Abstract

Eosinophils are present in large numbers in some squamous cell carcinomas of the oral cavity. Whilst it is proposed that they have an 'immuno-protective' effect, this remains unproven. The contradictory reports may be due to inconsistencies in eosinophil counting. Eighty-one cases of squamous cell carcinoma (SSC) of oral tongue were examined. Two methods of eosinophil counting were performed. In the first method (classical), the eosinophils were counted per 10 HPF. In the second method (our so-called density method), the highest eosinophil density per surface area was counted for each case. The two methods were correlated. Using the classical method a number of fields in cases ranked low, contained more than 10 eosinophils. Likewise, some moderate cases contained more than 100 eosinophils. There is poor correlation between the classical and density counts. Nevertheless, good correlation between the two methods could be achieved if the boundaries of the classical method are modified. Eosinophils invariably appear in clusters. We feel that an assessment of density may well be better than classical counting, and have more relationship with function.