【摘要】 目的 确定结直肠癌患者有效的预后参数及最佳预后指标。方法 对83例不同分期的结直肠癌肿瘤转移抑制基因Nm23的表达、巨噬细胞的浸润、人类白细胞抗原HLA-DR的表达、肿瘤的浸润深度、Dukes分期及其它临床指标进行单因素和多因素COX回归分析。结果 单因素分析得出的有关预后因素中，经多因素分析只有Dukes分期一项进入COX模型。结论 Dukes分期是结直肠癌患者最重要和最佳的预后指标。

The COX regression analysis of prognosis factors

    of colorectal carcinoma

YANG Chunkang, XU Dongpo, CHEN Daoda. Department of Surgery, First Affiliated Hospitalof Fujian Medical University, Fuzhou 350005

【 Abstract】 Objective To determine the effective prognostic parameter and the best prognostic index. Methods The prognostic relevance of clinical and pathological variables were evaluated in 83 patients with histologically proved colorectal carcinoma, (including the expression of Nm23, HLA-DR, the infiltration of macrophage) by the multivariate analysis of COX regression. Results Dukes' staging was an only parameter to enter the COX model. Conclusion The Dukes' staging is the most important, reliable and best varible factor in predicting the clinical outcome of colorectal cancer.

     ......