubio R, et al. Paradoxical responses in a cohort of HIV-1-infected patients with mycobacterial disease. Int J Tuberc Lung Dis
29. French MA. The immunopathogenesis of mycobacterial immune restoration disease. Lancet Infect Dis
30. Phillips P, Bonner S, Gataric N, et al. Nontuberculous mycobacterial immune reconstitution syndrome in HIV-infected patients: spectrum of disease and long-term follow-up. Clin Infect Dis
31. McNeil AC, Shupert WL, Iyasere CA, et al. High-level HIV-1 viremia suppresses viral antigen-specific CD4(+) T cell proliferation. Proc Natl Acad Sci USA
32. Iyasere C, Tilton JC, Johnson AJ, et al. Diminished proliferation of human immunodeficiency virus-specific CD4+ T cells is associated with diminished interleukin-2 (IL-2) production and is recovered by exogenous IL-2. J Virol
33. Tilton JC, Johnson AJ, Luskin MR, et al. Diminished production of monocyte proinflammatory cytokines during human immunodeficiency virus viremia is mediated by type I interferons. J Virol
34. Stone SF, Price P, Keane NM, et al. Levels of IL-6 and soluble IL-6 receptor are increased in HIV patients with a history of immune restoration disease after HAART. HIV Med
35. French MA. Disorders of immune reconstitution in patients with HIV infection responding to antiretroviral therapy. Curr HIV/AIDS Rep
36. Haubrich RH, Riddler S, Ribaudo H, et al. 14th Conference on Retrovirus and Opportunistic Infections, Los Angeles, 2007.
37. Bartlett JA, Fath MJ, Demasi R, et al. An updated systematic overview of triple combination therapy in antiretroviral-naive HIV-infected adults. AIDS
38. Vlahakis SR, Bennett SA, Whitehead SN, et al. HIV protease inhibitors modulate apoptosis signaling in vitro and in vivo. Apoptosis
39. Badley AD, Dockrell DH, Algeciras A, et al. In vivo analysis of Fas/FasL interactions in HIV-infected patients. J Clin Invest
40. Lu W, Andrieu JM. HIV protease inhibitors restore impaired T-cell proliferative response in vivo and in vitro: a viral-suppression-independent mechanism. Blood
41. Zhou H, Jarujaron S, Gurley EC, et al. HIV protease inhibitors increase TNF-alpha and IL-6 expression in macrophages: involvement of the RNA-binding protein HuR. Atherosclerosis
. 2007 May 23; [Epub ahead of print].