Abstract: Background　Mismatch repair（MMR） status of colorectal neoplasms is a biomarker for tumor screening and clinical management.It is routinely tested in pathological examination.Objective　To explore the problem and solution in the interpretation of immunohistochemistry （IHC） for MMR protein.Methods　A total of 3 357 patients received MMR IHC test in Sun Yat-sen University Cancer Center between November 2010 and December 2015 were enrolled.A retrospective analysis of IHC test for MMR protein was carried out and reassessment was made to screen out the cases with irregular and controversial interpretation，and summarized and reinterpret their test results.Results　One hundred and three patients with irregular interpretation were controversial：（1） IHC results in percentage form，such as MLH1 （80%+），were reported in 25 cases；（2） IHC results were described in semi-quantitative form，i.e.weak +，focal +，partial +，individual + and others.A total of 75 staining indexes in this form were obtained from 65 patients；（3） 14 staining index from 13 patients were fuzzy expression as +/-.After reevaluating，the results were as follows：（1） IHC results in 25 cases described in percentage form were all stained positive．（2） Among 75 semi-quantitative IHC results，69 were stained positive after reinterpretation，5 were stained negative，and one was still undecidable （loss of tumor tissue，no re-staining）．（3） Among 14 IHC results vaguely expressed as +/-，13 indexes were positive，and one index was negative after reinterpretation.Three patients showing four MMR proteins deficiency were also reevaluated.One patient resulted in four MMR proteins positive；another two patients showed MSH6 negative and paired MLH1-PMS2 negative，respectively.Conclusion　The reasons for the controversy in interpreting the results of immunohistochemistry of MMR proteins are mainly that the IHC results are described in percentage form，semi-quantitative form，fuzzy expression as +/- and negative staining of four MMR proteins.It should be noted that the results of immunohistochemistry of MMR proteins are only negative and positive，and there are very few missing expressions of all four proteins.Review of questionable staining results will help to improve the accuracy of interpretation.

Key words: Colorectal neoplasms；Adenocarcinoma；Mismatch repair, immunohistochemistry；Dissent and disputes；Outcome assessment

摘要： 背景　结直肠癌错配修复（MMR）状态可以指导肿瘤筛查和治疗决策，目前已成为病理检测的常规内容。目的　探讨MMR蛋白免疫组化结果判读中存在的问题与对策。方法　选取2010年11月—2015年12月在中山大学肿瘤防治中心就诊并进行MMR蛋白免疫组化检测的结直肠腺癌患者，共3 357例。回顾性分析其MMR蛋白免疫组化检测报告，并进行再评估，筛选其中判读不规范和存在争议的病例，总结并重读MMR蛋白免疫组化检测结果。结果　共发现103例患者的判读结果表述存在争议，主要为以下几种类型：（1）以百分数形式描述免疫组化结果，如MLH1（80%+）等，共计25例患者；（2）以半定量形式描述免疫组化结果，即表述为弱+、局灶+、部分+、个别+等半定量形式，共计65例患者的75个染色指标；（3）免疫组化结果模糊表述为+/-，共计13例患者的14个染色指标。再次阅片判读，重读结果如下：（1）25例以百分数形式描述的免疫组化结果，重读结果均为染色阳性。（2）75个以半定量形式描述的免疫组化结果，重读后69个指标为染色阳性，5个指标显示染色阴性，1个指标仍无法判定（肿瘤组织缺失，无法重染复核）。（3）14个免疫组化结果模糊表述为+/-的指标，重读后13个指标为染色阳性，1个指标为染色阴性。3例4个MMR蛋白染色结果均为阴性的患者，经重新切片染色后，1例4个MMR蛋白均染色阳性；1例为MSH6蛋白染色阴性；1例为MLH1、PMS2蛋白成对染色阴性。结论　MMR蛋白免疫组化结果判读存在争议的原因主要在于以百分数形式描述、以半定量形式描述、模糊表述为+/-以及4个MMR蛋白染色结果均为阴性等情况，应注意MMR蛋白免疫组化结果只有阴性和阳性，且极少存在4个蛋白全部表达缺失，复核有疑问的染色结果有助于提高判读准确性。

关键词: 结直肠肿瘤；腺癌；错配修复蛋白, 免疫组化；异议和争论；结果评价