From the Department of Breast Radiology, Veneto Institute of Oncology (IRCCS), Via Gattamelata 64, 35128 Padua, Italy (F.C., G.G.); Department of Radiology, Azienda Ospedaliera Universitaria Integrata Verona, Verona, Italy (S.M.); Breast Unit ULSS 9, Ospedale di Marzana, Verona, Italy (G.R., P.B., L.C.); and Department of Radiology, Ospedale Pederzoli, Peschiera del Garda, Verona, Italy (S.B.).
Radiology. 2021 Jan;298(1):49-57. doi: 10.1148/radiol.2020201246. Epub 2020 Nov 10.
Background Few results are available about subsequent outcomes after screening with digital breast tomosynthesis (DBT). Purpose To compare the diagnostic accuracy of a screening round with DBT plus synthetic mammography (SM) (hereafter, DBT+SM) and the repeat screening round with DBT with SM (hereafter, DBT+SM) or full-field digital mammography (FFDM) with FFDM screening. Materials and Methods This prospective study (Verona Pilot Study, clinical trial identification: 2015/1238) included women screened with DBT+SM between April 2015 and March 2017 and rescreened with DBT+SM or FFDM between April 2017 and March 2019. Screening performance (recall rate, cancer detection rate [CDR], and positive predictive value of recall [PPV1]) was compared with that obtained from 28 680 women screened with FFDM between 2013 and 2014 (control group). Cancer stages were compared between modalities and screening rounds. A χ test was used to evaluate differences. < .05 was indicative of a statistically significant difference. Results Of 34 638 women enrolled, 32 870 (median age, 58 years; age range, 52-71 years) underwent repeat screening-16 198 with DBT+SM and 16 672 with FFDM. The CDR was higher for repeat screening with DBT+SM than for the control group with FFDM (8.1 per 1000 women screened vs 4.5 per 1000 women screened, respectively; < .01) and was not significantly lower for repeat screening with FFDM (3.5 per 1000 women screened vs 4.5 per 1000 women screened, respectively; = .11). Compared with the control group, there was no difference in the recall rate at repeat screening with both DBT+SM (3.71% vs 3.40%, respectively; = .10) and FFDM (3.71% vs 3.69%, = .92), whereas PPV1 was higher only when repeat screening was performed with DBT+SM (23.8% vs 12.0%, < .01). At repeat screening, the proportion of cancers stage II or higher was 14.5% (19 of 131 cancers) with DBT+SM and 8.5% (five of 59 cancers) with FFDM, both of which were lower than the proportion in the control group with FFDM (30 of 110 cancers, 27.3%) ( ≤ .01). Conclusion At repeat screening, digital breast tomosynthesis plus synthetic mammography depicted more cancers than full-field digital mammography (FFDM) and found a lower number of stage II cancers compared with FFDM. © RSNA, 2020 See also the editorial by Bae in this issue.
背景 关于数字乳腺断层合成摄影术（digital breast tomosynthesis，DBT）筛查后的后续结果，相关研究结果寥寥无几。
目的 比较单次 DBT 加合成乳腺 X 线摄影术（synthetic mammography，SM）（以下简称 DBT+SM）筛查与 DBT 联合 SM 或全数字化乳腺 X 线摄影术（full-field digital mammography，FFDM）重复筛查的诊断准确性。
材料与方法 本前瞻性研究（Verona Pilot Study，临床试验注册号：2015/1238）纳入了 2015 年 4 月至 2017 年 3 月期间接受 DBT+SM 筛查且 2017 年 4 月至 2019 年 3 月期间接受 DBT+SM 或 FFDM 重复筛查的女性。比较了 DBT+SM 筛查的召回率、癌症检出率（cancer detection rate，CDR）和召回的阳性预测值（positive predictive value of recall，PPV1）与 2013 年至 2014 年期间接受 FFDM 筛查的 28680 例女性（对照组）的筛查结果。比较了不同检查方法和筛查轮次的癌症分期。采用 χ ²检验评估差异， ＜.05 表示具有统计学意义。
结果 在 34638 例入组的女性中，32870 例（中位年龄为 58 岁；年龄范围为 52-71 岁）接受了重复筛查，其中 16198 例接受了 DBT+SM 重复筛查，16672 例接受了 FFDM 重复筛查。与对照组 FFDM 筛查相比，DBT+SM 重复筛查的 CDR 更高（每 1000 例筛查女性中分别为 8.1 例和 4.5 例； ＜.01），而 FFDM 重复筛查的 CDR 并不显著更低（每 1000 例筛查女性中分别为 3.5 例和 4.5 例； =.11）。与对照组相比，DBT+SM（3.71%比 3.40%； =.10）和 FFDM（3.71%比 3.69%； =.92）重复筛查的召回率差异均无统计学意义，而 DBT+SM 重复筛查的 PPV1 更高（23.8%比 12.0%； ＜.01）。在重复筛查时，DBT+SM 组有 19 例（14.5%）癌症分期为Ⅱ期或更高，FFDM 组有 5 例（8.5%）癌症分期为Ⅱ期或更高，均低于对照组 FFDM 组的 30 例（27.3%）（ ＜.01）。
结论 在重复筛查时，DBT 加 SM 比 FFDM 检出更多的癌症，与 FFDM 相比，发现更少的Ⅱ期癌症。
版权所有©2020 美国放射学会。
参见本期社论，作者 Bae。
