Siminiak Natalia, Pasiuk-Czepczyńska Anna, Godlewska Antonina, Wojtyś Piotr, Olejnik Magdalena, Michalak Joanna, Nowaczyk Piotr, Gajdzis Paweł, Godlewski Dariusz, Ruchała Marek, Czepczyński Rafał
Department of Endocrinology, Metabolism and Internal Diseases, Poznan University of Medical Sciences, Poznań, Poland.
Cancer Prevention and Epidemiology Center, Poznań, Poland.
Front Oncol. 2022 Nov 24;12:941312. doi: 10.3389/fonc.2022.941312. eCollection 2022.
Full-field digital mammography (FFDM) is widely used in breast cancer screening. However, to improve cancer detection rates, new diagnostic tools have been introduced. Contrast enhanced mammography (CEM) and digital breast tomosynthesis (DBT) are used in the diagnostic setting, however their accuracies need to be compared.The aim of the study was to evaluate the diagnostic performance of CEM and DBT in women recalled from breast cancer screening program.
The study included 402 consecutive patients recalled from breast cancer screening program, who were randomized into two groups, to undergo either CEM (202 patients) or DBT (200 patients). All visible lesions were evaluated and each suspicious lesion was histopathologically verified.
CEM detected 230 lesions; 119 were classified as benign and 111 as suspicious or malignant, whereas DBT identified 209 lesions; 105 were classified as benign and 104 as suspicious or malignant. In comparison to histopathology, CEM correctly detected cancer in 43 out of 44 cases, and DBT in all 33 cases, while FFDM identified 15 and 18 neoplastic lesions in two groups, respectively. CEM presented with 97% sensitivity, 63% specificity, 70% accuracy, 38% PPV and 99% NPV, while DBT showed 100% sensitivity, 60% specificity, 32%, PPV, 100% NPV and 66% accuracy. The CEM's AUC was 0.97 and DBT's 0.99. The ROC curve analysis proved a significant (p<0.000001) advantage of both CEM and DBT over FFDM, however, there was no significant difference between CEM and DBT diagnostic accuracies (p=0.23).
In this randomized, prospective study CEM and DBT show similar diagnostic accuracy.
全视野数字化乳腺摄影(FFDM)广泛应用于乳腺癌筛查。然而,为提高癌症检出率,已引入新的诊断工具。对比增强乳腺摄影(CEM)和数字乳腺断层合成(DBT)用于诊断,但它们的准确性需要比较。本研究的目的是评估CEM和DBT在从乳腺癌筛查项目中召回的女性中的诊断性能。
该研究纳入了402例从乳腺癌筛查项目中连续召回的患者,这些患者被随机分为两组,分别接受CEM(202例患者)或DBT(200例患者)。对所有可见病变进行评估,每个可疑病变均经组织病理学证实。
CEM检测到230个病变;119个被分类为良性,111个为可疑或恶性,而DBT识别出209个病变;105个被分类为良性,104个为可疑或恶性。与组织病理学相比,CEM在44例病例中的43例中正确检测出癌症,DBT在所有33例病例中均正确检测出癌症,而FFDM在两组中分别识别出15个和18个肿瘤性病变。CEM具有97%的敏感性、63%的特异性、70%的准确性、38%的阳性预测值和99%的阴性预测值,而DBT显示100%的敏感性、60%的特异性、32%的阳性预测值、100%的阴性预测值和66%的准确性。CEM的曲线下面积(AUC)为0.97,DBT的为0.99。ROC曲线分析证明CEM和DBT相对于FFDM均具有显著优势(p<0.000001),然而,CEM和DBT的诊断准确性之间无显著差异(p=0.23)。
在这项随机、前瞻性研究中,CEM和DBT显示出相似的诊断准确性。