Amitotic cell divisions and tumour growth: an alternative model for cell kinetic compartments in solid tumours.

Previous studies on cytological assay of amitotic changes such as micronucleation or nuclear budding (akaryokinesis) and multinucleation (acytokinesis; Bhattathiri et al., Serial cytological assay of micronucleus induction--a new tool to predict human cancer radiosensitivity. Radiother Oncol 1996; 41: 139-142; Serial cytological assay of micronucleus induction--a new tool to predict human cancer radiosensitivity. Radiother Oncol. 1997; 41: 139-142; Radiation-induced acute immediate nuclear abnormalities in oral cancer cells. Serial cytologic evaluation. Acta Cytol 1998; 42: 1084-1090) had suggested that, in addition to predicting radiosensitivity, they may be related to proliferation characteristics of tumours. Hence the present study was undertaken to see if their pre-treatment frequency was related to the clinical growth characteristics of tumours. Smears from 121 untreated oral cancers were stained with Giemsa and the frequency, in percentage of total cells counted, of micronucleated or nuclear budded, binucleated and multinucleated cells taken as the akaryokinesis index (AKI), mitotic index (MI) and acytokinesis index (ACI), respectively were evaluated. The sum of AKI and MI was taken as the amitotic index (AMI), and the sum of AMI and MI as the cell division index (CDI). The tumours were divided to three groups according to size: <2 cm (Tsize1); 2-4 cm (Tsize2) and >4 cm (Tsize3). The tumours were divided to those with duration of <4 months and > or =4 months, this data having been collected from the patient. The differences in the frequency of the indices among the three size groups were analysed by Kruskall-Wallis one-way analysis of variance. Within each size group, the differences in frequency of the indices betwen the two duration groups was analysed by Mann-Whitney U test. Larger tumours had significantly higher CDI, the median frequency being 1.2, 2.29 and 3.28% in Tsize1, Tsize2 and Tsize3 tumours, respectively (P=0.0025). In the case of AMi, the frequency was significantly higher the median values being 1.0, 6.3 and 10.05%, respectively (P=0.0015). The difference was not significant in the case of the MI, the frequency being 0.69, 0.99 and 1.32%, respectively. AKI and ACI also showed significant increase. As regards relation to clinical growth, those tumours which reached larger size in shorter duration had higher frequency of MI. Those tumours which remained small in spite of a longer duration of growth had statistically significantly higher AKI whereas tumours which reached Tsize2 in shorter duration of growth had significantly higher ACI. The results suggests that cytological evaluation of mitoses and amitoses can be helpful in evaluating proliferation characteristics of solid tumours. Micronucleated and multinucleated cells are clonogenically dead, but not physically so, but continue to divide for a few times before dying. The significant increase in AMI with size suggests that induction of amitoses is an important mechanism of cell loss as tumours enlarge, probably induced by associated hypovascularity, and precedes necrosis. The high frequency of amitoses in untreated tumours suggest that proliferation markers such as Tpot, growth fraction, etc., as measured by techniques such as flowcytometry, thymidine and BrdUrd labelling, etc., which do not evaluate amitoses, may essentially be wrong. Based on the findings of the study, an alternative model for tumour cell kinetic compartmentalisation, which includes a non-clonogenic compartment in addition to clonogenic and quiscent compartments, is presented.