Histopathological and ultrastructural features of dermal telangiectasias in systemic sclerosis

Pathology

ISSN: 0031-3025 (Print) 1465-3931 (Online) Journal homepage: http://www.tandfonline.com/loi/ipat20

Histopathological and ultrastructural features of dermal telangiectasias in systemic sclerosis Jennifer G. Walker, John Stirling, Dimitra Beroukas, Kencana Dharmapatni, David R. Haynes, Malcolm D. Smith, Michael J. Ahern & Peter J. Roberts‐Thomson To cite this article: Jennifer G. Walker, John Stirling, Dimitra Beroukas, Kencana Dharmapatni, David R. Haynes, Malcolm D. Smith, Michael J. Ahern & Peter J. Roberts‐Thomson (2005) Histopathological and ultrastructural features of dermal telangiectasias in systemic sclerosis, Pathology, 37:3, 220-225 To link to this article: http://dx.doi.org/10.1080/00313020500033262

Published online: 06 Jul 2009.

Submit your article to this journal

Article views: 10

View related articles

Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=ipat20 Download by: [Universität Osnabrueck]

Date: 01 February 2016, At: 14:54

Pathology (June 2005) 37(3), pp. 220–225

ANATOMICAL PATHOLOGY

Histopathological and ultrastructural features of dermal telangiectasias in systemic sclerosis JENNIFER G. WALKER*, JOHN STIRLING{, DIMITRA BEROUKAS*, KENCANA DHARMAPATNI{, DAVID R. HAYNES{, MALCOLM D. SMITH*, MICHAEL J. AHERN* AND PETER J. ROBERTS-THOMSON*

Downloaded by [Universität Osnabrueck] at 14:54 01 February 2016

Departments of *Immunology, Allergy and Arthritis and {Anatomical Pathology, Flinders Medical Centre, and {Department of Pathology, University of Adelaide, Adelaide, South Australia

Summary Aims: To investigate the histological, ultrastructural and immunohistochemical features of the vascular lining of dermal telangiectasia, a characteristic clinical finding in scleroderma. Methods: Standard histological, electron microscopic and immunohistological techniques were used to examine dermal telangiectasias in five patients with limited scleroderma, the most common scleroderma variant in Caucasian populations. Results: The telangiectasias were dilated postcapillary venules located in the papillary and superficial reticular dermis. The vessel walls consisted of non-fenestrated endothelial cells surrounded by a variable number of pericytes and smooth muscle cells. There were no unique ultrastructural features. Thickened collagen fibres in the reticular or deep dermis were seen in all but one patient, although in variable and generally minimal quantities. Surrounding infiltrating inflammatory cells were scarce. No enhanced endothelial staining was obtained with antibodies directed against endoglin, endothelin, E-selectin and ICAM-1 suggesting a resting or inactivated state. Conclusion: The immunohistological and ultrastructural features of the lining endothelium of established telangiectasias in long-standing, limited scleroderma appear benign. It would be of interest to examine telangiectasias in the early phase of their formation. Alternatively, other explanations need to be explored in understanding the aetiopathogenesis of telangiectasia in scleroderma.

frequent finding in the more benign limited variant of scleroderma (also including the CREST subtype).2 This limited variant is the most common form of scleroderma seen in Caucasian populations. Telangiectasias, however, are not specific for scleroderma as they are also found in other systemic rheumatic disorders, the familial disorder hereditary haemorrhagic telangiectasia (HHT), following actinic skin damage, or occasionally in health.1 We have recently reported in detail the clinical features of telangiectasias in scleroderma.3 Here we report the histological and ultrastructural features of dermal telangiectasias obtained from patients with limited scleroderma. We were particularly interested in determining if there was altered expression of two endothelial proteins, endoglin (CD105) and endothelin (ET-1), both of these being implicated in the aetiopathogenesis of the vascular abnormalities observed in scleroderma.4 Endoglin is an auxiliary protein to one of the TGFb1 surface receptors on endothelial cells and mutations of this gene with reduced endoglin expression characterise the genetic defect in HHT.5 We have previously shown enhanced endoglin staining of endothelium in the dermal vessels in the early active phase of scleroderma.6 In contrast, endothelin is released from endothelial cells and is a potent vasoconstrictor.7 Circulating levels are increased in scleroderma and there is one report of increased endothelin staining of dermal vessels in scleroderma.8

Key words: Telangiectasia, systemic sclerosis, scleroderma, endothelium, endoglin, endothelin.

METHODS

Received 20 October 2004, revised 1 March, accepted 2 March 2005

This was a cross-sectional study of the telangiectasias obtained by multiple 3-mm punch biopsies of the visibly dilated vessels of the otherwise uninvolved skin of the dorsum of the hand from five patients with limited scleroderma. Two to three punch biopsies of individual mat blanching telangiectasias were taken for each patient. These patients fulfilled the American Rheumatology Association criteria for scleroderma and were all listed on the South Australian Scleroderma Register.9 All patients were female with a mean age of 74 years (range 69–79 years), mean disease duration (from first symptom of scleroderma) of 30 years (range 8–45 years) and four patients had a positive centromere antibody. For comparative purposes, skin was also obtained from 10 patients with scleroderma during their initial diagnostic work-up, as previously described.6 These patients consisted of six with diffuse cutaneous scleroderma (three male, three female; mean age 68 years) and four with

INTRODUCTION Telangiectasias are simply defined as visibly dilated capillaries or venules seen in the skin or mucosa. Multiple telangiectasias are particularly common in the systemic autoimmune disorder systemic sclerosis (scleroderma) where they are most commonly observed on the face and hands.1 Their number and size tend to increase with disease duration and their occurrence is strongly correlated with the presence of the anti-centromere antibody which is a

Patients and controls

ISSN 0031-3025 printed/ISSN 1465-3931 # 2005 Royal College of Pathologists of Australasia DOI: 10.1080/00313020500033262

Downloaded by [Universität Osnabrueck] at 14:54 01 February 2016

PATHOLOGICAL FEATURES OF DERMAL TELANGIECTASIAS

221

limited cutaneous scleroderma (two male, two female; mean age 65 years). Other skin biopsies were obtained from three patients with inflammatory skin disease (eczema, papular urticaria and lichen planus) and from three healthy controls undergoing plastic surgery. Other tissues used as positive controls included active rheumatoid synovium and tonsil (from tonsillectomy patients).

number of vessels observed in each section (score: 0, ,5% of vessels stained; 1, 6–25%; 2, 26–50%; 3, 51–75%; and 4, .76%). T-cell infiltrate (CD45Ro, positive memory T cells) was also scored semi-quantitatively as a percentage of positive staining cells compared with total number of cells (score: 1, 0–10% of cells stained; 2, 10–25%; 3, 25– 50%; 4, 50–75%; and 5, 75–100%.

Electron microscopy

Ethical approval

Tissues were fixed for 2–4 h at room temperature in 2.5% glutaraldehyde in 0.1 M phosphate buffer pH 7.4. Subsequently, specimens were washed in 0.1 M phosphate buffer, post-fixed for 1 h in 1% aqueous osmium tetroxide containing 1.66% potassium hexacyanoferrate, then dehydrated and embedded in epoxy resin using standard techniques.10

This project has the ethical approval of the Flinders Clinical Research Ethics Committee (No. 120-02).

Immunohistochemistry

Histological analysis of limited scleroderma telangiectasias H&E-stained sections from each patient were examined by an experienced skin pathologist. Features of solar elastosis (sun damage) were evident in all patients together with thickening of the stratum corneum (in keeping with their origin from acral skin). Telangiectatic vessels were visible in the papillary and superficial reticular dermis with varying size up to 1 mm in diameter. In one patient, a vessel was noted in the superficial reticular dermis with evidence of sub-intimal hyperplasia and this vessel appeared to branch into several ectatic capillary-like vessels. The telangiectatic endothelium appeared single cell in depth. Sub-intimal hyperplastic changes were apparent in the occasional small artery in two patients. Perivascular infiltrating inflammatory cells were scarce or absent. Thickened collagen fibres in the reticular or deep dermis were seen occasionally in four patients, one patient had a normal collagen component without evidence of thickening or increased collagen content. Calcification was not seen in any patient and there were no other histological features of scleroderma.

Formalin-fixed, wax-embedded tissue Tissues were fixed in 10% neutral buffered formalin at room temperature for a minimum of 24 h. The material was then dehydrated in a graded series of alcohol, cleared in xylene and embedded in paraffin wax. For immunohistochemistry, 4-mm sections were collected on slides coated with APTS (3aminopropyltriethoxysilane; Silenus, Australia) and dried overnight.

Frozen sections Synovial tissue was snap-frozen in an isopentane/liquid nitrogen mixture and stored in the 280uC freezer until required. Four-mm serial sections were cut on a cryostat (CM50; Leica, Germany), transferred to APTS-coated slides and dried overnight in a vacuum desiccator at 4uC. Dried sections were either used immediately or individually wrapped in foil and stored at 270uC until required.

Immunohistochemical labelling technique Cryosections were brought to room temperature, washed in Tris PBS pH 7.4 and endogenous peroxidase activity was inhibited using 10% H2O2 in methanol. A 20% normal donkey serum was applied for 60 min and sections were incubated with the primary antibody (Table 1) overnight at 4uC in a humidified chamber. After washing in Tris PBS, a secondary biotinylated antibody was applied for 30 min at room temperature (goat anti-rabbit antibody; Dako Laboratories, Australia; and donkey anti-mouse antibodies; Jackson Immunoresearch, USA). A standard ABC technique was performed (Vector Laboratories, USA) and signal was developed using 3,39-diaminobenzidine (DAB; Sigma, Australia). Sections were counterstained with haematoxylin and were DPX mounted.11 For control purposes, each specific antibody (Table 1) was tested on skin, synovium or tonsil as positive controls. Negative controls were incubated with Tris PBS containing 1% bovine serum albumin in place of the primary antibody.

Evaluation of staining intensity Staining intensity was assessed semiquantitatively as previously described.12 Cytokine and cell adhesion molecules which stained vessels were scored as a percentage of the total

TABLE

1

RESULTS

Electron microscopy Dilated vessels were variable in morphology (Fig. 1A). The endothelial lining of dilated vessels varied in thickness: in most vessels the endothelium was relatively thin but in a few (when nuclei were present) the cells bulged into the lumen, some with occasional villous projections. No thin fenestrated areas were found and there was a single layer of endothelial cells in all cases. The cytoplasm of endothelial cells often contained numerous intermediate filaments (presumed to be vimentin) and some were vesiculated; otherwise the endothelial cells were unremarkable. Some endothelial cells were in close contact with pericytes. Endothelial cells had a basal lamina which was variably reduplicated (Fig. 1B).

Characteristics of antibodies used for immunohistochemistry

Antibody (Cat. no.)

Isotype

Dilution

Ligand

Source

Tissue type

Anti-human endoglin (#33551A)

IgG1, k

1/30

CD105

Paraffin

Anti human UEA-1 (ES:250-2675)

Polyclonal

UEA-1

IgG2, k IgG1 IgG1 IgG1

1/400 1/1600 1/100 1/100 1/1600 1/30

PharMingen Becton Dickinson, USA Cell Marque, USA

IgG1

1/400

Anti-human Anti-human Anti-human Anti-human

CD45Ro E-selectin ICAM endothelin-1

Anti-human OPG (Mab805)

CD45Ro Dako, Australia CD62E Novacastra Laboratories, UK CD54 Novacastra Laboratories, UK Endothelin-1 (10% US Biological, USA cross-reactivity with GT-3) OPG (dimer) R and D Systems, USA

Paraffin Frozen Paraffin Frozen Frozen Frozen Frozen

222

WALKER et al.

Pathology (2005), 37(3), June

Downloaded by [Universität Osnabrueck] at 14:54 01 February 2016

A

B Fig. 1 Electron micrographs. (A) An enlarged vessel surrounded by pericytes (P). The nuclei (En) of the endothelial cells (E) have a variable profile (smooth to corrugated). The basal lamina adjacent to the endothelial cell (E) is mostly single layered. The perivascular stroma is filled with collagen fibres, seen in longitudinal and transverse section (COL). L, vessel lumen. (Bar51 mm; original magnification, 63600). (B) An enlarged vessel surrounded by welldifferentiated smooth muscle cells (SM). The nuclei (En) of the endothelial cells (E) have a variable profile (smooth to indented). The endothelial cell (E) lining in this instance is thin and the adjacent basal lamina is principally multi-layered. The perivascular stroma is filled with collagen fibres, seen in longitudinal and transverse section (COL). L, vessel lumen. (Bar53 mm; original magnification, 63900).

The perivascular stroma varied in density and contained collagen fibres, in two cases these were numerous and densely packed. The nuclei of the endothelial and perivascular cells ranged from mainly smooth to corrugated and/or indented in profile, some were deeply indented. Around some vessels there was a single discontinuous layer of pericytes (Fig. 1A), around others there were up to five discontinuous layers of smooth muscle cells (Fig. 1B). Areas of elastin were rare. Immunohistochemistry Paraffin-embedded tissue Skin sections were stained with antibodies to Ulex, endoglin and CD45Ro. Positive control tissue consisted of tonsil. Universal strong vascular staining was observed with the antibody to Ulex with no discernable differences between patients and controls. Telangiectasia endothelium stained equally as strongly as non-dilated vessels (Fig. 2A). Vascular staining with antibody to endoglin was either not apparent or weak for telangiectatic endothelium but moderate to strong for vessels seen in active sclerotic skin (characterised by prominent collagenosis with some inflammatory infiltrate) in both

limited and diffuse scleroderma and in patients with inflammatory skin disorders (Fig. 2B,C and 3). Weak vascular staining was observed in healthy controls. Staining for mature lymphocytes (CD45Ro) revealed scanty or absent cells in four biopsies and in the remaining biopsy moderate to strong staining was observed. Frozen tissues Skin sections from three patients and two healthy controls were stained with antibodies to Ulex, osteoprotegerin (OPG), E-selectin, ICAM and endothelin-1. For endothelin-1, rheumatoid synovium was used as a positive control (see Table 2). Universal strong vascular staining was observed with antibody to Ulex and OPG with no discernable difference between patients and controls. Telangiectasia endothelium stained equally as strongly as non-dilated vessels. Vascular staining with antibody to E-selectin was not apparent in patients whilst staining with antibody to ICAM was moderate in both groups with no discernable difference between the groups. Absent staining was observed with antibody to endothelin-1 in both patients and controls.

Downloaded by [Universität Osnabrueck] at 14:54 01 February 2016

PATHOLOGICAL FEATURES OF DERMAL TELANGIECTASIAS

A

B

223

C

Fig. 2 Immunohistochemical paraffin sections demonstrating: (A) prominent staining of wall of telangiectasia with antibody to Ulex (UEA-1) (original magnification, 6100); (B) absence of staining of wall of telangiectasia with antibody to endoglin (original magnification, 6100); and (C) positive staining with antibody to endoglin in small, non-dilated vessel (original magnification, 6400).

Fig. 3 Vascular staining of endoglin in skin biopsies assessed by semiquantitative analysis (scale, 0 to ++++). Each triangle represents a single patient. T, telangiectasia; OV, other non-dilated vessels in telangiectatic skin; SL, active sclerotic skin (limited disease); SD, active sclerotic skin (diffuse disease); INFL, skin from patients with inflammatory dermatoses; HC, skin from healthy controls.

DISCUSSION Multiple dermal telangiectasias are a characteristic clinical feature of scleroderma although their aetiopathogenesis is obscure. In the current study we have examined the histological, ultrastructural and immunohistochemical appearances of telangiectasia obtained by punch biopsy from the hands of five patients with long-standing limited scleroderma. Our aim was to identify any changes, particularly in the lining endothelium of the telangiectasia, which might provide insights into the causation of these vascular abnormalities and, hopefully, lead us to a better understanding of the pathology of scleroderma itself. Our study revealed that telangiectasias were dilated postcapillary venules situated in the papillary and superficial reticular dermis of the skin and, in our study, were up to

1 mm in diameter. Multiple dilated vessels were apparent in some patients, and in one appeared to originate from a small vessel with evidence of subintimal hyperplasia. These findings are very consistent with the studies of Braverman and colleagues who have performed three-dimensional reconstructions of macular telangiectasias in scleroderma and demonstrated a horizontal plexus of intercommunicating dilated vessels.13 Our electron microscopic studies confirmed that these vascular dilations were confined to post-capillary venules, as characterised by multiple laminated basement membranes in the vessel wall together with individual collagen fibrils throughout the wall. Bridging fenestrations in the endothelial cells were absent. The wall of the telangiectasia was thickened with additional cells with ultrastructural features of pericytes and smooth muscle cells consistent with the clinical observation that the telangiectatic vessels are contractile in response to direct stroking or adrenaline.14 No unique ultrastructural features of the lining endothelium was noted. Infiltration of mononuclear cells surrounding the telangiectasia was absent or scarce in four patients, although more evident in a single patient. Braverman, in contrast, found more evidence of an inflammatory infiltrate in his telangiectatic biopsies (from scleroderma patients) and hypothesised that these mononuclear cells were a source of angiogenic factors involved in the formation of the telangiectasia (and he also observed a heavier infiltrate in the telangiectasias of HHT).15 Our findings, however, do not lend strong support to this hypothesis. There was only minimal evidence of associated dermal fibrosis in our patients in accord with the apparent normality of skin texture on clinical examination of the dorsum of the hand. In many patients with limited scleroderma, sclerotic involvement of the skin of the dorsum of the hands is absent, particularly in long-standing disease after skin softening has occurred.4 In contrast,

224 TABLE

WALKER et al. 2

Pathology (2005), 37(3), June

Immunohistochemical staining of dermal vascular endothelium

Limited scleroderma patients 1 2 3 Healthy controls 1 2

Ulex

E-selectin

ICAM

Endothelin

Osteoprotegerin

+++ +++ ++++

0 0 0

+ + ++++

0 0 0

+++ ++++ +++

++++ +

++ +

++++ +

0 0

++++ ++++

Downloaded by [Universität Osnabrueck] at 14:54 01 February 2016

As assessed by semi-quantitative analysis (scale, 0 to ++++).

sclerodactyly is commonly present, even in long-standing disease. We could not identify enhanced staining of the lining endothelium of the telangiectasia with antibodies directed against markers of endothelial activation: E-selectin, ICAM, endoglin and endothelin. However, it should be emphasised that we were studying established telangiectasias in long-standing disease where evidence of prior activation may have been lost. In contrast, we have recently published our findings of prominent up-regulation of vascular endoglin in involved skin in the early fibrotic/ oedematous phase of scleroderma.6 These findings were not specific for scleroderma as they were also seen in dermal vessels from patients with inflammatory dermatoses.6 Prominent endothelial staining was observed in both patients and controls with an antibody directed against osteoprotegerin (OPG). OPG is a soluble decoy receptor for receptor activator of nuclear factor-kb ligand (RANKL) and belongs to the TNF receptor super-family.16 OPG is found closely associated with the Weibel–Palade bodies (containing von Willebrand factor) in the endothelium of the microcirculation and is actively secreted in response to inflammatory cytokines such as IL Ib and TNFa.17 We have recently reported that OPG staining is not apparent in activated endothelium as is found in the inflamed synovium of RA patients.18 Hence, the presence of normal staining in the endothelium of the telangiectasia would be consistent with a resting or inactivated endothelial lining. The aetiopathogenesis of telangiectasias is obscure.19 Factors which must be accounted for in any hypothesis include their apparent random distribution to the hands, face and mucosa of the nose, mouth and upper gastrointestinal tract, their occasional presence in health, other connective disorders or following actinic or irradiation skin damage, their common occurrence and accumulation with time in scleroderma and their expanding size over time.1,19 Venous hypertension as a cause seems unlikely in view of their rarity in the feet (where high venous pressures occur in the microcirculation). As mentioned above, inflammatory infiltrate and perivascular collagenosis is not a characteristic feature in the limited variant of scleroderma, the variant most commonly associated with telangiectasia.4 Whilst Raynaud’s phenomenon and telangiectasia are strongly associated with scleroderma, telangiectasia is not prominent in primary Raynaud’s disease. This observation suggests that telangiectasia is not a consequence of Raynaud’s vasoconstriction per se. Furthermore,

telangiectasias are common on the face where intermittent vasoconstriction is not clinically evident. Endothelial dysfunction and damage to the microcirculation is a cardinal feature of scleroderma and we can only surmise that the telangiectasias develop as a result of this damage. However, we have not noted in a previous study any significant correlation between number or size of telangiectasias and quantitative nailfold capillary dilation (microvascular change characteristic of scleroderma),19 although patient numbers in this study were small (n512), raising the possibility of insufficient power. Telangiectasia is also a cardinal feature of HHT, a genetic disorder due to a mutation in the genes coding for endoglin or ALK-1.20 In this familial disorder, the appearances and distribution of the telangiectasia are very similar to those found in scleroderma but there are subtle differences.19 For example, in HHT the telangiectasias tend to be punctate or papular (unlike the flat or mat type in scleroderma), are larger in size with thicker walls, and may occur under the nail (not seen in scleroderma). A prominent mononuclear infiltrate surrounds the telangiectasia in HHT and they have a greater tendency to bleed.15 It is intriguing that a point mutation of a single gene can cause the development of multiple telangiectasias that are very similar to those found in scleroderma, particularly so when the gene involved codes for a TGFb receptor protein (where TGFb is thought to have a pivotal role in the pathogenesis of scleroderma itself). Our recent immunohistochemical and cellular studies of endoglin in scleroderma have not revealed any apparent abnormalities of this receptor, either in the lining endothelium of involved sclerotic skin or, as in this study, in the lining endothelium of telangiectasia;6 however, our studies have been qualitative in nature rather than quantitative. A recent report has described alterations in endoglin density on fibroblasts,21 suggesting that a more detailed study concerning this protein in scleroderma may be rewarding. Another endothelial cytokine of interest in scleroderma pathogenesis is endothelin, a potent vasoconstrictor and fibrogenic protein.7 Vascular expression of endothelin and its receptor in active sclerotic skin appears enhanced, as are circulating levels.8 However, in the current study we were unable to implicate this cytokine in the pathogenesis of telangiectasia, with no apparent staining being observed in the lining endothelium of the vascular dilations. In conclusion, we have examined the histological, ultrastructural and immunohistochemical appearances of telangiectasias seen in the acral skin of five patients with

PATHOLOGICAL FEATURES OF DERMAL TELANGIECTASIAS

limited scleroderma. The endothelium of the telangiectasias appeared benign without evidence of activation (in contrast to that found in the microvasculature of active sclerotic skin). However, we did not examine telangiectasias in the early phase of their formation which may be more instructive. The understanding of the pathogenesis of the visibly dilated dermal vessels in scleroderma still remains obscure and will require further investigation. ACKNOWLEDGEMENTS We thank Dr A. Simpson, histopathologist, for her assistance in interpreting the skin pathology, Dr R. Czechowicz, dermatologist, for performing the skin biopsies, and Mrs M. Emin and Mrs C. Thomas for secretarial help.

Downloaded by [Universität Osnabrueck] at 14:54 01 February 2016

Address for correspondence: Professor P. J. Roberts-Thomson, Department of Immunology, Allergy and Arthritis, Flinders Medical Centre, Bedford Park, SA 5042, Australia. E-mail: [email protected]

References 1. Champion RH. Disorders of blood vessels. In: Champion RH, Burton JL, Ebling FJ, editors. Textbook of Dermatology. 5th ed. Oxford: Blackwell Scientific Publications, 1992; 1827–49. 2. Winterbauer RH. Multiple telangiectasis, Raynaud’s phenomenon, sclerodactyly and subcutaneous calcinosis: a syndrome mimicking hemorrhagic telangiectasia. Johns Hopkins Hosp Bull 1964; 111: 361. 3. Roberts-Thomson PJ, Mould TL, Walker JG, Smith MD, Ahern MJ. Clinical utility of telangiectasia of hands in scleroderma and other rheumatic disorders. Asian Pac J Allergy Immunol 2002; 20: 7–12. 4. Kahaleh MB, LeRoy EC. Autoimmunity and vascular involvement in systemic sclerosis. Autoimmunity 1999; 31: 195–214. 5. McAllister KA, Grogg KM, Johnson DW, et al. Endoglin, a TGFb binding protein of endothelial cells, is the gene for hereditary haemorrhagic telangiectasia type 1. Nature Genet 1994; 8: 345–51. 6. Dharmapatni AA, Smith MD, Ahern MJ, et al. The TGFb receptor endoglin in systemic sclerosis. Asian Pac J Allergy Immunol 2001; 19: 275–82. 7. Mayes MD. Endothelin and endothelin receptor antagonists in systemic rheumatic disease. Arthritis Rheum 2003; 48: 1190–9.

225

8. Vancheeswaran R, Azam A, Black C, Dashwood MR. Localization of endotheli-1 and its binding sites in scleroderma skin. J Rheum 1994; 21: 1268–76. 9. Roberts-Thomson PJ, Jones MJ, Hakendorf P, et al. Scleroderma in South Australia: epidemiological observations of possible pathogenic significance. Intern Med J 2001; 31: 220–9. 10. Woods AE, Stirling JW. Electron microscopy: the preparative techniques. Bancroft JD, Gamble M, editors. Theory and Practice of Histological Techniques. 5th ed. London: Churchill Livingstone, 2002; 679–700. 11. Smith MD, O’Donnell J, Highton J, Palmer DG, Rosenbilds M, Roberts-Thomson PJ. Immunohistochemical analysis of synovial membranes from inflammatory and non-inflammatory arthritidis: scarcity of CD5 positive B cells and IL-2 receptor bearing T cells. Pathology 1992; 24: 19–26. 12. Kraan MC, Smith MD, Weedon H, Ahern MJ, Breedveld FC, Tak PP. Measurement of cytokine and cell adhesion molecule expression in synovial tissue by digital image analysis. Ann Rheum Dis 2001; 60: 296–8. 13. Braverman IM, Keh-Yen A. Ultrastructure and three-dimensional reconstruction of several macular and papular telangiectases. J Invest Dermatol 1983; 81: 489–97. 14. Verel D. Telangiectasia in Raynaud’s disease. Lancet 1956; 271: 914–7. 15. Braverman IM, Keh-Yen A, Jacobson BS. Ultrastructure and threedimensional organization of the telangiectases of hereditary haemorrhagic telangiectasia. J Invest Dermatol 1990; 95: 422–7. 16. Schoppet M, Preissner KT, Hofbauer LC. RANK ligand and osteoprotegerin: paracrine regulators of bone metabolism and vascular function. Arterioscler Thromb Vasc Biol 2002; 22: 549–53. 17. Collin-Osdoby P, Rothe L, Anderson F, Nelson M, Maloney W, Osdoby P. Receptor activator of NF-KB and osteoprotegerin expression by human microvascular endothelial cells, regulation by inflammatory cytokines, and role in human osteoclastogenesis. J Biol Chem 2001; 276: 20659–72. 18. Haynes DR, Barg E, Crotti TN, et al. Osteoprotegerin expression in synovial tissue from patients with rheumatoid arthritis, spondyloarthropaties and osteoarthritis and normal controls. Rheumatology 2003; 42: 123–34. 19. Mould TL, Roberts-Thomson PJ. Pathogenesis of telangiectasia in scleroderma. Asian Pac J Allergy Immunol 2000; 18: 195–200. 20. Marchuk DA. Genetic abnormalities in hereditary haemorrhagic telangiectasia. Curr Opin Haematol 1998; 5: 332–8. 21. Connolly M, Venstrom K, Finley D, et al. Endoglin, a TGFb receptor III is upregulated in scleroderma fibroblasts. J Invest Dermatol 1999; 112: 587.