Joint Bone Spine 73 (2006) 733–735 http://france.elsevier.com/direct/BONSOI/
Original article
Bone mineral density in Marfan syndrome. A large case-control study Bertrand Mouraa, Florence Tubacha, Moana Sulpicea, Catherine Boileaub, Guillaume Jondeauc, Christine Mutib, Bertrand Chevallierd, Yasmine Ounnoughenee, Jean-Marie Le Parca,*, Multidisciplinary Marfan Syndrome Clinic Group a
Rheumatology Department, Ambroise-Paré Teaching Hospital, 9, avenue Charles-de-Gaulle, 92100 Boulogne-Billancourt, France b Genetics Department, Ambroise-Paré Teaching Hospital, 9, avenue Charles-de-Gaulle, 92100 Boulogne-Billancourt, France c Cardiology Department, Ambroise-Paré Teaching Hospital, 9, avenue Charles-de-Gaulle, 92100 Boulogne-Billancourt, France d Pediatrics Department, Ambroise-Paré Teaching Hospital, 9, avenue Charles-de-Gaulle, 92100 Boulogne-Billancourt, France e Ophthalmology Department, Ambroise-Paré Teaching Hospital, 9, avenue Charles-de-Gaulle, 92100 Boulogne-Billancourt, France Received 26 July 2005; accepted 30 January 2006 Available online 14 September 2006
Abstract Objective: To measure bone mineral density (BMD) in a group of patients meeting Gand criteria for Marfan syndrome, comparatively with a group of healthy controls. Methods: Dual-energy X-ray absorptiometry (DEXA) was used to measure BMD at the hip and wrist in 130 patients seen at the Multidisciplinary Marfan Clinic, Paris, France. Results were compared to values in the database of the absorptiometry machine (Hologic QDR100) and to values in 72 healthy height-matched controls including 35 whose body mass index (BMI) values were similar to those in the patients. Results: A history of fractures was noted in 32 (24.6%) patients. Z-score values were significantly decreased in the patients compared to the Hologic database values at the femoral neck (−1.190 ± 0.098, P < 0.0001) and wrist (−1.403 ± 1.06; P < 0.001). Patients had significantly lower BMD values at the femoral neck compared to the height-matched controls (0.841 ± 0.15 versus 1.010 ± 0.017; P < 0.0001). BMD values were also significantly lower in the patients compared to the controls of similar height and BMI. BMD values did not correlate with history of fractures or acetabular protrusion. In the patients, BMD values lower than −2.5 correlated with presence of dural ectasia. Conclusion: Men and women with Marfan syndrome have significant osteopenia independent from BMI. © 2006 Elsevier Masson SAS. All rights reserved. Keywords: Marfan syndrome; Bone mineral density; Body mass index
1. Introduction Marfan syndrome is an autosomal dominant condition due to mutations in the fibrillin gene located on chromosome 15. About 4 to 6/10,000 live born infants are affected. Abnormalities involve the musculoskeletal, cardiovascular, ocular, and neurological systems. Diagnostic criteria were published in 1988 and revised at the Gand Conference in 1996. Earlier diagnosis of Marfan syndrome has led to substantial improvements in the prevention and treatment of disease-related events, via both pharmacological and surgical means. Most notably, pro-
* Corresponding
author. Tel.: +33 1 49 09 56 75; fax: +33 1 49 09 58 65. E-mail address:
[email protected] (J.-M. Le Parc).
1297-319X/$ - see front matter © 2006 Elsevier Masson SAS. All rights reserved. doi:10.1016/j.jbspin.2006.01.026
gress has been made in preventing sudden death due to rupture of the aorta. The major clinical impact of musculoskeletal abnormalities in Marfan syndrome has been established by clinical and radiological studies. An increase in the fracture risk and a decrease in bone mineral density (BMD) values have been documented in most patients with Marfan syndrome, raising questions about the relationships between the disease and bone mass. In January 1995, a multidisciplinary clinic for Marfan syndrome patients was created at the Ambroise-Paré Teaching Hospital in Boulogne-Billancourt, France. In a study of clinical features and BMD values in 70 patients meeting Gand criteria for Marfan syndrome, we found significant osteopenia contrasting with a relatively low prevalence of fractures [1]. Here, we expand these findings by reporting data in a larger
734
number of patients and comparing them to values in healthy controls.
Table 2 BMD at the hip in patients with Marfan syndrome and in controls
2. Methods
BMD (mean ± S.D.) Femur Neck Ward triangle Trochanter
We included 130 adult patients seen consecutively at our Multidisciplinary Marfan Syndrome Clinic and meeting Gand criteria [2]. Patients were evaluated once a week as described elsewhere [3]. Body mass index (BMI) was computed in each patient. BMD values were measured at the hip and wrist by dual-energy X-ray absorptiometry (DEXA) using a Hologic QRD 1000 machine (Hologic, Waltham, MA). All BMD measurements were obtained by the same highly experienced investigator, and the results were reported in g/cm2. The coefficient of variation computed from repeated measurements of the phantom supplied by Hologic was 0.34% ± 3%. Because many patients had severe scoliosis or a history of spinal surgery, we did not measure BMD at the spine. BMD values in the patients were expressed as Z-scores and compared to those in the Hologic database. Then, we compared BMD values in the patients to those in 72 controls, including 27 young height-matched adults, 15 women and 12 men, recruited among hospital students and staff members who consented to the study and 45 adult men for whom BMD measurements at the hip were available (data supplied by Christian Roux, MD, Cochin Teaching Hospital, Paris, France). Because dural ectasia is common in patients with Marfan syndrome and may be related to BMD, we conducted separate analyses in Marfan patients with and without dural ectasia. 2.1. Statistical analysis Clinical features in cases and controls were compared using chi-square (χ2) and Student’s t tests. Correlation coefficients were computed between BMI and BMD, with the alpha risk set at 5%. BMD values were expressed as mean ± standard deviation (S.D.) and compared using Student’s t tests. For comparisons of BMD values in patients and in the Hologic database, we used the Z-score adjusted to 0 with the univariate Student’s test. BMD values in g/cm2 were compared between cases and healthy controls. Statistical tests were run using StatView 1.0 (Abacus Concept, Berkeley, CA). 3. Results Table 1 reports the main clinical characteristics in the patients and controls. Among the patients, 24.6% had a history of trauma-related or low-impact fractures, 85% had scoliosis,
Marfan (N = 130) 0.841 ± 0.15 0.781 ± 0.13 0.606 ± 0.13 0.620 ± 0.12
Controls (N = 72) 1.010 ± 0.17 0.876 ± 0.16 0.725 ± 0.15 0.765 ± 0.14
P value 0.0001 0.0003 0.0001 0.0001
Table 3 BMD at the hip in males with Marfan syndrome and in control males matched on BMI BMD values Mean (kg/m2 ± S.D.) Total hip (g/cm2 ± S.D.) Femoral neck (g/cm2 ± S.D.) Ward triangle (g/cm2 ± S.D.) BMD: bone mineral density.
Patients (N = 36) 24.2 ± 2.3 0.913 ± 0.8 0.830 ± 0.7 0.617 ± 0.5
Controls (N = 45) 23.1 ± 2.14 1.062 ± 0.4 0.895 ± 0.6 0.723 ± 0.5
P value < 0.05 0.0001 0.02 0.0007
17.7% had spondylolisthesis, 59.3% had grade II or higher acetabular protrusion, and 55.4% had dural ectasia. Z-scores at both the hip and wrist were significantly decreased in the patients compared to the Hologic database (total femoral neck, −1.190 ± 0.98 and total wrist, −1.404 ± 1.06 (P < 0.0001). Table 2 shows that BMD (g/cm2) at the femoral neck was significantly lower in the patients than in the height-matched controls with higher BMI values. We then compared femoral neck BMD values in patients whose BMI was 20 or higher and in 45 controls matched on BMI (mean BMI, 24.2 ± 2.3 in the patients and 23.1 ± 2.1 in the controls). As shown in Table 3, BMD values remained significantly lower in the patients. As expected, BMD correlated with BMI in our overall patient population, both at the wrist (r = 0.33, P < 0.0001) and at the hip (r = 0.36, P < 0.0001). Finally, patients with dural ectasia had significantly lower femoral neck BMD values than patients without dural ectasia (0.810 ± 0.14 g/cm2 and 0.865 ± 0.14 g/cm2, P = 0.017). 4. Discussion This is the first study comparing BMD values in over one hundred patients meeting Gand criteria for Marfan syndrome and in healthy controls. The results confirm that both men and women with Marfan syndrome have osteopenia at the wrist and hip. Bone mass was decreased compared to values in the Hologic database and to measurements made in heightmatched healthy adults. To evaluate the hypothesis that dural ectasia in Marfan syndrome is related to decreased bone strength, we compared BMD values in patients with and without dural ectasia. Dural ectasia was more common in patients with osteoporosis defined as hip BMD lower than −2.5 S.D.
Table 1 Clinical features of patients with Marfan syndrome and healthy controls Total N = 130 Age (years) 34.7 (10.7) Height (cm) 180.5 (9.7) BMI (kg/cm2) 20.9** (± 3.5) BMI: body mass index. *P < 0.0001; **P < 0.001.
Males N = 50 35.5 (10.2) 188.4* (7.4) 20.5 (3.5)
Females N = 70 34.1 (11) 175 (7.2) 20.5 (3.5)
Controls N = 72 35.7 (9.8) 180.9 (6.9) 23.2** (2.8)
Tall males N = 15 41.2 (6.4) 182.9 (3.6) 24.2** (2.3)
735
than in the other patients. However, the small number of cases in this comparison precludes definitive conclusions. As expected, BMD correlated with BMI in our patients. BMD correlates with BMI in the general population. However, when we compared a patient subgroup characterized by normal BMI values to controls matched on BMI, we found significantly lower BMD values at the hip in the patients, suggesting an association between Marfan syndrome and osteopenia. Our results agree with data reported by Kohlmeier et al. [4,5] but diverge from findings by Tobias et al. [6], Gray et al. [7] and, more recently, Giampietro et al. [8]. The patient selection criteria used in these studies differed from those used in our earlier study [1]. The clinical impact of our findings challenges the notion that Marfan syndrome is associated with a high fracture risk. A history of fractures was noted in 10% of the patients in our earlier study [1] and 24.6% in the study reported here. This higher prevalence is similar to that reported by Grahame and Pyeritz [9]. We did not separate trauma-related fractures from low-impact fractures, and we do not have epidemiological data on the prevalences of the various fracture types in the normal population in France. Methodological limitations of our study include the cross-sectional design, young patient age, and limited number of patients, particularly in each age group. In normal adults, BMD shows a close positive correlation with height, so that BMD values in populations of tall individuals are overestimated. This would be expected to lead to high BMD values in populations of patients with Marfan syndrome. To control for this source of bias, apparent BMD can be computed. Using this method, Kohlmeier et al. documented osteopenia in patients with Marfan syndrome [10]. Longitudinal studies would provide information on the relevance and impact of osteopenia in patients with Marfan syndrome. Peak acquired bone mass in adolescents with Marfan
syndrome deserves to be determined in order to assess bone mass changes during growth. In conclusion, in a group of 130 patients meeting Gand criteria for Marfan syndrome, we found significant osteopenia and a fairly high prevalence of fractures. The clinical impact of osteopenia in Marfan syndrome deserves to be evaluated in longitudinal studies. Currently available data are not sufficient to establish recommendations for preventing or treating bone loss in patients with Marfan syndrome. References [1] Le Parc JM, Molcard S, Tubach F. Bone mineral density in Marfan syndrome. Rheumatol 2001;40:358–9. [2] De Paepe A, Devereux RB, Dietz HC, Hennekam RC, Pyeritz RE. Revised diagnostic criteria for the Marfan syndrome. Am J Med Genet 1996;62:417–26. [3] Le Parc JM, Plantin P, Jondeau G, Goldschild M, Albert M, Boileau C. Bone mineral density in sixty adult patients with Marfan syndrome. Osteoporos Int 1999;10:475–9. [4] Kohlmeier L, Gasner C, Bachrach LK, Marcus R. The bone mineral status of patients with Marfan syndrome. J Bone Miner Res 1995;10:1550– 5. [5] Kohlmeier L, Gasner C, Marcus R. Bone mineral status of women with Marfan syndrome. Am J Med 1993;95:568–72. [6] Tobias JH, Dalzell N, Child AH. Assessment of bone mineral density in women with Marfan syndrome. Br J Rheumatol 1995;34:516–9. [7] Gray JR, Bridges AB, Mole PA, Pringle T, Boxer M, Paterson CR. Osteoporosis and the Marfan syndrome. Postgrad Med J 1993;69:373–5. [8] Giampietro PF, Peterson M, Schneider R, Davis JG, Raggio C, Myers E, et al. Assessment of bone mineral density in adults and children with Marfan syndrome. Osteoporos Int 2003;14:559–63. [9] Grahame R, Pyeritz RE. The Marfan syndrome: joint and skin manifestations are prevalent and correlated. Br J Rheumatol 1995;34:126–31. [10] Carter DR, Bouxsein ML, Marcus R. New approaches for interpreting projected bone densitometry data. J Bone Miner Res 1992;7:137–45.
