David H. Harpole, Jr., MD, FACS

Durham, NC

Non-small cell lung cancer (NSCLC) accounts for approximately 180,000 deaths annually in the United States and is the leading cause of cancer deaths in women and men. (1) One of the single most significant determinants for long-term survival is absence of local or distant metastasis at time of presentation. (2,3) The factors which control a primary tumor’s ability to metastasize locally to regional lymph nodes or to distant sites such as the brain or bone are poorly understood. Previous investigators have focused on the histologic characteristics of the primary tumor to include cell type, degree of differentiation and tumor size, all of which are correlated with metastatic propensity, but fail to explain biologic mechanisms of metastasis. It is known that for an epithelial tumor to spread, it must grow locally, overcome the confines of its associated basement membrane and extracellular matrix and subsequently enter lymphatic and vascular conduits which transport neoplastic cells to secondary target sites. (4)

The transformation of a cell to malignancy occurs via a process where genes are deleted or mutated so that their products are present in abnormal quantities or do not function properly. Over the last ten years, many of these molecular biologic pathways have been identified in cancer. The pathways can be placed in distinct groups by method of action: molecular genetic factors, proliferation factors and cellular differentiation factors. (5)

Molecular genetic markers are subdivided into proto-oncogenes and tumor suppresser genes. Proto-oncogenes can aid cellular transformation when only one of two alleles is mutated. Proto-oncogenes are activated either by gene amplification or mutation. If a gene is amplified, many copies are placed in the DNA, thereby intensifying the response. The erbB-2 (Her2/neu) gene is a proto-oncogene which codes for a transmembrane protein (p185). This gene has significant sequence homology with the gene for the epidermal growth factor receptor (erbB-1) and both are receptors which act through a tyrosine kinase signal pathway for growth. During DNA injury and repair in lung cancer, many duplicate copies of erbB-1 or erbB-2 are created, magnifying the strength of the signal growth. The signal is normally short-lived, after the GTPase lyses GTP. However, in adenocarcinoma, a mutation at codon 12 of the k-ras gene, (glycine replaced by valine in p21), inactivates the GTPase, leaving a continued signal for growth.

Tumor suppresser genes need both alleles and their products to be eliminated in some way to transform a cell. The first tumor suppresser gene described in the literature was the retinoblastoma gene (rb). Although rb allelic deletion is only seen in 10% - 15% of small cell lung cancers, greater than 90% have point mutations which encode for an abnormal rb protein. Mutation of deleted rb protein is observed in 20% to 30% of NSCLC. (6) Another tumor suppresser gene is p53, which has been associated with many solid tumors and the Li-Fraumeni syndrome. The p53 protein slows the cyclin cascade for cellular growth. A point mutation in one allele of the p53 gene produces an abnormal protein which binds to the normal p53 protein produced by the other normal allele. The bound p53 protein cannot modulate the cyclin pathway. In addition, most tumor suppresser gene control programmed cell death (apoptosis). The bcl-2 gene (originally observed in the 14;18 chromosomal translocation in B cell lymphoma), signals a senescent cell to die. When the gene is mutated in lung cancer, apoptosis is halted or delayed, leaving a viable colony of non-dividing tumor cells which is resistant to chemotherapy.

Retrospective radiographic determinations of tumor doubling times have demonstrated a decreased survival in cases of lung cancer with shorter doubling intervals. The process is a marker of tumor proliferation. Another method to define the rapidly dividing tumors utilizes DNA flow cytometry. Cellular DNA is measured for ploidy. Normal diploid (euploid) fraction is estimated compared to the amount of tumor with abnormal quantities of DNA (aneuploid). Flow cytometry can also estimate the percentage of cells which are in the cell cycle’s S-phase of DNA replication. Another method utilizes the identification of proliferation-associated proteins in the nucleus of tumor cells. KI-67, is 67 kD non-histone protein that is only expressed in the nucleus of cells which are near mitosis (cell cycles: late G1, S, and M). Immunohistochemical techniques allow an estimate of the percent of tumor nuclei which are actively proliferating as an index of the rate of tumor growth. (6)

Monoclonal antibodies have been developed against a number of cell surface receptors, such as the blood-group antigens. These markers of cellular differentiation can be mutated or eliminated with malignant transformation. Lung tumors which have lost the normal blood-group or HLA cell surface receptors are often more aggressive cancers. In addition, previously absent of juvenile proteins may now be expressed during rumor transformation, i.e., the alpha feto protein in germ cell tumors. The Lewis blood-group antigens (H/Le(y)/Le(b)) are such immature surface antigens that are expressed in lung cancer cells after deletion of normal blood-group antigens. (7)

Two processes define the metastatic potential of a tumor: local invasion (matrix metalloproteinases) and neovascularity (angiogenesis). A tumor must stimulate the production of enzymes to degrade the basement membrane and surrounding extracellular matrix to become an invasive cancer. An example of these enzymes which has been associated with invasion in NSCLC are the elastases and stromelysins. (8) Neovascularity of a tumor allows continued growth and the ability to disseminate tumors into the blood stream. Histologic evidence of blood vessel invasion is a crude marker of the angiogenesis of a tumor. Several autocrine and paracrine growth factors have been identified which are produced by the tumor to signal capillary growth. (9) Two examples are bFGF (basic fibroblast growth factor) and VEGF (vascular endothelial growth factor). Presence of these in tumors is associated with rapidly invasive cancers. (10-11)

Once a tumor has invaded locally, it may be disseminated by lymphatic or the blood vessels. Even though the tumor cells undoubtedly are shed throughout the body, NSCLC has a predictable pattern of metastasis. Animal investigations have verified Paget’s hypothesis, demonstrating that intravascular injection of tumors will not create wide-spread metastases unless the proper "seed and soil" are combined. The location of metastases in approximate order of frequency include bone (50%), brain (25%), liver (20%) and adrenal (20%). Selection of secondary target sites is influenced not only by mechanical consideration, i.e. the means of transport of cells to the secondary target site, but is believed to be controlled by yet undefined chemotactic factors and adhesion molecules. Investigations are underway to identify these factors. The neural cell adhesion molecule (NCAM) which is expressed in all small cell lung cancer and normal brain tissue may be a marker for the formation of CNS of adrenal metastasis from NSCLC. (12) No biologic markers have been identified for bony metastases.