AJR ARRS PQI
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Daeubler, B. F.
Right arrow Articles by Anderson, S. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Daeubler, B. F.
Right arrow Articles by Anderson, S. E.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?
Hotlight (NEW!)
Right arrow
What's Hotlight?
DOI:10.2214/AJR.05.0071
AJR 2006; 186:1246-1251
© American Roentgen Ray Society

Original Research

Alterations of the Thoracic Spine in Marfan's Syndrome

Bernd F. Daeubler1, Thierry Carrel2, Tomasz Kujawski2, Aurelia Schnyder2, Pia Zurmuehle2, Peter Vock1 and Suzanne E. Anderson1

1 Department of Radiology, University Hospital, Inselspital, Freiburgstrasse 20, Berne CH-3010, Switzerland.
2 Clinic for Cardiovascular Surgery, University Hospital, Inselspital, Berne, Switzerland.

Received January 13, 2005; accepted after revision March 14, 2005.

 
Address correspondence to B. F. Daeubler (Bernd.Daeubler{at}gmx.de).


Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
OBJECTIVE. The purpose of this study was to determine if the thoracic vertebral elements are altered in patients with Marfan's syndrome.

MATERIALS AND METHODS. Thirty patients underwent helical CT of the thorax because of suspected thoracic aortic dilatation and acute dissection. Thirteen had Marfan's syndrome and 17 did not. Two reviewers, unaware of the final diagnosis, evaluated the images by consensus for laminar thickness, foraminal width, dural sac ratios, and vertebral scalloping for T2-T12.

RESULTS. At T9-T12, dural sac ratios at the midcorpus level (p = 0.031) and foraminal width (p = 0.0124) were significantly greater in the patients with Marfan's syndrome than in the patients without. Dural sac ratios at lower endplate levels (p = 0.0685), laminar thickness (p = 0.951), and vertebral scalloping (p = 0.24) were not significantly greater in the patients with Marfan's syndrome than in the patients without.

CONCLUSION. Because the phenotypic expression of Marfan's syndrome is variable, information on the spine from thoracic studies in combination with major criteria may be helpful clinically.

Keywords: anatomy • collagen vascular disease • congenital malformations • dural ectasia • Marfan's syndrome • musculoskeletal imaging • spine


Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Marfan's syndrome, named after Antoine-Bernard Jean Marfan, a French pediatrician who lived from 1858 to 1942, is an autosomal-dominant disorder of connective tissue [1, 2]. The expression of Marfan's syndrome is highly variable [3]. Although the disorder results from mutations of the fibrillin-1 gene on the long arm of chromosome 15, the diagnosis is sometimes difficult to establish [2]. The diagnosis is based mainly on the Ghent nosology, originating from a multidisciplinary council of experts in 1996 and including major and minor clinical criteria [4].

Lumbosacral dural ectasia was considered one of the major criteria because it is limited to patients with Marfan's syndrome, neurofibromatosis, and Ehlers-Danlos syndrome [4]. For the dura to be involved as a major criterion, lumbosacral dural ectasia must be detected by CT or MRI [4]. Dural ectasia is defined as widening of the spinal canal and neural foramina, with evidence of posterior scalloping of the vertebral body and increased bilateral thinning of the cortex of the laminae and pedicles, or as the presence of a meningocele [4-6]. Alterations of lumbosacral vertebral elements in patients with Marfan's syndrome have already been reported [7, 8]. However, to our knowledge, a quantitative assessment with CT or MRI of the dural sac dimensions of the thoracic spine in patients with Marfan's syndrome has not been reported so far. The purpose of this study was to determine if the thoracic vertebral elements are altered in patients with Marfan's syndrome and to assess normal values for dural ectasia in the thoracic spine.


Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Study Population
From the database of the department of cardiovascular surgery, a list of patients with dilatation of the aortic root and type A aortic dissection was generated. The ethics committee of the institution did not require approval for this retrospective study at the time it was performed. As part of normal clinical routine, informed consent had been obtained from all patients who underwent the clinical and radiologic investigations. Thirteen of the patients had Marfan's syndrome and 17 did not. The diagnosis of Marfan's syndrome was established using the criteria of the Ghent nosology. Additional investigations included histologic and immunohistochemical examinations of the aortic wall, fibroblasts, cultured aortic tissue, and skin biopsy specimens. In addition, blood was analyzed for genetic assessment (gen-1 or gen-2 mutation). For all patients, clinical notes and laboratory results were reviewed. Other diseases such as autoimmune and collagen vascular diseases (Ehlers-Danlos) in the control group were excluded through histologic examination and biochemical screening. The control group had idiopathic aneurysms. All patients had undergone helical CT of the thorax because of a suspected aortic abnormality. Imaging was retrospectively reviewed, and the reviewers were unaware of the final diagnosis.


Figure 1
View larger version (88K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 1 —Process of measuring laminar thickness. LAM left = left lamina, LAM right = right lamina.

 
CT
All patients underwent helical CT of the thorax to rule out aortic dilatation in another segment and dissection. Enhanced helical CT was performed using an Asteion 4-MDCT scanner (Toshiba Medical Systems). The CT parameters included 4 x 2.0 mm detector collimation, a 12.0 mm/sec table speed, 120 kVp, a 0.75-sec gantry rotation, and 200 mAs. IV administration of 120 mL of iomeprol (Iomeron, Bracco), a nonionic contrast medium containing 300 mg I/mL, was performed with a flow rate of 4.0 mL/sec, whereas the scanning delay for evaluating the thoracic aorta was determined from bolus triggering in each patient individually. Image reconstruction and analysis were performed at a workstation.

Image Processing and Measurements
The acquired CT data were transferred to a computer workstation (Advantage 4.0, GE Healthcare) running on Ultra Sparc 60 hardware (Sun Microsystems) featuring two Ultra Sparc II 450-MHz central processing units and 2 GB of random-access memory. Two analysts together performed the sagittal reconstructions and determined the measurements of all patients by consensus. Neither analyst was aware of the final diagnosis.


Figure 2
View larger version (88K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 2 —Process of measuring foraminal width. FOR left = left neural foramen, FOR right = right neural foramen.

 
Laminar thickness—Laminar thickness was measured on the left and right sides at each vertebra from T2 through T12 on the axial images. At the workstation, the axial cut that yielded the largest value for each side at each vertebra from T2 through T12 was selected and the thickness of the lamina was measured. The thickness of each lamina was measured at its midpoint perpendicular to its axis in the axial plane. At each level, the thickness of the lamina was measured from the anterior cortex of the anterior surface to the posterior cortex of the posterior surface. This procedure is shown in Figure 1.

Width of neural foramina—The width of the neural foramina was measured on the left and on the right sides at each vertebra from T2 through T12. The foraminal width at the axial cut that yielded the largest value for each side at each thoracic vertebral level was measured at the workstation. The width was measured from the posterior cortex of the vertebral body to the anterior cortex of the superior articular process of the vertebral body below. This procedure is shown in Figure 2.

Vertebral scalloping—Vertebral scalloping is defined as an erosion through the central aspect of the vertebral body by a process [3, 9-11]. Therefore, sagittal helical CT reconstructions were performed at the workstation for the CT scans from T2 through T12. After the midsagittal plane of the sagittal reconstructions had been identified, the diameters of the superior and inferior endplates and of the center of the vertebrae were measured at each vertebra from T2 through T12. At each vertebral level, the values of the diameters at the superior and inferior endplates were averaged. Then, vertebral scalloping was calculated by subtracting the diameter at the midcorpus level from the calculated average diameter at the endplates. This procedure is shown in Figure 3A.


Figure 3
View larger version (141K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 3A —Measurement and calculation of vertebral scalloping and of dural sac ratios in 40-year-old man. Process of measuring and calculating vertebral scalloping. Vertebral scalloping value = (d1 + d3):2 - d2. d = diameter.

 
Dural sac dimensions—The dimensions of the spinal canal as a parameter of dural ectasia were assessed by calculating the dural sac ratios at the midcorpus lower endplate levels at each vertebral body from T2 through T12. For these measurements, sagittal reconstructions were performed at the workstation for the CT scans of each patient from T2 through T12. The midsagittal plane of the reconstructions was identified, and the anteroposterior vertebral body diameter of the inferior endplates and the anteroposterior dural sac diameter of each vertebra from T2 through T12 were measured at the midcorpus level and at the lower endplate level, respectively. This process is shown in Figure 3B for the midcorpus level and in Figure 3C for the lower endplate level. Dural sac ratios at the midcorpus level from T2 through T12 were calculated by dividing the dural sac diameter by the vertebral body diameter of the midcorpus level at each vertebra. This process was repeated at the lower endplate level from T2 through T12.


Figure 4
View larger version (113K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 3B —Measurement and calculation of vertebral scalloping and of dural sac ratios in 40-year-old man. Measurement of dural sac diameter (DSD) and vertebral body diameter (VBD) for calculating dural sac ratios at midcorpus level. This measurement was obtained for each vertebra from T2 through T12.

 

Figure 5
View larger version (112K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 3C —Measurement and calculation of vertebral scalloping and of dural sac ratios in 40-year-old man. Measurement of dural sac diameter (DSD) and vertebral body diameter (VBD) for calculating dural sac ratios at lower endplate level. This measurement was obtained for each vertebra from T2 through T12.

 
Data Analysis
To compare laminar thickness, foraminal width, vertebral scalloping, and dural sac dimensions between the patients with Marfan's syndrome and the patients without, we used Wilcoxon's rank sum test without Bonferroni adjustment. Because our data were few, the Student's t test could not be used. Bonferroni adjustment was not performed because our study was descriptive, hypothesizing that, between patients with Marfan's syndrome and patients without, there might be differences in the thoracic spine that should be examined in detail during further studies. For statistical analysis, thoracic levels were grouped into T2-T4, T5-T8, and T9-T12. For statistical purposes, the sums of the values of each parameter within these groups were obtained for patients with Marfan's syndrome and patients without, because the values for the single vertebra in this small series did not show significant differences and because we wished to reduce the great amount of expected data.

Laminar thickness—The laminar thicknesses of T2-T4 were added on the left and on the right sides for patients with Marfan's syndrome and patients without. The resulting sum for the patients with Marfan's syndrome was compared with the sum for the patients without Marfan's syndrome. This process was repeated at levels T5-T8 and T9-T12.

Width of neural foramina—For statistical analysis, the sum of the foraminal widths was calculated at T2-T4, T5-T8, and T9-T12 for the right side, the left side, and both sides together for the patients with Marfan's syndrome. These values were compared with the corresponding measurements for the patients without Marfan's syndrome.

Vertebral scalloping—The vertebral scalloping values, calculated from the average diameter at the endplates and the diameter at the midsagittal plane of each vertebra, were added for T2-T4, T5-T8, and T9-T12 and compared between the patients with Marfan's syndrome and the patients without.

Dural sac dimensions—The dural sac ratios at the midcorpus level in the patients with Marfan's syndrome were compared with the corresponding value in the patients without Marfan's syndrome at T2-T4, T5-T8, and T9-T12. This comparison was repeated for the lower endplate level.


Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Of the 13 patients with Marfan's syndrome, seven were female and six were male (age range, 17-52 years; mean, 35 years). Of the 17 patients without Marfan's syndrome, four were female and 13 were male (age range, 28-60 years; mean, 42 years).

Laminar Thickness
Laminar thickness was not statistically significantly different between the patients with Marfan's syndrome and the patients without. No differences were found within the groups T2-T4, T5-T8, or T9-T12 or in the total range, T2-T12 (p = 0.5193-0.9510; Table 1).


View this table:
[in this window]
[in a new window]

  		TABLE 1: Mean Sums of Laminar Thickness in Patients With and Without Marfan's Syndrome

 

Width of Neural Foramina
The neural foramina were found to be significantly wider in the patients with Marfan's syndrome than in the patients without, in the regions T9-T12 (p = 0.0124) and T2-T12 (p = 0.0153) (Table 2). In the regions T4-T8 and T5-T9, the difference between the two patient groups was not significant. Table 2 shows these differences in the sums of the widths of the neural foramina.


View this table:
[in this window]
[in a new window]

  		TABLE 2: Mean Sums of Foraminal Width in Patients With and Without Marfan's Syndrome

 

Vertebral Scalloping
The mean vertebral scalloping values were not significantly greater in the patients with Marfan's syndrome than in the patients without. For regions T4-T8, T5-T9, T9-T12, and T2-T12, the sums of the vertebral scalloping values did not differ between the two patient populations (p = 0.244-0.423) (Table 3).


View this table:
[in this window]
[in a new window]

  		TABLE 3: Mean Sums of Vertebral Scalloping in Patients With and Without Marfan's Syndrome

 

Dural Sac Dimensions
The sums of the dural sac ratios at the midcorpus level of T9-T12 were significantly greater in the patients with Marfan's syndrome than in the patients without (p = 0.031) (Table 4). The other regions and the whole range, T2-T12, did not show any statistically significant differences. The same calculations were done at lower endplate levels. They showed the greatest difference between the populations to be at T9-T12, although this difference was not significant.


View this table:
[in this window]
[in a new window]

  		TABLE 4: Mean Sums of Dural Sac Ratios at Midcorpus and Lower Endplate Levels in Patients With and Without Marfan's Syndrome

 


Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
Diagnosis of Marfan's syndrome is critical and ideally should take place as early as possible to allow elective surgical intervention, because dilatation and dissection of the ascending thoracic aorta are life threatening. Appropriate prevention or identification is essential for the individuals with this inherited disorder and for their families. Expression of the disease is highly variable, with more than 600 mutations currently known in the fibrillin-1 gene, which is the gene determining the main protein within the connective tissue of the human body [12]. Most recently, some mutations have been discovered on the fibrillin-2 gene [13-15]. These lead to a variety of connective tissue syndromes (Marfan's syndrome, Ehlers-Danlos syndrome, Shprintzen-Goldberg syndrome) [16, 17].

Although the location of the genetic defect of Marfan's syndrome on the long arm of chromosome 15 is known, the final diagnosis remains a clinical one, with the Ghent criteria being the current gold standard [18]. They were established in 1996 by experts in medical genetics who were revising the clinical criteria of the Berlin nosology of 1986 [19]. Only one major criterion is known for the neurologic and musculoskeletal systems— lumbosacral dural ectasia defined by either CT or MRI [4]. No minor criteria have been established for the neurologic and musculoskeletal systems. Dural ectasia is defined as a widening of the spinal canal and neural foramina, scalloping of the vertebral body posteriorly, increased thinning of the cortex of the laminae and the pedicles, or the presence of a meningocele [4-6, 9, 20].

The major criteria for the cardiovascular system are dilatation and dissection of the ascending aorta or the aortic root. Usually, the aortic root is examined by echocardiography, but CT or MRI may be necessary to evaluate more precisely the most cranial part of the ascending aorta and the aortic arch and descending aorta. The availability of CT or MR images allowed us to take a closer look at the thoracic vertebral elements, because the thoracic spine is included in the examination performed for the thoracic aorta. In our study, we tried to detect thoracic spine alterations typical of dural ectasia to gain additional objective radiologic features for the diagnosis of Marfan's syndrome. To our knowledge, no published studies have focused from the lumbar spine to the thoracic spine.

Our thoracic findings for Marfan's syndrome were more subtle than the reported lumbar findings [7, 8]. The results from our small series seem to support the theory that dural ectasia is expressed maximally at the level of the lumbosacral region, presumably because the hydrostatic pressure is highest at the bottom of a column [8].

Nevertheless, we could see that dural sac ratios at the midcorpus level for T9-T12 were significantly greater in the patients with Marfan's syndrome than in the patients without. At the lower endplate, the dural sac ratios differed the most for the same range (T9-T12), but the difference was not significant. Statistical analysis suggests this feature may be significant if a larger population is examined. We recognize that the expression of the tendency toward alterations of the spinal elements is much clearer at the lower spine levels and that these are not limited to one particular vertebra. This fact and the large data set warranted our grouping the thoracic levels into T2-T4, T5-T8, and T9-T12 for statistical comparison of the sums of each parameter, because values for a single vertebra did not show significant differences. Therefore, this grouping does not represent a limitation in the data analysis. Our data suggest a tendency toward higher dural sac ratios at the lower endplate levels. Oosterhof et al. [7] also detected a significant difference both at the midcorpus level and at the lower endplate level of the lumbar spine, when they compared populations with and without Marfan's syndrome.

Our results suggest that the width of the thoracic neural foramina is useful in Marfan's syndrome as a parameter for the diagnosis of dural ectasia. The neural foramina of the thoracic spine were wider, with high significance, in the Marfan's population than in the control group. This parameter has not been studied separately at the lumbosacral or thoracic spine level in another population. Our study suggests that widened neural foramina may be a helpful parameter in diagnosing Marfan's syndrome, even in the thoracic spine.

Sponseller et al. [8] found that laminar thickness at the lumbosacral spine was significantly smaller (p < 0.001) in patients with Marfan's syndrome (n = 16) than in individuals without Marfan's syndrome (n = 16). Our data do not show that the laminae of the thoracic spine are significantly thinner in the patients with Marfan's syndrome than in the patients without. Therefore, laminar measurements of the thoracic vertebrae may not be as good a predictor of Marfan's syndrome as are laminar measurements of the lumbosacral vertebrae. Here, it is important to take into account that Sponseller et al. [8] found a significant difference only from L5 through S1. This observation supports the theory that hydrostatic pressure, being highest at the bottom of a column, is responsible for the clearest expression of dural ectasia at the lumbosacral level of the spine [7, 8]. This observation may also explain the finding that vertebral scalloping was significantly greater in patients with dural ectasia at level S1 than in patients without dural ectasia. On the other hand, it also is important to take into account that vertebral scalloping is present in a variety of other conditions, such as neurofibromatosis, Ehlers-Danlos syndrome, spinal cord tumors, and homocystinuria [17, 20].

We are not aware of any previously published studies regarding CT measurements of the normal thoracic spine. For a detailed workup of the thoracic spine in patients with Marfan's syndrome, a larger case control study would be ideal, comparing a group of healthy volunteers with no aortic changes such as dilatation and dissection, a group of patients with aortic dissection or dilatation and Marfan's syndrome, and a second group of patients with aortic dissection and dilatation but no Marfan's syndrome. In the last of these groups, changes in collagen status should be examined in detail.

Although the population of this study was relatively small, we believe that, on thoracic CT performed for suspected dilatation and dissection of the ascending thoracic aorta, it may be interesting to review the spine for neuroforaminal width and dural sac ratios at the midcorpus level of T9-T12. Such a review offers the additional ability to detect dural ectasia, suggesting an ability to differentiate patients with Marfan's syndrome from those with idiopathic aortic dilatation and dissection. In view of the importance of preventing aortic dilatation and dissection in affected individuals and their families, our results may be helpful as they show an additional trend.


Acknowledgments
 
We thank Rene Burkhard, Department for Mathematic Statistics and Actuarial Science, University of Bern, Switzerland, for the statistical analysis.


References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Marfan's syndrome. Medcyclopaedia Web site. Available at: www.amershamhealth.com/medcyclopaedia/medical/volume%20v%202/marfans%20syndrome.asp. Accessed January 28, 2006
  2. Carter N, Duncan E, Wordsworth P. Bone mineral density in adults with Marfan syndrome. Rheumatology 2000;39 : 307-309[Abstract/Free Full Text]
  3. Mitcell GE, Lourie H, Berne AS. The various causes of scalloped vertebrae with notes on their pathogenesis. Radiology1967; 89:67 -74[Medline]
  4. De Paepe A, Devereux RB, Dietz HC, Hennekam RC, Pyeritz RE. Revised diagnostic criteria for the Marfan syndrome. Am J Med Genet 1996; 24:62:417 -426
  5. Pyeritz RE, Fishman EK, Bernhardt BA, Siegelman SS. Dural ectasia is a common feature of the Marfan's syndrome. Am J Hum Genet 1988; 43:726 -732[Medline]
  6. Fattori R, Nienaber CA, Descovich B, et al. Importance of dural ectasia in phenotypic assessment of Marfan's syndrome. Lancet 1999; 354:910 -913[CrossRef][Medline]
  7. Oosterhof T, Groenink M, Hulsman FJ, et al. Quantitative assessment of dural ectasia as a marker for Marfan syndrome. Radiology 2001;220 : 514-518[Abstract/Free Full Text]
  8. Sponseller PD, Ahn NU, Nallamshetty L, et al. Osseous anatomy of the lumbosacral spine in Marfan syndrome. Spine2000; 25:2797 -2802[CrossRef][Medline]
  9. Stern WE. Dural ectasia and the Marfan syndrome. J Neurosurg 1988; 69:221 -227[Medline]
  10. Nelson JD. The Marfan syndrome with special reference to congenital enlargement of the spinal cord. Br J Radiol1958; 31:561 -564[CrossRef][Medline]
  11. Smith MD. Large sacral dural defect in Marfan syndrome: a case report. J Bone Joint Surg Am 1993;75 : 1067-1070[Free Full Text]
  12. Vollbrandt T, Tiedemann K, El-Hallous E, et al. Consequences of cysteine mutations in calcium binding epidermal growth factor molecules of fibrillin-1. J Biol Chem 2004;279 : 32,924-32,931[Abstract/Free Full Text]
  13. Quondamatteo F, Reinhardt DP, Charbonneau NL, Pophal G, Sakai LY, Herken R. Fibrillin-1 and fibrillin-2 in human embryonic and early fetal development. Matrix Biol 2002;21 : 637-646[CrossRef][Medline]
  14. Gupta PA, Putnam EA, Carmical SG, et al. Ten novel FBN2 mutations in congenital contractural arachnodactyly: delineation of the molecular pathogenesis and clinical phenotype. Hum Mutat 2002; 19:39 -48[CrossRef][Medline]
  15. Robinson PN, Godfrey M. The molecular genetics of Marfan syndrome and related microfibrillopathies. J Med Genet2000; 37:9 -25[Abstract/Free Full Text]
  16. Collod-Beroud G, Beroud C, Ades L, et al. Marfan database (third edition): new mutations and new routines for the software. Nucleic Acids Res 1998; 26:229 -233[Abstract/Free Full Text]
  17. Hayward C, Brock DJ. Fibrillin-1 mutations in Marfan syndrome and other type-1 fibrillinopathies. Hum Mutat1997; 10:415 -423[CrossRef][Medline]
  18. Nallamshetty L, Ahn NU, Ahn UM, et al. Plain radiography of the lumbosacral spine in Marfan syndrome. Spine J2002; 2:327 -333[CrossRef][Medline]
  19. Beighton P, de Paepe A, Danks D, et al. International nosology of heritable disorders of connective tissue, Berlin, 1986. Am J Med Genet 1988; 29:581 -594[CrossRef][Medline]
  20. Villeirs GM, Van Tongerloo AJ, Verstraete KL, Kunnen MF, De Paepe AM. Widening of the spinal canal and dural ectasia in Marfan's syndrome: assessment by CT. Neuroradiology 1999;41 : 850-854[CrossRef][Medline]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?



This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Daeubler, B. F.
Right arrow Articles by Anderson, S. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Daeubler, B. F.
Right arrow Articles by Anderson, S. E.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?
Hotlight (NEW!)
Right arrow
What's Hotlight?

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS