Slaughter’s classic 1953 report on oral cancer proposed that UADT carcinogenesis is a process of “field cancerization” - the repeated exposure of a region’s entire tissue area to carcinogenic insult (eg, tobacco and alcohol), which increases the tissue’s risk for developing multiple independent premalignant and malignant foci. This concept (also called field carcinogenesis or condemned-mucosa syndrome) may explain the clinical occurrence of multiple primary and second primary tumors in HNSCC. The data indicate that second primaries, whether synchronous or metachronous, generally are of squamous histology, develop at a constant rate (4% to 7% of treated patients per year), are not treatment related, and occur in the aerodigestive field at risk, that is, the head and neck, the upper two-thirds of the esophagus, and the lung. These characteristics of second primaries support the field cancerization hypothesis. Recent data on p53 mutations provide strong molecular support for the field carcinogenesis concept. Despite epidemiologic studies having long associated tobacco and alcohol use with the development of HNSCC, the molecular targets of these agents remain to be identified. In recent investigations by Brennan and colleagues, significant tobacco and alcohol use was associated with a high frequency of p53 mutations. Their preliminary results suggest these p53 mutations occur at nonendogenous mutation sites. These findings suggest a role for tobacco in the molecular progression and field carcinogenesis process in head and neck cancer.

Despite intensive study, much of the complex fundamental biology of HNSCC remains poorly understood. Like other epithelial neoplasms, UADT carcinogenesis appears to evolve through a complex multistep process involving certain biomolecular changes that precede premalignant lesions, which in turn precede invasive cancer.

On the basis of animal studies, epithelial carcinogenesis has been divided into three phases: initiation, promotion, and progression. Although human neoplasia does not fit neatly into this tripartite framework, the framework serves as a useful model for understanding pharmacologic interventions. Pharmacologic interventions at each phase have attendant advantages and disadvantages. Chemoprevention’s greatest potential is in the promotion and progression phases of carcinogenesis.

Occurring within the three-phase model described above, multiple subtle steps of UADT carcinogenesis involve genetic alterations, dysregulated epithelial differentiation, abnormal proliferation, and altered regulatory effects. Although the earliest genetic changes precede the relatively simultaneous occurrences of altered differentiation and proliferation, evolving genetic changes occur as carcinogenesis progresses. One important focus of combined clinical and basic research is to establish specific probes and markers for these carcinogenic steps. These probes would help identify individuals at highest risk of UADT cancer and would act as intermediate end-point markers for early evaluations of the efficacy of chemopreventive agents.

Genetic Alterations
The degree of genetic damage is a composite of carcinogen exposure and inherent tissue sensitivity. Genomic changes accumulate, presumably in the entire carcinogen-exposed tissue. Clonal malignant foci develop only in specific sites, however, where tumorigenesis is possible. Nonspecific (random) genomic alterations indicated by micronuclei, sister-chromatid exchanges, and aneuploidy can occur in normal, premalignant, and malignant aerodigestive tract tissue. Although the fundamental genetic events associated with UADT cancers have not yet been established, several nonrandom chromosomal alterations (eg, alterations in chromosomes 1, 3, 5, 7, 8, 9, 10, 11, 13, and 17) have been detected in HNSCC (fresh-tissue and cell-line studies). Short-term cultures of oral premalignant cells may reveal early nonrandom cytogenetic changes.

Carcinogenesis is regulated by the balance between oncogenes and tumor-suppressor genes. This dynamic relationship is under intensive study in HNSCC. Aberrant expressions (amplifications or mutations) of specific families of cellular oncogenes such as myc, ras, neu, bcl, int, ems-1, cyclin D1 and hst-1 are associated with aerodigestive tract carcinogenesis. Up to one-third of cases of primary HNSCC demonstrate amplification of either the int-2 or hst-1 gene (both members of the fibroblast growth factor gene family) on chromosome 11q13. Little is known, however, about when these events take place during the multistep process or what role their gene products play in regulating growth and differentiation. The families of ras, neu, and n-myc oncogenes are amplified in the more advanced stages of oral squamous cell carcinoma (SCC), and activated oncogenes of these families in vitro alter the response to differentiation agents and promote uncontrolled cellular growth, aneuploidy, and tumorigenicity.

Differentiation Alterations
Dysregulation of differentiation is another hallmark of multistep carcinogenesis. Human oral and esophageal epithelium is stratified squamous in type. Oral mucosa is noncornified except for the mucosa on the gingiva and dorsal surface of the tongue, which undergo keratinization. Cytokeratins, a family of at least 19 intermediate-sized filaments that range from 40 to 68 kDa, are expressed in different complex patterns that correlate with distinct types of epithelial differentiation and with carcinogenic progression.

The spatial distribution of several cytokeratins, involucrin, transglutaminase I, and other differentiation antigens is also altered in dysplasia and carcinoma.

Proliferation Alterations
The third major phase of multistep carcinogenesis is dysregulated proliferation. The transition from normal epithelium to hyperplasia and dysplasia is associated with an increased growth fraction and cells proliferating beyond the basal layer. Older histologic assays have correlated this process with increased frequencies of mitotic figures; more recently, DNA flow cytometry and monoclonal-antibody probes to nuclear antigens (eg, Ki-67 and proliferating cell nuclear antigen [PCNA]) have revealed some strong positive correlations in UADT epithelium between abnormal suprabasal proliferation and later carcinogenic stages (severe dysplasia).

Altered Regulatory Effects
Epithelial carcinogenesis is also associated with the abnormal expression of cellular factors that regulate growth and development. The importance of these factors is inferred from their differential expression in normal and malignant tissues. For example, high expression of epidermal growth factor receptor (EGFR) is found in a significant fraction of experimental and human SCCs of the UADT. EGFR gene activation occurs early in experimental oral carcinogenesis, and foci of cells expressing EGFR can be found in human premalignant lesions. Similarly, high expression of transforming growth factor-α (TGF-α) has been associated with malignant transformation of a variety of tumors (including oral cancers), and is frequently accompanied by elevated levels of EGFR in SCC. The relationship between high TGF-a and high EGFR suggests that an autocrine loop mechanism drives the dysregulation of proliferation.

Molecular Biology
As advances in molecular biology and biotechnology continue at a rapid pace, our understanding of the initiation and progression of carcinogenic processes will follow. The most frequent molecular abnormalities in these cancers, to date, are mutations in the tumor-suppressor gene p53, which occur at a rate of approximately 40% to 70%. Mutations of p53 are present in premalignant areas, including carcinoma in situ or moderate to severe dysplasia, in approximately 20% of cases. No other oncogene or tumor-suppressor gene has been found to be mutated as frequently as p53 in head and neck tumor specimens. Other genes have been investigated in these cancers, including the tumor-suppressor gene Rb and several oncogenes, such as ras, myc, int-2, bcl-1, cyclin D1 and C-erb/neu; despite these studies, predictors of response to therapy, phenotypic behavior, and survival remain elusive.

Biotechnology and scientific advances have also allowed identification of areas of frequent chromosomal loss, termed loss of heterozygosity (LOH). In head and neck cancer, LOH is frequent among chromosomes 3p, 17p, 13q, 11q, 6p, 9p, and 14q. These studies may lead to the identification of putative tumor-suppressor genes in these malignancies as well as an understanding of the progression to malignancy.

The concepts of multistep carcinogenesis and field carcinogenesis have now been formulated in molecular biologic terms. The carcinogen-containing environment that initiated tumorigenesis has also affected the nontransformed surrounding tissues. Chromosome labeling and LOH studies have shown that LOH at 9p and abnormalities in chromosome 11 are present in histologically normal mucosa adjacent to tumors, supporting the field hypothesis.

Once a head and neck cancer has developed, its phenotypic behavior (the ability to invade, metastasize, respond to radiotherapy, and recur) is dependent on complex microenvironmental and biologic systems. Invasive and metastatic capabilities of a tumor depend on degradative enzymes, including collagenases, plasminogen activators (including urokinase), and cathepsin, as well as angiogenic factors, growth factors, cytokines, receptors, cell surface properties, and motility factors. The tumor milieu is also influenced by nerve fibers, stromal cells, and tumor-associated and tumor-infiltrating lymphocytes. All of these characteristics provide attractive targets for new and more specific therapeutic approaches.

Immunology
A variety of immunologic abnormalities occur in HNSCC, but precise cause-and-effect relationships between abnormalities and cancer remain unclear. Both cellular and humoral immunity have important implications for HNSCC prognosis and therapy.

Cell-mediated immunity may be involved in tumor control. Defective cell-mediated immunity as determined by natural killer (NK) cell activity, skin tests, and other measures correlates with advanced disease, early recurrence, and poor survival. Large variations in tumor-infiltrating lymphocyte (TIL) levels in HNSCC have been reported. TILs in HNSCC are primarily CD31 T cells (NK cells are rare), and 30% to 50% of TILs express human leukocyte antigen (HLA)-DR-activation antigens. The prognostic significance of tumor-infiltrating cell subsets is unclear.

Most recent work in HNSCC immunology has focused on humoral immune status. A large body of growing evidence suggests that humoral immunity plays a negative role in HNSCC. Increased serum levels of immunoglobulin (Ig) A, IgM, and immune complexes are associated with advanced disease and poor prognosis. Increased IgA can block many aspects of the host cell-mediated immune response, including cytotoxic effects of sensitized lymphocytes, NK cells, and macrophages, and IgG-mediated antibody-dependent cell-mediated cytotoxicity (ADCC). In contrast, increased IgE is associated with a good prognosis in HNSCC. The ratio of IgA to IgE, therefore, may be more predictive of prognosis than either IgA or IgE is individually. Posttreatment patterns of IgA and immune complexes differ; immune complex levels more accurately reflect tumor burden.

Although multiple and prominent immunologic deficits occur in HNSCC, immunologic therapeutic approaches (interferon-α [IFN-α], interleukin-2 [IL-2], plasmapheresis) have yielded only modest response rates and short response durations.

Biologic Goals
Several nonrandom chromosomal alterations, activated oncogenes, and LOH sites likely for tumor-suppressor genes are being actively investigated to develop informative biologic and predictive markers for HNSCC. The recent development of antibody probes for key gene products and molecular probes for gene transcripts should allow future study to identify the timing and sequence of gene alterations during multistep UADT carcinogenesis.

The mechanisms of HNSCC invasion and metastasis are not well established. Recent studies have identified specific membrane proteins, degradative enzymes, and binding sites required for SCC invasion and metastasis (in animal models) and associated with early recurrence (in clinical studies). This work opens the way to novel therapeutic approaches for future use of directed monoclonal antibodies/inhibitors to block key binding sites, thus preventing tumor invasion and metastasis.

Advances in molecular biology and biotechnology are also expanding our concept of the capabilities for molecular (or gene) therapy. Novel gene intervention strategies may be applicable for augmenting immune response, delivering a toxic gene or metabolite, alternating chemo or radiation sensitivity, or inducing cell cycle control or cell death. Further studies that demarcate critical molecular events in the HNSCC progression model may be essential for gene therapy prevention strategies.