Narayan Anand K, Keating Delia M, Morris Elizabeth A, Mango Victoria L
Memorial Sloan Kettering Cancer Center, 1275 York Avenue, New York, NY 10065, USA.
Clin Imaging. 2019 Jan-Feb;53:151-154. doi: 10.1016/j.clinimag.2018.09.016. Epub 2018 Oct 5.
The majority of recalls from screening mammography do not result in biopsy recommendation. The purpose of this study was to evaluate if calcifications recalled from screening mammography are more likely to result in biopsy recommendations than other findings.
IRB-approved electronic medical record search was performed to obtain a random sample of screening mammograms assigned BI-RADS 0 assessment during 2014-2015. Primary reason for recall was classified as mass, asymmetry, focal asymmetry, calcifications, or distortion. Primary outcome was biopsy performed after diagnostic work-up. Secondary outcome was proportion of biopsies performed that were positive for cancer, positive predictive value 3 (PPV3). Logistic regression was used to compare reasons for recall (calcifications vs other findings) with biopsy recommendation proportions.
Random database sampling yielded 402 screening examinations with BI-RADS 0 assessments with 449 total findings. Reasons for recall included calcifications (14.0%, 63/449), masses (15.8%, 71/449), asymmetries (50.8%, 228/449), focal asymmetries (14.3%, 64/449) and architectural distortions (5.1%, 23/449). Overall, 21.6% of recalls led to image-guided biopsy (87/402). Recalls for calcifications were more likely to result in biopsy compared with other types of findings (Adjusted OR 8.56, 95% CI 4.58 to 16.0, p < 0.001). No statistically significant differences were found in PPV3 proportions between calcification and non-calcification findings (p = 0.812).
Recalls for calcifications are much more likely to undergo biopsy compared with other findings. Increased biopsy rates for calcifications should be considered when recalling a patient from mammography screening in the context of practice specific positive predictive values and cancer detection rates.
大多数乳腺筛查钼靶检查召回病例未导致活检建议。本研究的目的是评估乳腺筛查钼靶检查召回的钙化灶是否比其他检查结果更有可能导致活检建议。
进行经机构审查委员会批准的电子病历搜索，以获取2014 - 2015年期间被指定为BI-RADS 0评估的乳腺筛查钼靶检查随机样本。召回的主要原因分为肿块、不对称、局灶性不对称、钙化或结构扭曲。主要结局是诊断性检查后进行的活检。次要结局是活检结果为癌症阳性的比例，即阳性预测值3（PPV3）。使用逻辑回归比较召回原因（钙化灶与其他检查结果）与活检建议比例。
随机数据库抽样产生了402例BI-RADS 0评估的筛查检查，共有449项检查结果。召回原因包括钙化（14.0%，63/449）、肿块（15.8%，71/449）、不对称（50.8%，228/449）、局灶性不对称（14.3%，64/449）和结构扭曲（5.1%，23/449）。总体而言，21.6%的召回病例导致了影像引导下的活检（87/402）。与其他类型的检查结果相比，因钙化灶召回更有可能导致活检（调整后的比值比8.56，95%可信区间4.58至16.0，p < 0.001）。钙化灶与非钙化灶检查结果的PPV3比例之间未发现统计学上的显著差异（p = 0.812）。
与其他检查结果相比，因钙化灶召回的病例进行活检的可能性要大得多。在根据特定实践的阳性预测值和癌症检出率进行乳腺钼靶筛查召回患者时，应考虑提高钙化灶的活检率。
