Nathanson S D, Schultz L, Tilley B, Kambouris A
Am Surg. 1986 Aug;52(8):428-33.
Five staging methods for colorectal cancer were compared to the Dukes method in order to assess which methods were most sensitive and specific. Three hundred fifty-two patients with resected carcinomas of the colon and rectum were followed from 6 to 11 years or until death. All patients were staged by intraoperative and pathological criteria. Survival curves were constructed for each stage. The pTNM system was most sensitive at predicting death. The Gastrointestinal Tumor Study Group (GITSG) system was the most specific at predicting survival. However, the Dukes system did no worse than either of these staging systems and predicted relatively few false-positive and false-negative cases. The sensitivity and correctness of the Dukes system was improved by adding information regarding percent of lymph node metastases, ulceration of the primary tumor, and adjacent organ invasion.
为了评估哪种分期方法最敏感且特异,对结直肠癌的五种分期方法与Dukes分期法进行了比较。352例接受结肠和直肠癌切除术的患者随访了6至11年或直至死亡。所有患者均根据术中及病理标准进行分期。为每个阶段构建了生存曲线。pTNM系统在预测死亡方面最敏感。胃肠道肿瘤研究组(GITSG)系统在预测生存方面最特异。然而,Dukes系统并不比这两种分期系统差,且预测的假阳性和假阴性病例相对较少。通过添加有关淋巴结转移百分比、原发肿瘤溃疡和邻近器官侵犯的信息,Dukes系统的敏感性和准确性得到了提高。