乳腺X线摄影BI-RADS 4级可疑恶性微钙化的风险分层评估

Risk stratification evaluation of suspicious malignant microcalcifications in BI-RADS category 4 of mammography

目的:结合微钙化的形态和分布描述,对乳腺X线摄影BI-RADS4级可疑恶性微钙化进行准确的风险分层。方法:回顾性分析2013年1月-2021年7月在哈尔滨医科大学附属第二医院经病理证实的乳腺X线摄影BI-RADS 4级可疑恶性微钙化患者1270例。以病理结果为金标准,采用χ^(2)检验或Fisher确切概率法分析可疑恶性微钙化的相关危险因素,计算并比较阳性预测值(PPV)。结果:患者的绝经状态、微钙化的形态及分布是BI-RADS 4级可疑恶性微钙化的相关危险因素(P<0.05),乳腺X线摄影BI-RADS 4级可疑恶性微钙化总体PPV为39.1%。各形态描述的PPV分别为:无定形钙化6.5%,粗糙不均质钙化22.7%,细小多形性钙化84.8%,细线状或细线分支状钙化91.1%。分布描述的PPV如下:区域分布32.4%,成簇分布36.9%,线样分布69.2%,段样分布80.3%;对于形态和分布描述组合,无定形钙化区域、成簇分布、线样或段样分布的PPV分别为4.1%、5.2%、40.9%;粗糙不均质钙化区域分布、成簇分布、线样或段样分布的PPV分别15.8%、24.5%、63.6%。结论:由于无定形、粗糙不均质和细小多形性微钙化的PPV存在差异,建议在形态描述进行亚分级,并结合可疑微钙化的形态和分布描述可进行准确的风险分层,尤其是对于绝经后的患者,恶性可能性更高。

Objective:To accurately stratify the risk of suspicious malignant microcalcifications in BI-RADS category 4 of mammography by combining their morphological and distribution characteristics.Methods:1270 cases of BI-RADS category 4 suspected malignant microcalcifications confirmed by pathology in the Second Affiliated Hospital of Harbin Medical University from January,2013 to July,2021 were retrospectively analyzed.Using the pathological results as the gold standard,the risk factors of suspected malignant microcalcifications were analyzed and the positive predictive value(PPV)was calculated and compared byχ^(2) test or Fisher's exact test.Results:Menopausal status,morphology and distribution of microcalcifications,were found to be significant risk factors for BI-RADS category 4(P<0.05),The overall PPV for BI-RADS category 4 suspicious malignant microcalcifications in mammography was 39.1%.The PPV for different morphological descriptions were as follows:amorphous calcifications,6.5%;coarse heterogeneous calcifications,22.7%;fine pleomorphic calcifications,84.8%;fine linear/fine linear branching calcifications,91.1%.The PPV for different distribution descriptions were as follows:regional distribution,32.4%;clustered distribution,36.9%;linear distribution,69.2%;segmental distribution,80.3%.For the combination of morphology and distribution descriptions,the PPV for amorphous calcifications in regional distribution,clustered distribution,and linear or segmental distribution were 4.1%,5.2%,40.9%respectively.PPVs for coarse heterogeneous calcifications in regional,clustered,and linear or segmental distribution were 15.8%,24.5%,63.6%respectively.Conclusion:Due to the varying PPVs of amorphous,coarse heterogeneous,and fine pleomorphic microcalcifications,it is recommended to further subcategorize the morphological descriptions.Combining the morphology and distribution descriptions of suspicious microcalcifications can enable accurate risk stratification,particularly for postmenopausal patients who have a higher likelihood of malignancy.