Published ahead of print on November 7, 2007, doi:10.1164/rccm.200708-1142OC
© 2008 American Thoracic Society doi: 10.1164/rccm.200708-1142OC
Chromosomal Aneusomy in Bronchial High-Grade Lesions Is Associated with Invasive Lung Cancer1 Department of Medicine, University of Colorado Health Sciences Center, Denver, Colorado; 2 Department of Medicine, University of Iceland Hospitals, Reykjavik, Iceland; 3 Pulmonary Division, Department of Medicine, Denver Veterans Affairs Medical Center, Denver, Colorado; 4 Department of Preventive Medicine and Biostatistics, University of Colorado Health Sciences Center, Denver, Colorado; 5 Pulmonary Division, Department of Medicine, Presbyterian/St. Luke's Health One Medical Center, Denver, Colorado; 6 Department of Pathology, University of Iceland Hospitals, Reykjavik, Iceland; 7 Department of Respiratory Medicine, British Columbia Cancer Agency, Vancouver, Canada; and 8 Department of Pathology, University of Colorado Health Sciences Center, Denver, Colorado Correspondence and requests for reprints should be addressed to Marileila Varella-Garcia, Ph.D., Department of Medicine/Medical Oncology, University of Colorado Health Sciences Center, Campus Box 6511, Mail Box 8117, Aurora, CO 80045. E-mail: marileila.garcia{at}uchsc.edu
Rationale: The development of lung cancer (LC) is accompanied by field changes in the airway mucosa that may have prognostic importance. Objectives: To compare patients with prevalent LC to control subjects regarding their histologic dysplasia scores and chromosomal aneusomy as measured by fluorescence in situ hybridization (FISH). Methods: The most advanced bronchial histology lesion was assessed from each of 44 LC cases and 90 cancer-free control subjects using a four-color FISH probe set encompassing the chromosome 6 centromere, 5p15.2, 7p12 (epidermal growth factor receptor), and 8q24 v-myc myelocytomatosis viral oncogene homolog (MYC) sequences. Histology grades were coded as dysplasia (moderate or severe) or carcinoma in situ (CIS). Measurements and Main Results: CIS was the highest histologic grade for 32 subjects, and dysplasia was the highest grade for 102 subjects (54 moderate, 48 severe). Chromosomal aneusomy was seen in 64% of the LC cases, but in only 31% of the control subjects (odds ratio [OR], 4.68; 95% confidence interval [CI]. 1.97–11.04). Among those with any level of dysplasia, the OR for positive FISH and LC was 2.28 (95% CI, 0.75–6.86). Among those with CIS, the OR for positive FISH and LC was 5.84 (95% CI, 1.31–26.01). Conclusions: Chromosomal aneusomy is associated with LC. Prospective examination of aneusomy as a precursor lesion that predicts LC is needed.
Key Words: chromosomal aneusomy • fluorescence in situ hybridization • carcinoma in situ • premalignancy
Lung cancer (LC) is the most common fatal cancer in the United States and most parts of the world. About 160,000 deaths are expected to occur annually in the United States alone, and the incidence of LC in developing countries continues to rise (1, 2). Only about 15% of patients with LC survive 5 years with the best diagnostic and treatment options currently available. The main reason for the poor outcome is that more than two-thirds of cases are diagnosed at an advanced stage and, even when diagnosed in stage I, up to 40% of cancers recur after surgical resection. Smoking is the overwhelming risk factor, and most studies indicate that about 90% of cases are due to smoking (2). Familial and genetic factors also affect the susceptibility to LC development (3, 4). To substantially improve the outcome for LC, new strategies of prevention and early detection are required (5). Screening with low-dose helical computed tomography scanning has been effective in finding small LC nodules that may be cured by surgery (6, 7), but the effect of this technique on survival is still uncertain (8, 9). Detecting radiographically occult intraepithelial bronchial lesions with sputum cytology and autofluorescence bronchoscopy (10, 11) could further contribute to improved survival in LC, as seen in other epithelial malignancies, such as cervical cancer and cancer of the colon (12, 13). The development of centrally located bronchogenic carcinoma likely occurs through a stepwise process of carcinogenesis that is reflected by advancing histologic grade of dysplasia to carcinoma in situ (CIS) and invasive cancer (14). These changes are accompanied by the accumulation of chromosomal abnormalities throughout the airway mucosa, a process known as field cancerization (14, 15). The distinction between high-grade dysplastic lesions is subjective (16), and histologic grade was shown to be an uncertain predictor of progression to invasive cancer, reflecting the variability in the biological behavior of these lesions (17). There is therefore a need for biomarkers that can be used to more accurately predict the development of invasive cancer from a preneoplastic lesion than can histologic grade alone. Chromosomal aneusomy has long been recognized as a hallmark for LC (18). The application of fluorescence in situ hybridization (FISH) technology to interphase cells has greatly facilitated the study of chromosomal changes in exfoliated cells and other airway epithelial samples (19). In previous studies, we have found that the detection of aneusomy by multitarget FISH assay increased the sensitivity of sputum cytology as a predictor of LC in high-risk subjects (20). In the present study, we investigated chromosomal aneusomy in situ in biopsies collected from heavy smokers by autofluorescence and white-light bronchoscopy, and correlated this biomarker with the histologic grade of the lesions and the presence of invasive LC. This study tested the hypotheses that chromosomal aneusomy is associated with histologic grade in bronchial epithelium, and that it is also associated with LC, accounting for histologic grade. Preliminary reports on the data contained in this article have been presented in abstract form at the Fourth International Chicago Symposium on Malignancies of the Chest and Head and Neck (21) and the 12th World Conference on Lung Cancer (22).
Subjects The study was performed using patient records and histologic sections that have been collected at the University of Colorado Cancer Center (UCCC) in Aurora, CO (106 individuals), the British Columbia Cancer Agency (BCCA) in Vancouver, Canada (19 individuals), and the University of Iceland Hospitals (UIH) in Reykjavik, Iceland (9 individuals). This study included all individuals in the UCCC database with CIS and severe dysplasia (SD) as the highest grade of mucosal abnormality, and 70% of those with moderate dysplasia (MD). The study population was specifically enriched with subjects with CIS lesions from the BCCA (Canada) and the UIH (Iceland). The subjects studied were all considered to be at high risk for LC based on a history of at least 30 pack-years of smoking and spirometric evidence of airflow obstruction documented by an FEV1/FVC ratio of less than 75% and an FEV1 of less than 70% of predicted. Former smokers were defined as having quit at least 1 year before the time of enrollment. Pack-years was defined as the average number of packs smoked per day multiplied by the numbers of years smoked. Subjects filled out a standard questionnaire, spirometry was performed, and results recorded. Flexible fiberoptic bronchoscopy was performed with both autofluorescence and white-light examination of the airways using either a Xillix LIFE II or OncoLIFE systems (Xillix, Richmond, BC, Canada), as previously described (10, 11), at the UCCC and BCCA sites; white-light examination alone was performed at the UIH. The BCCA cases had been diagnosed in a prospective study of early LC using autofluorescence and white-light bronchoscopy, or the subjects were enrolled as part of two National Cancer Institute–sponsored chemoprevention trials; in one case, the CIS biopsy was taken before chemoprevention, and in the other, after 6-month treatment with retinol. The Icelandic cases were identified from the computerized database of the Department of Pathology, UIH. Protocols for tissue acquisition by bronchoscopy were approved by the Colorado Multiple Institutional Review Board, the Research and Development Committee of the Denver Veterans Affairs Medical Center, the HealthOne Institutional Review Board, the BCCA–University of British Columbia Clinical Research Ethics Board, the National Bioethics Committee of Iceland, and the Icelandic Data Processing Commission. The determination of whether or not subjects had invasive LC at the time of bronchoscopy was made based on biopsy results, review of the clinical history, and matching to central cancer registries, vital statistics records, and national death indices at the respective institutions. The clinical information available on each individual included the following: (1) collection date of specimen studied; (2) collection site; (3) date of birth; (4) date of diagnosis of invasive LC (if present); (5) site of invasive LC; (6) histologic grade of preneoplastic lesion; (7) results of aneusomy detection by FISH analysis.
Histologic Analysis
FISH Analysis The H&E-stained sections and one 4-µm-thick unstained section per subject were transferred to the FISH laboratory, with the accompanying images identifying the lesions. The unstained slide was subjected to a FISH assay using the LAVysion DNA probe set (Vysis/Abbott Molecular, Des Plaines, IL), including four DNA targets, 6p11.1-q11 (Chromosome Enumeration Probe CEP6, labeled with SpectrumAqua), 5p15.2 (D523, D5S721 labeled in SpectrumGreen, encompasses the SEMA5A gene), 7p12 (labeled in SpectrumRed, encompasses the epidermal growth factor receptor gene), and 8q24.12-q24.13 (labeled in SpectrumGold, encompasses the v-myc myelocytomatosis viral oncogene homolog [MYC] gene). Initially, the slides were incubated for 2 hours at 56°C, deparaffinized in Citri-Solv (Fisher Scientific, Pittsburgh, PA), and washed in 100% ethanol for 5 minutes. The slides were then incubated in 2x sodium chloride, sodium citrate (SCC) buffer at 75°C for 15 minutes, digested in 0.25 mg/ml proteinase K/2x SSC at 45°C for 16 minutes, and washed in 2x SSC for 5 minutes before being dehydrated in an ethanol series. The probe set was applied according to the manufacturer's instructions to the selected hybridization areas, which were covered with 12-mm glass coverslips and sealed with rubber cement. Codenaturation and hybridization were performed in the Hybrite platform (Vysis), with temperature set for 1 minute at 85°C and 24 hours at 37°C. Posthybridization washes were performed with 2x SSC/0.3% NP40 at 72°C for 2 minutes, followed by wash in 2x SSC for 2 minutes at room temperature and dehydration in ethanol series. Chromatin was counterstained with 4',6-diamidino-2-phenylindole (0.3 µg/ml in Vectashield Mounting Medium; Vector Laboratories, Burlingame, CA). Analysis was performed on epifluorescence microscopes (Carl Zeiss MicroImaging, Thornwood, NY) equipped with single band pass interference filter sets for blue (4',6-diamidino-2-phenylindole), aqua (SpectrumAqua), green (fluorescein isothiocyanate), yellow (SpectrumGold), and red (Texas red). For documentation, images were acquired with a cooled charge-coupled device camera (Photometrics, Tucson, AZ) in monochromatic layers and merged and processed using the CytoVision software (Applied Imaging, Inc., San Jose, CA). The areas selected by the pathologist in the H&E-stained sections were identified in the hybridized slide, and 30–50 nuclei were scored per area. A total of 30 areas with normal histology was also included in the experiments and similarly analyzed. A cell was considered abnormal (aneusomic) when showing three or more signals for two or more of the DNA targets. A premalignant lesion was considered abnormal (aneusomic) based on 100% specificity for the test (e.g., when the frequency of abnormal cells was higher than the highest frequency of aneusomic cells found in areas with normal histology [6.7%]).
Statistical Analysis
The analyzed population was comprised of 104 males and 30 females, with a mean age of 63.7 years. Subjects had a mean smoking history of 62.3 pack-years, and the majority of them (56.7%) were current smokers (Table 1). The distribution of age, gender, and pack-years of smoking were not different between LC cases and control subjects. The cell types of the 44 LC cases included 6 adenocarcinoma, 24 squamous cell, 1 small cell LC, and 13 unspecified non–small cell LC. Of the 44 cases with LC, 41 were known to have prevalent LC at the time of diagnosis, and 3 were diagnosed within 5 years after the bronchoscopy (12, 30, and 45 mo).
The relationship between cancer status and both biomarkers, histologic grade and chromosomal aneusomy, is shown in Table 2. Differences between LC cases and control subjects are shown as crude and adjusted OR with 95% CIs. The proportion of subjects with LC increased from carriers of MD (18.5%) and SD (22.9%) to carriers of CIS lesions (71.9%). When the OR for SD and CIS were calculated using MD as a reference, we found no association for SD (OR, 1.27; 95% CI, 0.47–3.45), whereas CIS on histologic analysis was strongly associated with invasive LC (OR, 16.4; 95% CI, 5.02–53.72).
Chromosomal aneusomy (illustrated in Figure 1) was found in premalignant lesions of 56 subjects (41.8%), whereas the lesions in the remaining 78 subjects (58.2%) showed normal FISH results. Among the 44 LC cases, chromosomal aneusomy was found in 28 (63.6%), whereas, in control subjects, aneusomic lesions were seen in 28 of 90 cases (31.1%; OR, 4.68; 95% CI, 1.97–11.04). The proportion of subjects with chromosomal aneusomy increased from MD (22.2%) to SD (41.7%) and CIS lesions (75%), and showed a similar trend for cases and control subjects (Figure 2). The odds of the presence of aneusomy were over 4.5-times greater in LC cases than in those subjects who remained cancer free (Table 3). These odds were lower among those with MD (OR, 2.31; 95% CI, 0.27–20.04) and SD (OR, 3.07; 95% CI, 0.45–21.01), but increased among those with CIS (OR, 5.84; 95% CI, 1.31–26.01). Aneusomy was also significantly associated with smoking status in this study population (P = 0.04), but not with other demographic characteristics.
Of the 23 CIS lesions with a known LC site, 19 (83%) exhibited chromosomal aneusomy, compared with 5 out of 9 (55%) in CIS lesions occurring in patients without LC. It must be noted that 14 out of the 23 CIS lesions with a known cancer site were obtained at biopsy sites adjacent to an invasive LC diagnosed at the same bronchoscopy. Of these 14 CIS lesions, 11 (79%) were aneusomic. When the 14 cases of CIS with adjacent LC were excluded, 8 of the 9 cases (89%) with LC at a site distant (in a different lobe) from CIS exhibited chromosomal aneusomy.
In this study, we have investigated histologic preneoplasia grade and chromosome aneusomy assessed by FISH assay in bronchial biopsies as correlates of prevalent LC. Aneusomy was present in 63.6% of LC cases compared with 31.1% of control subjects (OR, 4.68; 95%, CI 1.97–11.04), implying a significant association with invasive LC. Chromosomal aneusomy was associated with LC in biopsies with any dysplasia, but was particularly strongly associated in biopsies with CIS. It is likely that different probe sets can be defined that will improve the test characteristics of the FISH assay. FISH analysis on tissue samples may add valuable information to conventional histology in the assessment of preneoplastic intraepithelial bronchial lesions, and, in the future, guide the need for monitoring of high-risk patients with aneusomic lesions. There are several limitations to this study. A total of 32 patients with CIS were available from three institutions; 23 had either a prevalent (20) or incident (3) LC. As 14 of 23 patients with CIS and invasive LC had the CIS at a site adjacent to the LC, it is possible that these subjects may have biased our results toward an association between chromosomal aneusomy and LC. We do not believe this to be the case, as similar fractions (11/14 vs. 8/9) of the patients with CIS and invasive LC at either adjacent or distant sites exhibited aneusomy. A larger study population will be required to evaluate this definitively. Two theories prevail regarding CIS adjacent to an invasive LC: that the CIS represents lateral spread of tumor cells within the epithelium without demonstrated invasion, or, alternatively, these CIS lesions may represent a region of advanced dysplasia within which an invasive carcinoma arose. A second limitation is that this study was largely cross-sectional. The question of whether aneusomy predicts the subsequent development of invasive LC in various histologic grades of premalignancy will require a prospective study. The difficulties of following premalignant lesions for an adequate length of time are illustrated by the fact that the present study represents samples from three sites, two of which (UCCC and BCCA) have performed bronchoscopy in a research setting for over 15 years each. As additional high-risk patients are enrolled and followed, adequate numbers of incident cancers may be accrued to address this in a prospective manner. A wide variety of chromosomal changes are thought to be important in the development of LC (15, 24). These involve tumor suppressor genes, oncogenes, cell regulatory genes, and DNA repair genes. Chromosomal changes have been known to affect almost every chromosome, and may be found in all regions of the bronchial mucosa. These changes have been identified in airway samples from healthy smokers, from individuals with premalignant dysplastic lesions and in nonmalignant bronchial mucosa from patients with LCs (25). Multifocal and unpredictable histologic and molecular changes in the airways of high-risk smokers have been described as "field cancerization," a phenomenon particularly reported in epithelial malignancies, such as oral squamous cell carcinoma, carcinoma of the pharynx, and LC (14). Multicolor interphase FISH analysis, as well as spectral karyotyping of metaphases of cultured bronchial epithelial cells, has enhanced the capacity to study chromosomal field changes in bronchial epithelium of high-risk smokers. In a recent study, FISH analysis of exfoliated cells was found to significantly increase the sensitivity of sputum cytology as a predictor of LC (26). FISH has also been used successfully in the diagnosis of breast cancer (27) and carcinoma of the bladder (28). The development of autofluorescence bronchoscopy has provided an opportunity to study the natural history of preneoplastic lesions in airway epithelium using visual assessment at bronchoscopy and histologic analysis. Up to the present, the histologic grade has been the only clinical biomarker to assess the propensity of preneoplastic change to progress to CIS or invasive cancer. In a recently published prospective follow-up study of 54 individuals with various grades of premalignant lesions, progression to CIS or invasive LC was observed in 30% (10/34) of individuals with low-grade lesions over a median follow-up time of 52 months (17), and in 39% of those with high-grade lesions over median follow-up of 39 months. The difference in progression rate between low- and high-grade lesions was not significant. Fluctuation of histologic grade on repeat, same-site biopsy was observed with 26% of low-grade lesions and 61% of high-grade lesions absent from follow-up biopsies. There are several possible explanations for this finding, including complete removal of small lesions by initial biopsy or regression of dysplasia. Direct observation of continuous, linear progression to malignancy from low- and even high-grade lesions at specific bronchial sites is not to be expected in most individuals, particularly during a short period of observation. The approach to the treatment of these premalignant early lesions will require better definition of the natural history of these lesions with longer follow-up and more comprehensive evaluation of correlative biomarkers. CIS lesions are rare in the clinical setting. Most longitudinal studies indicate that the risk of progression of such a lesion to invasive LC is substantial (29, 30). The development of biomarkers that more accurately predict the progression to invasive cancer is particularly important in this setting. On the basis of our cross-sectional data, the presence of chromosomal aneusomy might be hypothesized to denote a higher risk of developing invasive LC. Additional prospective studies of FISH analysis for chromosomal aneusomy are required to establish its potential role in the assessment and clinical management of preneoplastic bronchial lesions. Future studies in a larger population are warranted to assess the role of histologic grade and FISH in the evaluation of these lesions. Chromosomal aneusomy may also be one way to identify subjects with the lesions most suitable for chemoprevention trials.
The authors thank Jerry Haney (University of Colorado Cancer Center [UCCC] Tissue Procurement Core) and John Reeves (UCCC Cytogenetics Core) for technical assistance.
Supported in part by National Cancer Institute grants U01-CA85070 (W.A.F.), P01-CA58187, P30-CA46934, and the Landspitali-University Hospital Science Fund, Iceland. Originally Published in Press as DOI: 10.1164/rccm.200708-1142OC on November 7, 2007 Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Received in original form August 1, 2007; accepted in final form November 7, 2007
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
AT A GLANCE COMMENTARY
