Department of Radiology, University of Bonn, Sigmund-Freud-Str 25, D-53105 Bonn, Germany.
J Clin Oncol. 2010 Mar 20;28(9):1450-7. doi: 10.1200/JCO.2009.23.0839. Epub 2010 Feb 22.
We investigated the respective contribution (in terms of cancer yield and stage at diagnosis) of clinical breast examination (CBE), mammography, ultrasound, and quality-assured breast magnetic resonance imaging (MRI), used alone or in different combination, for screening women at elevated risk for breast cancer.
Prospective multicenter observational cohort study. Six hundred eighty-seven asymptomatic women at elevated familial risk (> or = 20% lifetime) underwent 1,679 annual screening rounds consisting of CBE, mammography, ultrasound, and MRI, read independently and in different combinations. In a subgroup of 371 women, additional half-yearly ultrasound and CBE was performed more than 869 screening rounds. Mean and median follow-up was 29.18 and 29.09 months.
Twenty-seven women were diagnosed with breast cancer: 11 ductal carcinoma in situ (41%) and 16 invasive cancers (59%). Three (11%) of 27 were node positive. All cancers were detected during annual screening; no interval cancer occurred; no cancer was identified during half-yearly ultrasound. The cancer yield of ultrasound (6.0 of 1,000) and mammography (5.4 of 1,000) was equivalent; it increased nonsignificantly (7.7 of 1,000) if both methods were combined. Cancer yield achieved by MRI alone (14.9 of 1,000) was significantly higher; it was not significantly improved by adding mammography (MRI plus mammography: 16.0 of 1,000) and did not change by adding ultrasound (MRI plus ultrasound: 14.9 of 1,000). Positive predictive value was 39% for mammography, 36% for ultrasound, and 48% for MRI.
In women at elevated familial risk, quality-assured MRI screening shifts the distribution of screen-detected breast cancers toward the preinvasive stage. In women undergoing quality-assured MRI annually, neither mammography, nor annual or half-yearly ultrasound or CBE will add to the cancer yield achieved by MRI alone.
我们研究了临床乳房检查（CBE）、乳房 X 线照相术、超声以及经过质量保证的乳房磁共振成像（MRI）各自的贡献（就癌症发生率和诊断时的分期而言），这些方法单独或不同组合用于筛查乳腺癌风险升高的女性。
前瞻性多中心观察性队列研究。687 名无症状、家族性风险升高（>或= 20%终生）的女性接受了 1679 次年度筛查，包括 CBE、乳房 X 线照相术、超声和 MRI，这些检查均独立进行且组合方式不同。在 371 名女性的亚组中，超过 869 次筛查进行了半年一次的超声和 CBE。平均和中位随访时间分别为 29.18 和 29.09 个月。
27 名女性被诊断患有乳腺癌：11 名导管原位癌（41%）和 16 名浸润性癌（59%）。3 名（11%）患者淋巴结阳性。所有癌症均在年度筛查期间检出；无间期癌；半年一次的超声未发现癌症。超声（每 1000 人中有 6.0 例）和乳房 X 线照相术（每 1000 人中有 5.4 例）的癌症发生率相当；如果两种方法联合使用，其发生率无显著增加（每 1000 人中有 7.7 例）。单独使用 MRI 的癌症发生率（每 1000 人中有 14.9 例）显著较高；增加乳房 X 线照相术无显著改善（MRI 加乳房 X 线照相术：每 1000 人中有 16.0 例），增加超声也无变化（MRI 加超声：每 1000 人中有 14.9 例）。乳房 X 线照相术的阳性预测值为 39%，超声为 36%，MRI 为 48%。
在家族性风险升高的女性中，经过质量保证的 MRI 筛查将筛查检出的乳腺癌分布向早期浸润前阶段转移。在接受年度质量保证 MRI 检查的女性中，乳房 X 线照相术、年度或半年一次的超声或 CBE 均不会增加 MRI 单独检查的癌症发生率。
