1970 年 9 巻 2 号 p. 135-148
Since the advance in technique has made it possible to obtain fresh specimens from the stomach, re-evaluation of the Papanicolaou's criteria for malignancy established chiefly by the use of the lavage method has become an important issue for the routine cytodiagnosis. The present paper is concerned with a morphological analysis of cancer cells which has been made in careful comparison with non-malignant cells. 195 specimens obtained from stomachs resected or biopsied for gastric cancer (90 cases) or non cancerous gastric lesions (105 cases) were subjected to the study. The results obtained were as follows:
1. As far as the nuclear size was concerned, cells with a nuclear diameter of more than 15, a consisted mostly of malignant cells and those with a nuclear diameter of more than 20 were all malignant cells. The standard deviation of nuclear diameter distribution in individual case was greater in cancer group than in noncancer group, indicating that the nuclear size of cancer cells was varied considerably in each case.
2. As for the nuclear pattern, an increase in chromatin quantity and an uneven dispersal of chromatin, which had been regarded as the most important criteria of malignancy, were found at a fairly high rate even in non-cancerous cells in the cases of gastritis and ulcer. On the other hand, fresh cancer cells showed, at a higher rate, a fine reticular structure of chromatin, which had been regarded so far as one of the criteria of non-malignancy. As far as the nucle olus was concerned, no finding specific for cancer cells was obtained in this series.
3. In individual specimens containing compara tively large number of cells, the nucleus and nucleolus were examined for following 8 morphological items: nuclear diameter in mode, irregularity in nuclear contour, irregular thickening of nuclear margin, hyperchromatism, irregularity in distribution of chromatin, coarseness in the pattern of chromatin, large nucleolus (more than 2μ) and irregular nucleolar contour. Decision as to whether most of the cases examined were benign or malignant was made by evaluating the frequency of each item alone or the combination of each item. Cases with cells showing hyperchromatism at a rate of less than 50 percent or with cells having large nucleolus at the same rate were judged as possibly benign. It was most noteworthy that these two items together with the measurement of nuclear diameter were available as automatic cytoscreener.
4. For practical purpose of cytoscreening, the cells were divided into three groups according to the nuclear diameter: cells less than 10μ, those 11-15, μ and those more than 16μ it in diameter, respectively. In each group, morphological characteristics of the nucleus were analysed.