Farshid Gelareh, Balleine Rosemary L, Cummings Margaret, Waring Paul
Division of Tissue Pathology, Institute of Medical and Veterinary Science, Adelaide, South Australia.
Am J Surg Pathol. 2006 Nov;30(11):1357-66. doi: 10.1097/01.pas.0000213273.22844.1a.
Women who have germline mutations in the BRCA1 gene are at substantially increased lifetime risk of developing breast and ovarian cancer but are otherwise normal. Currently, early age of onset of cancer and a strong family history are relied upon as the chief clues as to who should be offered genetic testing. Certain morphologic and immunohistochemical features are overrepresented in BRCA1-associated breast cancers but these differences have not been incorporated into the current selection criteria for genetic testing.
Each of the 4 pathologists studied 30 known cases of BRCA1- and BRCA2-associated breast cancer from kConFab families. After reviewing the literature, we agreed on a semiquantitative scoring system for estimating the chances of presence of an underlying BRCA1 mutation, based on the number of the reported prototypic features present. After a time lag of 12 months, we each examined a series of 62 deidentified cases of breast cancer, inclusive of cases of BRCA1-associated breast cancer and controls. The controls included cases of BRCA2-associated breast cancer and sporadic cases.
Our predictions had a sensitivity of 92%, specificity of 86%, positive predictive value of 61%, and negative predictive value of 98%. For comparison the sensitivity of currently used selection criteria are in the range of 25% to 30%.
The inclusion of morphologic and immunohistochemical features of breast cancers in algorithms to predict the likelihood of presence of germline mutations in the BRCA1 gene improves the accuracy of the selection process.
携带BRCA1基因种系突变的女性患乳腺癌和卵巢癌的终生风险大幅增加,但在其他方面则是正常的。目前,癌症发病的早期年龄和强烈的家族病史是确定谁应接受基因检测的主要线索。某些形态学和免疫组化特征在BRCA1相关乳腺癌中更为常见,但这些差异尚未纳入当前的基因检测选择标准。
4位病理学家每人研究了来自kConFab家族的30例已知的BRCA1和BRCA2相关乳腺癌病例。在查阅文献后,我们商定了一个半定量评分系统,根据所呈现的报告原型特征的数量来估计存在潜在BRCA1突变的可能性。在12个月的时间间隔后,我们每人检查了一系列62例身份不明的乳腺癌病例,包括BRCA1相关乳腺癌病例和对照。对照包括BRCA2相关乳腺癌病例和散发病例。
我们的预测敏感性为92%,特异性为86%,阳性预测值为61%,阴性预测值为98%。相比之下,目前使用的选择标准的敏感性在25%至30%之间。
在预测BRCA1基因种系突变可能性的算法中纳入乳腺癌的形态学和免疫组化特征可提高选择过程的准确性。