Print version ISSN 0102-3586
J. Pneumologia vol.29 no.6 São Paulo Nov./Dec. 2003
Histological features and survival in idiopathic pulmonary fibrosis*
Ester Nei Aparecida Martins ColettaI, Carlos Alberto de Castro PereiraII(TE SBPT); Rimarcs Gomes FerreiraIII; Adalberto Sperb RubinIV, Lucimara Sonja VillelaV, Tatiana MalheirosVI(TE SBPT); João Norberto StávaleIII
in Pathology and Assistant Professor in the Pathology Department of UNIFESP.
Doctor of Pathology at HSPE-SP
IIPh.D. and Doctor of Pulmonology at UNIFESP and in the Respiratory Diseases Department of HSPE-SP
IIIPh.D. and Assistant Professor in the Pathology Department of UNIFESP
IVPh.D. in Pulmonology and Doctor at the Pavilhão Pereira Filho, Santa Casa
VMedical Resident in the Pathology Department of HSPE-SP
VIPostgraduate student in Pulmonology at UNIFESP
Idiopathic pulmonary fibrosis was recently redefined as usual interstitial pneumonia
of unknown etiology. Consequently, the prognostic value of histological findings
needs to be reassessed.
OBJECTIVES: To correlate clinical, functional and histological findings with survival in patients with idiopathic pulmonary fibrosis.
METHODS: Fifty-one patients were evaluated. Mean age was 66 ± 8 years; 21 were female. There were 26 smokers or ex-smokers. Duration of symptoms, forced vital capacity, age, gender and smoking habits were recorded. All patients presented usual interstitial pneumonia verified through histology. Degree of honeycombing, established fibrosis, desquamation, cellularity, myointimal thickening of blood vessels and number of fibroblastic foci were graded according to the semiquantitative method.
RESULTS: Median duration of symptoms was 12 months and initial forced vital capacity was 72 ± 21%. Cox multivariate analysis revealed that survival correlated inversely and significantly (p 0.05) with duration of symptoms and fibroblastic foci score, as well as with myointimal thickening of blood vessels. Myointimal thickening involving less than 50% of blood vessels and limited numbers of fibroblastic foci were predictive of a greater survival. No correlation with survival was found for gender, age, forced vital capacity, inflammation or degree of cellularity.
CONCLUSION: Semiquantitative analysis of lung biopsies yields relevant prognostic information regarding patients with usual interstitial pneumonia.
Key words: Pulmonary fibrosis. Idiopathic pulmonary fibrosis. Interstitial lung disease.
in this paper:
FVC Forced vital capacity
UIP Usual interstitial pneumonia
NSIP Nonspecific interstitial pneumonia
IPF Idiopathic pulmonary fibrosis
CO Carbon monoxide
Since the publication of studies performed by Liebow in 1975, a variety of classifications have been proposed for the different types of idiopathic interstitial pneumonia.(1-3) Some authors have utilized the term idiopathic pulmonary fibrosis (IPF) to encompass the various manifestations.(2) However, in 2000, it was suggested that the term IPF be used to refer exclusively to those cases associated with usual interstitial pneumonia (UIP), as diagnosed by surgical biopsy.(4) In 1994, nonspecific interstitial pneumonia (NSIP) had been characterized as an entity apart from UIP.(5) The validity of this proposal has been confirmed by various studies, in which better prognoses were verified for NSIP than for UIP.(6-8)
A diagnosis of IPF is frequently accepted on the basis of clinical evidence. However, a definitive diagnosis cannot be obtained without surgical biopsy. The histological parameters considered essential to diagnosis include an interstitial process with heterogeneous distribution in time, presenting completely uncompromised areas of lung tissue juxtaposed with fibrotic areas, remodeling of the lung parenchyma and fibroblastic foci.(3,9,10)
In general, the evolution of an interstitial disease begins with an initial inflammation (alveolitis) and progresses into irreversible fibrosis. Based on this knowledge, it has been proposed that analysis of surgical biopsies should include quantification of the extent of inflammation and fibrosis.(11) In 1998, it was suggested that inflammation is irrelevant in IPF, and that the initial lesion is composed of fibroblastic foci.(3) In a recent study, the extent of granulation tissue, as well as that of interstitial "young connective tissue", was directly correlated with survival, whereas the degree of fibrosis was not.(12) Another study utilizing the same pathological scoring system, only the fibrosis score was predictive of survival.(13) The scoring system used in both studies scores fibroblastic foci both in the inflammation score (the "granulation tissue" score is the sum of such tissue found in alveoli, alveolar ducts and bronchioles) and in the fibrosis score, where they appear as "young connective tissue in the alveolar wall" and are added together with fibrosis characteristics, such as honeycombing. Another study, involving 85 patients, found no prognostic value in evaluating isolated fibroblastic foci in UIP.(14)
To the previously mentioned fibrosis score,(11) the authors of yet another study added vascular alterations (myointimal thickening) and smooth-muscle hyperplasia.(12) However, the importance of these isolated findings is debatable. Increased angiogenesis may promote fibrogenesis in IPF.(15) In addition, it has recently been suggested that the microvascular lesion is the initial event in the pathogenesis of IPF and may induce pulmonary fibrosis.(16) Extensive vascular remodeling occurs within areas of pulmonary fibrosis, which may indicate an advanced stage of the disease and a worse prognosis.
The objective of the present study was to determine the prognostic value of histological findings and other data, such as duration of symptoms, forced vital capacity (FVC), age, gender and smoking habits, correlated with survival in patients suffering from IPF as characterized by UIP pathological profile.
In the final analysis, 51 patients were included in this study. All were suffering from IPF as diagnosed by histological evaluation of biopsies and none had been previously treated for this condition. All cases originated from either the Escola Paulista de Medicina, the São Paulo Hospital do Servidor Público Estadual, or the Pavilhão Pereira Filho in Porto Alegre. After examination of the laminae, patients with other interstitial diseases (such as NSIP or hypersensitivity pneumonitis), patients who presented possible etiology for UIP (such as a history of hypersensitivity pneumonitis), or patients who suffered from a terminal lung disease, were excluded. Other exclusion criteria included collagen-related diseases, inadequate material or evolution data, possible etiology for UIP (such as a history of hypersensitivity pneumonitis), or death within the first 30 days after the biopsy. Following distribution analysis, three cases of comparable histology were also excluded due to discrepancies in length of survival (132, 179 and 191 months). In addition, two cases were excluded on the basis of other anomalous findings: precocity (38 years old) in one, and a disproportionate profusion of vascular lesions in the other.
Information regarding the deaths was obtained from the records departments of the institutions concerned through consultation with the São Paulo Municipal Program for the Improvement of Mortality Information (PRO-AIM).
In 11 cases, it was not possible to determine, retrospectively, the biopsy site. Of the remaining biopsies, 12 were obtained from the lingula and 28 from various other lobes.
Standard techniques involving hematoxylin and eosin (HE) staining and Masson trichrome staining were used to analyze the lung samples. Lung biopsies were examined by pathologists (Ester Nei Aparecida Martins and Rimares Gomes Ferreira) who have extensive experience in the diagnosis of interstitial pulmonary diseases and who were blinded as to the ends of the study. From the two reports, which included independent histological analysis of the laminae using the semiquantitative method,(11) mean values were calculated. When discrepancies occurred (defined as a greater than two point difference in a given score), a final consensus was reached. A scale from 1 to 10 (0 = absent; 1 = minimal; 2 = less than 25%; 3 = 25 to 49%; 4 = 50 to 75%; and 5+ = greater than 75%) was applied to the following alterations: extent and intensity of cellularity in the (interstitial) alveolar wall, as well as in the alveolar spaces (desquamation), "young" interstitial connective tissue in the alveolar walls (fibroblastic foci), established interstitial fibrosis (including honeycombing), cysts ("honeycomb" lung), metaplastic smooth muscle in the interstice (excluding smooth muscle associated with airways and vessels), and myointimal thickening of the vessels. The alveolar metaplasia score was not taken into account.
Final mean histological scores were assigned through analysis of various fields and subjective determination of predominance. The initial functional evaluation was made through measurement of FVC. Predicted values were based on data from a study of the general Brazilian population.(17) The survival curves were correlated with age, gender, duration of symptoms, smoking habits, FVC and histological findings. Survival was recorded until July 31, 2001.
The statistical analysis program SPSS version 10 (SPSS, Inc., Chicago, IL, USA) was used to compare data. Agreement between the independent analyses submitted by the two pathologists was quantified using the kappa coefficient. The resulting data has been previously published.(18) The distribution of the quantitative variables was verified using the Kolmogorov-Smirnov test. Means were compared with either Students t-test (for values displaying normal distribution) or the Mann-Whitney test (for those variables not meeting normality assumptions).
To investigate the relationships among the variables encompassed by the pathological score, the Spearman rank correlation coefficient (rho) was used. The associations among qualitative variables were evaluated using an x2 test. In order to evaluate the influence of the different variables on patient survival, Cox proportional hazards regression was utilized. Kaplan-Meier curves were employed to express general survival and variables transformed into categories, including various cutoff values. For curve comparison, the log-rank test was used. Values of p <0.05 were considered significant.
The median length of follow up was 26 months, ranging from 3 to 103 months. The overall findings and histological scores are shown in Tables 1 and 2. On average, the patients were elderly and presented moderate restrictive pulmonary dysfunction. Of necessity, all patient biopsies presented fibroblastic foci. All but one case evidenced honeycombing under microscopy.
Smokers or ex-smokers accounted for 26 of the 51 patients. These patients presented higher histology scores for desquamation (5.25 ± 1.58) than did the non-smokers (4.32 ± 0.84, p = 0.012). In other variables, including FVC, no differences were found between smokers and non-smokers.
Of the 51 patients studied, 25 (49%) died and 26 (51%) survived until the endpoint of the study. With the exception of one case, all of the deaths were attributed to fibrosis, either directly (respiratory insufficiency) or indirectly (pulmonary infection). Survival, estimated by the Kaplan-Meier survival curve, is depicted in Figure 1. Median survival was 48 months.
Histological findings correlated with survival were initially evaluated using Cox regression analysis of individual values. Only fibroblastic foci and myointimal thickening scores were found to correlate significantly (p < 0.05) with survival. The scores for honeycombing and established fibrosis (which includes honeycombing) correlated marginally with survival (p = 0.08 and p = 0.07, respectively).
Median survival for the group with a low fibroblastic foci scores (< 2, sparse foci) exceeded 70 months, compared with 29 months in the group (n = 37) with scores > 2 (logrank = 5.72, p = 0.017). Kaplan-Meier survival curves for these findings are shown in Figure 2. One typical fibroblastic focus is displayed in Figure 3. Figures 4a and 4b show representative examples of low (sparse) and high (profuse) fibroblastic scores. Fibroblastic foci score presented a direct correlation with FVC (rho = 0.43, p = 0.003).
For honeycombing, the best qualitative separation was obtained by differentiating between those patients who presented a score 3 (honeycombing in 50% or less of the field that was most compromised, as determined through biopsy) and those with a score > 3 (logrank = 2.98, p = 0.084). After five years, 72% of patients with a score 3 were surviving, compared to only 38% of those with a score > 3.
Within the group of 37 patients with a myointimal thickening score 3 (50% of vessels), 14 (38%) died. In the group with a score > 3 (n = 14), there were 11 deaths (79%; x2 = 6.74, p = 0.009). Median survival in patients with scores 3 exceeded 50 months, compared with 12 months in those with scores >3 (logrank = 10.86, p = 0.001). The Kaplan-Meier survival curve for these findings is shown in Figure 5. Figures 6a and 6b display representative examples of myointimal thickening scores.
When a multivariate model of Cox regression analysis was used to regress (against survival) the pathological variables fibroblastic foci, established fibrosis, honeycombing and myointimal thickening, all of which presented considerable (p < 0.10) interrelation, only myointimal thickening was shown to be significant (p = 0.036). Myointimal thickening scores were not significantly correlated with FVC percentage (rho = -0.18, p = 0.24), but were so with honeycombing scores (rho = 0.38, p = 0.006).
A simple qualitative system of sums (from 0 to 2) was designed, in which the points obtained from fibroblastic foci scores (0 = sparse; 1 = common) were added to those from myointimal thickening scores (0 = 50%; 1 = > 50%). Survival was significantly different among three groups, as is shown in Figure 7. Median survival in the group of patients with an overall score of 2 (n = 12) was only 9 months, compared with 51 months in patients with an overall score of 1 (n = 27). The group of patients with an overall score of 0 (n = 12) presented a greater length of survival than that of the other two groups. A similar qualitative scoring system, summing fibroblastic foci and honeycombing scores, also resulted in significant deviation among the three groups, although with less difference than in the fibroblastic foci + myointimal thickening scoring system.
When the variables previously described in the literature as being correlated with survival (FVC, duration of symptoms, gender, smoking habits) were individually evaluated with Cox regression analysis, only duration of symptoms was found to be inversely correlated with survival (p = 0.023).
The present study demonstrates that relevant prognostic information is provided through semiquantitative analyses of honeycombing, myointimal thickening and the number of fibroblastic foci in lung biopsies from idiopathic fibrosis patients.
Factors related to IPF have been explored by a number of authors. Discrepancies among these previous studies can be attributed to a variety of causes. Lack of uniformity in collection of study samples, as well as the presence of heterogeneous samples involving multiple pulmonary diseases (including idiopathic diseases and those of unknown etiology), is probably responsible for these incongruities. Neither was general uniformity maintained in the histological analyses. In most cases, only fibrosis and cellularity were evaluated, since the term cellularity was employed to describe and quantify both interstitial inflammation and intra-alveolar cell accumulation alike. In other studies, the term cellularity has been used to describe the extent and intensity of inflammation the alveolar wall. In the evaluation of degree of fibrosis, there has been a similar problem of non-uniformity.
In this study, rigid exclusion criteria were employed in the selection of IPF patients. Patients were excluded if they presented with collagen-related or other systemic diseases, had used potentially fibrogenic medications, or suffered from any other conditions known to provoke UIP. Data from patients who died as a result of the lung biopsy procedure were not considered. Selection bias may have been introduced by the biopsy prerequisite, although the average age of the patients indicates that the cases were representative. However, the low rate of mortality from other diseases suggests that patients with significant comorbidity were less frequently biopsied.
The method of sample collection was uniform: all biopsies were surgical and all were analyzed, with the exception of those presenting histology consistent with UIP. In the morphological analysis, the most significant findings were semi-quantified by previously reported scores.(11) However, individual components (rather than the proposed sum scores) of the various parameters were used.
Previous cases in which isolated lingular biopsies were performed were included. Postmortem lung samples from individuals who were free of interstitial disease revealed that more fibrosis and vascular alteration exists in the lingula than in other areas.(19) In patients suffering from interstitial disease, a biopsy of the lingula is usually representative.(20) Because cases with nonspecific alterations and cases of terminal lung disease were excluded, and because lesions characteristic of UIP were found, we believe that the lingular biopsies were acceptable.
Due to the redefinition of the term IPF, the results of this study were compared exclusively with those from more recent studies, which employ the term UIP to characterize IPF.(12,13,21,22)
In our analysis, gender was found to have no impact on survival, supporting data from a study of 238 cases by King et al.(12) However, in a study conducted by Flaherty et al,(21) women were found to have a longer length of survival. We also noted no influence of age on survival. King et al.(12) showed an inverse correlation of age with survival, whereas Flaherty et al.(21) showed no influence. A study by Nicholson et al.(22) also showed no correlation of age or gender with survival, although only eight women were included.
Longer duration of symptoms negatively influenced survival in patients evaluated in the present study. This could represent the earliest diagnosis in the natural history of the disease, the best treatment response during the initial phases of the disease, or both.
Smoking is recognized as a major risk factor for the development of IPF, although there is evidence that smokers have a longer survival.(12,21,23) This phenomenon was not observed in our study, in which smokers presented a greater accumulation of intra-alveolar macrophages as a direct result of tobacco use, as has been previously reported.(12) However, the extent and intensity of these macrophages has not been correlated with survival. The degree of other histological variables did not differ between smokers and non-smokers. The mechanism by which smoking exerts an influence over survival in IPF patients is unknown. It has been suggested that areas of active fibrotic focus may be more pronounced in smokers,(12) although no differences in fibroblastic foci scores were seen between the smokers and non-smokers evaluated in this study.
In contrast to a large recent study,(23) initial FVC values were not correlated with survival in our study. Estimated median survival of the patients in our study was 48 months. In previous studies by other authors, median survival was between 30 and 48 months.(12,21,23) In the present study, virtually all deaths were directly attributed to IPF. In this age group (66 ± 8 years), alternate causes of death are common(24) and, therefore, the cause of death must be taken into account when survival is analyzed.
In this study, the extent of areas containing fibroblastic foci influenced survival: sparse foci were associated with greater survival. Four previous studies have loosely correlated various biopsy findings with survival in IPF/UIP patients.(12,13,21,22) Two of those studies were conducted in the same center and utilized similar samples.(13,21) In both, fibrosis scores, as originally described by Cherniak,(11) were inversely correlated with survival. The biopsies were obtained from multiple sites in all cases. King et al.(12) observed longer survival in patients with lower numbers of fibroblastic foci. Cellularity and established fibrosis were not correlated with survival. The fact that absolutely no correlation with survival was found for degree of established fibrosis or honeycombing could be explained by the choice of biopsy site, i.e. that sites with advanced fibrosis were avoided. In our study, these sites were included and a closer correlation was observed. Nicholson et al. correlated biopsy findings, quantified by a new scoring system, with functional decline and survival in 53 patients with IPF or UIP. Higher rates of mortality and functional decline were independently related to higher fibroblastic foci scores.
An additional study specifically evaluated fibroblastic foci values, classified by three renowned pathologists as light, moderate and marked.(14) Higher scores were not associated with higher risk of mortality.
Pulmonary fibrosis is the common final stage in a diverse group of interstitial lung diseases. There are two routes for the development of diffuse pulmonary fibrosis: the inflammatory route (which is observed in virtually all interstitial lung diseases and which consists of an initial alveolitis phase, followed by the late fibrotic phase), and the epithelial route, which is observed in UIP.(25) The proposed hypothesis which is currently the most widely accepted is that IPF results from epithelial microlesions activated by alveolar epithelial cells and abnormal repair.(26) Epithelial cells release substances that induce migration and proliferation of fibroblasts, as well as alterations in the myofibroblastic phenotype. In the microenvironment of a lung with lesions, myofibroblasts induce apoptosis in the alveolar epithelial cells and irreversible alterations in the architecture of the extracellular matrix, resulting in erratic remodeling of the lung parenchyma.(25) Other authors have proposed that the apoptosis could be a signal event in UIP.(27)
Despite the fact that the greater part of the fibrosis seen in IPF is composed of bundles of "old" acellular collagen, small aggregations of fibroblasts and myofibroblasts are often identified as well. These aggregations, known as fibroblastic foci, are characterized by fusiform cells in a loosely organized myxoid matrix, in which they are in parallel alignment along the vertical axes of the alveolar septa. These foci represent the organization of prior foci of aggression, and may indicate active collagen synthesis at these sites. They also indicate that the fibrosis is active and, therefore, the number and extent of these foci could be representative of greater or lesser probability that the disease will progress.(3) Our results corroborate the epithelial hypothesis, as well as strongly suggesting that the number and extent of fibroblastic foci has definite prognostic value.
In this study, myointimal thickening involving more than 50% of blood vessels was shown to be highly predictive of increased mortality in IPF. In the study conducted by King et al.,(12) this correlation was not observed. This discrepancy could be explained by the fact that those authors avoided the most affected areas when selecting biopsy sites. It is well known that myointimal thickening of vessel walls is often seen in fibrotic areas. Extensive destruction of the pulmonary vascular bed is a result of pulmonary hypertension, a known predictor of a worse prognosis in pulmonary fibrosis.(23) A provocative recent study suggested that pulmonary fibrosis may arise from agression which is mainly vascular in nature.(16)
We conclude that extremely interesting prognostic information can be garnered from the evaluation of the number and extent of fibroblastic foci, the degree of honeycombing and, in particular, the consequent destruction of the pulmonary vascular bed, in biopsies obtained from patients suffering from IPF. Another study, involving a new cohort of patients, is now underway and should validate the findings of the present study.
1. Liebow AA. Definition and classification of interstitial pneumonias in human pathology. Prog Respir Res 1975;8:1-31. [ Links ]
2. Katzenstein ALA. Idiopathic interstitial pneumonia: classification and diagnosis. In: Churg A, Katzenstein A-LA, editors. The lung current concepts. Baltimore: Williams & Wilkins, 1993;1-31. [ Links ]
3. Katzenstein ALA, Myers JL. Idiopathic pulmonary fibrosis. Clinical relevance of pathology classification. State of the art. Am J Crit Care Med 1998;157:1301-15. [ Links ]
4. American Thoracic Society. Idiopathic pulmonary fibrosis: diagnosis and treatment. International consensus statement. Am J Respir Crit Care Med 2000;161:646-64. [ Links ]
5. Katzenstein ALA, Fiorelli RF. Nonspecific interstitial pneumonia/fibrosis. Histologic features and clinical significance. Am J Surg Pathol 1994; 18:136-47. [ Links ]
6. Bjoraker JA, Ryu JH, Edwin MK, Myers JL, Tazelaar HD, Schroeder DR, et al. Prognostic significance of histopathologic subsets in idiopathic pulmonary fibrosis. Am J Respir Crit Care Med 1998;157:199-203. [ Links ]
7. 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;158:1286-93. [ Links ]
8. 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. [ Links ]
9. Fukuda Y, Basset F, Ferrans VJ, Yamanaka N. Significance of early intraalveolar fibrotic lesions and integrin expression in lung biopsy specimens from patients with idiopathic pulmonary fibrosis. Hum Pathol 1995;26:53-61. [ Links ]
10. Myers JL, Katzenstein ALA. Epitelial necrosis and alveolar collapse in the pathogenesis of usual interstitial pneumonia. Chest 1988;94:1309-11. [ Links ]
11. Cherniack RM, Colby TV, Flint A, Thurlbeck WM, Waldron J, Ackerson L, et al. Quantitative assessment of lung pathology in idiopathic fibrosis. Am J Respir Dis 1991;144:892-900. [ Links ]
12. King TE Jr, Schwarz MI, Brown K, Tooze JA, Colby TV, Waldron JA Jr, et al. Idiopathic pulmonary fibrosis: relationship between histopathologic features and mortality. Am J Respir Crit Care Med 2001;164: 1025-32. [ Links ]
13. Gay SE, Kazerooni EA, Toews GB, Lynch JP 3rd, Gross BH, Cascade PN, et al. Idiopathic pulmonary fibrosis. Predicting response to therapy and survival. Am J Respir Crit Care Med 1998;157:1063-72. [ Links ]
14. Flaherty KR, Colby TV, Travis WD, Towes GB, Flint A, etal. Prognostic value of fibroblastic foci in patients with usual interstitial pneumonia. Chest 2001;120(Suppl):76-7. [ Links ]
15. Strieter RM, Belperio JA, Keane MP. CXC chemokines in angiogenesis related to pulmonary fibrosis. Chest 2002;120(Suppl 1):298-301. [ Links ]
16. Magro CM, Allen J, Pope-Harman A, Waldman J, Moh P, Rothrauff S, et al. The role of microvascular injury in the evolution of idiopathic pulmonary fibrosis. Am J Clin Pathol 2003;119:556-67. [ Links ]
17. Pereira CAC, Barreto SP, Simões JG, Pereira FWL, Gerstler JG, Nakatani J. Valores de referência para espirometria em uma amostra da população brasileira adulta. J Pneumol 1992;18:10-22. [ Links ]
18. Ferreira RG, Coletta ENAM, Giannotti Filho O. Avaliação de parâmetros histológicos na pneumonia intersticial usual (fibrose pulmonar idiopática). J Pneumol 2000;26:279-85. [ Links ]
19. Newman SL, Michel RP, Wang NS. Lingular lung biopsy: is it representative? Am Rev Respir Dis 1985;132:1084-6. [ Links ]
20. Flint A, Martinez FJ, Young ML, Whyte RI, Toews GB, Lynch JP 3rd. Influence of sample number and biopsy site on the histologic diagnosis of diffuse lung disease. Ann Thorac Surg 1995;60:1605-7. [ Links ]
21. Flaherty KR, Toews GB, Travis WD, Colby TV, Kazerooni EA, Gross BH, et al. Clinical significance of histological classification of idiopathic interstitial pneumonia. Eur Respir J 2002;19:275-83. [ Links ]
22. Nicholson AG, Fulford LG, Colby TV, Du Bois RM, Hansell DM, Wells AU. The relationship between individual histologic features and disease progression in idiopathic pulmonary fibrosis. Am J Respir Crit Care Med 2002;166:173-7. [ Links ]
23. King TE Jr, Tooze JA, Schwarz MI, Brown KR, Cherniack RM. Predicting survival in idiopathic pulmonary fibrosis: scoring system and survival model. Am J Respir Crit Care Med 2001;164:1171-81. [ Links ]
24. Panos RJ, Mortenson RL, Niccoli SA, King TE Jr. Clinical deterioration in patients with idiopathic pulmonary fibrosis: causes and assessment. Am J Med 1990;88:396-404. [ Links ]
25. Pardo A, Selman M. Molecular mechanisms of pulmonary fibrosis. Front Biosci 2002;7:1743-61. [ Links ]
26. Selman M, King TE, Pardo A. Idiopathic pulmonary fibrosis: prevailing and evolving hypotheses about its pathogenesis and implications for therapy. Ann Intern Med 2001;134:136-51. [ Links ]
27. Barbas-Filho JV, Ferreira MA, Sesso A, Kairalla RA, Carvalho CR, Capelozzi VL. Evidence of type II pneumocyte apoptosis in the pathogenesis of idiopathic pulmonary fibrosis (IFP)/usual interstitial pneumonia (UIP). J Clin Pathol 2001;54:132-8. [ Links ]
Submitted: 31/3/3. Accepted, after revision: 16/9/3.
* Study carried out in the Pathology Department of the Federal University of São Paulo (UNIFESP), the Anatomical Pathology Department of the São Paulo Hospital do Servidor Público Estadual (HSPE-SP), the UNIFESP and HSPE-SP pulmonology clinics, and the Pavilhão Pereira Filho, Santa Casa of Porto Alegre in the state of Rio Grande do Sul