iconstar paper   HIV Articles<  
Back grey arrow rt.gif
HIV Infection Is Associated With Worse Bone Material Properties, Independently of Bone Mineral Density [Microindentation measures Bone Material Strength index, new test for bone]- Brief Report
  Download the PDF here
Robert Guerri-Fernandez, MD, PhD,* Daniel Molina, MD, Judit Villar-Garcia, MD,* Daniel Prieto-Alhambra, MD, PhD, Leonardo Mellibovsky, MD, PhD, Xavier Nogues, MD, PhD, Alicia Gonzalez-Mena, MD,* Ana Guelar, MD,* Marta Trenchs-Rodriguez, MD,k Sabina Herrera-Fernandez, MD,* Juan Pablo Horcajada, MD, PhD,* Adolfo Diez-Perez, MD, PhD,* and Hernando Knobel, MD, PhD*
Compared with controls, patients with HIV had significantly lower BMSi [Bone Material Strength index], and thus worse bone material properties at the tissue level. Median BMSi was 84.5 (IQR 83-87) in HIV-infected patients and 90 (88.5-93) in controls (P < 0.001). Women had significantly lower BMSi in HIV population [85 (83-87) vs 80 (77-83); P = 0.0004]. These differences were not significant in controls 92 (88-96 in men) vs 89 (86-93 in women); P = 0.07.
The associations observed between HIV and BMSi in the univariate analyses were maintained in multivariable analysis (Table 2). In contrast, we observed no significant differences in BMD between cases and controls at any of the sites examined (total hip, femoral neck, and lumbar spine; Table 1).
Low bone mineral density (BMD) and increased fracture risk in HIV-infected individuals have been documented in an increasing number of studies.1-7 However, the mechanisms behind these associations are not clear: both the HIV infection itself and/or the toxicity of antiretroviral therapy (ART) toxicity have been implicated, but without conclusive evidence.
BMD and fracture risk have a close inverse correlation in the general population, particularly in untreated individuals.8 However, although ∼50% of women with nonvertebral fractures have a BMD T-score that exceeds the diagnostic threshold for osteoporosis (-2.5), less than half of patients who meet the criteria for osteoporosis will suffer a fracture.9 Therefore, dual-energy x-ray absorptiometry (DXA) provides limited information on bone health, and no information about other key components of bone quality, such as bone strength, composition, or microarchitecture, which are obtained through expensive or invasive methods.10,11
Microindentation is a technique that quantifies a patient's bone material properties at the tissue level and can be performed in the clinical setting. This technique discriminates between patients with and without fractures,4,12 and has been shown to detect bone tissue alterations in other situations where BMD is relatively preserved despite increased fracture risk, such as fragility fractures in patients with osteopenia,13 diabetes mellitus,14 or who are undergoing glucocorticoid treatment.15 Because the underlying mechanisms of the bone abnormalities observed in HIV-infected patients are not yet well characterized, we performed a bone microindentation study in a cohort of HIV-positive patients and a sample of age- and sex-matched controls. We hypothesized that HIV-infected patients have poorer bone tissue mechanical properties—Bone Material Strength index (BMSi) measured by microindentation. Our secondary objectives were to test for differences in BMD, markers of inflammation, and bone turnover between these groups, as well as the general safety of microindentation in this series.
In summary, for the first time, we describe in vivo measurement of bone material properties in HIV-infected patients. Microindentation captures changes in bone properties that are not measured by BMD. We found that HIV infection is associated with impaired bone material properties, independently of BMD.
In this study, we observe poorer bone material properties in HIV-infected patients than in controls, as measured by in vivo microindentation; this effect is independent of BMD and ART. Moreover, HIV-infected patients present higher levels of markers of inflammation, and this persistent inflammatory status has previously been found to be associated with bone fragility in other conditions.16,17
We found that bone density (by DXA) in our series of treatment-naive patients with HIV was similar to that in age-matched, uninfected controls after adjustment for confounding factors. The main difference between ours and previous studies is that we did not observe differences between these groups in their profile of traditional risk factors (i.e., alcohol consumption, corticosteroid treatment, body mass index, tobacco use, intravenous drug use, and malnutrition). This absence of differences may be because the profile of patients with HIV has changed over time, in that HIV is now mainly transmitted through sexual activity, whereas in the past, the transmission mechanism was mainly drug use.2,18-22
We also found no differences in bone turnover markers despite that HIV-infected patients showed lower BMSi values, independently of BMD. Microindentation induces the separation of mineralized collagen fibrils and initiates microcracks, which are likely to be the mechanism of fracture initiation.12 Hence, this technique measures the mechanical competence of bone tissue to resist the initiation and propagation of fractures,4,12,13 that is independent of bone turnover. This indicates that the material properties are impaired, although the amount of mineral, as measured by BMD, and also bone turnover markers, remain at a similar level than controls.
although BMD is considered the gold standard for evaluating bone health, in specific clinical situations, impaired quality of bone material is the most prominent feature rather than the amount of mineral,13-15 and bone fragility is only partially assessed by densitometry.9 Until recently, direct mechanical testing of bone strength has been cumbersome, requiring an invasive procedure to extract bone samples for ex vivo laboratory testing. Actually, microindentation has permitted direct measurement of the tissue's mechanical performance (i.e., strength vs fragility). As reported, microindentation is able to measure some changes in material properties that are not detected by DXA.
We recruited a total of 85 study subjects (50 HIV-infected patients and 35 controls) between January and October 2014. The time since diagnosis of infection ranged from 0 to 19 years (mean age at diagnosis, 34 years). We observed no remarkable difference in smoking, alcohol habits, and physical activity between cases and controls. Baseline characteristics of the sample are presented in Table 1.
We observed no differences in bone turnover markers. Patients with HIV had significantly lower levels of 25-hydroxy-vitamin D than controls (20.9 ng/mL vs 33.3 ng/mL; P = 0.015). In contrast, levels of high sensibility C-reactive protein (0.48 mg/dL vs 0.12 mg/dL; P = 0.005) and erythrocyte sedimentation rate (20 mm/h vs 3 mm/h; P = 0.0001) were significantly higher in HIV-infected individuals than in controls; similarly for D-dimer (289.6 ng/mL vs 137.2 ng/mL; P = 0.018), fibrinogen (400 mg/mL vs 317 mg/mL; P = 0.037), and beta-2 microglobulin (2.48 mg/L vs 1.438 mg/L; P = 0.0001), respectively.
Compared with controls, patients with HIV had significantly lower BMSi, and thus worse bone material properties at the tissue level. Median BMSi was 84.5 (IQR 83-87) in HIV-infected patients and 90 (88.5-93) in controls (P < 0.001). Women had significantly lower BMSi in HIV population [85 (83-87) vs 80 (77-83); P = 0.0004]. These differences were not significant in controls 92 (88-96 in men) vs 89 (86-93 in women); P = 0.07. The associations observed between HIV and BMSi in the univariate analyses were maintained in multivariable analysis (Table 2). In contrast, we observed no significant differences in BMD between cases and controls at any of the sites examined (total hip, femoral neck, and lumbar spine; Table 1).
We tested for prespecified interaction between HIV and microindentation with age, sex, nadir CD4 count, and time since diagnosis but found no interactions.
There were no complications in any patient during or after the microindentation procedure.

We conducted a cross-sectional study of outpatients of the Infectious Diseases Department at Hospital del Mar, Barcelona, Spain, and controls were healthy volunteers recruited from a Primary Care Center in Barcelona. The study was approved by the local Clinical Research Ethics Committee, and all participants gave written informed consent (Ethical Committee number 2013/5250/I).
All HIV-infected patients were viremic (HIV >40 copies/mL) and had no previous history of ART. Individuals were excluded from the study if they had a history of treatment with bone-active drugs, or if they had any condition that could interfere with bone metabolism, such as liver disease, alcoholism, malignancy, Cushing syndrome, hypogonadism, hyperthyroidism, hypopituitarism, hyperparathyroidism, chronic kidney disease, chronic obstructive pulmonary disease, hepatitis C or B, diabetes, neuropathic disease, and use of glucocorticoids, opioids, or intravenous drugs.
During the inclusion visit, we obtained a full medical and medication history, blood samples, hip and spine BMD by DXA [DXA-Hologic QDR4500SL (S/N 45329)], lateral x-ray (to rule out previous fractures) of the thoracic and lumbar spine, and microindentation test. The coefficient of variation (CV) for DXA measurements is 1.7% at the femoral neck and 1% in the spine.
Bone Microindentation Testing
Microindentation testing was performed using a handheld OsteoProbe Reference Point Indenter (Active-Life-Scientific, Santa Barbara, CA), which consists of a head unit with a displacement transducer, and an impact mechanism. After applying local anaesthesia (2% mepivacaine), a sterilized stainless steel probe with a 90º conical tip (375um diameter; <4 μm tip sharpness radius) is placed on the midpoint of the midshaft anterior tibial plateau and inserted through the soft tissue until it makes contact with the bone surface. Holding the device perpendicular to the bone surface, the operator displaces the device until, after a preload force of 10 N, the device's trigger mechanism releases a 30 N impact force. This displaces the test probe into bone, and the displacement transducer measures the indentation distance. After 8 repeated indentations separated by approximately 2 mm, the operator performs 5 additional indentations with the same probe on a cube of polymethylmethacrylate (PMMA) considered outliers, and ruled out, those measurements that were 1.5-fold the interquartile range (IQR) below or above the limits of the IQR. The software provides the BMSi, defined as 100 times the ratio between the harmonic mean indentation distance of the 5 impacts into the calibration phantom (PMMA) and that of the 8 impacts into the bone. This technique has previously been validated in humans.12-15 The procedure takes less than 5 minutes, causes minimal discomfort to the patient, and no complications have been observed in published studies.12-15 The coefficient of variation for microindentation is 3%.
Laboratory Tests
All measurements were performed at 8:00 AM and while fasting in all patients, to avoid variations due to circadian rhythms.
Serum concentration of specific markers was analyzed using the following assays: calcium, phosphate [Roche Diagnostics, Interassay CV (iCV) <10%]; bone turnover markers Amino propeptide of type 1 collagen (P1NP) and collagen type I cross-linked C-telopeptide (CTX) [Immunoenzymatic electrochemiluminescence, ECLIA (Roche Diagnostics) iCV <10%]; intact parathyroid (iPTH); Chemiluminescent immunoassay (CLIA) [iCV <10% (Siemens)], Serum 25-hydroxy-vitamin D (CLIA) [iCV <10%; Elecsys (Roche)]. High sensitivity C-reactive protein {CLIA [Immulite 2000 (Siemens)]}, erythrocyte sedimentation rate, beta-2 microglobulin (CLIA) [Immulite 2000 (Siemens)]. D-dimer [immunoturbidimetry (ACL TOP300)] and fibrinogen [Clauss method (ACL TOP300)].
Statistical Analysis
We compared quantitative and categorical variables between HIV-infected and control groups using 2-sample t tests and χ2 test, respectively. We performed univariate and multivariate linear regression to test for association between BMSi and HIV status. Factors with a univariate P value of <0.20 were included in the multivariable model. We explored interactions in the HIV group between baseline BMSi and sex, age, nadir CD4 count, and time since diagnosis.
Assuming a type I and II error of 0.05 and 0.2 in a 2-sided test, respectively, 15 controls and 22 patients with HIV were required to detect a statistically significant difference of ≥5 units of BMSi. The common standard deviation was assumed to be 5, and we expected a drop-out rate of 10%. Results with P < 0.05 (2-tailed) were considered statistically significant. Analyses were performed using Stata/IC 13.1.
  iconpaperstack View Older Articles   Back to Top   www.natap.org