Bone Turnover Markers and Cumulative Viral Load in HIV-Positive Patients

Presentation Number: SUN 359
Date of Presentation: April 2nd, 2017

Anne Wilson Wagstaff* and Amy Warriner
University of Alabama at Birmingham, Birmingham, AL


Background: As a result of antiretroviral therapy (ART), HIV has become a chronic disease that people live with for years/decades. HIV has been found to chronically impact the immune system and places patients at an increased risk of bone loss and fractures (1). The pathophysiology of low BMD in persons living with HIV (PLWH) appears to be multifactorial and is not fully understood. Our study investigates the relationship between cumulative viral load and bone turnover markers (BTM) in a population of PLWH to further elucidate the chronic effects of HIV on bone metabolism.

Methods: This retrospective longitudinal study examined a well-characterized cohort of PLWH. Patient serum samples were analyzed for N-midfragment of osteocalcin (OC) and N-terminal extension peptide of type 1 collagen (P1NP) as measures of osteoblast activity, and the b subform of the isoenzyme 5 of tartrate-resistant acid phosphatase (TRACP5b) and type 1 collagen C-telopeptide (CTX1) as measures of osteoclast activity. Viral load (VL) values measured within 12 months prior to BTM sample collection were used to calculate viremia copy-years (VCY), a published measure of cumulative viral load (2). BTM were compared to each other using Spearman Correlation Coefficients. Patients were placed into quartiles based on VCY values. ANOVA was used to examine the relationship between VCY quartiles and each BTM. We used multivariate regression analysis to examine the effect of VL, inflammatory markers, CD4 count, ART use, and other demographic data on each BTM.

Results: This study included 59 subjects (male=26, mean age=43.3). Some subjects had adequate data to contribute multiple observations totaling 63 observations. 56 were on ART (mean duration 40 months) with a mean CD4 count of 454 at baseline sample collection. Each individual BTM was significantly correlated with every other BTM (Spearman correlation coefficient p-values ranged from <0.001 to 0.0084). CD4 count was significantly correlated with VCY (r = -0.34, p-value = 0.007). ANOVA F and P values comparing VCY quartiles for each BTM are as follows: OC (F=3.55, P=0.020), P1NP (F=0.76, P=0.518), TRACP5b (F=2.24, P=0.094), and CTX1(F=3.27, P= 0.0278). For all four BTMs, the average BTM value increased from quartile 1 to quartile 2 and 3, but decreased for quartile 4. Male sex was associated with higher TRACP5b (p=0.0462). Age and smoking were associated with higher NMidOC (p=0.0177, p=0.0102).

Conclusion: BTM of osteoblast activity correlate with those of osteoclast activity among PLWH. Higher VCY correlated with lower CD4 counts. As VCY increases, BTM also initially appear to increase. However, in patients with the highest VCY, BTM are decreased. This demonstrates a non-linear relationship between cumulative viral load and bone turnover.

References: (1) McComsey et al., Clin Infect Dis. 2010 Oct 15;51(8):937-46 (2) Cole et al., Am J Epidemiol. 2010 Jan 15; 171(2): 198-205


Nothing to Disclose: AWW, AW