Statistical Analysis of Histopathological Parameters for Assessing Prognosis of Oral Squamous Cell Carcinoma Cases

本文对10例口腔鳞状细胞癌（SCC）福尔马林固定、石蜡包埋标本进行了DNA倍体分析，以选择活检标本的方法。两种方法中DNA直方图显示完全相同的病例较少。本研究收集了25例接受过放射治疗且有足够临床和随访资料的舌鳞状细胞癌患者，采用显微分光光度法测定其单细胞DNA含量，结果表明，显微分光光度法测定单细胞DNA含量的准确性较差。从石蜡包埋材料中分离肿瘤细胞，用DADI染色，将50-100个肿瘤细胞的核DNA含量制成DNA直方图。根据奥尔等人的标准，DNA倍体类型可分为4种类型。将具有单个二倍体（2C）峰的I型和具有二倍体（2C）和/或四倍体（4C）峰的II型总结为低倍性组。III型细胞的DNA含量在2C和4C之间，代表增殖细胞，IV型细胞的DNA分布范围从2C到超过8 C，这些类型被概括为高倍性组。两组生存曲线（Kaplan-Meier法）显示高倍体病例预后更差，但由于分析病例数不足，两组生存曲线经logrank检验或Cox-Mantel检验差异无统计学意义。高倍体组的死亡率（9/14）高于低倍体组（4/11）。

DNA ploidy analysis of formalin-fixed, paraffin embedded materials of 10 cases of oral squamous cell carcinoma(SCC) was performed by two different methods, microspectrophotmetry with 5 micron sections and cytofluorometry with isolated tumor cells to choose the method for the biopsy materials. There were few cases which showed same DNA histogram pattern in two methods. This was due to the insufficient accuracy in measurement of single cell DNA content with the microspectrophotometric method.Twenty-five cases with the tongue SCC,which had recieved radiation therapy and had sufficient clinical and follow-up data, were available for this study. Isolated tumor cells from the paraffinembedded materials were stained with DADI and DNA histogram was made of nuclear DNA content of 50-100 tumor cells. According to the criteria of Auer et al., DNA polidy patterns were divided into 4 types. Type I with a single diploid (2C)peak and type II with a diploid(2C) and/or a tetraploid(4C)peak were summarized as low ploidy group. Type III with a sizable amount of cells with DNA content between 2C and 4C representing proliferating cells and type IV showing a broad DNA distribution ranging from 2C to beyond 8C were summarized as high ploidy group. Survival curves (Kaplan-Meier method) of two groups demonstrated the more unfavorable prognosis in high ploidy cases, however, the number of analysed case was not sufficient so that the difference between two survival curves was not statistically significant with logrank test or Cox-Mantel test. Death was more frequently encountered in the high ploidy cases(9/14) than in the low ploidy cases(4/11).