Cytologic, histologic and clinical correlations of precancerous and cancerous oral lesions in 57,518 industrial workers of Gujarat, India

Abstract

A group of 57,518 industrial workers of Gujarat, India were screened for oral cancerous and precancerous oral lesions between 1967 and 1971. All subjects were 35 years of age or older and 95 per cent were males. The 27,841 oral lesions found were examined by cytologic scrapings and 13,230 were biopsied. Also, all lesions were followed clinically to assure a malignancy was not overlooked. There were 51 oral cancers diagnosed (.18% of the lesions and .09% of the entire study group). Computing all the cytologic smears, there was an overall accuracy exceeding 99 per cent. When assessing just the malignancies, the accuracy decreased to 80.4 per cent. Since there was a high degree of clinical suspicion on the part of the screeners, only one unsuspected cancer was discovered by cytology. There were four false positive interpretations; and 53 other specimens classified as "suspicious" subsequently were shown to be benign. Although exfoliative cytology has proved useful in assessing oral lesions as an adjunct to biopsy, the low frequency of oral cancer limits the value of this technique as a screening modality. Most false negatives have been associated with leukoplakic (hyperkeratotic) lesions. Therefore, in a persistent oral lesion, even though a cytologic scraping may not be suspicious or characteristic of malignancy, a biopsy should still be strongly considered.