INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Fibrosing alveolitis is a progressive lung disease with a poor prognosis. It can be idiopathic or associated with various autoimmune diseases. A typical feature of this disease entity is the activation of inflammatory cells in the lungs, increased production of cytokines and growth factors, and accumulation of collagen, which leads to fibrotic lung damage and impairment of gas exchange. Prediction of the progression of the disease is still controversial and poorly understood.

Collagens are synthesized primarily by fibroblasts as procollagen precursor molecules, and propeptides are cleaved in stoichiometric amounts during the secretion of the newly formed collagens (1). Pulmonary fibrosis is initially associated with increased accumulation of types III and I collagens, and subsequently mainly with increased type I collagen (2). Previous studies have suggested that propeptides of these collagens could be used in the assessment of interstitial lung diseases. The amino-terminal propeptide of type III procollagen (PIIINP) is the most widely investigated marker of synthesis of type III collagen, and in some instances of degradation of type III collagen (3). The carboxy-terminal propeptide of type I procollagen (PICP), on the other hand, reflects synthesis but not degradation of type I collagen fibers (8), and it may provide important information about the fibrotic process in the lungs.

The concentration of the amino-terminal propeptide of type III procollagen is increased during such conditions as induced wound healing (12) and hepatic cirrhosis (13). It can also be detected in the bronchoalveolar lavage fluid (BALF) of patients with sarcoidosis and fibrosing alveolitis (6). In a recent study, we found that PIIINP was increased in the BALF and epithelial lining fluid (ELF) of patients with sarcoidosis, and that it was present at higher levels in parenchymal than in nonparenchymal sarcoidosis, and in symptomatic than in asymptomatic patients (5). Thus, PIIINP may be an important marker of altered collagen metabolism in interstitial lung diseases.

Type I procollagen propeptides have been visualized with monoclonal antibodies in the lungs of patients with active fibrosis, but not in healthy lung (14), and an increased amount of the messenger RNA (mRNA) for type I procollagen appears to be associated with foci of highly activated fibroblasts (15). To our knowledge, no previous studies have been conducted on PICP in the BALF or ELF of patients with fibrosing alveolitis.

The objective of the present study was to compare the levels of PIIINP and PICP in fibrosing alveolitis, and to test the hypothesis that the levels of these markers may predict the course of this disease. One of the main goals was to assess the role of PICP in fibrosing alveolitis, since it could theoretically be a superior indicator of pulmonary fibrosis. Because many studies have measured these procollagen markers in either serum or lavage fluid only, we compared the concentrations of PIIINP and PICP in BALF, ELF, and serum. We followed up our patients prospectively for an average of 6 yr, to better assess the role of these markers of collagen metabolism in fibrosing alveolitis.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Study Population

The study included 18 patients admitted to the pulmonary department of Päivärinne Hospital in Muhos, Finland, between February 1990 and August 1992, who underwent bronchoscopy and bronchoalveolar lavage for suspected fibrosing processes of the lungs. Sixteen patients were considered to have idiopathic fibrosing alveolitis, one patient had rheumatoid arthritis, and one patient had scleroderma, both with pulmonary involvement. The histologic diagnoses were confirmed by biopsy in nine patients and at autopsy in one patient. The diagnoses of the remaining eight patients were based on the findings typical of fibrosing alveolitis in chest radiographs, in BALF, upon spirometry and in diffusion capacity, and on the exclusion of other pulmonary diseases. The characteristics of the study population are presented in Table 1. The mean follow-up time was 6 yr and 2 mo.

Seventeen control subjects had been investigated for minor respiratory symptoms. Their chest radiographs, lung function parameters, and BALF cell profiles were regarded as normal (Table 1). The amino-terminal propeptide of type III procollagen in BALF was detectable in two of the 17 control subjects, whereas PICP was not detectable in any of these subjects' BALF samples (Table 2). The values for the control group have been previously published (5).

Informed consent was obtained from each subject before the study began, and the Ethics Committee of Päivärinne Hospital approved the research protocol.

Procedures

The chest radiographic examinations used in the study included posteroanterior (PA) and lateral views. The chest radiographs were first interpreted by a specialist in pulmonary medicine, and then retrospectively by an experienced thoracic radiologist.

Pulmonary function tests, including FEV₁ and FVC, were made with a flow-volume spirometer. Diffusion capacity of carbon monoxide (DL_CO) and the specific diffusion coefficient for carbon monoxide (DL_CO/VA) were analyzed with the single breath technique.

Albumin, urea, and alanine aminotransferase (ALT) were measured with standard laboratory methods. Serum levels of PIIINP and PICP were measured with commercial radioimmunoassay (RIA) kits (Orion-Diagnostica Oy, Oulunsalo, Finland) (16, 17), using human antigens and specific polyclonal antibodies.

Fiberoptic bronchoscopy was performed under local anesthesia with lignocaine. The patients were premedicated with diazepam and atropine. A fiberoptic bronchoscope was wedged into the right middle lobe bronchus or into the left lingula. Saline solution was installed in 10 aliquots of 20 ml each. The recovery of the lavage fluid is shown in Table 1. After centrifugation (400 × g for 15 min), the cell-free supernatant was stored at 20° C for the assay of PIIINP and PICP. The total number of cells was counted in a Bürger hemocytometer, and cell viability was assessed by trypan blue exclusion as previously described (5).

For the RIA of PIIINP and PICP in BALF, we modified the original methods for serum assessment of these procollagens to make any further processing of the sample unnecessary (5). The results are expressed either as micrograms of propeptide per liter of BALF or ELF, as estimated by the urea method (18). The sensitivities of the respective assays are 0.2 µg/L in the serum sample of 200 µl for PIIINP and 1.2 µg/L in the serum sample of 100 µl for PICP. The intra- and interassay variations for PIIINP are 4.3% and 5.3%, and those for PICP are 3.2% and 6.6%, respectively.

Statistical Analysis

Nonparametric tests were used in the statistical evaluation of study data. Correlations were assessed with Spearman's correlation coefficient test, and the differences between groups were assessed with the Mann-Whitney U test and Fisher's exact probability test. Values of p  0.05 were considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We assayed the levels of PIIINP and PICP in BALF, ELF, and serum (Table 2, Figure 1). Two patients with fibrosing alveolitis had elevated serum ALT values, suggesting the possibility of a hepatic lesion. Because hepatic diseases might cause increases in serum PIIINP, the values for these two patients were excluded. The levels of PIIINP in serum (p < 0.05), BALF (p < 0.05), and ELF (p < 0.05), and the levels of PICP in BALF (p < 0.001) and ELF (p < 0.001), were higher in the patients with fibrosing alveolitis than in the controls. The level of PICP in serum did not differ significantly between the controls and the patients. PIIINP was detectable in the BALF of nine of 18 patients and two of 17 controls, and the corresponding figures for PICP were nine of 18 and none of 17. In fibrosing alveolitis, the average concentration of PIIINP in ELF was 13-fold greater and that of PICP was more than 4-fold greater than in serum.

View larger version (17K): [in this window] [in a new window]   Figure 1.   The individual and median values of PIIINP and PICP in BALF and ELF in patients with fibrosing alveolitis and in controls.

In fibrosing alveolitis, PICP in BALF (r = 0.65, p < 0.01) and ELF (r = 0.59, p < 0.05), but not PIIINP in BALF or ELF, had a significant negative correlation with DL_CO/VA. Levels of PIIINP or PICP in BALF or ELF did not correlate with FVC. The serum levels of PIIINP and PICP did not correlate with those in BALF or ELF. The concentrations of PIIINP and PICP in BALF correlated significantly with each other (r = 0.60, p < 0.01).

During the follow-up period, seven of 18 patients died of fibrosing alveolitis, three died of malignancy, and one died of an unknown cause. Detectable PIIINP in BALF had no prognostic significance if all the deaths were included, but predicted a poor prognosis in fibrosing alveolitis (Table 3). DL_CO, when assessed either in absolute or percentage values, was lower in the patients who died of pulmonary fibrosis (4.24 ± 1.3 mmol/min/kPa, 64.0 ± 8.3%) than in those who survived (5.81 ± 2.0 mmol/min/kPa, 85.7 ± 26.3%) (p < 0.05, Mann- Whitney U test). DL_CO/VA did not show any statistical difference between the two groups.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Our study shows that especially PICP in ELF and BALF, and to a lesser degree, PIIINP in BALF and serum, but not PICP in serum, are increased in fibrosing alveolitis, and can be used in the assessment of pulmonary involvement in patients with this disease.

The role of PIIINP in parenchymal lung diseases has been assessed in numerous studies, most of them conducted on patients with sarcoidosis, but the results even in sarcoidosis are contradictory. Milman and colleagues found no differences between BALF-PIIINP values in sarcoidosis patients and controls (10), and Low and associates reported no correlation of BALF-PIIINP with the clinical severity of sarcoidosis (6). Similarly, O'Connor and coworkers suggested that BALF- PIIINP is a poor indicator of the course and prognosis of sarcoidosis (11), and the results of Luisetti and colleagues indicated that PIIINP in serum is unable to predict the prognosis of patients with sarcoidosis (7). On the other hand, Pohl and associates showed that serum PIIINP was higher in patients with active sarcoidosis (9). Our recent study, in which levels of PIIINP in BALF and ELF were increased in sarcoidosis, showed these values to be higher in parenchymal than in nonparenchymal disease, but whether this finding was associated with fibrosis remained unclear (5).

Few studies of PIIINP in fibrosing alveolitis have been reported, and even their results are controversial. Asbestos exposure may be associated with the development of pulmonary fibrosis, and PIIINP has been shown to be increased in the BALF of asbestos-exposed sheep during the first months of exposure (19). Cavalleri and colleagues showed that serum levels of PIIINP were higher in workers who developed radiologically demonstrated asbestos-related interstitial fibrosis (20). On the other hand, it would appear that serum levels of PIIINP do not predict the development of pneumoconiosis in coal workers (21). PIIINP has been found to be significantly increased in the serum and BALF of patients with idiopathic fibrosing alveolitis (6, 22), scleroderma (25, 26), and rheumatoid arthritis (27) with pulmonary involvement, but the correlation between the severity of the disease and BALF-PIIINP has been variable (4, 6, 24, 28). In the present study, PIIINP was increased in the BALF and ELF of the patients with fibrosing alveolitis, but when these concentrations were compared with those in our recent study, they were significantly higher in sarcoidosis (8.7 µg/L and 785 µg/L, respectively) (5) than in fibrosing alveolitis (1.9 µg/L and 48.5 µg/L, respectively). These results are in good agreement with those of previous studies, which suggest that BALF-PIIINP reflects active inflammation but does not predict a fibrotic process in interstitial lung diseases (6, 7, 9, 11). Even though serum levels of PIIINP were significantly greater in our patients than in the controls, they did not correlate with BALF-PIIINP, suggesting that serum PIIINP cannot be recommended in the assessment of collagen metabolism in lung diseases. Our control group, which we have recently described (5), included patients with minor respiratory symptoms. They were younger than the patients this study. The results are, however, in agreement with those in one previous study in which BALF-PIIINP was found to be low or nondetectable in individuals without pulmonary involvement (29).

One of our goals was to test whether PIIINP can be used as a prognostic marker in patients with fibrosing alveolitis. The level of PIIINP in BALF has been shown to be increased in patients with adult respiratory distress syndrome (ARDS) who develop intraalveolar fibrosis (30), and recent studies have shown that a high concentration of PIIINP in BALF (31) or edema fluid (32) predicts pulmonary edema and a poor prognosis. Our patients were prospectively followed for an average of 6 yr. We are not aware of any corresponding studies. In our relatively small series, BALF-PIIINP was associated with active disease and a poor prognosis.

Very few studies have been done of PICP in lung parenchymal diseases, and PICP has not previously been assessed in BALF or ELF in fibrosing alveolitis. The carboxy-terminal propeptide of type I collagen in BALF or ELF is theoretically a better marker of lung collagen deposition than is PIIINP, since it may reflect ongoing collagen synthesis but not collagen degradation or inflammation (8). In the present study, PICP in BALF and ELF correlated negatively with DL_CO/VA, which may indicate an association between increased interstitial collagen deposition and the level of PICP in BALF. Furthermore, we found more than 4-fold higher concentrations of PICP in ELF than in serum, indicating active synthesis of type I procollagen in the lungs.

The role of S-PICP in various lung diseases is controversial. Bacchella and coworkers found that serum levels of PICP had no role in pulmonary sarcoidosis (33), whereas Kikuchi and associates reported increased serum levels of PICP in scleroderma patients with diffuse lung involvement (34). Our present and previous studies indicate that the serum level of PICP is notably similar in controls and in patients with sarcoidosis (5) and fibrosing alveolitis, and also that PICP in serum and BALF do not correlate with each other. Furthermore, a major part of serum-PICP is derived from bone (17). We therefore conclude that serum-PICP is a poor marker of disease activity in fibrosing alveolitis.

The present study shows that the procollagen propeptides PIIINP and PICP as measured in BALF and ELF, but not in serum, can be used in the assessment of fibrotic lung diseases. The PIIINP in BALF and ELF can be useful as a marker of disease activity, and may predict a poor prognosis in fibrosing alveolitis. By comparison, the level of PICP in BALF and ELF suggests collagen deposition but does not predict the course of the latter disease.