Structural Features of Interstitial Lung Disease in Systemic Sclerosis 1 , 2
NICHOLAS K. HARRISON, ALLEN R. MYERS, BRYAN CORRIN, GERALDINE SOOSAY, ANN DEWAR, CAROL M. BLACK, ROLAND M. DU BOIS, and MARGARET TURNER-WARWICK
Introduction
Patients with systemic sclerosis frequently develop interstitial lung disease, although pulmonary abnormalities are seldom apparent when scleroderma is first diagnosed. The investigation of patients with scleroderma should therefore allow identification of interstitial lung disease at an early stage in its natural history when morphologic features may give important clues to pathogenesis. This is in contrast to patients with "lone" cryptogenic fibrosing alveolitis, who necessarily present when lung disease givesrise to respiratory symptoms. Most reports of lung pathology in systemic sclerosis describe abnormalities found at autopsy. These include extensive interstitial fibrosis with cysticchanges, bronchiolectasis, and pleural thickening (1-5). Inflammatory cellsare also observed (6)but are often thought to represent terminal pathologic events rather than the underlying primary condition (7). These postmortem findings obviously represent advanced disease and add little to our understanding of pathogenesis. Detailed descriptions of antemortem lung pathology in systemic sclerosis have been confined to small numbers of patients. Alveolitis has been observed in these studies, but it is uncertain whether the inflammatory cell infiltrate precedes, accompanies, or follows fibrosis of the alveolar walls (8). The recognition that lung disease in systemic sclerosis resembles that of lone cryptogenic fibrosing alveolitis has led some workers to include systemic sclerosis patients in clinical studies of fibrosing alveolitis (9), but there has never been a formal comparison of lung histology between the two conditions. The ability of modern imaging techniques and lung function tests to detect early pulmonary abnormalities in systemic sclerosis provides an opportunity to examine lung structure at a time when morphologic features should giveinsight 706
SUMMARY Most studies of lung histology in systemic sclerosis have been based on autopsy specimens and consequently emphasize end-stage fibrotic disease. Although occasional pathologic descriptions of open-lung biopsies have recognized the presence of Inflammatory cells, suggesting a similarity to "lone" cryptogenic flbrosing alveolltls, the two conditions have never been formally compared. In this study we describe the morphologic features of 49 open-lung biopsies from 34 systemic sclerosis patients with Interstitial lung disease, many of whom had their lung disease diag· nosed at an early stage. None had pulmonary hypertension. Examination of lung tissue by light microscopy showed the earliest changes to include patchy lymphocyte and plasma cell Infiltration of the alveolar walls, Interstitial fibrosis, and increased macrophages but only occasional polymorphonuclear cells and lymphocytes In the alveolar spaces. Alveolltis was not observed without fibrosis. Comparison of 22 biopsies with a similar number from patients with lone cryptogenic fibrosing alveolitls, matched for age and sex, revealed no qualitative or quantitative differences, other than a higher prevalence of focal lymphoid hyperplasia (follicular bronchiolitis) In the systemic sclerosis patients than in the lone cryptogenic fibroslng alveolitis controls (23 and 5%, respectively). There was an Inverse correlation between the extent of interstitial Inflammation and patients' age (p < 0.05), disease duration (p < 0.05), and age at onset of systemic sclerosis (p < 0.01).There was also an inverse correlation between OLeo and Interstitial fibrosis (p < 0.01)and loss of lung architecture (p < 0.05).Ultrastructural studies of eight systemic sclerosis biopsies showed evidence of endothelial and epithelial injury together with interstitial edema and excess collagen deposition. Occasional mast cells were observed, often in close contact with Interstitial fibroblasts, but there were no tubuloretlcular structures or evidence of Immune complexes. Similar changes were also found in a further three biopsies that appeared normal by light microscopy. These results suggest that in systemic sclerosis pulmonary endothelial and/or epithelial injury may precede Inflammation and fibrosis. AM REV RESPIR DIS 1991; 144:706-713
into pathogenesis, and if the lung disease of systemicsclerosisis truly indistinguishable from that of lone cryptogenic fibrosing alveolitis, this should apply to the latter condition as much as to the first. To date there have been no studies of lung ultrastructure in systemicsclerosis,except for a solitary case report describing thickening of the common endothelial/epithelial basement membrane on the "thin" side of the alveolar wall (10). In the present study we describe the pathologic abnormalities in 49 open-lung biopsies from 34 systemic sclerosis patients with interstitial lung disease, and compare them with biopsies from a group of age- and sex-matched patients with lone cryptogenic fibrosing alveolitis. We also examine lung ultrastructure in systemic sclerosis with particular reference to biopsies showing minimal change in the belief that such changes may be relevant to early pathogenetic mechanisms.
Methods
Patients We studied 34 patients who fulfilled the American Rheumatism Association preliminary criteria for the diagnosis of systemicsclerosis (11) and who had undergone open-lung biopsy at the Brompton Hospital for diagnosis and staging of interstitial lung disease.
(Received in original form October 5, 1990 and in revised form April 3, 1991) 1 From the Departments of Thoracic Medicine and Lung Pathology, National Heart and Lung Institute, Brompton Hospital, and the Department of Rheumatology, Royal Free Hospital, London, United Kingdom, and the Department of Medicine, Temple University School of Medicine, Philadelphia, Pennsylvania. 2 Correspondence and requestsfor reprints should be addressed to Dr. Nicholas K. Harrison, Biochemistry Unit, Department of Thoracic Medicine, National Heart and Lung Institute, Emmanuel Kaye Building, Manresa Road, London SW3 6LR, UK.
707
MORPHOLOGY OF INTERSTITIAL WNG DISEASE IN SYSTEMIC SCLEROSIS
There were 26 women and 8 men; the mean (range) age was 45 (22 to 67) yr. One patient was black, two were from the Indian subcontinent, and 31 were white. A total of 20 had diffuse cutaneous scleroderma, 10 had limited cutaneous scleroderma, and 4 had CREST syndrome. Mean duration of the scleroderma was 7.7 (1to 31)yr. All patients had either abnormal chest radiographs or the characteristic features of fibrosing alveolitis on thinsection (3 mm) computed tomographic (CT) scan (12).Lung function tests were abnormal in many but not all patients. Mean values (0/0 of predicted normal) were FVC 72% (23 to 123%); TLC, 73% (31 to 112%); DLCO, 54% (17to 91%); and carbon monoxide gas transfer coefficient Kco, 83% (46 to 117%). None of the patients had clinical evidence of pulmonary hypertension.
Lung Biopsies A lower lobe biopsy was available for examination in all 34 patients. In 15 patients a second site, usually the right middle lobe, was also biopsied, giving a total of 49 biopsies available for analysis and enabling a comparison to be made between different lobes in the same individual. In 22 patients, lower lobe biopsies were compared with those from a similar number of age- and sex-matched patients with lone cryptogenic fibrosing alveolitis who also underwent open-lung biopsy at the Brompton Hospital for diagnosis and staging. Biopsies were undertaken when there was evidence of interstitial lung disease on CT scan. The purpose of open-lung biopsy was to confirm the diagnosis of interstitial lung disease and allow the clinician to plan therapy on the principle that commencing this as soon as possible may limit disease progression (9, 13, 14). All patients gave informed written consent to lung biopsy. Light Microscopy Paraffin-embedded sections of formalin-fixed tissue were stained with hematoxylin and eosin and by the combined elastin-van Gieson method. Biopsieswererandomized and scored by a pathologist who was unaware of the disease group, using a semiquantitative fourpoint scoring system for each of the following: interstitial inflammation, interstitial fibrosis, intraalveolar inflammation, type II epithelial cell hyperplasia, bronchiolization of alveoli, loss of alveolar architecture, and vascular abnormalities. Electron Microscopy Eleven lung biopsies from seven systemic sclerosis patients with only mild histologic abnormalities were selected for electron microscopy. In all biopsies, areas showing the least abnormalities at the light microscopic level werechosen. In three casesin whom both middle and lower lobes were examined the former appeared normal on light microscopy. In all cases small pieces of fresh lung tissue had been diced and fixed without delay in sodium cacodylate-buffered 2.5% glutar-
aldehyde for 4 to 18h, rinsed in buffer, postfixed in osmium tetroxide, and processed to Araldite", After selecting suitable areas from l-um-thick sections, ultrathin sections were cut, stained with uranyl acetate and lead citrate, and examined by transmission electron microscopy. Middle lobe biopsies from seven patients were scored, using a semiquantitative fourpoint scoring system, for the following abnormalities: Capillaries Endothelial swelling Thickening or reduplication of basement membrane Tubuloreticular structures Epithelium Type I cell swelling Type II cell hyperplasia Bare basement membrane Interstitium Edema Collagen Elastin
Data Handling and Statistics Parametric data are described as mean (range) and were compared by t test. Biopsy scores for different lobes in the same subject were compared by Wilcoxon test. Linear regression analysis was performed to assess correlations between pathologic scores and clinical data. The severity of pathologic abnormalities in systemic sclerosis and cryptogenic fibrosing alveolitis were compared for each of the seven variables by chi-square analysis. When possible, 4 x 2 tables were used, but when numbers were small groups were combined into 3 x 2 or 2 x 2 tables.
Results
Clinical Evaluation Of the 34 systemic sclerosis patients, seven had a normal chest radiograph and of these two had normal lung function tests (lung volumes> 80% of predicted, carbon monoxide diffusing capacity > 750/0 of predicted). CT scans were performed in 27 patients, and all showed abnormalities in the lower lobes. In eight patients, CT scans of the middle or upper lobes were normal, yet biopsies of these lobes showed histologic evidence of fibrosing alveolitis. In a further three patients, both CT scan and histologic examination of the right middle lobe were substantially normal, yet abnormalities were detected at that site by electron microscopy. These are described subsequently. There was no significant pathologic difference between the patients with abnormal chest radiographs and those with normal chest radiograph but thought to have earlier disease in whom
lung biopsy was justified on the basis of an abnormal CT scan. Bronchoalveolar lavage cell profiles were abnormal in all patients. These findings have been reported elsewhere (12, 15).
Histopathology Lower lobe biopsies from all patients with systemic sclerosis generally showed some degree of both inflammation and fibrosis. Appearances varied in extent from almost normal lung to end-stage disease with severe fibrosis and complete loss of lung structure (figures Ia to d). Intraalveolar inflammation consisted of increased numbers of macrophages together with occasional neutrophils, lymphocytes, and eosinophils. Lymphocytes and plasma cells predominated in the interstitium, and occasional small lymphoid aggregates were observed adjacent to bronchioles (follicular bronchiolitis) or the pleura (figure Ie). Middle or upper lobe biopsies were similar in appearance to those of the lower lobes. Abnormalities tended to be more marked in the lower lobes, but this difference was not statistically significant for any variable. In three patients the middle lobe biopsies were considered substantially normal: these biopsies were included in the electron microscope study. Pleural abnormalities were rare, consisting of a fibrinous pleurisy in two patients, diffuse pleural thickening in a third, and a collection of necrotizing granulomata in a fourth. These changes werethought to more likelyrelate to coincidental or past infection rather than to the systemic sclerosis disease process. Pulmonary arteries showed media hypertrophy and fibrous thickening of the intima. This was particularly evident where the parenchymal changes were severe (figure If): no vascular abnormalities were observed in areas where there was no parenchymal lung disease. Comparison with Cryptogenic Fibrosing Alveolitis Open-lung biopsies from 22 patients with systemicsclerosiswerecompared with the same number from a group of age- and sex-matched patients with lone cryptogenic fibrosing alveolitis. The clinical characteristics of the two groups are summarized in table 1. These patients were not matched for duration of pulmonary disease because in many systemicsclerosis patients CT abnormalities wereidentified before there were respiratory symptoms or abnormalities on chest radiograph.
708
HARRISON, MYERS, CORRIN, ET AL.
709
MORPHOLOGY OF INTERSTITIAL WNG DISEASE IN SYSTEMIC SCLEROSIS
With the exception of the pulmonary lymphoid tissue, there were no qualitative differences in the lung parenchyma between systemic sclerosisand lone cryptogenic fibrosing alveolitis, and when they were scored, chi-square analysis showed there were no significant quantitative differences in the extent of parenchymal or vascular disease between the two groups (table 2). Pleural abnormalities were not observed in any patient with lone cryptogenic fibrosing alveolitis and in only four of the 49 systemic sclerosis biopsies. Focal peribronchial lymphoid hyperplasia was seen in seven systemic sclerosis patients (five of those with age- and sex-matched controls) and two control patients with lone cryptogenic fibrosing alveolitis. In two systemic sclerosis patients and one fibrosing alveolitis control patient the lymphoid hyperplasia included the formation of follicles with germinal centers (follicular bronchiolitis). In addition, pleural lymph nodes were identified in two systemic sclerosis patients.
Fig. 1. Open-lung biopsies stained with hematoxylin and eosin showing the range of pulmonary abnormalities in systemic sclerosis. (a) Mild fibrosis and Iymphoplasmacytic infiltration of the alveolar walls (magnification: x130). (b) Moderate inflammatory changes consisting predominantly of lymphocyte and plasma cell infiltration of the alveolar walls and macrophage accumulation in the airspaces. Alveolar walls also show moderate epithelial Type II cell hyperplasia and fibrosis (magnification: x130). (c) Severe pulmonary fibrosis with loss of alveolar architecture (magnification: x80). (d) End-stage lung disease with severe fibrosis and complete loss of normal lung structure yet continuing active inflammation (magnification: x80). (e) Follicular bronchiolitis (magnification: x110). (f) Pulmonary artery showing medial hypertrophy and intimal fibrosis. These changes were observed only in areas of lung showing moderate or severe parenchymal disease. (Hematoxylin and eosin; magnification: x 240).
TABLE 1 CLINICAL CHARACTERISTICS OF 22 PATIENTS WITH SYSTEMIC SCLEROSIS AND 22 AGE- AND SEX-MATCHED PATIENTS WITH "LONE" CRYPTOGENIC FIBROSING ALVEOLITIS (CFA)
Systemic sclerosis (n = 22) Lone CFA (n = 22)
Lung Function (% of Predicted)
Age (yr)
Sex (F:M)
CXR (N/Ab)
FVC
TLC
46 ± 3
14:8
4/18
75 ± 6
75 ± 5
52 ± 4
79 ± 4
47 ± 3
14:8
0/22
69 ± 5
72 ± 5
37 ± 3"
66 ± 5
Definition of abbreviations: CXR = chest radiograph; N/Ab= normallabnormal.· • p < 0.05.
DLCO
Kco
Ultrastructural Abnormalities Ultrastructural abnormalities of endothelium, epithelium, and interstitium weredetected in all II biopsies examined, including the three that appeared substantially normal by light microscopy. In those three biopsies, endothelial cells were often swollen and electron lucent (figures 2 and 3) or, less commonly, pyknotic. In places the pyknotic cells had the appearance of being undermined and replaced by adjacent healthy endothelial cells (figure 4). Endothelial abnormalities were sometimes associated with thickening of the endothelial basement membrane (figure 4). Epithelial abnormalities consisted of TABLE 2 COMPARISON OF SCORES IN 22 PATIENTS AND CRYPTOGENIC
Interstitial inflammation Alveolar inflammation Interstitial fibrosis Type IJ cell hyperplasia Bronchiolization of alveoli Loss of lung architecture Blood vessels
MEAN LUNG PATHOLOGY SYSTEMIC SCLEROSIS CONTROLS WITH LONE FIBROSING ALVEOLITIS
Systemic Sclerosis
Lone CFA
Chi-square p Value
1.4
1.5
0.3
1.4
1.4
0.5
2.1
2.1
0.2
1.5
1.2
0.1
1.3
1.2
0.2
1.6
1.3
0.2
1.8
1.4
0.2
HARRISON, MYERS, CORRIN, ET AL.
710
mune complexes were observed, nor was there evidence of the tubuloreticular structures described in interstitial lung disease associated with other collagen vascular diseases and in viral pneumonias (16).
Figure 2. Electron micrograph of alveolar wall showing patchy cytoplasmic swelling of endothelial (star) and epithelial (arrow) cells. (Uranyl acetate and lead citrate; bar = 2 um.)
Clinicopathologic Correlations For the changes observed by light microscopy,regression analysis indicated a weak inverse correlation between the extent of interstitial inflammation and patients' age (r = - 0.36, p < 0.05), disease duration (r = - 0.34, p < 0.05), and age at onset of systemic sclerosis (r = - 0.48, p < 0.01). There was also a significant negative correlation between DLCO and interstitial fibrosis (r = - 0.46, p < 0.01) and loss of lung architecture (r = - 0.4, p < 0.05). For changes observed by electron microscopy, regression analysis did not show any significant correlations with clinical data. Discussion
patchy epithelial cell swelling (figure 2), focal loss exposing bare epithelial basement membranes (figure 5), and focal Type II pneumocyte proliferation. There was no clear difference in severity between endothelial and epithelial damage. Interstitial abnormalities consisted of areas of electron-lucent edema (figure 5) and increased interstitial collagen (figure 6) and elastin, largely confined to the
thick side of the air/blood barrier. Occasional mast cells were observed, often in intimate contact with interstitial fibroblasts (figure 3). Comparison of these appearances with the eight systemic sclerosis biopsies that had clear evidence of fibrosing alveolitis by light microscopy showed less edema and more fibrosis in the latter. No electron-dense deposits suggestive of im-
Figure 3. Capillary showing cytoplasmic endothelial swelling (star). A mast cell is seen in the interstitium in close contact with thin cytoplasmic processes of a fibroblast. (Uranyl acetate and lead citrate; bar = 2 um.)
All previous studies of lung pathology in systemic sclerosis involving significant numbers of patients used material obtained at autopsy. Consequently they emphasized advanced fibrotic disease and failed to identify early structural abnormalities. In contrast, the present study applied modern imaging techniques to detect interstitial lung disease at an early stage and therefore includes some asymptomatic patients who had normal chest radiographs and normal lung function tests. This provided us with the opportunity to examine the early structural abnormalities of the lung in systemic sclerosis. A striking observation was the extent of the ultrastructural changes. All 46 biopsies in which abnormalities were apparent by light microscopy showedpatchy inflammatory cellinfiltration of the interstitium and alveolar spaces accompanying fibrosis of the alveolar walls. This suggests that these processes occur together from the earliest stages of the disease. Interstitial infiltration by lymphocytes and plasma cells was one of the earliest abnormalities apparent by light microscopy. This contrasts with studies of early lung disease by bronchoalveolar lavage, which show increased proportions of neutrophils (15, 17). This discrepancy between histology and bronchoalveolar lavage is well known in lone cryptogenic fibrosing alveolitis. It is thought to reflect differential sampling of inflammatory cellsby the lavage procedure (18). Whether the interstitial lymphocytes are promoting or inhibiting fibrosis, or both, re-
MORPHOLOGY OF INTERSTITIAL WNG DISEASE IN SYSTEMIC SCLEROSIS
Figure 4. Pyknotic capillary endothelial cells are undermined by adjacent healthy cells, and there is thickening of the endothelial basement membrane (arrows). (Uranyl acetate and lead citrate; bar = 2 urn.)
mains uncertain, but might be investigated in the future by localization of lymphocyte subtypes and identification of their lymphokines. The parenchymal abnormalities weobserved in lung biopsies from patients with systemic sclerosis wereindistinguishable from those of patients with lone cryptogenic fibrosing alveolitis, with the excep-
tion of the focal lymphoid hyperplasia, which was more common in the systemic sclerosis patients. Focal lymphoid hyperplasia (follicular bronchiolitis) is also a feature of rheumatoid lung disease (19). We do not regard it as a component of the intestitiallung disease but as an independent feature of systemiccollagen vascular disease, which, when pres-
Figure 5. Alveolar epithelial cell loss has led to extensive denudation of the basement membrane (arrows). Residual alveolar epithelium is pyknotic (arrowheads). Interstitial edema is also evident (star). (Uranyl acetate and lead citrate; bar = 2 urn.)
711
ent, is a useful pointer to the possibility that fibrosing alveolitis is only one manifestation of such systemic disease. Apart from lymphoid hyperplasia, the interstitial lung disease of systemic sclerosis is identical to that seen in lone cryptogenic fibrosing alveolitis, supporting the concept that the interstitial lung disease of systemic sclerosis provides an opportunity to study the pathogenesis of fibrosing alveolitis at an early stage,often before the onset of respiratory symptoms. However, despite their pathologic identity, there is preliminary evidence to suggest that systemic sclerosis patients with fibrosing alveolitis have a more favorable prognosis than those with lone cryptogenic fibrosing alveolitis (20). If this is true, it is interesting to speculate on factors that may be modulating the disease in systemicsclerosis. It is known that men with cryptogenic fibrosing alveolitis have a worse prognosis than women (21), and the preponderance of women with systemic sclerosis may be one such factor. Other possibilities are that treatment for scleroderma, commenced before the development of interstitial lung disease, favorably affects the course of the latter or that inhibitors to collagen production are present in systemic sclerosis patients but not in those with lone cryptogenic fibrosing alveolitis. The ultrastructural abnormalities we observed in this study are also similar to those described previously in patients with lone cryptogenic fibrosing alveolitis (22). These have not been described in systemic sclerosis, however, and this is the first report of such changes in biopsies that appear normal by light microscopy. These early changes are important when considering the pathogenesis of fibrosing alveolitis. In particular, the lack of inflammatory cell infiltrate does not favor alveolitis preceding structural damage, although it is possible that such mediators as oxidants and fibroblast growth factors released from inflammatory cells in adjacent parts of the lung could be responsible for the damage and fibrosis we observed. Alternatively, if alveolitis precedes structural damage, our findings suggest this may be a "hit-andrun" phenomenon, the inflammation abating but leaving residual injury. Although we paid particular attention to tissue showing minimal changes, we were not able to determine whether endothelial or epithelial abnormalities predominated. Endothelial damage may appear less marked because of the ability of regenerating endothelial cellsto spread rapidly beneath adjacent damaged cells
712
HARRISON, MYERS, CORRIN,
Figure 6. The alveolar interstitium contains excess collagen. (Uranyl acetate and lead citrate; bar = 2 urn.)
(23, 24). In contrast, replacement of lost Type I epithelial cells requires the more prolonged process of proliferation and differentiation of Type II pneumocytes (25, 26). Injury to endothelial cells by circulating agents has been proposed as a pathogenetic mechanism in skin and other organs in systemic sclerosis (27). Injury to capillary endothelial cells is likely to account for the interstitial edema we observed. That increased permeability of the endothelial/epithelial barrier occurs early in the pathogenesis of systemic sclerosis lung disease is supported by observations that protein concentration in bronchoalveolar lavage fluid is increased (15) and clearance of radiolabeled, low-molecular-weight solutes is more rapid in these patients (12). Whether vascular leakage alone is sufficient to stimulate interstitial fibroblasts is uncertain, but it is increasingly recognized that circulating proteins can act as potent fibroblast mitogens (28, 29). Furthermore, other conditions associated with profound but transient vascular leakage, such as the adult respiratory distress syndrome, frequently progress to pulmonary fibrosis (30). Similarly, such conditions as mitral stenosis and pulmonary venoocelusive disease that are characterized by chronic edema of the lungs may also result in pulmonary fibrosis (31). We failed to identify any tubuloreticular structures in association with the endothelial abnormalities. These structures, which are composed of phospholipid and glycoprotein (32), were observed within damaged pulmonary endothelial cells
and peripheral blood lymphocytes in a study of patients with viral pneumonia and certain collagen vascular diseases (17), but lung tissue from patients with systemic sclerosis was not examined. Furthermore, the same study failed to identify tubuloreticular structures in lung biopsies from 30 patients with idiopathic pulmonary fibrosis (cryptogenic fibrosing alveolitis). It has been demonstrated that interferons can induce the formation of tubuloreticular structures within cells (33), and this has been proposed as a possible mechanism for their occurrence in viral pneumonias (17). The apparent absence of these structures in both systemic sclerosis and cryptogenic fibrosing alveolitis argues against a viral cause for these conditions. We also failed to identify any electrondense deposits suggestive of immune complexes. This has been a consistent finding in lone cryptogenic fibrosing alveolitis (22, 34-36), despite a few reports of their detection in early cases of this disease by immunofluorescent microscopy (37, 38). We did observe occasional mast cells, as described in other fibrotic lung diseases (39). This finding is consistent with similar observations in early skin lesions in systemic sclerosis (40), and it has been proposed that mast cells may represent a link between activated endothelial cells and skin fibroblasts in this condition (41). It is known that alveolar macrophages possess H 2 receptors and can release a factor that causes human lung mast cells to release histamine (42). Elevated levels
er AL.
of histamine have been observed in bronchoalveolar lavage fluid from patients with cryptogenic fibrosing alveolitis (43), and histamine can, under certain circumstances, stimulate fibroblast proliferation (44). Mast cells also release heparin, which binds to endothelial cells and stimulates them to release endothelial cell-derived growth factors, which in turn can stimulate fibroblast proliferation (45). Mast cells may therefore playa role in modulating the deposition of lung collagen in systemic sclerosis. Descriptions of the lung in systemic sclerosis often report characteristic alterations in the pulmonary blood vessels (46, 47). These include "onionskin" intimal proliferation comparable to that seen in renal and other arteries in this disease. Such specific appearances generally accompany pulmonary hypertension, which was not apparent clinically in any of our patients. When we identifiedintimal fibrosis and medial hypertrophy, they were found only in areas affected by parenchymal disease, that is, fibrosing alveolitis. These vascular abnormalities are those of endarteritis obliterans, as seen in any chronically diseased tissue. The clinicopathologic correlations demonstrated a significant inverse correlation between extent of interstitial inflammation and patients' age, disease duration, and age at onset of systemic sclerosis. These data suggest alveolitis is a more prominent feature in younger patients in the early stages of disease. Whether such findings have any bearing on prognosis is currently unknown. We also demonstrated an inverse correlation between DLco and both interstitial fibrosis and loss of alveolar architecture. A previous study of lung function suggested that DLCO was a predictor of outcome in systemic sclerosis (48). Our findings, together with these data, suggest that in the absence of pulmonary vascular disease, fibrosis of the alveolar walls and loss of alveolar architecture, rather than alveolitis, are determinants of disease outcome in systemic sclerosis. No correlations were observed between the clinical data and findings on electron microscopy. In particular, there was no correlation between the DLco and endothelial abnormalities, which might be expected if endothelial damage is important in the genesis and progression of the disease. Extensive endothelial abnormalities were a very early feature, but we also know they are present in more advanced disease.
MORPHOLOGY OF INTERSTITIAL LUNG DISEASE IN SYSTEMIC SCLEROSIS
In summary, this study has demonstrated histologic and ultrastructural abnormalities of the lungs in systemic sclerosis indistinguishable from those of lone cryptogenic fibrosing alveolitis and that these are frequently present at sites in the lung that appear normal by noninvasive techniques. Early features are interstitial edema and fibrosis accompanied by endothelial and epithelial cell injury, but not necessarily inflammation. The hypothesis that injury to endothelial and/or epithelium is the initiating event in fibrosing alveolitis requires further evaluation. References 1. Church RE, Ellis ARP. Cystic pulmonary fibrosis in generalised scleroderma. Lancet 1950;1:392-4. 2. Hayman LD, Hunt RE. Pulmonary fibrosis in generalised scleroderma. Dis Chest 1952; 21: 691-704. 3. Opie LH. The pulmonary manifestations of generalised scleroderma (progressive systemic sclerosis). Dis Chest 1955; 28:665-80. 4. WeaverAL, Divertie MB, Titus JL. Pulmonary scleroderma. Dis Chest 1968; 54:4-12. 5. D'Angelo WA, Fries JF, Masi AT,Shulman LE. Pathologic observations in systemic sclerosis. Am J Med 1969; 46:428-40. 6. Spain DM, Thomas AG. The pulmonary manifestations of scleroderma: an anatomicphysiologic correlation. Ann Intern Med 1950; 32:152-61. 7. Ashba JK, Ghanem MH. The lungs in systemic sclerosis. Dis Chest 1965; 47:52-64. 8. Rossi GA, Bitterman PB, Rennard SI, Ferrans VJ, Crystal RG. Evidence for chronic inflammation as a component of the interstitial lung disease associated with progressive systemic sclerosis. Am Rev Respir Dis 1985; 131:612-7. 9. Rudd RM, Haslam PL, Turner-Warwick M. Cryptogenic fibrosing alveolitis: relationship of pulmonary physiology and bronchoalveolar lavage to response to treatment and prognosis. Am Rev Respir Dis 1981; 124:1-8. 10. Wilson RJ, Rodnan GP, Robin ED. An early pulmonary physiologic abnormality in progressive systemic sclerosis (diffuse scleroderma). Am J Med 1964; 36:361-9. 11. 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; 23:581-90. 12. Harrison NK, Glanville AR, Strickland B, et al. Pulmonary involvement in systemic sclerosis: the detection of early changes by thin section CT scan, bronchoalveolar lavage and 99 m Tc-DTPA clearance. Respir Med 1989; 83:403-14. 13. Turner-Warwick M, Burrows B, Johnson A. Cryptogenic fibrosing alveolitis: response to corticosteroid treatment and its effects on survival. Thorax 1980; 35:593-9.
14. Watters LC, Schwartz MI, Cherniak RM, et al. Idiopathic pulmonary fibrosis: pretreatment bronchoalveolar lavage cellular constituents and their relationships with lung histopathology and clinical response to therapy. Am Rev Respir Dis 1987; 135:696-704. 15. Harrison NK, McAnulty RJ, Haslam PL, Black CM, Laurent GJ. Evidence for protein oedema, neutrophil influx and enhanced collagen production in lungs of patients with systemic sclerosis. Thorax 1990; 45:606-10. 16. Hammar SP, Winterbauer RH, Bockus D, Remington F, Sale GE, Meyers JD. Endothelial cell damage and tubuloreticular structures in interstitiallung disease associated with collagen vascular disease and viral pneumonia. Am Rev Respir Dis 1983; 127:77-84. 17. Wallaert B, Hatron BY,Grosbois JM, Tonnel AB, Devulder B, Voisin C. Subclinical pulmonary involvement in collagen vascular disease. Am Rev Respir Dis 1986; 133:574-80. 18. Haslam PL. Bronchoalveolar lavage. Semin Respir Med 1984; 6:55-70. 19. Corrin B. The lungs. Edinburgh: Churchill Livingstone, 1990; 203. 20. Alton EW, Kwan S, Turner-Warwick M. Pulmonary involvement in scleroderma; comparison with cryptogenic fibrosing alveolitis. Thorax 1986; 41:714P. 21. Turner-Warwick M, Burrows B, Johnson A. Cryptogenic fibrosing alveolitis: clinical features and their influence on survival. Thorax 1980; 35:171-80. 22. Corrin B. Dewar A. Rodriguez-Rosin R, Turner-Warwick M. Fine structural changes in cryptogenic fibrosing alveolitis and asbestosis. J Pathol 1985; 147:107-19. 23. Reidy MA. Schwartz SM. Endothelial regeneration. III. Time course of intimal changes after small defined injury to rat aortic endothelium. Lab Invest 1981; 44:301-8. 24. Vracko R. Benditt EP. Capillary basal lamina thickening and duplication in relationship to endothelial cell death and replacement. J Cell Biol 1970; 17:281-5. 25. Adamson IYR. Bowden DH. The type 2 cell as progenitor of alveolar epithelial regeneration. A cytodynamic study in mice after exposure to oxygen. Lab Invest 1974; 30:35-42. 26. Evans MJ. Cabral LJ, Stephens RJ. Freeman G. Transformation of alveolar type II cells following exposure to N0 2 • Exp Mol Pathol 1975; 22:142-50. 27. LeRoy EC. Pathogenesis of scleroderma (systemic sclerosis). J Invest Dermatol1982; 79(Suppl 1):87s-9s. 28. Brown LF. Dvorak AM, Dvorak HE Leaky vessels. fibrin deposition and fibrosis: a sequence of events common to solid tumors and to many other types of disease. Am Rev Respir Dis 1989; 140:1104-7. 29. Gray AJ, Reeves JT, Harrison NK, Winlove P, Laurent GJ. Growth factors for fibroblasts in the soluble products of clot formation. J Cell Sci 1990; 96:271-4. 30. Zapol WM. Trelstad RL, Coffey JW. Tsai I. Salvador RA. Pulmonary fibrosis in severe acute
713 respiratory failure. Am Rev Respir Dis 1979; 119:547-54. 31. Heard BE, Steiner RE, Herdan A, Gleason D. Oedema and fibrosis of the lungs in left ventricular failure. Br J Radiol 1968; 41:161-71. 32. Schaff Z. Barry DW, Grimley PM. Cytochemistry of tubuloreticular structures in lymphocytes of patients with systemic lupus erythematosus and in cultured lymphoid cells: comparison to a paramyxovirus. Lab Invest 1973; 29:577-86. 33. Rich SA. Human lupus inclusions and interferon. Science 1981; 213:772-5. 34. Gracey DR. Divertie MB, Brown AL. Alveolarcapillary membrane in idiopathic interstitial pulmonary fibrosis: electron microscopic study of 14 cases. Am Rev Respir Dis 1968; 98:16-21. 35. Leroy EP. Ultrastructure of alveoli in acute and chronic fibrosing alveolitis. Am J PathoI1972; 66:102a. 36. Coalson J J. The ultrastructure of human fibrosing alveolitis. Virchows Arch [A] 1982; 395:181-99. 37. Turner-Warwick M, Haslam PL, WeeksJ. Antibodies in some chronic fibrosing lung diseases. Clin Allergy 1971; 1:209-19. 38. Schwarz MI, Dreisen RB, Pratt DS, Stanford RE. Immunofluorescent patterns in the idiopathic interstitial pneumonias. J Lab Clin Med 1978; 91:929-38. 39. Kawanami 0, Ferrans VJ. Fulmer JD, Crystal RG. Ultrastructure of pulmonary mast cells in patients with fibrotic lung disorders. Lab Invest 1979; 40:717-34. 40. Hawkins RA. Claman HN. Clark RAF. Steigerwald JC. Increased dermal mast cell populations in progressive systemic sclerosis: a link in chronic fibrosis? Ann Intern Med 1985; 102:182-6. 41. Claman HN. On scleroderma. Mast cells, endothelial cells and fibroblasts. JAMA 1989; 262:1206-9. 42. Shulman ES, Liu MC. Proud D. MacGlashan DW. Lichtenstein LM. Plant M. Human lung macrophages induce histamine release from basophils and mast cells. Am Rev Respir Dis 1985; 131:230-5. 43. Haslam PL. Cromwell 0. Dewar A, TurnerWarwick M. Evidence of increased histamine levels in lung lavage fluids from patients with cryptogenic fibrosing alveolitis. Clin Exp Immunol1981; 44:587-93. 44. Jordana M, Befus AD. Newhouse MT, Bienenstock J. Gauldie J. Effects of histamine on proliferation of human adult lung fibroblasts. Thorax 1988; 43:552-8. 45. Gajdusek C. Release of endothelial cell-derived growth factor (ECDGF) by heparin. J Cell Physiol 1984; 121:13-21. 46. Naeye RL. Pulmonary vascular lesions in systemic sclerosis. Dis Chest 1963; 44:374-80. 47. Young RH, Mark GJ. Pulmonary vascular changes in scleroderma. Am J Med 1978; 64:998-104. 48. Peters-Golden M. Wise RA. Hocherg MC. Stevens MB. Wigley FM. Carbon monoxide diffusing capacity as a predictor of outcome in systemic sclerosis. Am J Med 1984; 77:1027-34.
NICHOLAS K. HARRISON, ALLEN R. MYERS, BRYAN CORRIN, GERALDINE SOOSAY, ANN DEWAR, CAROL M. BLACK, ROLAND M. DU BOIS, and MARGARET TURNER-WARWICK
Introduction
Patients with systemic sclerosis frequently develop interstitial lung disease, although pulmonary abnormalities are seldom apparent when scleroderma is first diagnosed. The investigation of patients with scleroderma should therefore allow identification of interstitial lung disease at an early stage in its natural history when morphologic features may give important clues to pathogenesis. This is in contrast to patients with "lone" cryptogenic fibrosing alveolitis, who necessarily present when lung disease givesrise to respiratory symptoms. Most reports of lung pathology in systemic sclerosis describe abnormalities found at autopsy. These include extensive interstitial fibrosis with cysticchanges, bronchiolectasis, and pleural thickening (1-5). Inflammatory cellsare also observed (6)but are often thought to represent terminal pathologic events rather than the underlying primary condition (7). These postmortem findings obviously represent advanced disease and add little to our understanding of pathogenesis. Detailed descriptions of antemortem lung pathology in systemic sclerosis have been confined to small numbers of patients. Alveolitis has been observed in these studies, but it is uncertain whether the inflammatory cell infiltrate precedes, accompanies, or follows fibrosis of the alveolar walls (8). The recognition that lung disease in systemic sclerosis resembles that of lone cryptogenic fibrosing alveolitis has led some workers to include systemic sclerosis patients in clinical studies of fibrosing alveolitis (9), but there has never been a formal comparison of lung histology between the two conditions. The ability of modern imaging techniques and lung function tests to detect early pulmonary abnormalities in systemic sclerosis provides an opportunity to examine lung structure at a time when morphologic features should giveinsight 706
SUMMARY Most studies of lung histology in systemic sclerosis have been based on autopsy specimens and consequently emphasize end-stage fibrotic disease. Although occasional pathologic descriptions of open-lung biopsies have recognized the presence of Inflammatory cells, suggesting a similarity to "lone" cryptogenic flbrosing alveolltls, the two conditions have never been formally compared. In this study we describe the morphologic features of 49 open-lung biopsies from 34 systemic sclerosis patients with Interstitial lung disease, many of whom had their lung disease diag· nosed at an early stage. None had pulmonary hypertension. Examination of lung tissue by light microscopy showed the earliest changes to include patchy lymphocyte and plasma cell Infiltration of the alveolar walls, Interstitial fibrosis, and increased macrophages but only occasional polymorphonuclear cells and lymphocytes In the alveolar spaces. Alveolltis was not observed without fibrosis. Comparison of 22 biopsies with a similar number from patients with lone cryptogenic fibrosing alveolitls, matched for age and sex, revealed no qualitative or quantitative differences, other than a higher prevalence of focal lymphoid hyperplasia (follicular bronchiolitis) In the systemic sclerosis patients than in the lone cryptogenic fibroslng alveolitis controls (23 and 5%, respectively). There was an Inverse correlation between the extent of interstitial Inflammation and patients' age (p < 0.05), disease duration (p < 0.05), and age at onset of systemic sclerosis (p < 0.01).There was also an inverse correlation between OLeo and Interstitial fibrosis (p < 0.01)and loss of lung architecture (p < 0.05).Ultrastructural studies of eight systemic sclerosis biopsies showed evidence of endothelial and epithelial injury together with interstitial edema and excess collagen deposition. Occasional mast cells were observed, often in close contact with Interstitial fibroblasts, but there were no tubuloretlcular structures or evidence of Immune complexes. Similar changes were also found in a further three biopsies that appeared normal by light microscopy. These results suggest that in systemic sclerosis pulmonary endothelial and/or epithelial injury may precede Inflammation and fibrosis. AM REV RESPIR DIS 1991; 144:706-713
into pathogenesis, and if the lung disease of systemicsclerosisis truly indistinguishable from that of lone cryptogenic fibrosing alveolitis, this should apply to the latter condition as much as to the first. To date there have been no studies of lung ultrastructure in systemicsclerosis,except for a solitary case report describing thickening of the common endothelial/epithelial basement membrane on the "thin" side of the alveolar wall (10). In the present study we describe the pathologic abnormalities in 49 open-lung biopsies from 34 systemic sclerosis patients with interstitial lung disease, and compare them with biopsies from a group of age- and sex-matched patients with lone cryptogenic fibrosing alveolitis. We also examine lung ultrastructure in systemic sclerosis with particular reference to biopsies showing minimal change in the belief that such changes may be relevant to early pathogenetic mechanisms.
Methods
Patients We studied 34 patients who fulfilled the American Rheumatism Association preliminary criteria for the diagnosis of systemicsclerosis (11) and who had undergone open-lung biopsy at the Brompton Hospital for diagnosis and staging of interstitial lung disease.
(Received in original form October 5, 1990 and in revised form April 3, 1991) 1 From the Departments of Thoracic Medicine and Lung Pathology, National Heart and Lung Institute, Brompton Hospital, and the Department of Rheumatology, Royal Free Hospital, London, United Kingdom, and the Department of Medicine, Temple University School of Medicine, Philadelphia, Pennsylvania. 2 Correspondence and requestsfor reprints should be addressed to Dr. Nicholas K. Harrison, Biochemistry Unit, Department of Thoracic Medicine, National Heart and Lung Institute, Emmanuel Kaye Building, Manresa Road, London SW3 6LR, UK.
707
MORPHOLOGY OF INTERSTITIAL WNG DISEASE IN SYSTEMIC SCLEROSIS
There were 26 women and 8 men; the mean (range) age was 45 (22 to 67) yr. One patient was black, two were from the Indian subcontinent, and 31 were white. A total of 20 had diffuse cutaneous scleroderma, 10 had limited cutaneous scleroderma, and 4 had CREST syndrome. Mean duration of the scleroderma was 7.7 (1to 31)yr. All patients had either abnormal chest radiographs or the characteristic features of fibrosing alveolitis on thinsection (3 mm) computed tomographic (CT) scan (12).Lung function tests were abnormal in many but not all patients. Mean values (0/0 of predicted normal) were FVC 72% (23 to 123%); TLC, 73% (31 to 112%); DLCO, 54% (17to 91%); and carbon monoxide gas transfer coefficient Kco, 83% (46 to 117%). None of the patients had clinical evidence of pulmonary hypertension.
Lung Biopsies A lower lobe biopsy was available for examination in all 34 patients. In 15 patients a second site, usually the right middle lobe, was also biopsied, giving a total of 49 biopsies available for analysis and enabling a comparison to be made between different lobes in the same individual. In 22 patients, lower lobe biopsies were compared with those from a similar number of age- and sex-matched patients with lone cryptogenic fibrosing alveolitis who also underwent open-lung biopsy at the Brompton Hospital for diagnosis and staging. Biopsies were undertaken when there was evidence of interstitial lung disease on CT scan. The purpose of open-lung biopsy was to confirm the diagnosis of interstitial lung disease and allow the clinician to plan therapy on the principle that commencing this as soon as possible may limit disease progression (9, 13, 14). All patients gave informed written consent to lung biopsy. Light Microscopy Paraffin-embedded sections of formalin-fixed tissue were stained with hematoxylin and eosin and by the combined elastin-van Gieson method. Biopsieswererandomized and scored by a pathologist who was unaware of the disease group, using a semiquantitative fourpoint scoring system for each of the following: interstitial inflammation, interstitial fibrosis, intraalveolar inflammation, type II epithelial cell hyperplasia, bronchiolization of alveoli, loss of alveolar architecture, and vascular abnormalities. Electron Microscopy Eleven lung biopsies from seven systemic sclerosis patients with only mild histologic abnormalities were selected for electron microscopy. In all biopsies, areas showing the least abnormalities at the light microscopic level werechosen. In three casesin whom both middle and lower lobes were examined the former appeared normal on light microscopy. In all cases small pieces of fresh lung tissue had been diced and fixed without delay in sodium cacodylate-buffered 2.5% glutar-
aldehyde for 4 to 18h, rinsed in buffer, postfixed in osmium tetroxide, and processed to Araldite", After selecting suitable areas from l-um-thick sections, ultrathin sections were cut, stained with uranyl acetate and lead citrate, and examined by transmission electron microscopy. Middle lobe biopsies from seven patients were scored, using a semiquantitative fourpoint scoring system, for the following abnormalities: Capillaries Endothelial swelling Thickening or reduplication of basement membrane Tubuloreticular structures Epithelium Type I cell swelling Type II cell hyperplasia Bare basement membrane Interstitium Edema Collagen Elastin
Data Handling and Statistics Parametric data are described as mean (range) and were compared by t test. Biopsy scores for different lobes in the same subject were compared by Wilcoxon test. Linear regression analysis was performed to assess correlations between pathologic scores and clinical data. The severity of pathologic abnormalities in systemic sclerosis and cryptogenic fibrosing alveolitis were compared for each of the seven variables by chi-square analysis. When possible, 4 x 2 tables were used, but when numbers were small groups were combined into 3 x 2 or 2 x 2 tables.
Results
Clinical Evaluation Of the 34 systemic sclerosis patients, seven had a normal chest radiograph and of these two had normal lung function tests (lung volumes> 80% of predicted, carbon monoxide diffusing capacity > 750/0 of predicted). CT scans were performed in 27 patients, and all showed abnormalities in the lower lobes. In eight patients, CT scans of the middle or upper lobes were normal, yet biopsies of these lobes showed histologic evidence of fibrosing alveolitis. In a further three patients, both CT scan and histologic examination of the right middle lobe were substantially normal, yet abnormalities were detected at that site by electron microscopy. These are described subsequently. There was no significant pathologic difference between the patients with abnormal chest radiographs and those with normal chest radiograph but thought to have earlier disease in whom
lung biopsy was justified on the basis of an abnormal CT scan. Bronchoalveolar lavage cell profiles were abnormal in all patients. These findings have been reported elsewhere (12, 15).
Histopathology Lower lobe biopsies from all patients with systemic sclerosis generally showed some degree of both inflammation and fibrosis. Appearances varied in extent from almost normal lung to end-stage disease with severe fibrosis and complete loss of lung structure (figures Ia to d). Intraalveolar inflammation consisted of increased numbers of macrophages together with occasional neutrophils, lymphocytes, and eosinophils. Lymphocytes and plasma cells predominated in the interstitium, and occasional small lymphoid aggregates were observed adjacent to bronchioles (follicular bronchiolitis) or the pleura (figure Ie). Middle or upper lobe biopsies were similar in appearance to those of the lower lobes. Abnormalities tended to be more marked in the lower lobes, but this difference was not statistically significant for any variable. In three patients the middle lobe biopsies were considered substantially normal: these biopsies were included in the electron microscope study. Pleural abnormalities were rare, consisting of a fibrinous pleurisy in two patients, diffuse pleural thickening in a third, and a collection of necrotizing granulomata in a fourth. These changes werethought to more likelyrelate to coincidental or past infection rather than to the systemic sclerosis disease process. Pulmonary arteries showed media hypertrophy and fibrous thickening of the intima. This was particularly evident where the parenchymal changes were severe (figure If): no vascular abnormalities were observed in areas where there was no parenchymal lung disease. Comparison with Cryptogenic Fibrosing Alveolitis Open-lung biopsies from 22 patients with systemicsclerosiswerecompared with the same number from a group of age- and sex-matched patients with lone cryptogenic fibrosing alveolitis. The clinical characteristics of the two groups are summarized in table 1. These patients were not matched for duration of pulmonary disease because in many systemicsclerosis patients CT abnormalities wereidentified before there were respiratory symptoms or abnormalities on chest radiograph.
708
HARRISON, MYERS, CORRIN, ET AL.
709
MORPHOLOGY OF INTERSTITIAL WNG DISEASE IN SYSTEMIC SCLEROSIS
With the exception of the pulmonary lymphoid tissue, there were no qualitative differences in the lung parenchyma between systemic sclerosisand lone cryptogenic fibrosing alveolitis, and when they were scored, chi-square analysis showed there were no significant quantitative differences in the extent of parenchymal or vascular disease between the two groups (table 2). Pleural abnormalities were not observed in any patient with lone cryptogenic fibrosing alveolitis and in only four of the 49 systemic sclerosis biopsies. Focal peribronchial lymphoid hyperplasia was seen in seven systemic sclerosis patients (five of those with age- and sex-matched controls) and two control patients with lone cryptogenic fibrosing alveolitis. In two systemic sclerosis patients and one fibrosing alveolitis control patient the lymphoid hyperplasia included the formation of follicles with germinal centers (follicular bronchiolitis). In addition, pleural lymph nodes were identified in two systemic sclerosis patients.
Fig. 1. Open-lung biopsies stained with hematoxylin and eosin showing the range of pulmonary abnormalities in systemic sclerosis. (a) Mild fibrosis and Iymphoplasmacytic infiltration of the alveolar walls (magnification: x130). (b) Moderate inflammatory changes consisting predominantly of lymphocyte and plasma cell infiltration of the alveolar walls and macrophage accumulation in the airspaces. Alveolar walls also show moderate epithelial Type II cell hyperplasia and fibrosis (magnification: x130). (c) Severe pulmonary fibrosis with loss of alveolar architecture (magnification: x80). (d) End-stage lung disease with severe fibrosis and complete loss of normal lung structure yet continuing active inflammation (magnification: x80). (e) Follicular bronchiolitis (magnification: x110). (f) Pulmonary artery showing medial hypertrophy and intimal fibrosis. These changes were observed only in areas of lung showing moderate or severe parenchymal disease. (Hematoxylin and eosin; magnification: x 240).
TABLE 1 CLINICAL CHARACTERISTICS OF 22 PATIENTS WITH SYSTEMIC SCLEROSIS AND 22 AGE- AND SEX-MATCHED PATIENTS WITH "LONE" CRYPTOGENIC FIBROSING ALVEOLITIS (CFA)
Systemic sclerosis (n = 22) Lone CFA (n = 22)
Lung Function (% of Predicted)
Age (yr)
Sex (F:M)
CXR (N/Ab)
FVC
TLC
46 ± 3
14:8
4/18
75 ± 6
75 ± 5
52 ± 4
79 ± 4
47 ± 3
14:8
0/22
69 ± 5
72 ± 5
37 ± 3"
66 ± 5
Definition of abbreviations: CXR = chest radiograph; N/Ab= normallabnormal.· • p < 0.05.
DLCO
Kco
Ultrastructural Abnormalities Ultrastructural abnormalities of endothelium, epithelium, and interstitium weredetected in all II biopsies examined, including the three that appeared substantially normal by light microscopy. In those three biopsies, endothelial cells were often swollen and electron lucent (figures 2 and 3) or, less commonly, pyknotic. In places the pyknotic cells had the appearance of being undermined and replaced by adjacent healthy endothelial cells (figure 4). Endothelial abnormalities were sometimes associated with thickening of the endothelial basement membrane (figure 4). Epithelial abnormalities consisted of TABLE 2 COMPARISON OF SCORES IN 22 PATIENTS AND CRYPTOGENIC
Interstitial inflammation Alveolar inflammation Interstitial fibrosis Type IJ cell hyperplasia Bronchiolization of alveoli Loss of lung architecture Blood vessels
MEAN LUNG PATHOLOGY SYSTEMIC SCLEROSIS CONTROLS WITH LONE FIBROSING ALVEOLITIS
Systemic Sclerosis
Lone CFA
Chi-square p Value
1.4
1.5
0.3
1.4
1.4
0.5
2.1
2.1
0.2
1.5
1.2
0.1
1.3
1.2
0.2
1.6
1.3
0.2
1.8
1.4
0.2
HARRISON, MYERS, CORRIN, ET AL.
710
mune complexes were observed, nor was there evidence of the tubuloreticular structures described in interstitial lung disease associated with other collagen vascular diseases and in viral pneumonias (16).
Figure 2. Electron micrograph of alveolar wall showing patchy cytoplasmic swelling of endothelial (star) and epithelial (arrow) cells. (Uranyl acetate and lead citrate; bar = 2 um.)
Clinicopathologic Correlations For the changes observed by light microscopy,regression analysis indicated a weak inverse correlation between the extent of interstitial inflammation and patients' age (r = - 0.36, p < 0.05), disease duration (r = - 0.34, p < 0.05), and age at onset of systemic sclerosis (r = - 0.48, p < 0.01). There was also a significant negative correlation between DLCO and interstitial fibrosis (r = - 0.46, p < 0.01) and loss of lung architecture (r = - 0.4, p < 0.05). For changes observed by electron microscopy, regression analysis did not show any significant correlations with clinical data. Discussion
patchy epithelial cell swelling (figure 2), focal loss exposing bare epithelial basement membranes (figure 5), and focal Type II pneumocyte proliferation. There was no clear difference in severity between endothelial and epithelial damage. Interstitial abnormalities consisted of areas of electron-lucent edema (figure 5) and increased interstitial collagen (figure 6) and elastin, largely confined to the
thick side of the air/blood barrier. Occasional mast cells were observed, often in intimate contact with interstitial fibroblasts (figure 3). Comparison of these appearances with the eight systemic sclerosis biopsies that had clear evidence of fibrosing alveolitis by light microscopy showed less edema and more fibrosis in the latter. No electron-dense deposits suggestive of im-
Figure 3. Capillary showing cytoplasmic endothelial swelling (star). A mast cell is seen in the interstitium in close contact with thin cytoplasmic processes of a fibroblast. (Uranyl acetate and lead citrate; bar = 2 um.)
All previous studies of lung pathology in systemic sclerosis involving significant numbers of patients used material obtained at autopsy. Consequently they emphasized advanced fibrotic disease and failed to identify early structural abnormalities. In contrast, the present study applied modern imaging techniques to detect interstitial lung disease at an early stage and therefore includes some asymptomatic patients who had normal chest radiographs and normal lung function tests. This provided us with the opportunity to examine the early structural abnormalities of the lung in systemic sclerosis. A striking observation was the extent of the ultrastructural changes. All 46 biopsies in which abnormalities were apparent by light microscopy showedpatchy inflammatory cellinfiltration of the interstitium and alveolar spaces accompanying fibrosis of the alveolar walls. This suggests that these processes occur together from the earliest stages of the disease. Interstitial infiltration by lymphocytes and plasma cells was one of the earliest abnormalities apparent by light microscopy. This contrasts with studies of early lung disease by bronchoalveolar lavage, which show increased proportions of neutrophils (15, 17). This discrepancy between histology and bronchoalveolar lavage is well known in lone cryptogenic fibrosing alveolitis. It is thought to reflect differential sampling of inflammatory cellsby the lavage procedure (18). Whether the interstitial lymphocytes are promoting or inhibiting fibrosis, or both, re-
MORPHOLOGY OF INTERSTITIAL WNG DISEASE IN SYSTEMIC SCLEROSIS
Figure 4. Pyknotic capillary endothelial cells are undermined by adjacent healthy cells, and there is thickening of the endothelial basement membrane (arrows). (Uranyl acetate and lead citrate; bar = 2 urn.)
mains uncertain, but might be investigated in the future by localization of lymphocyte subtypes and identification of their lymphokines. The parenchymal abnormalities weobserved in lung biopsies from patients with systemic sclerosis wereindistinguishable from those of patients with lone cryptogenic fibrosing alveolitis, with the excep-
tion of the focal lymphoid hyperplasia, which was more common in the systemic sclerosis patients. Focal lymphoid hyperplasia (follicular bronchiolitis) is also a feature of rheumatoid lung disease (19). We do not regard it as a component of the intestitiallung disease but as an independent feature of systemiccollagen vascular disease, which, when pres-
Figure 5. Alveolar epithelial cell loss has led to extensive denudation of the basement membrane (arrows). Residual alveolar epithelium is pyknotic (arrowheads). Interstitial edema is also evident (star). (Uranyl acetate and lead citrate; bar = 2 urn.)
711
ent, is a useful pointer to the possibility that fibrosing alveolitis is only one manifestation of such systemic disease. Apart from lymphoid hyperplasia, the interstitial lung disease of systemic sclerosis is identical to that seen in lone cryptogenic fibrosing alveolitis, supporting the concept that the interstitial lung disease of systemic sclerosis provides an opportunity to study the pathogenesis of fibrosing alveolitis at an early stage,often before the onset of respiratory symptoms. However, despite their pathologic identity, there is preliminary evidence to suggest that systemic sclerosis patients with fibrosing alveolitis have a more favorable prognosis than those with lone cryptogenic fibrosing alveolitis (20). If this is true, it is interesting to speculate on factors that may be modulating the disease in systemicsclerosis. It is known that men with cryptogenic fibrosing alveolitis have a worse prognosis than women (21), and the preponderance of women with systemic sclerosis may be one such factor. Other possibilities are that treatment for scleroderma, commenced before the development of interstitial lung disease, favorably affects the course of the latter or that inhibitors to collagen production are present in systemic sclerosis patients but not in those with lone cryptogenic fibrosing alveolitis. The ultrastructural abnormalities we observed in this study are also similar to those described previously in patients with lone cryptogenic fibrosing alveolitis (22). These have not been described in systemic sclerosis, however, and this is the first report of such changes in biopsies that appear normal by light microscopy. These early changes are important when considering the pathogenesis of fibrosing alveolitis. In particular, the lack of inflammatory cell infiltrate does not favor alveolitis preceding structural damage, although it is possible that such mediators as oxidants and fibroblast growth factors released from inflammatory cells in adjacent parts of the lung could be responsible for the damage and fibrosis we observed. Alternatively, if alveolitis precedes structural damage, our findings suggest this may be a "hit-andrun" phenomenon, the inflammation abating but leaving residual injury. Although we paid particular attention to tissue showing minimal changes, we were not able to determine whether endothelial or epithelial abnormalities predominated. Endothelial damage may appear less marked because of the ability of regenerating endothelial cellsto spread rapidly beneath adjacent damaged cells
712
HARRISON, MYERS, CORRIN,
Figure 6. The alveolar interstitium contains excess collagen. (Uranyl acetate and lead citrate; bar = 2 urn.)
(23, 24). In contrast, replacement of lost Type I epithelial cells requires the more prolonged process of proliferation and differentiation of Type II pneumocytes (25, 26). Injury to endothelial cells by circulating agents has been proposed as a pathogenetic mechanism in skin and other organs in systemic sclerosis (27). Injury to capillary endothelial cells is likely to account for the interstitial edema we observed. That increased permeability of the endothelial/epithelial barrier occurs early in the pathogenesis of systemic sclerosis lung disease is supported by observations that protein concentration in bronchoalveolar lavage fluid is increased (15) and clearance of radiolabeled, low-molecular-weight solutes is more rapid in these patients (12). Whether vascular leakage alone is sufficient to stimulate interstitial fibroblasts is uncertain, but it is increasingly recognized that circulating proteins can act as potent fibroblast mitogens (28, 29). Furthermore, other conditions associated with profound but transient vascular leakage, such as the adult respiratory distress syndrome, frequently progress to pulmonary fibrosis (30). Similarly, such conditions as mitral stenosis and pulmonary venoocelusive disease that are characterized by chronic edema of the lungs may also result in pulmonary fibrosis (31). We failed to identify any tubuloreticular structures in association with the endothelial abnormalities. These structures, which are composed of phospholipid and glycoprotein (32), were observed within damaged pulmonary endothelial cells
and peripheral blood lymphocytes in a study of patients with viral pneumonia and certain collagen vascular diseases (17), but lung tissue from patients with systemic sclerosis was not examined. Furthermore, the same study failed to identify tubuloreticular structures in lung biopsies from 30 patients with idiopathic pulmonary fibrosis (cryptogenic fibrosing alveolitis). It has been demonstrated that interferons can induce the formation of tubuloreticular structures within cells (33), and this has been proposed as a possible mechanism for their occurrence in viral pneumonias (17). The apparent absence of these structures in both systemic sclerosis and cryptogenic fibrosing alveolitis argues against a viral cause for these conditions. We also failed to identify any electrondense deposits suggestive of immune complexes. This has been a consistent finding in lone cryptogenic fibrosing alveolitis (22, 34-36), despite a few reports of their detection in early cases of this disease by immunofluorescent microscopy (37, 38). We did observe occasional mast cells, as described in other fibrotic lung diseases (39). This finding is consistent with similar observations in early skin lesions in systemic sclerosis (40), and it has been proposed that mast cells may represent a link between activated endothelial cells and skin fibroblasts in this condition (41). It is known that alveolar macrophages possess H 2 receptors and can release a factor that causes human lung mast cells to release histamine (42). Elevated levels
er AL.
of histamine have been observed in bronchoalveolar lavage fluid from patients with cryptogenic fibrosing alveolitis (43), and histamine can, under certain circumstances, stimulate fibroblast proliferation (44). Mast cells also release heparin, which binds to endothelial cells and stimulates them to release endothelial cell-derived growth factors, which in turn can stimulate fibroblast proliferation (45). Mast cells may therefore playa role in modulating the deposition of lung collagen in systemic sclerosis. Descriptions of the lung in systemic sclerosis often report characteristic alterations in the pulmonary blood vessels (46, 47). These include "onionskin" intimal proliferation comparable to that seen in renal and other arteries in this disease. Such specific appearances generally accompany pulmonary hypertension, which was not apparent clinically in any of our patients. When we identifiedintimal fibrosis and medial hypertrophy, they were found only in areas affected by parenchymal disease, that is, fibrosing alveolitis. These vascular abnormalities are those of endarteritis obliterans, as seen in any chronically diseased tissue. The clinicopathologic correlations demonstrated a significant inverse correlation between extent of interstitial inflammation and patients' age, disease duration, and age at onset of systemic sclerosis. These data suggest alveolitis is a more prominent feature in younger patients in the early stages of disease. Whether such findings have any bearing on prognosis is currently unknown. We also demonstrated an inverse correlation between DLco and both interstitial fibrosis and loss of alveolar architecture. A previous study of lung function suggested that DLCO was a predictor of outcome in systemic sclerosis (48). Our findings, together with these data, suggest that in the absence of pulmonary vascular disease, fibrosis of the alveolar walls and loss of alveolar architecture, rather than alveolitis, are determinants of disease outcome in systemic sclerosis. No correlations were observed between the clinical data and findings on electron microscopy. In particular, there was no correlation between the DLco and endothelial abnormalities, which might be expected if endothelial damage is important in the genesis and progression of the disease. Extensive endothelial abnormalities were a very early feature, but we also know they are present in more advanced disease.
MORPHOLOGY OF INTERSTITIAL LUNG DISEASE IN SYSTEMIC SCLEROSIS
In summary, this study has demonstrated histologic and ultrastructural abnormalities of the lungs in systemic sclerosis indistinguishable from those of lone cryptogenic fibrosing alveolitis and that these are frequently present at sites in the lung that appear normal by noninvasive techniques. Early features are interstitial edema and fibrosis accompanied by endothelial and epithelial cell injury, but not necessarily inflammation. The hypothesis that injury to endothelial and/or epithelium is the initiating event in fibrosing alveolitis requires further evaluation. References 1. Church RE, Ellis ARP. Cystic pulmonary fibrosis in generalised scleroderma. Lancet 1950;1:392-4. 2. Hayman LD, Hunt RE. Pulmonary fibrosis in generalised scleroderma. Dis Chest 1952; 21: 691-704. 3. Opie LH. The pulmonary manifestations of generalised scleroderma (progressive systemic sclerosis). Dis Chest 1955; 28:665-80. 4. WeaverAL, Divertie MB, Titus JL. Pulmonary scleroderma. Dis Chest 1968; 54:4-12. 5. D'Angelo WA, Fries JF, Masi AT,Shulman LE. Pathologic observations in systemic sclerosis. Am J Med 1969; 46:428-40. 6. Spain DM, Thomas AG. The pulmonary manifestations of scleroderma: an anatomicphysiologic correlation. Ann Intern Med 1950; 32:152-61. 7. Ashba JK, Ghanem MH. The lungs in systemic sclerosis. Dis Chest 1965; 47:52-64. 8. Rossi GA, Bitterman PB, Rennard SI, Ferrans VJ, Crystal RG. Evidence for chronic inflammation as a component of the interstitial lung disease associated with progressive systemic sclerosis. Am Rev Respir Dis 1985; 131:612-7. 9. Rudd RM, Haslam PL, Turner-Warwick M. Cryptogenic fibrosing alveolitis: relationship of pulmonary physiology and bronchoalveolar lavage to response to treatment and prognosis. Am Rev Respir Dis 1981; 124:1-8. 10. Wilson RJ, Rodnan GP, Robin ED. An early pulmonary physiologic abnormality in progressive systemic sclerosis (diffuse scleroderma). Am J Med 1964; 36:361-9. 11. 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; 23:581-90. 12. Harrison NK, Glanville AR, Strickland B, et al. Pulmonary involvement in systemic sclerosis: the detection of early changes by thin section CT scan, bronchoalveolar lavage and 99 m Tc-DTPA clearance. Respir Med 1989; 83:403-14. 13. Turner-Warwick M, Burrows B, Johnson A. Cryptogenic fibrosing alveolitis: response to corticosteroid treatment and its effects on survival. Thorax 1980; 35:593-9.
14. Watters LC, Schwartz MI, Cherniak RM, et al. Idiopathic pulmonary fibrosis: pretreatment bronchoalveolar lavage cellular constituents and their relationships with lung histopathology and clinical response to therapy. Am Rev Respir Dis 1987; 135:696-704. 15. Harrison NK, McAnulty RJ, Haslam PL, Black CM, Laurent GJ. Evidence for protein oedema, neutrophil influx and enhanced collagen production in lungs of patients with systemic sclerosis. Thorax 1990; 45:606-10. 16. Hammar SP, Winterbauer RH, Bockus D, Remington F, Sale GE, Meyers JD. Endothelial cell damage and tubuloreticular structures in interstitiallung disease associated with collagen vascular disease and viral pneumonia. Am Rev Respir Dis 1983; 127:77-84. 17. Wallaert B, Hatron BY,Grosbois JM, Tonnel AB, Devulder B, Voisin C. Subclinical pulmonary involvement in collagen vascular disease. Am Rev Respir Dis 1986; 133:574-80. 18. Haslam PL. Bronchoalveolar lavage. Semin Respir Med 1984; 6:55-70. 19. Corrin B. The lungs. Edinburgh: Churchill Livingstone, 1990; 203. 20. Alton EW, Kwan S, Turner-Warwick M. Pulmonary involvement in scleroderma; comparison with cryptogenic fibrosing alveolitis. Thorax 1986; 41:714P. 21. Turner-Warwick M, Burrows B, Johnson A. Cryptogenic fibrosing alveolitis: clinical features and their influence on survival. Thorax 1980; 35:171-80. 22. Corrin B. Dewar A. Rodriguez-Rosin R, Turner-Warwick M. Fine structural changes in cryptogenic fibrosing alveolitis and asbestosis. J Pathol 1985; 147:107-19. 23. Reidy MA. Schwartz SM. Endothelial regeneration. III. Time course of intimal changes after small defined injury to rat aortic endothelium. Lab Invest 1981; 44:301-8. 24. Vracko R. Benditt EP. Capillary basal lamina thickening and duplication in relationship to endothelial cell death and replacement. J Cell Biol 1970; 17:281-5. 25. Adamson IYR. Bowden DH. The type 2 cell as progenitor of alveolar epithelial regeneration. A cytodynamic study in mice after exposure to oxygen. Lab Invest 1974; 30:35-42. 26. Evans MJ. Cabral LJ, Stephens RJ. Freeman G. Transformation of alveolar type II cells following exposure to N0 2 • Exp Mol Pathol 1975; 22:142-50. 27. LeRoy EC. Pathogenesis of scleroderma (systemic sclerosis). J Invest Dermatol1982; 79(Suppl 1):87s-9s. 28. Brown LF. Dvorak AM, Dvorak HE Leaky vessels. fibrin deposition and fibrosis: a sequence of events common to solid tumors and to many other types of disease. Am Rev Respir Dis 1989; 140:1104-7. 29. Gray AJ, Reeves JT, Harrison NK, Winlove P, Laurent GJ. Growth factors for fibroblasts in the soluble products of clot formation. J Cell Sci 1990; 96:271-4. 30. Zapol WM. Trelstad RL, Coffey JW. Tsai I. Salvador RA. Pulmonary fibrosis in severe acute
713 respiratory failure. Am Rev Respir Dis 1979; 119:547-54. 31. Heard BE, Steiner RE, Herdan A, Gleason D. Oedema and fibrosis of the lungs in left ventricular failure. Br J Radiol 1968; 41:161-71. 32. Schaff Z. Barry DW, Grimley PM. Cytochemistry of tubuloreticular structures in lymphocytes of patients with systemic lupus erythematosus and in cultured lymphoid cells: comparison to a paramyxovirus. Lab Invest 1973; 29:577-86. 33. Rich SA. Human lupus inclusions and interferon. Science 1981; 213:772-5. 34. Gracey DR. Divertie MB, Brown AL. Alveolarcapillary membrane in idiopathic interstitial pulmonary fibrosis: electron microscopic study of 14 cases. Am Rev Respir Dis 1968; 98:16-21. 35. Leroy EP. Ultrastructure of alveoli in acute and chronic fibrosing alveolitis. Am J PathoI1972; 66:102a. 36. Coalson J J. The ultrastructure of human fibrosing alveolitis. Virchows Arch [A] 1982; 395:181-99. 37. Turner-Warwick M, Haslam PL, WeeksJ. Antibodies in some chronic fibrosing lung diseases. Clin Allergy 1971; 1:209-19. 38. Schwarz MI, Dreisen RB, Pratt DS, Stanford RE. Immunofluorescent patterns in the idiopathic interstitial pneumonias. J Lab Clin Med 1978; 91:929-38. 39. Kawanami 0, Ferrans VJ. Fulmer JD, Crystal RG. Ultrastructure of pulmonary mast cells in patients with fibrotic lung disorders. Lab Invest 1979; 40:717-34. 40. Hawkins RA. Claman HN. Clark RAF. Steigerwald JC. Increased dermal mast cell populations in progressive systemic sclerosis: a link in chronic fibrosis? Ann Intern Med 1985; 102:182-6. 41. Claman HN. On scleroderma. Mast cells, endothelial cells and fibroblasts. JAMA 1989; 262:1206-9. 42. Shulman ES, Liu MC. Proud D. MacGlashan DW. Lichtenstein LM. Plant M. Human lung macrophages induce histamine release from basophils and mast cells. Am Rev Respir Dis 1985; 131:230-5. 43. Haslam PL. Cromwell 0. Dewar A, TurnerWarwick M. Evidence of increased histamine levels in lung lavage fluids from patients with cryptogenic fibrosing alveolitis. Clin Exp Immunol1981; 44:587-93. 44. Jordana M, Befus AD. Newhouse MT, Bienenstock J. Gauldie J. Effects of histamine on proliferation of human adult lung fibroblasts. Thorax 1988; 43:552-8. 45. Gajdusek C. Release of endothelial cell-derived growth factor (ECDGF) by heparin. J Cell Physiol 1984; 121:13-21. 46. Naeye RL. Pulmonary vascular lesions in systemic sclerosis. Dis Chest 1963; 44:374-80. 47. Young RH, Mark GJ. Pulmonary vascular changes in scleroderma. Am J Med 1978; 64:998-104. 48. Peters-Golden M. Wise RA. Hocherg MC. Stevens MB. Wigley FM. Carbon monoxide diffusing capacity as a predictor of outcome in systemic sclerosis. Am J Med 1984; 77:1027-34.
