SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Objective

To describe the basis for entheseal-associated bone disease in the spondylarthritides, by analyzing microanatomic and histopathologic relationships between soft tissue, bone cortex, and adjacent trabeculae.

Methods

Serial sections from 52 entheses were examined; these entheses encompassed small and large insertions in the upper limb (n = 21), lower limb (n = 27), and spine (n = 4) from 60 cadavers. Enthesis microdamage (fissuring) as well as vascular and reparative changes were evaluated. Contact radiographs were used to ascertain the relationship between entheses and the trabecular network.

Results

At virtually all fibrocartilaginous entheses, the deep cortical boundary was extremely thin (typically 50–600 μm) or indistinguishable, and 96% of entheses had small holes in the cortical shell (typically 100–400 μm wide). Such regions were frequent sites of bone formation and renewal (96%) and microdamage (31%); these changes were more common in the lower limb. The presence of blood vessels near holes in the cortical shell was common; in 85% of attachments, blood vessels were present on the soft tissue side of the enthesis. Highly orientated trabeculae were more obvious in the lower limb than the upper limb (59% versus 29%).

Conclusion

The trabecular network supporting the cortical shell is an integral part of the enthesis, transferring load to an extensive skeletal region. In many cases, tendons/ligaments are anchored directly to such networks. This functional integration is associated with microdamage and repair at the hard tissue–soft tissue interface. These findings have implications for understanding bone involvement in SpA and for the SpA concept in general, especially the hypothesis that enthesis–bone architecture may be important in disease initiation.

The region where a tendon, ligament, or joint capsule is attached to the skeleton is known as an enthesis, and inflammation at this site is characteristic of ankylosing spondylitis and the related spondylarthritides (SpA) (1–3). Historically, enthesitis has been viewed as a focal, insertional problem of soft tissue (4), but the associated inflammatory reaction often extensively involves the neighboring soft tissue (5). Importantly, inflammation also affects the bone, because osteitis adjacent to fibrocartilaginous entheses or next to fibrocartilaginous synovial joints is a characteristic feature of SpA (6–8). Furthermore, prominent bone involvement (notably osteolysis and hyperostosis) is conspicuous in the SAPHO (synovitis, acne, pustulosis, hyperostosis, and osteitis) syndrome (9, 10), which is closely related to SpA and which has also been linked with enthesitis (11). However, the basis for this association remains unclear. Diffuse osteitis is also well recognized in psoriatic arthritis (12), a disease that is also strongly associated with enthesitis (13). In addition, diffuse osteitis occurs in chronic multifocal recurrent osteomyelitis, which is considered to be closely related to the SAPHO syndrome and SpA (14, 15). Thus, the basis for the diffuse bone-based changes that characterize a wide spectrum of related conditions is unclear (16).

Although the enthesis–organ concept helps explain the extended pattern of soft tissue disease around attachment sites (3, 17), our understanding of the pattern and extent of bone involvement adjacent to diseased entheses is far from complete. The purpose of the present study was to describe the structure and histopathology of bone at fibrocartilaginous entheses, i.e., at the sites where disease tends to localize (3). We analyzed the microanatomic and histopathologic relationships between the bone of the outer cortical shell, the adjacent trabecular networks, and the soft tissue components of the enthesis, in dissecting-room cadavers. Our findings highlight a close functional integration between entheses and the adjacent trabecular bone network, with prominent tissue microdamage and repair at such sites. This has important implications for understanding bone disease in SpA.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Specimens were obtained from 60 cadavers (35 men and 25 women; mean age 84 years [range 49–101 years]) that had been donated to Cardiff University for anatomic investigation under provision of the 1984 Anatomy Act and the 1961 Human Tissue Act. The cadavers had been perfused with an embalming fluid containing 4% formaldehyde and 25% alcohol, for dissection by students. The cadavers were selected simply according to the quality of preservation and the absence of gross abnormalities in the region of interest. Medical histories were not available. A total of 52 different anatomic sites of entheses were examined histologically and/or radiographically (Tables 1 and 2). At each site, 2–5 specimens were obtained. Each tendon/ligament was cut transversely, ∼1 cm from its enthesis. Two parallel saw cuts were then made deep into the bone (up to 2 cm from the surface, depending on the site), using a fine-toothed modeling saw. The cuts passed either side of the enthesis, along the long axis of the tendon. The size of the specimens varied according to the size of the tendon or ligament. The largest specimens were ∼5 cm long, ∼1 cm wide, and extended ∼2 cm below the bone surface. All radiographic analyses were conducted on the embalmed material prior to any subsequent processing of the tissue for histologic evaluation.

Table 1. Presence or absence of highly orientated trabeculae (anisotropy) and signs of histopathology in the cortical shell of upper limb and spine entheses
EnthesisAnisotropyHistopathology of the cortical shell
Bone formation or renewalLocal absence of cortical boneVascular invasion into soft tissueHorizontal fissuring at tidemark
Abductor pollicis brevis insertion+++++
Abductor pollicis longus insertion++++
Annulus fibrosus (L4/5)++++
Biceps brachii origin (short head) + coracobrachialis insertion+++
Biceps brachii insertion++++
Collateral ligaments of the proximal interphalangeal joints of fingers++
Common extensor muscle origin+++
Common flexor muscle origin+++
Coracoacromial ligament (coracoid attachment)+++
Extensor carpi radialis brevis insertion++++
Extensor carpi radialis longus insertion++++
Extensor carpi ulnaris insertion++
Extensor digitorum++++
Extensor pollicis longus insertion+++
Facet joint capsule+++
Flexor carpi ulnaris insertion + origin of hypothenar muscles+++
Flexor digitorum profundus insertion+++
Flexor pollicis brevis + abductor pollicis brevis insertion+++
Flexor pollicis longus insertion+++
Interspinous ligament (L5/S1)++++
Pronator teres insertion ++
Sacroiliac joint–interosseous ligament+++
Subscapularis insertion++++
Supraspinatus insertion+++
Triceps brachii insertion+++++
Table 2. Presence or absence of highly orientated trabeculae (anisotropy) and signs of histopathology in the cortical shell of lower limb entheses
EnthesisAnisotropyHistopathology of the cortical shell
Bone formation or renewalLocal absence of cortical boneVascular invasion into soft tissueHorizontal fissuring at tidemark
Achilles tendon insertion+++++
Adductor longus origin+++
Adductor brevis origin+++
Adductor magnus insertion (hamstring part)+++
Anterior cruciate ligament–tibial attachment++++
Anterior cruciate ligament–femoral attachment+++
Biceps femoris insertion++++
Extensor hallucis longus insertion++++
Flexor hallucis longus insertion+++
Gastrocnemius insertion (lateral head)++
Gluteus medius insertion++++
Gluteus minimus insertion+++
Iliopsoas insertion+++++
Medial collateral ligament of the knee–femoral attachment+++
Medial collateral ligament of the 1st metatarsophalangeal joint++++
Obturator internus insertion+++++
Patellar tendon origin+++
Patellar tendon insertion++++
Peroneus brevis insertion++++
Peroneus longus insertion+++
Pes anserinus insertion++
Popliteus tendon origin/lateral collateral ligament++++
Posterior cruciate ligament–tibial attachment++++
Posterior cruciate ligament–femoral attachment+++++
Quadriceps tendon insertion+++++
Sacroiliac joint–interosseous ligament+++
Sartorius origin+++
Tibialis anterior insertion++++

Histologic examination.

The specimens were postfixed in 10% neutral buffered formalin, decalcified with 5% nitric acid, dehydrated with graded alcohols, cleared in xylene, and embedded in paraffin wax. Serial longitudinal sections were cut at 8 μm, and 12 sections were mounted on slides at 1-mm intervals throughout each block. Consecutive slides were stained with Hall and Brunt's quadruple stain (18), Masson's trichrome, and toluidine blue.

Scoring system.

Enthesis microdamage (defined as fissuring at or near the hard tissue–soft tissue boundary), together with vascular and reparative response changes, were evaluated by examining histologic sections at 1-mm intervals at each of the different attachment sites. The results of these evaluations are shown in Tables 1 and 2. A positive result was recorded whenever a feature was observed in 1 or more specimens.

Evaluation of trabecular architecture.

In order to evaluate the relationship between the enthesis itself and the trabecular architecture at a distance of several millimeters from the actual insertion site, radiographs were obtained using a Faxitron MX-20 system (Faxitron, Wheeling, IL). Due to the nature of the study (i.e., a survey of a small number of samples from a large number of different entheses), no quantitative data are presented.

Statistical analysis.

The chi-square test was used to evaluate the significance of differences in the incidence of both trabecular anisotropy (highly orientated trabeculae) and horizontal fissures at the superficial boundary of the cortical shell, in entheses from the upper and lower limbs. P values less than 0.05 were considered significant.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

A substantial region of compact bone at any fibrocartilaginous enthesis was never observed (Figures 1a and b). This was in marked contrast to the thick cortex of compact bone at the site of insertion of the pronator teres; this single example of a fibrous enthesis was included in the study for comparative purposes only (Figure 1c). Thus, the cortical shell at fibrocartilaginous entheses was extremely thin and typically no more than 50–600 μm thick, i.e., of dimensions that included those of the underlying trabeculae with which it was continuous. In the pronator teres specimen shown in Figure 1c, the underlying cortex was ∼3 mm thick. At the sites of many fibrocartilaginous entheses, it was difficult to define any deep cortical border at all, and thus cortical thickness was difficult to measure. Consequently, even the largest of tendons or ligaments with fibrocartilaginous entheses (e.g., Achilles and patellar tendons) were effectively anchored to an extended trabecular network that formed a continuum at the hard tissue–soft tissue interface. The cortical shell was an intimate mixture of calcified fibrocartilage and bone (Figure 1d), the proportions of which varied both within and between entheses.

thumbnail image

Figure 1. Photomicrographs showing normal histology of bone at entheses. a and b, At fibrocartilaginous attachments, the cortical shell of bone (arrows) is extremely thin; in these examples of the Achilles tendon (AT) and the peroneus longus (PL), it is no thicker than the underlying trabeculae (T). Note that in peroneus longus, most of the trabeculae are orientated along the long axis of the tendon. c, In marked contrast, there is a thick layer of cortical bone (CB) at the fibrous enthesis of the pronator teres (PT). d, The thin cortical shell at a fibrocartilaginous enthesis (the Achilles tendon in this example) is an intimate mixture of bone (B) and calcified fibrocartilage (CF). Its outer limit is the tidemark (TM). Note the avascularity of the uncalcified enthesis fibrocartilage (UF) in this specimen. Toluidine blue stained in a and c; Masson's trichrome stained in b and d. Bars in ac = 3 mm; bar in d = 100 μm.

Download figure to PowerPoint

The entheses at which there was evidence of histopathologic change in or near the cortical shell are listed in Tables 1 and 2. Horizontal fissures at or near the superficial boundary of the cortical shell (the tidemark) (Figure 1d) were recorded in 31% of entheses. Comparable degrees of fissuring were evident in the upper and lower limbs (29% and 37%, respectively; P not significant). The fissuring was not artefactual, because at all entheses fissuring was associated with damage and/or repair in adjacent fibrocartilage (Figure 2b).

thumbnail image

Figure 2. Photomicrographs showing histopathology of the cortical shell of bone at entheses. a and b, Insertional tendon of iliopsoas. In the specimen shown in a, the cortical shell has split horizontally (arrows) at its superficial boundary (the tidemark [TM]), and the damage is associated with cell clustering (C) in the immediately adjacent soft tissue. This is interpreted as evidence of degeneration with ensuing repair. In the specimen shown in b, at the site of what is interpreted as an old horizontal tear at the level of the tidemark (arrowheads), there is a sudden change in the organization of the uncalcified enthesis fibrocartilage. In the bottom right corner of b, the fibrocartilage cells are arranged in longitudinal rows between parallel bundles of collagen fibers (FC1). However, as indicated by the arrowheads, there is an abrupt transition to a more cellular and disorganized repair fibrocartilage (FC2) that lies at the site of a small hole in the cortical shell, through which blood vessels (BV) are passing. This is an indication of reparative fibrosis. c and d, Small discontinuities in the cortical shell of entheseal bone (i.e., the local absence of cortical bone) were seen in virtually all entheses and are shown in c at the site of insertion of the tibialis anterior (TA) and in d in the Achilles tendon (AT). At both entheses, blood vessels (arrows) are conspicuous in the tendons and continuous with those in the bone marrow (M). Masson's trichrome stained. Bars in a and c = 100 μm; bars in b and d = 200 μm.

Download figure to PowerPoint

At virtually all entheses (96%), ranging in size from large to small and including those from the upper limbs, lower limbs, and the spine, cortical bone was locally absent in microscopic areas, so that soft tissue contacted marrow space directly through the holes. These holes ranged in size from ∼100 μm to 400 μm, as measured along the surface of the bone (i.e., at the soft tissue interface) (Figures 2c and d). Although enthesis fibrocartilage could be avascular (Figure 1d), in most of the fibrocartilaginous entheses examined in this study, blood vessels were present at sites of direct contact between tendon/ligament and adjacent vascular marrow spaces. Such invasion of entheses by thin-walled blood vessels and the presence of cellular connective tissue at the sites of microscopic holes in the cortical bone were features of 85% of entheses and sometimes gave the impression of marrow that had escaped through a restraining cortical shell (Figure 3c). It is thus noteworthy that there was marrow spillage into the interosseous ligament of the sacroiliac joint at its enthesis.

Signs of remodeling in the cortical shell and to a lesser extent in the neighboring trabeculae were present in the majority (96%) of entheses (Figures 3a–c and Tables 1 and 2). Thus, there was frequent evidence of osteoid formation, osteoblasts, and osteoclasts (Figure 3a), and the juxtaposition of woven and lamellar bone was a common feature of the cortical shell of many entheses and indicated the presence of bone of different ages (Figure 3b). Occasionally, however (in 4% of entheses), there was little evidence of bone turnover in association with microscopic holes in the cortical shell. Where this occurred, the underlying trabeculae were generally thin and their interconnectivity was low, which is suggestive of osteoporotic change. Although large holes (>1 mm wide) in the cortical shell were rare at entheses, where they were seen (e.g., in supraspinatus, flexor digitorum profundus insertion, and the tibial attachment of the anterior cruciate ligament), the tendon/ligament anchored secondarily to the underlying cancellous bone. Some of the larger holes were at the site of bone cysts. Enthesophytes most typically formed in the superficial (distal) part of the cortical shell in the diverse attachment sites at which they were observed. A common feature of many of the spurs was the paucity of trabecular bone within them and the frequent absence of bone from local areas of their cortex (Figure 3d).

thumbnail image

Figure 3. Photomicrographs showing histopathology of the cortical shell of bone at fibrocartilaginous entheses. a, Osteoid formation (O) at the site of a small hole in the cortical shell at the fibrocartilaginous (FC) insertion of the flexor digitorum profundus. Note the row of osteoblasts (arrows) and the prominent blood vessels (BV) in the region. b, Lamellar bone (LB) and woven bone (WB) in association with microscopic holes in the cortical shell. c, Bone marrow (M) sometimes spills out of small holes in the cortical shell into the soft tissue at fibrocartilaginous entheses, as seen in the attachment of the flexor digitorum profundus (arrow); this is a feature of remodeling. Note the osteoid formation that has begun to repair the hole. d, A hole at the tip of a bony spur at the tibial attachment site of the anterior cruciate ligament. The sides of this spur are covered with fibrocartilage. Masson's trichrome stained. Bar in a = 100 μm; bars in bd = 200 μm.

Download figure to PowerPoint

Finally, islands of cartilage were occasionally present in the marrow spaces between the trabeculae, adjacent to sites where there were holes in the cortical shell (notably in the annulus fibrosus and the origin of the common extensor tendon). This suggests that the biomechanics of the enthesis contributed to the development of cartilage within the bone, which is consistent with the close functional integration of these tissues.

A prominent feature of many entheses that was particularly evident radiographically was the greater number of spicules of cancellous bone per unit area beneath the cortical shell (Figure 4). Commonly, this affected the skeleton to a depth of ∼2–4 mm below the surface, showing how tendon/ligament attachment sites influence the structure of the adjacent trabecular bone, in keeping with Wolff's law. At large entheses, many trabeculae were orientated longitudinally, along the long axis of the tendon or ligament in the superficial part of the enthesis; these features were evident histologically (Figure 1b) but were more striking radiographically (Figures 4c and d). The trabeculae were often thicker and more densely packed here than in the deep part of the attachment site.

thumbnail image

Figure 4. Radiographic images showing the architecture of the cancellous bone at fibrocartilaginous entheses. At many attachment sites, including all of those illustrated here, there is a concentration of trabecular bone extending for ∼2–4 mm beneath the hard tissue–soft tissue interface. All of this bone should be regarded as part of the enthesis itself. a, Tibial attachment of the anterior cruciate ligament (ACL). b, Insertion of the infraspinatus tendon (IF). c, Insertion of the patellar tendon (PT). d, Insertion of the Achilles tendon (AT). At the enthesis of the Achilles and patellar tendons, note the parallel trabeculae (black arrows) orientated along the long axis of the tendon. Note also the contrasting, radiating trabeculae (RT) that anchor the parallel trabeculae at the attachments of the patellar tendon, the multidirectional arrangement of trabeculae at the attachment of infraspinatus, and the presence of a further system of trabeculae at the attachment of the Achilles tendon (white arrows), which stabilize those that run parallel to the long axis of the tendon.

Download figure to PowerPoint

An evaluation of the histologic sections suggested that highly orientated (i.e., anisotropic) trabeculae were significantly more common (P < 0.05) in the lower limb than in the upper limb (59% versus 29%) (Tables 1 and 2). At the tibial insertion of the patellar tendon, the parallel trabeculae were anchored by groups of radiating trabeculae, which ran at various angles to the long axis of the tendon and extended up to 4 mm into the patella (Figure 4c). Similar radiating trabeculae were seen at the attachment of the tendon of tibialis anterior to the medial cuneiform. Particularly at the entheses of small tendons and ligaments acting on multiaxial joints in the upper limb, trabecular orientation was more variable and multidirectional (Figure 4b). Therefore, the extensive bone remodeling and trabecular orientation, in large entheses in particular, provide further evidence for a strong functional integration of the enthesis and the underlying bone.

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Our findings on the structure and pathology of bone at a large number of entheses offer an explanation for the wide spectrum of target sites in SpA that have been clinically recognized but not conceptualized from a pathophysiologic perspective (16). The extreme paucity of cortical bone at fibrocartilaginous entheses is likely to be of pivotal importance. It means that the bony trabecular network is functionally integrated with the soft tissue attachment and dynamically involved in entheseal function. The implication is that load transfer between hard tissue and soft tissue affects a wider neighboring region than just the attachment site itself. This possibility is supported by our radiographic data showing that distinctive features in the trabecular bone architecture are maintained for a distance of several millimeters from the insertion sites. Although it is clearly impossible to determine where an enthesis ends on the bone side of the interface, we suggest that a wide trabecular network associated with each attachment site should be regarded as a part of the enthesis and fundamental to the dissipation of stress away from the hard tissue interface.

The present study shows that microdamage and/or bone formation and removal are very common at attachment sites, and that bone formation and removal are also observed in the adjacent trabecular bone network. Bone remodeling at entheses is probably integral to the periosteal bone reaction and to the spinal fusion that is characteristic of SpA. The presence of woven bone in conjunction with cartilage in areas of remodeling is suggestive of endochondral ossification, and chondrocytic cell clustering is indicative of chondrocytic metaplasia. All such evidence indicates that enthesis repair can occur in elderly individuals, although it is clearly impossible to make inferences about young persons based on these findings in cadavers of elderly individuals. Thus, whether the greater degree of activity in youth compared with old age is associated with an even higher level of microdamage and remodeling is an important question that cannot currently be answered.

The frequency and extent of microdamage and repair at insertions are likely to have implications for the etiopathogenesis of enthesitis/osteitis in SpA. Tissue repair responses are initially inflammatory in nature, and it is possible that the altered vascularity at insertions, at least in part, reflects this process. The altered vascularity of the enthesis–bone interface may also make this a site for the preferential deposition of adjuvant molecules, derived from the gut bacteria in the case of ankylosing spondylitis. The combined effects of these factors and others could predispose to innate immune system activation at these characteristic sites (3, 19). It is also possible that the degree of microdamage and vascularity adjacent to fibrocartilage could contribute to autoimmunity against this tissue, especially in HLA–B27–positive individuals, but this has yet to be shown conclusively (20). Observations in patients with SpA support the model in which class I major histocompatibility complex–associated diseases represent a fairly unique type of immunologically mediated pathology, in which adaptive and innate immune mechanisms contribute at similar degrees to disease expression (21).

We propose that the stress-dissipating role of the trabecular network at a fibrocartilaginous enthesis is complimentary to that of fibrocartilage on the soft tissue side of the attachment (3, 4, 22). In contrast, stress dissipation at a fibrous enthesis must be different, because not only is fibrocartilage absent, but also such attachments are characteristic of the shafts of long bones (23), i.e., regions where the mechanical integrity of the skeleton is promoted by a predominance of compact bone but the virtual absence of cancellous bone (24). It is likely that the greater area of the skeleton to which many fibrous entheses (e.g., pronator teres or deltoid) attach, compared with fibrocartilaginous entheses (e.g., the rotator cuff tendons) (24), is important in dissipating stress concentration.

Our findings also offer a novel explanation for the diffuse bone disease observed in patients with the SAPHO syndrome, because it is likely that deformation of the trabecular network is transmitted diffusely through the cancellous bone network, and that this is related to the pattern of osteitis noted at these sites (9, 10). Although chronic multifocal recurrent osteomyelitis is less strongly associated with SpA than is the SAPHO syndrome, the present study gives possible clues to a unifying microanatomic basis. We suggest that in SpA and the SAPHO syndrome, as in osteitis, aberrant responses to physiologic trabecular stressing contribute to disease localization by mechanisms that, until now, have been poorly understood.

The fact that microscopic holes in the cortical shell of fibrocartilaginous entheses are extremely common means that the fibrocartilage is in direct contact with bone marrow and its blood vessels in the majority of entheses, at least in older individuals. Thus, although fibrocartilage is correctly regarded as an avascular tissue in a healthy enthesis (22, 23, 25), in line with the avascularity requirement for chondrogenesis (26), the current evaluation of a large number of different entheses in elderly individuals challenges the view that the absence of vessels from fibrocartilaginous entheses is a universal rule.

We suggest that vascular invasion is probably secondary to the formation of small holes in the cortical shell or microdamage at an enthesis and is thus a sign of pathology. However, it is a feature that must be regarded almost as normal with increasing age. The invading vessels may be derived from the tendon or from the bone, and they are likely to be important in initiating enthesis repair. Blood vessel invasion from the bone side allows immune cells and stem cells to access enthesis fibrocartilage for healing and tissue repair responses. Clearly, the close juxtaposition of fibrocartilage and bone marrow provides an immediate capacity to repair stress-related damage, and marrow spillage into the soft tissue could trigger enthesitis. Our findings of new fibrocartilage contributing to the repair of horizontal fissuring are interpreted as evidence of bone marrow stem cell activity. Thus, small holes in the cortical shell could be advantageous, because they would promote tissue repair responses in regions where the cortical shell is thin and, thus, stress concentration is high. Whether such holes are also common in younger individuals remains unclear, and this is an important limitation of the present study that should be recognized.

The shoulder joint is nicely illustrative of the concept of bone involvement in SpA. Fibrocartilage is present at entheses adjacent to the glenohumeral joint, both at the attachments of the rotator cuff tendons (as shown in the present study) and at the acromial attachment of deltoid (27); therefore, such entheses have the thinnest of cortical bone shells. However, the more distal fibrous insertion of deltoid attaches to the shaft of the humerus and thus to a substantial zone of compact bone (23, 24). The close functional interrelationship and the structural continuity between some of the fibrocartilaginous entheses adjacent to the humeral head probably explain the diffuse bone edema noted by magnetic resonance imaging at that site, but not at the distal fibrous enthesis of deltoid (8).

In conclusion, this study explored the microanatomic basis for bone involvement in SpA and related conditions. It emphasizes that the trabecular network underlying the cortical shell should be considered as an integral part of the enthesis, and that this close functional integration is associated with significant microdamage and tissue repair at the soft tissue–bone interface. Taken together, these findings have implications for an improved understanding of bone involvement in SpA in general.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Dr. Benjamin had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study design. Drs. Benjamin, Emery, and McGonagle.

Acquisition of data. Drs. Benjamin, Toumi, and Redman.

Analysis and interpretation of data. Drs. Benjamin, Toumi, Suzuki, Emery, and McGonagle.

Manuscript preparation. Drs. Benjamin, Toumi, Emery, and McGonagle.

Statistical analysis. Dr. Toumi.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

We thank K. Hayashi for cutting sections of the spinal entheses.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES