Interstitial Vascularity in Fibrosing Alveolitis

Departments of Pathology and Thoracic Medicine, Interstitial Lung Disease Unit, Royal Brompton Hospital; National Heart and Lung Institute, Imperial College of Science, Technology, and Medicine; Department of Histochemistry, Royal Postgraduate Medical School, Hammersmith Hospital; Division of Academic Rheumatology, Royal Free Hospital, London; Academic Rheumatology, University of Nottingham Clinical Sciences Building, City Hospital, Nottingham, United Kingdom; and Division of Respiratory Diseases, University of Siena, Siena, Italy

	ABSTRACT

TOP ABSTRACT METHODS RESULTS DISCUSSION REFERENCES

 
The aim of this study was to evaluate interstitial vascularity in cryptogenic fibrosing alveolitis (CFA) and in fibrosing alveolitis associated with systemic sclerosis (FASSc). Open lung biopsies from eight patients with CFA, nine patients with FASSc, and normal lung from 12 patients undergoing surgery for lung cancer were studied. Markers for endothelial cells (CD34) and cell proliferation (proliferating cell nuclear antigen) were localized by sequential immunohistochemistry and quantified using computer-assisted analysis. Vascular distribution was evaluated at increasing distances (up to 160 µm) from the airspaces. Vessel density was markedly reduced in both FASSc (3.9%) and in CFA (4.5%) compared with control samples (20.4%, p < 0.0001). The percentage of tissue occupied by vessels decreased with increasing distance from alveoli in control samples but not in CFA or FASSc samples. Endothelial cell proliferation indices were increased in FASSc but not in CFA, compared with control samples (p = 0.006). In conclusion, there is net vascular ablation and redistribution of blood vessels in areas of interstitial thickening in both CFA and FASSc, which may contribute to gas exchange impairment.

　

Key Words: pulmonary fibrosis • scleroderma • pathologic neovascularization • endothelial proliferation

Cryptogenic fibrosing alveolitis (CFA), also known as idiopathic pulmonary fibrosis, a progressive and often fatal interstitial lung disease, is characterized by tissue damage and exuberant repair with an aberrant wound healing response, leading to severe disruption of pulmonary architecture (1). Angiogenesis plays an essential role in wound healing and may contribute to the fibroproliferation and extracellular matrix deposition observed in CFA. The observation of anastomoses between the pulmonary and systemic circulation in CFA has suggested a role for angiogenesis in this disease (2). Angiogenesis has been detected in a number of in vivo fibrotic models. Neovascularization associated with areas of pulmonary fibrosis has been reported in rat lung fibrosis induced by bleomycin (3). More recently, CFA lung tissue homogenates were found to be proangiogenic in an in vivo cornea assay (4). In a rat model, neutralization of proangiogenic chemokines attenuated both angiogenic and fibrotic responses to bleomycin (5).

Systemic sclerosis (SSc), a multisystem connective tissue disease characterized by vascular abnormalities, fibrosis, and humoral autoimmunity, is often associated with lung involvement. Vascular involvement in SSc is exemplified by the typical changes of nailfold capillaries, which include enlarged capillary loops and formation of new "bushy" capillaries in the early stages of SSc, vascular narrowing, and obliteration in more advanced disease (6). The most frequent clinical manifestation of lung involvement in SSc is fibrosing alveolitis, occurring in more than 70% of patients (7). Despite similarities between the two diseases, fibrosing alveolitis associated with systemic sclerosis (FASSc) is distinct from CFA in several regards including disease progression (8), histopathologic patterns of disease (9), bronchoalveolar cellularity (10), and morphologic appearance on computed tomographic scans of the chest (11). Furthermore, CFA is characterized by greater ventilation–perfusion mismatch than FASSc, after controlling for disease severity (12), suggesting differences in the extent of vascular involvement. However, differences in pulmonary vascular reorganization between CFA and FASSc have not been evaluated.

The aims of our study, therefore, were to evaluate the presence of angiogenesis and vascular distribution in the pulmonary interstitium in lung biopsies from patients with FASSc and CFA and to investigate the relationship between vascular distribution and gas exchange. To achieve these goals, we employed a double staining immunohistochemistry technique with quantification by digital image analysis to investigate the presence of angiogenesis and the distribution of blood vessels in lung biopsies of patients with CFA and FASSc.

	METHODS

TOP ABSTRACT METHODS RESULTS DISCUSSION REFERENCES

 
Patients/Control Subjects Biopsies
We studied surgical lung biopsies from eight patients with CFA, meeting the diagnostic criteria of the American Thoracic Society/European Respiratory Society Consensus Classification (13) and from nine with FASSc, meeting the criteria of the American College of Rheumatology (14) . Control biopsies consisted of normal lung from 12 patients undergoing cancer resection surgery. Surgical lung biopsies and resection specimens used as controls were submitted fresh to the laboratory and inflated with formalin in a comparable fashion; the former via injection into the parenchyma and the latter via the airways under pressure. Biopsies were classified (Dr. Nicholson) according to American Thoracic Society/European Respiratory Society consensus criteria (13): a pattern of "usual interstitial pneumonia" was observed for CFA biopsies and one of "nonspecific interstitial pneumonia" for all FASSc biopsies.

fig.ommitted TABLE 1. Clinical details of patients with cryptogenic fibrosing alveolitis and patients with fibrosing alveolitis associated with systemic sclerosis

 
Staining Procedures
Sections were stained sequentially for proliferating nuclei, endothelial cells, nonproliferating nuclei, and -smooth muscle actin. Proliferating nuclei were identified by staining for proliferating cell nuclear antigen (PCNA) with monoclonal antibody PC10, endothelium was stained using monoclonal antibody QBEND10 directed against CD34, nonproliferating nuclei with 1,2-diamidinophenylindole, as described previously (15), and smooth muscle with monoclonal antibody IA4 against -smooth muscle actin.

Quantification
Quantification was performed with the observer blinded to clinical details, using a Zeiss Axioskop photomicroscope (Zeiss, Welwyn Garden City, UK) with a x16 objective lens. Transmitted light and fluorescence images were captured using a 3-CCD (KY-F5JB) color video camera (JVC, Yokohama, Japan) and analyzed using a KS300 image analysis system (Imaging Associates, Thame, UK). Fields were chosen containing at least two airspaces interspersed by areas of significant interstitial fibrosis up to but not including areas of honeycomb/end-stage change.

The field area, including interstitial tissue and airspaces, was measured after bronchioles and large blood vessels staining positive for -smooth muscle actin had been excluded. Endothelial fractional area and tissue fractional area were defined as CD34-positive area and area occupied by tissue, respectively, divided by field area. Vascular density was defined as CD34-positive area divided by tissue area. Using computerized techniques (automatic delineation by iterative expansion of an airspace mask), measurements were made at 0 to 5 µm, 5 to 10 µm, 10 to 20 µm, 20 to 40 µm, 40 to 80 µm, 80 to 160 µm from the airspaces.

Endothelial cell proliferation indices were quantified in microscopic fields containing the highest proportion of PCNA-labeled nuclei (15). The proliferation index of endothelial cells was defined as the number of PCNA-positive endothelial cell nuclei divided by total endothelial cell nuclei. The epithelial cell proliferation index was defined as the number of positive cell nuclei per unit length of basement membrane (16).

Data Analysis
Unless stated otherwise, data are presented as medians and ranges. A p value less than or equal to 0.05 was considered statistically significant. Univariate relationships between arterial blood gases and percentage endothelial and tissue area included within 10 µm from airspaces, as well as between endothelial and epithelial proliferation indices, were quantified using Spearman's rank correlation. Differences between control, CFA, and FASSc biopsies were analyzed by repeated measures analysis of variance using a nested design, after log transformation of the data to satisfy assumptions of normality. The frequencies of PCNA-positive cells were compared by a logistic regression model (17) using a polytomous logistic regression with a robust variance estimate for clustered samples (18). Epithelial cell proliferation indices were compared using Wilcoxon's rank sum test. Analyses were performed using the statistical program Stata Release 7 (Stata Corp., College Station, TX).

	RESULTS

TOP ABSTRACT METHODS RESULTS DISCUSSION REFERENCES

 
Vessel Density and Distribution
The optimum number of fields per section and sections per biopsy were determined by evaluating CD34 staining in 10 fields each of 10 consecutive sections from one open lung biopsy. Six fields on one section per biopsy were sufficient to minimize the coefficient of variation and to give a SE ± 25% of the mean for vascular density.

Immunoreactivity for the endothelial antigen CD34 was identified beneath the epithelium lining alveolar air spaces in all samples  . A dense microvascular network was observed between alveolar epithelia in biopsies from control patients, whereas a sparser distribution of vessels was observed throughout the inter-alveolar fibrotic tissue in samples from patients with both FASSc and CFA. In control biopsies, staining for -smooth muscle actin was seen only in the media of large to small arteries down to a discontinuous layer in arterioles. However, in both CFA and FASSc, a complete layer of -smooth muscle actin-positive cells was present around many of the vessels in the areas of remodeling and fibrosis

fig.ommitted Figure 1. Endothelial cell proliferation and vascular distribution in lung biopsies from control patients (A) and from patients with FASSc (B) and (C). Sequential immunohistochemistry for the proliferation marker PCNA (black) and the endothelial cell antigen CD34 (red). Bold arrows: PCNA-immunoreactive endothelial cell nuclei. Fine arrows: PCNA-immunoreactive epithelial cells. Bars = 100 µm.

 
Fibroblastic foci were present in CFA (usual interstitial pneumonia) but not in FASSc (nonspecific interstitial pneumonia) biopsies, and were consistently characterized by an almost complete absence of vascularization  . However, immediately adjacent to and/or underlying some fibroblastic foci, an increase in the network of -smooth muscle actin-positive microvessels was observed .

fig.ommitted Figure 2. (A) and (B) Absence of vessels within fibroblastic foci (thick arrows). Immunohistochemistry for the endothelial cell antigen CD34, counterstained with haematoxylin. (C) Staining for -smooth muscle actin shows a layer of strongly actin-positive cells surrounding microvessels in areas of fibrovascular remodeling (thin arrows). Myofibroblasts within two fibroblastic foci (thick arrows) show weak cytoplasmic positivity.

 
The fraction of the area occupied by tissue rather than airspaces was higher in the FASSc and CFA groups compared with control groups (p < 0.0001) , as expected. Tissue areas were marginally higher in CFA than in FASSc (p = 0.12). By contrast, endothelial fractional area was reduced in patients with FASSc or CFA compared with control groups (p = 0.02). The percentage of tissue occupied by CD34-positive endothelium was markedly reduced in both FASSc (3.9%) and CFA (4.5%) biopsies compared with control biopsies (20.4%), with p values being less than 0.0001 and no significant differences between the two patient groups.

fig.ommitted TABLE 2. Area occupied by tissue and endothelial cells in control, fibrosing alveolitis associated with systemic sclerosis and cryptogenic fibrosing alveolitis biopsies

 
To investigate the distribution of microvessels within fibrotic lesions, we measured the amount of vascular tissue at increasing distances from the alveolar surface. As shown in  , in control biopsies, most of the CD34-positive endothelium was localized within 10 µm of the alveolar spaces, whereas, in samples from patients with CFA or FASSc, more vessels were identified farther than 10 µm from the nearest airspace. In control samples, tissue area and vascular densities decreased with increasing distance from the alveoli ( and ); in fibrotic tissue, vessel density differed little between tissue zones . Vessel distribution did not differ significantly between CFA and FASSc, although tissue and endothelial area in CFA tended to extend farther from the airspaces (p = 0.06 and 0.15, respectively).

fig.ommitted Figure 3. Altered distribution of blood vessels in CFA and FASSc biopsies compared with control biopsies. Points represent fraction (± range) of total endothelial area present in each distance from the airspaces (from within 5 µm up to beyond 80 µm from the alveolar spaces)

 

fig.ommitted TABLE 3. Distribution of tissue area at increasing distance from the alveolar surface

 

fig.ommitted TABLE 4. Distribution of vessel density at increasing distance from the alveolar surface

 
Endothelial Cell Proliferation Indices
Immunoreactivity for the proliferation marker PCNA was localized both to epithelial cells and to endothelial cells . Samples from control patients displayed only occasional PCNA-like immunoreactivity, whereas foci of epithelial and endothelial PCNA-immunoreactive nuclei were identified in samples from patients with both types of fibrosing alveolitis. The proliferation of endothelial cells was quantified by using the PCNA positivity of endothelial cell nuclei as an index. To use all individual cell observations, while adjusting for clustering of data within individual patients, a polytomous logistic regression model was constructed. Using this model, the odds ratio for endothelial PCNA positivity in comparison with control biopsies was 3.8 (95% confidence interval: 1.5–9.9, p = 0.006) in FASSc and 1.1 (95% confidence interval: 0.3–3.7, not significant) in CFA. As seen in  , the increased frequency of PCNA-positive endothelial cells in FASSc reflects the presence of a subgroup of patients with markedly increased proliferation indices, resulting from the presence of foci of endothelial proliferation. This observation was also made in a single CFA biopsy.

fig.ommitted   Figure 4. Increased frequency of endothelial cell proliferation indices (number of PCNA-positive endothelial cell nuclei/total number of endothelial nuclei) in FASSc biopsies compared with control biopsies (p = 0.006), without significant differences between CFA and control groups, using polytomous logistic regression.

 
In the disease groups, there was a correlation between endothelial and epithelial cell proliferation indices (r_s = 0.40, p = 0.04); however, no significant difference was detected in the epithelial proliferation index between the groups.

Arterial Gases in Relation to Morphologic Parameters
To verify whether the vascular redistribution observed in both CFA and FASSc lung biopsies could contribute to gas exchange impairment, the relationship between percentage of vessels included within 10 µm and alveolar–arterial oxygen gradient was evaluated. This parameter was chosen as an index of vascular derangement as the 10 µm range contains more than 80% of the CD34-positive endothelium in control biopsies.

Arterial gas data were available for 16 of 17 patients before biopsy (CFA n = 8, FASSc n = 8) but not for control subjects . Higher alveolar–arterial oxygen gradients were associated with lower percentage endothelial areas within 10 µm from airspaces (r_s= -0.51, p = 0.04)  as well as with the percentage tissue area within the same distance (r_s= -0.59, p = 0.02).

fig.ommitted   Figure 5. Significant correlation between higher alveolar–arterial oxygen gradients and lower percentage endothelial area included within 10 µm from the airspaces (endothelial area within 10 µm/total endothelial area) in 16 out of the 17 patients with fibrosing alveolitis (8 = CFA, open circles; 8 = FASSc, filled in circles) (p = 0.04).

 

	DISCUSSION

TOP ABSTRACT METHODS RESULTS DISCUSSION REFERENCES

 
The purpose of this study was to evaluate vascularity in the pulmonary interstitium of patients with CFA or FASSc. We found evidence of net interstitial vascular regression in both CFA and FASSc lung biopsies. Vascular density as well as total vascular area were reduced compared with control lungs in both types of fibrosing alveolitis. This is in agreement with the report of ablation of vessels in areas of honeycomb lung regardless of the cause of the pulmonary fibrosis (19) as well as the finding of a significant reduction in the mean capillary surface area in a study of nine patients with CFA compared with control subjects (20).

In addition to this reduction in vascular density, we report the novel finding of a substantial redistribution of microvessels within the pulmonary interstitium in pulmonary fibrosis. Whereas normal lungs were characterized by a sharp reduction in vascular density beyond the distance of 10 µm from alveolar spaces, a complete loss of this gradient was observed in pulmonary fibrosis, with similar vascular densities observed across the whole range of distances from airspaces.

We found no significant difference in overall vascular density or distribution between CFA and FASSc. We cannot however rule out the possibility that the two diseases differ in the degree of involvement of medium- or largesized vessels, as these were not evaluated in the present study.

Interestingly, in both CFA and FASSc lung biopsies, a layer of -smooth muscle actin staining was seen as part of the walls of many small vessels in areas of fibrosis. These vessels differed morphologically from normal arteries and arterioles, including lack of elastin layer, and most likely represent the vascular remodeling that accompanies disease progression. Clusters of such vessels were observed immediately adjacent to fibroblastic foci, suggesting that vascular remodeling is particularly active in these areas. By contrast, we observed an absence of vessels within fibroblastic foci, in agreement with the previously reported finding that capillarization is less frequent in fibroblastic foci of usual interstitial pneumonia than in the intraluminal fibromyxoid lesions of bronchiolitis obliterans organizing pneumonia (21). The link between fibroblastic foci and vascular remodeling merits further evaluation.

To our knowledge, this is the first study to assess the level of endothelial cell proliferation in pulmonary fibrosis. Control lungs were characterized by low levels of endothelial cell proliferation, similar to those found in other tissues such as the noninflamed human synovia (15). In FASSc, a subgroup of patients had foci of endothelial cell proliferation that were seen only in one of the CFA patients and in none of the control lungs.

Our findings may denote differences in degrees of angiogenesis between CFA and FASSc. Levels of circulating angiogenic factors such as vascular endothelial growth factor are higher in FASSc patients than in control subjects (22), and cutaneous new vessel formation is a feature of SSc (23). Sera from patients with SSc can enhance the production of proangiogenic cytokines by normal human peripheral blood mononuclear cells (24). However, the marked similarity in vascular distribution and density between CFA and FASSc, despite differences in the severity of fibrosis, suggests that new vessel formation is unlikely to differ radically between the two diseases.

The increased rate of endothelial cell proliferation observed in FASSc could reflect an earlier stage of fibrosis in this group or a difference in the involvement of endothelial cells in FASSc, possibly due to specific immunologic insults (25) or both. Recently, serum immunoglobulin G autoantibodies found in more than 90% of SSc patients have been observed to induce endothelial cell apoptosis through specific interaction with the cell surface integrin-NAG-2 protein (26). The increased endothelial proliferation observed in FASSc could be the result of a response to such an ongoing insult; in this case it would not necessarily result in new vessel formation but simply be an expression of increased vascular turnover.

The paucity of endothelial cell proliferative foci in CFA in the present study may reflect the inclusion of patients with advanced disease. Due to intrinsic differences in the course, evolution, and possibly in the pathogenesis of the two diseases, patients with FASSc are typically detected and, if judged necessary, undergo a biopsy, at an earlier stage than CFA. As expected, FASSc patients were found to have less severe disease, as judged by pulmonary function indices. In normal wound healing, angiogenesis occurs early and is followed by vascular regression, coinciding with the evolution of granulation tissue into scar tissue. Peripheral biopsies in severely fibrotic CFA may not capture evidence of earlier pathogenic events, including new vessel formation.

Our data indicate that in pulmonary fibrosis there is both ablation of vessels close to alveolar spaces and a substantial vascular redistribution leading to a greater proportion of vessels removed from areas of gas exchange. Even though our study, being a morphologic one, is not suited to specifically establish the relative contributions of the various mechanisms to gas exchange impairment in pulmonary fibrosis, it suggests that both vascular ablation and increased distance to be traveled by oxygen may have roles in the generation of hypoxemia in these patients. This is supported by the finding of an inverse correlation between the alveolar–arterial oxygen difference and the percentage of vessels within 10 µm from the airspaces.

In conclusion, both CFA and FASSc are characterized by substantial vascular redistribution, with a shift of interstitial vessels away from airspaces. Vascular remodeling was observed in the fibrotic areas of both diseases, with focal angiogenesis noted in a subgroup of SSc patients. The relevance of these findings to the pathophysiology of fibrosing alveolitis requires further investigation.

Received in original form February 21, 2002; accepted in final form October 2, 2002

	REFERENCES

TOP ABSTRACT METHODS RESULTS DISCUSSION REFERENCES

 

1. Gross TJ, Hunninghake GW. Idiopathic pulmonary fibrosis. N Engl J Med 2001;345:517–525.

2. Turner-Warwick M. Precapillary systemic–pulmonary anastomoses. Thorax 1963;18:225–237.

3. Peao MND, Aguas AP, DeSa CM, Grande NR. Neoformation of blood vessels in association with rat lung fibrosis induced by bleomycin. Anat Rec 1994;238:57–67.

4. Keane MP, Arenberg DA, Lynch JP, Whyte RY, Iannettoni MD, Burdick MD, Wilke CA, Morris SB, Glass MC, DiGiovine B, et al. The CXC chemokines, IL-8 and IP-10, regulate angiogenic activity in idiopathic pulmonary fibrosis. J Immunol 1997;159:1437–1443.

5. Keane MP, Belperio JA, Moore TA, Moore BB, Arenberg DA, Smith RE, Burdick MD, Kunkel SL, Strieter RM. IFN-gamma-inducible protein-10 attenuates bleomycin-induced pulmonary fibrosis via inhibition of angiogenesis. J Immunol 1999;163:5686–5692.

6. Maricq HR. Comparison of quantitative and semiquantitative estimates of nailfold capillary abnormalities in scleroderma spectrum disorders. Microvasc Res 1986;32:271–276.

7. Black CM, du Bois R. Organ involvement: pulmonary. In: Clements PJ, Furst DE, editors. Systemic sclerosis. Baltimore, MD: Williams and Wilkins; 1996. p. 299–331.

8. Wells AU, Cullinan P, Hansell DM, Rubens MB, Black CM, Newman-Taylor AJ, Du Bois RM. Fibrosing alveolitis associated with systemic sclerosis has a better prognosis than lone cryptogenic fibrosing alveolitis. Am J Respir Crit Care Med 1994;149:1583–1590.

9. Bouros D, Wells AU, Nicholson AG, Colby TV, Polychronopoulos V, Pantelidis P, Haslam PL, Vassilakis DA, Black CM, du Bois RM. Histopathologic subsets of fibrosing alveolitis in patients with systemic sclerosis and their relationship to outcome. Am J Respir Crit Care Med 2002;165:1581–1586.

10. Wells AU, Hansell DM, Haslam PL, Rubens MB, Cailes J, Black CM, du Bois RM. Bronchoalveolar lavage cellularity: lone cryptogenic fibrosing alveolitis compared with the fibrosing alveolitis of systemic sclerosis. Am J Respir Crit Care Med 1998;157:1474–1482.

11. Chan TY, Hansell DM, Rubens MB, du Bois RM, Wells AU. Cryptogenic fibrosing alveolitis and the fibrosing alveolitis of systemic sclerosis: morphological differences on computed tomographic scans. Thorax 1997;52:265–270.

12. Wells AU, Hansell DM, Rubens MB, Cailes JB, Black CM, du Bois RM. Functional impairment in lone cryptogenic fibrosing alveolitis and fibrosing alveolitis associated with systemic sclerosis: a comparison. Am J Respir Crit Care Med 1997;155:1657–1664.

13. American Thoracic Society/European Respiratory Society International Multidisciplinary Consensus Classification of the Idiopathic Interstitial Pneumonias. This Joint Statement of the American Thoracic Society (ATS), and the European Respiratory Society (ERS) was adopted by the ATS Board of Directors, June 2001 and by The ERS Executive Committee, June 2001. Am J Respir Crit Care Med 2002;165:277–304.

14. Subcommittee for scleroderma criteria of the American Rheumatism Association Diagnostic and Therapeutic Criteria Committee. Preliminary criteria for the classification of systemic sclerosis (scleroderma). Arthritis Rheum 1980;23:581–590.

15. Walsh DA, Wade M, Mapp PI, Blake DR. Focally regulated endothelial proliferation and cell death in human synovium. Am J Pathol 1998;152:691–702.

16. Salmon M, Liu YC, Mak JC, Rousell J, Huang TJ, Hisada T, Nicklin PL, Fan Chung K. Contribution of upregulated airway endothelin-1 expression to airway smoothmuscle and epithelial cell DNA synthesis after repeated allergen exposure of sensitized Brown-Norway rats. Am J Respir Cell Mol Biol 2000;23:618–625.

17. Bonassi S, Fontana V, Ceppi M, Barale R, Biggeri A. Analysis of correlated data in human biomonitoring studies: the case of high sister chromatide exchange frequency cells. Mutat Res 1999;438:13–21.

18. Rogers WH. Regression standard errors in clustered samples. Stata Tech Bull 1993;13:19–23.

19. Heath D, Gillund TD, Kay JM, Hawkins CF. Pulmonary vascular disease in honeycomb lung. J Pathol Bacteriol 1968;95:423–430.

20. Cassan SM, Divertie MB, Brown AL. Fine structural morphometry on biopsy specimens of human lung. 2: diffuse idiopathic pulmonary fibrosis. Chest 1974;65:275–278.

21. Lappi-Blanco E, Kaarteenaho-Wiik R, Soini Y, Risteli J, Paakko P. Intraluminal fibromyxoid lesions in bronchiolitis obliterans organizing pneumonia are highly capillarized. Hum Pathol 1999;30:1192–1196.

22. Kikuchi K, Kubo M, Kadono T, Yazawa N, Ihn H, Tamaki K. Serum concentrations of vascular endothelial growth factor in collagen disease. Br J Dermatol 1998;139:1049–1051.

23. Maricq HR, Harper FE, Khan MM, Tan EM, LeRoy EC. Microvascular abnormalities as possible predictors of disease subsets in Raynaud phenomenon and early connective tissue disease. Clin Exp Rheumatol 1983;1:195–205.

24. Majewski S, Skopinska-Rozewska E, Jablonska S, Polakowski I, Pawinska M, Marczak M, Szmurlo A. Modulatory effect of sera from scleroderma patients on lymphocyte-induced angiogenesis. Arthritis Rheum 1985;28:1133–1139.

25. Holt CM, Lindsey N, Moult J, Malia RG, Greaves M, Hume A, Rowell NR, Hughes P. Antibody-dependent cellular cytotoxicity of vascular endothelium: characterization and pathogenic associations in systemic sclerosis. Clin Exp Immunol 1989;78:359–365.

26. Lunardi C, Bason C, Navone R, Millo E, Damonte G, Corrocher R, Puccetti A. Systemic sclerosis immunoglobulin G autoantibodies bind the human cytomegalovirus late protein UL94 and induce apoptosis in human endothelial cells. Nat Med 2000;6:1183–1186.