| |
Bone Disease in HIV:
study finds, heavily pre-treated patients had Vitamin D deficiency & serum marker abnormalities & perhaps serum markers may be better for assessing patient risk than total BMD by DEXA
|
| |
| |
"Osteoprotegerin and bone turnover markers in heavily pretreated HIV-infected patients"
HIV Medicine
May 2005
E Seminari1, A Castagna1, A Soldarini2, L Galli1, G Fusetti1, F Dorigatti2, H Hasson1, A Danise1, M Guffanti1, A Lazzarin1 and A Rubinacci31
Infectious Diseases Department, 2Diagnostica e Richerca San Raffaele SpA, LABORAF and 3Bone Metabolic Unit, Vita/Salute University, San Raffaele Scientific Institute, Milan, Italy
".....This study showed that 40% of heavily pretreated patients with advanced HIV infection had an osteopenic/osteoporotic condition associated with a complex biochemical picture highly suggestive of a high bone turnover rate, which may have clinical relevance through increasing the risk of fractures regardless of bone mass...... Vitamin D insufficiency and the subsequent enhancement of PTH secretion should first be taken into account.....It is, therefore, conceivable that antiretroviral therapy itself may have a specific action on bone that increases bone cell turnover and possibly activates bone loss...... Whatever its origin, high bone turnover is associated with increased remodelling space and decreased BMD..... The absence of a substantial total BMD deficit does not reduce the clinical impact of the high bone turnover in the majority of HIV-infected patients...... Clinical assessment of osteoporosis by means of specific markers might therefore be cost-effective when risk factors for osteoporosis have to be determined in HIV-infected patients treated with antiretroviral therapy. The measurements of AP, BGP and Dpd significantly correlated with the presence of osteopenia..... the present study suggests that total body BMD in HIV-infected patients is not a fine indicator of bone disturbance..... In conclusion, with all its limitations (in particular the cross-sectional design, the small number of subjects analysed and the absence of a control group of untreated patients and/or HIV-negative subjects), this study showed that heavily pretreated patients with advanced HIV infection were characterized by vitamin D insufficiency, secondary hyperparathyroidism and highly active bone remodelling which, in addition to BMD, is considered an independent risk factor for bone fragility."
ABSTRACT
Objectives: To characterize osteoprotegerin (OPG) levels, bone remodelling and bone mineral density (BMD) in heavily pretreated HIV-infected patients on antiretroviral therapy, and to evaluate the clinical factors associated with bone density decline.
Methods: Heavily pretreated (>5 years) HIV-positive patients were enrolled in this cross-sectional, observational study, which was based on a total body bone densitometry examination and a comprehensive evaluation of bone and mineral parameters.
Results: Sixty-eight patients (55 male and 13 female) with a median age of 41 years (range 25-60 years) were included in the study.
Their antiretroviral treatment lasted for 82 months.
On the basis of the World Health Organization criteria, nine patients (13.2%) were osteoporotic [T-score<-2.5 standard deviation (SD)] and 19 patients (27.9%) were osteopenic (T-score between-1 and-2.5).
The principal outcomes associated with the presence of a low BMD were high OPG and lysylpyridinoline/creatinine ratio (Dpd) values.
Most of the patients (39 of 48; 81.25%) showed vitamin D insufficiency [Vitamin D (25(OH)D)<18 ng/mL] with secondary hyperparathyroidism (13 of 50 patients: 26%), which proved to be correlated to osteocalcin (BGP) levels [parathyroid hormone (PTH) vs. BGP: r=0.34; P<0.01].
There was an inverse correlation between T-scores and serum osteocalcin and alkaline phosphatase (AP) levels, on one hand, and Dpd, on the other.
High AP and Dpd values were associated with relative risks of 4.1 [95% confidence interval (CI)=1.01-17.6] and 7.2 (95% CI=1.67-31.03), respectively, of a pathological T-score.
Multivariate analysis revealed that the factors associated with the presence of osteopenia or osteoporosis were older age and lower body mass index.
Conclusions: About 40% of our heavily pretreated subjects with advanced HIV infection had a low BMD, and 56% (24 of 44 patients) showed a high bone turnover rate with marked osteoclast activation. High OPG levels may protect against bone resorption.
INTRODUCTION
Prolonged life expectancy exposes HIV-infected patients to metabolic challenges that may further reduce their quality of life and increase the economic burden of pandemic HIV infection [1]. As infected patients grow, live and age, the acquisition, maintenance and loss of bone may be affected by the virus itself [2-5], antiretroviral therapy [6-8], or the effects that both may have on other organ functions modulating bone metabolism [9,10]. Recent studies have found a very high prevalence of osteopenia/osteoporosis in cohorts of paediatric and adult HIV-infected patients [11-13].
As the activation of bone remodelling is a key factor in causing the cortical porosity and trabecular network deterioration that increase bone fragility, regardless of bone mineral density (BMD) [14], its presence and clinical relevance require to be investigated in fully characterized patient cohorts. The recent identification of the osteoprotegerin (OPG)/receptor activator of the nuclear factor kappa B (NF-kB) ligand (RANKL)/receptor activator of the NF-kB (RANK) system as the dominant mediator of osteoclastogenesis under osteoblastic control [15] has thrown some light on osteoblast/osteoclast cross-talk, and the genesis of the enhanced bone turnover observed in HIV-infected patients [10].
In the light of these considerations, we studied the OPG/RANKL ratio in a sample of heavily pretreated HIV-infected patients who underwent a complete assessment of bone formation [serum osteocalcin (BGP)], bone resorption [lysylpyridinoline /creatinine ratio (Dpd)] and bone mineral parameters [calcium, phosphorus, parathyroid hormone (PTH) and vitamin D]. We also considered clinical factors, immunological status, antiretroviral therapy history and bone mineral content as potential correlates of bone turnover status.
AUTHOR DISCUSSION
This study showed that 40% of heavily pretreated patients with advanced HIV infection had an osteopenic/osteoporotic condition associated with a complex biochemical picture highly suggestive of a high bone turnover rate, which may have clinical relevance through increasing the risk of fractures regardless of bone mass [14].
Our findings contrast with those of a study carried out before the antiretroviral therapy era in a heterogeneous group of immunocompromised, antiretroviral-naïve patients, in whom a number of histomorphometric parameters of bone formation and resorption may have been reduced by the presence of HIV [5], as well as the results of studies showing that BGP values were lower in naïve subjects with greater immune deficiency [2,3,17,18]. This pattern has been interpreted as the expression of the profound alteration in the cytokine microenvironment induced by HIV rather than a direct infection-related impairment of osteoblast activity and/or proliferation [19].
The controversy concerning bone turnover in HIV-infected patients may be resolved by considering that the high bone turnover typical of HIV-infected patients treated with antiretroviral therapy is of multifactorial origin.
Vitamin D insufficiency and the subsequent enhancement of PTH secretion should first be taken into account, because PTH is the major activator of bone remodelling. Vitamin D insufficiency or deficiency may be associated with a bone matrix liberalization defect, as suggested by high AP levels, and have exogenous, endogenous or drug-related origins [i.e. sunlight deprivation and diet lacking in calcium, which in HIV-infected patients may be related to a poor quality of life, and the inhibitory action of some protease inhibitors (PIs) of the cytochrome P450 enzyme system, which controls the serum levels of active vitamin D metabolites] [20,21]. The secondary hyperparathyroidism observed in our cohort could be the expression of a correction mechanism devoted to the maintenance of plasma calcium levels.
Antiretroviral therapy, which had its advent in 1997, has been considered a causative factor of osteopenia [1]. It is associated with a marked and significant increase in BGP levels over baseline conditions [3,17,22], increased biochemical markers of bone turn-over (such as AP and Dpd) [11,12], low BMI [11], and male gender [23], as was found in our study. It is, therefore, conceivable that antiretroviral therapy itself may have a specific action on bone that increases bone cell turnover and possibly activates bone loss. As the OPG/RANKL/RANK system is the dominant final mediator of osteoclastogenesis and regulates osteoblast/osteoclast cross-talk [15], our finding that OPG levels are high in HIV-infected individuals confirms previous observations [3], and suggests that HIV or antiretroviral therapy itself may act on the osteoblast and stromal cells that are essential inducers of osteoclast differentiation and producers of OPG. However, the significance of the high OPG levels remains unknown, because circulating levels may not be representative of local expression. By analogy with other bone metabolic states characterized by a high bone turnover and high OPG levels, such as postmenopausal osteoporosis [24], it is possible that high OPG levels indicate a protective response designed to counteract osteoclastogenesis activation by operating as a homeostatic system, in order to limit any more rapid bone loss. It is therefore intriguing to observe that high OPG levels were significantly associated with low BMD.
The osteoprotegerin ligand RANKL is a new member of the tumour necrosis factor receptor/tumour necrosis factor (TNFR/TNF) family that regulates osteoclast differentiation and activation [25]. Although there is evidence that some PIs enhance osteoclast activity [26], we did not find higher circulating RANKL levels in our osteopenic patients, nor any difference in the OPG/RANKL ratio between them and their osteopenic counterparts. This indicates that, at least in the studied cohort, osteoclastogenesis was not induced by HIV infection or antiretroviral therapy. It can, therefore, be speculated that the enhanced osteoclast activity reflected in the high Dpd values is the expression of an increase in cell life-span rather than in the number of cells.
Whatever its origin, high bone turnover is associated with increased remodelling space and decreased BMD. This association was only partly confirmed by our DXA measurements, which showed some bone mineral deficit in no more than 40% of the subjects. However, it should be pointed out that trabecular-rich sites, which are the most active metabolic areas and therefore the regions mainly affected by high bone turnover, had no detectable impact on total body BMD as measured in this study. Total body BMD reflects cortical bone, which constitutes 95% of the detected X-ray absorbance and may become significantly deficient only as a late outcome of altered bone and mineral metabolism. The absence of a substantial total BMD deficit does not reduce the clinical impact of the high bone turnover in the majority of HIV-infected patients, as it is an independent predictor of fragility fractures [14]. Clinical assessment of osteoporosis by means of specific markers might therefore be cost-effective when risk factors for osteoporosis have to be determined in HIV-infected patients treated with antiretroviral therapy. The measurements of AP, BGP and Dpd significantly correlated with the presence of osteopenia. Conversely, in agreement with other authors [27], the present study suggests that total body BMD in HIV-infected patients is not a fine indicator of bone disturbance.
In conclusion, with all its limitations (in particular the cross-sectional design, the small number of subjects analysed and the absence of a control group of untreated patients and/or HIV-negative subjects), this study showed that heavily pretreated patients with advanced HIV infection were characterized by vitamin D insufficiency, secondary hyperparathyroidism and highly active bone remodelling which, in addition to BMD, is considered an independent risk factor for bone fragility.
RESULTS
Sixty-eight patients consecutively undergoing clinical evaluation (55 male and 13 premenopausal female patients with a median age of 41 years; range 25-60 years) agreed to participate in the study by giving their oral informed consent. They had been receiving antiretroviral therapies for a median of 82 months (range 64-155 months), and had a median nadir CD4 lymphocyte count of 37.5 cells/muL (range 1-300 cells/muL) and a median baseline plasma viral load of 5 log10 HIV-1 RNA copies/mL (range 2.3-6.9 copies/mL). Thirty-seven patients (54.4%) had experienced at least one AIDS-defining event. Their biochemical characteristics are shown in Table 1. Most of the subjects had a high bone turnover, indicated by their high BGP and Dpd levels (24 of 44 patients; 56%), vitamin D insufficiency [25(OH)D <18 ng/mL; 39 of 48 patients; 81.25%] and secondary hyperparathyroidism (13 of 50 patients; 26%) (Fig. 1). PTH levels negatively correlated with vitamin D scores as expressed by 25(OH)D concentrations (r=-0283; P=0.04).
The bone mineral data show that 40 patients (58.8%) had normal T-scores (control group), and 28 (41.2%) had a T-score indicative of various degrees of bone loss (osteopenic group); of these, 19 (27.9%) had osteopenia and nine (13.2%) had osteoporosis. The two groups were different in terms of age (P=0.05); there were no significant differences in other epidemiological data (Table 2).
All of the osteoporotic patients were male (n=9; 61.8%); male patients had significantly lower T-scores (median -0.98 SD; range -4.1 to 1.8) than female patients (-0.1 SD; range -1.7 to 3) (P=0.03). There were no significant gender differences in terms of age (median 41 years, range 25-61 years for male patients; median 39 years, range 33-51 years for female patients), duration of therapy, nadir CD4 count or AIDS-defining events.
The main bone parameters associated with the presence of a low bone mineral content were high levels of OPG (osteopenic 5.0 vs. control group 4.2 pmol/mL; P=0.01) and Dpd (osteopenic 17.4 vs. control group 12.29 pmol/mL; P=0.05). There were no between-group differences in RANKL levels or the OPG/RANKL ratio (Table 1). The T-scores were negatively correlated with BGP (r=-0.3; P=0.03), AP (r=-0.3; P=0.03) and Dpd (r=-0.29, P=0.04); BGP and AP levels were significantly correlated with each other (r=0.34, P=0.01). Abnormal AP and Ddp values were, respectively, associated with relative risks of 4.1 [95% confidence interval (CI)=1.01-17.6] and 7.2 (95% CI=1.67-31.03) of having a pathological T-score.
In the multivariate analysis, logistic regression revealed that the factors associated with the presence of osteopenia/osteoporosis were age [odds ratio (OR)=1.108; 95% CI=1.014-1.226; P=0.03] and body mass index (BMI) (OR=0.748; 95% CI=0.562-0.946; P=0.03). The other parameters evaluated in the model (gender, AP, GGT, duration of antiretroviral therapy, and nadir and baseline CD4 lymphocyte counts) were not significant. The generalized linear model showed that male gender (F=10.2; beta=-1.5; P=0.003), BMI (F=4.8; beta=0.12; P=0.041), and pathological AP (F=4.33; beta=-0.7; P=0.05) and Dpd values (F=7.8; beta=-0.09; P=0.005) were significant and independent factors associated with T-scores. The other parameters evaluated in the model (age, PTH, RANKL, BGP and OPG) were not significant.
METHODS
Patients
Heavily pretreated HIV-positive patients (>5 years on antiretroviral treatments) who consecutively attended the Outpatient Clinic of the Department of Infectious Diseases, San Raffaele Hospital (Milan, Italy) between November 2001 and June 2002, and who agreed to undergo bone densitometry, were included in this cross-sectional study. Blood samples were taken at the time the patients underwent dual-energy X-ray absorptiometry (DXA). Their clinical data were obtained from an electronic database.
Assays
Levels of plasma and urinary calcium and creatinine, plasma phosphate, alkaline phosphatase (AP), alanine aminotransferase (ALT), total bilirubin, gamma-glutamyl transpeptidase (GGT) and triglycerides were measured colorimetrically (Roche Diagnostici, Monza, Italy) using a Hitachi 747 auto analyser (Hitachi, Tokyo, Japan); bone alkaline phosphatase (bone AP) was measured using agarose gel electrophoresis (REP, Helena Lab, Inc, Beaumont, TX, USA). Serum ionized calcium was measured by means of a selective electrode (AVL; Roche Diagnostici, Monza, Italy) whose intra- and interassay coefficients of variation (CVs) are, respectively, 2% and 3%.
Intact PTH was quantified using a two-site chemiluminescence enzyme immunometric assay (Nichols Institute, San Juan Capistiano, CA). Both the within- and between-assay CVs for the PTH values within the reference range were less than 6%. Intact serum BGP was measured using a chemiluminescence assay (Diasorin, Saluzzo, Italy), and vitamin D [25(OH)D] was measured after extraction with a radioimmuno assay (RIA) (Diasorin, Saluzzo, Italy), in both cases using a LIAISON analyser (Byk-Sangtec Diagnostica GmBH, Dietzenbach, Germany): the intra- and interassay CVs were, respectively, 4% and 5%. The blood levels of OPG and RANKL were determined using a sandwich-type enzyme-linked immunosorbent assay (ELISA; Biomedica AG, Wiens, Austria); the inter- and intra-assay CVs were less than 10%, and the detection limit was 0.14 pmol/L. Dpd was measured in fasting morning urine samples using ion-pair reverse-phase high-performance liquid chromatography (HPLC), after hydrolysis of the diluted urine (250 muL of urine sample plus 250 muL of HCl, 12 N, at 105°C for 16 h) and partition chromatography extraction, as previously described [16]. The analysate was detected fluorimetrically (excitation (exc.) 295 nm; emission (em.) 400 nm) on the basis of its natural fluorescence. The within-run precision of the HPLC method was assessed by analysing 50 replicates of a hydrolysed urine specimen, and the CV was 5%. The urinary markers were expressed as ratios of the creatinine concentration.
BMD measurement
A DXA scanner (Lunar DPXL, Madison, WI) was used to measure whole-body lean and adipose mass, as well as total BMD, using the manufacturer's software (version 6.10). The CV for BMD was 0.6%. The data were expressed as areal density. T-scores for the bone measurements were calculated according to the European Lunar Device database (MEC, Minster, OH, USA) in order to show the magnitude of the standard deviation from the mean values for young controls. The World Health Organization (WHO) criteria were used to classify the patients as having osteoporosis (BMD T-score<-2.5 standard deviation (SD)), osteopenia (BMD T-score between-1 and -2.5 Standard deviation SD) or normal BMD (BMD T-score >-1 SD) [17].
Statistical analysis
The patients with pathological (osteopenic and osteoporotic) T-scores were considered as one group and compared with those with normal T-scores. The values of some laboratory parameters were stratified into two classes (normal vs. pathological) on the basis of their normal ranges. Univariate analysis was performed using the chi2 test or Fisher's exact test for categorical variables, as appropriate, and Wilcoxon's two-sample rank test for continuous variables. The latter was used to compare the mean values of some measures between the T-score classes (pathological vs. normal) or to compare the mean T-score values between genders and bone laboratory parameter classes. T-scores were used in the overall analysis in order to avoid a gender-related bias. Spearman correlation coefficients were also calculated to test for linear relationships among the continuous variables. The multivariate analysis applied the logistic regression model in order to assess the independent contribution of some risk factors to outcome (the presence of bone mineral alterations). The generalized linear model was also applied to estimate the independent and significant risk factors that may affect T-score variation. Each estimate is described using the beta-value and its corresponding F-value and P-value.
|
|
| |
| |
|
|
|
