Because of the logistical and practical problems that make cancer prevention trials using cancer incidence as an endpoint virtually impossible to conduct for the majority of cancer types, there is a desperate need for valid intermediate markers of cancer risk to serve as surrogate endpoints in chemoprevention trials. A long and continually growing list of potential markers has been developed in the recent past. Unfortunately very few, if any, of them have been subjected to the usual quality control requirements for a laboratory test before being applied to clinical settings.
Modulation of micronuclei frequency has been reported in a number of chemoprevention trials involving the oral cavity, esophagus, lung, and lower GI tract; however, we have focused our efforts primarily on applying the assay to exfoliated buccal mucosal cells, since much of the published data deal with this site, and oral cancer prevention is the theme of one of our chemoprevention trials. After standardizing the definition of a micronucleus by literature review and direct exchange of slides and photographs with other investigators active in the field, we obtained smears from normal subjects, smokers with or without leukoplakia, and tobacco chewers with or without leukoplakia. Our summarized findings follow: (1) Micronuclei represent only one of numerous cytological abnormalities in exfoliated buccal cells that are manifest particularly in tobacco chewers. These include a high frequency of anucleate, binucleate, and multinucleated cells, abnormal shapes and sizes of nuclei, etc.
(2) Intra-observer variability in the micronucleus count, assessed by counting the same group of slides on multiple occasions by the same observer, is in the range of k 3 per 1000 nucleated cells, i e . , 0.3%. Clearly, the impact of this can be extremely significant, particularly if the overall micronuclei frequency is low. (3) It is often difficult to find 1000 nucleated cells on a smear, especially if it is obtained from a tobacco chewer at the site of quid placement or at the site of a leukoplakic lesion. A large percent, often >50-70% of the exfoliated cells from such sites, will be anucleate. (4) Micronucleated cell frequency is very low in non-chewers. It is usually <0.5% (5 per 1000 nucleated cells), and frequently no micronucleated cells are present. (5) Smoking alone does not cause a significant increase in micronuclei frequency, irrespective of whether leukoplakia is present or not. In other words, nonchewers, with or without leukoplakia, who constitute a majority of the subjects of chemoprevention trials in the Western world, do not have the strikingly increased micronuclei frequencies reported by Stich et al. from studies conducted in Asia. In fact, the micronuclei frequency in this group is no different from lesion-free subjects. (6) Micronuclei frequency is increased in tobacco chewers, most significantly at the site of placement of the quid. These smears also have numerous other cytologic abnormalities, with micronuclei frequency being increased to about 0.8-1% (8-10 cells per 1000 nucleated cells).