Stenkvist B, Bengtsson E, Eriksson O, Jarkrans T, Nordin B, Westman-Naeser S
J Clin Pathol. 1983 Apr;36(4):392-8. doi: 10.1136/jcp.36.4.392.
The inter- and intraobserver reproducibilities of the histopathological systems of breast cancer classification suggested by the World Health Organisation (WHO), the Armed Forces Institute of Pathology (AFIP) and Ackerman have been analysed. The reproducibilities of the three classification systems were only "fair" to "moderate" and no correlation with the five-year recurrence rate was found. Our results indicate that these classification systems are without biological significance and are useless for prognosis in the individual patient. When the tumours were classified according to degree of differentiation (high, moderate, low) or graded according to WHO (which includes both differentiation and nuclear atypia), however, there was a significant correlation with the five-year recurrence rate. Yet even such "reduced" subdivisions are of no value in judging prognosis for the individual patient at the time of diagnosis; rather, they are useful only in the follow-up analysis of groups of patients.
对世界卫生组织(WHO)、武装部队病理研究所(AFIP)以及阿克曼所提出的乳腺癌组织病理学分类系统的观察者间及观察者内可重复性进行了分析。这三种分类系统的可重复性仅为“一般”到“中等”,且未发现与五年复发率存在相关性。我们的结果表明,这些分类系统没有生物学意义,对个体患者的预后毫无用处。然而,当根据分化程度(高、中、低)对肿瘤进行分类,或根据WHO(包括分化和核异型性)进行分级时,与五年复发率存在显著相关性。即便如此,这种“简化”的细分在诊断时判断个体患者的预后仍毫无价值;相反,它们仅在对患者群体的随访分析中有用。