HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Fong, K. M. Articles by Minna, J. D. Search for Related Content PubMed PubMed Citation Articles by Fong, K. M. Articles by Minna, J. D.

J Thorac Cardiovasc Surg 1999;118:1136-1152
© 1999 Mosby, Inc.

BASIC SCIENCE REVIEW

MOLECULAR PATHOGENESIS OF LUNG CANCER

From the Hamon Center for Therapeutic Oncology Research,^c University of Texas Southwestern Medical Center, Dallas, Tex, The Prince Charles Hospital,^a Brisbane, Australia, and the Department of Clinical Preventive Service,^b Nagoya University Hospital, Nagoya, Japan.

Supported by Lung Cancer SPORE P50 CA70907.

Address for reprints: John D. Minna, MD, Hamon Center for Therapeutic Oncology Research, University of Texas Southwestern Medical Center, Dallas, TX 75235-8593 (E-mail: minna{at}simmons.swmed.edu ).

	Abstract

Top Abstract Introduction Molecular alterations in lung... Specific molecular alterations... Molecular alterations in normal... Inherited lung cancer... Clinical implications References

 
Lung cancer is the largest cancer killer of men and women in the united states. In addition to the progress made from antismoking primary prevention measures, new tools to help treat patients with lung cancer are emerging from the rapid advances in knowledge of the molecular pathogenesis of lung cancer. These tools include molecular and cellular biology and are starting to provide an insight into how the tumor cell, by altering oncogenes and tumor suppressor genes, achieves growth advantage, uncontrolled proliferation and metastatic behavior via disruption of key cell-cycle regulators and signal transduction cascades. Moreover, new knowledge is being developed in terms of the molecular definition of individual susceptibility to tobacco smoke carcinogens. These tools are being translated into clinical strategies to complement surgery, radiotherapy, and chemotherapy and also to assist in primary and secondary prevention efforts. This review summarizes current knowledge of the molecular pathogenesis of lung cancer. From this we know that respiratory epithelial cells require many genetic alterations to become invasive and metastatic cancer. We can detect cells with a few such changes in current and former smokers, offering the opportunity to intercede with a biomarker-monitored prevention and early detection effort. This will be coupled with new advances in computed tomography–based screening. Finally, because the molecular alterations are known, new mechanism-based therapies are being developed and brought to the clinic, including new drugs, vaccines, and gene therapy, which also must be integrated with standard therapies.

	Introduction

Top Abstract Introduction Molecular alterations in lung... Specific molecular alterations... Molecular alterations in normal... Inherited lung cancer... Clinical implications References

 
Although much is known about the causes and clinical features of lung cancer, including outcome according to TNM stages and histologic features, only recently have physicians been able to delve into the molecular basis of lung cancer pathogenesis and relate these to clinical factors. Here we summarize the major findings resulting from the explosion of knowledge from molecular studies of lung cancer, and we attempt to provide some insight as to how these findings may affect clinical practice in the future. All of these efforts are predicated on continued efforts in preventing smoking initiation and assisting in smoking cessation. They also need to be coupled to recent advances in computed tomography–guided screening for early diagnosis.

	Molecular alterations in lung cancer cells (Table I)

Top Abstract Introduction Molecular alterations in lung... Specific molecular alterations... Molecular alterations in normal... Inherited lung cancer... Clinical implications References

 
Markers of genetic instability
Lung cancer is the end stage of multiple-step carcinogenesis, in most cases driven by genetic and epigenetic damage caused by chronic exposure to tobacco smoke carcinogens. The genetic instability in human cancers appears to exist at two levels: at the chromosomal level, including large-scale losses and gains, and at the nucleotide level, including single or several base changes. ¹ Lung cancers harbor many numeric chromosome abnormalities (aneuploidy) and structural cytogenetic abnormalities including deletions and nonreciprocal translocations. Chromosomal instability leading to aneuploidy is associated with the loss of function of a mitotic checkpoint. Exactly how this loss comes about in lung cancer is not known. Tumor DNA copy number aberrations can be more finely mapped by means of newer molecular cytogenetic techniques such as comparative genomic hybridization, which also demonstrated multiple abnormalities in lung cancer. ^2,3

	Specific molecular alterations in lung cancer cells

Top Abstract Introduction Molecular alterations in lung... Specific molecular alterations... Molecular alterations in normal... Inherited lung cancer... Clinical implications References

 
Growth stimulation and oncogenes
Autocrine and paracrine growth stimulatory loops exist in lung cancers as a consequence of the expression of growth factors, regulatory peptides, and their receptors by either the cancerous or adjacent normal cells. Several but not all components of these stimulatory pathways are proto-oncogene products.

Gastrin-releasing peptide (GRP)/bombesin (BN) autocrine loop
GRP is the 27 amino acid mammalian homologue of the amphibian peptide BN. GRP/BN functions include a role in lung development and repair. ¹⁰ Immunohistochemical studies show that approximately 20% to 60% of SCLC cancers express GRP/BN, whereas non-SCLC (NSCLC) express GRP/BN less frequently. ¹¹ The human GRP/BN receptor subtypes belong to the G-protein–coupled receptor superfamily and include GRP-, neuromedin B–, and BN subtype–3 receptors; all of these can be expressed in SCLC, NSCLC, and in some biopsy specimens of bronchial epithelium of smokers. ^12,13 However, lung cancers so far have not been found to have mutations of either GRP/BN or the GRP/BN receptor; thus the mechanisms for "reactivation" of these embryonic regulatory loops are yet unidentified. The GRP/BN autocrine loop is an important growth stimulatory loop in lung cancer, particularly SCLC. The in vitro formation of soft agar clones and the in vivo growth of nude mouse xenografts of SCLC cell lines are inhibited by a neutralizing monoclonal antibody directed against GRP/BN, as well as by antagonists of BN. ^14,15 A clinical trial of the anti-BN monoclonal antibody has shown some antitumor activity in previously treated patients with SCLC. ¹⁶ This system appears to play an early role in pathogenesis, since there is an increased likelihood of expression of the GRP receptor messenger RNA in the respiratory epithelium of some individuals with a history of prolonged tobacco exposure, and expression of the GRP receptor mRNA is accompanied by responsiveness of these respiratory epithelial cells in vitro to the mitogenic effects of BN-like peptides. ¹³ These effects also appear to persist after smoking cessation.

ERBB family
NSCLCs but not SCLCs often demonstrate abnormalities of the neuregulin receptors ERBB2 and ERBB1, a family of transmembrane receptor tyrosine kinases. On ligand binding, ERBB receptors homodimerize or heterodimerize, thereby inducing intrinsic kinase activities that initiate intracellular signal transduction cascades including the MAP kinase pathway. ERBB2 (HER2/neu) is highly expressed in approximately 30% of NSCLCs, especially adenocarcinomas. ^17,18 Transfection experiments suggest that ERBB2 overexpression contributes to tumorigenicity in immortalized human bronchial epithelial cells. ¹⁹ An anti-ERBB2 monoclonal antibody inhibited the in vitro growth of ERBB2 expressing NSCLC cell lines. ²⁰ High ERBB2 levels are associated with the multiple drug resistance phenotype ²¹ and increased metastatic potential in NSCLC, ²² which may help explain the poor clinical outcome linked to ERBB2 overexpression reported by some investigators. ^23,24 Monoclonal antibodies against the ERBB2 receptor (Herceptin) have entered clinical trials combined with chemotherapy in both breast cancer and lung cancer.

ERBB1 (epidermal growth factor receptor) regulates epithelial proliferation and differentiation and is usually activated in lung cancer cells by overexpression by an unknown mechanism. The production of ERBB1 ligands such as epidermal growth factor and transforming growth factor ß by lung cancer cells expressing cognate receptors indicates an autocrine loop. ^18,25 More common in NSCLC, ERBB1 activation may be related to tumor stage and differentiation. ^26,27 Monoclonal antibodies against the ERBB1 receptor (C225, ImClone Systems Incorporated, New York, NY) are entering clinical trials combined with chemotherapy. In addition, several tyrosine kinase inhibitors that have some selectivity as epidermal growth factor receptor (ERBB1) blockers (CP358774, ZD1839) are also entering clinical trials.

Other membrane tyrosine kinases
Hepatocyte growth factor stimulates epithelial cells to proliferate, move, and undergo complex differentiation programs. During fetal lung development, hepatocyte growth factor acts as a mesenchyme-derived morphogenic factor. ²⁸ The constitutively low levels of hepatocyte growth factor increase in response to lung or distant injury. ²⁹ Hepatocyte growth factor stimulates mitogenesis and/or motogenesis of human bronchial epithelial, alveolar type II, and SCLC cells in vitro. The receptor for hepatocyte growth factor is the MET proto-oncogene; it is generally expressed in normal lung and both SCLC and NSCLC; hepatocyte growth factor, however, is expressed mostly in NSCLCs, suggesting an autocrine loop specific for NSCLC. ^30,31 Clinically, high hepatocyte growth factor levels were associated with a poor outcome in patients with resectable NSCLC. ³² The KIT proto-oncogene encodes the tyrosine kinase receptor, CD117. It is co-expressed together with its ligand, stem cell factor, in many SCLCs, ^33,34 representing another autocrine loop that may provide a growth advantage or mediate chemo-attraction. Other putative loops involve insulin-like growth factor 1, insulin-like growth factor 2, and the type I insulin-like growth factor receptor, which are frequently co-expressed in both SCLC and NSCLC, ³⁵ as well as platelet-derived growth factor and its receptor. ³⁶

Nicotine and opioid receptors
Lung cancer cells express nicotine receptors and a variety of opioid receptors, as well as being able to produce several opioid peptides. ³⁷ Intriguingly, opioids can inhibit lung cancer cell growth and cause apoptosis. ³⁸ This leads to the apparent paradox whereby lung cancer cells express a negative opioid autocrine regulatory loop. Regardless, nicotine acting through specific, high-affinity, nicotinic acetylcholine receptors on lung cancer cells can antagonize the apoptotic effect of opioids. Thus it is possible that inhaled nicotine from smoking may play a role in lung cancer pathogenesis by inhibiting apoptosis, for example, in precursor lesions. ³⁸

RAS
The RAS proto-oncogene family (KRAS, HRAS, and NRAS) encodes plasma membrane proteins and is activated in some lung cancers by point mutations, and RAS mutations result in inappropriate prolonged signaling for continued cell division. KRAS is the most frequently activated RAS gene in lung cancer, usually by mutations at codon 12 but occasionally codons 13 and 61. KRAS mutations affect approximately 20% to 30% of lung adenocarcinomas and 15% to 20% of all NSCLC but rarely SCLCs. ¹¹ There is subtype heterogeneity; for example, KRAS mutations are present in parenchymal but not bronchial adenocarcinomas, ³⁹ and goblet-cell subtypes of adenocarcinoma have the highest frequency of KRAS mutations. ⁴⁰ KRAS mutations correlate with smoking ⁴¹ and most mutations comprise the G-T transversions expected from bulky DNA adducts caused by the polycyclic hydrocarbons and nitrosamines in tobacco smoke. ⁴² The presence of KRAS mutations in a patient’s tumor appears to portend a poor prognosis in NSCLC, although this is debated. ^13,43-46 However, KRAS mutations were not associated with in vitro resistance against a range of chemotherapeutic agents in NSCLC cell lines. ⁴⁷ Moreover, neither chemotherapy sensitivity nor survival correlated with KRAS mutations in a prospective study of advanced lung adenocarcinoma. ⁴⁸ To be active in the cell, RAS has to have a lipid modification (farnesylation) regulated by an enzyme (farnesyltransferase). Several farnesyltransferase inhibitors (from Bristol Myers Squibb, Janssen, and Merck) are currently in clinical trials against lung cancer.

MYC
RAS signaling ultimately activates nuclear proto-oncogene products like MYC, which on heterodimerization transcriptionally activates downstream genes, which drives cells to grow. The myc family comprises MYC, MYCN, and MYCL (originally isolated from a lung cancer ⁴⁹). MYC is the most frequently activated and affects SCLC and NSCLC, whereas the others usually only affect SCLC. Activation occurs by gene amplification (~20-115 copies per cell) or by transcriptional dysregulation, both of which lead to protein overexpression. Richardson and Johnson ¹¹ concluded from 17 studies that 18% to 31% of SCLCs had amplification of one MYC family member. Conversely, only 8% to 20% of NSCLCs were affected. MYC amplification appears to occur more frequently in chemotherapy-treated patients and the "variant" SCLC subtype ⁵⁰ and may correlate with adverse survival. Last, in vitro growth inhibition of an SCLC cell line by all-trans -retinoic acid was associated with increased neuroendocrine differentiation, increased MYCL, and decreased MYC expression. ⁵¹

Tumor suppressor genes (TSGs)

p53
p53 maintains genomic integrity in the face of DNA damage from or ultraviolet irradiation and carcinogens. DNA damage or hypoxia results in a rapid increase of p53, which acts as a sequence-specific transcription factor regulating downstream genes including p21, MDM2, GADD45, and BAX, thereby helping to regulate the G1/S cell cycle transition, G2/M DNA damage checkpoint, and apoptosis. Dysfunction of p53 allows the inappropriate survival of genetically damaged cells, setting the stage for the accumulation of multiple mutations and the subsequent evolution of a cancer cell. p53 plays a critical role in lung cancer; its chromosome 17p13 locus is frequently hemizygously deleted, and mutational inactivation of the remaining allele occurs in 75% or more of SCLCs and about 50% of NSCLCs. ^42,52,53 p53 mutations in lung tumors correlate with cigarette smoking and are mostly the G-T transversions expected of tobacco smoke carcinogens. ⁴² Furthermore, benzo(a )pyrene, a major tobacco smoke carcinogen, selectively forms adducts at the major p53 mutational hot spots in bronchial epithelial cells. ⁵⁴ Missense p53 mutations can prolong the protein half-life, leading to easily detected mutant p53 protein by immunohistochemical studies. ⁵⁵ However, other types of p53 mutations do not correlate with immunohistochemical staining. Studies have shown abnormal p53 expression by immunohistochemistry in 40% to 70% of SCLCs and 40% to 60% of NSCLCs (squamous cell carcinomas higher than adenocarcinomas). ^56-58 Although the prognostic value of p53 immunohistochemistry or mutational abnormalities is controversial, ⁵⁹ mutations have been linked to response to cisplatin-based chemotherapy in NSCLC ⁶⁰ and response to radiotherapy. ⁶¹

Can a mutant p53 be replaced with a normal wild-type copy with beneficial effect? The in vitro reintroduction of wild-type p53 into lung cancer cells that also have various other genetic abnormalities besides p53 blocks tumor cell growth by inducing apoptosis. ⁶² Direct injection, using a retroviral vector containing wild-type p53 , into tumors (either by bronchoscopy or by a computed tomography–guided needle) in 9 patients with NSCLC in whom conventional treatment had failed led to tumor regression in a third of patients. ⁶³ Other promising alternative strategies for delivering p53 as gene replacement therapy include liposomal-p53 complexes delivered endobronchially and adenovirus-mediated p53 transfer given locally, endobronchially, or even systemically. ⁶⁴

Antibodies against the p53 protein develop in some 15% to 25% of patients with lung cancer, indicating that mutant p53 protein overexpression can lead to a humoral immune response. Whereas the development of p53 antibodies appears associated with squamous histology, the value of the antibodies as diagnostic or prognostic markers is controversial. ^65-68 Nonetheless, p53 antibodies appear to be reduced during chemotherapy. ⁶⁹ DNA from tumor cells can be released into patients’ plasma. Thus additional future early detection strategies may involve screening by means of polymerase chain reaction technology for the presence of DNA in patient plasma that contains mutant p53 sequences. ⁷⁰

There are other proteins homologous to p53, including p51 at chromosome loci 3q28 and p73 at lp36, both of which can induce growth suppression and apoptosis. Mutations of p51 appear to be infrequent in lung cancer. ⁷¹ Similarly, p73 mutations are absent or infrequent in lung cancers despite frequent allele loss of the 1p36 region (42%). ^72,73 In fact, it may be that p73 "activation" is important since wild-type p73 is highly expressed in lung tumors. ⁷⁴

p21 is a p53-responsive gene that inhibits cyclin/ cyclin-dependent kinase (CDK) complexes at the G1 phase. Although not somatically mutated in lung cancer, p21 was overexpressed in 65% to 75% of NSCLCs, especially in well-differentiated tumors. ⁷⁵ One NSCLC study reported that p21 expression was linked to a favorable outcome, ⁷⁶ whereas another suggested that concordant expression of p21 and transforming growth factor ß1 predicts better survival than discordant expression. ⁷⁷

The MDM2 oncogene product inhibits p53 function by blocking its regulation of target genes and also enhances proteasome-dependent degradation of p53. Conversely, p53 regulates (increases) the expression of MDM2 by directly binding and activating the MDM2 promoter. The MDM2 protein is overexpressed in 25% of NSCLCs, ⁷⁸ and MDM2 expression without abnormal p53 expression (mutation) has been reported to be a favorable prognostic factor. Thus different ways of inactivating the p53 pathway may have different clinical outcomes.

p16-cyclin D1-CDK4-RB pathway

The retinoblastoma gene, RB
The pl6-cyclin D1-CDK4-RB pathway is central to controlling the G1-S transition of the cell cycle, and its components are functionally altered or mutated in many cancers. The major growth suppressing function of RB is by blocking G1-S progression. Inactivation of both RB alleles at chromosome region l3ql4 is common in lung cancers, ⁷⁹ with protein abnormalities detected in about 90% of SCLCs and 15% to 30% of NSCLCs. ^58,80-82 Functional RB loss can include deletion, nonsense mutations, or splicing abnormalities, frequently leading to a truncated RB protein. Whether absent RB expression is associated with poor prognosis in NSCLCs is controversial. ^58,82-84 Functionally, in vitro reintroduction into tumor cells of a wild-type RB suppresses SCLC growth. ⁸⁵ The inherited germline form of mutant RB is associated with the development of the childhood disease retinoblastoma, and relatives of retinoblastoma patients carrying germline RB mutation are about 15 times more likely to die of lung cancer than the general population. ⁸⁶ Thus in this rare inherited disorder there appears to be an example of genetic predisposition to lung cancer. Two RB- related genes also have been implicated in lung cancer including, p107 and pRB2/p130, where decreased expression of these proteins is associated with more aggressive histologic behavior. ⁸⁷

Cyclin D1 and CDK4
Cyclin D1 inhibits the activity of RB by stimulating its phosphorylation by CDK4. Thus cyclin D1 overexpression is an alternative mechanism to RB mutation for disrupting the pl6-cyclin D1-CDK4-RB pathway. Cyclin D1 was overexpressed in 25% to 47% of primary NSCLCs and in some cases has been associated with poor prognosis. ^88-90 Transfection of a cyclin D1 antisense construct into lung cancer cell lines (and thus inhibiting any tumor-related cyclin D1 production) induces destabilization of RB and retards growth. ⁹¹ Although amplification of CDK4 has been reported in other human malignancies, its role in lung cancer has not yet been studied. However, a CDK inhibitor, flavopiridol, is in clinical trials against lung and other cancers.

p16
Chromosome 9p allele loss is observed in many lung cancers. ⁹² p16 (p16 ^INK4 CDKN2) is situated at 9p21 and frequently undergoes allele loss and mutation in lung cancer. ⁴ As p16 regulates RB function by inhibiting CDK4 and CDK6 kinase activity, p16 inactivation is thus another tumor strategy for disrupting the pl6-cyclin D1-CDK4-RB cell cycle control pathway. p16 abnormalities are frequent in NSCLC but rare in SCLC (where the pathway is usually disrupted by an RB mutation). Homozygous deletion or point mutations occur in 10% to 40% of NSCLCs. ^93,94 Hypermethylation in the 5' CpG island of the retained allele may also cause loss of expression of pl6 from the remaining allele in lung cancer. ^95,96 These mechanisms contribute to the loss of p16 expression in NSCLC, functionally analogous to the preferential RB inactivation in SCLC, with the common end result of pl6-cyclin D1-CDK4-RB pathway disruption. ^84,97-101 Co-inactivation of RB and p16 in any one tumor is rare, but cyclin D1 overexpression can coexist with these abnormalities in the same tumor. ¹⁰² Notably, 10% to 30% of NSCLCs appear normal for RB and p16, indicating involvement of cyclin D1, CDK4, or other pathway members in these cases. Perhaps 30% to 50% of early-stage primary NSCLCs do not express p16. p16 alteration and loss of function in lung cancer may be associated with tumor progression, clinical stage, and survival, although not all studies concur. ^{58,84,101,103,104}

p19^ARF
The p16 locus also encodes a second alternative reading frame protein, p19^ARF, that overlaps with p16. The amino acid sequence of these two alternatively spliced messages are different. However, they both appear to be important in growth regulation. p19^ARF binds to the MDM2-p53 complex and prevents p53 degradation, resulting in p53 activation. Immunohistochemical analysis suggests that the p19^ARF protein expression was more frequently lost in tumors with neuroendocrine features. ¹⁰⁵ Thus one genetic locus at 9p21 produces two products, p16 and p19^ARF, both of which play a critical role in growth regulation: p16 with the RB pathway and p19^ARF with the p53 pathway.

Reduced expression of another CDK inhibitor gene, p27^KIP1, was correlated with poor prognosis in NSCLC. ^106,107 In contrast, most SCLCs exhibit increased p27^KIP1 staining, suggesting a possible link with neuronal differentiation. ¹⁰⁷ p57^KIP2 at chromosome region 11p15 is imprinted with maternal expression and p57^KIP2 expression is down-regulated by selective loss of the maternal alleles in some lung cancers. ¹⁰⁸

Other candidate TSGs
Cytogenetics and allelotyping have revealed many hemizygous and some homozygous deletions at multiple chromosomal regions in lung cancers. In the classic TSG paradigm, the presence of underlying TSGs is suggested as the target of these genetic losses. The chromosomal regions showing hemizygous deletions include 1p, 1q, 2q, 3p, 4p, 4q, 5q, 6p, 6q, 8p, 8q, 11p, 11q, 14q, 17q, 18q, and 22q. ^109-118 Although several of these chromosomal arms contain known or candidate TSGs (such as APC at 5q21, WT1 at 11p13, and NF2 at 22q12), these genes are not known to be mutated in lung cancer.

Very frequent hemizygous loss of one chromosome 3p allele occurs in lung cancer, ¹¹⁹ implicating multiple TSGs in at least three regions including 3p25-26, 3p21.3-22, and 3p14-cen. ¹²⁰ Moreover, five separate homozygously deleted regions at 3p12-13, 3p14.2, and 3p21 occur in several lung cancer cell lines. ⁴ One candidate is FHIT at 3p14.2, which undergoes frequent hemizygous and occasional homozygous deletion in lung cancer cells and encodes a dinucleoside hydrolase. Lung cancer cells frequently (40%-80%) express abnormal mRNA transcripts of FHIT but nearly always also express wild-type FHIT transcripts. ^121,122 However, unlike classic TSG inactivation, FHIT point mutations are rare, ^121,122 and abnormal transcripts can be found in normal lung tissue. ¹²³ Notwithstanding, Fhit is expressed in normal lung but absent Fhit protein occurs in primary lung tumors. ¹²⁴ Moreover, FHIT allele loss is more common in smokers than nonsmokers ¹²⁵ and may be associated with a poorer survival in NSCLC. ¹²⁶ Furthermore, reintroduction of exogenous wild-type FHIT suppressed tumorigenicity of lung cancer cell lines in nude mice, ^127,128 although others have reported that FHIT transfection does not suppress tumor growth of human cancer cell lines. ¹²⁹ The 3p21.3 region has been extensively examined for putative TSGs, ¹¹⁹ particularly at a 600-kb region homozygously deleted in three SCLC cell lines, ^130,131 while another 800-kb deletion region exists at 3p21. ¹³²

Protein tyrosine phosphatases may play a role in tumor suppression because of their ability to antagonize the growth-promoting protein kinases. A candidate is the phosphatase encoded by the PTEN gene at chromosome 10q23. ¹³³ PTEN is mutated in a few primary lung cancers, and several lung cancer cell lines show homozygous deletions of this gene. ¹³⁴ Another candidate at 10q, region 25.3-26.1, is DBMT1, which is frequently down-regulated and occasionally homozygously deleted in lung cancer. ¹³⁵

11q23-24 chromosome is a region containing frequent allelic loss including 71.4% of lung cancer cell lines, with two distinct minimal regions of loss. ¹³⁶ One region contains the gene encoding the beta isoform of the A subunit of the human protein phosphatase 2A (PPP2R1B). Recently, mutations were described in lung and colon cancers, suggesting that PPP2R1B functions as a TSG. ¹³⁷

Apoptosis
Tumor cells often escape the physiologic response (programmed cell death or apoptosis) to cellular and DNA damage. Expression of the BCL-2 anti-apoptotic proto-oncogene is relatively higher in squamous cell carcinoma (25%-35%) than in adenocarcinoma (~10%). ^138-141 It also correlates with neuroendocrine differentiation and is higher in SCLCs (75%-95%) than in NSCLCs. ^138,142 The latter is interesting inasmuch as SCLCs are more responsive to chemotherapy, which usually results in tumor apoptosis. Clinically, there is a trend toward longer survival in SCLCs that express BCL-2, ¹⁴² but this relationship is controversial in NSCLC. ^{139-141,143,144} BAX, which counteracts BCL-2 and promotes apoptosis, is a downstream transcription target of p53. Expression of BAX and BCL-2 are inversely related in neuroendocrine cancers; most typical and atypical carcinoids have low BCL-2 and high BAX expression, with the reverse in most SCLCs and large cell neuroendocrine cancers. ⁵⁶ There are multiple anti-apoptosis approaches that are in preclinical development or in early clinical trials. These include antisense BCL2 in SCLC (to inhibit the translation of BCL2 mRNA and down-regulate BCL2 protein expression), BCL-xL antisense in NSCLC, and a bispecific BCL2-BCLxL antisense to target both SCLC and NSCLC.

The receptor Fas (CD95) and the Fas ligand (FasL) also play key roles in the initiation of apoptotic pathways. FasL can induce apoptosis in activated T cells, a mechanism thought to contribute to immune privilege sites. Thus it is notable that lung cancer cell lines and tumors express FasL, and co-culture of lung cancer lines with a Fas-sensitive human T-cell line induces T-cell apoptosis. ¹⁴⁵ Meanwhile, cell surface Fas expression has been shown to be reduced in adenocarcinomas, which may also account for resistance to Fas-mediated apoptosis. ¹⁴⁶ Thus it appears that lung cancers can make FasL but cannot respond to it because they lack the receptor, Fas. However, the FasL they produce can function to inactivate T cells, providing for a mechanism of escape from a patient’s immune response.

Telomerase activity
Human telomeres are specialized structures located at the ends of chromosomes comprising TTAGGG tandem repeats bound to specific proteins. During normal cell division, the absence of telomerase activity is associated with progressive telomere shortening, perhaps representing an intrinsic "cellular clock," leading to cell senescence and normal cell "mortality." In contrast, germ cells, some stem cells, and most cancer cells have telomerase activity, which is able to compensate for telomere shortening by replacing the hexameric repeats, leading to potential cellular "immortality." The telomerase holoenzyme is composed of both RNA and protein (enzymatic) subunits. Both of these components are required for activity and can be overexpressed in cancer, with the final measure being telomerase enzyme activity. Nearly all SCLCs and 80% to 85% of NSCLCs have high levels of telomerase activity. ^147,148 High telomerase activity was associated with increased cell proliferation rates and advanced stage in NSCLCs. ¹⁴⁸ Additionally, telomerase activity and/or dysregulated RNA expression are frequent in carcinoma in situ lesions, implicating involvement early in lung cancer development. ¹⁴⁹

Aberrant methylation
Methylation of the promoter regions of genes is frequently used to regulate gene expression with hypermethylation of the promoter region, resulting in loss of expression. Abnormalities of DNA methylation occur consistently in human neoplasia, and promoter region hypermethylation in 5' CpG islands may transcriptionally silence and thus inactivate TSGs such as the RB, VHL, and p16 genes. In NSCLCs, p16 hypermethylation contributes to down-regulation of p16 expression and occurs at an early stage in lung cancer development. ^95,96,150 A series of genes have been found to undergo promoter methylation in lung cancer but not in the associated normal lung. In a small series of 22 patients, the following methylation frequencies were observed in NSCLC but not in the normal lung tissue: p16 (41%), death-associated protein (DAP) kinase (23%), glutathione S transferase (GSTP1) (9%), MGMT (27%), or any of the markers (68%). ¹⁵¹ In addition, DNA containing these methylated sequences could be detected in patients’ sera at the following frequencies when the tumors were positive for the methylated marker: p16 (33%), DAP kinase (80%), GSTP1 (50%), MGMT (66%), or any marker (73%). ¹⁵¹ Because the methylated DNA sequences can be found even when they represent a small fraction within total normal DNA, they are very attractive candidates for early molecular detection tools and for following chemoprevention studies. Other regional sites of hypermethylation have been found in lung cancer, including sites at 3p, 4q34, 10q26, and 17p13, although the precise gene targets at these sites are uncertain and the significance is not yet apparent. ^152,153 Methylation also plays a role in mediating genomic imprinting, which is a gamete-specific modification causing differential expression of the two alleles of a gene. Loss of genomic imprinting of the IGF2 and H19 genes at 11p15 occurs in some lung cancers, and in the case of IGF2 this could lead to IGF2 expression and lung cancer growth. ^154,155

Angiogenesis and metastases
Tumors cannot exceed l- to 2-mm³ volume without new blood vessels developing; thus they require angiogenic factors early in their pathogenesis. Microvessel density may be regarded as a morphologic measure of tumor angiogenesis and has been variably linked to metastases and impaired survival in lung cancer. ^156-159 Tumor angiogenesis is complex and controlled by a diverse family of inducers and inhibitors governing angiogenesis, regulating endothelial cell proliferation and migration. ¹⁶⁰ Vascular endothelial growth factor (VEGF) and basic fibroblast growth factor are among the more important angiogenesis inducers. Expression of VEGF was associated with intratumoral microvessel density and tumor cell proliferation in squamous lung cancer. ^161-163 Expression was higher in lung cancers with nodal metastasis than in those without ¹⁶⁴ and was associated with poor prognosis in NSCLC. ^162,165 In squamous cell carcinomas, the VEGF receptor, Flt-1, is frequently expressed, suggesting a possible autocrine role of VEGF. ¹⁶⁵ Mutant p53 acts synergistically with hypoxia to induce VEGF expression ¹⁶⁶ whereas wild-type p53 up-regulates the expression of the anti-angiogenic protein TSP-1. ¹⁶⁷ Basic fibroblast growth factor is expressed in about 70% of NSCLCs, but its prognostic significance is controversial. ^168,169 In addition, expression of the angiogenic platelet-derived endothelial cell growth factor is correlated with tumor angiogenesis and a worse prognosis in N0-stage NSCLCs. ¹⁷⁰ Angiogenic CXC chemokines such as interleukin 8 (IL-8) and epithelial neutrophil attractant 78 (ENA-78), and angiostatic chemokines such as gamma interferon–inducible protein 10 (IP-10), may also be relevant in lung cancer. ¹⁷¹ IL-8 may be an angiogenic factor for SCLC, and IL-8 inhibition by passive immunization reduces tumorigenesis of NSCLC xenografts in SCID mice. ¹⁷² In contrast, IP-10 was shown to inhibit NSCLC tumorigenesis and spontaneous metastases. ¹⁷³

There are multiple new therapeutic strategies in lung cancer directed against angiogenesis in general and the VEGF system in particular. Currently, ongoing clinical trials are testing a recombinant, "humanized," monoclonal antibody against VEGF (rhuMVEGF, Genentech, Inc, South San Francisco, Calif) that blocks the binding of VEGF to its receptor. The "humanization" process takes a mouse monoclonal antibody and replaces human sequences within it so that no immune response is elicited when the antibody is given to patients. Other preclinical studies or clinical trials involve a monoclonal antibody directed against VEGF receptor 2 (KDR/FLK1, ImClone Systems Incorporated, New York, NY), a tyrosine kinase inhibitor (SU5416, SUGEN, Inc, South San Francisco, Calif), and a ribozyme (Angiozyme) directed against the VEGF receptor 1 (Flt-l) mRNA and thus inactivating the receptor by inactivating the mRNA.

Tumor metastasis is a complex multistep host-dependent process. Cell adhesion molecules have been implicated in epithelial differentiation, carcinogenesis, and metastasis. E-cadherin is responsible for the organization, maintenance, and morphogenesis of epithelial tissues, and reduced expression has been associated with tumor dedifferentiation, increased lymphogenous metastasis, and poor survival in NSCLC. ¹⁷⁴ The integrins that bind to the major basement membrane components (collagen and laminin) are down-regulated in NSCLC. ¹⁷⁵ Reduced integrin 3 expression is linked with poor prognosis in adenocarcinomas. ¹⁷⁶ Others have reported that the v6 isoform of the CD44 lymphocyte homing receptor correlates with adverse prognosis in NSCLC. ¹⁷⁷ The degradation of the extracellular matrix and basement membrane, particularly type IV collagen, by tumor cells is essential for invasion and metastasis. Type IV collagen expression of adenocarcinomas is progressively lost with increasing tumor dedifferentiation. ¹⁷⁸ The matrix metalloproteinase enzymes that degrade stroma and thus could lead to tumor invasion and metastasis include collagenases, gelatinases, stromelysins, membrane-type matrix metalloproteinases, matrilysin, and metalloelastase. Gelatinase A expression has been shown in both SCLCs (~50%) and NSCLCs (~65%). ¹⁷⁹ The activation of gelatinase A in lung cancer was correlated with membrane-type matrix metalloproteinase mRNA expression, whose product is one of the pro-gelatinase activators. ¹⁸⁰ Stromelysin-3 has been shown to be more strongly expressed by stromal elements in primary NSCLCs than in adjacent normal lung specimens. ¹⁸¹ Several drug companies have matrix metalloproteinase inhibitors (Prinomastat from Agouron Pharmaceuticals, Inc, [La Jolla, Calif] and BAY129566 from Bayer AG [Leverkusen, Germany]) currently in clinical trials.

	Molecular alterations in normal and preneoplastic at-risk respiratory cells

Top Abstract Introduction Molecular alterations in lung... Specific molecular alterations... Molecular alterations in normal... Inherited lung cancer... Clinical implications References

 
The concept of cancers arising from precursor lesions occurring in a multistep fashion has long been suggested and underpinned by the notion of somatically acquired genetic alterations in preneoplastic cells, which subsequently evolve into invasive cancer by clonal expansion. Morphologically distinct preneoplastic changes (hyperplasia, metaplasia, dysplasia, and carcinoma in situ) can be observed in bronchial epithelium before overt, invasive cancer develops. The sequential morphologic changes that occur in preneoplastic lesions are better described for central squamous cell carcinomas than for adenocarcinomas or SCLCs. Some reversal of morphology can occur after smoking cessation even though the elevated risk of lung cancer does not completely return to baseline, raising the possibility that there may be persistent areas of at-risk bronchial epithelium, perhaps from being irreversibly genetically altered. These preneoplastic cells and even macroscopically normal bronchial epithelium adjacent to cancers can contain genetic abnormalities identical to those in invasive cancer cells. These include allele loss at several loci (3p, 9p, 8p, 17p), MYC and RAS up-regulation, cyclin D1 expression, p53 immunoreactivity, BCL-2 overexpression, and DNA aneuploidy. ^182-186 Examination of microdissected, preneoplastic lesions suggests that the earliest change is allele loss at chromosome regions 3p, then 9p, 8p, 17p (with p53 mutation), 5q, and then RAS mutations. ^187-190 However, topographic analyses by others have indicated that KRAS activation also can occur at early stages. ¹⁸⁵ Moreover, a potential precursor lesion of adenocarcinomas, atypical alveolar hyperplasia, can also have KRAS mutations. ^39,191

Recent studies have suggested that the genetic changes found in invasive cancers and preneoplasia can also be identified in morphologically normal appearing bronchial epithelium from current or former smokers. ^190,192,193 In general, such genetic changes are not found in the bronchial epithelium from true, lifetime never smokers. The earliest change appearing in histologically normal epithelium appears to be allele loss at one of several different 3p sites. Approximately 50% of histologically normal specimens from current or former smokers show 3p allele loss. There is also a progressive increase in the frequency and size of the areas undergoing allele loss as the specimens progress from normal, to hyperplasia or metaplasia (mildly abnormal), to dysplasia, to carcinoma in situ. ^190,193 Moreover, some smokers exhibited allele loss at multiple sites, a molecular picture frequently observed in carcinoma in situ and invasive cancers. ¹⁹³

These observations are consistent with the multistep model of carcinogenesis and a "field cancerization" process, whereby the whole tissue region is repeatedly exposed to carcinogenic damage (tobacco smoke) and is at risk for the development of multiple, separate, clonally unrelated foci of neoplasia. The widespread aneuploidy that occurs throughout the respiratory tree of smokers supports this notion. ¹⁹⁴ Interestingly, in patients with lung cancer, geographically separate regions of bronchial mucosa may demonstrate loss of the same allele of a polymorphic marker ("allele-specific loss"). Although this could be consistent with a clone populating a large part of the respiratory tree, other mechanisms are not excluded. ^187,188 However, the presence of the same somatic p53 gene point mutation at widely dispersed preneoplastic lesions in a smoker without invasive lung cancer is consistent with the clonal expansion of a single progenitor clone throughout the respiratory tree. ¹⁹⁵

	Inherited lung cancer susceptibility

Top Abstract Introduction Molecular alterations in lung... Specific molecular alterations... Molecular alterations in normal... Inherited lung cancer... Clinical implications References

 
The major classes of carcinogens in tobacco smoke are the polycyclic hydrocarbons (such as benzo[a]pyrene), the nitrosamines, and the aromatic amines. Tobacco smoke carcinogens may be activated enzymatically to chemically reactive electrophiles that form carcinogen DNA adducts. It has been hypothesized that an individual’s susceptibility to cancer may partially be affected by the balance between the capacity to activate inhaled pro-carcinogens (phase I enzymes) and the capacity to detoxify carcinogens (phase II enzymes). ¹⁹⁶ It is increasingly recognized that genetic polymorphisms common in the population can affect each of these processes. Thus an individual’s susceptibility to lung cancer could be determined by these inherited factors coupled with exposure through cigarette smoking. Familial factors predisposing to lung cancer have long been suggested. ^197,198 The phenotypic variations that are thought to be important in lung cancer include polymorphisms at the P450 gene loci, such as CYP1A1, CYP2D6, and the GST gene cluster. ^199,200 Early evidence suggests that different metabolic enzymes may be associated with tumor subtype susceptibility to various tobacco smoke carcinogens. ²⁰¹ Another indication of individual variation was the difference in the number of chromosome breaks in peripheral blood lymphocytes after they were exposed in vitro to benzo(a)pyrene diol-epoxide, a very carcinogenic derivative of benzopyrene. Lung cancer cases and controls had lymphocytes exposed to benzo(a)pyrene diol-epoxide in vitro and then had the chromosomes assayed by fluorescent in situ hybridization for deletions in chromosome region 3p21.3. More deletions were found in the lung cancer cases than in the controls such that the phenotype of high numbers of 3p21.3 deletions was associated with almost a 14-fold increase in risk of having lung cancer. ²⁰² This could prove to be a very useful marker for risk assessment of which individuals are likely to have lung cancer develop. It also provides a correlation between 3p deletions in the target tissue (lung) compared with a surrogate tissue (lymphocytes) after exposure to a respiratory carcinogen.

	Clinical implications

Top Abstract Introduction Molecular alterations in lung... Specific molecular alterations... Molecular alterations in normal... Inherited lung cancer... Clinical implications References

 
We have given examples of how the new molecular tools may change clinical practice in the future and these are summarized in Table II. For instance, the potential of molecular epidemiology is increasingly recognized, with the aim of identifying individual genetic susceptibility factors to lung cancer and individuals at the highest risk for the development of lung cancer. This will be particularly important as we begin large-scale spiral computed tomographic lung cancer screening trials. ^203,204 Next, somatic mutational events occur in airways exposed to tobacco smoke carcinogens and may lead to further refinement of risk prediction in genetically susceptible individuals. However, any such molecular markers used for this and other purposes such as intermediate markers for chemoprevention have to be tested and validated in prospective clinical trials. Apart from this, early detection of lesions remains a high priority for lung cancer workers, ²⁰⁵ and highly sensitive molecular methods are now able to detect changes in one cancer cell among thousands. Bodily fluids are suitable for molecular analysis, but the challenge will be to identify a panel of markers that are highly sensitive and specific for lung cancer. Refinement of histologic diagnosis and categorization of lung cancer subtypes may also be improved with increasing knowledge of subtype-specific molecular changes. Moreover, the molecular basis for tests such as F-18 fluorodeoxyglucose positron emission tomography (FDG-PET) scan may lead to more advanced refinements in their use. ^206,207 We have also given examples of some links to prognosis, but in general most studies have until now been relatively small, retrospective, and quite variable in methodology. Currently, there is no molecular marker that has major clinical prognostic predictive value. Clinically valid conclusions will need to come from large, adequately powered prospective studies. Finally, there is much therapeutic optimism for a variety of rationally based new therapies. Thus, in the next millennium, the molecular basis of lung cancer should provide us with the new tools to counter the clinical nihilism that affects lung cancer treatment. It is likely that these tools will complement existing strategies for lung cancer control.

	References

Top Abstract Introduction Molecular alterations in lung... Specific molecular alterations... Molecular alterations in normal... Inherited lung cancer... Clinical implications References

1. Lengauer C, Kinzler KW, Vogelstein B. Genetic instabilities in human cancers. Nature 1998;396:643-9. [Medline]
   
2. Levin NA, Brzoska PM, Warnock ML, Gray JW, Christman MF. Identification of novel regions of altered DNA copy number in small cell lung tumors. Genes Chromosomes Cancer 1995;13:175-8. [Medline]
   
3. Petersen I, Bujard M, Petersen S, Wolf G, Goeze A, Schwendel A, et al. Patterns of chromosomal imbalances in adenocarcinoma and squamous cell carcinoma of the lung. Cancer Res 1997;57:2331-5. [Abstract/Free Full Text]
   
4. Sekido Y, Fong KM, Minna JD. Progress in understanding the molecular pathogenesis of human lung cancer. Biochem Biophys Acta 1998;1378:F21-59. [Medline]
   
5. Ryberg D, Lindstedt BA, Zienolddiny S, Haugen A. A hereditary genetic marker closely associated with microsatellite instability in lung cancer. Cancer Res 1995;55:3996-9. [Abstract/Free Full Text]
   
6. Rosell R, Pifarré A, Monzó M, Astudillo J, López-Cabrerizo MP, Calvo R, et al. Reduced survival in patients with stage-I non-small-cell lung cancer associated with DNA-replication errors. Int J Cancer 1997;74:330-4. [Medline]
   
7. Adachi J, Shiseki M, Okazaki T, Ishimaru G, Noguchi M, Hirohashi S, et al. Microsatellite instability in primary and metastatic lung carcinomas. Genes Chromosomes Cancer 1995;14:301-6. [Medline]
   
8. Chevillard S, Radicella JP, Levalois C, Lebeau J, Poupon M-F, Oudard S, et al. Mutations in OGG1, a gene involved in the repair of oxidative DNA damage, are found in human and kidney tumors. Oncogene 1998;16:3083-6. [Medline]
   
9. Esteller M, Hamilton SR, Burger PC, Baylin SB, Herman JG. Inactivation of the DNA repair gene O6-methylguanine-DNA methyltransferase by promoter hypermethylation is a common event in primary human neoplasia. Cancer Res 1999;59:793-7. [Abstract/Free Full Text]
   
10. Spurzem JR, Rennard SI, Romberger DJ. Bombesin-like peptides and airway repair: a recapitulation of lung development. Am J Respir Cell Mol Biol 1997;16:209-11. [Medline]
    
11. Richardson GE, Johnson BE. The biology of lung cancer. Semin Oncol 1993;20:105-27. [Medline]
    
12. Fathi Z, Way JW, Corjay MH, Viallet J, Sausville EA, Battey JF. Bombesin receptor structure and expression in human lung carcinoma cell lines. J Cell Biochem Suppl 1996;24:237-46. [Medline]
    
13. Siegfried JM, Gillespie AT, Mera R, Casey TJ, Keohavong P, Testa JR, et al. Prognostic value of specific KRAS mutations in lung adenocarcinomas. Cancer Epidemiol Biomarkers Prev 1997;6:841-7. [Abstract/Free Full Text]
    
14. Cuttitta F, Carney DN, Mulshine J, Moody TW, Fedorko J, Fischler A, et al. Bombesin-like peptides can function as autocrine growth factors in human small-cell lung cancer. Nature 1985;316:823-6. [Medline]
    
15. Halmos G, Schally AV. Reduction in receptors for bombesin and epidermal growth factor in xenografts of human small-cell lung cancer after treatment with bombesin antagonist RC-3095. Proc Natl Acad Sci U S A 1997;94:956-60. [Abstract/Free Full Text]
    
16. Kelley MJ, Linnoila RI, Avis IL, Georgiadis MS, Cuttitta F, Mulshine JL, et al. Antitumor activity of a monoclonal antibody directed against gastrin-releasing peptide in patients with small cell lung cancer. Chest 1997;112:256-61. [Abstract/Free Full Text]
    
17. Weiner DB, Nordberg J, Robinson R, Nowell PC, Gazdar A, Greene MI, et al. Expression of the neu gene-encoded protein (P185^neu) in human non-small cell carcinomas of the lung. Cancer Res 1990;50:421-5. [Abstract/Free Full Text]
    
18. Rachwal WJ, Bongiorno PF, Orringer MB, Whyte RI, Ethier SP, Beer DG. Expression and activation of erbB-2 and epidermal growth factor receptor in lung adenocarcinomas. Br J Cancer 1995;72:56-64. [Medline]
    
19. Noguchi M, Murakami M, Bennett W, Lupu R, Hui F Jr, Harris CC, et al. Biological consequences of overexpression of a transfected c-erb B-2 gene in immortalized human bronchial epithelial cells. Cancer Res 1993;53:2035-43. [Abstract/Free Full Text]
    
20. Kern JA, Torney L, Weiner D, Gazdar A, Shepard HM, Fendly B. Inhibition of human lung cancer cell line growth by an anti-p185HER2 antibody. Am J Respir Cell Mol Biol 1993;9:448-54.
    
21. Tsai C-M, Chang K-T, Wu L-H, Chen J-Y, Gazdar AF, Mitsudomi T, et al. Correlations between intrinsic chemoresistance and HER-2/neu gene expression, p53 gene mutations, and cell proliferation characteristics in non-small cell lung cancer cell lines. Cancer Res 1996;56:206-9. [Abstract/Free Full Text]
    
22. Yu D, Wang S-S, Dulski KM, Tsai C-M, Nicolson GL, Hung M-C. c-erb B-2/neu overexpression enhances metastatic potential of human lung cancer cells by induction of metastasis-associated properties. Cancer Res 1994;54:3260-6. [Abstract/Free Full Text]
    
23. Kern JA, Slebos RJC, Top B, Rodenhuis S, Lager D, Robinson RA, et al. C-erb B-2 expression and codon 12 K-ras mutations both predict shortened survival for patients with pulmonary adenocarcinomas. J Clin Invest 1994;93:516-20.
    
24. Pfeiffer P, Clausen PP, Andersen K, Rose C. Lack of prognostic significance of epidermal growth factor receptor and the oncoprotein p 185HER-2 in patients with systemically untreated non-small-cell lung cancer: an immunohistochemical study on cryosections. Br J Cancer 1996;74:86-91. [Medline]
    
25. Rusch V, Baselga J, Cordon-Cardo C, Orazem J, Zaman M, Hoda S, et al. Differential expression of the epidermal growth factor receptor and its ligands in primary non-small cell lung cancers and adjacent benign lung. Cancer Res 1993;53:2379-85. [Abstract/Free Full Text]
    
26. Tateishi M, Ishida T, Mitsudomi T, Kaneko S, Sugimachi K. Immunohistochemical evidence of autocrine growth factors in adenocarcinoma of the human lung. Cancer Res 1990;50:7077-80. [Abstract/Free Full Text]
    
27. Damstrup L, Rygaard K, Spang-Thomsen M, Poulsen HS. Expression of the epidermal growth factor receptor in human small cell lung cancer cell lines. Cancer Res 1992;52:3089-93. [Abstract/Free Full Text]
    
28. Ohmichi H, Koshimizu U, Matsumoto K, Nakamura T. Hepatocyte growth factor (HGF) acts as a mesenchyme-derived morphogenic factor during fetal lung development. Development 1998;125:1315-24. [Abstract]
    
29. Yanagita K, Matsumoto K, Sekiguchi K, Ishibashi H, Niho Y, Nakamura T. Hepatocyte growth factor may act as a pulmotrophic factor on lung regeneration after acute lung injury. J Biol Chem 1993;268:21212-7. [Abstract/Free Full Text]
    
30. Harvey P, Warn A, Newman P, Perry LJ, Ball RY, Warn RM. Immunoreactivity for hepatocyte growth factor/scatter factor and its receptor, met, in human lung carcinomas and malignant mesotheliomas. J Pathol 1996;180:389-94. [Medline]
    
31. Olivero M, Rizzo M, Madeddu R, Casadio C, Pennacchietti S, Nicotra MR, et al. Overexpression and activation of hepatocyte growth factor/scatter factor in human non-small-cell lung carcinomas. Br J Cancer 1996;74:1862-8. [Medline]
    
32. Siegfried JM, Weissfeld LA, Singh-Kaw P, Weyant RJ, Testa JR, Landreneau RJ. Association of immunoreactive hepatocyte growth factor with poor survival in resectable non-small cell lung cancer. Cancer Res 1997;57:433-9. [Abstract/Free Full Text]
    
33. Sekido Y, Takahashi T, Ueda R, Takahashi M, Suzuki H, Nishida K, et al. Recombinant human stem cell factor mediates chemotaxis of small-cell lung cancer cell lines aberrantly expressing the c-kit protooncogene. Cancer Res 1993;53:1709-14. [Abstract/Free Full Text]
    
34. Krystal GW, Hines SJ, Organ CP. Autocrine growth of small cell lung cancer mediated by coexpression of c-kit and stem cell factor. Cancer Res 1996;56:370-6. [Abstract/Free Full Text]
    
35. Quinn KA, Treston AM, Unsworth EJ, Miller MJ, Vos M, Grimley C, et al. Insulin-like growth factor expression in human cancer cell lines. J Biol Chem 1996;271:11477-83. [Abstract/Free Full Text]
    
36. Antoniades HN, Galanopoulos T, Neville-Golden J, O’Hara CJ. Malignant epithelial cells in primary human lung carcinomas coexpress in vivo platelet-derived growth factor (PDGF) and PDGF receptor mRNAs and their protein products. Proc Natl Acad Sci U S A 1992;89:3942-6. [Abstract/Free Full Text]
    
37. Maneckjee R, Minna JD. Opioid and nicotine receptors affect growth regulation of human lung cancer cell lines. Proc Natl Acad Sci U S A 1990;87:3294-8. [Abstract/Free Full Text]
    
38. Maneckjee R, Minna JD. Opioids induce while nicotine suppresses apoptosis in human lung cancer cells. Cell Growth Differ 1994;5:1033-40. [Abstract]
    
39. Cooper CA, Carey FA, Bubb VJ, Lamb D, Kerr KM, Wyllie AH. The pattern of K-ras mutation in pulmonary adenocarcinoma defines a new pathway of tumour development in the human lung. J Pathol 1997;181:401-4. [Medline]
    
40. Tsuchiya E, Furuta R, Wada N, Nakagawa K, Ishikawa Y, Kawabuchi B, et al. High K-ras mutation rates in goblet-cell-type adenocarcinomas of the lungs. J Cancer Res Clin Oncol 1995;121:577-81. [Medline]
    <