| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Clinical Studies |
Departments of Medicine (E.S., M.R., S.J.S., D.M., K.A., V.A.), Surgery (R.E.M.), and Radiology (R.B.S.), College of Physicians and Surgeons, and Division of Biostatistics, School of Public Health (S.H.L.), Columbia University, New York, New York 10032; and the Department of Medicine, University of Heidelberg (M.J.S.), Heidelberg, Germany
Address all correspondence and requests for reprints to: Elizabeth Shane, M.D., Department of Medicine, Columbia University College of Physicians and Surgeons, 630 West 168th Street, New York, New York 10032.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
We conclude that rapid bone loss is primarily confined to the initial year after transplantation. During the first 6 months, bone loss is accompanied by alterations in markers of bone turnover consistent with biochemical uncoupling of bone formation and resorption. Greater exposure to glucocorticoids, lower serum concentrations of vitamin D metabolites and testosterone, and higher bone turnover were associated with more rapid bone loss.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
-2.0) was present at the lumbar spine
in 28% and at the femoral neck in 20% of patients (2). In addition,
35% of the patients had one or more vertebral fractures (2). Other
investigators have reported prevalence rates for vertebral fractures
after cardiac transplantation ranging from 18–50% (3, 4, 5, 6) and annual
incidence rates ranging from 5–18% (7, 8, 9). In an effort to elucidate the etiology and pathogenesis of this form of osteoporosis, a prospective longitudinal study of bone mineral density (BMD), fracture incidence, and indexes of mineral metabolism after cardiac transplantation was initiated. A report on the natural history of and risk factors for fracture during the first year after transplantation in 47 patients in this study was published recently (10). Atraumatic fractures occurred in 36% of the patients (54% of women and 30% of men), an incidence rate similar to the 35% prevalence rate of fractures in cardiac transplant recipients observed previously (2). The vast majority of the fractures involved the spine or ribs, sites composed predominantly of cancellous bone. Of those patients who fractured, 85% did so within the first 6 months. Pretransplant femoral neck BMD was significantly lower in women who fractured, whereas there was no difference in pretransplant BMD between men who fractured and those who did not. In contrast, the amount of bone loss during the first 6 months after transplantation was significantly greater in men who fractured, whereas women who did and did not fracture did not differ in this regard. Many patients fractured despite having normal pretransplant BMD. No pretransplant densitometric or biochemical measurement reliably predicted fracture after transplantation in the individual patient. The two groups also did not differ with respect to exposure to glucocorticoids or cyclosporine A.
Although several investigators have reported on changes in BMD over the initial year after cardiac transplantation (4, 7, 8, 9), there are few data on bone turnover markers. Similarly, data on patterns of bone loss and fracture at time points beyond 1 yr (11, 12) are limited. Moreover, probably because of the male preponderance of arteriosclerotic heart disease, there are few data on changes in BMD in women after cardiac transplantation. In this report, we present a more detailed analysis of alterations in BMD and biochemical indexes of mineral metabolism in men and women followed for the first 3 yr after cardiac transplantation.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
We consecutively enrolled 77 adult men and women with end-stage
cardiac disease, ranging in age from 25–70 yr, who underwent cardiac
transplantation between January 1, 1991, and December 31, 1993.
Exclusion criteria included disorders known to affect bone and mineral
metabolism (primary hyperparathyroidism, multiple myeloma,
thyrotoxicosis, serum creatinine above 2.5 mg/dL), current or prior
therapy with bisphosphonates, therapy with thyroid hormone causing
suppressed TSH, significant exposure to glucocorticoids before
transplantation (prednisone or its equivalent, 
10 mg/day for 3
months), previous organ transplant, and current therapy with estrogen
or calcitonin. Seven patients who died within 3 months of
transplantation were excluded from further analyses. The cardiac
diagnoses of the 70 remaining patients included 39 patients with
ischemic cardiomyopathy, 23 with dilated cardiomyopathy, 3 with
congenital and 3 with valvular heart disease, 1 with hypertrophic
cardiomyopathy, and 1 with atrial myxoma.
During the first year, 4 patients died from complications related to transplantation. Eight patients withdrew from the study; 2 developed intercurrent malignancy, and 6 did not wish to participate. Antiresorptive therapy (injectable calcitonin or cyclic etidronate) was initiated in 17 patients, 9 who had sustained symptomatic vertebral fractures and 8 whose lumbar spine T score had fallen below -2.5 within the first 6 months after transplantation. Fifty-seven patients completed the initial year after cardiac transplantation, 41 without and 16 with antiresorptive therapy.
During the second year, 2 patients died. Six patients withdrew from the study, 2 because they relocated and 4 because they did not wish to participate. Antiresorptive therapy was begun during the second year in 1 patient because of a fracture and in 2 because of a T score below -2.5. Thus, 50 patients completed the second year, 30 without and 20 with antiresorptive therapy.
During the third year, 3 patients died, 6 withdrew, and 1 was excluded because he developed multiple myeloma. Two patients had not completed the third year at the time of this analysis. Thirty-eight patients completed the third year, 23 without and 15 with antiresorptive therapy.
As antiresorptive agents may affect rates of bone loss and biochemical indexes of mineral metabolism, data from patients who required antiresorptive therapy at any point in the course of the 3-yr study were analyzed separately after therapy was initiated.
Study design
Each patient was evaluated with bone densitometry (BMD) of the lumbar spine, right hip, and the one third site of the nondominant forearm, radiographs of the spine, and biochemical indexes of mineral metabolism 5 ± 1 months before transplantation. Six patients had their initial bone density measurement performed 2 weeks after transplantation. After transplantation, all patients received 500 mg elemental calcium (as the carbonate salt) twice daily and a multivitamin with 400 IU vitamin D. Biochemical measurements were performed 1 month after transplantation, and BMD and biochemical measurements were repeated 3, 6, 12, 18, 24, 30, and 36 months after transplantation. A standard set of spinal radiographs was repeated 12, 24, and 36 months after transplantation and analyzed by one of us (R.B.S.) for incident fractures.
Immunosuppressive regimen
After transplantation, all patients received glucocorticoids, cyclosporine A, and azathioprine. The glucocorticoid regimen consisted of iv methylprednisolone (500 mg intraoperatively and 125 mg every 8 h during the first 24 h), oral prednisone (100 mg daily, tapering rapidly to 30 mg by 2 weeks and gradually to 10 mg by 4–6 months after transplantation). Rejection was managed by high dose iv or oral glucocorticoids, followed by rapid tapering; the precise regimen depended upon severity. Cyclosporine A was prescribed at a dose of 2 mg/kg, iv, for 24 h; the dose was adjusted by serum levels to maintain a concentration of 200–300 ng/mL during the first 6 months. Azathioprine was prescribed at a dose of 3 mg/kg/day and adjusted to maintain the white blood cell count between 4000–6000/mm3.
Daily doses of glucocorticoids (expressed in prednisone equivalents) and cyclosporine A were recorded 1, 3, 6, 12, 18, 24, 30, and 36 months after transplantation. Cumulative doses of prednisone were calculated, and serum trough cyclosporine A levels were recorded at the same time intervals. By 1 yr after transplantation, the mean prednisone dose was 9.0 ± 0.5 mg daily, and the calculated mean cumulative prednisone exposure was 9.0 ± 0.3 g. No patient was completely withdrawn from glucocorticoids during the first year after transplantation. The average dose of cyclosporine A 12 months after transplantation was 417 ± 25 mg/day, and the mean cyclosporine A level was 189 ± 12 ng/mL.
Measurements
BMD was measured by dual energy x-ray absorptiometry using a
QDR-1000 bone densitometer (Hologic, Waltham, MA). The lumbar spine
measurement represented the average of three vertebrae, usually the
second, third, and fourth. If one of these vertebrae was fractured, L1
was analyzed in its place. In our laboratory, the in vitro
reproducibility of this machine using an anthropomorphic spine phantom
is 0.51%. The short term in vivo coefficient of variation
in a group of middle-aged women with low bone mass is 0.68% for the
lumbar spine (L2–L4) and 1.36% for the proximal femur. Bone density
was expressed as grams per cm2 and as standardized T and Z
score analyses, which compare individual bone density determinations to
those, respectively, of a young and an age-matched normal population of
the same gender. According to the criteria defined by the WHO, T scores
more than 2.5 SD below the mean (-2.5) represent
osteoporosis, whereas T scores between -1.0 and -2.5 represent
osteopenia (13).
Anterior-posterior and lateral radiographs of thoracic and lumbar spine were obtained as previously described (10). Vertebral height was measured with a straight edge. Fractures were classified according to the method of Eastell et al. (14) and included wedge, biconcave, and compression deformities. An incident fracture was diagnosed if there was a 20% decrease in any vertebral height compared to the baseline radiograph.
Serum for biochemical analyses was obtained in the morning under fasting conditions whenever possible. Eight of the 70 patients (2 women and 6 men) had their initial biochemical evaluation performed after transplantation, and these data were excluded from analysis of pretransplant biochemistries. Serum calcium, phosphate, alkaline phosphatase, creatinine, and blood urea nitrogen were measured by standard autoanalyzer techniques (Technicon Instruments, Tarrytown, NY). Serum osteocalcin (bone Gla protein) (15) was measured by RIA (16). Intact PTH was measured by immunoradiometric assay (17). Serum concentrations of 25-hydroxyvitamin D (25OHD) and 1,25-dihydroxyvitamin D [1,25-(OH)2D] were measured as previously described (18). Serum total testosterone was measured by solid phase RIA (19).
A 24-h urine collection was analyzed for calcium by atomic absorption
spectrophotometry and for creatinine using standard autoanalyzer
techniques. Hydroxyproline excretion was measured by the method of
Kivirikko et al. (20). Hydroxypyridinium cross-links of type
1 collagen, pyridinoline and deoxypyridinoline (21), were measured by
high performance liquid chromatography (22). Urinary markers are
expressed as creatinine ratios throughout. The normal ranges for all
biochemical measurements are given in Table 2
.
|
All data are presented ± SEM.
Annualized rates of bone loss
To evaluate rates of bone loss, expressed as grams per cm2, measured over intervals less than 1 yr on the same temporal scale, all differences between consecutive visits were transformed to annualized rates by multiplying the 3-month interval changes by 4- and 6-month interval changes by 2.
Rates of bone loss as percent change
To express all bone loss values as an annualized percent loss from baseline, the difference between two consecutive values was divided by the earlier value and multiplied by the appropriate factor (4 for 3-month intervals and 2 for 6-month intervals) to produce annualized percentage rates.
Descriptive statistics
The temporal pattern of bone loss after cardiac transplantation was described by graphing the annualized differences between consecutive samples in both raw score units (grams per cm2) and percent change units for each of the target BMD measurement sites over the study period.
Missing BMD values
To avoid dropping patients from multivariate analyses due to a missed visit, the BMD values from missed visits that occurred between two captured data points were imputed according to the method of Buck (23). A regression equation predicting the missing time values from the earlier time value and the change from the earlier value to the later value was calculated for all subjects possessing all three data points. The earlier value and the change from the earlier value to the later value for the patient with the missing intermediate value were then entered into the predictive equation, and the predicted value was used to fill the missing value. For example, if a patient had a baseline lumbar spine BMD of 1.02 g/cm2, a missing 3 month value, and a 6 month value of 0.92 g/cm2, and a regression of all subjects with complete data for these time points yielded the equation y = 1.00 + 0.02(baseline BMD) + 0.63(6 month BMD - baseline BMD), then the 3 month imputed value for the patient in question would be y = 1.00 + 0.02(1.02) + 0.63(0.92–1.02) = 0.957. Thirty-nine of 377 (10.3%) lumbar spine and 42 of 374 (11.2%) femoral neck BMD determinations were imputed. The imputation process only slightly altered the results of a repeated measures ANOVA comparison of bone loss at three time points during the first year after transplantation (baseline to 3 months, 3–6 months, and 6–12 months): femoral neck bone loss without imputed values (F(2,149) = 3.92; P < 0.03) vs. with imputed values (F(2,186) = 2.98; P < 0.05), lumbar spine bone loss without imputed values (F(2,150) = 20.53; P < 0.001) vs. with imputed values (F(2,186) = 11.53; P < 0.001), and radius bone loss without imputed values (F(2,141) = 1.17; P < 0.32) vs. with imputed values (F(2,186) = 2.47; P < 0.09).
Inferential statistics
Two-way ANOVA with time of sample and gender, employing least squares means for an unbalanced design with different numbers of observations in each cell, were computed separately for the following effects: time (12, 24, and 36 months), gender (male and female), and time by gender interaction; antiresorptive therapy and time by therapy interaction; time (0, 3, 6, and 12 months), gender (male and female), and time by gender interaction; and antiresorptive therapy and time by therapy interaction. Overall model effects were evaluated at the P < 0.05 significance level. For those models exceeding this criterion, main effects of time, gender, therapy, and the interaction were evaluated at the P < 0.05 significance level. For those effects exceeding these criteria, post-hoc comparisons of means were performed to test the apriori hypotheses that consecutive time points differed from each other, that male and female subjects differed at each time point, and that treated and untreated patients differed at each time point. All post-hoc comparisons were evaluated at the P < 0.005 level of significance.
These data were also analyzed by repeated measures ANOVA, which limits the dataset to the 38 individuals (23 untreated and 15 treated) who completed months 3–36. The pattern of statistical significance was the same whether parallel group analysis or repeated measures ANOVA was used. Therefore, the parallel group analysis is reported herein, as this method permits analysis of all subjects followed after transplantation.
To assess the influence of postoperative immunosuppressive therapy and
biochemical indexes of mineral metabolism on the course of bone loss,
daily prednisone dose (milligrams), cumulative prednisone dose (grams),
daily cyclosporine A dose (milligrams), and trough serum cyclosporine A
levels (nanograms per mL) and the biochemical determinations (Table 2
)
were each entered into the above ANOVA models as continuous covariates,
and F tests were computed. All analyses employed Proc General Linear
Models from the SAS Institute (Cary, NC).
The study was conducted at the Irving Center for Clinical Research and the Cardiac Transplant Unit of Columbia-Presbyterian Medical Center (New York, NY). It was approved by the Institutional Review Board of Columbia-Presbyterian Medical Center. Written informed consent for participation was obtained from all subjects.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
All patients were New York Heart Association class III or IV at the time of the initial evaluation. The group included 52 men, aged 54 ± 2 yr; 15 postmenopausal women, aged 56 ± 2 yr; and 3 premenopausal women, aged 34 ± 4 yr. No postmenopausal woman was taking hormone replacement therapy. Two premenopausal women were amenorrheic, with normal serum gonadotropin and PRL concentrations; both resumed regular menses 2 months after transplantation. The majority of the patients were Caucasian; 4 women and 2 men were African-American. Because of the relatively small number of premenopausal women and African-Americans, analyses were not stratified by menopausal status or race. The majority of patients (86%) received high doses of loop diuretics as part of the pretransplant medical regimen. No patient was taking supplemental calcium or multivitamins before transplantation.
BMD before transplantation
As expected, mean pretransplant BMD, expressed as grams per cm2, was significantly lower in women than in men at the lumbar spine, femoral neck, and forearm (data not shown). The mean femoral neck T score was low in both men and women (-1.344 ± 0.20 vs. -2.01 ± 0.35, respectively) and did not differ significantly. The mean lumbar spine T score was low in women and normal in men (-1.676 ± 0.35 vs. -0.852 ± 0.21, respectively; P < 0.05).
Bone loss after transplantation in untreated patients
Rates of bone loss during the 3 yr after transplantation in
patients who did not receive antiresorptive therapy are presented in
Fig. 1. During the first year, the mean rate of bone
loss was 7.3 ± 0.9% at the lumbar spine and 10.5 ± 1.1%
at the femoral neck. Rates of bone loss slowed significantly at both
sites (0.9 ± 0.9% and 0.1 ± 1.0%, respectively;
P < 0.001 compared to year 1) during the second year.
During the third year, lumbar spine BMD increased by an average of
2.4 ± 0.8%/yr (P < 0.02), whereas no change was
observed at the femoral neck. Men and women did not differ with respect
to the amount or pattern of bone loss (data not shown).
|
As the major declines in BMD were observed at the lumbar spine and
femoral neck during the first year, we examined the pattern of bone
loss within that period in greater detail (Fig. 2).
Rates of bone loss were annualized as described in Materials and
Methods. At the lumbar spine (Fig. 2A), the group lost bone at an
annualized rate of 19.5 ± 2.6% during the first 3 months,
significantly greater than the annualized rate of bone loss documented
during the second 3 months (8.9 ± 2.9%; P <
0.001). On the average, there was no further bone loss at the lumbar
spine during the second half of the first year.
|
) differed from that of
the lumbar spine. The rate of bone loss tended to be greater during the
first compared to the second 3-month period (18.1 ± 3.0%
vs. 11.5 ± 3.2%; P < 0.09). In
contrast to the lumbar spine, femoral neck bone loss continued during
the second half of the first year at an annualized rate of 8.2 ±
1.3%.
We observed previously that the risk of fracture after transplantation
was greater in women (10) and, therefore, had hypothesized that women
might differ from men with respect to the rate or pattern of bone loss.
However, stratification of rates of bone loss by gender during the
first year after transplantation (Fig. 3) revealed the
amount and pattern of bone loss to be similar at both sites in men and
women. These data are expressed as grams per cm2 to avoid
the effect of the lower pretransplant BMD to magnify rates of bone loss
in women when expressed as percent change.
|
Immunosuppressive therapy is likely to be important in the
pathogenesis of bone loss and fractures after cardiac transplantation.
Therefore, we examined rates of bone loss during the first year after
transplantation with respect to daily and cumulative prednisone doses,
daily cyclosporine A doses, and trough serum cyclosporine A levels. In
general, greater glucocorticoid exposure was associated with higher
rates of bone loss. During the first 3 months after transplantation,
higher cumulative prednisone doses were associated with more femoral
neck bone loss (r = 0.421; P < 0.003). During the
second half of the first year, rates of femoral neck and lumbar spine
bone loss were related to the daily prednisone dose (milligrams per
day) at 3, 6, and 12 months (Table 1), rather than to
the cumulative prednisone exposure. In contrast, higher 3 month serum
levels of cyclosporine A were associated with less femoral neck bone
loss between 3–6 months (r = -0.369; P < 0.02).
No relationship was detected between daily doses or serum levels of
cyclosporine A and rates of bone loss during any other time period.
|
Serial biochemistries are shown in Table 2. Mean
values for all parameters were normal before transplantation.
Consistent with the well known nephrotoxic effects of cyclosporine A
(24), there was a significant decline in renal function, which
stabilized after 6 months. The mean 25OHD concentration rose
significantly, perhaps due to the inclusion of vitamin D in the
posttransplant regimen. However, the mean 1,25-(OH)2D
concentration fell by 3 months and remained significantly below
baseline throughout the first year. The decline in
1,25-(OH)2D concentrations correlated with deteriorating
renal function (data not shown).
In men, a transient, but significant, decline in mean serum testosterone was apparent by the first month. Three months after transplantation, serum testosterone levels remained slightly, but not significantly, below baseline concentrations. At all later time points, mean values were similar to baseline concentrations.
Biochemical markers of bone turnover also changed after transplantation
(Table 2 and Fig. 4). The serum osteocalcin
concentration fell by 80% at 1 month after transplantation, remained
significantly suppressed at 3 months, and correlated negatively with
daily prednisone dose; the strongest relationship was observed at 3
months (r = -0.419; P < 0.003). Greater
suppression of serum osteocalcin during this period was associated with
higher rates of bone loss at both the femoral neck (r = 0.416;
P < 0.02) and the lumbar spine (r = 0.402;
P < 0.02). Osteocalcin returned to baseline levels by
6 months after transplantation, concomitant with tapering of prednisone
treatment. The decline in osteocalcin was mirrored by significant
increases at 1 and 3 months in all bone resorption markers (urinary
calcium, hydroxyproline, pyridinoline, and deoxypyridinoline); all
returned to baseline by 6 months.
|
To elucidate the pathogenesis of the bone loss, we examined relationships between rates of bone loss during the first year and serial biochemical determinations. Concentrations of both vitamin D metabolites were related to rates of femoral neck bone loss during the first year after transplantation. The influence of vitamin D status on bone loss differed by metabolite and by the time period during which bone loss occurred. Lower rates of femoral neck bone loss during the first 3 months after transplantation were associated with higher body stores of vitamin D, as indicated by serum 25OHD concentrations 3 months after transplantation (r = -0.418; P < 0.005). Lower rates of femoral neck bone loss between 6–12 months after transplantation were associated with higher concentrations of 1,25-(OH)2D at 12 months after transplantation (r = -0.335; P < 0.05). No relationship was detected between vitamin D metabolites and lumbar spine bone loss.
In men, significant relationships were detected between serum testosterone and rates of femoral neck bone loss during both the first 3 months and the second 6 months of the first year. Rates of decline of BMD during the first 3 months were lower in men with higher pretransplant serum testosterone concentrations (r = -0.396; P < 0.02). Similarly, rates of decline of femoral neck BMD between 6–12 months after transplantation were also lower in men with higher 6 month (r = -0.621; P < 0.01) and 12 month (r = -0.485; P < 0.07) serum testosterone concentrations. A similar trend was observed at the lumbar spine site.
Relationships between markers of bone turnover and rates of bone loss
between 3–6 months after transplantation are shown in Table 3. In general, increased bone turnover was associated
with higher rates of bone loss. However, baseline bone turnover markers
did not predict rates of bone loss, and only the 6 month osteocalcin
level correlated with the decline in BMD between 3–6 months. Urinary
calcium and hydroxyproline excretion demonstrated more consistent
relationships with the rates of bone loss than did serum
osteocalcin and urinary deoxypyridinoline. Other serial measurements of
bone turnover markers and calculated changes in bone resorption markers
did not correlate with rates of bone loss during the first 3 months or
the second half of the year.
|
Seventeen patients sustained atraumatic fractures during the first year after transplantation. A detailed analysis of the natural history of and risk factors for fracture in these patients has been published recently (11). During the second year after transplantation, three patients sustained atraumatic fractures. One patient, who had previously sustained multiple vertebral fractures during the first year, fractured twice at 30 and 36 months after transplantation.
Antiresorptive therapy after transplantation
Those patients who required antiresorptive therapy (for fractures
or T score <-2.5) at any time during the first posttransplant year
had significantly lower pretransplant lumbar spine and femoral neck BMD
than untreated patients (Table 4). Their age and the
patterns and rates of bone loss at the femoral neck (Fig. 5A) and the lumbar spine (data not shown) during the
first year did not differ significantly between treated and untreated
patients. During the second and third years, rates of femoral neck bone
loss were the same regardless of antiresorptive therapy (data not
shown). At the lumbar spine (Fig. 5B), however, patients receiving
antiresorptive therapy gained bone mass during year 2, whereas
untreated patients did not (0.027 ± 0.01 vs.
-0.012 ± 0.01 g/cm2; P = 0.014).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
By providing data on fracture incidence during the second and third years after cardiac transplantation, this study extends earlier observations by us (10) and others (8, 12) on the occurrence of fracture during the first year and confirms that fractures are concentrated during the first year. However, to some extent, the lower incidence of fracture during the second and third years may be related to institution of antiresorptive therapy in those patients who developed fractures or extremely low bone mass during the first year and may not entirely reflect the natural history of fracture after cardiac transplantation.
The observation that lumbar spine bone loss is greatest during the first 6 months after cardiac transplantation is in general agreement with the findings of other studies of cardiac transplant recipients (4, 8, 10). Muchmore et al. (4) found that vertebral BMD stabilized after 3 or 6 months in all age groups studied. Sambrook and colleagues (8) also observed that lumbar spine bone mass fell rapidly in 25 patients followed with dual energy x-ray absortiometry for 1 yr after cardiac transplantation, and that all the bone loss occurred during the first 6 months. The hip and forearm were not measured in either study. A later study by these investigators (12) demonstrated recovery of lumbar spine BMD during the third year, as observed in this study. The extent of the bone loss reported here is also in agreement with other reports (8, 9, 11), as is the observation that pretransplant vertebral BMD is lower than that in age- and sex-matched control subjects in a significant number of patients (4, 9).
In addition to establishing the incidence and natural history of bone loss and fracture after cardiac transplantation, a major goal of this study was to elucidate the pathogenesis of the decline in BMD. Immunosuppression agents are obvious etiological candidates. The associations between glucocorticoids, bone loss, and vertebral fractures are well established (25). Most patients taking high dose, long term glucocorticoids will experience bone loss, predominantly involving skeletal areas composed of cancellous bone. The most rapid bone loss occurs during the first 12–18 months of therapy, after which the rate of decline slows. Glucocorticoids directly suppress osteoblast function and impair collagen synthesis and new bone formation (26), an effect reflected biochemically by low serum concentrations of osteocalcin (27, 28), an important clinical marker of bone formation and turnover (29). There is also evidence of increased bone resorption (30) during the early phase of glucocorticoid therapy, possibly related to inhibition of gastrointestinal calcium absorption and stimulation of renal calcium excretion or suppression of gonadal steroidogenesis (25). However, a recent longitudinal study did not demonstrate an increase in biochemical markers of bone resorption (hydroxyproline) after initiation of glucocorticoid therapy (27).
Cyclosporine A has also been implicated in the development of transplantation osteoporosis (1). The work of Epstein and colleagues has shown that rats exposed to cyclosporine A demonstrate rapid and severe bone loss, characterized histologically by increases in both bone resorption and formation (31, 32, 33, 34). This bone loss can be prevented by agents that inhibit bone resorption (35, 36). Moreover, rat studies indicate that cyclosporine A appears to counteract the suppressive effects of glucocorticoids on the osteoblast (37). The effects of cyclosporine alone have been difficult to characterize in human subjects, primarily because cyclosporine A is generally administered in combination with glucocorticoids. However, there is evidence that cyclosporine A does cause increased bone turnover in humans (38, 39). Thus, the question of whether cyclosporine A contributed to the bone loss and alterations in bone turnover markers observed in our patients is of interest.
In this study, the temporal association of bone loss with higher doses of glucocorticoid and cyclosporine A supports a causative role for immunosuppressive therapy. Several additional observations indicate that glucocorticoid therapy is important in the pathogenesis of the bone loss. These include the positive correlations between the 3-month cumulative prednisone dose and early femoral neck bone loss, between daily prednisone dose and rates of bone loss during the second half of the first year, the negative relationships between daily prednisone dose and osteocalcin levels, and the observation that the degree of suppression of osteocalcin levels was related to early bone loss. However, these observations contrast with those of other investigators (8, 9, 11, 40), who found no significant relationship between glucocorticoid exposure and BMD. Moreover, the limited strength of the regressions described above suggests that other factors are also important in the pathogenesis of bone loss in this study. However, cyclosporine A, another potential cause, was not associated with higher rates of bone loss at any point during the first year or with serum osteocalcin levels or urinary markers of bone resorption 1 and 3 months after transplantation. Moreover, the observation that higher 3 month cyclosporine A levels were associated with lower rates of femoral neck bone loss during the subsequent 3 months could be construed as evidence that cyclosporine A may counteract the deleterious effects of glucocorticoids on the skeleton, as has been observed in the rat (37). Thus, the results of this study suggest that glucocorticoids may be of greater importance than cyclosporine A in the pathogenesis of early bone loss after cardiac transplantation.
The recovery of lumbar spine BMD in year 3 may have been related to the gradual reduction in glucocorticoid dosage accomplished during this period. Although most patients remained on triple immunosuppressive therapy, by 3 yr after transplantation the mean daily prednisone dose was 7.6 ± 0.7 mg, and 65% of the patients were receiving less than 7.5 mg. As daily doses below 7.5 mg are less likely to be associated with bone loss (25), the gradual reduction in prednisone dose may be related to the stabilization or gain in BMD. It is also possible that lower prednisone doses might permit cyclosporine to stimulate bone formation, as has been observed in the rat model (31, 32, 33, 37), and thus to play a role in the directional shift observed in this study.
The marked suppression of serum osteocalcin levels during the initial 3 months, accompanied by increases in all markers of bone resorption, indicates that the early posttransplant period is characterized by a biochemical profile consistent with uncoupling of bone formation and resorption. Under such circumstances, severe bone loss might be expected to occur. This pattern of change is similar to that recently reported by Sambrook et al. in a smaller number of patients (41). The pathogenesis of these changes in markers of bone turnover is probably multifactorial. Certainly, glucocorticoids alone could be directly responsible for both the decline in osteocalcin and the increase in urinary calcium. However, a longitudinal study of bone turnover markers after initiation of high dose glucocorticoid therapy found a similar marked decrease in osteocalcin, but no increase in hydroxyproline excretion (27). Thus, the increase in resorption markers may be related to other factors. PTH, another potential cause of increased bone resorption, was not associated with markers of bone resorption during the first year. Although no independent relationship was detected between cyclosporine A exposure and bone resorption markers, it remains possible that cyclosporine A could have directly stimulated osteoclastic bone resorption during this period.
Most of the literature indicates that excretion of collagen cross-links is a more specific marker of bone resorption than hydroxyproline excretion (42). Thus, the observation that hydroxyproline excretion correlated with rates of bone loss to a greater extent than did deoxypyridinoline (or pyridinoline) was of interest. In this regard, a recent study of markers of bone turnover in hyperandrogenic women receiving GnRH agonist therapy also showed that hydroxyproline, but not pyridinium cross-link, excretion fell after institution of hormone replacement therapy (43). Hydroxyproline is a nonspecific index of total body collagen turnover that may reflect both the synthesis and degradation of all fibrillar collagens, whereas pyridinium cross-link excretion is relatively specific for bone collagen (42). As glucocorticoids may affect both the synthesis and degradation of all types of collagen, the rise in hydroxyproline excretion after transplantation might thus reflect an increase in total body collagen turnover as well, in addition to increased bone resorption. The increased mobility of cardiac transplant recipients after successful transplantation may also lead to an increase in total body collagen turnover, which may be better reflected by hydroxyproline than pyridinium cross-link excretion.
Biochemical influences of bone loss after cardiac transplantation also included vitamin D status. The results of this investigation suggest that more robust body stores of vitamin D are associated with lower rates of bone loss after cardiac transplantation. Similarly, higher 1,25-(OH)2D levels during the second half of the year were associated with lower rates of bone loss. No other biochemical influences on bone loss were observed during the first year.
The interpretation of these results is limited by the relatively large number of patients who died or withdrew from the study, particularly those who required antiresorptive therapy for fractures and/or low bone mass. These patients had lower pretransplant BMD, but the pattern and rates of bone loss were generally similar to those of untreated patients. The only exception was observed during the second year, when treated patients gained and untreated patients lost lumbar spine BMD. Had it been possible to observe these more severely affected individuals without therapy, rates of bone loss might have been even more striking. In addition, the statistical power to detect significant relationships between biochemical and hormonal indexes might have been greater had it not been necessary to exclude the biochemistries of treated patients. Patients who completed the second and third years without therapeutic intervention may thus not be representative of the group of cardiac transplant recipients in its entirety. This study, designed to investigate the natural history and biochemical correlates of bone loss after cardiac transplantation and to lay the groundwork for rational studies to prevent this complication, was not intended to serve as a therapeutic trial. Antiresorptive therapy was not randomized or blinded and was begun of necessity at varying intervals after transplantation when fractures occurred or osteoporosis (by densitometric criteria) developed. Thus, the efficacy of antiresorptive therapy should not be judged on the basis of these results. In this regard, we have recently demonstrated that antiresorptive therapy begun shortly after transplantation prevents posttransplantation bone loss (44).
In summary, this study documents that rapid and severe bone loss ensues after cardiac transplantation and continues for 6–12 months. Men and women lose bone at similar rates. Despite continuing therapy with lower doses of cyclosporine A and glucocorticoids, bone loss essentially ceases after the end of the first year. However, perhaps because of their lower pretransplantation BMD, similar rates of bone loss are more devastating in women, as earlier analyses indicated that women were more likely to fracture (10). These results and those of Sambrook and colleagues (41) indicate that the bone loss is accompanied by biochemical evidence of uncoupled bone formation and resorption, particularly during the first 6 months after transplantation. Abnormal vitamin D status and, in men, hypogonadism also are associated with higher rates of bone loss. These results indicate that efforts to prevent osteoporosis should begin immediately after transplantation and continue for at least the first posttransplant year.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Current address: Veterans Administration Medical Center, Medical
Service III, One Veterans Plaza, San Juan, Puerto Rico
00927-5800.
Received August 13, 1996.
Revised December 11, 1996.
Revised January 21, 1997.
Accepted January 27, 1997.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  H. Zhou, W. Mak, Y. Zheng, C. R. Dunstan, and M. J. Seibel Osteoblasts Directly Control Lineage Commitment of Mesenchymal Progenitor Cells through Wnt Signaling J. Biol. Chem., January 25, 2008; 283(4): 1936 - 1945. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. M. Maalouf and E. Shane Osteoporosis after Solid Organ Transplantation J. Clin. Endocrinol. Metab., April 1, 2005; 90(4): 2456 - 2465. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Shane, V. Addesso, P. B. Namerow, D. J. McMahon, S.-H. Lo, R. B. Staron, M. Zucker, S. Pardi, S. Maybaum, and D. Mancini Alendronate versus Calcitriol for the Prevention of Bone Loss after Cardiac Transplantation N. Engl. J. Med., February 19, 2004; 350(8): 767 - 776. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Vincenti Immunosuppression Minimization: Current and Future Trends in Transplant Immunosuppression J. Am. Soc. Nephrol., July 1, 2003; 14(7): 1940 - 1948. [Full Text] [PDF]
|
|
|
|
 |
|
 D. M. Biskobing COPD and Osteoporosis Chest, February 1, 2002; 121(2): 609 - 620. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Y. Lee, S. W. Cho, E. S. Oh, K. W. Oh, J. M. Lee, K. H. Yoon, M. I. Kang, B. Y. Cha, K. W. Lee, H. Y. Son, et al. The Effect of Bone Marrow Transplantation on the Osteoblastic Differentiation of Human Bone Marrow Stromal Cells J. Clin. Endocrinol. Metab., January 1, 2002; 87(1): 329 - 335. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Canalis and A. Giustina Glucocorticoid-Induced Osteoporosis: Summary of a Workshop J. Clin. Endocrinol. Metab., December 1, 2001; 86(12): 5681 - 5685. [Full Text] [PDF]
|
|
|
|
 |
|
 I. R. Reid Glucocorticoid Osteoporosis: Reid IR Glucocorticoid osteoporosis J Intensive Care Med 1999, 14231-242 J Intensive Care Med, September 1, 1999; 14(5): 231 - 242. [PDF]
|
|
|
|
 |
|
 R. B. Staron, R. Greenspan, T. T. Miller, J. P. Bilezikian, E. Shane, and N. Haramati Computerized Bone Densitometric Analysis: Operator-dependent Errors Radiology, May 1, 1999; 211(2): 467 - 470. [Abstract] [Full Text]
|
|
|
|
 |
|
 J. Cremer, M. Struber, I. Wagenbreth, J. Nischelsky, S. Demertzis, T. Graeter, C. Abraham, and A. Haverich Progression of steroid-associated osteoporosis after heart transplantation Ann. Thorac. Surg., January 1, 1999; 67(1): 130 - 133. [Abstract] [Full Text] [PDF]
|
|
| |||||||
