Lung Cancer 32 (2001) 19 – 25 www.elsevier.nl/locate/lungcan
Fluorescence bronchoscopy in the detection of preinvasive bronchial lesions in patients with sputum cytology suspicious or positive for malignancy Kiyoshi Shibuya a, Takehiko Fujisawa a,*, Hidehisa Hoshino a, Masayuki Baba a, Yukio Saitoh a, Toshihiko Iizasa a, Makoto Suzuki a, Mizuto Otsuji a, Kenzo Hiroshima b, Hidemi Ohwada b a
Department of Surgery, Institute of Pulmonary Cancer Research, Chiba Uni6ersity School of Medicine, 1 -8 -1 Inohana, Chuo-ku, Chiba 260 -8670, Japan b Department of Pathology, Institute of Pulmonary Cancer Research, Chiba Uni6ersity School of Medicine, 1 -8 -1 Inohana, Chuo-ku, Chiba 260 -8670, Japan Received 28 March 2000; received in revised form 14 September 2000; accepted 21 September 2000
Abstract Background: a new strategy in the treatment of squamous cell carcinoma of the tracheobronchial tree is the detection and eradication of preinvasive bronchial lesions before they become invasive cancers. It is, however, difficult to detect preinvasive lesions by conventional white-light bronchoscopy alone. Purpose: we conducted a detailed investigation on the use of fluorescence bronchoscopy in the detection of preinvasive bronchial lesions in patients with sputum cytology suspicious or positive for malignancy. Methods: 64 participants with sputum cytology suspicious or positive for malignancy were examined with both white light and fluorescence bronchoscopy (LIFE group). Earlier to this study, before fluorescence bronchoscopy became available in our institute, 48 participants having sputum cytology suspicious or positive for malignancy were examined with white light bronchoscopy alone (control group). Biopsy specimens for pathological examinations were taken of all abnormal areas discovered by white light or fluorescence bronchoscopy examination. Results: in sputum cytology suspicious or positive for malignancy, the diagnosis of preinvasive bronchial lesions was greatly enhanced in the LIFE group as compared with the control group (45 vs. 7 lesions). The percentage of participants with preinvasive bronchial lesions was also significantly higher in the LIFE group than in the control group (40.6 vs. 12.5%, P= 0.00087, respectively). Conclusions: our study suggests that the use of fluorescence bronchoscopy in addition to conventional white-light examination could greatly enhance the detection and localization of preinvasive bronchial lesions in patients with sputum cytology suspicious or positive for malignancy. © 2001 Elsevier Science Ireland Ltd. All rights reserved.
Abbre6iations: LIFE, light-induced fluorescence endoscopy. * Corresponding author. Tel.: + 81-43-2227171, ext. 5464; fax: +81-43-2262172. E-mail address:
[email protected] (T. Fujisawa). 0169-5002/01/$ - see front matter © 2001 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 1 6 9 - 5 0 0 2 ( 0 0 ) 0 0 2 0 0 - 2
20
K. Shibuya et al. / Lung Cancer 32 (2001) 19–25
Keywords: Preinvasive bronchial lesions; Fluorescence bronchoscopy; Sputum cytology
1. Introduction During the past decade, incidences of lung cancer and its related death rates have dramatically increased in Japan. In 1999, lung cancer became the leading cause of death from cancer for both men and women, exceeding that of gastric cancer. Detection of early hilar lung cancers, especially carcinoma in situ and microinvasive carcinoma, from sputum cytologic examination were expected to be a cure. In patients with carcinoma in situ and microinvasive carcinoma, however, repeated examinations over many months may be required before localization is achieved. Moreover, it is exceedingly difficult to detect dysplastic lesions in the bronchial epithelium by conventional white light bronchoscopy alone [1]. This is due to the absence of any gross mucosal abnormalities in the case of dysplasia and carcinoma in situ. Centrally arising squamous cell carcinoma of the tracheobronchial tree, especially in heavy smokers, is thought to develop in multiple stages; from squamous metaplasia to dysplasia, followed by carcinoma in situ, and finally invasive cancer [2,3]. Therefore, a new strategy for treatment of squamous cell carcinoma of the tracheobronchial tree is the detection and eradication of preinvasive bronchial lesions (dysplasia and carcinoma in situ) before they become invasive cancer [4,5]. A fluorescence bronchoscopy system has been developed to detect and localize dysplastic lesions, carcinoma in situ and early hilar lung cancers [1,4,6]. When the bronchial surface is illuminated by a blue light such as that from a helium-cadmium laser, there is a reduction in fluorescence intensity in tissue that is abnormal [7,8]. The monitor image can visually display this reduction with normal tissue appearing green, and precancerous or cancerous lesions appearing brown or red-brown, thereby, distinguishing it from normal bronchial tissue. Currently, sputum cytologic examination provides the only method of detection and screening of early hilar lung cancer [9,10]. In this study, a mass screening of sputum cytologic examinations
was made for a high-risk group of participants. We then conducted a detailed investigation on the use of fluorescence bronchoscopy in the detection of preinvasive bronchial lesions in patients with sputum cytology suspicious or positive for malignancy. Preinvasive bronchial lesions were defined as lesions that were dysplasia and carcinoma in situ according to recent WHO criteria [11].
2. Material and method
2.1. Study population Sputum examinations in mass screening were done for a high-risk group of participants. The group consisted of the following; (1) those 50 years or older with 30 or more pack-years; (2) those 40 years or older who have had bloody sputum within the past 6 months; and (3) those 40 years or older with a high risk for occupational disease (asbestos exposure and chloro methyiethyl ether exposure so on). Approximately 10 000 participants were examined sputum cytology in mass screening each year and 20–30 patients with suspicious or positive for malignancy were defined. From 1997 through 1999, 64 participants with sputum cytology suspicious or positive for malignancy were referred to our institute and examined with both white light and fluorescence bronchoscopy (LIFE group). From 1994 through 1996, before fluorescence bronchoscopy became available in our institute, 48 participants with sputum cytology suspicious or positive for malignancy were referred to our institute and examined with white light bronchoscopy alone (control group). Criteria for cytological diagnosis in mass screening by sputum cytology according to the Japan Lung Cancer Society (Table 1) [12]. Sputum cytology suspicious or positive for malignancy were defined as lesions that had ‘severe dysplastic cells or suspicious for malignant tumor’or ‘cells of malignant tumor’ according to this criteria.
K. Shibuya et al. / Lung Cancer 32 (2001) 19–25 Table 1 Criteria for cytological diagnosis in mass screening by sputum cytology Cytological diagnoses A
Definition
No histiocytes in sputum (insufficient specimen) Normal Only normal epithelial cells Proliferative basal cells Mild dysplasia Cilliated columnar cells Moderate dysplasia Columnar cells with changes of nuclear size, or strong staining Severe dysplasia or cells suspicious Suspicious for for malignant tumor malignancy Cells of malignant tumor Positive for malignancy
B
C
D E
Characteristics of the participants are shown in Table 2. In the LIFE group, there were 61 males and 3 females ranging in age from 47 to 85 years old with a mean of 68 years. Their smoking history in pack-years ranged from 0 to 175 with a mean of 57. Fifty-two were current smokers, ten were ex-smokers and two were non-smokers. In the control group, the distribution of age, sex, pack-years and smoking status were not significantly different.
21
2.2. White light and fluorescence bronchoscopy Bronchoscopic examination was performed by one of two investigators (K. Shibuya and T. Fujisawa) who were trained in the use of fluorescence bronchoscopy. Conventional white-light examinations were first performed with patients under local anesthesia with sedation by intravenous injection (diazepam) and O2 inhalation. Areas with abnormal findings were recorded for subsequent biopsy. Fluorescence bronchoscopy examination with the use of the LIFE lung-system (Xillix LIFE-Lung Fluorescence Endoscopy System; Xillix Technologies Corp; Richmond, BC, Canada) was then carried out. The definition of ‘Early hilar lung cancer’ is as follows according to the Japan Lung Cancer Society [13]. Clinical criteria (1) plain chest X-ray films, tomography and CT scans are normal; (2) no lymph node metastasis or distant metastasis is observed in routine clinical screening, such as chest X-ray, CT scans, abdominal CT, abdominal echo, brain CT, or bone scintigraphy. Bronchoscopic criteria (1) location in subsegmental or more proximal bronchi or trachea; (2) the peripheral margin of the tumor is visible bronchoscopically; (3) the tumor size is less than 2 cm in greatest dimension; (4) squamous cell carcinoma is identified histologically. Tumor growing beyond this criteria are considered to be invasive. Biopsy specimens for pathological examination were taken of all suspicious or abnormal areas discovered by white-light
Table 2 Characteristics of the participantsa Control group (n =48)
LIFE group (n =64)
P value
Age, mean 9S.D. (range)
669 7 (51–84)
68 98 (47–85)
0.3043
Sex Male Female
47 1
61 3
0.4243
Pack-years, mean 9 S.D. (range)
55 9 23 (0–125)
57 931 (0–175)
0.6792
Smoking status Current Former None
46 0 2
52 10 2
0.1991
a
S.D., standard deviation; LIFE, light-induced fluorescence endoscopy
22
K. Shibuya et al. / Lung Cancer 32 (2001) 19–25
bronchoscopy examination, fluorescence bronchoscopy examination, or both. In addition, one or more biopsies were done in normal areas. All of the biopsied specimens were diagnosed by two expert pulmonary pathologists at our institute (K. Hiroshima and H. Ohwada). Dysplasia and CIS were classified according to recent WHO criteria. Carcinoma in situ that involved the entire thickness of epithelium and shows the cellular and architectural features of carcinoma without penetration of the subepithelial basement membrane. Dysplasias fall short of full thickness involvement that characterizes carcinoma in situ [11]. Informed consent was obtained from each patient prior to investigation.
Table 3 The number of carcinoma, preinvasive bronchial lesions and others in both the LIFE and control group
Invasive lung cancer Early hilar lung cancer Laryngo-pharyngeal cancer Carcinoma in situ Dysplasia Squamous metaplasia Hyperplasia, bronchitis, normal epithelium
10 6
10 7
2
4
0 7 2 13
3 42 19 127
2.3. White light bronchoscopy alone
Total
40
212
Before introducing the LIFE-lung system, conventional white-light examination alone was performed with patients under local anesthesia and O2 inhalation. If abnormal bronchoscopic findings were not detected, differential bronchial brushing was systematically done to all segmental bronchi. All lobal brushing specimens were assembled for cytological diagnosis. Biopsy specimens were taken of abnormal areas for pathological diagnosis. We performed the white-light bronchoscopy again, when the abnormality was not confirmed with cytopathological examinations.
2.4. Statistical analysis The Fisher’s exact test was applied to a contingency table analysis to determine the association of preinvasive bronchial lesions to the LIFE group and the control group. P values less than 0.05 were considered as statistically significant.
3. Results
3.1. The number of carcinoma, prein6asi6e bronchial lesions and others in both the LIFE group and control group Total of 40 biopsy specimens were obtained from the control group (Table 3). Ten biopsies
Control group (94–96)
LIFE group (97–99)
were found to have invasive lung cancers, six biopsies were early lung cancers, seven biopsies were dysplasia, two biopsies were squamous metaplasia and 13 were hyperplasia or bronchitis or normal bronchial epithelium. Two biopsies were laryngo-pharyngeal cancer. In the LIFE group, total of 212 biopsy specimens were obtained. Ten biopsies were found to have invasive lung cancers, seven biopsies were early lung cancers, three biopsies were carcinoma in situ, 42 biopsies were dysplasia, 19 biopsies were squamous metaplasia and 127 were hyperplasia or bronchitis or normal bronchial epithelium. Four biopsies were laryngopharyngeal cancer. Among 212 biopsies, 45 lesions were revealed to be preinvasive bronchial lesions. Diagnoses of preinvasive bronchial lesions were increased in the LIFE group (45 vs. 7 lesions). In the control group, one invasive lung cancer and one early hilar lung cancer have synchronous dysplasia. However, in the LIFE group, four invasive lung cancers and two early hilar lung cancers have synchronous dysplasia. In the control group, 23 patients required one examination (ten invasive lung cancer, six early hilar lung cancer and two laryngo-pharyngeal cancer), ten patients required two examinations, seven patients required three examinations, seven patients required four examinations and one patient re-
K. Shibuya et al. / Lung Cancer 32 (2001) 19–25 Table 4 Detection of each bronchial lesions as abnormal or normal areas by white light or fluorescence bronchoscopy in the LIFE group White light Fluorescence (abnormal/normal) (abnormal/normal) Invasive lung cancer Early hilar lung cancer Laryngo-pharyngeal cancer Carcinoma in situ Dysplasia Squamous metaplasia Hyperplasia, bronchitis, normal epithelium Total
10/0 7/0
10/0 7/0
4/0
4/0
2/1 29/13 11/8
3/0 38/4 16/3
52/75
84/43
115/97
162/50
quired seven examinations. In the LIFE group, all patients required only one examination.
3.2. Detection of each bronchial lesions as abnormal or normal areas by white light or fluorescence bronchoscopy in the LIFE group Using white light examination alone, two of three carcinoma in situ and 29 of 42 dysplasias were detected (Table 4). With the addition of fluorescence bronchoscopy all three carcinoma in situ and 38 of 42 dysplasias were defined. Therefore, the sensitivity for preinvasive bronchial lesions was 68.9% by white light bronchoscopy and 91.1% with the addition of fluorescence bronchoscopy. In relation to specificity, white light bronchoscopy revealed 56.8% and fluorescence bronchoscopy 31.5%, respectively.
23
3.3. The frequency of participants with prein6asi6e bronchial lesions In sputum cytology suspicious or positive for malignancy, the percentage of participants with preinvasive bronchial lesions was significantly higher in the LIFE group than in the control group (40.6 vs. 12.5%, P=0.00087, respectively) (Table 5).
4. Discussion Sputum cytologic examination of the high-risk group has often been used for the detection of early hilar lung cancer which could not be detected by chest X-ray. Invasive or early hilar lung cancer found by sputum cytologic examination can generally be localized with conventional white light bronchoscopy. Carcinoma in situ and dysplasia, however, are difficult to detect and localize even with repeated conventional white light examinations [8,14]. This is due to the absence of any gross mucosal abnormalities. In the present study, we examined the use of fluorescence bronchoscopy in the detection of preinvasive bronchial lesions (dysplasia and carcinoma in situ) in patients with sputum cytology suspicious or positive for malignancy. In the control group, only seven dysplasias were defined, using white light bronchoscopy. It was more difficult in the control group than in the LIFE group to accurately determine subtle changes in the mucus membrane of the bronchus by white light bronchoscopy alone. In the control group, bronchoscopy was invariably repeated in subjects in whom there were no abnormal findings from initial white light bronchoscopy, but there were still very few cases in which subtle abnormal findings could be seen in the mucus membrane of
Table 5 The frequency of participants with preinvasive bronchial lesions
Preinvasive bronchial lesion (+) Preinvasive bronchial lesion (−)
Control group (n =48) 94–96
LIFE group (n =64) 97–99
P value
6 (12.5%) 42 (87.5%)
26 (40.6%) 38 (59.4%)
0.00087
24
K. Shibuya et al. / Lung Cancer 32 (2001) 19–25
the bronchus. As Table 3 shows, there were no more than 15 sites (two squamous metaplasia and 13 hyperplasia, bronchitis or normal bronchial epithelium), where the lesions were more benign than preinvasive bronchial lesions. However, as can be seen with the LIFE group, performing fluorescence bronchoscopy further enhances the white light bronchoscopy process. Even with white light bronchoscopy alone, the number of subtle abnormal sites that can be detected is increasing. In any event, it is thought that in order to determine by white light bronchoscopy alone subtle changes in the bronchial mucus membrane that indicate preinvasive bronchial lesions, it is necessary to always take special care, when performing bronchoscopy. In the LIFE group, however, three carcinoma in situ and 38 dysplasias were diagnosed using fluorescence bronchoscopy. Moreover, the percentage of participants with preinvasive bronchial lesions was found to be significantly higher in the LIFE group than in the control group. It became possible to diagnose preinvasive bronchial lesions in the LIFE group that were undetectable with white light bronchoscopy alone. Kurie et al. has reported that LIFE bronchoscopy did not improve the detection of dysplasia or squamous metaplasia [15]. This could be due to a different patient population with a lower possibility of preinvasive bronchial lesions than the population in this and other earlier studies [6]. Patients with sputum cytology suspicious or positive for malignancy in mass screening were more likely to have preinvasive bronchial lesions than patients with known or diagnosed lung cancer and radiologically suspected lung cancer [16]. Since most invasive or early invasive lung cancers are visible under white-light bronchoscopy, the detection of these lesions was not the aim of the fluorescence bronchoscopy examination. The clinical indication for fluorescence bronchoscopy is, therefore, not necessary for all cases of suspected lung cancer [17,18]. We found instead that in a high-risk population, fluorescence bronchoscopy was more effective than white-light bronchoscopy in detecting preinvasive bronchial lesions. The aim of fluorescence bronchoscopy is to detect preinvasive bronchial lesions and eradicate them before they
become invasive [4,5]. In this study preinvasive bronchial lesions were detected with a high frequency in the LIFE group. It is our opinion that in patients with sputum cytology suspicious or positive for malignancy, the use of fluorescence bronchoscopy is an essential tool for the detection and localization of preinvasive bronchial lesions. Thus, in the LIFE group, preinvasive bronchial lesions were definitively depicted as abnormal fluorescent images, as has been earlier reported [6,16,19]. Specificity, however, was found to be low. Since LIFE revealed many milder sites, such as bronchitis and squamous metaplasia, by abnormal fluorescence, we intend to perform more biopsies. Although specificity tends to be low, we wish to emphasize an increase in sensitivity, which should be useful in performing fluorescence bronchoscopy. In the future, these problems may be solved by utilizing a quantitative fluorescence device as a next-generation fluorescent diagnostic system, which can be displayed on a television monitor in real time. The study of preinvasive bronchial lesions as diagnosed by fluorescence bronchoscopy is also expected to contribute to translational research at the gene level [20]. This would give insight into the process of the multi step carcinogenesis of the bronchial epithelium. It is also hoped that a course of preventive medicine for lung cancer will be established whereby patients with localized dysplasia detected by fluorescence bronchoscopy can be carefully monitored and motivated in the effort to stop smoking along with the possible administration of chemopreventive agents. In conclusion, we have found that the use of fluorescence bronchoscopy in addition to conventional white-light examination can enhance the detection and localization of preinvasive bronchial lesions in patients with sputum cytology suspicious or positive for malignancy.
Acknowledgements Supported in Part by Grants-in-Aids for Scientific Research (C) 11671305 from the Ministry of Education, Science, Sports and Culture of Japan.
K. Shibuya et al. / Lung Cancer 32 (2001) 19–25
References [1] Lam S, MacAulay C, LeRiche JC, Ikeda N, Palcic B. Early localization of bronchogenic carcinoma. Diagn Ther Endosc 1994;1:75–8. [2] Franklin W. The biology of bronchial premalignancy. Semin Respr Crit Care 1996;17:309–21. [3] Thiberville L, Payne P, Vielkinds J, et al. Evidence of cumulative gene losses with progression of premalignant epithelial lesions to carcinoma of the bronchus. Cancer Res 1995;55:5133– 9. [4] Lam S, MacAulay C. Endoscopic localization of preneoplastic lung lesions. In: Martinet Y, Hirsch FR, Martinet N, Vignaud J-M, Mulshine JL, editors. Clinical and Biological Basis of Lung Cancer Prevention. Basel: Birkhauser Verlag, 1997:231–8. [5] George PJ. Fluorescence bronchoscopy for the early detection of lung cancer. Thorax 1999;54:180–3. [6] Lam S, Kennedy T, Unger M, et al. Localization of bronchial intraepithelial neoplastic lesions by fluorescence bronchoscopy. Chest 1998;113:696–702. [7] Hung J, Lam S, Le Riche JC, Palcic B. Autofluorescence of normal and malignant bronchial tissue. Lasers Surg Med 1991;11:99– 105. [8] Palcic B, Lam S, Hung J, MacAulay C. Detection and localization of early lung cancer by imaging techniques. Chest 1991;99:742– 3. [9] Saito Y, Nagamoto N, Ota S, et al. Results of surgical treatment for roentgenographically occult bronchogenic squamous cell carcinoma. J Thorac Cardiovasc Surg 1992;104:401– 7. [10] Bechtel JJ, Kelley WR, Petty TL, Patz DS, Saccomanno G. Outcome of 51 patients with roentgenographically occult lung cancer detected by sputum cytologic testing: a community hospital program. Arch Intern Med 1994;154:975– 80. [11] Travis WD, Colby TV, Corin B, Shimosato Y, Brambilla E. Preinvasive lesions. In: Travis WD, Colby TV, Corin
[12]
[13]
[14] [15]
[16]
[17]
[18]
[19]
[20]
.
25
B, Shimosato Y, Brambilla E, editors. International Histological Classification of Tumours: Histological Typing of Lung and Pleural Tumours, third ed. Berlin: Springer, 1999:2– 5. The Japan Lung Cancer Society. Cytological Findings. In: The Japan Lung Cancer Society. Classification of Lung Cancer, first English edn. Tokyo: Kanehara and Co., 2000, pp. 81 – 9. The Japan Lung Cancer Society. Bronchoscopy of Early Lung Cancer. In: The Japan Lung Cancer Society. Classification of Lung Cancer, first English edn. Tokyo: Kanehara and Co., 2000, pp. 35 – 6. Edell ES, Cortese DA. Bronchoscopic localization and treatment of occult lung cancer. Chest 1989;96:919– 21. Kurie JM, Lee JS, Morice RC, et al. Autofluorescence bronchoscopy in the detection of squamous metaplasia and dysplasia in current and former smokers. J Natl Cancer Inst 1998;90:991– 5. Ikeda N, Honda H, Katsumi T, et al. Early detection of bronchial lesions using lung imaging fluorescence endoscope. Diagn Ther Endosc 1999;5:85– 90. Fujisawa T. Principles of autofluorescence bronchoscopy. In: Fujisawa T, editor. Autofluorescence BronchoscopyPrinciples and Clinical Applications. Tokyo: Kanehara and Co, 1999:12– 6. O’Neil KM, Johnson BE. Lights flicker on fluorescence bronchoscopy in patients at risk for lung cancer. J Natl Cancer Inst 1998;90:953– 5. Vermylen P, Pierard P, Roufosse C, et al. Detection of bronchial preneoplastic lesions and early lung cancer with fluorescence bronchoscopy: a study about its ambulatory feasibility under local anaesthesis. Lung Cancer 1999;25:161– 8. Lam S, MacAulay C. Fluorescence imaging-translational research opportunities in lung cancer. In: Fujisawa T, editor. Autofluorescence Bronchoscopy-Principles and Clinical Applications. Tokyo: Kanehara and Co, 1999:1– 6.