Prevalence and factors associated with idiopathic hypercalciuria in HIV patients on combination antiretroviral therapy

INTRODUCTION

Idiopathic hypercalciuria is a metabolic abnormality, defined as any level of urine calcium that exceeds net intestinal absorption, leading to a net loss of calcium (250 mg/day for women and 300 mg/day for men, or 4 mg/kg body weight over a 24-hour period in either sex) and frequently to calcium kidney stone formation. ¹ Generally the high urine calcium excretion is accompanied by normal serum calcium, normal or slightly low serum phosphate and no evidence of bone disease, but there is some evidence reported in the literature that idiopathic hypercalciuria might affect bone density. ²

There are three major subtypes of hypercalciuria ³ : (1) absorptive hypercalciuria in which a primary increase in intestinal calcium absorption may result in increased urine calcium; (2) resorptive hypercalciuria caused by an increase in bone turnover; and (3) renal hypercalciuria in which a primary defect in renal tubule calcium excretion leads to increased calcium loss with a compensatory gut hyperabsorption or bone mobilization. In HIV-infected patients several factors may contribute to a dysfunction of these three compartments complicating the scenario. Nephrotoxic drugs, high systemic acid load and a persistent inflammatory state may influence gut absorption, renal leakage and bone turnover.

Several studies conducted on HIV-negative patients show that hypercalciuria may affect bone density; this issue is of particular interest because HIV-infected subjects have a higher prevalence of osteoporosis compared with seronegative ones, ⁴ even if the contributions of antiretroviral therapy (ART) and immunological and virological factors remain unclear. One paediatric cross-sectional study conducted in Venezuela ⁵ found hypercalciuria in 16/26 HIV-infected children; even if the analysis did not focus on factors associated with this abnormality, the authors stress the importance of an early detection of kidney tubular dysfunction, especially for its implications on growth retardation and bone metabolism.

In order to evaluate the prevalence and characteristics associated with hypercalciuria, we performed a cross-sectional study in a sample of HIV-positive patients on combination ART (cART).

METHODS

We enrolled all consecutive HIV patients attending our outpatient clinic for cART-related toxicity management who agreed to enter this observational study. For each participant we evaluated HIV clinical features, biochemical parameters, cART regimens, biochemical markers of bone metabolism (C-terminal telopeptide of type 1 collagen, N-terminal telopeptide of type 1 collagen, parathormone [PTH], osteocalcin, 25-OH vitamin D, phosphataemia, calciuria) and bone density (measured by DXA). All these measurements were routinely performed in our outpatient clinic. Urinary calcium was measured by 24-hour urinary collection. We did not perform total serum calcium level or albumin level, preferring instead measurement of ionized calcium, which is not influenced by serum protein concentration. Other urinary electrolytes were not collected.

In evaluating idiopathic hypercalciuria we ruled out all secondary causes of increased urine calcium excretion, such as primary hyperparathyroidism, sarcoidosis, Cushing's syndrome, cancer, excess of vitamin D, hyperthyroidism, glucocorticoid use, Paget's disease and renal tubular acidosis.

The cut-off value for the definition of hypercalciuria was 4 mg/kg. This threshold was applied to the total urine calcium calculated on 24-hour urine collection. Glomerular filtration rate (GFR) was estimated with the Cockroft–Gault formula. A GFR below 90 mL/minute was considered to be impaired. ⁶

Associations between selected factors and occurrence of hypercalciuria were evaluated with a t-test and a χ ² test, for continuous and categorical variables, respectively. The Cochrane–Mantel–Haenszel method was used to account for sex and age, if appropriate, in the multivariate analysis. Odds ratios were calculated to indicate the strength of the association between calciuria and selected variables, adjusting for gender and age (<44 and ≥45).

RESULTS

Two hundred and seventeen patients were available for the study. According to blood test results and clinical histories we did not identify any secondary causes of hypercalciuria in the sample, so all patients were included in the analysis. The mean age was 46.3 (SD 8.0) and 58.5% were men; risk factors for HIV acquisition showed intravenous drug use in 21.9%, heterosexual transmission in 52.6% and homosexual transmission in 21.9%. Mean CD4+ cell count was 557/mm³ (SD 274). In all, 55.4% of patients were on tenofovir, and 42.7% of the study population was treated with a regimen containing both tenofovir and a protease inhibitor (PI). Despite normal values of creatinine (<1.20 mg/dL in men and <0.95 mg/dL in women) in 95.5% of the sample, impaired glomerular function was detected in 32.5% of patients (GFR < 90 mL/minute). If a stricter cut-off was used (GFR < 60 mL/minute), only four patients had impaired renal function (1 with and 3 without calciuria, respectively; Fisher's exact test P = 1.0). Hypophosphataemia (serum phosphate <2.50 mg/dL) was detected in 17 patients (12.7%) out of 134 who had this measured. Hypercalciuria occurred in 67 patients: the prevalence was 30.9% (95% confidence interval 27.4–37.0). The majority of patients (79.2%) showed low levels of 25 OH-vitamin D (<30 ng/mL). Considering season variability, higher prevalence of hypovitaminosis D was detected in winter (93.0 %), followed by spring (89.7%), autumn (63.6%) and finally summer (41.7%). However, season did not significantly affect the relation between 25-OH vitamin D and hypercalciuria in the multivariate analysis. According to the World Health Organization classification, we detected 14.9% of patients with vertebral osteoporosis and 36.6% with vertebral osteopaenia; only 48.6% of patients had normal lumbar spine T-score values. Bone turnover markers were within normal limits in the majority of patients.

Multivariate analysis (see Table 1) showed that a high level of ionized calcium (>1.30 mg/dL) and a normal level of PTH (15–65 pg/mL) were associated with hypercalciuria. The relation with other bone turnover biochemical markers was not statistically significant. Moreover, the T-scores (lumbar spine, right and left femur) were not related to hypercalciuria, perhaps reflecting the normal bone turnover. Tenofovir alone or with a PI and immunovirological status (CD4 cell count and viral load) were not associated with increased calcium excretion.

DISCUSSION

The main result of this study is the remarkably high prevalence of hypercalciuria in HIV-positive patients, whereas this abnormality occurs in less than 10% of the general population. ⁷ The main strengths of this study are its large sample size and the accurate information about cART, which enabled us to analyse whether a relation existed between hypercalciuria and individual drugs or overall associations. The occurrence of hypercalciuria in subjects with normal values of PTH may indicate an absorptive form of hypercalciuria not necessarily involving bone reserves. In fact, there are two distinct mechanisms of intestinal calcium absorption: an active transcellular transport route promoted by the active form of vitamin D that up-regulates the synthesis of specific calcium channels, and a passive transport related to dietary calcium intake. When calcium intake and ionized serum calcium levels are low, PTH and vitamin D work together to normalize this deficit ⁸ . The lack of a statistical association with vitamin D levels and the significant association with normal values of PTH observed in our cohort seems to lead towards an intestinal dysregulation not dependent on effective serum calcium levels. Moreover, we did not identify any association between hypercalciuria and tenofovir use or hypophosphataemia. These data cannot exclude a primary role of the kidney; however the pathway would be different from the classical pathogenesis of tenofovir-induced toxicity. Hence, if we move away from the hypothesis of kidney dysregulation, the increased levels of ionized serum calcium found in few patients but significantly associated with hypercalciuria may result from intestinal hyperabsorption (possibly by hyperexpression of the vitamin D receptor or increased passive transport) leading to a greater calcium filter load and a suppressed PTH level.

Furthermore, primary tubular dysfunction related to HIV replication was not supported by our study, where hypercalciuric patients had mostly undetectable viral loads (76.0% versus 24.0%).

Our study has several limitations. Firstly, the lack of a control group meant we could only perform an internal comparison. Secondly, data about all factors that may influence calcium serum levels such as sodium and phosphate balance or dietary factors are missing. Thirdly, data about acidosis are lacking, though this metabolic complication can increase the level of ionized calcium.

In conclusion, the results of this study show a high prevalence of idiopathic hypercalciuria in a group of cART-experienced patients: the consequences and the exact causes of this metabolic complication are not known. Further investigations such as the measurement of 1,25-OH vitamin D, phosphorus and sodium balance are needed to better understand the complex pathogenesis of hypercalciuria.