| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Guest Access | Sign In via User Name/Password
|
|
|
|||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Department of Respirology (Drs. Stephenson and Tullis), Adult CF Centre, the Departments of Endocrinology (Dr. Jamal) and Radiology (Drs. Dowdell and Pearce), St. Michael’s Hospital, and the Research Institute (Dr. Corey), The Hospital for Sick Children, University of Toronto, Toronto, ON, Canada.
Correspondence to: Anne Stephenson, MD, St. Michael’s Hospital, Respirology, 30 Bond St, 6th Floor, Bond Wing, Toronto, ON, M5B 1W8 Canada; e-mail: stephensona{at}smh.toronto.on.ca
Abstract
Study objectives: The objectives of this study were to determine the prevalence of morphometric vertebral fractures in a large cohort of adult cystic fibrosis (CF) patients, and to examine the association between fractures and bone mineral density (BMD).
Design: Cross-sectional retrospective study.
Setting: A tertiary care academic hospital.
Patients: Adult CF patients who had undergone BMD testing and chest radiography within 1 month of each other.
Measurements and results: BMD was measured by dual-energy x-ray absorptiometry (DXA) at the lumbar spine (LS) and femoral neck (FN). Vertebral fractures were diagnosed using lateral chest radiographs. Several clinical and biochemical variables were assessed as correlates. Sixty subjects (36%) had z scores between –1.0 and –2.5, and 15 subjects (9%) had z scores of < –2.5. Twelve patients (7.2%) had 19 morphometric fractures. The mean BMD at the LS was 1.266 g/cm2 in the fracture group and 1.112 g/cm2 in the nonfracture group (p = 0.0002). The mean BMD at the FN was 1.129 g/cm2 in the fracture group and 0.987 g/cm2 in the nonfracture group (p = 0.0006). Both FEV1 and body mass index were significantly associated with BMD at both the LS and the FN.
Conclusion: Seven percent of adult patients with CF had vertebral fractures as determined by morphometry. Subjects in the fracture group had both clinically and statistically higher BMD as measured by DXA. Our findings raise the intriguing possibility that BMD may not be useful in identifying CF patients with fractures.
Key Words: adults • bone densitometry • cystic fibrosis • osteoporosis • vertebral fractures
Cystic fibrosis (CF) is an autosomal genetic disease causing thick tenacious secretions and recurrent pulmonary infections. In the 1960s, patients with CF died at a median age of 10 years, whereas in the 2000s the median age of survival is into the fourth decade of life.1 Published studies consistently show that CF patients have lower bone mineral density (BMD) as measured by dual-energy radiograph absorptiometry (DXA) when compared to gender-matched population-based norms. The prevalence of low BMD in CF patients varies between 50% and 90%, depending on the subgroup being studied.2345 Although it is likely to be multifactorial, the etiology of low BMD in CF patients includes malabsorption of vitamins and minerals,67 hormonal deficiencies,8 corticosteroids,3 inflammatory cytokines,9 nutritional status,4 and possibly CF transmembrane conductance regulator genotype.10
BMD is a surrogate marker for fracture risk, which is a clinically more important outcome. Small studies3111213 have documented vertebral deformities in CF patients, with those who are awaiting lung transplant being at highest risk. In postmenopausal women, age and low BMD (T score, < 2.5), as measured by DXA testing, predict fracture risk14; however, this is not the case in other populations. Jamal et al15 studied individuals with dialysis-dependent renal failure and found no association between DXA testing and fractures in this population. Similarly, Mitra et al16 were unable to show a significant association between BMD and fractures in patients with ankylosing spondylitis. It is uncertain whether data from the postmenopausal literature can be extrapolated to CF patients. Individuals with CF are much younger, and the pathophysiology of their bone disease is quite different from that found in elderly postmenopausal women. The purpose of this study was to determine the prevalence of vertebral fractures in a large cohort of adult CF patients and to examine the relationship between DXA BMD measurements and fractures.
Materials and Methods
All patients attending the adult CF clinic at St. Michael’s Hospital in Toronto who had undergone DXA testing and chest radiography as a part of routine care within 1 month of each other were included in the study. The study was approved by the research ethics board at St. Michael’s Hospital, and as the study was a retrospective review individual patient consent was not required. In all patients, a diagnosis of CF was made by sweat testing/genotyping together with typical clinical manifestations. We excluded patients who were receiving therapy currently or had received prior therapy with bisphosphonate, and we excluded those patients who had received a lung transplant. BMD was measured using DXA (Lunar Prodigy; GE Healthcare; San Diego, CA) testing at St. Michael’s Hospital. All tests were performed with the same unit, and each unit was calibrated daily with a phantom control for longitudinal instrument stability. Bone density was measured at the lumbar spine (LS) [L1-L4] as well as the femoral neck (FN) and was expressed as the number of grams of bone mineral per square centimeter of bone. The data were also expressed as a z score and a T score. A T score is the number of SDs below the mean score for young adult control subjects, while the z score reflects the number of SDs below the mean of age-matched control subjects. Osteopenia is defined according to the World Health Organization (WHO) as being present if the T score is between –1.0 and –2.5. Osteoporosis is defined as a T score that is < –2.5.17
Lateral radiographs were examined for thoracic and lumbar morphometric vertebral fractures as well as nonvertebral fractures by a musculoskeletal radiologist who was blinded to the DXA results. Vertebral fractures were identified using semiquantitative morphometry.18 A fracture was defined as a decrease of at least 25% in any vertical height measurement (ie, anterior, mid, and posterior) when compared to adjacent vertebrae. Body mass index (BMI) and FEV1, expressed as a percentage of normal predicted values, were measured at the time of DXA testing. Blood was drawn on the day of the DXA scan for biochemistry measurements, which included ionized calcium, phosphate, magnesium, 25-OH vitamin D (radioimmunoassay; DiaSorin; Stillwater, MN), 1,25-dihyroxyvitamin D (calf-thymus receptor assay; Varian; Palo Alto, CA), growth hormone (Ultrasensitive human growth hormone Access assay; Beckman Coulter; Brea, CA), bioavailable testosterone (radioimmunoassay), thyroid-stimulating hormone (Access HYPERsensitive human thyroid-stimulating hormone assay; Beckman Coulter), parathyroid-stimulating hormone (intact parathyroid hormone SP immunoradiometric assay; DiaSorin), osteocalcin (Osteocalcin radioimmunoassay; DiaSorin), insulin-like growth factor-1 (IGF-1 immunoradiometric assay; Nichols Institute Diagnostics; San Clemente, CA), creatinine, and estrogen (Access Estradiol assay; Beckman Coulter). Glucocorticoid exposure was determined from clinic charts. Other variables such as pancreatic status, the presence of CF-related diabetes, sputum bacteriology, and pulmonary function measurements were obtained from the Toronto CF Database.
Statistical analysis was performed using a statistical software package (SAS, version 8e; SAS Institute; Cary, NC). Descriptive statistics were calculated to describe the overall patient population in the study. The Pearson correlation coefficient was used to determine the association between BMD and other continuous variables. Bone density measurements were then modeled using multiple linear regression to determine how the significantly associated variables predicted bone density. The Student t test was used to compare the mean differences in demographics, BMD at the LS and FN, biochemistry, nutritional status, and pulmonary function testing among subjects with and without fractures. Logistic regression was used to examine the association between the presence of vertebral fractures and BMD by DXA at the LS (L1-L4) and the FN, accounting for potential confounders such as FEV1, BMI, gender, and age. Data were also analyzed separately for men and women. A p value of < 0.05 was considered to be significant.
Results
One hundred sixty-seven adult CF patients (male patients, 62%) were included in the study. A total of 311 clinic patients underwent DXA testing, but, of those patients, 144 did not have a corresponding chest radiograph that had been obtained within 1 month of the DXA and, therefore, were not included in this analysis. The characteristics of the subjects (Table 1 ) in this study did not differ from those in the general clinic population. Overall, 75 of 167 patients (45%) were found to have reduced BMD at any site. Sixty of those subjects (36%) had osteopenia, while 15 subjects (9%) had osteoporosis. When analyzed individually, FEV1 (percent predicted) and BMI were positively associated with BMD at the LS (r = 0.29, p = 0.0002; and r = 0.25, p = 0.001, respectively) and the FN (r = 0.37, p = <0.0001; and r = 0.28, p = 0.0003, respectively). In multiple regression, FEV1 (percent predicted) and BMI continued to be significant predictors of BMD at the LS when adjusted for age and gender (p = 0.005 and 0.02, respectively). FEV1 (p = 0.0002), BMI (p = <0.0001), and age (p = <0.0001) were significant predictors of BMD at the FN. No significant interactions were detected.
|
|
We found that 7% of our patients had vertebral fractures. This is comparable to that reported in untreated, postmenopausal, osteoporotic women1920 despite the fact that the CF population is much younger. Other published studies31112 in the adult CF population have documented the prevalence of vertebral fractures to be as high as 25 to 51%, which is considerably higher than the findings in our study. Possible explanations for this include the fact that lateral radiographs may underestimate the prevalence of vertebral fractures as the vertebral bodies may be obscured by overlying parenchymal abnormalities. However, both Aris et al3 and Conway et al11 used lateral chest radiographs to document vertebral compression fractures with prevalence rates of 51% and 26%, respectively. Furthermore, Kim et al21 showed that chest radiographs are adequate to document clinically significant vertebral compression fractures, although these fractures often go unreported.
Another possibility for these discrepant results is that our CF patients differ significantly from other study populations. Specifically, the use of high-fat, high-calorie diets in CF patients has been standard practice in Toronto for the past 3 decades. The overall mean BMI in our study population was 22.5 kg/m2 compared to a range of 18 to 21 kg/m2 seen in other studies.341213 Consequently, our subjects had higher BMIs during peak bone mass acquisition. Maintaining normal body weight during this time period may play an important role in bone development, and this, in turn, may translate into a lower risk of future fractures. Another key difference in our study population was the severity of pulmonary disease. The mean FEV1 in our population was 62% predicted, whereas the mean FEV1 in other studies varied between 33% and 51%.3111213 Severity of lung disease has been consistently associated with reduced BMD,41112 which may also impact on bone quality and fracture risk. Although the pathophysiology of bone disease in CF patients is not clearly understood, the way to prevent fractures in this population may be to concentrate on preserving pulmonary function and maintaining normal nutritional status rather than considering medical therapies such as treatment with bisphosphonate.
Several cross-sectional studies31213 have evaluated the relationship between fractures and BMD in CF patients, although no significant relationships have been documented to date. An unexpected finding in our study was the paradoxical significant relationship seen between BMD and fractures in the CF group studied. Subjects with fractures had higher BMD than did those who did not have vertebral fractures. Based on studies14 in postmenopausal women, low BMD increases the risk of fractures. Although vertebral fractures themselves can falsely elevate BMD on DXA testing,22 BMD was higher not only at the LS, but also at the hip, suggesting that this increase was not secondary to the vertebral fractures themselves. Furthermore, the difference in BMD between the two groups was not only statistically significant, but also clinically significant. Within the literature on postmenopausal women, improvements in BMD of 2.5 to 5% at the LS and 5 to 9% at the hip in those patients treated with bisphosphonate compared to placebo are considered to be clinically important.23 Our study showed a 14% difference in BMD between the two groups at both the LS and the FN.
Only 12 subjects were found to have 19 vertebral fractures in our study. One might expect that if no relationship between DXA and fractures was found, the study may be underpowered to adequately detect this association. However, both a clinically and statistically significant relationship was detected, therefore, it is hard to explain this based on the small number of outcomes. If patients in the nonfracture group had extremely low BMDs, one might argue that the cortex of the vertebral bodies was poorly visualized, making it difficult to accurately measure the vertebral height, and, hence, vertebral compression fractures might be missed in this group of patients. However, the mean BMD in the nonfracture group was 1.112 g/cm2 (T and z score, –0.9) and 0.987 g/cm2 (T and z score, –0.3), respectively, at the LS and FN. These values do not meet the criteria for osteopenia, making it extremely unlikely that fractures were missed due to low bone density in the nonfracture group. Other possible reasons for our counterintuitive finding include the notion that the individuals with fractures had poor bone quality, which was not appreciated with DXA testing. This highlights the fact that BMD, although reduced, may not accurately reflect the fracture risk in this population. Although no biologically plausible reason for the fracture group to have higher BMD is apparent, it is possible that bone and mineral metabolism in CF patients is altered in certain subgroups, creating poor quality bone in those individuals. Bone quality assessed by bone biopsy has only been examined in two published studies to date. Haworth et al24 published a retrospective study of postmortem bone specimens from 15 patients with CF. Eleven of the subjects had received a lung transplant and were receiving long-term glucocorticoid therapy. Not surprisingly, BMD was reduced in this population. The results were confounded by the fact that the subjects were examined after transplantation and the biopsy specimens were obtained from a postmortem examination. This limits the generalizability of the results to the general CF population. A study by Elkin et al25 examined 20 healthy CF patients and concluded that there was evidence of reduced bone formation rather than increased bone resorption. It is also interesting to note that the patients with fractures had lower serum 25-OH vitamin D levels compared to the nonfracture group, although this difference was not statistically significant. Vitamin D deficiency in CF patients is common, and its role in bone disease in this population is unclear. What constitutes a clinically and physiologically important difference is serum vitamin D levels is unknown; however, the possibility that vitamin D deficiency affects bone quality and may contribute to fracture risk is plausible. Further research is needed to better understand the pathophysiology of bone disease in CF patients, and additional longitudinal studies on fractures in this population are required to confirm the findings of this study.
It is important to recognize the possible limitations of our study. The results are based on cross-sectional data and prevalent fractures. Because of this, no comments can be made about the incidence of new fractures and the ability of DXA to predict future fractures. We have reported on the association between BMD and fractures using WHO criteria, but the application of WHO criteria to BMD in CF patients has not been validated. Also, subjects were not formally asked about a history of traumatic vertebral fractures in this study. There were slightly more male patients in the fracture group, although this was not statistically significant (p = 0.71). It is well-known that young men experience more traumatic fractures than young women, presumably from engaging in higher risk activities. If the vertebral fractures in our subjects were secondary to trauma rather than to osteoporosis, the relationship between BMD and fractures may be obscured. However, these patients had been followed up regularly (approximately every 3 to 4 months) since infancy, making it unlikely that a history of trauma significant enough to cause a vertebral fracture would not be documented in the patients’ charts. In an attempt to address this issue, the charts of the 12 subjects with vertebral fractures were reviewed, and no documentation of traumatic vertebral fractures was found for any of the patients. Along the same vein, individuals in the fracture group had slightly higher BMI and FEV1 values, although this was not statistically significant. It is possible that this allowed these subjects to engage in more high-risk activities that would potentially increase BMD, while also increasing their risk of fractures. A limitation of this study is that activity level was not formally assessed; therefore, it is difficult to comment on the effect this may have had on both BMD and the risk of fracture. Future studies are needed to evaluate the important effect of activity level on fracture risk in this population.
Fractures were identified using morphometric techniques, which have limitations. Not all deformities that meet the criteria using morphometry are due to vertebral fractures. Other spinal disorders such as congenital abnormalities or osteoarthrosis can alter the vertebral shape, although these conditions are rare in the general population and are not known to be associated with CF. All subjects who participated in this study are followed up regularly at a tertiary care hospital, which may limit the generalizability of the results. However, there was a broad range of disease severity, based on FEV1 and BMI, represented within the study population, thus enhancing the generalizability of the findings.
Although low BMD is common in CF patients, the clinical relevance of this finding is not clear. There is some evidence to suggest that individuals with CF are at increased risk for fractures, but the best method for identifying those who are at the highest risk is uncertain. To date, no studies in CF patients have been able to demonstrate that low BMD on DXA discriminates between fracture and nonfracture groups. Currently, screening DXA scans have been recommended by a Cystic Fibrosis Foundation Consensus Conference26 on all CF patients > 8 years of age if ideal body weight is < 90%, FEV1 is < 50% predicted, glucocorticoids (
5 mg/d for 90 d/yr) were being administered, or there was delayed puberty or a history of fractures. If the T or z score is 
– 2.0, the recommendation is to consider bisphosphonate therapy and to monitor the patient with annual DXA testing. Based on the current literature, it is not clear whether DXA testing is the correct or appropriate modality to assess fracture risk in this population. Further prospective studies need to be performed to assess the usefulness of DXA as well as other techniques such as quantitative CT scanning or ultrasound in identifying those patients who are at the highest risk for fractures.
Footnotes
Abbreviations: BMD = bone mineral density; BMI = body mass index; CF = cystic fibrosis; DXA = dual-energy x-ray absorptiometry; FN = femoral neck; LS = lumbar spine; WHO = World Health Organization
The authors declare no conflicts of interest.
Received for publication December 14, 2005. Accepted for publication February 3, 2006.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
