Automated Breast Density Measurements From Chest Computed Tomography Scans.

To develop an automated method for quantifying percent breast density from chest computed tomography (CT) scans. A naïve Bayesian classifier based on gray-level intensities and spatial relationships was developed on CT scans from 10 patients diagnosed with Hodgkin lymphoma (HL) and imaged as part of routine clinical care. The algorithm was validated on CT scans from 75 additional HL patients. The classifier was developed and validated using a reference dataset with consensus manual segmentation of fibroglandular tissue. Accuracy was evaluated at the pixel-level to examine how well the algorithm identified pixels with fibroglandular tissue using true and false positive fractions (TPF and FPF, respectively). Quantitative estimates of the patient-level CT percent density were contrasted to each other using the concordance correlation coefficient, ρc, and to subjective ACR BI-RADS density assessments using Kendall's τb. The pixel-level TPF for identifying pixels with fibroglandular tissue was 82.7% (interquartile range of patient-specific TPFs 65.5%-89.6%). The pixel-level FPF was 9.2% (interquartile range of patient-specific FPFs 2.5%-45.3%). Patient-level agreement of the algorithm's automated density estimate with that obtained from the reference dataset was high, ρc = 0.93 (95% CI 0.90-0.96) as was agreement with a radiologist's subjective ACR-BI-RADS assessments, τb = 0.77. It is possible to obtain automated measurements of percent density from clinical CT scans.

Comparison of Background Parenchymal Enhancement at Contrast-enhanced Spectral Mammography and Breast MR Imaging.

Purpose To assess the extent of background parenchymal enhancement (BPE) at contrast material-enhanced (CE) spectral mammography and breast magnetic resonance (MR) imaging, to evaluate interreader agreement in BPE assessment, and to examine the relationships between clinical factors and BPE. Materials and Methods This was a retrospective, institutional review board-approved, HIPAA-compliant study. Two hundred seventy-eight women from 25 to 76 years of age with increased breast cancer risk who underwent CE spectral mammography and MR imaging for screening or staging from 2010 through 2014 were included. Three readers independently rated BPE on CE spectral mammographic and MR images with the ordinal scale: minimal, mild, moderate, or marked. To assess pairwise agreement between BPE levels on CE spectral mammographic and MR images and among readers, weighted κ coefficients with quadratic weights were calculated. For overall agreement, mean κ values and bootstrapped 95% confidence intervals were calculated. The univariate and multivariate associations between BPE and clinical factors were examined by using generalized estimating equations separately for CE spectral mammography and MR imaging. Results Most women had minimal or mild BPE at both CE spectral mammography (68%-76%) and MR imaging (69%-76%). Between CE spectral mammography and MR imaging, the intrareader agreement ranged from moderate to substantial (κ = 0.55-0.67). Overall agreement on BPE levels between CE spectral mammography and MR imaging and among readers was substantial (κ = 0.66; 95% confidence interval: 0.61, 0.70). With both modalities, BPE demonstrated significant association with menopausal status, prior breast radiation therapy, hormonal treatment, breast density on CE spectral mammographic images, and amount of fibroglandular tissue on MR images (P < .001 for all). Conclusion There was substantial agreement between readers for BPE detected on CE spectral mammographic and MR images. © RSNA, 2016.

Radial Scar at Image-guided Needle Biopsy: Is Excision Necessary?

Optimal management of a lesion yielding radial scar (RS) without epithelial atypia on breast biopsy is controversial. In this single-institution study spanning 17 years, 53 patients with this biopsy diagnosis were evaluated in terms of clinical, radiologic, and pathologic features and outcomes. RSs were categorized as either "incidental" or as the "targeted" lesion according to defined criteria. Of 48 patients who underwent surgical excision after a diagnosis of RS on biopsy, only 1 had an "upgrade" diagnosis of malignancy (2%). No "incidental" RS was associated with the presence of malignancy on surgical excision. Meta-analysis of 20 RS excision studies demonstrated an overall upgrade rate of 10.4%, with a higher rate in patients with a diagnosis of RS with atypia (26%). The upgrade rate for RS without atypia was 7.5% overall. The lower rate of upgrade to malignancy in this study (2%) is likely related to the thorough radiologic-pathologic review undertaken. In the setting of multidisciplinary agreement and careful radiologic-pathologic correlation, it may be appropriate for patients with a biopsy diagnosis of RS without atypia to forego surgical excision in favor of imaging follow-up.

Intravoxel incoherent motion diffusion-weighted MRI at 3.0 T differentiates malignant breast lesions from benign lesions and breast parenchyma.

PURPOSE: To study the differentiation of malignant breast lesions from benign lesions and fibroglandular tissue (FGT) using apparent diffusion coefficient (ADC) and intravoxel incoherent motion (IVIM) parameters. MATERIALS AND METHODS: This retrospective study included 26 malignant and 14 benign breast lesions in 35 patients who underwent diffusion-weighted MRI at 3.0T and nine b-values (0-1000 s/mm(2) ). ADC and IVIM parameters (perfusion fraction fp , pseudodiffusion coefficient Dp , and true diffusion coefficient Dd ) were determined in lesions and FGT. For comparison, IVIM was also measured in 16 high-risk normal patients. A predictive model was constructed using linear discriminant analysis. Lesion discrimination based on ADC and IVIM parameters was assessed using receiver operating characteristic (ROC) and area under the ROC curve (AUC). RESULTS: In FGT of normal subjects, fp was 1.1 ± 1.1%. In malignant lesions, fp (6.4 ± 3.1%) was significantly higher than in benign lesions (3.1 ± 3.3%, P = 0.0025) or FGT (1.5 ± 1.2%, P < 0.001), and Dd ((1.29 ± 0.28) × 10(-3) mm(2) /s) was lower than in benign lesions ((1.56 ± 0.28) × 10(-3) mm(2) /s, P = 0.011) or FGT ((1.86 ± 0.34) × 10(-3) mm(2) /s, P < 0.001). A combination of Dd and fp provided higher AUC for discrimination between malignant and benign lesions (0.84) or FGT (0.97) than ADC (0.72 and 0.86, respectively). CONCLUSION: The IVIM parameters provide accurate identification of malignant lesions.

Impact of menopausal status on background parenchymal enhancement and fibroglandular tissue on breast MRI.

OBJECTIVE: To evaluate the effect of menopausal status on the background parenchymal enhancement (BPE) and amount of fibroglandular tissue (FGT) on breast MRI. METHODS: Retrospective review identified 1,130 women who underwent screening breast MRI between July and November 2010. In 28 of these women, breast MRI was performed both at one time point while pre- and one time point while post-menopausal (median interval 49 months). Two independent readers blinded to menopausal status used categorical scales to rate BPE (minimal/mild/moderate/marked) and FGT (fatty/scattered/heterogeneously dense/dense). Consensus was reached when there was disagreement. The sign test was used to assess changes in rating categories, and the Spearman rank and Fisher's exact tests were used to measure correlations and associations between variables. RESULTS: Significant proportions of women demonstrated decreases in BPE and FGT on post-menopausal breast MRI (P = 0.0001 and P = 0.0009). BPE category was unchanged in 39 % (11/28) and decreased in 61 % (17/28) of women. FGT category was unchanged in 61 % (17/28) and decreased in 39 % (11/28) of women. Age, reason for menopause, or interval between MRIs had no significant impact on changes in BPE and FGT. CONCLUSION: On MRI, BPE, and FGT decrease after menopause in significant proportions of women; BPE decreases more than FGT. KEY POINTS: On MRI, background parenchymal enhancement and fibroglandular tissue both decrease after menopause. The reduced postmenopausal enhancement is more marked in parenchyma than fibroglandular tissue. Background enhancement and fibroglandular tissue seen on MRI are under hormonal influence.

Cumulative imaging radiation exposure following breast-conservation therapy.

INTRODUCTION: Radiation from medical imaging may induce cellular damage and increase the risk of cancer. While health care workers are restricted to an annual dose of 50 milliSieverts (mSv), the exposure to patients is not typically recorded. After breast-conservation therapy (BCT), patients are subjected to screening mammography, diagnostic breast imaging, and systemic surveillance imaging (SSI). Our objectives are to determine the cumulative radiation exposure of breast cancer survivors after completion of BCT, and to compare exposure levels in two historical cohorts. We also evaluated the indications of SSI. METHODS: We performed a retrospective study of 68 patients with stage I or II breast cancer who received BCT in 1997 or 2002. Cumulative radiation exposure during follow-up from all imaging attributable to the breast cancer diagnosis was recorded, including both breast and non-breast imaging. The indications for SSI were recorded. RESULTS: In the first 5 years after BCT, patients received a median annual dose of 0.92 mSv with no difference between the 1997 and 2002 cohorts. A median of 90% of radiation exposure was due to mammography. From 1997 to 2002, the percentage of patients receiving computed tomography (CT) scans increased. Additional SSI occurred in 65% of patients, with the majority of tests ordered in the asymptomatic patient. Patients with nodal positivity were more likely to receive SSI (p = 0.03). CONCLUSIONS: In the first 5 years after BCT, annual radiation exposure due to imaging was low. However, it seems prudent to consider the risks of radiation exposure when ordering potentially low-yield screening studies in asymptomatic patients.

Complete excision of the MRI target lesion at MRI-guided vacuum-assisted biopsy of breast cancer.

OBJECTIVE: The purpose of our study was to determine the frequency of complete removal of the imaging target at MRI-guided vacuum-assisted biopsy of breast cancer and to assess the residual cancer rate at surgery in these lesions. MATERIALS AND METHODS: With the approval of our institutional review board, retrospective review was performed of 416 consecutive lesions that underwent MRI-guided 9-gauge vacuum-assisted biopsy, of which 76 (18%) yielded cancer. Medical and pathology records were reviewed. RESULTS: Vacuum-assisted biopsy histology in 76 cancers was ductal carcinoma in situ in 39 (51%) and invasive cancer in 37 (49%). Median MRI lesion size in these 76 cancers was 1.2 cm (range, 0.4-8.0 cm). The median number of samples obtained was 12 (range, 6-24 samples). Among 76 cancers, the MRI target was completely excised in 23 (30% [95% CI, 20-42%]), sampled in 52 (69% [57-79%]), and possibly missed in one (1% [0-7%]). Complete MRI target excision rather than sampling was significantly more likely in lesions < or = 1 cm than in lesions >1 cm (18/34 = 53% vs 5/41 = 12%; p < 0.001). Surgery, performed in 67 of 76 cancers, showed residual cancer in 55 (82% [71-90%]). The residual cancer rate was significantly lower if the MRI target was completely excised rather than sampled (14/22 = 64% vs 40/44 = 91%; p < 0.02). CONCLUSION: Complete excision of the MRI target occurred in 30% of breast cancers diagnosed at MRI-guided vacuum-assisted biopsy. Among cancers in which the MRI target was percutaneously excised, surgery yielded residual cancer in 64%. Complete removal of the MRI target does not ensure complete histologic excision of the cancer.

Imaging histologic discordance at MRI-guided 9-gauge vacuum-assisted breast biopsy.

OBJECTIVE: The purpose of this study was to determine the frequency of discordance at MRI-guided vacuum-assisted biopsy and to assess the cancer rate in discordant lesions. MATERIALS AND METHODS: With institutional review board approval, retrospective review was performed of a database of 342 lesions that had MRI-guided vacuum-assisted biopsy during a 39-month period. Biopsy samples were obtained in a 1.5-T magnet using a 9-gauge MRI-compatible vacuum-assisted biopsy device. Medical and pathology records were reviewed to determine the number of discordant lesions and surgical outcome. Statistical analysis was performed. RESULTS: Among 342 lesions that had MRI-guided vacuum-assisted biopsy, results were discordant in 24 (7% [95% CI, 3-14%]) lesions. The discordance rate was significantly (p < 0.001) higher among MRI target lesions that were possibly missed rather than sampled. A trend was seen (p < 0.06) toward a higher discordance rate in MRI lesions that were sampled rather than excised at MRI-guided vacuum-assisted biopsy. Subsequent surgery in 20 discordant lesions yielded cancer in six (30% [12-54%]), including ductal carcinoma in situ (DCIS) in two and invasive carcinoma in four (three ductal and one lobular, all with DCIS). The cancer rate among discordant lesions was significantly higher in postmenopausal women than in premenopausal women (p <0.05). CONCLUSION: Imaging-histologic discordance was found in 7% of lesions that had MRI-guided vacuum-assisted biopsy. Among discordant lesions, surgical excision revealed cancer in 30%. Imaging-histologic correlation is essential after MRI-guided vacuum-assisted biopsy to avoid delay in the diagnosis of breast cancer.

Underestimation of DCIS at MRI-guided vacuum-assisted breast biopsy.

OBJECTIVE: The study objective was to assess the rate of underestimation of ductal carcinoma in situ (DCIS) at MRI-guided 9-gauge vacuum-assisted breast biopsy. MATERIALS AND METHODS: An institutional review board-approved retrospective review was performed of 373 consecutive lesions that had undergone MRI vacuum-assisted breast biopsy. In 34 lesions with subsequent surgery, vacuum-assisted breast biopsy yielded DCIS without frank microinvasion or invasion. DCIS underestimates were lesions for which vacuum-assisted breast biopsy yielded DCIS without frank microinvasion or invasion at biopsy and surgery yielded invasive cancer. Records and pathology findings were reviewed. RESULTS: Among 34 lesions, vacuum-assisted breast biopsy histology was DCIS in 29 and DCIS with possible microinvasion in five. Of 29 lesions yielding DCIS at MRI vacuum-assisted breast biopsy, surgical excision revealed invasive cancer in five (17%; 95% CI, 6-36%). The DCIS underestimation rate was significantly higher in lesions 6 cm or larger versus smaller lesions (60% vs 8%, p = 0.02). MRI lesion type, kinetics, number of specimens, menopausal status, and target sampling versus excision did not significantly affect underestimation. Of five lesions yielding DCIS with possible microinvasion at MRI vacuum-assisted breast biopsy, surgery revealed invasive carcinoma in four (80%; 95% CI, 28-99%). DCIS underestimation was significantly more likely if MRI vacuum-assisted breast biopsy showed possible microinvasion than if it did not (80% vs 17%, p =0.01). CONCLUSION: Underestimation occurred in 17% of lesions yielding DCIS and in 80% of lesions yielding DCIS with possible microinvasion at MRI vacuum-assisted breast biopsy. DCIS underestimation was significantly more likely in lesions measuring 6 cm or larger. No other patient or lesion factors significantly affected DCIS underestimation at MRI vacuum-assisted breast biopsy.

MR imaging findings in the contralateral breast of women with recently diagnosed breast cancer.

OBJECTIVE: The purpose of this study was to determine the frequency and positive predictive value of biopsy performed on the basis of MR imaging findings in the contralateral breast in women with recently diagnosed breast cancer. MATERIALS AND METHODS: We performed a retrospective review of records of 1336 consecutive breast MR imaging examinations over a 2-year period. Of these examinations, 223 imaged the asymptomatic, mammographically normal contralateral breast in women whose breast cancer was diagnosed within 6 months preceding MR imaging. Records of these 223 examinations were reviewed to determine the frequency of recommending contralateral breast biopsy and the biopsy results. RESULTS: Contralateral breast biopsy was recommended in 72 (32%) of 223 women and performed in 61 women. Cancer occult to mammography and physical examination was detected by MR imaging in 12 women, constituting 20% (12/61) of women who underwent contralateral biopsy and 5% (12/223) of women who underwent contralateral breast MR imaging. Among these 12 cancers, six (50%) were ductal carcinoma in situ (DCIS) and six (50%) were infiltrating carcinoma. The median size of infiltrating carcinoma was 0.5 cm (range, 0.1-1.0 cm). Contralateral biopsy revealed benign (n = 31) or high-risk (n = 18) lesions in 49 women, constituting 80% (49/61) of women who underwent contralateral biopsy and 22% (49/223) of women who underwent contralateral MR imaging. CONCLUSION: In women with recently diagnosed breast cancer, MR imaging of the contralateral breast led to a biopsy recommendation in 32%. Cancer was found in 20% of women who underwent contralateral breast biopsy and in 5% of women who underwent contralateral breast MR imaging.

To excise or to sample the mammographic target: what is the goal of stereotactic 11-gauge vacuum-assisted breast biopsy?

OBJECTIVE: This study was undertaken to determine whether complete percutaneous excision rather than sampling of the mammographic target conveys any significant advantage or disadvantage at stereotactic 11-gauge vacuum-assisted biopsy. MATERIALS AND METHODS: A retrospective review was performed of 788 consecutive solitary lesions in which the mammographic target was excised (n = 466) or sampled (n = 322) at stereotactic 11-gauge vacuum-assisted biopsy. Medical records and histologic findings were reviewed to determine the frequency of sparing surgery, discordance, histologic underestimation, rebiopsy, complete histologic removal of cancer, and complications. Statistical comparisons were made using the Fisher's exact test. RESULTS: Complete excision rather than sampling of the mammographic target was associated with a significantly lower frequency of discordance (1/466, 0.2% vs 8/322, 2.5%; p = 0.004) and a trend toward fewer ductal carcinoma in situ underestimates (4/59, 6.8% vs 12/60, 20.0%; p = 0.07). Complete histologic removal of cancer was significantly more likely if the mammographic target was excised rather than sampled (19/91, 20.9% vs 7/106, 6.6%; p = 0.006); however, among 91 cancers in which the mammographic target was excised, surgery revealed residual cancer in 72 (79.1%). Complete excision rather than sampling of the mammographic target yielded no significant differences in the frequency of sparing surgery, atypical ductal hyperplasia underestimates, rebiopsy, or complications. CONCLUSION: Complete excision rather than sampling of the mammographic target was associated with lower frequencies of discordance and ductal carcinoma in situ underestimation but had no other advantage or disadvantage. Among cancers in which the mammographic target was excised, surgery revealed residual cancer in almost 80%.

Breast lesions detected on MR imaging: features and positive predictive value.

OBJECTIVE: The purpose of this study was to analyze features of breast lesions detected on MR imaging that had subsequent biopsy and to determine the positive predictive value (PPV) of these features. MATERIALS AND METHODS: Retrospective review was performed of 100 consecutive solitary MR imaging-detected breast lesions that had MR imaging-guided needle localization and surgical excision. We described lesions, using terms found in a proposed breast MR imaging lexicon. Histologic findings were reviewed. RESULTS: Carcinoma was identified in 25 lesions (25%), including ductal carcinoma in situ (DCIS) in 13 (52%) and infiltrating carcinoma in 12 (48%). Carcinoma was found in 15 (25%) of 60 masses versus 10 (25%) of 40 nonmass lesions; most malignant masses (73%) were infiltrating carcinoma, whereas most malignant nonmass lesions (90%) were DCIS. The features with the highest PPV were spiculated margin (80% carcinoma), rim enhancement (40% carcinoma), and irregular shape (32% carcinoma) for mass lesions and segmental (67% carcinoma) or clumped linear and ductal enhancement (31% carcinoma) for nonmass lesions. Visually assessed kinetic patterns were not significant predictors of carcinoma, but washout was present in 70% of infiltrating carcinomas versus 9% of DCIS lesions (p < 0.01). Carcinoma was present in 17 (19%) of 88 lesions classified as suspicious versus eight (67%) of 12 lesions classified as highly suggestive of malignancy (p = 0.001). CONCLUSION: Among MR imaging-detected breast lesions referred for biopsy, carcinoma was found in 25%, of which half were DCIS. Features with the highest PPV were spiculated margin, rim enhancement, and irregular shape for mass lesions and segmental or clumped linear and ductal enhancement for nonmass lesions. Final assessment categories were significant predictors of carcinoma.

Preoperative MR imaging-guided needle localization of breast lesions.

OBJECTIVE: MR imaging of the breast can depict cancer that is occult on mammography and at physical examination. Our study was undertaken to determine the ease of performance and the outcome of MR imaging-guided needle localization and surgical excision of breast lesions. MATERIALS AND METHODS: Retrospective review revealed 101 consecutive breast lesions that had preoperative MR imaging-guided needle localization with commercially available equipment, including a 1.5-T magnet with a breast surface coil, a dedicated biopsy compression device, and MR imaging-compatible hookwires. Imaging studies and medical records were reviewed. RESULTS: Histologic findings in these 101 lesions were carcinoma in 31 (30.7%), high-risk lesions (atypical ductal hyperplasia or lobular carcinoma in situ) in nine (8.9%), and benign lesions in 61 (60.4%). Fifteen (48.4%) of 31 carcinomas were ductal carcinoma in situ, and 16 (51.6%) were infiltrating carcinoma (size range, 0.1-2.0 cm; median, 1.2 cm). Carcinoma was found in 16 (45.7%) of 35 lesions detected in women with synchronous cancer, 10 (32.3%) of 31 lesions detected on MR imaging for problem solving, and five (14.3%) of 35 lesions detected on MR screening. The time range to perform MR imaging-guided localization was 15-59 min (median time, 31 min). Complications encountered in three cases were retained wire fragments in two and breakage of the wire tip in one. CONCLUSION: MR imaging-guided needle localization can be performed quickly and safely with commercially available equipment. The positive predictive value of MR imaging-guided needle localization (30.7%) was comparable to that reported for mammographically guided needle localization and was highest in women with synchronous breast cancer.