The White Lesion That Kills — Aneuploid Dysplastic Oral Leukoplakia
http://www.100md.com
《新英格兰医药杂志》
Perhaps the earliest link between oral leukoplakia and cancer was made by James Paget, for whom Paget's disease was named; he also recognized the connection between oral leukoplakia and smoking. Leukoplakia is a clinical term that refers to an oral mucosal white patch that will not rub off and is not attributable to any other known disease. It is considered to be potentially malignant, with a transformation rate in various studies and locations ranging from 0.6 to 18 percent. Clinically, oral leukoplakia is in the same spectrum of disease as the more sinister red or speckled lesion erythroplakia, which has a much higher transformation rate and is more often found on biopsy to be squamous-cell carcinoma.
The leukoplakia we are discussing here should be distinguished from oral hairy leukoplakia, the lesion associated with Epstein–Barr virus and seen in immunosuppressed persons, predominantly but not exclusively those with human immunodeficiency virus infection. White patches in the mouth may have other causes, including chronic trauma such as friction and mucocutaneous diseases such as white sponge nevus, lichen planus, or lupus erythematosus. However, the lesion that is the subject of the article by Sudb? et al. in this issue of the Journal (pages 1405–1413) is more commonly associated with tobacco-related habits — notably, smoking and the use of some other forms of tobacco, including moist snuff (smokeless tobacco) and pan (betel nut). Another form of precancer, proliferative verrucous leukoplakia, is predominantly found in those who do not use tobacco, but it has a high rate of transformation to cancer.
Here is where the histopathological changes known as dysplasia enter the picture. On biopsy, most leukoplakias show histologic features of epithelial dysplasia, including basal-cell hyperplasia, bulbous rete ridges, loss of polarity of basal cells, abnormal keratinocyte stratification, an increased nuclear:cytoplasmic ratio, nuclear hyperchromatism and pleomorphism, abnormal keratinization, and abnormally increased numbers of mitoses and suprabasal mitoses (see Figure). Unfortunately, the recognition of these features is subjective, and there is wide variation in the degree of interexaminer agreement and intraexaminer reproducibility in the diagnosis. Yet a microscopical diagnosis of oral epithelial dysplasia is still one of the few indicators of an increased risk of oral cancer that we have, despite the reality that most dysplasias do not progress to cancer.
Figure. Microscopical Appearance of Normal Oral Mucosa (Panel A; Hematoxylin and Eosin, x100) and of Oral Epithelial Dysplasia with Hyperkeratosis, Irregular and Bulbous Rete Processes, and Numerous Atypical Keratinocytes (Panel B; Hematoxylin and Eosin, x40).
Transformation rates, like those associated with a clinical diagnosis of leukoplakia, vary, ranging from 16 to 33 percent. Differences in the transformation rate may reflect differences in criteria for histologic grading, clinical follow-up, and populations of patients. Therefore, a search has long been under way for molecular markers that may indicate an increased likelihood of malignant transformation. The possibilities that have been explored include DNA aneuploidy, loss of heterozygosity as assessed on the basis of microsatellite markers, mutations in the p53 gene and aberrant expression of p53 protein, inappropriate expression of other oncogenes such as cyclin D1, and differentiation markers such as keratins, blood-group antigens, and other cell-surface carbohydrates. Some assays for these markers have shown promise in modeling the risk of cancer, but their use has been hampered by technical difficulties that make their wide-scale application challenging. Others have proved to be of limited predictive value for individual persons, thus necessitating a search for new markers of risk and methods of risk assessment.
As demonstrated by Sudb? et al., aneuploidy is emerging as one of the most promising prognostic indicators for the study of oral cancer. The authors have previously shown that the DNA content of dysplastic oral leukoplakia predicts the development of oral squamous-cell carcinoma. In these studies, nuclear DNA content or ploidy was classified as diploid (normal nuclear content), tetraploid (double the normal amount of DNA and number of chromosomes), or aneuploid (an amount of DNA that was not an exact multiple of the diploid number). DNA was measured by means of optical examination of cell nuclei obtained from thick paraffin sections of biopsy samples. Taking advantage of the superb comprehensive Cancer Registry of Norway to follow for several years the same group of patients with dysplasia on biopsy, the authors previously showed that oral squamous-cell carcinoma was significantly more likely to develop in persons with dysplastic leukoplakias showing aneuploidy than in those with diploidy or tetraploidy. The only other predictor of squamous-cell carcinoma, according to multivariate analysis, was tobacco use.
In the present study, Sudb? et al. show that further oral cancers consistently developed in patients whose dysplastic leukoplakia was characterized by aneuploidy; this group also had poorer survival than the other two groups. Furthermore and most intriguingly, the removal of lesions with histologically clear margins did not lead to an improved outcome, challenging a central dogma that oral dysplasia can be managed surgically. The investigators begin to answer the question that has rarely even been asked — not which oral white lesions become cancer, but rather which oral white lesions are likely to kill. They conclude that it is aneuploidy that best predicts poor survival.
However, several questions remain. For example, it is not clear how practical and reproducible the complex methods that Sudb? et al. have used will be in other settings or in the clinical rather than the research arena. Presumably, other and larger studies will be conducted to confirm or refute the intriguing findings. Why should cancers arising from aneuploid dysplastic oral leukoplakia be more lethal than cancers arising from diploid or tetraploid oral leukoplakia? What is happening in the epithelium that appears to be clinically and histologically normal in these persons? Should we change our views about pathogenesis? Until now, we have associated aneuploidy with invasion or at least carcinoma in situ. We must now recognize that it occurs earlier, during a process that we currently call dysplasia, and its initiation may be a turning point or sentinel event in the transformation from normal to malignant oral epithelium. There is already evidence that aneuploidy is not merely a byproduct of a cell gone wrong but that the process itself most likely contributes to neoplasia. What is the role of the assay for aneuploidy in the assessment of proliferative verrucous leukoplakia or even lichen planus? The observation that the lethality associated with aneuploid dysplastic oral leukoplakia appears not to be ameliorated by excision suggests that the procedure may be of little more than palliative or cosmetic value. Are its effects perhaps analogous to those of retinoids on epithelial dysplasia — an improvement in the clinical appearance that often does not reflect a change in the underlying molecular abnormalities or an improvement of the clinical outcome?
The primary and most perplexing question that the article by Sudb? et al. raises is what treatment we can offer persons with aneuploid dysplastic oral leukoplakia. Is this situation hopeless, or will effective chemopreventive or chemotherapeutic agents be developed that can interfere with the molecular and metabolic events leading to aneuploidy? One promising area may be the use of inhibitors of cyclooxygenase-2, an enzyme that Sudb? et al. have shown in another study1 to be more frequently overexpressed in aneuploid oral dysplasia than in diploid or tetraploid dysplasia. The future assessment of oral leukoplakia may involve the routine assessment of ploidy in persons with oral leukoplakia. It is even possible that this could lead to the screening of apparently clinically normal mucosa in persons or populations thought to be at risk.
Where do we go from here? The study by Sudb? and colleagues raises important questions about our current approaches to the diagnosis and management of dysplastic oral leukoplakia. Although the investigators address current tobacco use in the population they studied, it remains unclear how smoking cessation can modify the natural history of the lesions. Until that question is answered, there seems to be no reason to reduce our current emphasis on the cessation of both smoking and the use of smokeless tobacco. Moreover, tobacco use causes or exacerbates a number of other oral diseases, including periodontal disease, candidiasis, and xerostomia. Histopathological examination to establish the presence or absence of epithelial dysplasia or carcinoma will continue to have an important role in the assessment of oral leukoplakia. However, in the light of this study, the importance attached to the establishment of histologically clear surgical margins in the excision of dysplasia must be reassessed, for the implication is that the genetic aberrations within the epithelium may not be reflected in the clinical or microscopical appearance. What we do know from the work presented here is that the determination of the prognosis for persons with oral leukoplakia will never be quite the same again.
Source Information
From the Departments of Stomatology (D.G., R.C.K.J.) and Pathology (R.C.K.J.), University of California at San Francisco, San Francisco.
References
Sudbo J, Ristimaki A, Sondresen JE, et al. Cyclooxygenase-2 (COX-2) expression in high-risk premalignant oral lesions. Oral Oncol 2003;39:497-505.(Deborah Greenspan, B.D.S.)
The leukoplakia we are discussing here should be distinguished from oral hairy leukoplakia, the lesion associated with Epstein–Barr virus and seen in immunosuppressed persons, predominantly but not exclusively those with human immunodeficiency virus infection. White patches in the mouth may have other causes, including chronic trauma such as friction and mucocutaneous diseases such as white sponge nevus, lichen planus, or lupus erythematosus. However, the lesion that is the subject of the article by Sudb? et al. in this issue of the Journal (pages 1405–1413) is more commonly associated with tobacco-related habits — notably, smoking and the use of some other forms of tobacco, including moist snuff (smokeless tobacco) and pan (betel nut). Another form of precancer, proliferative verrucous leukoplakia, is predominantly found in those who do not use tobacco, but it has a high rate of transformation to cancer.
Here is where the histopathological changes known as dysplasia enter the picture. On biopsy, most leukoplakias show histologic features of epithelial dysplasia, including basal-cell hyperplasia, bulbous rete ridges, loss of polarity of basal cells, abnormal keratinocyte stratification, an increased nuclear:cytoplasmic ratio, nuclear hyperchromatism and pleomorphism, abnormal keratinization, and abnormally increased numbers of mitoses and suprabasal mitoses (see Figure). Unfortunately, the recognition of these features is subjective, and there is wide variation in the degree of interexaminer agreement and intraexaminer reproducibility in the diagnosis. Yet a microscopical diagnosis of oral epithelial dysplasia is still one of the few indicators of an increased risk of oral cancer that we have, despite the reality that most dysplasias do not progress to cancer.
Figure. Microscopical Appearance of Normal Oral Mucosa (Panel A; Hematoxylin and Eosin, x100) and of Oral Epithelial Dysplasia with Hyperkeratosis, Irregular and Bulbous Rete Processes, and Numerous Atypical Keratinocytes (Panel B; Hematoxylin and Eosin, x40).
Transformation rates, like those associated with a clinical diagnosis of leukoplakia, vary, ranging from 16 to 33 percent. Differences in the transformation rate may reflect differences in criteria for histologic grading, clinical follow-up, and populations of patients. Therefore, a search has long been under way for molecular markers that may indicate an increased likelihood of malignant transformation. The possibilities that have been explored include DNA aneuploidy, loss of heterozygosity as assessed on the basis of microsatellite markers, mutations in the p53 gene and aberrant expression of p53 protein, inappropriate expression of other oncogenes such as cyclin D1, and differentiation markers such as keratins, blood-group antigens, and other cell-surface carbohydrates. Some assays for these markers have shown promise in modeling the risk of cancer, but their use has been hampered by technical difficulties that make their wide-scale application challenging. Others have proved to be of limited predictive value for individual persons, thus necessitating a search for new markers of risk and methods of risk assessment.
As demonstrated by Sudb? et al., aneuploidy is emerging as one of the most promising prognostic indicators for the study of oral cancer. The authors have previously shown that the DNA content of dysplastic oral leukoplakia predicts the development of oral squamous-cell carcinoma. In these studies, nuclear DNA content or ploidy was classified as diploid (normal nuclear content), tetraploid (double the normal amount of DNA and number of chromosomes), or aneuploid (an amount of DNA that was not an exact multiple of the diploid number). DNA was measured by means of optical examination of cell nuclei obtained from thick paraffin sections of biopsy samples. Taking advantage of the superb comprehensive Cancer Registry of Norway to follow for several years the same group of patients with dysplasia on biopsy, the authors previously showed that oral squamous-cell carcinoma was significantly more likely to develop in persons with dysplastic leukoplakias showing aneuploidy than in those with diploidy or tetraploidy. The only other predictor of squamous-cell carcinoma, according to multivariate analysis, was tobacco use.
In the present study, Sudb? et al. show that further oral cancers consistently developed in patients whose dysplastic leukoplakia was characterized by aneuploidy; this group also had poorer survival than the other two groups. Furthermore and most intriguingly, the removal of lesions with histologically clear margins did not lead to an improved outcome, challenging a central dogma that oral dysplasia can be managed surgically. The investigators begin to answer the question that has rarely even been asked — not which oral white lesions become cancer, but rather which oral white lesions are likely to kill. They conclude that it is aneuploidy that best predicts poor survival.
However, several questions remain. For example, it is not clear how practical and reproducible the complex methods that Sudb? et al. have used will be in other settings or in the clinical rather than the research arena. Presumably, other and larger studies will be conducted to confirm or refute the intriguing findings. Why should cancers arising from aneuploid dysplastic oral leukoplakia be more lethal than cancers arising from diploid or tetraploid oral leukoplakia? What is happening in the epithelium that appears to be clinically and histologically normal in these persons? Should we change our views about pathogenesis? Until now, we have associated aneuploidy with invasion or at least carcinoma in situ. We must now recognize that it occurs earlier, during a process that we currently call dysplasia, and its initiation may be a turning point or sentinel event in the transformation from normal to malignant oral epithelium. There is already evidence that aneuploidy is not merely a byproduct of a cell gone wrong but that the process itself most likely contributes to neoplasia. What is the role of the assay for aneuploidy in the assessment of proliferative verrucous leukoplakia or even lichen planus? The observation that the lethality associated with aneuploid dysplastic oral leukoplakia appears not to be ameliorated by excision suggests that the procedure may be of little more than palliative or cosmetic value. Are its effects perhaps analogous to those of retinoids on epithelial dysplasia — an improvement in the clinical appearance that often does not reflect a change in the underlying molecular abnormalities or an improvement of the clinical outcome?
The primary and most perplexing question that the article by Sudb? et al. raises is what treatment we can offer persons with aneuploid dysplastic oral leukoplakia. Is this situation hopeless, or will effective chemopreventive or chemotherapeutic agents be developed that can interfere with the molecular and metabolic events leading to aneuploidy? One promising area may be the use of inhibitors of cyclooxygenase-2, an enzyme that Sudb? et al. have shown in another study1 to be more frequently overexpressed in aneuploid oral dysplasia than in diploid or tetraploid dysplasia. The future assessment of oral leukoplakia may involve the routine assessment of ploidy in persons with oral leukoplakia. It is even possible that this could lead to the screening of apparently clinically normal mucosa in persons or populations thought to be at risk.
Where do we go from here? The study by Sudb? and colleagues raises important questions about our current approaches to the diagnosis and management of dysplastic oral leukoplakia. Although the investigators address current tobacco use in the population they studied, it remains unclear how smoking cessation can modify the natural history of the lesions. Until that question is answered, there seems to be no reason to reduce our current emphasis on the cessation of both smoking and the use of smokeless tobacco. Moreover, tobacco use causes or exacerbates a number of other oral diseases, including periodontal disease, candidiasis, and xerostomia. Histopathological examination to establish the presence or absence of epithelial dysplasia or carcinoma will continue to have an important role in the assessment of oral leukoplakia. However, in the light of this study, the importance attached to the establishment of histologically clear surgical margins in the excision of dysplasia must be reassessed, for the implication is that the genetic aberrations within the epithelium may not be reflected in the clinical or microscopical appearance. What we do know from the work presented here is that the determination of the prognosis for persons with oral leukoplakia will never be quite the same again.
Source Information
From the Departments of Stomatology (D.G., R.C.K.J.) and Pathology (R.C.K.J.), University of California at San Francisco, San Francisco.
References
Sudbo J, Ristimaki A, Sondresen JE, et al. Cyclooxygenase-2 (COX-2) expression in high-risk premalignant oral lesions. Oral Oncol 2003;39:497-505.(Deborah Greenspan, B.D.S.)