© 2002 American Thoracic Society
Histopathology and prediction of survival in usual interstitial pneumoniaTo the Editor :In the September issue of the AJRCCM (1), King and colleagues published a study regarding the relationship between histopathologic features and mortality in idiopathic pulmonary fibrosis (IPF). It was concluded that the extent of fibroblastic foci present on lung biopsy predicts survival, and current smokers had a markedly improved survival compared with never or former smokers. In this regard, two issues require further attention. At the first glance, the fact that smokers had a better survival compared with never smokers seems very interesting, as it is the first study in its kind showing a favorable outcome of smoking in IPF. This is contradictory to the case-control studies that suggested smoking as a risk factor for IPF (2–4). These studies, however, are not contradictory to the current hypothesis that looks at IPF as a form of abnormal wound healing in the lung characterized by fibroblast–myofibroblast migration and proliferation, decreased myofibroblast apoptosis, and increased activity and responses to fibrogenic cytokines (5), as the effect of smoking on myofibroblast apoptosis is not well-defined. Nevertheless, the review of lung function in the three groups of current smokers, former smokers, and never smokers in King and colleagues' study reveals that the baseline lung function in these three groups were not similar, as reflected by FEV1 of 82, 83, and 68% predicted, respectively. In fact, the Kaplan–Meier survival analysis shows the survival benefit in current smokers is not sustained, and after 120 months the survival in current smokers falls below never smokers. This likely is the point where current smokers developed terminal lung disease, a point that never smokers have already reached in the beginning of the curve. It seems that during the first 120 months never smokers died more than current smokers because they had a worse FEV1 to begin with. Second, the current pathologic criteria for diagnosis of usual interstitial pneumonia requires presence of heterogeneity in lung, including inflammation and mature fibrosis, as well as fibroblastic foci interspersed with areas of completely normal lung. Considering that even open lung biopsies are blind, biopsy specimens cannot represent the extent of fibroblastic activity of the entire lungs, as it is unclear what part of the lung is represented by the biopsy. Thus, one cannot make prognostic implications based on biopsy samples. It seems that although surgical biopsy is required for the diagnosis, it cannot be accurate for comparison of disease activity in different patients due to heterogenous nature of pathology.
Baylor College of Medicine Houston, Texas REFERENCES
From the authors: We have shown that survival time is longer among current smokers with IPF when compared with former smokers or never smokers (1, 2). Data from others have suggested a similar relationship (3). This finding does not contradict previous studies showing a high percentage of ever smokers among persons with IPF (3, 4) or that cigarette smoking is a risk factor for the development of IPF (5). The role of cigarette smoking in the development of IPF or as a factor in the progression of disease is not mutually exclusive. Thus, it appears that smoking is a risk factor for developing IPF; however, among those who develop it, current smokers live longer. It is very difficult (and probably not appropriate) to make too many inferences about the tails of a Kaplan–Meier curve. After 10 years, very few people are still alive, especially among the former smokers. Why current smokers with IPF live longer is unclear. Among current smokers, either "lead time" bias (overestimation of survival duration due to earlier detection as a result of concerns about their smoking or the development of nonspecific pulmonary symptoms) or "length" bias (overestimation of survival duration due to the relative excess of slowly progressing cases) may have played a role in our results. However, we showed that there are other findings that explain much of the difference in survival. Current cigarette smokers had different clinical presentations, radiographic manifestations, pattern of lung function abnormalities, and histopathological appearances compared with former or never smokers (1, 2). When compared with never smokers, current smokers: (1) are younger, (2) have less extensive crackles on chest examination and more frequent digital clubbing, (3) are more likely to have signs of pulmonary hypertension on chest radiograph, (4) have less reduction in lung volumes and FEV1, (5) have lesser degrees of lung stiffness, and (6) have better exercise tolerance associated with better gas exchange. It is possible that current smokers have concomitant emphysema and that this influences lung function and alters the prognosis. Why this "protective" effect is not seen among former smokers is unknown. In addition, the extent and severity of the specific histopathological features of usual interstitial pneumonia (UIP) are different among current smokers compared with former or never smokers. Importantly, there were no differences in the histopathology between never smokers and former smokers. Current smokers showed less overall cellularity but a greater extent and severity of alveolar space cellularity—the latter reflecting increased inflammation as a result of smoking. The fibrotic changes were similar among the smoking groups. However, the extent and severity of granulation/connective tissue was less in current smokers. Limited data exist regarding the effect of cigarette smoke on fibroblast function. However, cigarette smoke does inhibit lung fibroblast proliferation as well as chemotaxis, and thereby may impair lung repair after lung injury (6). Certainly there are limitations in the use of lung biopsy for assessing the disease activity of UIP. However, despite these limitations, this study shows that by taking multiple biopsies from two or more sites and using a semiquantitative assessment of specific features of the pathology of UIP, we could identify histopathological features that predict survival in IPF (UIP) (1). Moreover, these findings suggest that we should refocus our efforts at controlling the ongoing epithelial damage and repair process associated with persistent fibroblastic proliferation rather than on stopping inflammation.
a University of California, San Francisco San Francisco, California REFERENCES
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
