The Molecular Biology of Thoracic Malignancies


As Dr. Blake Cady eloquently stated, tumor biology is the principle determinant of outcome in surgical oncology.

“Biology is King, Selection is Queen, and the technical details of surgical procedures are the Princes and Princesses of the realm who frequently try to overthrow the powerful forces of the King and Queen, usually to no long-term avail, although with some temporary apparent victories.”

Reflections such as this and our clinical experience highlight that our ability to make a significant, lasting impact on the outcomes of most cancer patients rests in our capacity to understand and influence tumor biology.

A revolution in the care of patients with thoracic malignancies is upon us. The clinical application of molecular diagnostic and analytic technologies is growing rapidly, and recent studies have begun to unfold their true potential for thoracic oncology patients. In particular, genetic science has a possible role in guiding therapy (based on both an individual's risk of recurrence and death and a tumor's genetic features) and counseling patients (with regards to risk of recurrence and prognosis). Indeed, as our understanding of tumor biology continues to evolve, effective personalized oncology is becoming a reality. In this chapter, we provide a brief review of our current understanding of the molecular mechanisms underlying oncogenesis and metastases of thoracic malignancies. We also highlight the current role of epigenetics, proteomics, biomarkers, and personalized oncology in the care of patients with thoracic malignancies.

Mechanisms of Oncogenesis and Metastasis

The progression of a cell from normal to overt malignancy is the end result of a complex sequence of genetic events influenced by a number of environmental and biological factors. The precise mechanism of oncogenesis of thoracic malignancies is complex and incompletely understood ( Fig. 45-1 ). However, a few salient points have emerged from the literature.

FIGURE 45-1, Overview of the pathways involved in oncogenesis.

Environmental Exposure

First, most thoracic malignancies are the end result of the effect of an environmental exposure (e.g., tobacco smoke, alcohol use, chronic gastroesophageal reflux, asbestos) on a cell's genome. Such exposures can cause alterations in DNA structure (e.g., pyrimidine dimers and double-strand breaks), which can alter DNA replication and transcription. Environmental exposures can also induce mutations if errors are introduced by DNA repair mechanisms while attempting to correct mistakes in DNA structure. Other carcinogens (e.g., asbestos), however, cause chromosomal loss, gain, and/or rearrangements before causing genetic mutations.

Genetic Mutations

Second, mutated genes contribute to oncogenesis by selectively conferring a growth advantage to cancer cells. Such mutations can be either inherited (germline) mutations or acquired (somatic) mutations. Proto-oncogenes are dominant genes that encode proteins that are responsible for cell-cycle progression, cell differentiation, or regulation of apoptosis (programmed cell death). When proto-oncogenes are mutated (by point mutations, translocations, gene amplifications, gene insertions, or gene deletions), they are called oncogenes, which are either constitutively active or abnormally active (under conditions that the normal wild-type gene is not).

Tumor-suppressor genes are the genetic brakes on aberrant cell proliferation by regulating cell division, DNA repair, and apoptosis. In contrast to oncogenes, tumor-suppressor genes are recessive genes—both copies need to be inactivated for their effect to be lost. One allele is typically lost through loss of heterozygosity, and the other is typically lost by a mutation or epigenetic modification (discussed later). Mutations in tumor-suppressor genes confer a growth advantage by removing a normal checkpoint in cellular proliferation.

Mismatch repair genes are recessive genes involved in repairing mistakes occurring primarily during DNA replication. Because microsatellites (tandem repeats in DNA sequences) are especially vulnerable to DNA replication mistakes, microsatellite instability is a hallmark of defective DNA repair. To appreciate the interaction among proto-oncogenes, tumor suppressor genes, and mismatch repair genes, these genes have also been subclassified as gatekeepers, caretakers, and landscapers.

Gatekeepers are oncogenes and tumor suppressor genes that control cell growth and differentiation. In contrast, caretakers (i.e., tumor suppressor genes and mismatch repair genes) are responsible for maintaining the genomic integrity of the cell. Mutations in caretaker genes promote oncogenesis indirectly by inciting genetic instability, which increases the rate of mutations of other genes (including gatekeepers). If these mutations result in a selective advantage toward cell proliferation (or repression of apoptosis), tumor formation may ensue. Finally, landscapers are not directly involved in cellular growth. Rather, they foster a stromal environment conducive to unregulated cell proliferation. Indeed, although normal stroma creates an environment with innate antitumor properties, inherited or acquired mutations in landscaper genes facilitate tumor growth by removing the natural antagonism of the stroma toward oncogenesis (e.g., by altering cell adhesions, cellular signaling, and altered innate immunity) or by altering the microenvironment to a more oncogenic state (e.g., through matrix metalloproteinases and angiogenesis).

Heterogeneity of Mutations

Third, not all genetic mutations that accrue during tumorigenesis are equipotent—some mutations involved in cancer progression are more significant than others. As oncogenesis progresses, cancer cell proliferation becomes increasingly reliant on these more potent (driver) mutations, a phenomenon known as oncogene addiction . Because of a tumor's reliance on the products of these critical oncogenes for growth and survival, this dependence is a point of weakness (an Achilles heel in the seemingly impenetrable armor of unrestrained genetic activity) and, therefore, a potential therapeutic target.

Driver mutations, which comprise a minority of mutations in cancer, are mutations that have been positively selected during oncogenesis by conferring a growth advantage to the cancer cell. They “drive” cancer cell proliferation. In contrast, passenger mutations, which comprise the majority of mutations in cancer cells, confer no such advantage. They are simply the expected functional consequence of normal clonal and subclonal expansion. In other words, passenger mutations are simply “along for the ride.” To have a positive therapeutic effect on tumor biology, one must sort through the passenger mutations and devise means to target the driver mutations effectively.

Most genetic mutations found in tumors are clonal (they are present in all cells within the tumor). However, subclonal mutations are an important source of intratumoral, intermetastatic, intrametastatic, and interpatient tumoral heterogeneity. This source of heterogeneity has significant treatment implications.

Intratumoral heterogeneity refers to heterogeneity of tumor cells within the same tumor. Each time a tumor cell divides, the progeny cells acquire new mutations that distinguish them from their common progenitor cells. Therefore, in the process of cell replication down the genetic tree of oncogenesis, cells that are generationally divergent will also be more genetically distinct. From a practical standpoint, this form of heterogeneity is usually not clinically significant in patients with localized disease because the primary tumor is often surgically removed. However, this divergence down the genetic tree of oncogenesis spawns the seeds of intermetastatic heterogeneity.

Intermetastatic heterogeneity refers to heterogeneity among spatially distinct metastatic lesions in the same patient. Most patients who die from cancer die from an overwhelming tumor burden in metastatic sites that cannot be removed surgically and that fail systemic chemotherapy or radiation therapy. As such, this form of heterogeneity has significant clinical implications. At one end of the spectrum, if each metastatic site is sufficiently genetically distinct to respond differently to chemotherapeutic agents, then long-term survival is nearly impossible to achieve. Fortunately, most intermetastatic heterogeneity is confined to passenger mutations. The mutations required for metastasis are present in the progenitor cells before the seeds of metastasis are released.

Similar to the heterogeneity within the primary tumor that develops during the process of tumor cell propagation (intratumoral heterogeneity), intrametastatic heterogeneity is the end result of genetic mutations that occur as a founding cell of a metastatic site divides. Such mutations are the grounds for drug resistance. Metastatic deposits may initially respond to treatment that targets mutations that occurred in progenitor cells before metastases occurred, because many of the cells in a metastatic lesion have the same susceptibility to chemotherapy, evidenced by a decrease in the size (or complete resolution) of the lesion noted on posttreatment medical imaging. However, hundreds or thousands of the cells with acquired drug resistance are likely still present, perhaps below the resolution of medical imaging or clinical examination. Recurrence in such situations is inevitable.

There is also heterogeneity of the same tumor type between different patients (interpatient heterogeneity) . This form of heterogeneity is responsible for the clinical observation that “no two cancer patients are the same.” As a result, uniformly applying the same systemic treatment to all patients with the same type of tumor does not result in the same response rates for each patient. Even if the end result is the same in two different patients (e.g., a particular driver mutation that leads to lung cancer), the mutations in each patient are likely sufficiently distinct such that they alter the coding sequence of the gene and hence the translated protein product of the gene in unique ways. Even small differences in a protein product can affect the conformation of the protein and hence its susceptibility to treatment. This heterogeneity is likely the result of host factors (e.g., inherited mutations that predispose to developing cancer, mutations that affect the pharmacokinetics of chemotherapeutic agents, environmental factors within a host that facilitate tumor propagation) and distinct mutations between different tumors that develop during clonal and subclonal expansion.

DNA Aneuploidy

One of the most common observations in solid tumors is DNA aneuploidy (abnormal chromosome number and content). Aneuploidy can result from aberrant mitotic divisions, chromosome cohesion defects, improper attachments of chromosomes to microtubules, weakened mitotic checkpoint signaling, or mutated mitotic checkpoint genes. The contribution of aneuploidy to oncogenesis is controversial. Aneuploidy may simply facilitate (rather than directly cause) oncogenesis. For instance, populations of cells that are capable of redistributing chromosomes such that there is an increased probability of loss of heterozygosity (and hence loss of function of tumor-suppressor genes) or duplication of a mutated oncogene allele would confer a growth advantage. Alternatively, aneuploidy (generated either spontaneously or by a carcinogen) could catalyze a chain reaction of genetic instability and therefore initiate carcinogenesis. Regardless of its precise role in oncogenesis, aneuploidy is clinically significant, because it may be a mechanism of acquired resistance to chemotherapy.

Dysregulation of Cellular Proliferation

Mutations in proto-oncogenes and tumor-suppressor genes promote uncontrolled cellular growth if they lead to increased cellular proliferation, decreased apoptosis, or both. Cell proliferation is directly controlled by cyclins, a group of proteins that interact with cyclin-dependent kinases (CDKs) and are controlled by cyclin-dependent kinase inhibitors (CDKIs). Increased expression of cyclins and decreased activity (or loss) of CDKIs contribute to cell cycle dysregulation.

Cell proliferation is also controlled by telomeres, noncoding, tandemly repeated DNA sequences at the ends of chromosomes that allow DNA to be completely replicated all the way to the end of the coding sequence, thereby preventing enzymatic degradation and stabilizing the chromosome. A small amount of telemetric DNA is lost with cell cycle. Once this DNA reaches a critical point, a signal prevents further cell division. Telomerase is a ribonucleoprotein enzyme complex (containing a reverse transcriptase subunit and an RNA subunit) that maintains telomere length and hence evades replicative senescence. In contrast to most somatic cells (which have no telomerase activity), many cancer cells have increased telomerase activity. Consequently, telomerase activity may be a marker of progression for a benign to malignant process and may be a therapeutic target.

There are multiple interconnected complex pathways of cell-cycle control. Noncoding ribonucleic acids (ncRNA) are functional RNA molecules that are not translated into proteins and are important mediators of these pathways. One such ncRNA is microRNA (miRNA), single-stranded, short (about 22-nucleotide) ncRNAs that are important regulators of mRNA translation and degradation by binding at partially complementary sites in the untranslated regions of target mRNAs. miRNAs contribute to tumorigenesis by directly and indirectly (through epigenetic regulation) influencing translation of tumor-suppressor genes and oncogenes and by triggering a toll-like receptor–mediated tumorigenesis inflammatory response. miRNAs are in turn regulated by endogenous RNAs, which share miRNA response elements with mRNA, thereby vying for miRNA binding. As such, endogenous RNAs can remove the repressive effect of miRNAs on gene expression. The result is a complex network of classic genetic and epigenetic regulation of gene expression by miRNAs and endogenous RNAs. Dysfunctional endogenous RNAs can promote oncogenesis by removing the inhibitory effect of miRNA on oncogenes.

Cancer Stem Cells

Most cancers arise from a single cell that has undergone genetic mutations. Such cells, known as cancer stem cells, have the ability to self-renew, proliferate, and produce the heterogeneous lineages of cancer cells within a tumor. It is unclear whether cancer stem cells are true pluripotent stem cells, adult tissue-specific stem cells that acquire oncogenic mutations, or mature cells influenced by their microenvironment to undergo de-differentiation. The process of tumor progression from cancer stem cells is also unclear. It may involve aberrant differentiation of cancer stems cells or it may involve cell-cell fusion between cancer stem cells and differentiated heterogeneous somatic cells, which can then undergo further cell division into various permutations of phenotypically unique cancer stem cells, mutated differentiated cells, or fused cells. Both pathways account for cellular heterogeneity and aneuploidy noted in tumors.

There is growing evidence of lung cancer stem cells. Normal stem cells and lung cancer cells share several proliferative cell signaling pathways (e.g., K-ras, hedgehog, phosphate and tensin homolog [PTEN], Akt, and phosphoinositide 3-kinase [PI3K]). Furthermore, lung cancers often share cell-surface markers with lung stem cells, which may explain the distribution of lung cancer histologies along the tracheal-bronchiole-alveoli axis. For instance, squamous cell carcinomas, which tend to occur in the central airways, and stem cells of the pseudostratified epithelium of the central airways both contain keratin-expressing basal cells.

There is also growing evidence of esophageal cancer stem cells. Several studies on murine models demonstrated evidence of phenotypic stem-cell surface markers (e.g., α-6 integrin, p75ntr, and CD44 ). Based on evidence found in murine models, pluripotent bone marrow progenitor cells contribute to Barrett esophagus. Given the relative low turnover of lung and esophageal epithelial tissue compared with hematopoietic cells and keratinocytes, stem cells research in lung and esophageal basic science research has been somewhat limited.

The cancer stem cell theory has important therapeutic implications. The success of existing systemic chemotherapy relies on a stochastic model in which all cells in a tumor have the same oncogenic potential. However, if there is only a small subset of cells with proliferative potential (i.e., cancer stem cells), treatment failure is likely. In fact, non–small-cell lung cancer (NSCLC) patients with CD133 + lung cancer stem cells are less responsive to platinum treatment. Potential mechanisms for resistance of cancer stem cells to chemoradiation therapy include their relative dormant nature compared with nontumorigenic cancer cells (which tend to replicate more rapidly), ability to survive in hypoxic microenvironments, upregulation of chemotherapy efflux mechanisms, and upregulation of DNA repair pathways.

Epigenetics

In basic genetics, the expression of a gene (the phenotype ) is directly associated with the DNA sequence of the gene (the genotype ). However, this seemingly simple association between phenotype and genotype is often more complex. There are other heritable changes in gene expression not accompanied by alterations in the DNA sequence. These processes, known as epigenetics, provide insight into the complex link between the genotype and expressed phenotype, which may differ from the phenotype anticipated by basic genetics. Epigenetic changes may explain why cells in the human body share the same genome yet have vast (yet stable) differences in gene expression.

The most studied area of epigenetics is DNA methylation. DNA methylation occurs in normal cells, predominantly on cytosines, which are adjacent to guanines and separated by only a phosphodiester bond. This dinucleotide sequence ( CpG, shorthand for 5′ cytosine-phosphate-guanine) is distinct from cytosine-guanine DNA base pairing. High-density CpG regions (known as CpG islands ) span the promoter region of many (approximately 40%) genes and are often constitutively unmethylated in normal tissues. When these islands are methylated by DNA methyltransferases, transcription is repressed. During the process of tumorigenesis and the resultant accumulation of genetic defects, there is a progressive loss of total genomic methylation, an increased frequency of hypermethylation at CpG islands, and an increase in histone modification.

One of the epigenetic hallmarks of cancer cells is global genomic hypomethylation compared with the methylation content in normal cells. Hypomethylation occurs at highly repeated, interspersed DNA sequences (which results in transcriptional interference of neighboring genes and cancer-associated gene insertions) ; centromeric satellite DNA ; and single-copy transcriptional control sequences. The contribution of DNA hypomethylation to tumorigenesis is independent of effects of DNA hypermethylation at CpG islands. It is neither a prelude nor a consequence of DNA hypermethylation.

The mechanistic basis for the contribution of hypomethylation to tumorigenesis is poorly understood but may include genomic instability (thereby facilitating genetic recombination) and dysfunctional transcriptional control elements (thereby increasing oncogene expression).

DNA hypermethylation in tumorigenesis usually affects the CpG islands at the promoter regions of tumor suppressor genes (upwards of 80% of patients with lung cancer) and transcription factors, which can upregulate expression of oncogenes or downregulate expression of tumor suppressor genes. Although it is less well understood, hypermethylation may also affect non–CpG island promoters and gene target sites. In general, promoters with hypermethylated CpG islands cannot initiate transcription unless it is overridden by alternative signals (e.g., demethylation or chromatin modulation). Consequently, epigenetic silencing of tumor-suppressor genes leads to upregulation of cell survival and proliferation pathways. In contrast to genetic mutations, which often affect proliferation pathways at a single foci, multiple epigenetic events can affect a single signaling pathway and in turn affect other pathways, thereby promoting tumorigenesis in a more integrated manner.

Histones, the major protein component of chromatin, also undergo epigenetic modification, predominately through changes in lysine acetylation (via histone acetyl transferases and deacetylases), arginine and lysine methylation (via histone methyltransferases), and serine phosphorylation (via histone kinases). Such covalent modifications can alter the physical properties of chromatin, thereby influencing its interactions with catalytic enzymes and hence its control of gene activity. There are also noncovalent modifications that contribute to chromatin variation, including chromatin-remodeling complexes, variations in histone proteins, and nucleosome (segments of DNA wound around histone proteins) remodeling. Epigenetic chromatin modifiers such as the polycomb complexes of proteins, which are important for long-term transcriptional repression, may link epigenetic gene silencing with cancer stem cells.

DNA hypomethylation, DNA hypermethylation, histone modification, and likely undiscovered epigenetic pathways are not mutually exclusive. There is significant functional crosstalk among these pathways, producing an intricate network of signals that influence gene expression. Epigenetic changes in cancer may serve as biomarkers and potential therapeutic targets.

Proteomics

As a result of posttranslational modifications, genomic studies of DNA and RNA often poorly correlate with protein expression and hence the phenotype of the cell or tissue. The proteome is the protein complement of the genome. Hence, proteomics is the study of the structure and function of proteins expressed by the genome of a biological system. Because of posttranslational modifications and the resultant heterogeneous nature of proteins, human proteomics is more complex than human genomics.

Most proteomic approaches involve homogenization of a particular sample of interest followed by separation of the proteins in the sample of two-dimensional polyacrylamide gel electrophoresis (2D-PAGE). The samples are first separated by their isoelectric point and then by their molecular weight (hence, “2D” PAGE). The proteins on these gels are displayed as spots, which can be identified by Western blot analysis and comparing the results with protein databases or by digesting the gel spots with trypsin and then performing mass spectrometry on the sample. Alternative techniques to 2D-PAGE include high-performance liquid chromatography and gel-free isotope-labeling techniques to separate peptides from a sample.

Regardless of the separation technique, mass spectroscopy is the primary technique used in proteomics to identify proteins and peptides in a sample. Protein identification is based on measurement of their molecular weight and charge (m/z ratio). To obtain these measurements, samples are ionized (i.e., by matrix-assisted laser desorption ionization or by electrospray ionization), travel through a mass analyzer that separates ions based on their m/z ratio (e.g., using time-of-flight, Fourier transformation, or ion traps), and then are processed by a mass detector. A second mass spectrometer can be added to this sequence (known as tandem mass spectroscopy ) to identify the amino-acid sequence of a peptide. The resultant data are compared against protein databases to identify the protein in the sample. As an alternative to performing proteomic analysis in two steps, multidimensional protein identification technology uses high-performance liquid chromatography in-line with tandem mass spectrometry.

Understanding the proteome will provide tremendous insight into the downstream cellular response to perturbations in the genome of thoracic malignancies.

Metabolomics

Metabolomics is the quantitative description of all endogenous metabolites in a biological sample (e.g., tissue, blood, urine) simultaneously. It allows for a global assessment of changes in phenotype in the context of the immediate environment of the cell by reflecting changes in enzymatic gene regulation and altered kinetics of metabolic enzymes. Metabolites include those involved in cellular respiration (e.g., metabolic intermediates of glycolysis, anaerobic metabolism, the citric acid cycle), amino acid metabolism (i.e., alterations in the levels of essential and nonessential fatty acids and their metabolites), fatty acid metabolism (i.e., the relative abundance of unsaturated and saturated fatty acids), and nucleotide metabolism. Most metabolomic platforms are based on spectroscopic techniques (e.g., nuclear magnetic resonance and mass spectroscopy). Multivariate metabolomic analyses are performed in three major steps. First, pattern recognition (also known as group clustering ) is used to decipher differences in spectral patterns between two groups (e.g., cancer and normal specimens). Next, the spectral region of interest identified in the first step of the analysis is linked to a specific metabolite based on its nuclear magnetic resonance chemical shift. The final step involves ascertaining associations between the metabolite of interest and a particular clinical outcome. Although studies on cancer metabolomics are still in the preclinical exploratory phase, metabolomics have a potential role in cancer diagnosis and assessment of response to therapy by providing relatively inexpensive, rapid, and automated methods to assay biomarkers in biological samples.

Clinical Application of Nucleic Acid and Protein Biomarkers

A biomarker is a surrogate indicator of a biological process that can be measured qualitatively or quantitatively. DNA biomarkers include mutations (in oncogenes, tumor-suppressor genes, and mismatch repair genes), single nucleotide polymorphisms, DNA copy number changes, chromosomal translocations, microsatellite instability, and epigenetic changes (e.g., differential promoter methylation). RNA biomarkers include messenger RNA (mRNA) and miRNA. Protein biomarkers include proteins that are expressed on cell surfaces or shed into serum. For patients with thoracic malignancies, nucleic acid and protein biomarkers may have diagnostic, prognostic, and predictive roles in patient care and may serve as therapeutic targets.

In general, detection of nucleic acid and protein biomarkers in tumor tissue primarily has prognostic, predictive, and therapeutic implications. In contrast, detection of these biomarkers in other tissues and body fluids primarily has diagnostic implications for cancer screening and for detection of occult malignancy.

Detection in Tumor Tissue

Lung Cancer

p53

p53 is a tumor-suppressor gene that plays a central role in cell-cycle regulation, apoptosis, and DNA repair. As such, inactivation of (or mutations in) p53 lead to derangements in cell-cycle progression, evasion of apoptosis, and genomic instability due to impaired DNA repair mechanisms. Given its vital role in controlling cellular proliferation, it is not surprising that p53 mutations are among the most common mutations in all human malignancies. As much as 50% of NSCLCs and 90% of small-cell lung cancers contain p53 mutations or deletions. A systematic review of 56 studies found that p53 mutations in NSCLC are associated with worse overall survival for all stages and for both adenocarcinoma and squamous cell carcinoma histologies. In that study, there were insufficient data to determine the prognostic significance of p53 mutations in small-cell lung cancer.

ERCC1

Platinum-based chemotherapy is the gold standard for neoadjuvant and adjuvant chemotherapy for patients with resectable NSCLC and for treating patients with unresectable NSCLC. The mechanism of action of platinum chemotherapeutic agents (i.e., cisplatin, carboplatin, and oxaliplatin) is formation of DNA adducts. These adducts disrupt DNA replication.

Excision repair cross-complementing group 1 (ERCC1) is part of a DNA repair pathway that recognizes and removes platinum-induced DNA adducts, therefore conferring resistance to platinum-based chemotherapy. ERCC1 expression is common (present in 44% of tumors in one study ) and is an important prognostic factor (independent of treatment). Low ERCC1 levels are associated with a worse prognosis and identify patients who may benefit from platinum-based chemotherapy. In contrast, high ERCC1 levels are associated with a better prognosis and identify patients who are less likely to benefit from platinum-based chemotherapy. One possible explanation is that high levels of ERCC1 indicate intact DNA repair pathways and hence greater genomic stability. In one study of 51 NSCLC patients who underwent resection for cure, high-level ERCC1 expression was associated with an absolute improvement in median overall survival of 60 months.

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here