Fischmann A, Pietsch-Breitfeld B, Müller-Schimpfle M, Siegmann K, Wersebe A, Rothenberger-Janzen K, Claussen C D, Janzen J
Abteilung Radiologische Diagnostik, Universitätsklinikum Tübingen.
Rofo. 2004 Apr;176(4):538-43. doi: 10.1055/s-2004-812932.
To perform a statistical evaluation of microcalcifications (MC) from suspicious breast lesions detected by radiography and histopathology.
Histological and radiological detection of calcifications were compared from 116 biopsies in 96 women. Lesions with identical description of calcifications detected in histopathology and radiography were considered concordant, patients with obvious discrepancies discordant. If histological and radiological groups of calcifications were identical in number but differed in location, the case was considered pseudo-concordant.
Histopathology classified 24 of 116 lesions as malignant and 92 as benign. A total of 3196 core biopsies were examined, 851 of these contained groups of calcifications or single calcifications. Both modalities detected 579 calcifications, with 169 exclusively detected by radiography and 103 exclusively by histopathology. In 35 cases (30 %) radiologic and pathologic results were concordant, in 6 cases pseudo-concordant (4 %) and in 75 cases (65 %) discordant. The case-based Kappa coefficient was - 0.09 (- 0.24 to 0.07). The 122 calcifications not detected by histopathology were few or single calcifications at the edge of the core that were probably lost during processing, 18 were possible artefacts. Six cores contained calcium oxalate, 3 contained milk of calcium. In 6 cases malignant disease was found after the first examination, hence the cores were not searched thoroughly for the missing calcifications. In the remaining 14 cases, no calcifications were found despite complete processing of the tissue. In 49 of 103 cases of radiologically undetected microcalcifications, the retrospect analysis showed dense tissue areas that probably contained the calcification. The remaining 54 cases contained calcifications, which were too small to be detected radiologically.
Discordant results from pathological and radiological examinations of biopsies can mainly be explained by calcifications at the edge of the specimen lost during processing, which are therefore not detected in histopathology, and calcifications too small to be visualized radiologically.
对通过放射成像和组织病理学检测到的可疑乳腺病变中的微钙化（MC）进行统计学评估。
比较了96名女性116次活检中钙化的组织学和放射学检测结果。组织病理学和放射成像中对钙化描述相同的病变被视为一致，有明显差异的患者则为不一致。如果钙化的组织学和放射学分组数量相同但位置不同，则该病例被视为假一致。
组织病理学将116个病变中的24个分类为恶性，92个为良性。共检查了3196次芯针活检，其中851次包含钙化群或单个钙化。两种方法均检测到579处钙化，其中169处仅通过放射成像检测到，103处仅通过组织病理学检测到。在35例（30%）中，放射学和病理学结果一致，6例（4%）为假一致，75例（65%）为不一致。基于病例的Kappa系数为-0.09（-0.24至0.07）。组织病理学未检测到的122处钙化是芯针边缘的少量或单个钙化，可能在处理过程中丢失，18处可能是伪像。6个芯针含有草酸钙，3个含有钙乳。在6例中，首次检查后发现为恶性疾病，因此未对芯针进行彻底搜索以寻找缺失的钙化。在其余14例中，尽管对组织进行了完整处理，但未发现钙化。在103例放射学未检测到的微钙化病例中，回顾性分析显示49例存在可能包含钙化的致密组织区域。其余54例含有太小而无法通过放射学检测到的钙化。
活检的病理和放射学检查结果不一致主要可归因于处理过程中丢失的标本边缘钙化，因此在组织病理学中未检测到，以及太小而无法通过放射学显示的钙化。
