Laura Liberman

Beth Israel Deaconess Medical Center, Boston, MA, United States

Are you Laura Liberman?

Claim your profile

Publications (153)540.71 Total impact

  • [Show abstract] [Hide abstract]
    ABSTRACT: To determine the frequency of cancelled stereotactic biopsy due to non-visualisation of calcifications, and assess associated features and outcome data. A retrospective review was performed on 1,874 patients scheduled for stereotactic-guided breast biopsy from 2009 to 2011. Medical records and imaging studies were reviewed. Of 1,874 stereotactic biopsies, 76 (4 %) were cancelled because of non-visualisation of calcifications. Prompt histological confirmation was obtained in 42/76 (55 %). In 28/76 (37 %) follow-up mammography was performed, and 7/28 subsequently underwent biopsy. Of 27 without biopsy, 21 (78 %) had follow-up. Nine cancers (9/49, 18 %) were found: 6 ductal carcinoma in situ (DCIS), 3 infiltrating ductal carcinoma (IDC). Of 54 patients with either biopsy or at least 2 years' follow-up, 9 (17 %) had cancer (95 % CI 8-29). Cancer was present in 7/42 (17 %, 95 % CI 7-31 %) lesions that had prompt histological confirmation (DCIS = 5, IDC = 2) and in 2/28 (7 %, 95 % CI 0.8-24 %) lesions referred for follow-up (DCIS = 1, IDC = 1). Neither calcification morphology (P = 0.2), patient age (P = 0.7), breast density (P = 1.0), personal history (P = 1.0) nor family history of breast cancer (P = 0.5) had a significant association with cancer. Calcifications not visualised on the stereotactic unit are not definitely benign and require surgical biopsy or follow-up. No patient or morphological features were predictive of cancer. • Half of cancelled stereotactic biopsies were due to non-visualisation of calcified foci. • This reflects the improved detection of calcifications by digital mammography. • Calcifications too faint for the stereotactic technique require alternative biopsy or follow-up • 17 % of patients with biopsy or at least 2 years' follow-up had cancer. • No patient/morphological features were found to aid selection for re-biopsy vs. follow-up.
    European Radiology 11/2013; · 4.34 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Focal extravasated mucin (EM) with benign or atypical epithelium is a rare finding at breast core needle biopsy (CNB) and usually prompts surgical excision to rule out mucin-producing carcinoma. In the largest detailed series to date, we assessed surgical outcomes in lesions yielding EM with atypical or nonatypical epithelium at CNB. With IRB approval, we retrospectively reviewed 28 consecutive atypical and nonatypical CNBs with EM that underwent surgical excision at our center over a 22-year period. CNB imaging and pathologic findings were concordant if pathology sufficiently explained the radiologic features of the lesions. Pathologic findings in CNB and excision specimens were correlated. Statistical analysis was performed. CNBs sampled mammographic calcifications in 25/28 (89%) women and a mass in 3/28 (11%). All cases had concordant pathologic and imaging findings. At CNB, the epithelium associated with EM was atypical in 18/28 (64%) lesions and nonatypical in 10 (36%). Cancer (one mucinous carcinoma; three ductal carcinoma in situ) was present in 4/28 excision specimens (14%; 95% confidence intervals [CI], 4%-33%). All carcinomas were in lesions with epithelial atypia at CNB (4/18; 22%; 95% CI, 6%-48%) versus none (0/10; 0%; 95% CI, 0%-31%) in nonatypical lesions at CNB; this difference was not statistically significant (p = 0.3). Surgery is warranted for lesions yielding EM with atypia at CNB due to the high (22%) prevalence of cancer. Our data suggest that surgical excision of lesions yielding EM without epithelial atypia at CNB may not be necessary provided that imaging and pathologic findings are concordant.
    The Breast Journal 03/2013; · 1.83 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: BACKGROUND: No consensus exists on the need to excise breast lesions that yield classic lobular carcinoma in situ (LCIS) or atypical lobular hyperplasia (ALH) (known together as classic lobular neoplasia [LN]) as the highest risk lesion at percutaneous core-needle biopsy (CNB). Here, the authors report findings from 72 consecutive lesions with LN at CNB and prospective surgical excision (EXB). METHODS: Lesions that yielded LN at CNB at the authors' center have been referred for EXB since June 2004, regardless of imaging-histologic concordance. A lesion was "concordant" if histologic findings provided sufficient explanation for imaging. An upgrade consisted of ductal carcinoma in situ and/or invasive carcinoma at EXB. Statistical analysis, including 95% confidence intervals (CIs), was performed. RESULTS: Between June 2004 and May 2009, CNB of 85 consecutive lesions yielded LN without other high-risk histologies. Eighty of 85 lesions (94%) underwent prospective EXB. Seventy-two of 85 lesions (90%; 42 LCIS, 30 ALH) had concordant imaging-histologic findings. EXB yielded low-grade carcinoma in 2 of 72 cases (3%; 95% CI, 0%-9%). In both patients, stereotactic, 11-gauge, vacuum-assisted biopsy of calcifications yielded calcifications in benign parenchyma and ALH. CNB results were discordant in 8 of 80 lesions (10%; 4 LCIS, 4 ALH), and EXB yielded cancer in 3 of those 8 lesions (38%; 95% CI, 9%-76%). The upgrade rate was significantly higher for discordant lesions versus concordant lesions (38% vs 3%; P < .01). CONCLUSIONS: Prospective excision of LN identified carcinoma in 3% (95% CI, 0%-9%) of concordant cases versus 38% (95% CI, 9%-76%) of discordant cases. The current data provide an unbiased assessment of the upgrade rate of LN diagnosed at CNB. Cancer 2012. © 2012 American Cancer Society.
    Cancer 11/2012; · 5.20 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: The objective of our study was to determine the frequency of cancer at surgery in breast lesions yielding papilloma at MRI-guided 9-gauge vacuum-assisted biopsy (VAB) and to determine whether any features are associated with cancer upgrade. For this study, 1487 MRI-guided vacuum-assisted biopsies performed from January 2004 to March 2011 were reviewed. Lesions yielding papilloma were identified and classified as papilloma with or without atypia. Surgical findings were reviewed to determine the cancer rate. Statistical analysis was performed and 95% CIs were calculated. Papilloma was identified in 75 of the 1487 MRI-guided vacuum-assisted biopsies (5%). These 75 papillomas occurred in 73 women with a median age of 49 years (age range, 27-70 years). Of the 75 papillomas, 25 (33%) had atypia and 50 (67%) did not on core needle biopsy. Subsequent surgery of 67 of the 75 papillomas (89%) yielded ductal carcinoma in situ (DCIS) in four (6%; 95% CI, 2-15%). Surgery yielded DCIS in two of 23 papillomas with atypia (9%; 95% CI, 1-28%) at MRI-guided VAB and in two of 44 papillomas without atypia (5%; 95% CI, 0.4-16%) at MRI-guided VAB; these cancer rates did not differ significantly (p = 0.6). Postmenopausal status (p = 0.04) and histologic size of less than 0.2 cm (p = 0.04) had a significant association with the cancer upgrade rate. Papilloma with or without atypia was found in 5% of patients who underwent MRI-guided VAB during the study period. Surgery revealed cancer in 6%. DCIS was found at surgery in 9% of lesions yielding papilloma with atypia versus 5% of lesions yielding papilloma without atypia. For lesions yielding papilloma with or without atypia at MRI-guided VAB, surgical excision is warranted.
    American Journal of Roentgenology 10/2012; 199(4):W512-9. · 2.90 Impact Factor
  • Radiology 05/2012; 263(2):618. · 6.34 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: The objective of this study was to evaluate the impact of tamoxifen treatment on amount of fibroglandular tissue (FGT), background parenchymal enhancement (BPE), and cysts on breast MRI. Retrospective search identified 96 women with breast cancer who had a breast MRI both before and during adjuvant tamoxifen therapy between 2002 and 2008. After exclusion of all irradiated breasts, 88 women were eligible. Two readers blinded to tamoxifen treatment status independently rated level of BPE, amount of FGT, and cysts using a 4‐point categorical scale: BPE––Minimal, Mild, Moderate, Marked; FGT––Fatty, Scattered, Heterogeneously Dense (HD), Dense; Cysts––Minimal, Mild, Moderate, Marked. A consensus interpretation was reached in cases of disagreement. During tamoxifen, there was a significant shift from higher to lower degree BPE, cysts, and FGT compared with before tamoxifen. BPE, cysts and FGT decreased in 68% (60/88), 38% (33/88), and 40% (35/88) of women during tamoxifen (p Document Type: Research Article DOI: http://dx.doi.org/10.1111/tbj.12002 Affiliations: 1: Department of Radiology, Memorial Sloan-Kettering Cancer Center, New York, NY; 2: Department of Epidemiology and Biostatisitics, Memorial Sloan-Kettering Cancer Center, New York, NY Publication date: November 1, 2012 $(document).ready(function() { var shortdescription = $(".originaldescription").text().replace(/\\&/g, '&').replace(/\\, '<').replace(/\\>/g, '>').replace(/\\t/g, ' ').replace(/\\n/g, ''); if (shortdescription.length > 350){ shortdescription = "" + shortdescription.substring(0,250) + "... more"; } $(".descriptionitem").prepend(shortdescription); $(".shortdescription a").click(function() { $(".shortdescription").hide(); $(".originaldescription").slideDown(); return false; }); }); Related content In this: publication By this: publisher In this Subject: Obstetrics & Gynecology , Oncology By this author: King, Valencia ; Kaplan, Jennifer ; Pike, Malcolm C. ; Liberman, Laura ; David Dershaw, D. ; Lee, Carol H. ; Brooks, Jennifer D. ; Morris, Elizabeth A. GA_googleFillSlot("Horizontal_banner_bottom");
    The Breast Journal 01/2012; 18(6). · 1.83 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Guidelines do not support utilization of high technology radiologic imaging (HTRI) for surveillance after curative treatment for early stage breast cancer. Surveillance, Epidemiology, and End Results (SEER)-Medicare linked data were used to identify 25,555 women diagnosed with stage I-II breast cancer between 1998 and 2003 who survived ≥ 48 months from diagnosis without evidence of second primary or recurrent cancer in this interval. HTRI utilization (computerized tomography scanning (CT), bone scan (BS), breast magnetic resonance imaging, and positron emission tomography scans) was measured in months 13-48 post-diagnosis. Cases were individually matched to 75,669 female Medicare enrollees without cancer. Factors associated with HTRI utilization were evaluated. Forty percent of women with stage I-II breast cancer and 25% of controls had ≥ 1 HTRI during the surveillance interval (P < 0.001). High utilization rates were observed for CT (30%) and BSs (19%). The proportion of women who had a CT during the surveillance period increased in both cancer survivors and controls. Among breast cancer cases age <80, higher comorbidity index, stage II disease, and more recent diagnosis were independently associated with receipt of HTRI. Paralleling patterns observed in controls, HTRI utilization for surveillance following diagnosis of early stage breast cancer has steadily increased among Medicare beneficiaries. Strategies to foster judicious utilization of HTRI should be a priority.
    Breast Cancer Research and Treatment 09/2011; 131(2):663-70. · 4.47 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: To determine the rate of canceled magnetic resonance (MR) imaging-guided breast biopsies due to nonvisualization of the lesion and to assess associated features and outcome data for these cases. With the approval of the institutional review board, a HIPAA-compliant retrospective review, in which the requirement for informed consent was waived, was performed for 907 patients scheduled for MR imaging-guided breast biopsy from 2004 to 2008. In 70 patients, MR imaging biopsy was canceled due to lesion nonvisualization. Medical records and imaging studies were reviewed to identify patient, parenchymal, lesion features and outcome data. Statistical analysis was performed with the Fisher exact test. The 95% confidence interval (CI) was calculated. Cancellation of MR-guided biopsy due to lesion nonvisualization occurred in 8% (70 of 907) of patients and in 8% (74 of 911) of lesions. Factors associated with a significantly higher cancellation rate included marked and moderate versus mild and minimal background parenchymal enhancement (38 of 316 [12%] vs 32 of 591 [5%], P = .001), extremely and heterogeneously dense versus scattered fibroglandular densities and fatty parenchymal volume (64 of 712 [9%] vs six of 195 [3%], P = .006), and lesions 1 cm or less in size (52 of 520 [10%] vs 22 of 391 [6%], P = .02).The rate of cancellation per year was highest in the first year, with a decrease in subsequent years (14 of 102 [14%] vs 56 of 805 [7%], P = .025). A significantly lower rate was found in women with synchronous breast cancer (nine of 240 [4%] vs 61 of 667 [9%], P = .007), and a significantly higher rate was found in women with a history of cancer (35 of 315 [11%] vs 35 of 592 [6%], P = .01). Among 58 women who had MR imaging follow-up, no cancers were identified. Among three women who underwent mastectomy after cancellation, one had ductal carcinoma in situ in the same quadrant as the MR-depicted lesion. The cancer detection rate among 61 women who underwent either MR imaging or pathologic follow-up was 2% (one of 61) (95% CI: 0.4%, 9%). MR imaging-guided breast biopsy was canceled due to lesion nonvisualization in 8% of the patients. Although the cancer detection rate among the lesions for which biopsy was canceled is low (95% CI: 0%, 9%), short-term follow-up MR imaging is prudent.
    Radiology 08/2011; 261(1):92-9. · 6.34 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: To compare total choline concentrations ([Cho]) and water-to-fat (W/F) ratios of subtypes of malignant lesions, benign lesions, and normal breast parenchyma and determine their usefulness in breast cancer diagnosis. Reference standard was histology. In this HIPPA compliant study, proton MRS was performed on 93 patients with suspicious lesions (>1 cm) who underwent MRI-guided interventional procedures, and on 27 prospectively accrued women enrolled for screening MRI. (W/F) and [Cho] values were calculated using MRS data. Among 88 MRS-evaluable histologically-confirmed lesions, 40 invasive ductal carcinoma (IDC); 10 invasive lobular carcinoma (ILC); 4 ductal carcinoma in situ (DCIS); 3 invasive mammary carcinoma (IMC); 31 benign. No significant difference observed in (W/F) between benign lesions and normal breast tissue. The area under curve (AUC) of receiver operating characteristic (ROC) curves for discriminating the malignant group from the benign group were 0.97, 0.72, and 0.99 using [Cho], (W/F) and their combination as biomarkers, respectively. (W/F) performs significantly (P < 0.0001;AUC = 0.96) better than [Cho] (AUC = 0.52) in differentiating IDC and ILC lesions. Although [Cho] and (W/F) are good biomarkers for differentiating malignancy, [Cho] is a better marker. Combining both can further improve diagnostic accuracy. IDC and ILC lesions have similar [Cho] levels but are discriminated using (W/F) values.
    Journal of Magnetic Resonance Imaging 04/2011; 33(4):855-63. · 2.57 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Background parenchymal enhancement on breast MRI refers to normal enhancement of the patient's fibroglandular tissue. The aim of this study was to determine the effect of background parenchymal enhancement on short-interval follow-up, biopsy, and cancer detection rate on baseline screening MRI in a high-risk group. Two hundred fifty baseline high-risk screening MRI examinations were reviewed. For each, the background parenchymal enhancement pattern was recorded (minimal, ≤ 25%; mild, 26-50%; moderate, 51-75%; and marked, > 75%), as were BI-RADS category, biopsy rate, and final pathology result. Results were compared for each enhancement category. Of the 250 MRI examinations, 24.8% showed minimal enhancement; 34%, mild; 24%, moderate; and 17.2%, marked enhancement. Women with minimal enhancement had a significantly higher number of BI-RADS categories 1 and 2 examinations (64.5%) than women with mild (38.8%), moderate (40%), or marked (25.6%) enhancement. The BI-RADS category 3 rate was 43.6% overall and was significantly lower for women with minimal enhancement (27.4% vs 47.1% for women with mild, 45.0% for women with moderate, and 58.1% for women with marked enhancement). At follow-up, 86.2% of the BI-RADS 3 lesions were converted to BI-RADS category 1 or 2 and 13.8% were converted to BI-RADS 4, with a malignancy rate of 0.9% for lesions undergoing short-interval follow-up. There was no significant difference in biopsy rate or cancer detection rate among enhancement categories. Mild, moderate, and marked background parenchymal enhancement is associated with a significantly lower rate of BI-RADS categories 1 and 2 assessments and a significantly higher rate of BI-RADS category 3 assessments than minimal enhancement. There was no significant difference in biopsy rate or cancer detection rate among the enhancement categories.
    American Journal of Roentgenology 01/2011; 196(1):218-24. · 2.90 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: PURPOSE To determine impact of preoperative staging breast MRI on outcomes of early breast cancer patients (AJCC stage 0, I and II) undergoing breast conservative therapy (BCT) and radiation therapy (RT). METHOD AND MATERIALS With IRB approval, 1224 staging breast MRI exam records from 2000-2004 were retrospectively reviewed. Among these, 174 staging MRI exams were performed in women with AJCC 0, I or II cancer diagnosed by percutaneous biopsy undergoing BCT and RT. A control group of 174 women who had BCT & RT but no MRI were identified from 3645 patients (same time period), matched by age, histopathology, staging and surgeon. Median follow-up after treatment was 5.5 (range 1.8-9.4) years. For both groups the following were compared: final margin, lymph node status, lymphovascular involvement, extensive intraductal component (EIC) status, hormone receptor status, adjuvant therapy, breast density, and whether cancer was mammographically occult. Outcomes of ipsilateral recurrence, contralateral new cancer and distant metastasis after treatment were compared using Kaplan-Meier methods and the log-rank test. RESULTS Women who underwent preoperative breast MRI, compared to those who did not, were significantly more likely to have extremely dense breasts (28% vs 6%, P<0.0001) and mammographically occult cancer (24% vs 9%, p<0.0003). The two groups had identical rates of negative final margins, positive lymph nodes, lymphovascular involvement, EIC status, hormone receptor positivity and systemic adjuvant therapy. In patients who were staged with MRI there were significantly fewer total operations following initial diagnosis (227 vs 252, p=0.04). For women undergoing preoperative MRI compared to those who did not, no significant difference in actuarial local/regional ipsilateral recurrence (3.6% vs.4.9%, p=0.4), contralateral cancer rate (2.0% vs 3.2%, p=0.6), or distant metastasis rate (2.3% vs 5.3%, p=0.8) was observed. CONCLUSION Preoperative staging MRI decreased total number of surgical procedures in women with early stage breast cancer undergoing BCT and RT. Preoperative staging MRI did not decrease ipsilateral recurrence, contralateral cancer or distant metastasis rates. CLINICAL RELEVANCE/APPLICATION Preoperative MRI appears to be used more in extremely dense breasts and mammographically occult cancer, therefore retrospective outcome analysis is difficult. A prospective randomized trial is needed.
    Radiological Society of North America 2010 Scientific Assembly and Annual Meeting; 11/2010
  • [Show abstract] [Hide abstract]
    ABSTRACT: The purpose of this article is to determine the cancer detection and biopsy rate among women who have breast MRI screening solely on the basis of a personal history of breast cancer. This retrospective review of 1,699 breast MRI examinations performed from 1999 to 2001 yielded 144 women with prior breast cancer but no family history who commenced breast MRI screening during that time. Minimal breast cancer was defined as ductal carcinoma in situ (DCIS) or node-negative invasive breast cancer < 1 cm in size. Of 144 women, 44 (31% [95% CI, 15-29%]) underwent biopsies prompted by MRI examination. Biopsies revealed malignancies in 17 women (12% [95% CI, 7-18%]) and benign findings only in 27 women (19% [95% CI, 13-26%]). Of the 17 women in whom cancer was detected, seven also had benign biopsy results. In total, 18 malignancies were found. One woman had two metachronous cancers. MRI screening resulted in a total of 61 biopsies, with a positive predictive value (PPV) of 39% (95% CI, 27-53%). The malignancies found included 17 carcinomas and one myxoid liposarcoma. Of the 17 cancers, 12 (71%) were invasive, five (29%) were DCIS, and 10 (59%) were minimal breast cancers. Of 17 cancers, 10 were detected by MRI only. The 10 cancers detected by MRI only, versus seven cancers later found by other means, were more likely to be DCIS (4/10 [40%] vs 1/7 [14%]; p = 0.25) or minimal breast cancers (7/10 [70%] vs 3/7 [43%]; p = 0.26). We found that breast MRI screening of women with only a personal history of breast cancer was clinically valuable finding malignancies in 12%, with a reasonable biopsy rate (PPV, 39%).
    American Journal of Roentgenology 08/2010; 195(2):510-6. · 2.90 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The objective of our study was to assess the experiences and preferences of radiology residents with respect to breast imaging. We surveyed radiology residents at 26 programs in New York and New Jersey. Survey topics included plans for subspecialty training, beliefs, and attitudes toward breast imaging and breast cancer screening and the likelihood of interpreting mammography in the future. Three hundred forty-four residents completed the survey (response rate, 62%). The length of time spent training in breast imaging varied from no dedicated time (37%) to 1-8 weeks (40%) to more than 9 weeks (23%). Most respondents (97%) agreed that mammography is important to women's health. More than 85% of residents believed that mammography should be interpreted by breast imaging specialists. Respondents shared negative views about mammography, agreeing with statements that the field was associated with a high risk of malpractice (99%), stress (94%), and low reimbursement (68%). Respondents endorsed several positive attributes of mammography, including job availability (97%), flexible work schedules (94%), and few calls or emergencies (93%). Most radiology residents (93%) said that they were likely to pursue subspecialty training, and 7% expressed interest in breast imaging fellowships. Radiology residents' negative and positive views about mammography seem to be independent of time spent training in mammography and of future plans to pursue fellowship training in breast imaging. Systematic assessment of the plans and preferences of radiology residents can facilitate the development of strategies to attract trainees to careers in breast imaging.
    American Journal of Roentgenology 06/2010; 194(6):1680-6. · 2.90 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Screening mammography rates vary geographically and have recently declined. Inadequate mammography resources in some areas may impair access to this technology. We assessed the relationship between availability of mammography machines and the use of screening. The location and number of all mammography machines in the United States were identified from US Food and Drug Administration records of certified facilities. Inadequate capacity was defined as <1.2 mammography machines per 10,000 women age 40 or older, the threshold required to meet the Healthy People 2010 target screening rate. The impact of capacity on utilization was evaluated in 2 cohorts: female respondents age 40 or older to the 2006 Behavioral Risk Factor Surveillance System survey (BRFSS) and a 5% nationwide sample of female Medicare beneficiaries age 65 or older in 2004-2005. About 9% of women in the BRFSS cohort and 13% of women in the Medicare cohort lived in counties with <1.2 mammography machines per 10,000 women age 40 or older. In both cohorts, residence in a county with inadequate mammography capacity was associated with lower odds of a recent mammogram (adjusted odds ratio in BRFSS: 0.89, 95% CI: 0.80-0.98, P < 0.05; adjusted odds ratio in Medicare: 0.86, 95% CI: 0.85-0.87, P < 0.05), controlling for demographic and health care characteristics. In counties with few or no mammography machines, limited availability of imaging resources may be a barrier to screening. Efforts to increase the number of machines in low-capacity areas may improve mammography rates and reduce geographic disparities in breast cancer screening.
    Medical care 02/2010; 48(4):349-56. · 3.24 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: PURPOSE To identify lesion factors that may be associated with the use and pathologic outcome of bracketing wires at MRI-guided localization and surgical excision of breast cancer. METHOD AND MATERIALS Retrospective review was performed of records of 263 consecutive malignant lesions that had pre-operative MRI-guided needle localization. We identified lesions in which bracketing wires were placed pre-operatively. MRI findings, wire number and location, cancer histology, and pathologic margin status were recorded. Statistical analysis was performed with the Fisher Exact Test. RESULTS Among 263 cancers that had MRI-guided localization, one wire was used in 173 (66%) and bracketing wires were used in 90 (34%). The number of bracketing wires used was two in 52 (58%) of 90 lesions, three in 28 (31%), and four in ten (11%). Of 90 bracketed lesions, maximal diameter was <1 cm in 4 (4%), 1-<2 cm in 13 (14%), 2-<3 cm in 19 (21%), 3-<4 cm in 21 (23%), and >4 cm in 33 (37%). The median MRI lesion size was 3.7 (range, 0.6-9.0) cm for bracketed vs. 1.7 (range, 0.4-6.0) cm for non-bracketed lesions. Bracketed lesions (vs. those that were not bracketed) were significantly more likely to be >2 cm (73/90=81% vs. 48/173=28%, p<0.001), proven cancers (75/90=83% vs. 83/173=48%, p<0.001), and non-masses (75/90=83% vs. 81/173=47%, p<0.001). The likelihood of positive margins did not differ significantly between bracketed and non-bracketed lesions (65/90=72% vs. 126/173=73%, p=0.97). Among 90 bracketed lesions, the positive margin rate was significantly higher if the MRI target was >1 cm (65/86=74% vs. 0/4=0%, p<0.01) or if the lesion had an extensive intraductal component (20/22=91% vs. 45/68=66%, p<0.05). For bracketed lesions, the location of wires, presurgical histology, nodal status, and MRI lesion type had no significant impact on margin status. CONCLUSION Bracketing wires were placed at MRI-guided needle localization in approximately one-third of cancers, usually for larger lesions with non-mass components that were pre-operatively proven to be malignant. Even among cancers that had bracketing wires placed at MRI-guided needle localization, more than two thirds had positive margins. CLINICAL RELEVANCE/APPLICATION Placement of bracketing wires at MRI-guided needle localization to delineate the borders of the MRI target usually does not yield clear histologic margins of resection.
    Radiological Society of North America 2009 Scientific Assembly and Annual Meeting; 12/2009
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Breast tumor volume measured on MRI has been used to assess response to neoadjuvant chemotherapy. However, accurate and reproducible delineation of breast lesions can be challenging, since the lesions may have complicated topological structures and heterogeneous intensity distributions. In this article, the authors present an advanced computerized method to semiautomatically segment tumor volumes on T1-weighted, contrast-enhanced breast MRI. The method starts with manual selection of a region of interest (ROI) that contains the lesion to be segmented in a single image, followed by automated separation of the lesion volume from its surrounding breast parenchyma by using a unique combination of the image processing techniques including Gaussian mixture modeling and a marker-controlled watershed transform. Explicitly, the Gaussian mixture modeling is applied to an intensity histogram of the pixels inside the ROI to distinguish the tumor class from other tissues. Based on the ROI and the intensity distribution of the tumor, internal and external markers are determined and the tumor contour is delineated using the marker-controlled watershed transform. To obtain the tumor volume, the segmented tumor in one slice is propagated to the adjacent slice to form an ROI in that slice. The marker-controlled watershed segmentation is then used again to obtain a tumor contour in the propagated slice. This procedure is terminated when there is no lesion in an adjacent slice. To reduce measurement variations possibly caused by the manual selection of the ROI, the segmentation result is refined based on an automatically determined ROI based on the segmented volume. The algorithm was applied to 13 patients with breast cancer, prospectively accrued prior to beginning neoadjuvant chemotherapy. Each patient had two MRI scans, a baseline MRI examination prior to commencing neoadjuvant chemotherapy and a 1 week follow-up after receiving the first dose of neoadjuvant chemotherapy. Blinded to the computer segmentation results, two experienced radiologists manually delineated all tumors independently. The computer results were then compared with the manually generated results using the volume overlap ratio, defined as the intersection of the computer- and radiologist-generated tumor volumes divided by the union of the two. The algorithm reached overall overlap ratios of 62.6% +/- 9.1% and 61.0% +/- 11.3% in comparison to the two manual segmentation results, respectively. The overall overlap ratio between the two radiologists' manual segmentations was 64.3% +/- 10.4%. Preliminary results suggest that the proposed algorithm is a promising method for assisting in tumor volume measurement in contrast-enhanced breast MRI.
    Medical Physics 10/2009; 36(10):4359-69. · 2.91 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: OBJECTIVE: The purpose of this study was to determine the characteristics of fellowship programs offering breast imaging training and their success in filling positions. CONCLUSION: A comprehensive list of fellowship training opportunities in breast imaging may help potential applicants identify relevant programs.
    American Journal of Roentgenology 03/2009; 192(2):403-7. · 2.90 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: 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. 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. 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). 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.
    American Journal of Roentgenology 10/2008; 191(4):1198-202. · 2.90 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: To retrospectively determine the false-negative rate and the underestimation rate of ultrasonography (US)-guided 14-gauge core-needle breast biopsy (CNB) in nonpalpable lesions, with validation at surgical excision histologic examination and with stability during clinical and imaging follow-up. Informed consent was waived by the institutional review board for this retrospective review of 1352 cases. In 1061 cases, patients underwent surgical excision of lesions visible at US subsequent to US-guided 14-gauge CNB. Follow-up of another 291 benign lesions at US-guided 14-gauge CNB histologic examination showed stability during clinical and imaging follow-up for at least 2 years. US and histologic findings were reviewed and compared for agreement. A false-negative finding was defined as pathologically proved cancer for which biopsy results were benign. The false-negative rate was defined as the proportion of all breast cancers with a diagnosis of benign disease at US-guided 14-gauge CNB. The underestimation rate was defined as an upgrade of a high-risk lesion at US-guided 14-gauge CNB to malignancy at surgery. US 14-gauge CNB yielded 671 (63.2%) malignant, 86 (8.1%) high-risk, and 304 (28.7%) benign lesions. Each of the 291 benign lesions without surgery remained stable during follow-up. The agreement of US-guided 14-gauge CNB results, surgical excision findings, and follow-up results was 95.8% (kappa = 0.93). False-negative findings were encountered in 11 (0.8%) of 1352 cases, and the false-negative rate was 1.6% (11 of 671 malignancies). All false-negative findings were prospectively identified owing to discordance between imaging results and US-guided 14-gauge CNB histologic findings. The underestimation rate was 31.4%. US-guided 14-gauge CNB is an alternative to surgical excision for assessing nonpalpable breast lesions.
    Radiology 08/2008; 248(2):406-13. · 6.34 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Sentinel lymph node biopsy (SLNB) in the setting of prophylactic mastectomy (PM) remains controversial. In the current study, recent experience with PM was described and the value of preoperative magnetic resonance imaging (MRI) was analyzed in selecting patients for PM with or without SLNB. Between January 1999 and January 2006, 529 patients underwent 613 PMs. Both preoperative magnetic resonance imaging (MRI) and SLNB were performed selectively at the discretion of the surgeon. Occult cancer was identified in 33 of 613 PMs (5%) (10 invasive and 23 ductal carcinoma in situ cases). PM with SLNB was performed in 393 of 529 patients (74%), 178 of whom underwent MRI. Of these, occult cancer was found in 6 of 178 patients (3%), all of whom had negative SLNB. Preoperative MRI was concordant with PM in 4 of 6 cases with occult carcinoma. The remaining 215 of 393 patients (55%) underwent PM with SLNB without MRI. Occult cancer was found in 18 of 215 patients (8%); 3 had positive SLNB. Overall, PM with SLNB spared 4 of 393 patients (1%) from axillary lymph node dissection (ALND). Among 136 patients undergoing PM alone, 57 had preoperative MRI. MRI detected 5 cancers and PM revealed an additional 4 occult carcinomas not detected by MRI. Overall, 9 of 136 patients (7%) undergoing PM alone were found to have occult cancer, 3 of which were invasive, raising the decision of reoperation with ALND. Occult cancer was identified in 5% of PMs. PM with or without SLNB spared only 4 of 393 patients (1%) from undergoing ALND, whereas PM alone identified unsuspected invasive disease in 3 of 136 patients (2%). When performed, MRI accurately ruled out the presence of an invasive cancer in the prophylactic breast, suggesting that MRI can be used to select patients for PM without SLNB.
    Cancer 04/2008; 112(6):1214-21. · 5.20 Impact Factor

Publication Stats

5k Citations
540.71 Total Impact Points

Institutions

  • 2012
    • Beth Israel Deaconess Medical Center
      • Department of Radiology
      Boston, MA, United States
  • 1993–2011
    • Memorial Sloan-Kettering Cancer Center
      • • Department of Radiology
      • • Breast Service
      • • Department of Pathology
      New York City, NY, United States
  • 2003
    • Medical University of Vienna
      • Universitätsklinik für Radiodiagnostik
      Vienna, Vienna, Austria