INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Transplantation-related osteoporosis is a significant and growing clinical problem (1). Lung transplantation shares with other transplant fields mandatory, life-long immunosuppressive therapy, which is a major risk factor for bone loss. In addition, by comparison, lung transplant recipients often suffer more serious comorbidities that may contribute further to accelerated bone mineral density (BMD) loss (2). Declining BMD after lung transplant is compounded by significant pretransplant bone deficits due to longstanding respiratory illnesses and their treatments (3, 4). In fact, patients awaiting lung transplantation have more severe BMD deficits than those awaiting transplantation of other organs. Patients with cystic fibrosis (CF) who may suffer from delayed puberty, pancreatic insufficiency, and chronic malnutrition and infectious disease, are the most severely affected group (5). As improvements in the management of respiratory illness have doubled the median survival of patients with CF over the past few decades, osteoporosis, and its sequelae, have been reported with increasing frequency across the globe (11, 12). For these reasons, osteoporosis in patients with CF after lung transplant is almost universally present and may complicate the postoperative course.

Early therapeutic trials to improve posttransplant osteoporosis have met with mixed results. Recent efforts in the fields of heart and liver transplantation have proven more successful (14). Our goal in this randomized, nonblinded, clinical therapeutic trial was to determine if pamidronate, a second-generation bisphosphonate, when combined with calcium and vitamin D, could improve BMD in patients with CF following lung transplantation compared with vitamin D and calcium alone.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patients

Thirty-four ambulatory, white adults (18-38 yr old) with CF (18 men, 16 women) were recruited 1-12 mo after lung transplantation at the University of North Carolina. The interval of 1-12 mo posttransplant was chosen to optimize the number of available patients for the study realizing that the rate of bone loss may vary during this period. Exclusionary criteria for this trial included primary graft failure or other postoperative morbidities that precluded long-term survival, renal insufficiency (serum creatinine > 3.0 mg/dl), and pregnancy. All subjects signed a consent form approved by the Committee on Human Research. The diagnosis of CF had been confirmed prior to transplant by an elevated sweat chloride test, pancreatic insufficiency, a compatible clinical history, and by pathological examination of explanted lung tissue.

Clinical Protocol

Patients were stratified by sex and severity of osteopenia using a spine T-score of 3.0 (based on previous data) as the cut-point between groups and, then, randomized in a "blocks of four" design to receive calcium carbonate (1 g/d) and ergocalciferol (800 IU/d) by mouth with or without pamidronate 30 mg intravenously every 3 mo for 2 yr. The study was unblinded, but the radiologist who interpreted the DEXA scans was unaware of the treatments that each patient received. The pamidronate infusions were administered in our General Clinical Research Center (GCRC) (94% of the infusions) or, if the patient lived out of town and was unable to return, by a home health agency (6% of the infusions). Twenty-four hours after the first infusion, the patient was assessed for cellulitis, thrombophlebitis, bone pain, and fever and 24 h later, a complete blood count and serum calcium, phosphorous, and magnesium levels were determined. Compliance with medications was determined by patient interview.

The primary end point was spine BMD as measured by DEXA and secondary end points were femur BMD, serum and urine bone markers, vertebral and long bone fractures, and kyphosis angles. Data were collected from the medical record and shadow files for variables that may potentially confound the results of the study. These included corticosteroid use (prednisone and methylprednisolone), serum immunosuppressant (cyclosporin A and tacrolimus) and creatinine levels, hospitalizations, and beginning and end of treatment body mass index (BMI) and lung function (FEV₁). The daily prednisone doses and levels of cyclosporin, tacrolimus, and creatinine from Day 1 and 15 of each study month were averaged for each patient.

Immunosuppression and Other Medications

Our immunosuppressant protocol, which has been described elsewhere (3), includes cyclosporin A and prednisone in tapering doses, and azathioprine. Methylprednisolone and prednisone were used to manage acute allograft rejection and tacrolimus and mycophenolate were used to help manage chronic graft rejection (BOS).

Bone Densitometry

Spine BMD was the primary end point. Spine (L1-L4) and femur measurements, expressed as grams of bone mineral (hydroxyapatite) per square centimeter of bone, were made on all patients at baseline and every 6 mo for 2 yr using dual-energy X-ray absorptiometry (Hologic QDR 1000/W; Waltham, MA) (17). Quality control was maintained by scanning an anthropomorphic spine phantom daily. The coefficient of variation for our Hologic 1000/W is 0.3% and the reference limits for variation are ± 1.5%. The results were expressed as T-scores (Z-scores were similar), indicating the number of standard deviations below expected peak bone mass for normal young adults.

Radiographic Endpoints

Baseline and end-of-study lateral chest radiographs were analyzed at the end of the study for thoracic spine curvature (i.e., kyphosis angle using a modification of the method of Cobb [18]), and vertebral fractures. High kilovoltage chest rays afford similar sensitivity to lateral thoracic and lumbar spine films for the detection of vertebral fractures. The vertebral fractures were determined by measuring anterior and posterior vertebral body height and expressing the difference divided by the posterior height as a percentage (19). The posteroanterior chest radiograph was examined for rib fractures (i.e., focal deformities in the rib contour with varying degrees of reparative bone formation). Clinical fractures (required radiograph confirmation and casting or splinting) were recorded.

Serum and Urine Biochemical Measurements

Fasting serum and urine were collected at baseline and 3, 12, and 24 mo later on all patients and 2 and 14 d following the first pamidronate infusion in patients to confirm that pamidronate was capable of suppressing N-telopeptide levels for early feedback on drug bioactivity. Baseline serum was measured immediately for calcium and creatinine and serum were stored at 80° C for parathyroid hormone (PTH), 25-hydroxyvitamin D (25-OHD) and 1,25-dihydroxyvitamin D [1,25-(OH)₂D] determinations. Urine was measured for creatinine with an automatic analyzer (Hitachi 911; Boehringer Mannheim, Indianapolis, IL) at University of North Carolina Hospitals and, then, stored at 80° C for N-telopeptides of type I collagen (NTx) and deoxypyridinolines (Dpd).

Markers of bone formation. Serum was measured in duplicate for osteocalcin with a two site radioimmunoassay for human osteocalcin that measures both intact and N-terminal and mid-fragment osteocalcin (CIS-US, Bedford, MA) (20). Normal levels for males and females (ages 20-40 yr) are 10.7-37.0 and 7.7-39.4 ng/ml, respectively. Serum was also measured for bone-specific alkaline phosphatase using an enzyme-linked immunoassay that has a low (16%) cross-reactivity with the liver isoenzyme (Metra Biosystems, Mountain View, CA) (21). Normal levels for males over the age of 25 and premenopausal females are 15.0-41.3 U/L and 11.6-30.6 U/L, respectively.

Markers of bone resorption. Urine cross-linked N-telopeptides of type I collagen (Ostex International) and free deoxypyridinoline ("Pyrilinks-D," Metra Biosystems, Mountain View, CA) were measured in duplicate using immunoassays (22, 23). The normal ranges for cross-linked N-telopeptides and free deoxypyridinoline for controls are 3-65 nmol BCE/mmol creatinine (ages 25-49 yr) and 0-7.5 nmol/mmol creatinine, respectively.

Calcium, creatinine, vitamin D, and PTH levels. Serum calcium and creatinine levels were measured by an automatic analyzer with normal ranges of 8.4-10.8 mg/dl and 0.7-1.1 mg/dl, respectively. Serum PTH (CIS-US), 25-OHD, and 1,25-(OH)₂D (both INCSTAR, Stillwater, MN) were measured as previously described (24) with normal values of 12-60 pg/ml, 14-52 ng/ml, and 15-45 pg/ml, respectively.

Statistical Analysis

Comparisons of baseline variables between the control and pamidronate groups were tested by t-tests. An intention-to-treat principle was used in analyses of the treatment end points. Differences in the primary end point, spine BMD, and one of the secondary end points, femur BMD, were compared between controls and the pamidronate arm in two ways: (1) rates of BMD change per year were calculated by linear regression analysis for each subject and the averages of each treatment arm were compared with two-sample t-tests (25) and (2) using repeated measures analysis of variance for the effects of time, treatment, and time × treatment interaction. Because similar results were obtained using both analyses, we chose to include the p values corresponding to the linear regression method. In separate analyses (two sample t-tests), potential confounding variables were compared to exclude differences in treatment arms (Table 2). Longitudinal changes in the biochemical markers of bone metabolism and calcium homeostasis were compared between treatment arms with separate repeated measures of analyses of variance for each end point. Fracture rates were calculated as the number of patients with new fractures divided by the number of patient-years in each study arm. Poisson regression was then used to calculate the p values to compare the fracture rates between the two treatment arms. All statistical analyses were performed with SAS (v6.12; SAS Institute, Cary, NC) (26). A p < 0.05 in a two-tailed test was considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patient Characteristics

Forty-four adult patients with CF were transplanted during the course of this study, but seven died immediately postoperatively and, therefore, were not eligible for this trial. The remaining 37 patients consented to participate. Three patients who died during the course of the study were excluded from the final analysis of baseline characteristics, but their presence, when included in separate analyses, did not significantly affect the results. All three died (one each from sepsis, acute respiratory distress syndrome, and obliterative bronchiolitis) before the first primary end point measurement. Thirty-four patients completed the study. Patient baseline characteristics are shown in Table 1. Thirteen patients in the pamidronate arm and 12 in the control arm were osteoporotic at one or more sites at baseline using the WHO criteria of a T-score < 2.5 (27). All other patients were osteopenic at one or more sites.

Medical Management and Clinical Outcomes

Seven patients in each arm were switched from cyclosporin A to tacrolimus and from azathioprine to mycophenolate for obliterative bronchiolitis during the course of this study. Immunosuppressant doses and levels, pulmonary and renal function, nutritional status as measured by BMI, and hospitalization rate did not differ significantly between the treatment groups (Table 2)

Treatment Safety

Over the course of the study, more than 100 pamidronate infusions were administered at the University of North Carolina without thrombophlebitis, cellulitis, bone pain, or fever. No one dropped out of this study due to adverse drug effects. Forty-eight hour postinfusion serum calcium, phosphorous, magnesium, and creatinine, and white blood cell counts were not significantly changed when compared with preinfusion values. No episodes of hypocalcemia (serum calcium < 7.8 mg/ dl), but three episodes of mild hypervitaminosis D (serum 25-OHD > 55 ng/ml, all of which resolved spontaneously) occurred during the study.

Efficacy in Improving Bone Mineral Density: The Primary End Point

The pamidronate-treated patients (n = 16) experienced significant increases in spine (p = 0.015) and femur BMD (p = 0.01) compared with the group treated with vitamin D and calcium alone (n = 18) (Figure 1). By the end of the study, 14 and 12 (of 16) patients had increased BMD at the spine and femur, respectively, in the pamidronate group compared with 10 and 8 (of 18) patients, respectively, in the control group. The control group did not demonstrate significant changes in spine or femur BMD over the study period. Within and across treatment groups, patient age, sex, and entry time after transplant did not influence the BMD responses to therapy.

View larger version (19K): [in this window] [in a new window]   Figure 1.   Mean ± SE change in BMD (expressed as a percent of baseline) over time in subjects on pamidronate plus calcium and vitamin D (dashed line) compared with calcium and vitamin D alone (solid line) for spine (top, p = 0.015) and femur (bottom, p = 0.01). The baseline time point represents the initiation of therapy.

Fractures and Kyphosis Angles

Eleven patients (four in the pamidronate arm and seven in the control arm) had 16 vertebral fractures and two patients (both in the control arm) had three rib fractures by radiograph review at the beginning of the study. Six control subjects and three pamidronate-treated patients had new long bone fractures and one control subject and three pamidronate-treated patients had new vertebral fractures during the study (p >> 0.1 between arms). All but two patients who experienced fractures during the course of the study had BMD values that were higher at the end of the study. The composite fracture rate for the 34 study patients was 13 per 68 patient-years. Mean ± SD kyphosis angles, although elevated in both treatment arms at baseline (Table 1), did not change across the study period (pamidronate: 0.0 ± 5.9° versus control: 1.4 ± 4.5°, p >> 0.1).

Bone Biomarkers

As expected, NTx responded to pamidronate infusions and remained suppressed while pamidronate therapy continued (Figure 2). Fourteen, 90, and 365 days after the first infusion, mean ± SD cross-linked N-telopeptides levels fell 57.1 ± 23.7%, 48.8 ± 38.2%, and 53.7 ± 39.0%, respectively, compared with baseline whereas cross-linked N-telopeptides levels in control subjects were unchanged (p < 0.001 comparing baseline, 90, and 365 day data to pamidronate). Mean ± SD deoxypyridinoline levels followed a trend similar to the cross-linked N-telopeptides (pamidronate versus control: baseline: 15.0 ± 14.5 versus 10.5 ± 5.6; 90 d: 8.1 ± 6.3 versus 8.6 ± 5.2, 1 yr; 6.0 ± 3.6 versus 11.2 ± 7.2 nmol/mmol creatinine, p < 0.001). Osteocalcin levels were very low immediately after transplantation and rose dramatically in both treatment arms (p < 0.001 for effect of time; p > 0.1 for effect of treatment) (Figure 2). Mean serum bone-specific alkaline phosphatase levels were in the normal range and did not change significantly based on treatment or time (data not shown).

View larger version (33K): [in this window] [in a new window]   Figure 2.   (Top) Mean ± SE levels of N-telopeptides of type I collagen (NTx) over time on pamidronate (hatched bars) and control (open bars) demonstrating significant declines in NTx on pamidronate compared with control (p < 0.001). Control measurements were not made at 0.5 mo. (Bottom) Mean ± SE levels of osteocalcin levels over time in pamidronate (hatched bars) and control (open bars) subjects showing a significant time effect (p < 0.001). No treatment effect was present. The baseline time point in both graphs represents the initiation of therapy.

Serum Calcium, Vitamin D, and PTH Levels

Baseline serum calcium, 25-OHD, 1,25-(OH)₂D, and PTH levels are shown in Table 1. All variables increased significantly over time irrespective of treatment group (p > 0.1 for effect of treatment and p < 0.001 for the effect of time). Mean ± SD serum 25-OHD levels rose to 40.1 ± 9.1 and 37.9 ± 9.3, respectively, by the end of the study in the pamidronate and control groups. Mean ± SD PTH levels rose from baseline (pamidronate: 40.7 ± 19.7 pg/L; control: 44.2 ± 20.4 pg/L) to the end of the study (59.5 ± 26.2 pg/L; control: 68.0 ± 36.7 pg/L) (p < 0.01 for effect of time, p > 0.1 for effect of treatment). For the group as a whole, mean ± SD serum creatinine levels, as a result of cyclosporin treatment, increased over the study (0.9 ± 0.3 to 1.8 ± 0.4 mg/dl, p < 0.001 for effect of time, p > 0.1 for effect of treatment).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The results of this study demonstrate that pamidronate is safe and effective in improving BMD following lung transplant in CF patients, a scenario that represents one of the most severe forms of posttransplant osteoporosis. The benefit of pamidronate was evident despite the high immunosuppressant levels required following transplantation of the lung as compared with that of other solid organs. This study also is the first to demonstrate that patients with CF with osteoporosis can be successfully treated and safety monitoring confirmed the lack of serious adverse effects in the use of pamidronate in the posttransplant setting. It is possible that corticosteroid usage "protected" our patients from the bone pain that has been described by Haworth and coworkers in nontransplanted patients with CF (28). These results strongly suggest that posttransplant osteoporosis is a treatable medical complication and patients with symptomatic osteoporosis, which currently is a relative contraindication to lung transplantation by international guidelines (29), need not be excluded necessarily from the opportunity for a potentially life-saving intervention.

We chose to study patients with CF because a large majority suffer from low BMD and, simultaneously, represent a relatively homogeneous population, which allows for better control of potentially confounding variables. Because there were no longitudinal data on BMD after lung transplant in any patient population when this study was started, we chose the control treatment (vitamin D and calcium) largely to determine the natural history of BMD in patients with CF after transplant. We believed that a number of patients might gain BMD, despite the immunosuppressant therapy, due to increases in physical activity and weight gain. We chose pamid - ronate for the treatment arm, because it seemed to have advantages over other antiosteoporotic medications at the time this study began. Bioavailability was not a concern as it is with oral bisphosphonates vis-à-vis CF and transplantation (30) and the long half-life of pamidronate was a significant benefit. The disadvantage of an intravenous route of administration was minimized in our patients as the drug was often given through permanent venous access lines. Although this study was not powered to demonstrate efficacy in reducing the sequelae of osteoporosis, the ultimate goal of the treatment of posttransplant osteoporosis is to reduce fractures and kyphosis to improve the quality of life of transplant recipients.

The beneficial effects of pamidronate in our study compare favorably with other studies of posttransplantation osteoporosis. Valero and coworkers were the first to demonstrate that posttransplant osteoporosis could be treated successfully with ethidronate (14). They reported increases in vertebral BMD of 8.2% in 23 patients treated for 1 yr while control subjects lost 3.4%. More recently, Shane and coworkers demonstrated that pamidronate (a single 60 mg infusion) followed by four cycles of etidronate plus calcitriol for 1 yr significantly slowed declines in spine (0.2 ± 0.9% versus 6.8 ± 1.0%) and femoral neck (2.7 ± 1.4% versus 10.6 ± 1.1%) BMD and reduced fractures (3 versus 17, p < 0.02) in 18 patients after cardiac transplantation when compared with historic controls (31). In contrast, Van Cleemput and coworkers were not able to demonstrate a benefit for etidronate alone after cardiac transplantation (32). Calcitonin has been shown to improve vertebral BMD (+6.4%) in a randomized, controlled, 1-yr trial in liver transplant recipients (14), but other studies of calcitonin after transplant have been less convincing. Monofluorophosphate resulted in 12.5% and 30% increases in vertebral BMD in cardiac transplant recipients after 1 and 2 yr, respectively (16), but the quality of accrued bone after this therapy has been debated. Patients treated with calcidiol or alphacalcidiol have weaker responses (16, 32). Hormone replacement has not been studied rigorously in transplant recipients (1, 33).

Although osteoporosis is very common after lung transplantation (3), only two studies on BMD have attempted to address natural history and therapy. Ferrari and coworkers measured BMD within 6 mo of transplant and, again, 6 and 12 mo after transplant in 14 patients with various lung diseases and found mean ± SD declines in spine and femur BMD of 4.0 ± 1.7% and 0.8 ± 2.0%, respectively, after 6 mo and 2.1 ± 2.3% and +1.0 ± 1.9%, respectively, after 1 yr (33). All patients were treated with vitamin D and calcium and five patients were treated with hormone replacement therapy or monofluorophosphate for declining BMD at 6 mo, a practice that improved BMD by 1 yr. More recently, Shane and coworkers treated 30 patients with a variety of underlying diseases with antiresorptive drugs with resulting mean ± SD declines in spine and femoral neck BMD of only 1.3 ± 1.2% and 2.8 ± 1.4%, respectively in the first posttransplant year (34). Even on therapy, 11 patients sustained an alarming 54 atraumatic fractures during the study.

At study entry, our patients had severely deranged bone metabolism with accelerated bone resorption (cross-linked N-telopeptides and deoxypyridinoline levels averaging at or above the upper limits of normal of our assays) and markedly suppressed bone formation (mean osteocalcin levels below the normal range of the assay). Control patients demonstrated ongoing bone hyperresorption throughout the study, whereas pamidronate effectively lowered cross-linked N-telopeptides and deoxypyridinoline levels. Osteocalcin levels rose, similar in nature but larger in magnitude, to other studies of osteoporosis after transplantation (14, 15), to levels 10-fold above entry values irrespective of treatment. These data suggest that improving health or declining corticosteroid doses had a salutary effect on bone formation. Lastly, the rise in serum PTH levels probably resulted from cyclosporin nephropathy rather than osteomalacia (given the rise in serum 25-OHD levels through the study).

Several potential limitations of this study exist. First, one of its strengths, a relatively uniform patient population, is also a weakness, because the results may not extrapolate, particularly in terms of degree of benefit, to patients without CF following lung transplantation. Patients with CF, being younger than other patients with end-stage lung disease, may have had intrinsically better bone formation capabilities as evidenced by the large rebound in serum osteocalcin levels over time and the failure of the control patients to lose significant BMD despite glucocorticoids and cyclosporin. Nonetheless, the effect of pamidronate on BMD, even if less in other patient populations, should still be positive. Second, the number of patients in this study is relatively small and the number of potentially confounding variables (given the complexity of the post-lung transplant course) is relatively large. Therefore, it is possible that under different circumstances, the results would have been less favorable. Third, this study was not blinded and despite efforts to have primary end points analyzed by individuals unaware of the study treatments, it is possible that bias could have been introduced. During the design of the study, we did not believe that the risk/benefit ratio of sham infusions justified their use, as the primary end point was being analyzed in a "blinded" manner.

In summary, our results demonstrate a very favorable impact of pamidronate on BMD after lung transplantation in patients with CF, a relatively homogeneous group with severe posttransplant osteoporosis. Although our sample size was too small to detect a fracture benefit, the increase in BMD seen in the pamidronate group might be expected to reduce fractures. However, the increased activity level of our patients may have mitigated the ostensibly favorable impact of pamidronate on BMD vis-à-vis fracture risk after transplant. More aggressive surveillance for, and targeted management of, osteoporosis before lung transplantation will probably mitigate the clinical ramifications of low BMD. No doubt, further experience with pamidronate or other new generation bisphosphonates will be necessary to determine their long-term usefulness in the posttransplant setting.