Minimal Invasive Biopsy Results of ”Uncertain Malignant Potential” in Digital Mammography Screening: High Prevalence but also High Predictive Value for Malignancy

Artificial Intelligence for Indication of Invasive Assessment of Calcifications in Mammography Screening

PURPOSE

Lesion-related evaluation of the diagnostic performance of an individual artificial intelligence (AI) system to assess mamographically detected and histologically proven calcifications.

MATERIALS AND METHODS

This retrospective study included 634 women of one screening unit (July 2012 - June 2018) who completed the invasive assessment of calcifications. For each leasion, the AI-system calculated a score between 0 and 98. Lesions scored > 0 were classified as AI-positive. The performance of the system was evaluated based on its positive predictive value of invasive assessment (PPV3), the false-negative rate and the true-negative rate.

RESULTS

The PPV3 increased across the categories (readers: 4a: 21.2 %, 4b: 57.7 %, 5: 100 %, overall 30.3 %; AI: 4a: 20.8 %, 4b: 57.8 %, 5: 100 %, overall: 30.7 %). The AI system yielded a false-negative rate of 7.2 % (95 %-CI: 4.3 %: 11.4 %) and a true-negative rate of 9.1 % (95 %-CI: 6.6 %; 11.9 %). These rates were highest in category 4a, 12.5 % and 10.4 % retrospectively. The lowest median AI score was observed for benign lesions (61, interquartile range (IQR): 45-74). Invasive cancers yielded the highest median AI score (81, IQR: 64-86). Median AI scores for ductal carcinoma in situ were: 74 (IQR: 63-84) for low grade, 70 (IQR: 52-79) for intermediate grade and 74 (IQR: 66-83) for high grade.

CONCLUSION

At the lowest threshold, the AI system yielded calcification-related PPV3 values that increased across categories, similar as seen in human evaluation. The strongest loss in AI-based breast cancer detection was observed for invasively assessed calcifications with the lowest suspicion of malignancy, yet with a comparable decrease in the false-positive rate. An AI-score based stratification of malignant lesions could not be determined.

KEY POINTS

· The AI-based PPV3 for calcifications is comparable to human assessment.. · AI showed a lower detection performance of screen-positive and screen-negative lesions in category 4a.. · Histological subgroups could not be discriminated by AI scores..

CITATION FORMAT

· Weigel S, Brehl AK, Heindel W et al. Artificial Intelligence for Indication of Invasive Assessment of Calcifications in Mammography Screening. Fortschr Röntgenstr 2023; 195: 38 - 46.

High-risk breast lesions: role of multi-parametric DCE-MRI in detection and histopathological upgrade prediction

Background

High-risk breast lesions represent 3–21% of all breast lesions and are non-obligate precursors of malignancy. Various studies have evaluated the value of DCE-MRI including DWI and ADC mapping in the detection of high-risk breast lesions, differentiating them from malignant lesions and predicting upgrade risk after surgical excision reducing misdiagnosis and overtreatment. This study is a retrospective review of all image-guided breast biopsy procedures performed in 2021 at our institution, identified 68 patients with histopathologically proven high-risk breast and available MRI examinations with no concurrent ipsilateral malignancy. Image analysis of MRI examinations included morphological criteria, enhancement pattern, dynamic curves, DWI and ADC mapping assessment. Since our knowledge of high-risk breast lesions is still growing, this study is important to evaluate the merits of DCE-MRI in the assessment of high-risk breast lesions, to allow optimization of treatment, better limit it to those women at risk, and avoid overtreatment in women at low risk.

Results

The mean ADC value of high-risk breast lesions was not significantly different from that of malignant breast lesions (p value = 0.015). Non-mass enhancement and type III enhancing curve proved to be significant indicators of high-risk breast lesions upgrade in surgical pathology. Cut-off average ADC value for differentiating upgraded high-risk lesions from non-upgraded high-risk lesions proved 1.24 mm2/sec with sensitivity and specificity of 94 & 100%, respectively.

Discussion

Management of high-risk breast lesions is important in the screening setting, as they are non-obligate precursors of malignancy, and also function as risk indicators. Frequency and upgrade rates of high-risk lesions detected by MRI provide a reference for clinical management. DCE-MRI has a high negative predictive value in predicting the upgrade risk of high-risk lesions. In this study, non-mass enhancement and type III curve were proven to be indicators of high-risk lesion upgrade. Limitations of the study included small number of patients and limited follow-up period.

Conclusions

The use of multi-parametric DCE-MRI including DWI and ADC mapping provides complementary information to detect high-risk breast lesions and predict their upgrade rate.

A ten-year, single-center experience: Concordance between breast core needle biopsy/vacuum-assisted biopsy and postoperative histopathology in B3 and B5a cases

Purpose

To determine the concordance rate between core needle biopsy/vacuum-assisted biopsy (CNB/VAB) and postoperative histopathology in B3 (lesions of uncertain malignant potential) and B5a (in situ) lesions found on mammograms or ultrasound.

Material and methods

2,029 consecutive biopsies performed over 10 years for patients who underwent mammograms or ultrasounds. For CNB 14G needle and for VAB 8G/10G needles were used. In all biopsies, we identified the age, BI-RADS®, histopathological biopsy results, B-category, nuclear grade for DCIS and postoperative histopathology results in B3 and B5a cases from the biopsy.

Results

The B-categories from CNB/VAB were as follows: B2 42.2 percent (n = 856), B3 4.5 percent (n = 91), B5a 5.7 percent (n = 115), and B5b 47.6 percent (n = 967). In the B3-category, 72/91 patients underwent surgical excision, with a concordance rate of 83.3 percent (n = 60/72) and a discordance rate of 16.7 percent (n = 12/72) to postoperative histopathology. From the discordant cases, 67.7 percent (n = 8/12) showed DCIS and 32.3 percent (n = 4/12) showed invasive breast cancer. The BIRADS of the discordant cases was 4b in 41.7 percent (n = 5/12) and 5 in 58.3 percent (n = 7/12). The PPVs for malignancy of B3 lesions were 0.21, with no statistical significance between subgroups. In the B5a-category, 101 of 115 patients underwent surgery in our hospital, with a concordance rate of 80.2 percent (n = 81/101) and a discordance rate of 19.8 percent (n = 20/101) to postoperative histopathology. From the discordant cases, 55 percent (n = 11/20) showed invasive breast carcinoma of no special type (NST).

Conclusion

Our concordance rate for B3 (83.3 percent) and B5a (80.2 percent) lesions in the biopsies to postoperative histopathology is matching to previously published literature. Surgical excision is our recommendation for lesions biopsied with a B3 category in the histopathology and a BIRADS category of (4b, 4c and 5). The PPVs for malignancy of B3 lesions showed no statistical significance between subgroups. Also, the nuclear grade of DCIS was not statistically significant in terms of upgrade into invasive breast cancer.

Our Radiological Experience on B3 Lesions: Correlation Between Mammographic and MRI Findings With Histologic Definitive Result

INTRODUCTION

The purpose of this study was to evaluate mammographic and magnetic resonance imaging (MRI) features in B3 lesions.

PATIENTS AND METHODS

From 2011 to 2018, 139 patients with histologically proven B3 lesions who underwent mammography or/and MRI, were retrospectively reviewed. B3 lesions were classified in: atypical ductal hyperplasia (ADH), lobular neoplasia (LN), papillary lesion (PL), radial scar (RS), flat epithelial atypia (FEA), phyllodes tumor (PT), or mesenchymal lesion. Imaging features evaluated were: the presence of microcalcifications, mass and architectural distortions on mammograms and type of margins (circumscribed, irregular, spiculate), enhancement (mass-like, non-mass-like), size (≤ 15 mm, > 15 mm), and kinetics curves (I, II, III) on MRI. The definitive histologic results confirmed benign lesion or were upgraded to malignancy, and the positive predictive value was calculated.

RESULTS

Histologic classification of B3 lesions counted 45 (32.37%) ADH, 12 (8.63%) LN, 25 (17.99%) PL, 5 (3.61%) RS, 31 (22.31%) FEA, 20 (14.39%) PT, and 1 (0.70%) mesenchymal lesion. One hundred seven patients had mammography, and 38 had MRI. In 90 (65%) cases, the histologic diagnosis confirmed B3, in 15 (11%) cases, benign lesion, and in 34 (24%) cases, malignancies were found, with best positive predictive value for mesenchymal tumor (1), ADH (0.36), and FEA (0.4). Significant correlations comparing core needle biopsy groups and microcalcifications (P = .016) and presence of mass (P = .002) and comparing definitive histology with the presence of mass (P = .023), were found. Regarding MRI, the morphology correlated with core needle biopsy groups (P = .038); morphology (P = .024), dimension (P = .040), and kinetic curve (P = .005) correlated with malignancy.

CONCLUSIONS

The B3 category includes different entities, with various risk of malignancy; their heterogeneity is associated with specific mammographic and MRI features, although further confirmations are needed.

Malignancy rates of B3-lesions in breast magnetic resonance imaging - do all lesions have to be excised?

BACKGROUND

Approximately 10% of all MRI-guided vacuum-assisted breast biopsies (MR-VAB) are histologically classified as B3 lesions. In most of these cases surgical excision is recommended. The aim of our study was to evaluate the malignancy rates of different B3 lesions which are visible on MRI to allow a lesion-adapted recommendation of further procedure.

METHODS

Retrospective analysis of 572 consecutive MR-VAB was performed. Inclusion criteria were a representative (=successful) MR-VAB, histologic diagnosis of a B3 lesion and either the existence of a definite histology after surgical excision or proof of stability or regression of the lesion on follow-up MRI. Malignancy rates were evaluated for different histologies of B3 lesions. Lesion size and lesion morphology (mass/non-mass enhancement) on MRI were correlated with malignancy.

RESULTS

Of all MR-VAB 43 lesions fulfilled the inclusion criteria. The malignancy rate of those B3 lesions was 23.3% (10/43). The highest malignancy rate was found in atypical ductal hyperplasia (ADH) lesions (50.0%; 4/8), 33.3% (2/6) in flat epithelial atypia (FEA), 28.6% (2/7) in lobular intraepithelial neoplasia (LIN) and 12.5% (2/16) in papillary lesions (PL). All 6 complex sclerosing lesions were benign. Mass findings were significantly more frequently malignant (31.3%, 10/32; p < 0.05) than non-mass findings (0/11). Small lesions measuring 5-10 mm were most often malignant (35.0%; 7/20). All large lesions (> 20 mm) were not malignant (0/10). Intermediate sized lesions (11-20 mm) turned out to be malignant in 23.1% (3/13).

CONCLUSIONS

The malignancy rate of B3 lesions which were diagnosed after MR-VAB was 23.3%. ADH, FEA and LIN showed considerable malignancy rates (50%, 33% and 29%) and should therefore undergo surgical excision. None of the cases, which were diagnosed as radial scars, non-mass enhancement or larger lesions (> 20 mm) were malignant. Here, a follow-up MRI seems to be advisable to avoid unnecessary operations.

TRIAL REGISTRATION

Retrospective study design, waived by the IRB.

Lesions of "uncertain malignant potential" in the breast (B3) identified with mammography screening

Background

Core needle biopsy (CNB) is a standard diagnostic procedure in the setting of breast cancer screening. However, CNB may result in the borderline diagnoses of lesion of uncertain malignant potential (B3). The aim of this study was to access the outcome of lesions diagnosed as B3 category in a large series of screen-detected cases to evaluate the rates of malignancy for the different histological subtypes.

Methods

We identified all CNBs over a six-year period (2009-2015) in a breast cancer screening unit in Germany. A total of 8.388 CNB’s were performed for screen detected breast lesions. B3 diagnosis comprised 4.5% (376/8.388). Of the 376 patients who were diagnosed as B3, 299 underwent subsequent excision biopsy with final excision histology.

Results

Out of 376 patients diagnosed with B3 lesions, the prevalence of different histopathology showed 161 (42.8%) patients with atypical ductal hyperplasia (ADH), 98 (26.1%) with flat epithelial atypia (FEA), 50 women (13.3%) showed lobular neoplasia (LN), in 40 (10.6%) patients papillary findings and in 27 patients (7.2%) a radial scar complex. Final excision histology was benign in 74% (221/299) and malignant in 26% (78/299) of the patients. Lesion specific positive predictive values (PPV) for a subsequent diagnosis of in situ or invasive carcinoma were as follows: ADH 40%, FEA 20.5%, papillary lesion 13.5%, radial scar 16.6%, LN 0%.

Conclusion

Our results show that approximately one-third of core needle biopsies of screen detected breast lesions classified as B3 are premalignant or malignant on excision.

Lesions of uncertain malignant potential of the breast (B3) are heterogeneous in respect to risk of malignancy.

An Update of Mucinous Lesions of the Breast

Mucinous lesions of the breast include a variety of benign and malignant epithelial processes that display intracytoplasmic or extracellular mucin, including mucocelelike lesions, mucinous carcinoma, solid papillary carcinoma, and other rare subtypes of mucin-producing carcinoma. The most important diagnostic challenge is the finding of free-floating or stromal mucin accumulations for which the significance depends on the clinical, radiologic, and pathologic context. This article emphasizes the differential diagnosis between mucocelelike lesions and mucinous carcinoma, with a brief consideration of potential mimics, such as biphasic and mesenchymal lesions with myxoid stroma ("stromal mucin") and foreign material.

Lesions with unclear malignant potential (B3) after minimally invasive breast biopsy: evaluation of vacuum biopsies performed in Switzerland and recommended further management

Background

Histopathological B3 lesions after minimal invasive breast biopsy (VABB) are a particular challenge for the clinician, as there are currently no binding recommendations regarding the subsequent procedure.

Purpose

To analyze all B3 lesions, diagnosed at VABB and captured in the national central Swiss MIBB database and to provide a data basis for further management in this subgroup of patients.

Material and Methods

All 9,153 stereotactically, sonographically, or magnetic resonance imaging (MRI)-guided vacuum-assisted breast biopsies, performed in Switzerland between 2009 and 2011, captured in a central database, were evaluated. The rate of B3 lesions and the definitive pathological findings in patients who underwent surgical resection were analyzed.

Results

The B3 rate was 17.0% (1532 of 9000 biopsies with B classification). Among the 521 lesions with a definitive postoperative diagnosis, the malignancy rate (invasive carcinoma or DCIS) was 21.5%. In patients with atypical ductal hyperplasia, papillary lesions, flat epithelial atypia, lobular neoplasia, and radial scar diagnosed by VABB, the malignancy rates were 25.9%, 3.1%, 18.3%, 26.4%, and 11.1%, respectively.

Conclusion

B3 lesions, comprising 17%, of all analyzed biopsies, were common and the proportion of malignancies in those lesions undergoing subsequent surgical excision was high (21.5%).

Mucocele-Like Lesions of the Breast: Rate of Upstaging and Cancer Development

BACKGROUND

Breast mucocele-like lesions (MLLs) are rare breast lesions. There is a paucity of literature on the rate of pathologic upstaging of MLL to in situ or invasive breast cancer. We sought to review our experience with MLL at a tertiary care center and review published literature.

METHODS

A retrospective review was undertaken of all patients with breast MLL at the Mayo Clinic, Rochester, from 1998 to 2016. Descriptive statistics were performed.

RESULTS

Overall, 30 patients were diagnosed with MLL, with a mean age of 54.6 years (range 51-58). Calcifications on screening mammogram were the most common method of MLL detection (n = 26, 86.7 %), with core needle biopsy (CNB) being performed in 29/30 patients. MLL with atypia was identified in 14 patients (48 %), and surgical excision was performed in 26 patients. Disease was upstaged to invasive disease in one patient (3.8 %) who presented with a palpable mass. With a median duration of follow up of 3.5 years [interquartile range 1-7.1], two patients developed contralateral invasive breast cancer (both patients had a remote history of invasive breast cancer on that side), and one patient developed contralateral in situ disease. Review of the literature demonstrates a 2.9 % upgrade rate of MLL without atypia on excisional biopsy.

CONCLUSIONS

The upstage rate of MLL identified on CNB to invasive disease at surgical excision in this series was 3.8 %. In cases with MLL presenting as calcifications, no cases were upstaged in this series. Across the literature, upstage rates of MLL without atypia on CNB are low, indicating excisional biopsy may be avoided in these cases.

B3-lesions of the breast and cancer risk - an analysis of mammography screening patients

The use of mammography screening, followed by needle core biopsy (NCB), is associated with an increasing amount of invasive procedures. A considerable amount of specimens must be classified as lesions with uncertain malignant potential (B3-lesion). In these cases, an open biopsy is indicated for further diagnosis. We evaluated patients with B3-lesions to determine the risk of malignancy corresponding to the histopathological NCB results and the type of radiological lesion identified. A total of 95 patients participating in the German mammography screening program with a B3-lesion following NCB (104 B3-lesions in total) were included in our analysis. We analyzed the correlation between the initial histopathological findings from the NCB specimen and cancer risk. We further analyzed the correlations of malignant results with the type of mammographic lesion. In 23 cases (22%), histopathological examination following excision revealed a malignant lesion, including invasive and in situ carcinoma. The positive predictive value of the subgroups of B3-lesions ranged between 0.11 and 0.31; the B3-lesion associated with the highest cancer risk was the atypical ductal hyperplasia; however, no significant difference was observed between the B3-lesion subgroups (P=0.309) regarding the risk of malignancy. Comparing the different types of mammographic findings, such as radiological mass or microcalcifications, there was no significant difference in the risk for malignancy (P=0.379). The different types of B3-lesions did not exhibit differences in the risk for malignancy, and the morphological type of mammographic lesion does not appear to be correlated with cancer risk; therefore, our results underline the need for open biopsy in patients with B3-lesions following NCB.

Use of an Additional Diagnostic Work-up Following a Treatment Recommendation from the Preoperative Conference of the Mammography Screening Units

Objective: If a focus of suspicion is classified as being B 3-5 by a punch biopsy as part of a mammography screening, a recommendation for further action to be taken will be given in the preoperative conference of the screening unit. As part of this investigation, these treatment recommendations were compared with the final therapeutic approach taken at a certified breast centre. Furthermore, it was investigated whether and which additional examinations were performed on patients, depending on compliance with the recommended treatment. Material and Method: The data from 272 breast cancer patients from the years 2007, 2008 and 2009 was analysed. The patients took part in the screening programmes of four screening units in the German mammography screening programme, in one federal state. In addition, the data from each patient from one screening unit was analysed in two further federal states. Results: In total, the most recently conducted intervention deviated from the treatment recommendation from the preoperative conference in the screening unit in 77 out of 272 patients (28.3 %). Of these, there were 50 recommendations for open biopsy which ultimately resulted in breast-conserving surgery, which is not to be evaluated as an error, as the bioptic result was supplemented by the open biopsy. Additional examinations were performed in patients with deviating treatment recommendation in 39 cases (50.6 %) and in patients without deviating treatment recommendation in 66 cases (34.0 %). The additional examinations carried out included additional punch biopsies (most frequent) and MRI scans, but also additional ultrasounds or a mammography. Conclusions: Additional examinations lead to a change in treatment in a higher percentage of patients in comparison with the initial screening including assessment. An exact reexamination of the findings obtained in the screening is therefore preoperatively necessary in order to guarantee optimum treatment.