Summaries of the Sixth Conference on Retroviruses and Opportunistic Infections:
Summaries of the Sixth Conference on Retroviruses and Opportunistic Infections: Part IV
conference coverage
Note: The following summaries represent a selection of papers from those presented at the 6th Conference on Retroviruses and Opportunistic Infections held on Jan. 31-Feb. 4, 1999, in Chicago, IL. It is important to recognize that some of these summaries are extracted only from the published abstract and it is possible that some of the material presented at the conference may have differed. The abstracts and posters, as well as other information presented at the conference, are available on the internet at www.retroconference.org. —Stan Deresinski, MD, FACP
Metabolic Complications of Therapy
Hypothesis and Observations
Patients receiving protease inhibitors (PIs) are at risk of developing abnormalities of lipid metabolism resulting in hyperlipidemia and abnormal body fat distribution. The latter is associated with lipodystrophy, a loss of subcutaneous fat resulting from increased lipolysis (triglyceride hydrolysis), decreased adipogenesis (differentiation of stem cells into fat cells), and decreased lipogenesis (synthesis of triglycerides).
Retinoids are vitamin A derivatives whose systemic administration commonly causes hypertiglyderidemia. The retinoid, all-trans retinoic acid (ATRA), is bound by cytosolic retinoic acid binding protein-1 (CRABP-1), which allows its presentation to CYP3A isoforms that catalyze the conversion of ATRA to cis-9-retinoic acid. The latter is the only known ligand for the retinoid X receptor (RXR); RXR regulates the synthesis of gene products involved in fat metabolism via interactions with peroxisome proliferator activated receptor gamma (PPAR-g) and retinoic acid receptor (RAR). In essence, ligand binding to RXR or PPAR-g inhibits apoptosis of adipocytes and upregulates adipogenesis preferentially in peripheral rather than central fat.
A 12-amino acid sequence in the HIV-1 protease has approximately 60% homology with a C-terminal region of CRABP-1.1 In vitro studies have demonstrated that some PIs displace ATRA from CRABP. This may result in decreased cis-9-retinoic acid and, consequently, impairment of RXR regulation of fat metabolism. Furthermore, there is increased binding of ATRA to RAR as a consequence of its displacement from CRABP-1. RAR agonists are known to prevent adipogenesis and to cause dyslipidemia in rats and, as indicated above, pharmacologic administration of retinoids results in dyslipidemia in humans. It is of interest that retinoids also cause epidermal abnormalities, including ingrown toenails, since indinavir administration was reported to be a risk factor for development of ingrown toenails. (Abstract 681.)
An alternative mechanism by which PIs lead to lipodystrophy may be interference with RXR signalling. In vitro studies found that nelfinavir, saquinavir, and ritonavir each inhibited adipogenesis of murine cell lines and increased lipolysis in vitro. These effects were not seen with amprenavir or indinavir. Nelfinavir, saquinavir, and ritonavir each inhibit retinoid effects in vitro by inhibiting signalling of RXR-dependent pathways. Indinavir, in contrast, appears to cause displacement of ATRA from CRABP, making ATRA available to activate RAR-responsive genes. (Abstracts 665, 666.) These complex results are consistent with the hypothesis that PI-related lipodystrophy and lipid abnormalities may involve altered retinoid signalling, but that the mechanism and even occurrence of this phenomenon may differ among the PIs. (See Figure.)
To make things even more complicated, a number of patients, both male and female, with lipodystrophy who were being treated with only RT inhibitors were described. None, however, developed significant hyperlipidemia. (Abstracts 653, 654, 660, 662.)
Changes in serum cholesterol and triglycerides were retrospectively examined in patients treated with an indinavir-containing regimen or with NRTIs alone. Indinavir therapy was associated with an increase in mean cholesterol at 24 weeks from 157 mg/dL to 177 mg/dL; less than 5% reached more than 240 mg/dL. Triglycerides did not significantly change. Thus, the increase in serum cholesterol seen with indinavir did not usually warrant pharmacological intervention. (Abstract 648.)
The major concern with regard to the development of hyperlipidemia is, of course, the occurrence of atherosclerotic vascular disease. Three studies examined coronary artery events in a total of almost 5000 patients receiving PI therapy; none found a statistically significant increased risk. (Abstracts 656-658.) Continued observation with longer follow-up is, however, necessary to definitively assess this issue.
Glucose metabolism may also be affected by PI therapy. A retrospective analysis found that, on average, PI therapy had no effect on glucose levels but was associated with a mean increase in serum cholesterol of 31 mg/dL and in triglycerides of 111 mg/dL. (Abstract 644.) A French study reported that 30% of patients receiving a PI were glucose intolerant, a frequency much greater than that previously reported by other investigators. (Abstract 642.)
Management
Many patients who have developed lipodystrophy or other adverse effects while receiving PI-based regimens have requested a change in therapy. Seventeen NNRTI-naïve patients with plasma HIV RNA less than 200 copies/mL on a PI-containing regimen had the PI replaced with a NNRTI (15 nevirapine, 2 efavirenz). After a mean follow-up of eight weeks, all patients had plasma HIV RNA less than 80 copies/mL and no change in mean CD4+ T cell count. Improvement in fat maldistribution was reported to have occurred in all patients. (Abstract 381.)
Discontinuation of PI therapy with substitution of nevirapine for six months was associated with improvement in metabolic and fat distribution abnormalities in most patients while plasma HIV RNA remained less than 2.6 log10. There was no benefit observed in switching from another PI to nelfinavir. (Abstract 668.)
Another 23 patients with plasma HIV RNA less than 200 copies/mL and lipodystrophy elected to discontinue their PI and substitute nevirapine. After a median follow-up of seven months, there was a 21% decrease in serum cholesterol, a 56% decrease in triglycerides, and a 16% decrease in fasting glucose. Twenty-one (91%) reported a partial improvement in body habitus. One patient developed a detectable viral load (546 copies/mL). (Abstract 670.)
Thirteen NNRTI-naïve patients with plasma HIV RNA less than 500 copies/mL, but with metabolic abnormalities while receiving indinavir, had their PI replaced with efavirenz. In the six patients who had reached 12 weeks, switching to efavirenz was associated with a mean 2.1 kg increase in weight, and a mean reduction in abdominal circumference of 1.7 cm. However, fasting serum cholesterol increased from 6.15 mmol/L to 8.9 mmol/L (P < 0.05) and in triglycerides from 6.7 mmol/L to 10.7 mmol/L. Viral load remained undetectable in five of five. Thus, switching from indinavir to efavirenz may modestly improve body habitus but may also further elevate lipid levels. (Abstract 669.)
A European study also found that patients with lipodystrophy on virologically successful PI-containing regimens who switched to ddi/d4T/nevirapine maintained viral suppression, reduced cholesterol from a mean of 230 mg/dL to 196 mg/dL, and improved self-reported perception of abnormal fat distribution when evaluated at 12 weeks. (Abstract LB14.)
Nineteen patients meeting National Cholesterol Education Program criteria for drug intervention were treated with gemfibrozil 600 mg bid and atorvastatin, starting at 10 mg qd, with careful escalation and monitoring of CK. At six months, cholesterol decreased by a mean of 30% and triglycerides by a mean of 60%. Despite potential pharmacokinetic interactions of atorvastatin with PIs (predicted increase in atorvastatin levels), no toxicity was observed. (Abstract 671.)
In a controlled randomized trial, metformin administration for eight weeks to patients with central adiposity while receiving a PI was associated with a mean weight decrease of 2.1 kg, mean decrease in waist-to-hip ratio from 0.91 to 0.86, and a mean decrease in the visceral to abdominal fat ratio from 0.81 to 0.72. There was a significant decrease in triglycerides as well as in testosterone. Mean plasma insulin levels decreased. (Abstract 672.)
Troglitazone, a thiazolidinedione PPAR-g activator, was administered to six patients with PI-associated diabetes mellitus with improvement in glucose levels. Lipid levels initially rose, followed by a return to baseline. "Heterogenous" effects on liposystrophy were observed. (Abstract 673.)
RhHGH was administered to six patients with fat maldistribution while on PI-based therapy. Four patients completing three months of therapy had reduction in "buffalo hump" and abdominal girth but no change in peripheral lipodystrophy. There was a 5-10% gain in fat-free mass. There was no consistent effect on blood lipids. (Abstract 675.)
Bone Metabolism
Evaluation of 17 (16 asymptomatic) patients who had received HAART for a mean of 11.4 months revealed a decreased mean total body bone mineral density but with little decrease in density in the lumbar spine and hip. (Abstract 679.) Eight (1.3%) of 600 clinic patients, all receiving PIs, developed avascular necrosis of bone (mostly femoral head) during a recent 12-month period. (Abstract 680.)
Complications of HIV Infection and Their Prevention
Response to NonHIV Vaccines
A significant proportion of stable HIV-infected children older than 2 years of age with well preserved CD4% have low titers of antibody to tetanus toxoid, despite previous vaccination. (Abstract 57b.)
Thirty-one patients with baseline CD4 count of 100-300 cells/mm3 and who had then received HAART for 48 weeks, received a series of immunizations. Seventy-nine percent responded to hepatitis A vaccine. There was a marked improvement in in vitro responses to tetanus toxoid as well as to the neoantigen, KLH. The number of CD4+ memory cells was associated with responses to recall antigens while the total number of CD4+ cells and of naïve cells correlated with neoantigen responses. In contrast, there were no significant correlations with viral load. (Abstract 329.)
Patients with a mean CD4+ T-cell count of 404 cells/mm3 while receiving HAART received influenza vaccine. A four-fold or greater rise in HIV titer to influenza antigen was observed in 52.4%. Three of 11 patients with baseline plasma HIV RNA less than 50 copies/mL had a transient elevation of viral load, while all five patients with baseline viral load of 50-400 copies/mL had an increase. All three patients with baseline viral load more than 400 copies/mL had a rise in both HIV RNA and DNA. In contrast to results observed in controls, immunization against HAV with VAQTA in 90 HIV-infected individuals resulted in seroconversion in 63% at week four and 96% at week 28 and was well tolerated. (Abstract 195.)
Two doses of live attenuated varicella vaccine (Oka; Merck) were administered 12 weeks apart to 42 asymptomatic HIV-infected children (age 1-8 years) with no antibody to varicella-zoster virus and with little or no apparent immunological impairment. There were no serious adverse events, although fever occurred in 20% after the first injection and in 5% after the second. VZV antibody and VZV-specific lymphocyte proliferation were detected in 56% and 82%, respectively, after the second injection. Subsequently, eight varicella exposures were noted; one exposed vaccinee developed mild varicella. (Abstract 440.)
Mycobacterial Infection
In a randomized trial (ACTG 223), treatment of MAC bacteremia with clarithromycin/ethambutol/rifabutin was associated with improved survival and microbiological efficacy when compared to treatment with either clarithromycin/ethambutol or clarithromycin/rifabutin. (Abstract 249.) However, the response rates in all arms were extraordinarily low (40-50%), compared to other reported results, raising questions about methodology and applicability. Furthermore, these results contrast with those of a previous comparison of these two regimens (albeit with a dose of 450 mg rather than 300 mg of rifabutin daily) in which no difference was detected between the two treatment arms.
In patients who received preventive therapy with one of two equally effective regimens (INH for 12 months or PZA/RFP for 2 months) the risk factors for the development of tuberculosis were absence of antiretroviral therapy, CD4 count less than 200 cells/mm3, and a PPD skin test reaction of more than 15 mm in duration. (Abstract 448.) While the last mentioned may initially seem paradoxical, a large skin test reaction to PPD is indicative of a greater likelihood of prior exposure to M. tuberculosis than is a smaller reaction which may well be due to exposure to other antigenically-related mycobacteria.
The treatment of tuberculosis in HIV-infected patients is complicated by the potential for pharmacokinetic drug interactions between rifampin or, to a lesser extent, rifabutin and some PIs and NNRTIs. In four patients given a regimen including rifabutin (300 mg daily), mean serum nelfinavir levels two hours after a 1000 mg dose were 3.25 ± 1.17 mcg/mL (predicted: 5-11 mcg/mL). Indinavir levels two hours after a 1200 mg dose in four recipients also receiving rifabutin were 4.80 ± 0.97 mcg/mL, compared to an expected value of 2-5 mcg/mL. In 10 patients, rifabutin levels increased from baseline after introduction of HAART (from 0.12 + 0.03 mcg/mL to 0.22 ± 0.03 mcg/mL), although they remained below suggested target levels of 0.3-0.9 mcg/mL. Therapy of tuberculosis was successful in 28 of 28 patients and viral load decreased to less than 400 copies/mL in 133 (46%) of 28. Thus, it appears that rifabutin can be safely and effectively used in patients receiving indinavir, but that use of nelfinavir may be associated with low plasma concentrations of the PI. (Abstract 449.)
Mycoses
PIs inhibit the production of aspartyl protease by Candida albicans at concentrations achievable in vivo and inhibit fungal replication both in vitro and in a non-HIV-infected rodent model of Candida vaginitis. (Abstract 184.) It has recently been demonstrated that this enzyme is a virulence factor for C. vaginitis.2
CD40L administration had significant antifungal effect in a CD4-depleted murine model of histoplasmosis. (Abstract 188.)
A retrospective review of 1660 cases found that mortality was three-fold higher in patients with PCP when corticosteroids were administered according to CDC guidelines. (Abstract 697.)
Another in a series of studies has demonstrated the apparent safety of discontinuation of primary or secondary PCP prophylaxis in patients responding to HAART. In this instance, no episodes of PCP were observed after prophylaxis discontinuation in 171 patients whose CD4+ T cell count had remained above 200 cells/mm3 for more than three months and who were followed for a mean of approximately 6.6 months. (Abstract LB7.)
Cytomegalovirus Infection
In contrast to previously reported findings indicating a much higher prevalence of CMV infection, dilated fundoscopic examination of 505 patients with CD4 count less than 100/mm3 or 101-200/mm3 after a nadir less than 50 cells/mm3 found evidence of asymptomatic CMV retinitis in only nine (1.6%). The mean CD4 count of those nine patients was 21.5 cells/mm3 (range, 10-48 cells/mm3). (Abstract 453.)
Institution of HAART was associated with a mean increase in antibody titer against the major neutralizing target of CMV, gB antigen, from 1:116 to 1:1644 (P > 0.0001) and resolution of asymptomatic CMV viremia. This is consistent with previous evidence indicating that low levels of antibody to this antigen are associated with disseminated CMV infection. In patients receiving only NRTIs, CMV viral load increased, antibody levels were unchanged, and some developed CMV disease. In contrast, CMV DNA in blood became undetectable in patients receiving HAART, none of whom developed CMV disease over a mean of 14 months. (Abstract 454.)
The cumulative incidence of CMV retinitis in 172 CMV-seropositive patients who initiated PI-based therapy when their CD4 counts were less than 100 cells/mm3 was 5% at one year and 6% at two years. The only predictor of the development of CMV retinitis at the time of initiation of PI therapy was the presence of CMV viremia by PCR. The 12-month Kaplan-Meier CMV retinitis event rate was 38% in those CMV-PCR positive and only 2% in those who were negative (P = 0.002). Two-thirds of cases occurred in the first three months of PI therapy. (Abstract 251.)
Patients with CMV encephalitis commonly have ependymal enhancement detected on MRI scan of the brain. Involvement of the fornix, a periventricular efferent bundle of fibers connecting the hippocampus to the diencephalon, may also be seen in patients with this infection. Enhanced signal in the fornix, which is seen posterior to the inferior portion of the lateral ventricle, on T2-weighted brain MR scans ("fornix white line") was present in 15 (50%) of 30 patients with CMV retinitis and in 15 (20%) of 76 HIV-infected controls without retinitis. Those with moderate-to-severe neuropsychological impairment were more likely to have a fornix white line. A fornix white line of moderate-to-high intensity was present on the final MR scan in five (50%) of 10 patients with CMV encephalitis at postmortem examination but in only two (9.5%) of those without encephalitis (95% CI, 1.6-82). Thus, the fornix white line may be an early manifestation of encephalitis in patients with CMV retinitis. (Abstract 414.)
Additional studies have added to the conclusion that secondary prophylaxis may be safely discontinued in selected patients with previous CMV retinitis now receiving HAART. (Abstracts 455, 456.) Although there is, as yet, no consensus, patients who are candidates may include those on HAART whose CD4 counts have increased to more than 75-100 cells/mm3, with good control of HIV. The absence of CMV viremia is likely also important, although one report indicates that the risk of disease in CMV viremic patients receiving HAART remains low. (Abstract 458.) Recovery of neutralizing antibody to the gB antigen as well as cellular immune responses to CMV antigens may prove to be useful in the future. (Abstracts 454, 250.)
Progressive Multifocal Leukoencephalopathy (PML)
A retrospective analysis of patients with progressive multifocal leukoencephalopathy failed to demonstrate significant benefit from administration of cidofovir. (Abstract 417.)
Malignancies
Examination of chemokine receptor gene variants revealed an association between the development of nonHodgkin’s lymphoma and the SDF1-3’A variant. Heterozygotes had an approximately two-fold increased risk while homozygotes had an approximately four-fold increased risk of development of this malignancy. (Abstract 246.) On the other hand, the SDF1-3’A/3’A homozygous genotype does not appear to play a role in disease progression. (Abstract 259.)
In contrast to previous reports, thallium-201 SPECT scanning was not useful in the evaluation of CNS mass lesions in AIDS patients because of relatively low specificity and sensitivity for the diagnosis of lymphoma. (Abstract 416.)
Serological evidence of HHV-8 infection was found in 16% of males having sex with men and was associated with a greater number of sex partners, use of amyl nitrate, and a history of prior sexually transmitted disease. (Abstract 197.)
Reasonably effective systemic therapy for Kaposi’s sarcoma include ABV (adriamycin, bleomycin, vincristine) or liposomal doxorubicin and many consider the latter as first-line therapy. Paclitaxel is an effective second-line therapeutic agent for this indication. Doclitaxel is a related drug and administration of 60 mg/m2 every three weeks with G-CSF on days 4-10 was associated with an 80.5% partial response rate but no complete responses. (Abstract 204.) Topical application of alitretinoin gel was associated with responses, by ACTG criteria, in 35-49% of patients with cutaneous KS. (Abstract 205.)
HIV Encephalopathy
Matrix metalloproteinases (MMP) are gelatinases that degrade type IV collagen, a component of the extracellular matrix of the blood-brain barrier. These enzymes are secreted by microglia and perivascular macrophages. Microglia also secrete enzyme inhibitors and the balance between enzyme activity and inhibition determines the turnover of extracellular matrix. The concentrations of MMPs targeting critical blood brain barrier proteins, including collagen type IV, are increased in the cerebrospinal fluid of patients with HIV dementia. (Abstract 282.) Activated monocytes and macrophages secrete MMP-7 and MMP-9. (Abstracts 282, 283.)
Increased concentrations of MMP-9 (gelatinase B) in CSF were confirmed in another study examining HIV-infected patients and the concentrations were found to be similar to those previously reported in the CSAF of patients with multiple sclerosis or meningitis. (Abstract 411.)
Some isolates of HIV-1 induce apoptosis in primary brain cultures; this characteristic may be determined by the ability to use CXCR4 and by the V3 region of gp120. (Abstract 287.) CXCR4-mediated signal transduction may play a role in neuronal injury. (Abstracts 288, 289.)
Wasting
High-plasma viral load is associated with suppressed muscle protein synthesis and overexpression of myostatin (GDF-8), a muscle-specific growth and differentiation factor expressed by a gene in the TGF-b superfamily, whose absence in knock-out mice is associated with marked muscle hypertrophy. (Abstract 242.)
Hepatitis Viruses
In HCV-infected patients well matched by duration of disease, alcohol use, and parental risk factor, HIV infection was not independently associated with cirrhosis or progression of hepatic fibrosis. (Abstract 190.) In a separate study, coinfected patients had similar hepatic damage compared to age-matched controls with chronic HCV infection alone. (Abstract 191.)
HAART had no consistent effect on HCV infection in coinfected individuals; failure of HAART, however, was associated with a rise in plasma HCV levels along with HIV levels. (Abstracts 193, 192.)
Coinfection with GBV-C was associated with slower progression of HIV disease. (Abstract 194.)
A retrospective analysis of 63 HBV-coinfected patients who received lamivudine 150 mg bid for two years found that 57.1% had persistent clearance of HBV DNA from plasma, 14.3% never had a negative HBV DNA assay, and 33.3% cleared it but relapsed. (Abstract 196.)
References
1. Carr A, et al. Pathogenesis of HIV-1-protease inhibitor-associated peripheral lipodystrophy, hyperlipidaemia, and insulin resistance. Lancet 1998;351: 1881-1883.
2. De Bernardis F, et al. Evidence that members of the secretory aspartyl proteinase gene family, in particular SAP2, are virulence factors for Candida vaginitis. J Infect Dis 1999;179:201-208.
Subscribe Now for Access
You have reached your article limit for the month. We hope you found our articles both enjoyable and insightful. For information on new subscriptions, product trials, alternative billing arrangements or group and site discounts please call 800-688-2421. We look forward to having you as a long-term member of the Relias Media community.