© 2002 American Thoracic Society
Are all high-grade preinvasive lesions premalignant, and should they all be treated?To the Editor :We were interested to read the article by Bota and colleagues on the surveillance of preinvasive bronchial epithelial lesions (1). Although we share their hope that this approach will reduce lung cancer mortality, we do not agree that their study now provides evidence for the immediate treatment of carcinoma in situ (CIS). Their stated aim was to assess the evolution of preinvasive lesions by bronchoscopic surveillance. However, by offering treatment to severe dysplasia and CIS, it is impossible to assess their malignant potential. We believe that it is clinically and ethically justifiable to conduct more prolonged surveillance of such lesions. Firstly, there is evidence to suggest that the majority will not progress to malignancy. The post-mortem studies conducted by Auerbach and colleagues demonstrated CIS in 75% of heavy smokers (2). As only 10–20% of heavy smokers develop clinically significant lung cancers and as CIS is only implicated in squamous cell carcinoma, the authors suggested that the majority of these lesions were unlikely to progress. Furthermore, if somatic genetic changes are accepted as a surrogate for bronchial abnormalities, then two recent studies confirm the Auerbach observations by demonstrating a much higher proportion of molecular changes in random bronchial biopsies from smokers than could be expected to proceed to malignancy (3, 4). Secondly, it should be appreciated that treatment strategies aimed at eradicating CIS may prejudice subsequent management. Bota and colleagues make no recommendations on the management of CIS that persists despite endobronchial treatment (1). The next logical step would be surgical excision, although it has been argued that the risks are unacceptably high when there is no certainty that CIS will progress (5). In addition, these patients have a substantial risk of developing separate lung cancers, with the result that prior surgery may render them unfit for curative treatment. Thirdly, even if one accepts the hypothesis that all CIS lesions are premalignant, the question arises as to whether there is any advantage in detecting and treating them at such an early stage. Similar questions were raised by the NCI trials in which sputum cytology led to the detection of early stage central lung cancers without any apparent reduction in eventual lung cancer mortality. If the recommendations made by Bota and colleagues become incorporated into clinical practice, considerable time and resources will be spent detecting and treating high-grade lesions without objective evidence of clinical benefit. We believe that it will first be necessary to obtain more direct evidence of their malignant potential, identify markers of invasive disease, and then demonstrate the clinical value of detecting and treating genuinely premalignant lesions in large randomized controlled clinical trials.
a University of Cambridge Cambridge, United Kingdom REFERENCES
From the authors: The answer to the question raised by Banerjee and colleagues as to whether or not all high-grade preinvasive bronchial lesions should be treated, has been previously addressed in our article (1). In our experience, the answer is clearly no. Our study on intraepithelial bronchial lesions using autofluorescence endoscopy has demonstrated that the majority of lesions below carcinoma in situ (CIS) will spontaneously regress over a 2-year period. This includes high-grade severe dysplasia lesions, which we have found to persist or progress in only one third of the cases. We have also shown that spontaneous evolution of bronchial CIS dramatically differs from the other intraepithelial lesions, since they display very little tendency to regress on short-term follow-up, or reoccur after transient regression at 24 months. We agree with Banerjee and colleagues that the probability and the time to progression from CIS to invasive cancer could not be deduced from our study. This was due to the fact that we decided to treat each persistent severe dysplasia and CIS at the intraepithelial stage. Indeed, data has recently been reported showing a very high probability of progression of bronchial CIS to invasive cancer (2). The 6-month outcome of CIS was studied in nine patients who were not treated on first intent but only when neoplastic changes persisted or progressed, either by increase in size or histologically. Six of these patients were treated with endobronchial therapy after endoscopic reevaluation. Despite this treatment, progression to invasive carcinoma was observed in four patients. In a seventh patient with persistent CIS who refused therapy, progression to invasive carcinoma was observed. Therefore, a progression to carcinoma was documented in five of nine CIS patients (56%) after 6 months follow-up. The arguments presented by Banerjee and colleagues to support prolonged bronchial CIS surveillance without intervention seem debatable in the light of these recent data. In his post-mortem study, Auerbach and colleagues applied the term "carcinoma in situ" to all lesions composed entirely of atypical cells without cilia present (3). According to the recent WHO classification (4), such definition will include "modern" CIS as well as severe dysplasia. Indeed, the prevalence of true CIS detected by fluorescence endoscopy in the bronchial tree of high-risk patients could be estimated at 5% or less of the patients (5), a figure far from the 40–75% reported in the Auerbach study. This is in fact compatible with a real malignant potential for the majority of CIS lesions. As progression to cancer results from the accumulation of an unknown number of molecular abnormalities (6), the recent findings regarding some molecular damage in noncancerous bronchial epithelium of heavy smokers was not unexpected. Therefore, this cannot be considered an argument against treating CIS. Finally, time and resources spent for prolonged and repeated endoscopic surveillance until invasion eventually occurs from CIS would appear at least as costly as immediate endobronchial treatment. Safe and inexpensive treatment that has proven to produce long-term complete response in CIS, such as cryotherapy or electrocaughtery, is currently available (7). Taking these data into consideration, we feel that a simple surveillance program of bronchial CIS without intervention is only justifiable as part of a controlled clinical trial.
a Hopital Charles Nicolle and INSERM EPI 9906 Rouen, France REFERENCES
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
