Abstract
Fibrosing alveolitis associated with systemic sclerosis (FASSc) has a better prognosis than idiopathic pulmonary fibrosis. In view of recent reports that idiopathic nonspecific interstitial pneumonia (NSIP) has a better prognosis than idiopathic usual interstitial pneumonia (UIP), we classified histologic appearances of surgical lung biopsies performed in 80 patients with FASSc. NSIP (n = 62, 77.5%), subcategorized as cellular NSIP (n = 15) and fibrotic NSIP (n = 47) was much more prevalent than UIP (n = 6), end-stage lung disease (ESL, n = 6), or other patterns (n = 6). There were 25 deaths (NSIP 16/62, 26%; UIP/ESL 6/12, 50%). Five-year survival differed little between NSIP (91%) and UIP/ESL (82%); mortality was associated with lower initial carbon monoxide diffusing capacity (DlCO) and FVC levels (p = 0.004 and p = 0.007, respectively). Survival and serial FVC and DlCO trends did not differ between cellular and fibrotic NSIP. Increased mortality in NSIP was associated with lower initial DlCO levels (p = 0.04), higher BAL eosinophil levels (p = 0.03), and deterioration in DlCO levels during the next 3 years (p < 0.005). We conclude that NSIP is the histopathologic pattern in most patients with FASSc. However, outcome is linked more strongly to disease severity at presentation and serial DlCO trends than to histopathologic findings.
Pulmonary involvement occurs more frequently in systemic sclerosis than in other connective tissue diseases (1) and is a significant cause of morbidity and mortality (2). The most common pulmonary manifestations of systemic sclerosis are fibrosing alveolitis, which occurs in approximately 80%, and pulmonary arterial hypertension, seen in up to 50% of patients (3). We have previously shown (4, 5) that fibrosing alveolitis associated with systemic sclerosis (FASSc) has a better prognosis than lone cryptogenic fibrosing alveolitis (CFA), alternatively termed “idiopathic pulmonary fibrosis” (IPF). The classification of idiopathic interstitial pneumonia has been refined recently (6–8); the histopathologic pattern of nonspecific interstitial pneumonia (NSIP), now recognized as a subgroup of the idiopathic interstitial pneumonias, has a prognosis intermediate between usual interstitial pneumonia (UIP) and desquamative interstitial pneumonia/respiratory bronchiolitis interstitial lung disease (DIP/RBILD) (9–13) or cryptogenic organizing pneumonia (14). NSIP is known to occur in collagen vascular disorders (15), but its prevalence and relationship to clinical parameters, response to treatment, and prognosis are not known. In particular, the better prognosis in FASSc, compared with CFA, has not been evaluated in relation to the new histopathologic classification.
The aims of this retrospective study were to examine histopathologic subsets of interstitial pneumonia in systemic sclerosis patients with pulmonary involvement (FASSc), and to compare the prognostic value of histologic classification with that of clinical indices, including pulmonary function tests and bronchoalveolar lavage (BAL) cellularity.
Between January 1985 and December 1997, 476 patients with systemic sclerosis were investigated at the Royal Brompton Hospital, with surgical lung biopsies performed in 94 patients. It was unit policy that biopsies should be taken when possible, if interstitial lung disease was thought to be clinically significant. In 80 patients, clinical information and histopathologic material was available. Interstitial lung disease was evident on computed tomography (CT) in 78 cases; in two cases, biopsied before CT, pulmonary fibrosis was present on chest radiography. Patients fulfilled American Rheumatism Association preliminary criteria for systemic sclerosis (16); scleroderma was classified as limited or diffuse (17). Patients with overlap syndromes or environmental causes of pulmonary fibrosis (identified by a standardized history) were excluded. Current, ex- (history of smoking at least one cigarette/day for a minimum of 1 year, cessation for at least 3 months), and nonsmokers were identified. Treatment constituted a minimum daily dose of prednisolone of 10 mg, with or without an immunosuppressive agent, or D-penicillamine. Follow-up was obtained to death or to January 31, 2001 (censored in seven cases). The study was approved by the local Ethics Committee.
Pulmonary function tests (PFT), performed using ATS guidelines (18), were expressed as percentage predicted values (19). Spirometric volumes (Ohio water-sealed spirometer; Ohio Instruments, Atlanta, GA) plethysmographic volumes (whole body plethysmograph; Fenyvens and Gut, Basel, Switzerland) and carbon monoxide diffusing capacity (DlCO) (single-breath technique, PK Morgan respirometer; PK Morgan Ltd, Chatham, Kent, UK) were evaluated. “Significant” changes in PFT were defined using ATS and European Respiratory Society (ERS) criteria (20): ⩾ 10% change in VC (or at least ⩾ 200-ml change), ⩾ 15% change in single-breath DlCO (or at least 3 ml/minute/mm Hg). Lesser changes were designated “marginal.” PFTs were categorized to pathophysiologic patterns, per ERS recommendations (19).
BAL was performed as part of routine clinical management, according to recommended guidelines (21) and previous reports (22, 23). Abnormal BAL percentage values were: neutrophils > 4%; eosinophils > 2%; lymphocytes > 14%.
Sections stained with haematoxylin and eosin, and elastin van Gieson, were evaluated independently by two histopathologists blinded to clinical data (A. G. N., T. V. C.), according to recognized patterns of interstitial pneumonia (7, 11); differences were resolved by joint review. UIP was distinguished by temporal heterogeneity (variation in age of patchy fibrosis, ranging from fibroblastic proliferation, with abundant plump spindle cells and little intervening collagen, to established fibrosis, characterized by poorly cellular hyalinized collagen). In NSIP, fibrosis and/or inflammation was patchy or diffuse, but temporally uniform; cases were subdivided into cellular NSIP (cellular interstitial pneumonia, little or no fibrosis) and fibrotic NSIP (mature collagen, absent or sparse fibroblasts, with or without derangement of architecture) (15). Appearances of uniform honeycombing and complete loss of lung architecture were designated “end-stage lung.”
Analysis was performed using STATA software (Stata Corp., Santa Monica, CA). Data are expressed as means (SD), or medians (ranges), depending on distribution. Group comparisons were made using the Wilcoxon rank-sum test, or chi-squared statistics. Survival was evaluated using proportional hazards regression (24). A p value < 0.05 was considered significant.
Eighty patients undergoing open or thoracoscopic biopsy made up the study population; 67 were white and 13 were Indian or Afro-Caribbean. The kappa coefficient of agreement between the two histopathologists, for the five histologic subgroups (shown in Table 1)
Histologic Subset | No. of Subjects | Type of Scleroderma (Limited/Diffuse) | Mean Duration of Dyspnea at Biopsy (mo) |
|---|---|---|---|
| NSIP | 62 (77.5%) | 43/19 | 11 |
| UIP | 6 (7.5%) | 4/2 | 28 |
| ESL | 6 (7.5%) | 5/1 | 24 |
| Miscellaneous* | 6 (7.5%) | 4/2 | 12 |
Clinical information in the 74 patients with NSIP or UIP/ESL is shown in Table 2
Age at presentation, median (SD), | |
|---|---|
| yr (range, yr) | 46 ± 11 (23–69) |
| Sex: M/F ratio | 13/61 |
| Smoking history | |
| Current smokers | 4 (5%) |
| Exsmokers | 21 (28%) |
| Nonsmokers | 49 (66%) |
| Duration of respiratory symptoms before | 13 ± 13 |
| referral (mo: mean ± SD, range) | (0–60) |
| Type of scleroderma | |
| Limited | 52 (70%) |
| Diffuse | 22 (30%) |
| Respiratory history | |
| Chronic exertional dyspnea | 66 (89%) |
| Chronic cough | 26 (35%) |
| No respiratory complaints | 8 (11%) |
| Respiratory findings | |
| Bilateral predominantly basal crackles | 63 (85%) |
| Finger clubbing | 2 (3%) |
At presentation, spirometric and plethysmographic volumes were normal in 23 patients, with abnormalities seen in 51 patients (restrictive pattern, n = 42; obstructive pattern, n = 2; mixed pattern, n = 7). The percentage predicted DlCO (mean ± SD = 51.2 ± 14.5%) was decreased in all but two patients. The percentage predicted FVC (mean ± SD = 75.6 ± 19.5%) was decreased in 43 of 74 patients (58%). Overall, only one patient had normal initial FVC and DlCO levels (> 80% predicted). Among 43 patients with abnormal initial FVC levels, the proportion with NSIP (n = 35, 81%) and UIP/end-stage lung disease (ESL) (n = 8, 19%) did not differ from the proportions of NSIP and UIP/ESL in the whole population. Patients with UIP/ESL had lower DlCO levels than those with NSIP (41.8% ± 11.1% versus 53.1% ± 14.4%, p = 0.01), but FVC levels did not differ significantly between the two subgroups (68.0 ± 28.4% in UIP versus 77.1 ± 17.3% in NSIP, p = 0.14). DlCO and FVC levels at presentation did not differ significantly between cellular and fibrotic NSIP.
BAL data at first presentation were available in 67 of 74 patients with UIP/ESL or NSIP (Table 3
UIP/ESL | NSIP | Cellular NSIP | Fibrotic NSIP | |
|---|---|---|---|---|
| Subjects, n | 10 | 57 | 12 | 45 |
| Alveolar macrophages | 82.5 | 78 | 76.5 | 79 |
| 28–97 | 46–95 | 60–92 | 46–95 | |
| Lymphocytes | 6 | 8 | 13.5 | 6 |
| 1–22 | 0–45 | 6–30 | 0–45 | |
| Neutrophils | 5 | 5 | 2.5 | 6 |
| 1–55 | 1–41 | 1–12 | 1–41 | |
| Eosinophils | 2.5 | 4 | 3 | 5 |
| 0–4 | 0–19 | 0–10 | 0–19 |
Figure 1. Percentage neutrophil and eosinophil counts in 73 patients undergoing BAL. A BAL neutrophilia and/or eosinophilia was seen in the majority of patients.
[More] [Minimize]
Figure 2. Percentage lymphocyte differential counts in 57 patients with NSIP. BAL lymphocyte counts were higher in cellular than in fibrotic NSIP, p = 0.007.
[More] [Minimize]Twenty-two patients (30%) died during a median follow-up of 74.5 months (range, 1–16 years), including 16 of 62 patients with NSIP (26%), and 6 of 12 patients with UIP/ESL (50%; UIP, n = 4; ESL, n = 2). Four patients (5%) developed lung cancer during follow-up (adenocarcinoma, n = 3; small cell lung carcinoma, n = 1); one patient was an exsmoker and three were nonsmokers (all four patients died of the malignancy). Proportional hazards analysis showed that mortality increased with decreasing initial FVC levels (p = 0.007) and decreasing initial DlCO levels (p = 0.004), but did not vary with age, sex, or smoking status (which did not differ significantly between the two groups), or BAL findings (percentage neutrophil, eosinophil, or lymphocyte counts). The pattern of cutaneous involvement (limited versus diffuse scleroderma) was not linked to outcome. Survival did not differ between NSIP and UIP/ESL (p = 0.33), and this finding persisted after controlling for age, sex, and initial FVC and DlCO levels (analyzed in separate models), and when NSIP patients were compared separately with UIP patients (four deaths) and with ESL patients (two deaths). As shown in Figure 3
Figure 3. Survival compared between NSIP and UIP/ESL. No significant difference was disclosed.
[More] [Minimize]Survival did not differ significantly between cellular and fibrotic NSIP. Decreasing DlCO levels at presentation were associated with increasing mortality (p = 0.04) but age, sex, and FVC levels were not linked to survival. Survival was re-examined with the inclusion of BAL data, in the 57 patients undergoing BAL. There was a significant association between increasing BAL eosinophil levels and increasing mortality (p = 0.03) (Figure 4)
Figure 4. Survival in NSIP in 57 patients with (n = 36) and without (n = 21) increased BAL percentage eosinophil counts. Mortality was positively linked to increasing BAL eosinophil percentages, p = 0.03.
[More] [Minimize]As shown in Table 4
FVC (1 yr) (n = 61) | DlCO (1 yr) (n = 61) | FVC (3 yr) (n = 57) | DlCO (3 yr) (n = 57) | |
|---|---|---|---|---|
| Overall change (median) | −1.7% | −0.3% | −2.5% | −1.8% |
| Overall change (range) | −25.2 to 30.7% | −43.6 to 52.7% | −29.6 to 64.2% | −44.2 to 43.4% |
| Significant improvement | n = 10 | n = 9 | n = 13 | n = 12 |
| Significant decline | n = 15 | n = 8 | n = 11 | n = 8 |
| Change not significant | n = 37 | n = 44 | n = 33 | n = 37 |
When pulmonary function trends were evaluated using proportional hazards analysis, changes in FVC at 1 and 3 years, and changes in DlCO at 1 year, were not predictive of survival. However, changes in DlCO at 3 years were linked to survival, both from the date of initial PFT (p = 0.002) and from the date of follow-up PFT at 3 years (p = 0.003) (Figure 5)
Figure 5. Survival from the date of follow-up pulmonary function tests at 3 years in 57 patients with NSIP. Survival was highest in patients with a significant (DlCO +, n = 12) or marginal (DlCO +/−, n = 12) improvement in DlCO, and higher in patients with a marginal decline in DlCO (DlCO −/+, n = 25) than in those with a significant decline in DlCO (DlCO −, n = 8).
[More] [Minimize]In this study, we describe the histopathologic subsets of FASSc in a large series of patients. The most interesting finding was the high prevalence of NSIP. Histologic subsets have been described in association with dermatomyositis (25, 26) and rheumatoid arthritis (27), but have not been well characterized in pulmonary fibrosis associated with systemic sclerosis. In terms of the histologic pattern of NSIP, the series of Katzenstein and Fiorelli had connective-tissue diseases in only 10 of 64 patients (16%), including patients with rheumatoid arthritis, systemic lupus erythematosus, systemic sclerosis, polymyositis, and Sjogren's syndrome (15). By contrast, the prevalence of NSIP in idiopathic interstitial pneumonia is now well documented. Bjoraker and coworkers (10) in a retrospective analysis of 104 patients with a previous diagnosis of IPF found that only 14% had NSIP compared with 62% with UIP. However, Nagai and colleagues (13) found that 31 of 111 patients (28%) presenting with idiopathic interstitial lung disease had NSIP, although many patients had the radiographic features of organizing pneumonia, rather than IPF. Recently, NSIP was identified by Travis and coworkers (11) in 29% of patients with idiopathic interstitial lung disease and by Nicholson and colleagues (12) in 35% of patients presenting with cryptogenic fibrosing alveolitis. Thus, NSIP appears to be substantially more prevalent in systemic sclerosis than in idiopathic disease.
The problem of selection bias is common to all histopathologic series of patients with interstitial lung disease. Specifically, it was theoretically possible that in the present series, the performance of lung biopsy might have been reserved for patients with atypical CT findings, leading to an overestimation of the frequency of NSIP. However, it is unlikely that this was a major factor because the prevalence of NSIP was virtually identical before and after December 1990. The distinction between typical and atypical UIP appearances using CT was not made in our unit before 1991 and was not made with confidence until several years later. Broadly, it was unit policy to biopsy patients when possible, if they were considered to have “significant interstitial lung disease,” based on symptoms and pulmonary function indices. The largest groups of patients excluded were those with no evidence of lung involvement and a larger subgroup who were considered to have trivial or minor pulmonary fibrosis, this conclusion being reached on a patient-by-patient basis. Although patients too compromised to undergo biopsy (including those with severe pulmonary vascular disease) and others who declined biopsy were necessarily excluded, these amounted to small subgroups. The authors were not aware of basing the decision to biopsy on atypical as opposed to typical radiologic appearances, although minor bias cannot be wholly excluded.
Our findings differ from those in idiopathic interstitial pneumonia in one further respect. Idiopathic NSIP has a better survival than idiopathic UIP (9–12), although differences in outcome were more marked in earlier series (9, 10) than in recent reports (11, 12). In the present study survival did not differ significantly between NSIP and UIP/ESL, with both subgroups having a relatively good outcome (5-year survival exceeding 80%) compared with that documented in idiopathic UIP. However, this observation must be interpreted with great caution. Even with the inclusion of six patients with ESL, there were only 12 patients in the UIP/ESL subgroup. ESL has not been classified separately in previous studies of IPF; however, although appearances of ESL are closer to UIP than to NSIP, it is not known whether survival in ESL and UIP are equivalent. It is also possible that “lead time bias” (due to earlier evaluation than in idiopathic disease, due to the presence of systemic disease) obscured differences in outcome between UIP/ESL and NSIP. However, the functional severity of disease was adjusted for in analysis and mean DlCO levels at presentation in UIP patients in the present study (41.8% predicted) were virtually identical to those previously reported at presentation in idiopathic UIP patients at our institution (43.5% predicted) (12).
A further problem is the difficulty in identifying the contribution made by interstitial lung disease to overall mortality. The development of lung cancer may be ascribable to pulmonary fibrosis; patients with systemic sclerosis are known to have an increased risk of cancer (28–30), in contrast to patients with idiopathic NSIP, in which a malignant predisposition has yet to be reported. Genetic abnormalities at the level of microsatellite DNA have been found in BAL specimens of patients with FASSc (31). Loss of pulmonary reserve due to interstitial lung disease may contribute to mortality from pulmonary vascular and cardiac disease. In addition, if advanced pulmonary fibrosis is associated with more severe systemic disease, the severity of lung disease may be a marker of nonrespiratory mortality. For all these reasons, analyses of outcome must consist of analyses of overall mortality, as in studies of IPF, rather than mortality primarily due to progression of respiratory disease, although this may obscure prognostic distinctions between UIP and NSIP.
Thus, the finding that mortality in NSIP did not differ from UIP or from UIP/ESL in the present study must be viewed as inconclusive, pending further large histopathologic studies. The more impressive finding was the lack of major decline in serial pulmonary function tests over a 3-year period after presentation in patients with NSIP. It has recently been reported that cyclophosphamide therapy is associated with a substantial reduction in the prevalence of a poor respiratory outcome in systemic sclerosis (32). A number of markers of lung disease were linked to outcome in the present study, both in the whole cohort (DlCO and FVC levels at presentation) and in patients with NSIP (DlCO levels at presentation, BAL eosinophil percentages, and changes in DlCO during the first 3 years of follow-up). However, these findings do not necessarily indicate that the severity and progression of parenchymal lung disease contributed to eventual mortality. Severe depression of DlCO may denote pulmonary vascular disease rather than advanced pulmonary fibrosis. In the present study, 3-year trends in FVC levels had no long-term prognostic significance, suggesting that interstitial lung disease in systemic sclerosis may be arrested or slowed by treatment, and that the link between mortality and DlCO trends might have reflected progression of pulmonary vascular disease.
The observation that increased BAL eosinophil levels were associated with a higher mortality in NSIP is also difficult to interpret. In previous studies of systemic sclerosis (32, 33), a BAL neutrophilia was associated with subsequent declines in pulmonary function tests in untreated patients. In the present study, increases in BAL neutrophil and eosinophil percentages were seen in the majority of patients with NSIP, but were not predictive of subsequent functional trends, despite the previous observation that a BAL neutrophilia is associated with extensive fibrotic disease on HRCT in systemic sclerosis (22). It appears that BAL abnormalities may not identify future progression of pulmonary fibrosis, provided that treatment is instituted. One possible explanation of the link between BAL eosinophil levels and mortality is that a BAL eosinophilia may denote more aggressive systemic disease. Although this speculation is unproven, eosinophilic infiltration is well recognized in systemic sclerosis in extrapulmonary sites (34).
The interobserver agreement between pathologists was higher than in two previous series of idiopathic interstitial pneumonia, classified by the same two pathologists (12, 35), reflecting, in part, increasing concordance of diagnostic criteria after discussion of discrepant observations in earlier studies. However, in the view of the observers the higher kappa coefficient of agreement in the present study also reflected the paucity of biopsies with very severe fibrotic disease. The difficulty in discriminating between UIP and advanced NSIP3 was a major source of observer variation in earlier work (12), reinforcing the designation of a separate category of “end-stage lung.” In previous studies (12, 35), these appearances were categorized as UIP, and were grouped with UIP in the present study for analyses of outcome. However, the classification of end-stage lung disease remains problematic; we agree with the suggestion (7) that a term such as “end-stage lung” or “interstitial fibrosis with honeycomb change” is appropriate.
It is unclear whether some patients with NSIP progress to ESL or even to UIP, as has been suggested, based on the finding in idiopathic disease that NSIP and UIP may be found in different biopsy sites in the same patients (36, 37). Progression from NSIP to UIP might explain the subgroup of patients with idiopathic NSIP with a poor outcome (12), but appears less likely to occur in systemic sclerosis judging from the low prevalence of UIP in the present series. However, progression from NSIP to UIP in a small subset of patients, cannot be wholly excluded; the significantly lower gas transfer levels in UIP patients would be compatible with this possibility.
In conclusion, we have found that the great majority of patients with fibrosing alveolitis associated with systemic sclerosis have a histologic pattern of NSIP rather than UIP, in contrast to patients with idiopathic interstitial pneumonia. However, although a number of markers of lung disease were linked to survival, the histopathologic distinction between cellular and fibrotic NSIP had no prognostic significance, and there was little overall functional decline in treated patients with NSIP in the 3 years after presentation.
The authors thank Dr. Katerina Antoniou for her assistance in data collection.
| 1. | Minai OA, Dweik RA, Arroliga AC. Manifestations of scleroderma pulmonary disease. Clin Chest Med 1998;19:713–731. |
| 2. | D'Angelo WA, Fries JF, Masi AT, Shulman LE. Pathologic observations in systemic sclerosis (scleroderma): a study of fifty-eight autopsy cases and fifty-eight matched controls. Am J Med 1969;46:428–440. |
| 3. | Arroliga AC, Podel DN, Mathay RA. Pulmonary manifestations of scleroderma. J Thorac Imaging 1992;7:30–45. |
| 4. | 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. |
| 5. | Wells AU, Hansell DM, Rubens MB, Cailes J, 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;156:1657–1664. |
| 6. | Katzenstein AL, Myers JL. Idiopathic pulmonary fibrosis: clinical relevance of pathologic classification. Am J Respir Crit Care Med 1998; 157:1301–1315. |
| 7. | Katzenstein ALA, Myers JL. Nonspecific interstitial pneumonia and the other idiopathic interstitial pneumonias: classification and diagnostic criteria. Am J Surg Pathol 2000;24:1–3. |
| 8. | Bouros D. Current classification of idiopathic interstitial pneumonias. Monaldi Arch Chest Dis 2000;55:450–454. |
| 9. | Daniil ZD, Gilchrist FG, Nicholson AG, Hansell DM, Harris J, Colby TV, du Bois RM. A histologic pattern of nonspecific interstitial pneumonia is associated with better prognosis than usual interstitial pneumonia in patients with cryptogenic fibrosing alveolitis. Am J Respir Crit Care Med 1999;160:899–905. |
| 10. | Bjoraker JA, Ryu JH, Edwin MK, Myers JL, Tazelaar HD, Schroeder DR, Offord KP. Prognostic significance of histopathologic subsets in idiopathic pulmonary fibrosis. Am J Respir Crit Care Med 1998;157: 199–203. |
| 11. | Travis WD, Matsui K, Moss J, Ferrans VJ. Idiopathic nonspecific interstitial pneumonia: prognostic significance of cellular and fibrosing patterns: survival comparison with usual interstitial pneumonia and desquamative interstitial pneumonia. Am J Surg Pathol 2000;24:19–33. |
| 12. | Nicholson AG, Colby TV, du Bois RM, Hansell DM, Wells AU. The prognostic significance of the histological pattern of interstitial pneumonia in patients presenting with the clinical entity of cryptogenic fibrosing alveolitis. Am J Respir Crit Care Med 2000;162:2213–2217. |
| 13. | Nagai S, Kitaichi M, Itoh H, Nishimura K, Izumi T, Colby TV. Idiopathic nonspecific interstitial pneumonia/fibrosis: comparison with idiopathic pulmonary fibrosis and BOOP. Eur Respir J 1998;12:1010–1019. |
| 14. | Cottin V, Donsbeck AV, Revel D, Loire R, Cordier JF. Nonspecific interstitial pneumonia. Individualization of a clinicopathologic entity in a series of 12 patients. Am J Respir Crit Care Med 1998;158:1286–1293. |
| 15. | Katzenstein AL, Fiorelli RF. Non-specific interstitial pneumonia/fibrosis. Histologic features and clinical significance. Am J Surg Pathol 1994;18:136–147. |
| 16. | 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. |
| 17. | LeRoy EC, Black C, Fleischmajer R, Jablonska S, Krieg T, Medsger TA Jr, Rowell N, Wollheim F. Scleroderma (systemic sclerosis): classification, subsets and pathogenesis. J Rheumatol 1988;15:202–205. |
| 18. | American Thoracic Society. Standardization of spirometry, 1994 update. Am J Respir Crit Care Med 1995;152:1107–1136. |
| 19. | Quanjer PH, Tammeling GJ, Cotes JE, Pedersen OF, Peslin R, Yernault JC. Lung volumes and forced ventilatory flows: report Working Party Standardisation of Lung Function Tests. European Community for Steel and Coal: official statement of the European Respiratory Society. Eur Respir J 1993;6:5–40. |
| 20. | American Thoracic Society. Idiopathic pulmonary fibrosis: diagnosis and treatment. International consensus statement. Am J Respir Crit Care Med 2000;161:646–664. |
| 21. | Technical recommendations and guidelines for bronchoalveolar lavage (BAL). Report of the European Society of Pneumology Task Group. Eur Respir J 1989;2:561–585. |
| 22. | 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. |
| 23. | Wells AU, Hansell DM, Rubens MB, Cullinan P, Haslam P, Black CM, du Bois RM. Fibrosing alveolitis in systemic sclerosis: bronchoalveolar lavage findings in relation to computed tomographic appearance. Am J Respir Crit Care Med 1994;150:462–468. |
| 24. | Cox DR, Oakes D. Analysis of survival data. London: Chapman and Hall Ltd; 1988. |
| 25. | Tazelaar HD, Viggiano RW, Pickersgill J, Colby TV. Interstitial lung disease in polymyositis and dermatomyositis: clinical features and prognosis as correlated with histologic findings. Am Rev Respir Dis 1990;141:727–733. |
| 26. | Douglas WW, Tazelaar HD, Hartman TE, Hartman RP, Decker PA, Schroeder DR, Ryu JH. Polymyositis-dermatomyositis-associated interstitial lung disease. Am J Respir Crit Care Med 2001;164:1182–1185. |
| 27. | Hakala M, Paakko P, Huhti E, Tarkka M, Sutinen S. Open lung biopsy of patients with rheumatoid arthritis. Clin Rheumatol 1990;9:452–460. |
| 28. | Rosenthal AK, McLaughlin JK, Gridley G, Nyren O. Incidence of cancer among patients with systemic sclerosis. Cancer 1995;76:910–914. |
| 29. | Roumm AD, Medsger TA Jr. Cancer and systemic sclerosis: an epidemiological study. Arthritis Rheum 1985;28:1336–1340. |
| 30. | Bouros D, Hatzakis K, Labrakis H, Zeibecoglou K. Association of malignancy with disease causing interstitial pulmonary changes. Chest (In press) |
| 31. | Vassilakis D, Pantelidis P, Bouros D, Siafakas NM, du Bois RM. Evidence of genetic alterations in BAL cells of patients with fibrosing alveolitis-associated with scleroderma. Eur Respir J 2000;14:526s. |
| 32. | White B, Moore WC, Wigley FM, Xiao HQ, Wise RA. Cyclophosphamide is associated with pulmonary function and survival benefit in patients with scleroderma and alveolitis. Ann Intern Med 2000;132:947–954. |
| 33. | Behr J, Vogelmeier C, Beinert T, Meurer M, Krombach F, Konig G, Fruhmann G. Bronchoalveolar lavage for evaluation and management of scleroderma disease of the lung. Am J Respir Crit Care Med 1996;154:400–406. |
| 34. | Gustafsson R, Fredens K, Nettelbladt O, Hallgren R. Eosinophil activation in systemic sclerosis. Arthritis Rheum 1991;34:414–422. |
| 35. | MacDonald SL, Rubens MB, Hansell DM, Copley SJ, Desai SR, du Bois RM, Nicholson AG, Colby TV, Wells AU. Nonspecific interstitial pneumonia and usual interstitial pneumonia: comparative appearances at and diagnostic accuracy of thin-section CT. Radiology 2001; 221:600–605. |
| 36. | Flaherty KR, Travis WD, Colby TV, Ooews GB, Kazerooni EA, Gross BH, Jain A, Strawderman RL, Flint A, Lynch JP. Martinez FJ. Histopathologic variability in usual and nonspecific interstitial pneumonias. Am J Respir Crit Care Med 2001;164:1722–1727. |
| 37. | Nicholson AG, Wells AU. Nonspecific interstitial pneumonia: Nobody Said It's Perfect. Am J Respir Crit Care Med 2001;164:1553–1554. |
