(Chest. 2004;125:90S-94S.)
© 2004
American College of Chest Physicians
Premalignant Evolution of Lung Cancer*
Gilles F. Filley Lecture
Wilbur A. Franklin, MD
* From the University of Colorado Health Sciences Center, Denver, CO.
Correspondence to: Wilbur A. Franklin, MD, University of Colorado Health Sciences Center, 4200 E 9th Ave, Box B216, Denver, CO 80262; e-mail: wilbur.franklin{at}uchsc.edu
 |
Introduction
|
|---|
Lung cancer continues to be the most lethal of human tumors and is expected to account for > 160,000 deaths in 2004.1 This high mortality has been attributed to the late stage of the disease at detection and to aggressive biological behavior. Because of this aggressive behavior, there is a relatively short window of opportunity in which to effectively treat the disease. However, if the preinvasive changes in progenitor cells that lead to carcinoma are recognized as part of the process of malignant transformation, then the window of opportunity for effective intervention is considerably widened. In such a paradigm, premalignancy is regarded as the disease, and carcinoma as the end point. The objective of this article is to review the evidence for the existence of premalignant lesions in the lower airways, to determine who develops the lesions and how to find them, and to describe the molecular and cellular properties of those lesions.
|
Premalignant Lesions in the Lower Airways
|
|---|
The idea that lung cancer is preceded by premalignant changes in the epithelium of the central airways is not new and was the subject of a series of reports in the 1950s and 1960s by Auerbach and coworkers.23 Although these studies were carried out nearly 50 years ago, they remain as relevant today as they were when they were reported. In these studies, the frequency of bronchial cellular atypia was evaluated in serial sections of the lungs that were removed at autopsy from 1,522 adult smokers and control nonsmoking patients.2 Nearly 42,000 bronchial cross- sections were evaluated by the leading pathologists of the time. All of the smokers evaluated had epithelial lesions. Atypical cells were present in 93% of sections from current smokers but were present in only 1.2% of sections from individuals who had never been smokers. The histologic abnormalities in smokers’ lungs were not only frequent, but were often multifocal and independent of age, sex, place of residence, and the presence of pneumonia. A significant number of individuals who were former smokers in this study3 also had epithelial atypia, while atypia was rare in individuals who never smoked.
Although these studies confirmed the presence of widespread histologic lesions in the airways of smokers without cancer, the categorization of these lesions was highly descriptive and not easily transferable to a widely applicable diagnostic classification. Another difficulty was that the study was an autopsy study and consequently could only assess the prevalence of bronchial lesions at a single time point. The timing of the appearance of these lesions and their power to predict which individuals would develop lung cancer at some future time point could not be estimated.
Attempts to better define the pathogenesis of bronchial premalignancy have been thwarted by the invisibility of the cellular lesions and their random distribution throughout the airways. With the introduction of fiberoptic bronchoscopy in 1970, it might have been expected that premalignant lesions would be more accessible, but this has not been the case. Premalignant dysplasias in the central airways proved to be invisible by white light bronchoscopy, and the lesions remained a purely microscopic finding. In 1991, however, fluorescence bronchoscopy was introduced to visualize these lesions.4 In this methodology, airways are illuminated by laser light. Fluorescence emitted from the mucosal surface is partially quenched in areas of dysplasia, a result that can be enhanced by video signal processing, so that the latest model fluorescence bronchoscopes produce clear images of areas of quenching and, by extrapolation, dysplasia. Fluorescence bronchoscopy has enhanced the ability to identify these early lesions and to access lesional tissue for molecular studies and biomarker discovery. In a direct comparison between white light and fluorescence bronchoscopy,5 fluorescence bronchoscopy was found to be nearly three times more effective in identifying dysplastic foci in the lung than white light bronchoscopy. A downside of the technology is that it has relatively low specificity.
With improving instrumentation, a need for more reproducible histologic grading criteria for premalignant changes has developed. The most recent consensus classification for premalignant changes was published by the World Health Organization in 19996 and has been validated by interobserver studies.7 The classification contains seven categories, including normal histology, reserve cell hyperplasia, squamous metaplasia, mild, moderate, and severe dysplasia, and carcinoma in situ, in addition to three grades of dysplasia (ie, mild, moderate, and severe). In addition to alterations in the microscopic appearances of the bronchial epithelium, changes in the stromal support tissues of the epithelial cells have been described. It has become evident that neoangiogenesis occurs in the bronchial lining early in lung carcinogenesis and is reflected in the sprouting of capillaries into dysplastic squamous mucosa (Fig 1
), a lesion that has been referred to as angiogenic squamous dysplasia (ASD).8 ASD is preferentially associated with squamous cell carcinoma rather than adenocarcinoma (Table 1
).9 It is associated with elevated levels of vascular endothelial growth factor expression in the bronchial mucosa and possibly of other angiogenic signaling molecules.
|
The Susceptible Population
|
|---|
As mentioned above, premalignant squamous lesions of the central airways are frequent in smokers and rare in never-smokers. Precisely how much smoking is necessary to develop premalignant changes is not known, but the frequency of these lesions would be expected to increase with increasing tobacco exposure. In addition to smoking, airway obstruction is a strong indicator for the subsequent development of lung cancer, and the risk for lung cancer increases in proportion to the extent of the obstruction.10 Finally, it is likely that atypical cells from the premalignant lesions are exfoliated into the sputum of individuals harboring the lesions. In the large National Cancer Institute sputum screening trial conducted at the Mayo Clinic and Johns Hopkins University,11 the frequency of subjects with atypical but not malignant cells in the sputum was 1.4%, suggesting that premalignant cells could be shed into the sputum. Therefore, the levels of premalignant cells in the sputum could be an indicator of premalignant airway disease as well as of invasive carcinoma. Even prior to the National Cancer Institute study, Saccomanno et al12 had used a detailed sputum cytology grading system, which largely coincides with the current World Health Organization histology classification for grading, to show that atypical cells of lesser grade than frank carcinoma may be present in the sputum for several months to years prior to the development of invasive carcinoma.
To maximize the chances for finding premalignant or early malignant changes in the lower airways, the Colorado Specialized Program of Research Excellence in Lung Cancer program has used three criteria, including the following: (1) > 30 pack-years of cigarette smoking; (2) FEV1 < 70% predicted; and (3) moderate dysplasia or worse after sputum cytology.
To date, > 2,500 patients have been screened. Among patients in the group with both 30 pack-years of smoking and airway obstruction, 19% have sputum atypia. We have found that two thirds of the individuals with moderate dysplasia or worse after sputum cytology have lower airway lesions graded as moderate or worse after a histologic evaluation of bronchial biopsy specimens (Table 2 ).
|
Molecular Changes in the Airways During Lung Carcinogenesis
|
|---|
Although morphologic changes are the current standard for the diagnosis and documentation of premalignant changes in the lower airways, an increasing body of evidence indicates that molecular abnormalities could be used to identify and treat individuals who are at high risk for invasive lung cancer. Like other tumors, lung cancer is now thought to be the result of a stepwise accumulation of molecular abnormalities in benign precursor cells. These abnormalities occur as a direct result of DNA damage caused by the carcinogens in cigarette smoke. Many of the specific molecular lesions that occur in the premalignant lower airway epithelium are illustrated in Figure 2
and are described in more detail below. Often the changes shown in Figure 2 are found simultaneously in the same cells. Many of these molecular lesions have been evaluated for their power to predict prevalent and incident lung carcinoma. Most of these potential biomarkers have been confounded by the long lead time between the appearance of the abnormality in epithelial cells and the occurrence of the invasive carcinoma.
View larger version (82K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2. Sequence of morphological and molecular genetic changes that results in invasive lung cancer. Initial DNA damage consists of bulky adduct formation caused by interaction with the carcinogens in tobacco smoke. This may hinder DNA repair mechanisms, resulting in mitotic recombination and mutation. Recombination in the presence of a mutation may result in homozygous mutation. When mutation occurs in checkpoint genes such as p53, chromosomal instability may result in aneuploidy with numerous possible chromosomal rearrangements. The sequence of molecular changes coincides generally with morphological progression from normal or hyperplastic epithelium (upper left; hematoxylin and eosin, original magnification x400), through dysplasia (upper middle; hematoxylin and eosin, original magnification x400), to invasive carcinoma (upper right; hematoxylin and eosin, original magnification x200).
|
|
DNA Adducts
An initiating molecular event in the multistep process leading to lung carcinoma is the formation of DNA adducts in bronchial epithelial DNA. DNA adducts are physical complexes that form between DNA and the reactive metabolites of carcinogenic compounds in tobacco smoke and industrial pollutants. Two of the best studied and most potent adduct formers are the polyaromatic hydrocarbons, benzo[a]pyrene, and nicotine-derived nitrosoamine ketone.13 P450 enzymes convert these compounds into several reactive intermediates by oxidation. The reactive species are largely converted to water-soluble compounds that are excreted into the urine, but a small proportion of the reactive intermediates escape and bind directly with DNA forming bulky adducts. The formation of DNA adducts activates elaborate DNA repair mechanisms. Unrepaired DNA adducts may be bypassed or may cause miscoding by DNA polymerase, inducing mutations that seriously impair the transmission of accurate genetic codes to daughter cells. There is some evidence that adduct formation may occur preferentially at endogenously methylated CpG islands and may account for the high frequency of specific G-T transversions that are found in the tumors of smokers.14
Adducts may be measured in the lung tissue and peripheral blood DNA by a variety of technologies, including high-performance liquid chromatography with fluorescence detection,15 and 32P postlabeling and immunolabeling,16 and there is substantial evidence that tobacco exposure is associated with the adduct burden. However, it is not yet clear whether adduct levels are additive to smoking history as a risk factor or simply are a reflection of exposure to tobacco carcinogens.
Loss of Heterozygosity
Adducts may interfere with enzymes that are responsible for the repair of double-stranded DNA breaks, which could result in the recombination of homologous DNA strands during the repair process and loss of heterozygosity (LOH), another change that is observed early in lung carcinogenesis. In this process, chromosomal loci that normally harbor two different polymorphic alleles exhibit the loss of one or both of these alleles. The loss of both alleles (or homozygous deletion) with associated loss of gene expression is much less common than the loss of a single allele, but the loss of only one allele may affect gene expression if the retained allele is mutated or inactivated by methylation. Therefore, LOH has been interpreted to indicate the genomic sites of tumor suppressor genes. However, it is important to recognize that LOH does not necessarily imply the physical loss of a gene but rather the reduction of a particular polymorphic locus to homozygosity. The compelling evidence for this is that LOH detected in tumors by molecular mechanisms is most often associated with an increase in gene copy number at the same locus demonstrated by fluorescence in situ hybridization (FISH).17
LOH has been documented throughout the genome in most lung carcinomas.18 Chromosomal regions 3p and 9p have been especially well-studied, and are frequently lost in both carcinoma as well as in the premalignant epithelium. Numerous candidate tumor suppressor genes have been identified in these regions, but none are uniformly lost in lung cancer. The early appearance of LOH even in smokers with normal epithelium, the technical requirement for microdissection for the test to be interpretable, and the inconstancy of LOH at specific sites all mitigate against its use as a clinical screening assay.
Methylation
Another mechanism for gene silencing is the methylation of cytosine bases in promoter regions of the genome. In the healthy individual, methylation is responsible for the regulation of several different biological functions, including genomic imprinting (in which only one parental gene copy is expressed), x-chromosome inactivation, silencing of selfish genetic elements (eg, transposons, retrotransposons, and viruses), reduction of transcriptional noise and embryonic development.19 In cancers, DNA is generally hypomethylated but is hypermethylated at important specific loci, which contribute to the neoplastic phenotype and are therefore regarded as tumor suppressor genes. Hypermethylation of promoter regions is the result of the aberrant activity of endogenous DNA methyl transferases through mechanisms that are still not well-understood.20 Gene silencing through methylation depends on which sites in the promoter regions are methylated and on the density of the methylated sites. For these reasons, testing for biologically meaningful methylation is not straightforward and is most reliable when confirmed by gene expression data.
Currently, the most widely used and straightforward assay for methylation is the methylation-specific polymerase chain reaction. In this test, metabisulfite is used to convert unmodified cysteine bases to uracil while sparing methylated bases.21 Uracil pairs with thymidine in hybridization reactions, and, thus, primers can be designed that detect only the protected (methylated) cysteine bases. This robust method has permitted the analysis of the methylation status of a number of genes that may be important in lung carcinogenesis, including p16, DAP kinase, GSTP1, MGMT, H-cadherin, RASSF1, APC, PAX-5, and RARß. Although individual genes may be methylated in a minority of tumors, most tumors are methylated at several tumor suppressor loci, suggesting an important role for accumulated gene silencing by methylation in lung carcinogenesis. Currently, considerable effort is being devoted to identifying which and how many methylated genes are most highly predictive of epithelial transformation to invasive carcinoma.
Aneuploidy
A consequence of accumulated DNA damage is chromosomal instability and missegregation leading to aneuploidy. Aneuploidy involving virtually all chromosomes is a consistent feature of invasive lung carcinoma. Although chromosomal rearrangements are massive, no consistent single chromosomal abnormality has yet been discovered. It is nevertheless possible and quite feasible to identify highly aneuploid cells in mixtures with normal cells using multicolor FISH. In a recent study,22 multicolor FISH analysis using a commercially available kit incorporating probes for 5p15.2, 6 centromere, 7p12 (EGFR), and 8q23 (myc) chromosomes detected tumor cells in all pretreatment sputum samples from 20 patients undergoing lung resection for carcinoma. Aneuploidy could be a significant tool for the detection of individuals with early-stage carcinoma or late-stage premalignancy, but tests to detect this molecular lesion in mixed cell samples, such as those in sputum, are technically complicated and are not now amenable to mass screening trials.
|
Future Directions
|
|---|
Despite the accumulation of much new fundamental information regarding cellular and molecular carcinogenesis, transferring this information into clinical practice has been and will continue to be a major challenge. A major pressing need is for biomarkers of early lung cancer. One potentially productive approach to the discovery of such biomarkers is the application of microarray technology to identify genes that are selectively overexpressed by tumor cells in comparison with normal tissues. Such an approach has already led to the identification of several genes that are potentially useful as biomarkers that are highly overexpressed in tumor but are silent in lung tissue.23 Many of these genes prove to be cancer/testis genes, genes that are expressed only by germ cell tumors and by certain other tumors, including those in melanoma (ie, MAGE proteins) and lung cancer. These genes are switched on by demethylation, and many are encoded on the long (q) arm of the x chromosome. These potential biomarkers are currently being tested singly and in combination for their ability to supplement newer helical CT imaging technology for early detection.
Invasive lung cancer, even early-stage invasive lung cancer, is a fundamentally different process than premalignant airway disease. The biomarkers that may be useful for the detection of invasive carcinoma, which are the products of a highly unbalanced and unstable genome, may not be useful for the detection and monitoring of premalignant disease. The identification of new biomarkers by high-throughput methods such as oligonucleotide arrays is complicated by the minute amount of RNA that is available from premalignant lesions and by the invisibility of these lesions when viewed by white light bronchoscopy. The fact that it has been easier to identify and sample premalignant changes through fluorescence bronchoscopy offers the prospect that premalignant lesions in the lower airways can be mapped and revisited to obtain samples for biomarker discovery, and to assess the effect of chemopreventive treatment and ablation procedures. Several agents targeting specific pathways (including ErbB, eicosanoid, and prenylation pathways) appear to have minimal side effects and can be tested for their chemopreventive effectiveness. The standard for evaluating the effects of these drugs currently is a change in the histology of the airway epithelium. In developing trials to test such agents, it will be important to properly sample the airway epithelium, and to obtain surrogate specimens such as peripheral blood and urine to evaluate not only morphology and the targeted molecular pathways, but also to identify potential molecular profiles and specific biomarkers that may better define premalignant disease. It is hoped and expected that the application of microarray technology will provide predictive gene expression profiles, and possibly individual biomarkers, that will identify more advanced premalignant airway lesions and permit more selective chemopreventive treatment to be offered to high-risk patients.
|
Footnotes
|
|---|
Abbreviations: ASD = angiogenic squamous dysplasia; FISH = fluorscence in situ hybridization; LOH = loss of heterozygosity
Part of the work described in this manuscript was supported by grants U01-CA85070, from the Early Detection Research Network, and P50-CA58157, from the Specialized Program of Research Excellence in Lung Cancer.
|
References
|
|---|
- Jemal, A, Tiwari, RC, Murray, T, et al (2004) Cancer statistics, 2004. CA Cancer J Clin 54,8-29[Abstract/Free Full Text]
- Auerbach, O, Stout, AP, Hammond, EC, et al Changes in bronchial epithelium in relation to sex, age, residence, smoking and pneumonia. N Engl J Med 1962;267,111-119[Medline]
- Auerbach, O, Stout, AP, Hammond, EC, et al Bronchial epithelium in former smokers. N Engl J Med 1962;267,119-125
- Lam, S, Hung, JY, Kennedy, SM, et al Detection of dysplasia and carcinoma in situ by ratio fluorometry. Am Rev Respir Dis 1992;146,1458-1461[ISI][Medline]
- Hirsch, FR, Prindiville, SA, Miller, YE, et al Fluorescence versus white-light bronchoscopy for detection of preneoplastic lesions: a randomized study. J Natl Cancer Inst 2001;93,1385-1391[Abstract/Free Full Text]
- Travis, WD, Colby, TV, Corrin, B, et al Histological typing of tumours of lung and pleura. Sobin, LH eds. World Health Organization international classification of tumours 1999 Springer-Verlag. New York, NY:
- Nicholson, AG, Perry, LJ, Cury, PM, et al Reproducibility of the WHO/IASLC grading system for pre-invasive squamous lesions of the bronchus: a study of inter-observer and intra- observer variation. Histopathology 2001;38,202-208[CrossRef][ISI][Medline]
- Keith, RL, Miller, YE, Gemmill, RM, et al Angiogenic squamous dysplasia in bronchi of individuals at high risk for lung cancer. Clin Cancer Res 2000;6,1616-1625[Abstract/Free Full Text]
- Keith, RL, Miller, YE, Kennedy, TC, et al Angiogenic squamous dysplasia (ASD): a biomarker highly associated with squamous cell lung carcinoma [abstract]. Lung Cancer 2003;41,S161
- Tockman, MS, Anthonisen, NR, Wright, EC, et al Airways obstruction and the risk for lung cancer. Ann Intern Med 1987;106,512-518[ISI][Medline]
- Frost, JK, Ball, WC, Jr, Levin, ML, et al Early lung cancer detection: results of the initial (prevalence) radiologic and cytologic screening in the Johns Hopkins study. Am Rev Respir Dis 1984;130,549-554[ISI][Medline]
- Saccomanno, G, Saunders, RP, Archer, VE, et al Cancer of the lung: the cytology of sputum prior to the development of carcinoma. Acta Cytol 1965;9,413-423[ISI][Medline]
- Hecht, SS Cigarette smoking and lung cancer: chemical mechanisms and approaches to prevention. Lancet Oncol 2002;3,461-469[CrossRef][ISI][Medline]
- Pfeifer, GP, Denissenko, MF, Olivier, M, et al Tobacco smoke carcinogens, DNA damage and p53 mutations in smoking- associated cancers. Oncogene 2002;21,7435-7451[CrossRef][ISI][Medline]
- Kriek, E, Rojas, M, Alexandrov, K, et al Polycyclic aromatic hydrocarbon-DNA adducts in humans: relevance as biomarkers for exposure and cancer risk. Mutat Res 1998;400,215-231[ISI][Medline]
- Santella, RM Immunological methods for detection of carcinogen-DNA damage in humans. Cancer Epidemiol Biomarkers Prev 1999;8,733-739[Free Full Text]
- Varella-Garcia, M, Gemmill, RM, Rabenhorst, SH, et al Chromosomal duplication accompanies allelic loss in non-small cell lung carcinoma. Cancer Res 1998;58,4701-4707[Abstract/Free Full Text]
- Girard, L, Zochbauer-Muller, S, Virmani, AK, et al Genome-wide allelotyping of lung cancer identifies new regions of allelic loss, differences between small cell lung cancer and non-small cell lung cancer, and loci clustering. Cancer Res 2000;60,4894-4906[Abstract/Free Full Text]
- Jeltsch, A Beyond Watson and Crick: methylation and molecular enzymology of DNA methyltransferases. Chembiochem 2002;3,274-293[CrossRef][Medline]
- Baylin, SB Mechanisms underlying epigenetically mediated silencing in cancer. Semin Cancer Biol 2002;12,331-337[CrossRef][ISI][Medline]
- Herman, JG, Graff, JR, Myohanen, S, et al Methylation-specific PCR: a novel PCR assay for methylation status of CpG islands. Proc Natl Acad Sci U S A 1996;93,9821-9826[Abstract/Free Full Text]
- Santos Romeo, M, Sokolova, IA, Morrison, LE, et al Chromosomal abnormalities in non-small cell lung carcinomas and in bronchial epithelia of high-risk smokers detected by multi-target interphase fluorescence in situ hybridization. J Mol Diagn 2003;5,103-112[Abstract/Free Full Text]
- Sugita, M, Geraci, M, Gao, B, et al Combined use of oligonucleotide and tissue microarrays identifies cancer/testis antigens as biomarkers in lung carcinoma. Cancer Res 2002;62,3971-3979[Abstract/Free Full Text]
TOP
Introduction
Premalignant Lesions in the...
