Renal Osteodystrophy in the First Decade of the New Millennium: Analysis of 630 Bone Biopsies in Black and White Patients

Patients

Bone biopsies were done from 2003 to 2008 in patients volunteering to participate in various research protocols. There were 316 patients from the United States and 314 from Europe. Only baseline biopsies were included in the study. All patients received routine dialysis-support medications for at least 6 months, including phosphate binders, active vitamin D, or calcimimetics at the discretion of the treating nephrologists. The study was conducted in adherence to the Declaration of Helsinki. Inclusion criteria were patient age 18 years or older, chronic maintenance dialysis for at least 6 months, and signed informed consent. Exclusion criteria included failed kidney transplant during the past 6 months; uncontrolled systemic illnesses or organic diseases with potential influence on bone metabolism (except diabetes mellitus), such as active or chronic liver disease, malabsorption, malignancy, and thyroid dysfunction; history (during past 12 months) of treatment with pharmacologic agents known to affect bone metabolism (except vitamin D compounds or calcimimetics), such as bisphosphonates, fluoride, calcitonin, glucocorticoids or other immunosuppressive agents, hormone-replacement therapy, and selective estrogen receptor modulators; chronic alcoholism and/or drug addiction; lack of double tetracycline labeling; and bone aluminum accumulation.

Biochemical determinations

Blood was drawn at time of biopsy. Blood was collected in tubes coated with EDTA (lavender tube) for determination of plasma parathyroid hormone (PTH) and in plastic tubes (red-top tube) for determination of serum calcium, phosphorus, and alkaline phosphatase levels. After centrifugation, blood was aliquoted in cryovials and initially stored at −80°C. Serum calcium, phosphorus, and alkaline phosphatase levels were determined using routine laboratory techniques. Plasma PTH levels were obtained using intact Nichols Allegro assay (Nichols Institute Diagnosis, San Juan Capistrano, CA, USA) until April 2006 and the total PTH assay (Scantibodies, Inc, Santee, CA, USA) thereafter. These two assays use comparable antibodies and provide equivalent results.⁽²¹⁾ All plasma PTH levels were determined by the central research laboratory.

Bone biopsy, mineralized bone histology, and histomorphometry

Anterior iliac crest bone biopsies were done after tetracycline labeling under local anesthesia and conscious sedation. The labeling schedule consisted of a 2-day oral administration of tetracycline hydrochloride (500 mg bid) followed by a drug-free interval of 10 days and subsequent oral administration of demeclocycline hydrochloride (300 mg bid) for 4 days. Bone biopsies were performed 3 to 4 days after completing the second label. Bone samples were obtained with the one-step electric drill technique (Straumann Medical, Waldenburg, Switzerland) or a 8-gauge Monoject needle (Sherwood Medical, St Louis, MO, USA). With the electric drill, one bone sample was obtained measuring 5 mm in diameter and 2 to 3 cm in length. With the 8-gauge needle, at least two samples were taken each measuring 3 mm in diameter and 2 to 3 cm in length.^(3,22) Bone samples were processed undecalcified, and sections were stained with the modified Masson–Goldner trichrome stain,⁽²³⁾ the aurin tricar-boxylic acid stain,⁽²⁴⁾ and solochrome azurin.⁽²⁵⁾ Unstained sections were prepared for phase-contrast and fluorescent light microscopy. Histomorphometric analysis of bone was done at standardized sites in cancellous bone and cortical bone using the semiautomatic method (Osteoplan II; Kontron, Munich, Germany) at ×200 magnification.

Bone turnover was classified as low when activation frequency (Ac.f.) was less than 0.49/year and/or bone-formation rate/bone surface (BFR/BS) was less than 1.80 mm³/cm² per year, high when Ac.f. was above 0.72/year and/or BFR/BS was above 3.80 mm³/ cm² per year, and normal when Ac.f. was between 0.49 and 0.72/ year and/or BFR/BS was 1.80 to 3.80 mm³/cm² per year. Defective mineralization was identified when osteoid thickness in lamellar bone was above 20 μm and mineralization lag time in lamellar bone was above 50 days. Cancellous bone volume/tissue volume (BV/TV) was classified as low when it was below 16.8%, normal when it was between 16.8% and 22.9%, and high when it was above 22.9%. Cortical thickness was considered to be low when values were below 0.52 mm, normal when values were between 0.52 and 1.65 mm, and high when values were above 1.65 mm. Cortical porosity was low when values were below 1.9%, normal when values were between 1.9% and 10%, and high when values were above 10%. Trabecular thickness (Tb.Th) was classified as low when it was below 99 μm, normal when it was between 99 and 142 μm, and high when it was above 142 μm. Trabecular separation (Tb.Sp) was classified as low when it was below 280 μm, normal when it was between 280 and 658 μm, and high when it was above 658 μm. Normal ranges were obtained from bone samples taken in healthy individuals covering the gender, race, and age range of the study population.^(3,26)

Statistical analysis

Results are given as means ±SEM for continuous variables or percentage for categorical variables. Differences between the various groups were performed using a t test or one-way analysis of variance with a Bonferroni post-hoc test for continuous variables and a chi-square test for categorical variables. Logistic regression for prediction of low bone turnover or low bone volume in black and white patients was done with the following variables entered in the model: age, dialysis vintage, gender, plasma PTH levels, serum calcium and phosphorus levels, treatment with vitamin D, phosphate binders, and dialysate calcium. All computations were done using SPSS Version 7.5 (SPSS, Inc, Chicago, IL, USA).

Patient characteristics

Six hundred and fifty-one patients were enrolled. Twenty-one patients were excluded from the study because of stainable bone aluminum (n =4), inadequate sample size and/or quality (n =7), or absence of tetracycline double labeling (n =10); thus 630 bone samples were analyzed. Patients included 329 men and 301 women. Their mean age was 55 ±1 years (19 to 84 years). Six hundred patients were on hemodialysis, and 30 patients underwent peritoneal dialysis. Dialysis vintage was 51 ±1.8 months (3 to 240 months). There were 87 black patients, 543 white patients, 141 diabetics (type 2: n =139; type 1: n =2), 109 patients treated with active vitamin D metabolites—iv calcitriol in 72 patients (mean dose 1.45 ±0.13 μg/d), paricalcitol in 32 patients (mean dose 5.69 ±0.59 μg/d), and doxercalciferol in 5 patients (mean dose 2.92 ±0.35 μg/d)—and 429 patients treated with calcium-containing phosphate binders. Only 4 patients received calcimimetics. Dialysate calcium concentration was 2.5 mEq/L in 371 patients and 3.5 mEq/L in 259 patients.

All black patients were from the United States. There were no differences between white American and European patients in biochemical parameters or patient characteristics such as gender distribution, percentage of diabetics, or treatment with vitamin D or calcium-based phosphate binders, except that white European patients were slightly older (58 ±0.77 years versus 53 ±1.00 years, p <.05) and on dialysis for a longer period of time (58 ±2 months versus 46 ±5 months, p <.05). There were no differences in patient characteristics and biochemical results between males and females, diabetics and nondiabetics, and patients with or without active vitamin D treatment. Also, no differences were found between patients treated with or without calcium-containing phosphate binders except for moderately lower plasma parathyroid hormone (PTH) levels in patients receiving calcium-containing phosphate binders (285 ±18 pg/ mL versus 391 ±32 pg/mL, p <.002). When patients were classified by race, black patients were younger, were on dialysis for a shorter period of time, were more likely to be treated with active vitamin D analogues, and were more likely to receive non-calcium- and non-aluminum-containing phosphate binders (Table 1). Serum phosphorus, alkaline phosphatase, and PTH levels were higher in blacks than in whites (Table 1).

Bone turnover (T)

Overall, 58% of patients exhibited low bone turnover, 18% normal turnover, and 24% high turnover. There was no difference in distribution of turnover abnormalities between male and female patients, diabetics and nondiabetics, and patients treated with vitamin D and not. There were more patients with low bone turnover dialyzed with 3.5 mEq/L of calcium bath than with 2.5 mEq/L (71.8% versus 48.5%, p <.01). When patients were analyzed according to race, there was a statistically significant difference in the distribution of these abnormalities ( p <.001; Fig. 1). White patients exhibited predominantly low bone turnover, whereas high bone turnover was the prominent feature in blacks (Fig. 1). White and black patients with high bone turnover were younger than patients with low turnover (Table 2). In both races, patients with high bone turnover had higher blood phosphorus and PTH levels (Table 2). At low, normal, or high bone turnover in both races there was no difference in gender, dialysis vintage, presence of diabetes, and serum calcium and alkaline phosphatase levels (Table 2). However, at any level of plasma PTH, bone turnover was lower in blacks than in whites; that is, the ratio PTH/Ac.f. was higher in blacks than in whites (2702 ±1564 versus 720 ±41, p <.001). Logistic regression analysis revealed that PTH levels, age, and calcium dialysate content were predictors of low bone turnover in white patients ( p <.001). In black patients, no significant predictor of bone turnover could be established. Gender, diabetes, vitamin D treatment, or calcium-containing phosphate binder treatment were not predictive of bone turnover in blacks and whites.

Mineralization (M)

Only 21 patients (17 whites and 4 blacks) showed a mineralization defect characterized by wide osteoid seams (>20 μm) and delayed mineralization lag time (>50 days). None of these patients exhibited stainable bone aluminum or iron. Patients with defective mineralization were on dialysis for longer periods of time and had lower serum calcium levels and higher levels of alkaline phosphatase and intact PTH (Table 3). There were no other significant differences (Table 3). Patients with mineralization defects had either low (n =10), normal (n =8), or high bone turnover (n =3). No apparent clinical reason was documented to explain the mineralization defect.

Volume (V)

In white patients, there was approximately the same proportion of patients with low, normal, and high cancellous bone volume, whereas high cancellous bone volume was the predominant feature in black patients (Fig. 2). Cortical parameters were available in 418 patients. The majority of black patients had normal cortical thickness but high porosity, whereas there was approximately the same number of white patients with low or normal cortical thickness and normal or high porosity (Figs. 3 and 4).

In both races, changes in cortical and cancellous bone were not different when analyzed by gender, presence of diabetes, dialysis vintage, and treatment with active vitamin D or phosphate binders. Black and white patients with high cancellous bone volume and/or thick cortices were significantly younger than patients with low bone volume (Tables 4 and 5). There was no age difference between patients with normal or high cortical porosity (Table 6). In both races, there were no differences in dialysis vintage, serum calcium level, and alkaline phosphatase level between patients with low, normal, or high cancellous bone volume, cortical thickness, or porosity (Tables 4 through 6). In contrast, in both races, serum phosphorus and PTH levels were higher in patients with high cancellous bone volume (Table 4).

Logistic regression identified in white patients low PTH levels and female gender as predictors of low bone volume ( p <.01), whereas only older age predicted low bone volume in black patients ( p <.01).

Architecture (A)

Trabecular thickness was low in 37% of patients, normal in 40%, and high in 13%. Trabecular separation was normal in most patients (78%), high in 6%, and low in 16%. There was no difference in trabecular thickness and separation when analyzed by race, gender, presence by diabetes, dialysis vintage, and treatment with active vitamin D or phosphate binders.

Interactions among turnover, mineralization, volume, and architecture

Both black and white patients with high bone turnover had increased cortical porosity (69% and 67%, respectively, p <.01). Patients with increased cortical porosity had significantly higher erosion depth than patients with normal porosity (15.90 ±0.77 μm versus 13.70 ±0.73 μm, p <.03).

There was no recognizable relationship between defective mineralization and bone turnover, cancellous bone volume, or cortical thickness. Most white patients with low cancellous bone volume or thin cortices had low bone turnover (73% and 68%, respectively). More than 80% of black and white patients with low cancellous bone volume had thin trabeculae ( p <.001; Fig. 5). Patients with thin trabeculae had a lower bone-formation rate/bone surface (2.17 ±0.12 mm³/cm² per year) than patients with normal or increased trabecular thickness (3.31 ±0.19 mm³/ cm² per year and 4.15 ±0.41 mm³/cm² per year, respectively, p <.001), whereas there was no difference in parameters of bone resorption (erosion depth 13.88 ±0.37 μm versus 15.41 ± 0.41 μm and 16.99 ±0.95 μm, respectively). There was no interaction between trabecular separation and cancellous bone volume in both racial groups.