Department of Pathology, Vancouver General Hospital, University of British Columbia, Vancouver, BC, Canada.
Am J Surg Pathol. 2012 Oct;36(10):1503-8. doi: 10.1097/PAS.0b013e31825d5371.
Most reports on antibodies that claimed to separate mesothelioma from pulmonary adenocarcinoma originated from academic centers or specialized immunohistochemistry laboratories, but little is known about how such stains perform in general practice laboratories. The Canadian Immunohistochemistry Quality Control program circulates tissue array slides to laboratories across Canada; these are stained and then interpreted by the local laboratory and by a set of experienced reviewers. For Canadian Immunohistochemistry Quality Control run 16, tissue array slides from 16 pulmonary adenocarcinomas and 6 mesotheliomas were stained in 36 different laboratories for CEA, CD15, CK5/6, and calretinin. A total of 736 results (cores) were interpretable. If 3 of 4 staining results concordant with the diagnosis was accepted as definitive, 166/192 (86.4%) mesothelioma cores and 461/544 (84.7%) adenocarcinoma cores were correctly diagnosed. However, if 4 of 4 concordant markers were required, then 93/192 (48.4%) mesothelioma cores and 265/544 (48.7%) adenocarcinoma cores were correctly diagnosed. Only 3/192 (1.6%) mesothelioma cores were incorrectly classified as carcinomas and 8/544 (1.5%) of adenocarcinoma cores incorrectly classified as mesotheliomas on the basis of the immunoprofile (ie, 3 of 4 or 4 of 4 marker results were discordant with the diagnosis). We conclude that, in a study based on results from nonspecialized laboratories, the combination of CEA, CD15, calretinin, and CK5/6, used as a panel, has a very low false-positive rate when separating pulmonary adenocarcinomas from mesotheliomas; however, single negative or incorrect results are common, therefore the panel is only useful diagnostically if 3 of 4 correct results are deemed acceptable for diagnosis.
大多数声称能够区分间皮瘤和肺腺癌的抗体报告都来自学术中心或专门的免疫组织化学实验室，但对于这些染色剂在常规实验室中的表现知之甚少。加拿大免疫组织化学质量控制计划向加拿大各地的实验室分发组织阵列载玻片；这些载玻片由当地实验室和一组经验丰富的审阅者进行染色和解释。在加拿大免疫组织化学质量控制运行 16 中，来自 16 个肺腺癌和 6 个间皮瘤的组织阵列载玻片在 36 个不同的实验室中用 CEA、CD15、CK5/6 和钙视网膜蛋白进行染色。共获得 736 个可解释的结果（核心）。如果接受 4 个染色结果中有 3 个与诊断一致作为确定结果，那么 166/192（86.4%）的间皮瘤核心和 461/544（84.7%）的腺癌核心被正确诊断。然而，如果需要 4 个一致的标志物，则 93/192（48.4%）的间皮瘤核心和 265/544（48.7%）的腺癌核心被正确诊断。只有 3/192（1.6%）的间皮瘤核心被错误地归类为癌，而 8/544（1.5%）的腺癌核心被错误地归类为间皮瘤，这是基于免疫组化结果（即，4 个标志物中有 3 个或 4 个标志物结果与诊断不一致）。我们的结论是，在一项基于非专业实验室结果的研究中，当将肺腺癌与间皮瘤区分开来时，CEA、CD15、钙视网膜蛋白和 CK5/6 的组合作为一个面板，具有非常低的假阳性率；然而，单个阴性或不正确的结果很常见，因此只有当 4 个正确结果中有 3 个被认为可接受用于诊断时，该面板才具有诊断价值。
