Skeletal Radiology

, Volume 39, Issue 7, pp 645–654

Herniation pits and cystic-appearing lesions at the anterior femoral neck: an anatomical study by MSCT and µCT

Authors

    • Department of RadiologyTrauma Center Murnau
  • Ulrich Esch
    • Department of RadiologyTrauma Center Murnau
  • Ahmed Nabil Abdulazim
    • Biomechanics LaboratoryParacelsus University Salzburg and Trauma Center Murnau
  • Peter Augat
    • Biomechanics LaboratoryParacelsus University Salzburg and Trauma Center Murnau
Scientific Article

DOI: 10.1007/s00256-009-0792-9

Cite this article as:
Panzer, S., Esch, U., Abdulazim, A.N. et al. Skeletal Radiol (2010) 39: 645. doi:10.1007/s00256-009-0792-9

Abstract

Objective

To determine distinguishing features between herniation pits (HPs) and other cystic-appearing lesions at the anterior femoral neck in multi-slice computed tomography (MSCT) and micro-computed tomography (microCT) examinations.

Materials and methods

Institutional review board approval was obtained to examine 37 proximal femora of 23 cadaveric specimens (mean age available in 19 cadavers, 83 years; range 68–100 years; 9 female, 8 male, 6 unknown). All 37 femora were investigated by MSCT. 23 femora, which revealed cystic-appearing lesions at the anterior femoral neck in MSCT examinations, were additionally examined by microCT. Cystic-appearing lesions were categorized by their location, sclerotic margin, demarcation and shape in MSCT with assessment of inter-observer agreement. Detailed cortical and trabecular properties were evaluated in microCT examinations.

Results

There were seven HPs in three femora. There were a number of abnormalities potentially imitating HPs, including focal osteoporosis (13 in 13 femora), degenerative changes (5 in 4 femora) and trabecular restructuring (5 in 4 femora) at the anterior femoral neck. HPs were differentiated on the basis of their subchondral/subcortical location, completely surrounding sclerosis, clear demarcation and round-to-oval shape in MSCT. Because of their location and their microscopic appearance, HPs seem to resemble intra-osseous ganglia at the anterior femoral neck.

Conclusion

HPs have to be differentiated from other cystic appearing lesions at the anterior femoral neck to avoid overestimation of their incidence in the context of diagnosis of femoroacetabular impingement.

Keywords

Spiral computed tomographyBone cystsGanglion cystFemur neckHip joint

Introduction

Herniation pits (HPs) of the femoral neck were first described by Pitt et al. in 1982 [1]. He described them as well delineated, round-to-oval radiolucencies in the proximal superior quadrant of the femoral neck, just beneath the anterior cortex. He further suggested that these cavities are formed by herniation of soft tissues through erosions or perforations [2], created by abrasive action of the overlying hip capsule.

In contrast to earlier publications, in which HPs were generally recognized as an incidental finding [1, 3], in current literature they are mentioned in the context of femoro-acetabular impingement (FAI) [49] and are discussed as a possible radiographic indicator for FAI [4, 9]. Because FAI leads to premature osteoarthritis of the hip [911], recognition and consideration of HPs are gaining increasing clinical relevance.

Advanced high spatial resolution of multi-slice computed tomography (MSCT) in routine skeletal diagnostic imaging increasingly depicts osseous lesions and disturbances of trabecular architecture in the anterior part of the femoral neck. However, not all of these lesions are necessarily HPs. They may also result from focal osteoporosis, degenerative changes or subchondral cysts [1218]. Because of their potential clinical relevance, it is essential to differentiate HPs from other cystic-appearing lesions in MSCT.

Micro-computed tomography (microCT) is a non-invasive method to study bone architecture at a microscopic level providing three dimensional (3D) images, which can be easily explored in various planes [19]. Since its first description as a tool for the direct analysis of 3D trabecular bone structure in 1989 [19], continuing improvements in scan resolution allows 3D analysis with quantifying aspects in trabecular and cortical bone [20, 21].

The aim of our study was to identify and classify all osseous lesions at the anterior femoral neck with MSCT as a clinically relevant diagnostic tool. The MSCT based classification was then verified using high resolution microCT imaging of each individual lesion.

Material and methods

Institutional review board approval was obtained to investigate proximal femora of cadavers from the anatomical dissecting room. Thirty-seven proximal femora without fracture or prosthesis from 23 cadavers were used. Information about age was available in 19 cadavers (mean age 83 years; range 68–100 years). Nine of these specimens were known to be male and eight were known to be female.

MSCT (LightSpeed VCT (64 detector rows), General Electrics, Milwaukee) was performed on all 37 femora with a slice thickness of 0.625 mm, interval of 0.625 mm, pitch of 0.984, 120 kV and 100 mA in bone algorithm, similar to a clinical set of scan parameters. Those femora that revealed any osseous lesions were further analyzed by microCT imaging. MicroCT (µCT 80 ScancoMedical Bassersdorf, Switzerland) scans of the head and femoral neck were performed with a slice thickness of 50 μm, slice increment of 50 μm, 70 kV, 114 μA and integration time of 200 ms in cone-beam mode.

MSCT data sets were analyzed on the PACS-based interface (J-Vision, Tiani PACS Software, Vienna, Austria) of the CT workstation. Multi-planar reconstructions with slice thickness of 0.625 mm were created in a paraxial plane (parallel to the axis of the femoral neck), a paracoronal plane (perpendicular to the paraxial plane, parallel to the axis of the femoral neck) and a parasagittal plane (perpendicular to the paraxial plane through the axis of the femoral neck). All femora were examined for the occurrence of any osseous lesions. Each individual lesion was classified using the scheme in Fig. 1 according to well-known descriptions from the literature.
https://static-content.springer.com/image/art%3A10.1007%2Fs00256-009-0792-9/MediaObjects/256_2009_792_Fig1_HTML.gif
Fig. 1

Diagram illustrating steps of categorization of osseous lesions in MSCT scans. Subchondral cysts were defined by their localization at the femoral head [1218]. Degenerative lesions were differentiated by their location within new bone formation [1215]. HPs were characterized by their completely surrounding sclerosis and clear demarcation [1, 22, 23], while focal osteoporosis was characterized by none or partial surrounding sclerosis and predominantly missing demarcation [12]

In order to characterize the osseous lesions, the following criteria were applied:
  • Existence and degree of marginating sclerosis (complete, partial, none)

  • Demarcation (clear, not clear) in all three reconstruction planes

  • Shape (round/oval, long, wedge-shaped)

  • Cortical breaks

The location of each lesion was classified in the parasagittal plane, in which the anterior, almost semicircular half of the femoral neck was subdivided into six equal portions between superior and inferior (inferior defined as 6 o’clock, anterior as 9 o’clock, superior as 12 o’clock).

MicroCT examinations were evaluated on the workstation (Image Processing Language, ScancoMedical AG, Switzerland) using the parasagittal slices. Additional 3D models with sectional planes through each lesion were generated. Each individual osseous lesion was analyzed for their trabecular and cortical features. The maximal antero-posterior, medio-lateral, and cranio-caudal diameter of the lesion, the maximal thickness of the marginating sclerosis, the minimal and maximal thickness of the overlying cortex, and the maximal contact length of the lumina of the cystic lesion and the overlying cortical line were measured

The following characterization criteria were applied:
  • Existence and degree of marginating sclerosis (complete, partial, none)

  • Demarcation of the lesion in the three orientations (clear, not clear)

  • Communication of the lesion’s lumen with adjacent voids (yes, no)

  • Communication between neighbouring lumina (yes, no)

  • Content (trabeculae, bony fragments)

  • Porosity of the cortical structures overlying the lesion (yes, no)

  • Cortical breaks in the overlying cortex (number, maximal width)

MSCT and microCT examinations were analyzed by one senior radiologist with 10 years of experience and emphasis on musculoskeletal radiology. Independent reading of MSCT examinations was performed by the chief radiologist of the department with emphasis on musculoskeletal radiology for 22 years. Inter-observer variability was determined by calculating the percentage agreement for each classification.

Results

The MSCT scans revealed 30 osseous lesions at the anterior femoral neck and three osseous lesions at the femoral head in 23 femora (12 right femora, 11 left femora). Five femora showed multiple lesions with varying characteristics. The most profound features distinguishing between the different categories of osseous lesions were the location with respect to the subchondral interface, the degree of sclerosis and the demarcation (Table 1). None of the femora showed any evidence of malignant or benign bone tumors, metastases or tumor-like lesions. No signs of destruction or post-mortem degeneration were observed resulting from the formalin fixation.
Table 1

MSCT characteristics of cystic-appearing lesions at the anterior femoral neck. Data presented as percentages

 

Osseous lesions

Femoral neck

Femoral head

Herniation pits

Focal osteoporosis

Trabecular restructuring

Degenerative changes

Subchondral cysts

n = 7

n = 13

n = 5

n = 5

n = 3

Location

  Subchondral/subcortical

100%

100%

100%

 

100%

  Inside new bone formation

   

100%

 

  7–8a

14%

 

20%

  

  8–9a

29%

55%

 

20%

33%

  9–10a

57%

38%

20%

40%

 

  10–11a

 

8%

40%

40%

 

  11–12a

  

20%

 

67%

Sclerosis

  Complete

100%

  

100%

100%

  Partial

 

62%

100%

  

  None

 

38%

   

Demarcation

  Paraaxial

100%

31%

40%

60%

100%

  Parasagittal

100%

31%

20%

20%

100%

  Paracoronal

100%

0%

20%

20%

100%

Shape

Paraaxial

  Round/oval

100%

70%

100%

20%

100%

  Long

   

80%

 

  Wedge-shaped

 

30%

   

Parasagittal

  Round/oval

100%

30%

40%

20%

100%

  Long

  

40%

80%

 

  Wedge-shaped

 

70%

20%

  

Paracoronal

  Round/oval

100%

39%

80%

40%

100%

  Long

 

15%

20%

60%

 

  Wedge-shaped

 

46%

   

Cortical breaks

57%

38%

20%

60%

100%

aSectors of the anterior half of the femoral neck respectively head, 12 = superior, 9 = anterior, 6 = inferior

A total of seven lesions were categorized as HPs in three femora, one femur containing two, one containing four lesions. All HPs showed a completely surrounding sclerotic margin, round-to-oval in shape and clear demarcation in all reconstruction planes (Fig. 2ac). A total of 13 subchondral/subcortical lesions were assigned as focal osteoporosis (Fig. 3ac). They had none or only partially marginating sclerosis which changed shape in different reconstruction planes, and was predominantly missing demarcation. Five subchondral/subcortical lesions in four femora with partial sclerotic margin, partially changing shape in different planes and only partially clear demarcation could not be assigned as HPs or focal osteoporosis and were considered nonspecific trabecular restructuring (Fig. 4ac). Lesions which resulted from a sclerotic double line or were located within appositional bone formation were assigned as degenerative changes. They showed a surrounding sclerotic margin and round or oval-to-long shape and only partially clear demarcation (Fig. 5ac). We found five of these lesions in four femora.
https://static-content.springer.com/image/art%3A10.1007%2Fs00256-009-0792-9/MediaObjects/256_2009_792_Fig2_HTML.gif
Fig. 2

Multi-planar reconstructions of MSCT scans in the a paraxial, b parasagittal and c paracoronal direction show a HP subchondral/subcortical at the anterior femoral neck with round-to-oval shape, completely surrounding sclerosis and clear demarcation in all reconstruction planes; more detailed illustration in the d parasagittal microCT scan and the e 3D reconstruction of the microCT examination (arrow) with broad cortical defect and marginating reactive bone formation

https://static-content.springer.com/image/art%3A10.1007%2Fs00256-009-0792-9/MediaObjects/256_2009_792_Fig3_HTML.gif
Fig. 3

Multi-planar reconstructions of MSCT scans in the a paraxial, b parasagittal and c paracoronal direction show subchondral/subcortical focal osteoporosis at the anterior femoral neck with round-to-oval shape, only partial to no sclerotic margin and lack of clear demarcation in the paraxial plane; more detailed illustration in the d parasagittal microCT scan and the e 3D reconstruction of the microCT examination (arrow) as area created by sparse or absent trabeculae

https://static-content.springer.com/image/art%3A10.1007%2Fs00256-009-0792-9/MediaObjects/256_2009_792_Fig4_HTML.gif
Fig. 4

Multi-planar reconstructions of MSCT scans in the a paraxial, b parasagittal and c paracoronal direction show subchondral/subcortical trabecular restructuring at the anterior femoral neck with oval-to-long shape, partially marginating sclerosis and lack of clear demarcation in the paraxial and parasagittal plane; more detailed illustration in the d parasagittal microCT scan and the e 3D reconstruction of the microCT examination (arrow) with incomplete demarcation and partial new bone formation

https://static-content.springer.com/image/art%3A10.1007%2Fs00256-009-0792-9/MediaObjects/256_2009_792_Fig5_HTML.gif
Fig. 5

Multi-planar reconstructions of MSCT scans in the a paraxial, b parasagittal and c paracoronal direction show degenerative cystic-appearing changes at the anterior femoral neck with round-to-long shape and completely surrounding sclerosis with predominantly clear demarcation; more detailed illustration in the d parasagittal microCT scan and the e 3D reconstruction of the microCT examination (arrow) showing the lesion inside appositional bone formation created by double lining

Apart from the lesions at the anterior femoral neck, we found three subchondral cysts at the femoral head in two femora. These lesions were characterized by a surrounding sclerotic margin, round-to-oval shape and clear demarcation in all reconstruction planes (Fig. 6ac).
https://static-content.springer.com/image/art%3A10.1007%2Fs00256-009-0792-9/MediaObjects/256_2009_792_Fig6_HTML.gif
Fig. 6

Multi-planar reconstructions of MSCT scans in the a paraxial, b parasagittal and c paracoronal direction show a subchondral cyst at the femoral head with round-to-oval shape, surrounding sclerosis and clear demarcation; more detailed illustration in the d parasagittal microCT scan and the e 3D reconstruction of the microCT examination (arrow) with cortical defect and marginating reactive bone formation

Evaluating the MSCT scans of all 37 femora independently, the second reader found no additional osseous lesions. Inter-observer reproducibility was 87.8% concerning the description of the sclerotic margin and differed between 78.8% (parasagittal), 93.9% (paracoronal) and 97.0% (paraxial) concerning the description of shape in the three reconstruction planes. Inter-observer reproducibility was 100.0% in describing the demarcation and the presence of cortical breaks as well as in defining the lesions’ location. The assignment to the five different categories showed complete agreement between both readers.

MicroCT examinations revealed more detailed and sometimes additional morphological information of cortical and spongy bone properties of the different lesions (Table 2). The HPs (Fig. 2d,e) were confirmed by their appearance with clear demarcation of all lesions and complete surrounding sclerotic margin including the cortical and the subchondral lamella, respectively. In six lesions a cortical break was present. All HPs revealed small bony fragments inside the lumen. Three lesions contained trabeculae, predominantly in the periphery. MicroCT examinations did not reveal additional lesions with characteristics of HPs.
Table 2

Micro-CT characteristics of cystic-appearing lesions at the anterior femoral neck. Data presented as percentages and mean values (range) in mm

 

Ossous lesions

Femoral neck

Femoral head

Herniation pits

Focal osteoporosis

Trabecular restructuring

Degenerative changes

Subchondral cysts

n = 7

n = 13

n = 5

n = 5

n = 3

Sclerosis

 Complete

86%

    

 Partial

14%

 

100%

80%

100%

 None

 

100%

 

20% (1/5)

 

 Max. thickness in mm

0.60 (0.45–0.83)

 

0.73 (0.43–1.03)

0.42 (0.35–0.50)

0.41 (0.39–0.43)

Demarcation

 Antero-posterior

100%

38%

60%

80%

33%

 Medio-lateral

100%

23%

100%

60%

100%

 Cranio-caudal

100%

15%

20%

60%

0%

Size in mm

 Antero-posterior

3.73 (2.38–5.13)

4.84 (3.08–5.86)

4.92 (3.57–5.67)

3.13 (2.67–4.25)

2.40 (2.31–2.48)

 Medio-lateral

3.66 (2.00–5.20)

4.79 (2.60–6.63)

3.58 (2.26–5.52)

3.54 (1.84–4.79)

2.32 (1.87–2.78)

 Cranio-caudal

4.91 (3.25–7.35)

5.07 (2.60–6.05)

3.37 (2.10–5.45)

8.08 (4.35–1.10)

3.08 (2.65–3.50)

Cortical thickness in mm

 Minimal

0.42 (0.21–0.63)

0.21 (0.11–0.36)

0.34 (0.20–0.67)

0.17 (0.11–0.25)

0.19 (0.11–0.25)

 Maximal

0.74 (0.53–1.00)

0.55 (0.20–0.99)

0.75 (0.51–1.12)

0.43 (0.35–0.60)

0.42 (0.25–0.65)

Cortical breaks

86%

92%

0%

100%

100%

 Number

0.86 (0–1)

3.15 (0–14)

 

2.00 (1–5)

1.67 (1–3)

 Maximal width in mm

0.90 (0.10–1.47)

0.37 (0.07–2.00)

 

0.42 (0.07–1.20)

1.22 (1.08–1.42)

Cortical contact

 Maximal in mm

2.85 (1.90–4.01)

3.12 (1.19–4.77)

2.18 (0.97–3.37)

2.84 (2.00–4.16)

2.08 (1.85–2.5)

Trabeculae

43%

100%

80%

60%

0%

Bony fragments

100%

15%

40%

0%

0%

 Maximal length in mm

0.32 (0.10–0.65)

0.38 (0.10–0.65)

0.36 (0.36–0.36)

  

 Maximal width in mm

0.25 (0.10–0.40)

0.13 (0.10–0.16)

0.11 (0.11–0.11)

  

Lesions categorized as focal osteoporosis in MSCT scans differed in their appearance in the microCT scans. All of them showed complete sclerosis, which was not apparent on the MSCT scans. Most of the lesions had cortical breaks and all lesions contained trabeculae.

Lesions in the category of trabecular restructuring (Fig. 4d,e) confirmed partial surrounding sclerosis in all cases. Clear demarcation was present in only one to three cases in different planes. Cortical breaks were not found in any lesion. Four lesions contained trabeculae and two lesions contained bony fragments.

Lesions categorized as degenerative changes (Fig. 5d,e) showed partial sclerosis in four cases, while one case showed no sclerosis at all, in contrast to the MSCT findings. Clear demarcation in the three different planes varied between three and four cases in the different planes. Cortical breaks were found in all cases, ranging from one to five in number. Three lesions contained trabeculae; however, bony fragments were never found.

The subchondral cysts (Fig. 6d,e) only showed partial sclerosis in microCT examinations and demarcation was not clear in all planes. Cortical breaks existed in all cases, ranging from one to three in number. Subchondral cysts contained neither trabeculae nor any bony fragments.

All lesions demonstrated a connection of the lumina with the voids of the adjacent spongy structures. Cortical porosity was observed in 79% of all lesions. The extent of the lesions’ contact to the cortical interface or the subchondral lamella could not be used as a distinguishing feature between categories.

Comparison of MSCT and microCT evaluations revealed that sclerosis was more prominent in MSCT than in microCT. More than half of the completely sclerotic lesions in MSCT emerged as only partially sclerotic lesion in the microCT investigations. All partially sclerotic focal osteoporosis lesions showed no sclerosis in microCT scans. Moreover, lesions representing focal osteoporosis, which were conspicuous in MSCT, were detectable in microCT, but not necessarily conspicuous. Both phenomena were most likely related to partial volume effects in MSCT examinations despite high spatial resolution.

Discussion

In this study, we determined distinguishing features between herniation pits (HP) and other cystic lesions at the anterior femoral neck utilizing multi-slice CT and microCT. Cystic lesions were divided into four different categories. Only seven out of a total of 30 cystic lesions demonstrated unambiguous signs of HPs while the majority of lesions were more likely to be associated with focal osteoporosis or degenerative changes. In the context of FAI diagnosis, in which HPs are discussed as possible anatomical markers, their differentiation is important to avoid overestimation of their incidence.

Lesions, which we classified as HPs in MSCT examinations, showed the typical characteristics of HPs as defined by Pitt et al. [1] and described in case reports [22, 23]. These characteristics are a round-to-oval subcortical/subchondral radiolucency at the anterior femoral neck with a sclerotic rim and sometimes cortical breaks. The usual location is the superior quadrant of the anterior femoral neck, and a less frequent location the inferior quadrant [23, 24].

To our knowledge, this is the first study using microCT for the investigation of HPs. MicroCT evaluation depicted detailed cortical and trabecular characteristics of HPs that are normally restricted to histological examinations. Osseous characteristics of HPs have been previously described by histological examination [1, 4, 22, 23]. The varying presence and size of cortical breaks in the overlying cortex of HPs [1, 23] is in accordance with our results. Increased cortical thickness in our specimens corresponds to reactive new bone formation in the cortex [1]. Reactive bone changes in the periphery [1] were observed as surrounding sclerosis in all lesions defined as HPs. All HPs in our specimens contained fragments of bone inside, as observed by Crabbe [22].

The size of HPs in our specimens was in accordance with the definition of Pitt et al. [1], describing HPs as having less than 1 cm in maximum diameter. The minimum size of HPs has been identified by Leunig et al. [4] with greater than 3 mm in AP radiographs and Panzer et al. [24] with at least 2 mm in MSCT. The smallest diameter in our collective was 2.0 mm.

Based on their osseous characteristics, HPs could be clearly differentiated from focal osteoporosis without discrete margins and without reactive changes in the periphery. Similar findings of focal osteoporosis deep within the femoral head were already observed by Resnick et al. [12]. In our study, focal osteoporosis was associated with the largest lesion size. However, the size of the lesions did not turn out to be a distinguishing feature. HPs also could be clearly distinguished from degenerative changes created by osteophytic rims [1215]. As described by Resnick et al. [12], these lesions often are spindle-shaped with orientation along the paraxial plane.

The five lesions, which we termed trabecular restructuring, did not fulfil the criteria of HPs and did not correspond to focal osteoporosis or degenerative changes. We do not know exactly what kind of trabecular disturbance they represent. Possibly they constitute an early stage or an atypical feature of HPs.

Cortical porosity, defined by the presence of canals inside the cortex, was observed in 79% of all lesions. These canals consist of Haversian and Volkmann’s canals with a diameter in the order of 50 μm [25]. Therefore clear differentiation from cortical breaks was possible in our microCT examinations.

Articular or juxta-articular cysts are in principle divided into two entities in the literature. One entity represents subchondral cysts, which we found in three cases. They occur within the pressure zone [12, 13, 16], are associated with osteoarthritis of the involved joint [12, 14, 15, 17, 18], commonly show communication with the articular cavity [12, 14, 18, 26] and reveal sclerotic margins [12]. The other entity corresponds to cystic defects of bone not related to joint lesions and predominantly labelled as intra-osseous ganglion [16]. Intra-osseous ganglia appear in non-weight-bearing portions of the joint [18, 26], are located in close relationship to the articular cartilage [18, 27] and are not associated with degenerative changes [18, 26, 27]. Histologically, the osseous tissue adjacent to intra-osseous ganglia was found to be sclerotic with thickened trabeculae as a consequence of active remodelling [26].

Our categorization criteria (Fig. 1) allowed clear differentiation between HPs at the anterior femoral neck and subchondral cysts at the weight-bearing zone of the femoral head, despite showing very similar morphological criteria. However, it seems reasonable to suggest, that HPs represent intra-osseous ganglia because of their location subcortically/subchondrally in the non-weight-bearing area and their clear demarcation with sclerotic border and sometimes cortical breaks. This assumption would also support the statement of Leunig et al. [4]. In contrast to publications where intra-osseous ganglia have been reported to occur usually as single lesions [18, 27], we found HPs with multiplicity.

There are study limitations to be considered. The number of investigated femora was small. Four out of the seven HPs that were found, occurred in one femur. The femora of cadavers from the anatomical dissecting room represented a population with above-average age, which accounts for the frequent osteoporotic changes. The study was based on osseous characteristics and did not provide information on adjacent soft tissue structures. The evaluation was not hypothesis based and therefore did not allow statistical analysis. In our study, MSCT was the clinically relevant method used to categorize of lesions, and microCT imaging was performed for corroboration. For the evaluation of osseous changes, which was the main topic of our study, microCT seems to be comparable with histological examinations because of its high spatial resolution.

In conclusion, HPs have to be differentiated from other cystic-appearing lesions at the anterior femoral neck to avoid overestimation of their incidence in context with the diagnostic of FAI. The rule for how to identify a HP in MSCT is to search for a round-to-oval subcortical/subchondral lesion at the anterior femoral neck, which is clearly demarcated in all reconstruction planes by a completely surrounding sclerotic margin.

Acknowledgements

The authors thank Prof. Dr. Reinhard Putz and Prof. Dr. Magdalena Müller-Gerbl, Department of Anatomy, LMU Munich, Germany, for providing the anatomical specimens and Julia Körber, Institute of Biomechanics, Trauma Center Murnau for proofreading of the manuscript.

Copyright information

© ISS 2009