icon-    folder.gif   Conference Reports for NATAP  
  14th CROI
Conference on Retroviruses and Opportunistic Infections Los Angeles, California
Feb 25- 28, 2007
Back grey_arrow_rt.gif
Osteopenia in HIV: risk factors for screening
  Session 149 - Osteopenia and Osteoporosis
Monday, February 26, 2007
Carl J. Fichtenbaum, M.D. University of Cincinnati College of Medicine

There were three papers presented in this session. In the first one, Overton and colleagues report a survey of bone density in the SUN study (Study of Unnatural Complications of HIV). This is a 4 center study of "clinic" populations. The baseline data presented here is really a cross-sectional look at where things stand. Osteopenia was very common and osteoporosis was not rare. These data suggest it may be time to screen selected persons with DEXA scans to identify those at risk for fractures. The multi-variate analyses suggest there are some factors that a clinician could use to identify who needs screening: older males (age >46 years); with low body-mass index (<22.6 kg/m2). However, before you start screening everyone, the authors need to do some more work here identifying the sensitivity and specificity of using the above criteria which, by the way, apply only to cohorts that are generally using HAART. In the second paper by Tebas and colleagues (Abstract 837), they analyzed whether there was any difference in the progression of bone loss using different ART regimens (nelfinavir vs. Efavirenz). Turns out that it does not matter. Bone loss occurs with the use of HAART and/or the continuation of HIV infection. Remember that it is hard to tease out whether persistent inflammation associated with HIV induces bone loss and/or HAART directly induces bone loss. This study suggests that no regimen is without association with bone loss. The final report by Modarressi et al (Abstract 838) was a preliminary study using an in vitro model to show that ritonavir induces genes responsible for bone resorption and this can be reversed by the addition of interferon gamma. It is not at all clear how this model applies to the in vivo situation (in patients) and much more work would need to be done to confirm these findings.
Factors Associated with Low Bone Mineral Density in a Large Cohort of HIV-infected US Adults: Baseline Results from the SUN Study

Et Overton*1, K Mondy1, T Bush2, L Conley2, E Kojic3, K Henry4, J Hammer5, K Wood6, K Lichtenstein7, J Brooks2, and SUN Study Investigators 1Washington Univ Sch of Med, St Louis, MO, US; 2CDC, Atlanta, GA, US; 3The Miriam Hosp, Providence, RI, US; 4HIV Prgm, Hennepin County Med Ctr, Univ of Minnesota, Minneapolis, US; 5Denver Infectious Disease Consultants, Rose Med Ctr, CO, US; 6Cerner Corp, Vienna, VA, US; and 7Univ of Colorado Hlth Sci Ctr, Denver, US
Background: Low bone mineral density (BMD) is a common metabolic complication associated with HIV infection. With increasing survival in the era of HAART, there is continued interest in the factors influencing low BMD in HIV-infected individuals.
Methods: The SUN Study prospectively follows a cohort of HIV-infected patients at clinics in Denver, Minneapolis, Providence, and St. Louis. At baseline, DEXA bone densitometry and body composition, clinical data, and fasting laboratory values were obtained on all subjects. Results were compared to persons from the National Health and Nutrition Examination Study III (NHANES) matched for age, race, gender, and BMI.
Results: Data were available for 562 matched pairs. Participants had a mean age of 41.0 years, 22% were female, and 27% were African American. Among SUN subjects, 79% were currently on HAART and 60% had an undetectable HIV viral load. Osteopenia (T-score -1.0 to -2.5) and osteoporosis (T-score ≦ -2.5) were identified in 262 (47%) and 59 (11%) subjects, respectively. In univariate analyses, low BMD (T-score ≦ -1.0) was associated with older age, male gender, unemployment, lower body mass index, higher visceral to subcutaneous fat ratio, duration since HIV diagnosis and any stavudine use (all p ≦0.01). Osteoporosis alone was associated with older age, non-white race, lower body mass index, current tobacco use, unemployment, lower nadir CD4 cell count, and longer duration since HIV diagnosis (all p ≦0.05). In multivariate analyses, low BMD was associated with older age, male gender, lower body mass index, unemployment, and stavudine use; osteoporosis was associated with older age, non-white race, lower body mass index, longer duration since HIV diagnosis, and unemployment (all p <0.05).When compared to the NHANES cohort, the SUN Study cohort had significantly lower mean T-scores at the femoral neck (-0.79 vs -0.36, p ≦0.01). At this site, the prevalence of low BMD was higher in the SUN Study cohort (47% vs 31%, p ≦0.01).
Conclusions: In the HAART era, low BMD remains a significant complication of HIV infection, more prevalent than found in the general population. In this large cohort of HIV-infected persons, the majority of whom are virologically controlled, the ramifications of low BMD will become a concern as they age. HIV- and ART-related factors that may affect BMD loss warrant longitudinal study.


Initiation of ART Is Associated with Bone Loss Independent of the Specific ART Regimen. The Results of ACTG A5005s

Pablo Tebas*1, T Umbleja2, M Dubˇ3, R Parker2, K Mulligan4, R Roubenoff5, G Robbins2, S Grinspoon2, and ACTG
1Univ of Pennsylvania, Philadelphia, US; 2Harvard Univ, Boston, MA, US; 3Indiana Univ, Indianapolis, US; 4Univ of California, San Francisco, US; and 5Tufts Univ, Boston, MA, US
Background: Osteopenia is very frequent among HIV-infected individuals. Both HIV infection and its treatment play a role in its development. However, longitudinal data of the effects of initiating ART on bone mineral content are very limited. The goals of this study were to measure bone mineral content among ART-na•ve individuals, to evaluate the longitudinal effects of the initiation of ART on bone, and to study the relationships of bone changes with other metabolic parameters.
Methods: In A5005s, the metabolic substudy of AIDS Clinical Trials Group (ACTG) 384, whole-body DEXA scans were performed at entry and every 16 weeks thereafter in 157 subjects. Participants were randomized to receive nelfinavir (NFV), efavirenz (EFV), or both drugs combined with open-label zidovudine (AZT) and lamivudine (3TC) or didanosine (ddI) and stavudine (4dT) (NRTI groups). Percentage change in total bone mineral content was the primary outcome variable. Results: After the initiation of ART, there was a small but statistically significant decrease in total bone mineral content at 16 weeks (median: -0.44%, IQR -1.96 to 1.07%, p = 0.02). Total bone mineral content continued to decrease. At 48 weeks, the decrease was 1.02% (IQR -3.63 to 1.24%, p <0.01), while at 96 weeks, the decrease was 2.25% (IQR -4.75 to 0.35%, p <0.01). Even though the decline in total bone mineral content in the EFV group was slightly less using a mixed models analysis limited to participants remaining on original treatment, the difference between the NFV and EFV groups was not statistically significant (p = 0.08), nor was the effect of nucleoside reverse transcriptase inhibitor (NRTI) assignment (p = 0.60). Higher baseline CD4 cell count was associated with a slower decline in total bone mineral content after ART initiation (p = 0.002). Baseline total bone mineral content correlated weakly with HDL cholesterol (R = -0.24, p <0.01) and percentage of free testosterone (R = 0.26, p <0.01). CD4 and HIV RNA levels were not associated with baseline total bone mineral content. Correlations between changes in total bone mineral content and other metabolic parameters were generally not significant.
Conclusions: The initiation of ART was associated with a modest but statistically significant bone loss that was independent of the regimen used in our study. The amount of bone loss after ART initiation was greater in individuals with lower baseline CD4. These findings suggest that the treatment effect of ART on bone was not a direct toxic effect, but might be mediated through the antiviral or immunological changes associated with the initiation of ART.
Microarray Analysis of Primary Human Osteoclast Differentiation and Activity Identifies Signaling Pathways Altered by Low Levels of Ritonavir, and Restored by Pharmacologic Levels of IFN-g: Relationship to HIV Therapy-mediated Osteopenia

Rozbeh Modarresi*, Z Xiang, and J Laurence
Cornell Univ, New York, NY, US
Background: Enhanced bone demineralization occurs in advanced HIV disease and may be accelerated by certain ART drugs. We reported that T-cell production of RANKL, the obligate cytokine for osteoclastogenesis, is induced by HIV-1 gp120, with positive feedback between RANKL secretion and HIV replication. We also found that low pharmacologic levels (1 to 5 _M) of HIV protease inhibitors (PI) ritonavir (RTV) and saquinavir (SQV), but not other ART drugs, increased RANKL-mediated osteoclasts differentiation and activity in vitro, and blocked the physiologic suppression of RANKL via degradation of TRAF6 by physiologic levels (1 ng/mL) of interferon-gamma (IFN-_), without altering other proteasome functions. We sought to define the pathways involved in the marked specificity of these processes.
Methods: Non-adherent peripheral blood mononuclear cells were prepared from a healthy donor and cultured in the presence of physiologic levels of RANKL (50 ng/mL), M-CSF (100 ng/mL), and IFN-_ (1 ng/mL) (control) for 24 hours or 5 days. Select cultures were exposed to RTV (5 _M), or therapeutic levels of IFN-_ (10 ng/mL), or both. RNA was amplified, biotin-labeled, and subjected to array hybridization using U133A 2.0 Affymetrix chips. Experiments were repeated 2 or 3 times, and supported by real-time polymerase chain reactions (RT-PCR). Results: When expression of 470 genes was altered by RTV, and then re-affected by addition of IFN-_, 24-hour results were most informative. Key genes for osteoclast gene transcription (NFATc1) and WNT signaling were decreased 3- to 33-fold by RTV, and reversed by IFN-_, while genes for apoptosis, cytokines linked to suppression of osteoclast activity (interleukin [IL] -12, IL-18) and MAPK genes involved in osteoclast inhibition were increased 2.5- to 8-fold by RTV, and reversed by IFN-_. Ephrin-B2 and -B3 genes and their signaling pathways, involved in osteoclast/osteoblast coupling, were enhanced by RTV and decreased by IFN-_, adding a new dimension to effects of anti-HIV drugs on bone. Conclusions: Concentrations of RTV present in all protease inhibitor (PI) -boosted regimens significantly alter expression of genes critical to osteoclast differentiation, activity, and cooperation with osteoblasts, and can be corrected by IFN-_. This suggests a means of possible therapeutic intervention in at least 1 metabolic complication of HIV PI therapy.