Volumes could be written -- indeed, volumes have been written -- on whether antiretrovirals individually or en masse directly lower BMD. The bottom line: some do, but no one suggests avoiding or delaying antiretroviral therapy because it sends people to the hospital with fragility fractures.
The National Osteoporosis Foundation pointedly omits antiretrovirals from its 16-drug list of medications that make osteoporosis more likely (Table 2). And new HIV bone guidelines advise following national protocols on when to start antiretroviral therapy because treatment benefits "far outweigh the potential negative long-term effects on bone mass and metabolism, and fracture risk."15 For people with a high fracture risk (previous fragility fracture, T score at or below -2.5, FRAX at or above 20%), the guidelines suggest avoiding tenofovir disoproxil fumarate (TDF) and protease inhibitors (PIs) in first-line regimens and switching from those drugs if treatment has begun. (For findings on tenofovir alafenamide -- TAF -- and BMD, see "Antiretrovirals: PIs, TDF and TAF").
Perhaps because many people start therapy with TDF, a PI, or both, BMD does drop 2% to 6% in the first 2 years of treatment, the bone guidelines acknowledge.15 But after that not much happens, according to a 6-year study comparing 44 men with HIV and 37 without HIV.16 The HIV group averaged 49 years in age and the control group 46. The groups did not differ significantly in weight or body mass index, but a significantly higher proportion of the HIV group smoked (37% versus 3%, P < 0.001). Men with HIV had been infected for an average 8 years.
HIV-positive men had DXA scans when antiretrovirals began and 2 and 6 years later. HIV-negative controls also had three DXA scans over 6 years. Six-year change in total body BMD did not differ significantly between men with and without HIV (+0.3% and +0.5%, P = 0.15), and the same proved true for total hip BMD (-0.6% and -0.1%, P = 0.8). Men with HIV gained significantly more lumbar spine BMD than did seronegative men through 6 years (+5.3% versus +0.3%, P < 0.001).
Anyone still wavering on whether to start (or continue) antiretroviral therapy in someone with a high fracture risk should consider the plentiful reports linking low CD4 count or high viral load to lower bone density or fracture. At least two studies tied higher viral load to lower BMD. The older study involved 161 consecutive HIV-positive people in Italy who completed a rigorous demographic questionnaire and testing for bone-related disease.17 Twenty-two of 48 antiretroviral-naive people (46%) and 58 of 113 taking antiretrovirals (51%) had osteopenia or osteoporosis -- a slim difference. Three classic risk factors -- female gender, older age, and lower body mass index -- independently predicted osteopenia or osteoporosis. And every 10-fold higher viral load at DXA scanning doubled chances of osteopenia or osteoporosis (adjusted odds ratio [aOR] 1.97, 95% CI 1.16 to 3.34, P = 0.01) (Figure 1). Compared with antiretroviral-naive people, the treated group did not run a higher risk of low bone density. A more recent analysis pooled data from 796 participants in AIDS Clinical Trials Group (ACTG) studies in which everyone had whole-body DXA before starting first-line therapy and 96 weeks later.18 Participants lost an average 2% of BMD through 96 weeks of treatment. The same three classic risk factors identified in the Italian study17 -- female gender, older age, and lower body mass index -- emerged as independent predictors of a greater 96-week BMD drop. Higher pretreatment viral load independently predicted greater BMD loss (-0.56% lower per 10-fold higher viral load, P = 0.02) (Figure 1). Treatment with TDF or a PI also made losing more BMD more likely.
Figure 1. An ACTG analysis linked every 10-fold higher HIV load before antiretroviral therapy (ART) to 0.56% lower bone mineral density (BMD).18 In a series of 161 consecutive HIV patients in Italy, every 10-fold higher viral load before or during ART doubled chances of osteopenia or osteoporosis.17 (Osteoporosis images from Servier PowerPoint Library.)
A half-dozen studies link lower baseline18,19 or nadir20-23 CD4 count to low BMD or fracture. In a single-center Australian case-control study of HIV patients with and without fractures, a current CD4 count below 500 cells/mm3 doubled the risk of fragility fracture.19 A count below 200 cells/mm3 (versus above 500) inflated fracture risk almost 7-fold. In the U.S. HIV Outpatient Study (HOPS), a nadir CD4 count below 200 raised the risk of incident fracture 60%.20 Together these studies show that uncontrolled HIV infection -- signaled by high viral loads or low CD4 counts -- threaten bone health more than antiretroviral therapy.
Neither the National Osteoporosis Foundation nor the CDC counts hepatitis virus infection as an osteoporosis risk factor, though the Foundation more broadly lists "liver disease." But plentiful evidence from recent meta-analyses and other studies confirms that either HCV or HBV adds to the bone-thinning risk with HIV. Testing for these viruses should be routine for people with HIV, regardless of bone risk, because both can be treated and HBV can be prevented by vaccination. The added bone risk with hepatitis virus infection should offer an extra prod to test HIV patients for these viruses.
Two recent meta-analyses established a tie between HCV/HIV coinfection and osteoporosis or fracture compared with HIV alone or no infection. The more recent analysis considered studies published up to April 2013 that assessed endpoints of BMD or incident fracture in HCV/HIV-coinfected people compared with HIV-monoinfected people or HCV/HIV-negative individuals.24 The investigators found 13 studies -- six considering BMD and seven fracture -- involving 427,352 people. Compared with HIV-monoinfected individuals (but not uninfected people), HCV/HIV-coinfected people had a doubled chance of low bone density (pooled OR 1.98, 95% CI 1.18 to 3.31). Fracture risk proved more than 50% higher with coinfection than with HIV alone (pooled relative risk [RR] 1.57, 95% CI 1.33 to 1.86). Fractures were more than twice as likely with HCV/HIV coinfection than with neither infection in cohort studies (pooled RR 2.46, 95% CI 1.03 to 3.88) or in cross-sectional studies (pooled OR 2.30, 95% CI 2.09 to 2.33).
Exploring 15 studies presented through 2013 -- nine focused on BMD and six on fracture -- another team reached similar conclusions.25 HCV/HIV-coinfected people had 63% higher odds of osteoporosis than HIV-monoinfected people (OR 1.63, 95% CI 1.27 to 2.11). Coinfected individuals had a 77% higher overall fracture incidence than HIV-monoinfected people (pooled incidence rate ratio [IRR] 1.77, 95% CI 1.44 to 2.18) and almost a tripled fracture incidence compared with uninfected people (pooled IRR 2.95, 95% CI 2.17 to 4.01). This analysis confirmed the importance of classical bone risk factors in these populations -- older age, lower BMI, smoking, and alcohol and substance abuse.
These investigators noted that people with HCV/HIV may run a higher fracture risk because of behaviors that boost chances of bone-breaking trauma.25 But they doubted such dangerous behavior completely explains the inflated risk because adjustment for alcohol and substance use did not eliminate the higher fracture risk with HCV. Rather, they observed, research shows that chronic liver disease upsets healthy bone remodeling.26-28 Other research suggests chronic inflammation resulting from HIV infection promotes bone resorption (breakdown).29 Reason suggests that the added inflammation due to HCV coinfection would compound that process.
Less work addresses the interplay between HBV infection -- with or without HIV -- and bone risk. A recent study compared incident hip fracture in four U.S. Medicaid populations -- 4156 people dually treated for HBV and HIV, 2015 treated only for HBV monoinfection, 96,253 treated for HIV monoinfection, and 746,794 HBV/HIV-negative people.30 Through 5 years the HBV/HIV group had a 37% higher risk of hip fracture (aHR 1.37, 95% CI 1.03 to 1.83) than HIV-monoinfected people and a 35% higher risk than uninfected people (aHR 1.35, 95% CI 1.03 to 1.84).
No comments have been made.
The content on this page is free of advertiser influence and was produced by our editorial team. See our content and advertising policies.