Incidental Histopathologic Patterns: Possible Evidence of `Field Cancerization' Surrounding Skin Tumors

To the Editor: The `field cancerization' theory of carcinogenesis states that all the epithelium of the upper aerodigestive tract has been mutagenized, presumably as a result to exposure to carcinogens such as tobacco smoke or nitrates, and therefore is at increased risk for multiple epithelial tumors (1,2). We propose that this theory applies equally well to the epidermis of people affected by skin cancers as they are at high risk for second primary skin cancers (3–9) and have had (excessive) exposure to mutagens such as UV light and human papillomavirus (HPV) (10,11). Confirmation of widespread genetic damage to sun-damaged skin is evidenced at the histologic level by the frequent presence of precursor lesions such as solar (actinic) keratoses associated with squamous cell carcinoma (12) or melanocytic (dysplastic) nevi associated with malignant melanoma (13). However, the epidermis surrounding most skin cancers does not exhibit histologic evidence of carcinogenic events. Epidermolytic hyperkeratosis (EHK) (Fig. 1), focal acantholytic dyskeratosis (FAD) (Fig. 2) and pronounced acantholysis as found in Hailey-Hailey disease (HH) (Fig. 3) are histopathologic patterns that signal the presence of a specific autosomal dominant genodermatosis or loss of heterozygosity. The effect if these genetic alterations originate at conception or at an early stage of embryogenesis is to produce a generalized or segmental/systematized linear cutaneous disorder, respectively (14–20). Moreover, each of these histopathologic patterns is reflective of mutations in specific genes: EHK-genes for keratin 1 or 10 (16), FAD-ATP2A2 gene, encoding a Ca2+ pump (21) and HH-ATP2C1 gene, also encoding a calcium pump (18). In addition to generalized and systematized presentations, EHK, FAD, and HH can be seen as incidental histologic findings in biopsies taken for other diagnostic or therapeutic purposes (22–25). These incidental findings may well represent random, acquired mutations of the above genes. Alternatively, they could be secondary to epigenetic factors (e.g., cytokine and or adhesion molecule dysregulation). Underscoring the possibility of acquired mutations is the report of FAD secondary to carcinogen exposure in the oral mucosa of rats (26).FIG. 1.: Epidermolytic hyperkeratosis (EHK) is characterized by granular and vacuolated degeneration of keratinocytes in the spinous and granular layers and overlying compact hyperkeratosis. Mutations in genes for keratin 1 or 10 (16) have been described for this pathologic pattern.FIG. 2.: Focal acantholytic dyskeratosis (FAD) is characterized by suprabasalar clefting with overlying acantholytic and necrotic (dyskeratotic) keratinocytes through out all levels of the epidermis. Mutations in the ATP2A2 gene, encoding a Ca2+ pump (21) have been described for this pathologic pattern.FIG. 3.: Acantholysis as found in Hailey-Hailey disease (HH) is characterized by suprabasalar vesicle formation and acantholytic cells involving at least the lower half of the epidermis. This incidental histologic finding has more acantholysis than is typical of pemphigus and none of the basal layer atypia and overlying parakeratosis found in an acantholytic solar (actinic) keratosis. Mutation in the ATP2C1 gene that encodes a calcium pump (18) has been described for this pathologic pattern.To seek evidence supportive of the hypothesis that incidental histopathologic patterns are markers of genetic instability, therefore field cancerization, we examined sequential skin biopsies and excisions for the prevalence of the previous three incidental histopathologic patterns. If cancer is the result of an accumulation of genetic defects, evidence of this genetic damage should be more frequently found in older individuals and in regions affected by cancer. In this study, we scrutinized 1,606 sequential skin specimens from the Albany Medical College Dermatopathology Service for the presence of incidental EHK, FAD, and HH and documented the age of the patient, site, and associated skin disorder. Only routinely processed histologic material was examined (one hematoxylin and eosin stained slide per tissue block with 1 to 4 profiles of tissue per slide). Applying the chi-squared test, we determined whether these incidental histopathologic patterns occurred significantly more frequently in older individuals, in association with malignant skin tumors, or in sun-exposed skin than would be expected by chance. See Table 1.TABLE 1: Frequency of incidental histologic findings in 1606 consecutive skin specimensIncidental histologic reaction patterns were relatively uncommon, affecting 2.6% of all skin specimens in this survey, and were found in the adjacent nonlesional skin, not within the epidermis affected by the primary lesion except for the cases of scar and keloid. The presence of these focal histopathologic patterns was significantly more common surrounding malignant skin cancers than benign skin tumors, in benign skin tumors than inflammatory or fibrotic conditions and in older than younger individuals. No differences in frequency of these patterns among sun-exposed, partially sun-exposed and non-sun–exposed skin were observed. Sun-exposed skin (head and neck, back, chest, arms, and lower legs) was the most frequent source of pathologic material. If incidental EHK, FAD, and HH represent clonal populations of mutated keratinocytes, their higher frequency in older individuals and surrounding skin cancers could be a reflection of genetic instability, thus a histologic marker of field cancerization. Supporting the possibility that these incidental histopathologic patterns represent clonal expansions of mutated keratinocytes is a report of multiple, dissimilar p53 gene mutations identified in clonally expanded, non-neoplastic keratinocytes surrounding skin cancers (27). Of note, among the benign skin tumors, dysplastic melanocytic nevi were most frequently associated with these incidental histopathologic patterns—a finding corroborated by a prior study that implicated EHK as a marker for melanocytic atypia (25). Although it is disputed whether dysplastic melanocytic nevi represent precursor lesions to malignant melanoma, clinically, these melanocytic nevi have been found to be significant markers of risk for malignant melanoma (28). The high percentage (2.4%) of incidental histopathologic patterns found in 20 to 40 year-olds is likely a reflection of the higher frequency of dysplastic melanocytic nevi in this age group (29–32). In conclusion, the incidental histopathologic patterns of EHK, FAD, and HH may not be coincidental at all, but a consequence of clonal expansion of mutated keratinocytes because of long-term exposure to mutagens such as UV light and HPV. Their frequent presence in the skin around neoplastic skin tumors could be the reflection of a field cancerization, i.e. multiple genetic defects in tissue most of which do not result in cancer. The development of cancers in these regions would be expected to be higher than in areas without these reaction patterns because of the greater likelihood of a second mutation leading to malignancy in the mutated population (33,34). Confirmation of this supposition awaits molecular analysis of microdissected keratinocytes exhibiting these histopathologic patterns. If these keratinocytes do indeed contain base-pair mutations associated with these histopathologic patterns (16,18,21), these changes likely reflect collateral genetic damage and are not a direct risk for cancer as the generalized and systematized (congenital) forms of these disorders are rarely reported with cancer (35–37). J. Andrew Carlson, M.D. Diane Scott, M.D. Jacqueline Wharton, M.D. Stewart Sell, M.D.