Berg Wendie A, Vargo Adrienne, Lu Amy H, Berg Jeremy M, Bandos Andriy I, Hartman Jamie Y, Zuley Margarita L, Ganott Marie A, Kelly Amy E, Nair Bronwyn E, Chough Denise M, Reginella Ruthane F, Wallace Luisa P, Harnist Kimberly S, Carlin Beatrice A, Cohen Cathy S, Hakim Christiane M, Sumkin Jules H
Department of Radiology, University of Pittsburgh School of Medicine, 300 Halket St, Pittsburgh, PA 15213.
UPMC Magee-Womens Hospital of UPMC, Pittsburgh, Pa.
Radiology. 2025 Jun;315(3):e242634. doi: 10.1148/radiol.242634.
Background Many state laws require insurance coverage for supplemental screening MRI in women at elevated risk for breast cancer, but MRI capacity is limited. Purpose To evaluate the impact of contrast-enhanced mammography (CEM) on incremental cancer detection rate (ICDR), false-positive rate (FPR) due to recall, and positive predictive value of biopsies performed (PPV3) when added to digital breast tomosynthesis (DBT) in women eligible for screening MRI. Materials and Methods From March 2021 to December 2022, 615 eligible women were prospectively recruited and consented to a single screening CEM examination with clinical DBT. Two radiologists interpreted each imaging study: Reader 1 recorded DBT findings first, and reader 2 recorded CEM findings first. Incremental cancer detection rate, cancer type and/or nodal status, FPR, PPV3, interval cancer rate, and areas under receiver operating characteristic curve (AUCs) based on forced Breast Imaging Reporting and Data System assessments were evaluated. Results Six hundred one women (mean age, 56 years [range: 30-75 years]) completed CEM. Twelve of the 601 women (2.0%) were diagnosed with 16 malignant lesions; cancers in five of the 12 women (42%) were detected by reader 1 at DBT, and one was detected at DBT by reader 2 (also seen on CEM by both observers). Cancers in the other six women were identified only with CEM (ICDR, six per 601 or 10.0 per 1000 women [95% CI: 3.3, 18.3]): five had invasive disease, all node negative, with a median lesion size of 0.7 cm (range: 0.4-1.1 cm); three were lobular. The FPR of combined DBT plus CEM was 127 of 589 (21.6%) for reader 1, which was an increase of 13.4% (95% CI: 10.8, 16.4) over DBT alone at 48 of 589 (8.1%). Despite increased FPR, the addition of CEM improved the overall AUC for reader 1 to 0.92 versus 0.73 for DBT alone ( = .016). Among the 601 women, 50 (8.3%) were recommended for biopsy according to CEM by reader 1, and six of the 50 (12%) were diagnosed with cancer. At the lesion level, the PPV3 was seven of 62 (11%) for biopsies prompted only with CEM by reader 1. There were no interval cancers at 1 year. Conclusion A significant increase in the detection of early-stage breast cancer was achieved using CEM after DBT. Despite substantially increasing the FPR, adding CEM to DBT improved the overall AUC by 0.19. ClinicalTrials.gov registration no.: NCT04764292 © RSNA, 2025 .
许多州法律要求为乳腺癌高危女性的补充筛查MRI提供保险覆盖，但MRI能力有限。目的：评估对比增强乳腺造影（CEM）在符合筛查MRI条件的女性中添加到数字乳腺断层合成（DBT）时对增量癌症检出率（ICDR）、召回导致的假阳性率（FPR）以及活检的阳性预测值（PPV3）的影响。材料与方法：2021年3月至2022年12月，前瞻性招募了615名符合条件的女性，并同意进行一次临床DBT的筛查CEM检查。两名放射科医生解读每项影像研究：读者1先记录DBT结果，读者2先记录CEM结果。评估了基于强制乳腺影像报告和数据系统评估的增量癌症检出率、癌症类型和/或淋巴结状态、FPR、PPV3、间隔期癌症率以及受试者操作特征曲线下面积（AUC）。结果：601名女性（平均年龄56岁[范围：30 - 75岁]）完成了CEM。601名女性中有12名（2.0%）被诊断出16个恶性病变；12名女性中有5名（42%）的癌症由读者1在DBT时检测到，1名由读者2在DBT时检测到（两名观察者在CEM上也都能看到）。其他6名女性的癌症仅通过CEM发现（ICDR，每601名中有6名，即每1000名女性中有10.0名[95% CI：3.3，18.3]）：5名患有浸润性疾病，均无淋巴结转移，病变大小中位数为0.7 cm（范围：0.4 - 1.1 cm）；3名是小叶癌。读者1的DBT加CEM联合检查的FPR为589例中的127例（21.6%），比单独DBT时589例中的48例（8.1%）增加了13.4%（95% CI：10.8，16.4）。尽管FPR增加，但添加CEM使读者1的总体AUC从单独DBT时的0.73提高到了0.92（P = 0.016）。在601名女性中，读者1根据CEM建议50名（8.3%）进行活检，其中6名（12%）被诊断为癌症。在病变层面，读者1仅因CEM提示进行活检的PPV3为62例中的7例（11%）。1年内无间隔期癌症。结论：在DBT后使用CEM可显著提高早期乳腺癌的检出率。尽管大幅增加了FPR，但在DBT中添加CEM使总体AUC提高了0.19。ClinicalTrials.gov注册号：NCT04764292 © RSNA，2025 。
