Effects of Alendronate on Bone Quality and Remodeling in Glucocorticoid-Induced Osteoporosis: A Histomorphometric Analysis of Transiliac Biopsies



Effects of alendronate (ALN) on bone quality and turnover were assessed in 88 patients (52 women and 36 men aged 22–75 years) who received long-term oral glucocorticoid exposure. Patients were randomized to receive oral placebo or alendronate 2.5, 5, or 10 mg/day for 1 year and stratified according to the duration of their prior glucocorticoid treatment. Transiliac bone biopsies were obtained for qualitative and quantitative analysis after tetracycline double-labeling at the end of 1 year of treatment. As previously reported in glucocorticoid-induced osteoporosis, low cancellous bone volume and wall thickness were noted in the placebo group as compared with normal values. Alendronate treatment was not associated with any qualitative abnormalities. Quantitative comparisons among the four treatment groups were performed after adjustment for age, gender, and steroid exposure. Alendronate did not impair mineralization at any dose as assessed by mineralization rate. Osteoid thickness (O.Th) and volume (OV/BV) were significantly lower in alendronate-treated patients, irrespective of the dose (P = 0.0003 and 0.01, respectively, for O.Th and OV/BV); however, mineral apposition rate was not altered. As anticipated, significant decreases of mineralizing surfaces (76% pooled alendronate group; P = 0.006), activation frequency (–72%; P = 0.004), and bone formation rate (–71%; P = 0.005) were also noted with alendronate treatment. No significant difference was noted between the changes observed with each dose. Absence of tetracycline label in trabecular bone was noted in approximately 4% of biopsies in placebo and alendronate-treated groups. Trabecular bone volume, parameters of microarchitecture, and resorption did not differ significantly between groups. In conclusion, alendronate treatment in patients on glucocorticoids decreased the rate of bone turnover, but did not completely suppress bone remodeling and maintained normal mineralization at all alendronate doses studied. Alendronate treatment did not influence the osteoblastic activity, which is already low in glucocorticoid-induced osteoporosis.


Glucocorticoid-induced osteoporosis is a major cause of secondary osteoporosis in men and women. Importantly, osteoporosis and associated fractures, especially of the spine, hip, and rib, are the most common complication of glucocorticoid use and result in significant pain, disability, and morbidity.(1) Glucocorticoids produce bone loss by two major mechanisms: (1) direct suppression of osteoblastic bone formation, resulting from a decrease in osteoblastic activity at the cellular level; and (2) increased osteoclast-mediated bone resorption, likely resulting from secondary hyperparathyroidism caused by glucocorticoid-induced impairment of intestinal calcium absorption. In some patients, hypogonadism also contributes to excess bone resorption and is mediated both by direct and indirect (via the pituitary) effects of glucocorticoids on gonadal hormone production.(2–4) The effects of glucocorticoids on bone tissue are partly reversible, as observed in patients with Cushing's syndrome.(5) A wide variety of therapies such as calcium, vitamin D metabolites, calcitonin, etidronate, or fluoride have been studied and proposed to treat glucocorticoid-induced osteoporosis.(4) However, studies with these agents have generally been on a relatively small number of patients and have shown only modest or equivocal efficacy.(6) Moreover, bone biopsies have not been obtained in these studies to assess the effects of these agents at the tissue level.

Bisphosphonates are analogs of inorganic pyrophosphate that are taken up by bone, bind to the bone mineral surface, and inhibit osteoclastic bone resorption.(7) Alendronate sodium is a potent inhibitor of bone resorption that has been extensively studied in the treatment of postmenopausal osteoporosis and in the prevention of postmenopausal bone loss.(8–11) It has been shown to increase bone mineral density (BMD) of the spine, hip, and total body.(8,12–15) These changes in bone density are associated with a significant reduction of fracture incidence at both vertebral and non-vertebral sites including the hip.(8,9,16–18) Histomorphometric analyses have reported a decrease in bone turnover,(19–20) which has been confirmed by biochemical markers of bone turnover.(12,13,20,21) In men and women on glucocorticoids, alendronate has been shown to increase BMD in the spine, hip, and total body.(22)

Bone histomorphometry is a powerful tool in its ability to assess the quality of bone and to evaluate the effects of treatment on bone mineralization and microarchitecture. In addition, histomorphometry enables quantitative assessment of treatment-related changes in several indices of bone remodeling, at the cell and tissue levels. The purpose of this study was to evaluate by histomorphometry the effects of alendronate on transiliac bone biopsies taken from patients receiving chronic glucocorticoid treatment. The objectives were to evaluate alendronate safety (bone quality) and efficacy (effect on bone turnover) at the bone tissue level after 1 year of treatment.


Description of clinical trials

Two studies were conducted:(22) one in the United States (U.S., 15 centers) and the other in five countries around the world (Multinational, 22 centers). In both studies, glucocorticoid-treated patients were randomized to receive either daily placebo or alendronate 2.5 (Multinational study only), 5, or 10 mg/day for 1 year, and the study design and BMD results have been previously reported.(22) Of the 232 patients enrolled in the U.S. study, 23 patients consented to have a bone biopsy at the end of treatment (from 5 of the 15 centers). Sixty-five of the 245 patients enrolled in the Multinational study consented to biopsy (from 9 of the 22 centers). A total of 88 iliac biopsies obtained in 36 men (mean age 53.5 ± 15.6 years) and 52 women (mean age 53.1 ± 15.6 years) were analyzed.

To enter these studies, patients had to be receiving an oral glucocorticoid at an average daily dose equivalent to at least 7.5 mg of prednisone and considered very likely (>90% chance) to require glucocorticoids at or above this minimum dose for the next year. Patients were receiving glucocorticoids for either rheumatic disorders, asthma, dermatologic, intestinal, or renal diseases. At baseline, all biochemical markers were in or close to the normal range and serum 25-OH vitamin D concentrations were between 10 and 80 ng/ml. Information regarding calcium intake was collected at baseline using a food-frequency questionnaire, which provided an estimated average of the daily intake of calcium from dietary sources, as well as from the concomitant use of supplements. Calcium (800–1000 mg) and vitamin D (250–500 IU) supplements were provided for all patients. Baseline mean calcium intake (including supplements) was 1584 mg/day. To ensure a balanced enrollment of patients with a wide range of prior glucocorticoid exposure, patients were stratified according to duration of prior glucocorticoid use at entry (<4months, 4–12 months, >12 months), irrespective of their bone mineral density.

These studies were sponsored by Merck Research Laboratories (Rahway, NJ, U.S.A.), which also provided the alendronate tablets and matching placebo. All patients provided informed consent for the overall study and separate consent for the biopsy procedure both before the initiation of treatment and again before administering the tetracycline for the biopsy labeling. In all cases, the studies had prior approval from the appropriate Institutional Review Board/ethics committee.


Biopsy procedure: Transiliac bone biopsies were obtained using a trephine needle (Meunier modification of Bordier trephine; Lepine, Lyon, France) with a minimal internal diameter of 7.5 mm. Before biopsy, patients received either tetracycline 250 mg four times daily or deme-clocycline 300 mg twice daily (for 2 days on; 12 days off; 2 days on), with the last dose timed to occur 4 to 6 days before the bone biopsy procedure. Tetracycline is incorporated onto bone surfaces as they undergo mineralization. The length of individual tetracycline-labeled surfaces are added together and then reported as a percentage of the total cancellous surface (mineralizing surface; MS/BS). This measurement provides an assessment of the rate of bone formation at the tissue level. Bone specimens were transported in 70% ethanol to the laboratory responsible for the central analysis of biopsies (P. J. Meunier, Lyon, France).

Biopsy processing: Biopsies were dehydrated in graded alcohols and embedded in methylmethacrylate. Serial sections were cut at three different levels sufficiently far apart (at least 250 μm) to avoid replicate sampling of a single bone remodeling unit. Some sections were stained with Goldner's trichrome, whereas others were left unstained for evaluation under ultraviolet light. Solochrome cyanin R staining was used for the assessment of wall thickness (W.Th). Only biopsies that provided a large enough area of intact cancellous bone were assessed quantitatively, because smaller samples can provide misleading data related to sampling errors. For example, the rate of bone turnover is reflected histomorphometrically by the number of active remodeling sites, and, on average, is proportional to the MS. Although MS generally correlates with the rate of bone turnover, a sample containing few surfaces may either grossly underestimate or overestimate the actual rate of turnover in the bone as a whole. To limit such sampling errors, which are more likely to occur when the rate of bone turnover is low (as is generally observed in patients receiving glucocorticoids), an adequate specimen was prospectively defined as one, with a minimum of either 20 mm2 of intact cancellous tissue area for samples containing two or more tetracycline labels, or 40 mm2 for samples with fewer than two labels. From the 88 specimens obtained, 31 incomplete or crushed biopsies were excluded from quantitative histomorphometric analysis but were included in the qualitative histological assessment. Five other specimens mistakenly taken at the top of the iliac crest were also excluded from the quantitative analysis. Thus, 52 specimens were measured and analyzed.

Qualitative examination: The appearance of the cellular components and the presence of tetracycline labels, woven bone, or marrow fibrosis, or any other noteworthy features were assessed qualitatively in all 88 biopsies.

Quantitative analysis: The entire cancellous tissue area and the endocortical bone surface of each section were analyzed separately. To confirm the predefined hypothesis (derived from preclinical studies) that therapeutic doses of alendronate would not impair mineralization in humans, the following endpoints were measured/calculated from Goldner-stained sections (except as noted) in cancellous and endocortical bone separately and named in accordance with ASBMR Committee nomenclature:(23)

  • Osteoid volume/bone volume (OV/BV) in cancellous bone, expressed as a percentage of cancellous bone volume.

  • Osteoid surface/bone surface (OS/BS), expressed as a percentage of cancellous and endocortical bone surface.

  • Osteoid thickness (O.Th) expressed in μm.

  • Mineral apposition rate (MAR) on unstained sections under ultraviolet light expressed in μm/day. The rate of mineralization is assessed by measurement of the distance between the midpoints of two consecutive tetracycline labels. Dividing this distance by the known time interval between tetracycline labels provides the MAR. Osteomalacia is characterized by an increase in unmineralized bone (osteoid thickness and volume) and a decrease in the rate of mineralization (MAR).

To estimate the effects of alendronate on bone turnover at the site of biopsy and to investigate the mechanism of action by which alendronate induced the observed increase in bone density, the following parameters were measured on Goldner-stained sections (excepted wall thickness measured on Solochrome Cyanin R–stained sections) in cancellous and endocortical bone separately:

  • Eroded surface (ES/BS), the percentage of the total bone surface that includes both active (with presence of osteoclasts) and inactive (without osteoclasts) eroded surfaces.

  • Osteoclast surface (Oc.S/BS), the percentage of the total bone surface that consists of active eroded surfaces

  • Osteoclast number (N.Oc/BS) per mm of bone surface.

  • Erosion depth (E.De) derived after estimation of the topography of the previous trabecular bone surface resorption cavities on an image analyzer.(24) All resorption cavities were measured.

  • Eroded volume (EV/BV), the amount of bone eroded as a percentage of cancellous bone volume was measured as before.(24)

  • Mneralizing surface, an estimate of the rate of bone formation, was expressed as a percentage of the total bone surface (MS/BS). The extent of the mineralizing surface was calculated as the length of double-labeled surface plus half of the single-labeled surface.(23)

  • Wall thickness (W.Th) of cancellous packets measured under polarized light. Only completed packets bearing no osteoblast or osteoclast were measured.

  • Bone formation rate (BFR/BS) was calculated as MS/BS times MAR and expressed in μm3/μm2/day.

  • Adjusted apposition rate (Aj.AR) was calculated as BFR/OS and expressed in μm/day.

  • The formation period (FP) was calculated as W.Th/Aj.AR. The active formation period (FPa+) was derived from W.Th/MAR. FP and FP(a+) were expressed in days.

  • The activation frequency (Ac.f), which represents the probability that a new cycle of remodeling will be initiated at any point of the bone surface, was calculated as (BFR/BS)/W.Th and expressed per year.

  • Mneralization lag time (Mlt), which expressed the delay of the mineralization onset, was calculated as O.Th/Aj.AR.

To evaluate the effects of alendronate on bone structure and microarchitecture, the following parameters were measured on cancellous bone area:

  • Cancellous bone volume (BV/TV), which represents the percentage of cancellous bone tissue including mineralized bone and osteoid.

  • Trabecular thickness (Tb.Th, μm), trabecular separation (Tb.Sp, μm), and trabecular number (/mm) were calculated from cancellous and trabecular area and perimeter.(25)

All thickness/depth results (O.Th, MAR, E.De, W.Th, Tb.Th, Tb.Sp) were corrected for obliquity of sections by multiplying by π/4.

All parameters were measured by using a semiautomatic (Ibas 1; Leica, Heerbrugg, Switzerland) or an automatic (visiolab 5000; Biocom, Les Ulis, France) analyzer. Intra-and interobserver coefficients of variation for parameters measured were less than 6% for BV/TV, OS/BS, OV/BV, ES/BS, MS;(26) 2.1% for W.Th; and 6% for EV/BV and E.De (unpublished data).

Table Table 1.. Baseline Patient Characteristics
Treatment (n)Placebo (15)ALN 2.5 mg (10)ALN 5 mg (11)ALN 10 mg (16)P Value
  1. a Results expressed as means (SEM).

  2. b Kruskall–Wallis test. NS, not significant.

  3. c χ2 test.

Age (years)a47.6 (3.8)54.4 (4.7)47.8 (5.4)50.8 (3.3)0.7b
Stratum (n)    0.4c
 <4 months2345 
 4–12 months4116 
 >12 months9665 
Sex (n)    0.4c
BMD (T-score)a
 Lumbar spine−1.31 (0.43)−1.31 (0.55)−1.48 (0.48)−1.21 (0.35)NSb
 Femoral neck−1.82 (0.33)−1.57 (0.32)−1.73 (0.45)−1.60 (0.31)NSb
Fractures (n)

Statistical analyses and hypotheses

The results were expressed as means ± SEM after adjustment for the age, sex, and stratum of prior glucocorticoid administration. The normality of the distribution was tested. The distribution was normal after a logarithm transformation for OS/BS, O.Th, ES/BS, Oc.S/BS, N.Oc/BS, EV/BV, W.Th, BFR/BS, Aj.AR, FP, FP(a+), Mlt, Ac.f, and Tb.Sp.

Differences among the four treatment groups were tested by one-way analysis of variance (ANOVA) using age, gender, and stratum as covariate. The adjustment was made by a covariance analysis. If the ANOVA was significant (P < 0.05), the comparison between two groups was performed by a Newman–Keuls test. For MS/BS, no transformation was available and comparison between groups was performed by a nonparametric Kruskall–Wallis test. If the difference was significant, the comparisons between two groups were performed by a Mann–Whitney U test.

It was hypothesized that alendronate would have an inhibitory effect on bone turnover in iliac cancellous bone, as shown by a decrease in MS relative to placebo. It was also hypothesized that in patients with GIOP, treatment with alendronate for up to 1 year would neither impair mineralization, nor be associated with any qualitative abnormalities (such as woven bone or marrow fibrosis). With a sample size of 15 patients per group, the power to detect a 3% difference in MS/BS was equal to 88%, assuming a standard deviation of 2.5% for the variable MS/BS.


Baseline characteristics (age, glucocorticoid stratum, gender, prevalent fracture status, and BMD) of patients who underwent bone biopsy were comparable to patients who did not undergo biopsy in the overall population within each study. Inadequate biopsies were predominantly the result of the samples' being either crushed or incomplete as a result of the biopsy procedure. Baseline characteristics of patients with an adequate biopsy for bone histomorphometry were similar in the four treatment groups (Table 1). The median cumulative glucocorticoid intake (equivalent of prednisone) during the study was similar in the placebo group (3.30 g, range 1.48 to 17.78) compared with the pooled alendronate group (3.36 g, range 1.53 to 11.34).

Effects on bone quality

The qualitative assessment of bone revealed no abnormalities in biopsies of alendronate-treated patients. In all alendronate-treated patients, newly formed bone retained its normal lamellar structure without woven bone, and there was no evidence of marrow fibrosis. Two biopsies from placebo-treated patients had mild or moderate marrow fibrosis. Absence of tetracycline label in trabecular bone was noted in biopsies from 1 of 23 patients (4.3%) on placebo and 3 of 65 patients (4.6%) receiving alendronate.

Effects on trabecular bone

Mineralization: Table 2 shows the results for the three primary predefined endpoints for assessment of mineralization. Osteoid thickness (O.Th) and osteoid volume (OV/BV) were significantly lower in alendronate-treated patients relative to those who received placebo, but no significant difference was found between alendronate doses. No change in mineral apposition rate (MAR) was observed with alendronate treatment. Mineralization lag time tended to increase with alendronate dose, but the difference was not significant.

Table Table 2.. Effects of Alendronate on Static and Dynamic Mineralization Parameters
Treatment (n)Placebo (15)ALN 2.5 mg (10)ALN 5 mg (11)ALN 10 mg (16)
  1. Adjusted mean (SEM). Comparison between the four treatment groups was performed by ANOVA. Comparison between two groups was performed by a Newman–Keuls test.

  2. a Statistical tests were performed after logarithm transformation.

  3. * P < 0.001 versus placebo.

O.Th (μm)a8.9 (0.7)5.9 (1.0)*6.1 (0.4)*5.8 (0.2)*
OV/BV (%)1.6 (0.4)0.5 (0.2)*0.7 (0.3)*0.3 (0.1)*
MAR (μm/day)0.54 (0.03)0.56 (0.05)0.58 (0.08)0.55 (0.03)
Mlt (days)a57.7 (8.0)72.7 (19.7)123.2 (35.1)236.9 (90.0)
Table Table 3.. Effects of Alendronate on Bone Remodeling
Treatment (n)Placebo (15)ALN 2.5 mg (10)ALN 5 mg (11)ALN 10 mg (16)
  1. Adjusted mean (SEM). Comparison between the four treatment groups was performed by ANOVA or Kruskall–Wallis test (MS/BS). Comparison between two groups was performed by a Newman–Keuls or Mann–Whitney U (MS/BS) tests.

  2. a Statistical tests were performed after logarithm transformation.

  3. * P < 0.05; P < 0.01 versus placebo.

MS/BS (%)4.4 (1.0)0.9 (0.6)1.7 (1.0)0.6 (0.3)
OS/BS (%)a9.9 (2.2)5.0 (1.6)7.1 (2.2)5.3 (0.9)
ES/BS (%)a2.3 (0.4)2.2 (0.4)2.2 (0.5)2.6 (0.5)
EV/BV (%)a0.44 (0.1)0.40 (0.1)0.44 (0.1)0.50 (0.1)
Oc.S/BS (%)a0.075 (0.03)0.07 (0.03)0.067 (0.02)0.097 (0.04)
N.Oc/BS (mm−1)a0.016 (0.006)0.014 (0.007)0.013 (0.005)0.019 (0.008)
Max. EDe (μm)15.0 (1.3)13.4 (1.0)15.4 (1.1)16.2 (1.0)
Mean EDe (μm)9.4 (0.8)8.4 (0.5)9.2 (0.4)10.3 (0.7)
BFR/BS (μm3/μm2/day)a0.027 (0.006)0.010 (0.004)0.007 (0.004)*0.006 (0.003)*
Aj.AR (μm/day)a0.251 (0.040)0.137 (0.040)0.087 (0.020)0.123 (0.042)*
Ac.f (/year)a0.312 (0.064)0.117 (0.04)0.085 (0.042)*0.067 (0.039)
FP (days)a214.4 (29.3)316.6 (81.1)719.8 (244.5)*1319.8 (551.3)*
FP(a+) (days)a59.9 (3.9)57.9 (7.7)60.6 (6.1)61.5 (4.1)

Bone turnover: As shown in Table 3, all resorption parameters were similar in patients treated with alendronate and placebo. For each alendronate dose, mean values for osteoid surfaces tended to decrease, but this difference was not significant when compared with the placebo group.

The mineralizing surface, bone formation rate, and activation frequency each showed marked and significant decreases with alendronate 10 mg. This was associated with an increase in the total formation period. However, the active formation period remained unchanged. The mean value of adjusted apposition rate decreased with the highest dose of alendronate (Table 3), confirming the results noted at the tissue level. Bone volume, mean wall thickness, and parameters of microarchitecture were unchanged (Table 4).

Effects on endosteal bone

The effects of alendronate were similar on endosteal and trabecular bone surfaces. The magnitude of decreases in O.Th, MS, BFR/BS, and Ac.f were similar in these two compartments. At 10 mg, relative to placebo, mineralizing surface was lower by 86 and 84%, whereas activation frequency was lower by 78 and 71% in trabecular and endosteal bone, respectively (Table 5).


The main pathophysiologic mechanism underlying glucocorticoid-induced osteoporosis is a decrease in osteoblastic activity, which results in a reduced wall thickness with a shortening of the period of time during which osteoblasts synthesize bone matrix.(2,3,27) The depressed bone formation is often combined with an increase in bone resorption and rate of bone turnover, which may be attributed to a secondary hyperparathyroidism and hypogonadism. In combination, these effects at the cellular and tissue level result in bone loss and an increased risk of fracture.(4)

The purpose of the present study was to evaluate the effects of alendronate in glucocorticoid-treated patients. All patients were treated with calcium (mean total intake 1584 mg/day) and vitamin D. The studied population was heterogeneous, including men and pre- and postmenopausal women aged 17 to 83 years. Furthermore, patients were receiving glucocorticoids for various underlying diseases, some of which may affect bone directly, and thus contribute to the development of osteoporosis, which may have multiple etiologies in an individual patient. The patients were stratified based on prior duration of glucocorticoid use, irrespective of their baseline bone mineral density. The age, gender, and glucocorticoid use were similar in the four treatment groups. Because alendronate effects on bone may vary according to the patient's age, gender, or glucocorticoid exposure, all analyses were performed after adjustment for these three factors, which took into account the variations related to these factors. The differences observed were the result of only the alendronate treatment.

Table Table 4.. Effects of Alendronate on Cancellous Bone Volume and Bone Microarchitecture Parameters
Treatment (n)Placebo (15)ALN 2.5 mg (10)ALN 5 mg (11)ALN 10 mg (16)
  1. Adjusted mean (SEM). Comparison between the four treatment groups was performed by ANOVA. Comparison between two groups was performed by a Newman–Keuls test. No significant difference was found between the four treatment groups.

  2. a Statistical tests were performed after logarithm transformation.

BV/TV (%)15.5 (0.9)13.6 (2.4)14.4 (2.1)16.3 (1.3)
Tb.Th (μm)97.1 (5.4)99.8 (8.9)95.1 (6.3)106.5 (6.2)
Tb.Sp (μm)a548.1 (26.2)726.0 (89.5)683.1 (96.9)583.4 (50.9)
Tb.N (/mm)1.27 (0.05)1.05 (0.11)1.10 (0.10)1.20 (0.10)
W.Th (μm)a30.6 (0.8)30.0 (1.0)31.2 (0.9)31.4 (1.0)
Table Table 5.. Effects of Alendronate on Bone Remodeling on Endocortical and Cancellous Bone
Treatment (n)Placebo (15)ALN 2.5 mg (10)ALN 5 mg (11)ALN 10 mg (16)
  1. Adjusted mean (SEM). Comparison between the four treatment groups was performed by ANOVA. Comparison between two groups was performed by a Newman–Keuls test.

  2. a Statistical tests were performed after logarithm transformation.

  3. *P < 0.05; P < 0.01; P < 0.0002 versus placebo.

Endocortical bone
O.Th (μm)10.2 (1.3)7.1 (1.0)*6.8 (0.7)*5.3 (0.3)
MS/BS (%)5.1 (1.3)1.5 (0.7)*3.7 (3.4)*0.8 (0.3)
OS/BS (%)a15.1 (3.8)7.2 (2.7)11.8 (3.9)6.7 (1.0)
BFR/BS (μm3/μm2/day)a0.037 (0.008)0.015 (0.005)0.029 (0.016)0.011 (0.003)
Ac.f (/yr)*0.394 (0.085)0.156 (0.048)0.349 (0.198)0.113 (0.032)*
Cancellous bone
O.Th (μm)8.9 (0.7)5.9 (1.0)*6.0 (0.4)*5.8 (0.2)
MS/BS (%)4.3 (1.0)0.9 (0.6)1.6 (1.0)*0.6 (0.3)
OS/BS (%)a10.2 (2.2)5.0 (1.6)6.9 (2.2)5.3 (0.9)
BFR/BS (μm3/μm2/day)a0.027 (0.006)0.010 (0.004)0.007 (0.004)0.006 (0.003)
Ac.f (/year)a0.314 (0.064)0.117 (0.044)0.086 (0.042)0.075 (0.039)

Effects of glucocorticoids on bone

The expected effects of glucocorticoids on bone were observed in the placebo group: cancellous bone volume, wall thickness, and mineralizing surface were significantly lower than values in normal age-matched controls.(28–30) The effects of glucocorticoids were similar for men and women, except for mineralizing surface. In the placebo group, MS/BS were lower in women than in men (2.75 ± 1.45 and 5.71 ± 1.22%, respectively), as previously reported in normal controls.(31) This finding is consistent with those previously reported in untreated corticosteroid-induced osteoporosis.(5,27) Mineralization rate was slightly decreased when compared with control values. However, these parameters were higher than those previously reported by our own laboratory from studies of patients with more severe glucocorticoid-induced osteoporosis, as defined by radiological evidence of vertebral fractures.(2,5,27) The main histologic feature of corticosteroid-induced osteoporosis is decreased bone formation. Increased bone resorption, decreased osteoblast proliferation and activity, sex-steroid deficiency, and secondary hyperparathyroidism have all been proposed as mechanisms of the corticosteroid-induced bone loss.(3,4) Recently, a suppressive effect of corticosteroids on osteoblastogenesis and a promotion of osteoblast apoptosis have been suggested.(32)

In a recent study,(19) the effects of alendronate treatment have been evaluated in 231 women with postmenopausal osteoporosis, including 71 who received placebo. In the present study, only 3 patients in the placebo group were postmenopausal women not receiving estrogen replacement therapy. Despite the small number of patients, the comparison of mean values of bone parameters between glucocorticoid-treated and nontreated postmenopausal women showed that trabecular thickness was 20% lower and trabecular number 20% higher in glucocorticoid-treated patients than in postmenopausal women with osteoporosis. These observations tend to confirm that glucocorticoid-induced osteoporosis is characterized by a greater thinning of trabeculae, but better connectivity than can be found in postmenopausal osteoporosis.(33)

Effects of alendronate

Qualitative effects: The qualitative assessment of bone biopsies in glucocorticoid-treated patients showed that alendronate did not induce any qualitative bone structure abnormalities such as woven bone or marrow fibrosis, consistent with our previous observations in postmenopausal osteoporosis.(19)

Effects on mineralization: The decrease in osteoid volume without any change in mineral apposition rate confirmed the decreased rate of bone turnover previously shown with biochemical markers during alendronate treatment.(22) This was associated with decreases in osteoid volume and osteoid thickness that are accounted for by a decrease in the rate of bone turnover. In a steady state, the rate at which osteoid newly appears at any point on the surface is the same as the frequency of remodeling activation, and the average lifespan of the osteoid seam at any point on the surface is the same as the formation period.(34,35) The proportionate decline in MS/BS and OV/BV without change in MAR indicates the lack of a mineralization defect in association with alendronate treatment of glucocorticoid-induced osteoporosis and is consistent with the histomorphometric results from studies of postmenopausal osteoporosis(19,20) and Paget's disease.(36,37) These observations are consistent with the expected effects of a treatment-related decrease in the rate of bone turnover and the absence of any morphological or dynamic evidence of an impairment of mineralization. These effects of alendronate were similar in women and in men.

Effects on bone turnover: As expected, alendronate markedly decreased the rate of bone turnover. The mineralizing surfaces were reduced (relative to placebo) by 85% and the activation frequency by 79% at the dose of 10 mg/day. Consequently, the bone formation rate at the tissue level was considerably decreased. No significant difference among daily doses of alendronate was noted. However, the mean values of bone formation rate and activation frequency tended to decrease with increasing doses. The magnitude of the reduction in bone turnover was similar to that previously reported in elderly postmenopausal women with osteoporosis.(20) However, in both postmenopausal and glucocorticoid-treated populations, biochemical markers of bone turnover have shown maximum decreases in bone resorption (∼70%) and formation (∼40%) within 3 to 6 months of the alendronate treatment initiation, with no further decrease thereafter with continued treatment.(13,38) The differences in the proportional degree of suppression in turnover assessed by biochemical markers relative to histomorphometry of iliac cancellous bone is most likely related to the greater targeting of bisphosphonates to sites of high bone turnover and high blood flow.

Absence of detectable tetracycline label in cancellous bone was noted in 4–5% of biopsies, in both placebo and alendronate-treated groups, almost certainly reflecting the sampling limitations inherent in bone histomorphometry. These data confirm that even under conditions of depressed bone formation (as can be found with glucocorticoid treatment), alendronate reduced but did not totally suppress bone turnover. Similar results have been previously reported in postmenopausal osteoporosis(19) and in animal studies of alendronate.(39)

Effects on bone resorption parameters: Alendronate had no effect on histomorphometric bone resorption parameters, including osteoclast surface, eroded surface, eroded volume, and osteoclast number. In contrast, alendronate significantly inhibited urinary excretion of bone collagen breakdown products (a biochemical marker of bone resorption).(22) Similar results were found in postmenopausal studies.(8,19,21,38) The lack of decrease in OcS and ES/BS suggests that the predominant effect of alendronate is not only to reduce activation frequency, but also to decrease osteoclast function (and rate of bone turnover), rather than to decrease osteoclast numbers; in other words, reducing bone resorption through slowing down function. Furthermore, a prolongation of the osteoblastic phase of the remodeling cycle or a decreased resorption rate may result from the lower bone turnover rate. However, the relatively low values for osteoclast surface and eroded surface observed in this and most other histomorphometric studies limit the power to detect intergroup differences and/or treatment effects in these endpoints, suggesting the possibility of a Type II error. Still, the extent of osteoclast surface and eroded surface appeared to remain unchanged, but the resorption activity decreased as measured by urinary pyridinoline.

Effects on structure unit: At the basic structural unit level of bone, alendronate had no detectable effect on either mean wall thickness, which represents the amount of bone deposited in a resorption cavity, or erosion depth, although the methodology for the latter measurement remains to be fully validated. The total formation period tended to increase, but it was most likely because of an increase in the inactive period related to the low bone remodeling.

Effects on secondary mineralization: A higher degree of mineralization of bone matrix, resulting from the marked slowing of bone turnover, probably contributes to the increases in BMD observed in the alendronate-treated patients in this study.(40) As the “life span” of osteons and trabecular packets is augmented, secondary mineralization can progressively increase for a longer period of time. This hypothesis has been confirmed in minipigs by quantitative back-scattered electron microscopy(41) and by quantitative micro-radiography in baboons(40) and in humans.(42)

In summary, these data show that alendronate treatment decreases bone turnover with no adverse effects on bone structure and mineralization in glucocorticoid-treated patients. Although alendronate was given during concurrent glucocorticoid exposure, which is characterized by low bone formation, its effect on bone remodeling was similar to that previously reported in postmenopausal osteoporosis. These results confirm that with alendronate treatment, the bone formed was normal lamellar bone of good quality.


We thank Christine Fourneret for typing the manuscript.