Search

Purpose The purpose of this study was to create a deep learning algorithm to infer the benign or malignant nature of breast nodules using two-dimensional B-mode ultrasound data initially marked as BI-RADS 3 and 4. Materials and methods An ensemble of mask region-based convolutional neural networks (Mask-RCNN) combining nodule segmentation and classification were trained to explicitly localize the nodule and generate a probability of the nodule to be malignant on two-dimensional B-mode ultrasound. These probabilities were aggregated at test time to produce final results. Resulting inferences were assessed using area under the curve (AUC). Results A total of 460 ultrasound images of breast nodules classified as BI-RADS 3 or 4 were included. There were 295 benign and 165 malignant breast nodules used for training and validation, and another 137 breast nodules images used for testing. As a part of the challenge, the distribution of benign and malignant breast nodules in the test database remained unknown. The obtained AUC was 0.69 (95% CI: 0.57–0.82) on the training set and 0.67 on the test set. Conclusion The proposed deep learning solution helps classify benign and malignant breast nodules based solely on two-dimensional ultrasound images initially marked as BIRADS 3 and 4.

To evaluate the diagnostic accuracy of breast MRI in identifying lesions requiring excision for patients with suspicious nipple discharge but normal mammograms and ultrasounds. Between September 2013 and May 2019, 106 female participants (mean age 57.9 years) were consecutively included in this prospective multicenter study; 102 were retained for analysis. MRI was considered negative in the absence of suspicious enhancement and positive in cases of ipsilateral abnormal enhancement (BI-RADS 3 to 5). Final diagnoses were based on histological findings of surgical or percutaneous biopsies or at 1-year follow-up. We considered all lesions requiring excision found on pathology (papilloma, atypia, nipple adenomatosis, or cancer) as positive results. We considered spontaneous resolution of the discharge at 1 year as a negative result. MRI showed ipsilateral abnormal enhancement in 54 patients (53%) revealing 46 lesions requiring excision (31 benign papillomas, 5 papillomas with atypia, 2 nipple adenomatosis, and 8 cancers) and 8 benign lesions not requiring excision. No suspicious enhancement was found in the remaining 48 participants (47%). Forty-two were followed up at 1 year with spontaneous resolution of the discharge and six underwent surgery (revealing 2 benign papillomas). MRI diagnostic accuracy for the detection of a lesion requiring excision was as follows: sensitivity 96%, specificity 85%, positive predictive value 85%, and negative predictive value 96%. In patients with suspicious nipple discharge and normal mammogram and ultrasound, MRI demonstrates excellent performance to identify lesions for which excision is required. Normal MRI indicates it is safe to propose follow-up only management, thus avoiding unnecessary duct excision. ClinicalTrials.gov NCT02819362 • Breast MRI can be useful for the management of patients with suspicious nipple discharge and negative mammogram and ultrasound. • MRI detected a lesion requiring excision in 46 participants (45%) with unexplained discharge. • If breast MRI is negative, follow-up is a safe alternative for these patients.

Second reading is an important part of breast cancer organized screening program. Image quality control and detection of non-diagnosed cancer by first reader are the two goals of this process. In France, 6 % of all screening cancer are diagnosed by second reading, actually done on screen film. With the technologic evolution (Digital breast tomosynthesis, Artificial intelligence) and societal digitalization, this process need to evolve. After some report about organization and results for second reading in France and outside, current and future shortcomings, proposition from professionals involved in breast cancer screening are made to improve this public health program.

INTRODUCTION: Organized and individual breast screening have been accompanied by an increase in the detection of "atypical breast lesions "(ABL). Recently, the NOMAT multicenter study proposed a predictive model of the risk of developing breast cancer after detection of an ABL in order to avoid surgical removal of "low-risk" lesions. It also aimed to provide information on psychological experience, in particularly anxiety, to assist in the shared medical decision process. METHODS: Three hundred women undergoing surgery for ABL were included between 2015 and 2018 at 18 French centers. Women completed questionnaires before and after surgery assessing their level of anxiety (STAI-State, STAI-Trait), their level of tolerance to uncertainty, their perceived risk of developing a breast cancer, and their satisfaction with the management care. RESULTS: One hundred nighty nine patients completed the STAI-Status before and after surgery. Overall, a decrease in anxiety level (35.4 vs 42.7, p

e12557 Background: Triple-negative breast cancer (TNBC) tends to exhibit aggressive behavior lacking from targeted therapies. Tumor-associated-macrophages (TAMs as M2 and M1 macrophages) are interest targets in TNBC to approach the tumor patient immunity. The mitochondrial translocator protein (TSPO), a sensitive marker for macrophages, could be interesting for TNBC micro-environment stratification. We performed a multicenter imaging pilot study (NCT04320030) using non-invasive [18F]DPA-714 (DPA), a TSPO PET radioligand, aiming at assessing immunity to define the immunotherapy possibilities in TNBC patients. Methods: All patients underwent TSPO genotyping (LAB: Low Affinity Binder, MAB: Mixed and HAB: High), [18F]FDG (FDG) and DPA PET-CT, macrophages immunochemistry and in vitro TSPO autoradiography using both tritiated PK11195 (PK) and DPA. Thorax PET-CT was acquired just after injection (DPA1) to 30 min, then from 45 min (DPA2) to 60 min, followed by a whole body PET-CT. Groups of interest (Macrophages immunochemistry and Genotype) were compared using Kruskal-Wallis test. Correlation between continuous parameters were determined using Spearman’s Rho. Results: A total of 13 TNBC patients were included. No adverse events occurred after DPA injection. For TSPO genotyping, 2/13 (15.5%) patients were LAB, 6/13 (46%) MAB and 5/13 (38.5%) HAB. All of them showed FDG and DPA uptake regardless of the patient's genotype without significant difference between the 3 groups. The DPA kinetics showed a tracer accumulation between DPA1 and DPA2, whatever patient genotypes, observed on SUVmax (p = 0.0015), SUVmean (p = 0.0015) and TL-DPA (p = 0.0024). A correlation between FDG and DPA1 either DPA2 tumor volume was shown, respectively (p = 0.0252 and 0.0067) as well as between the TLG (FDG) and the TL-DPA1 (p = 0.0346) regardless of the patient's genotype. Macrophages immunochemistry showed a tendency difference for M2 percentage (p = 0.09) between the 3 genotypes groups. In vitro binding ratio (PK/DPA), measured for both radioligands, on adjacent tumor slices, was 1.77 (1.35-2.21). A difference in this binding ratio was observed between MAB and HAB (Mean 1.68±0.36) compared to LAB 3.04±0.23 (p = 0.0367), suggesting that DPA binding is sensitive to TSPO genotype, unlike PK binding. Conclusions: Despite the limited population, all TNBC tumor patients were DPA positive with an accumulation over time. The total glycolysis and total DPA uptake in the tumor were correlated, which may link tumor aggressiveness and DPA uptake. These data were confirmed by a better sensitivity for HAB and MAB vs LAB. DPA seemed to be a promising tracer to explore non-invasively TNBC patient’s immunity. Clinical trial information: NCT04320030.

International audience; Background: Diagnosis of atypical breast lesions (ABLs) leads to unnecessary surgery in 75-90% of women. We have previously developed a model including age, complete radiological target excision after biopsy, and focus size that predicts the probability of cancer at surgery. The present study aimed to validate this model in a prospective multicenter setting.- methods: Women with a recently diagnosed ABL on image-guided biopsy were recruited in 18 centers, before wire-guided localized excisional lumpectomy. Primary outcome was the negative predictive value (NPV) of the model.Results: The NOMAT model could be used in 287 of the 300 patients included (195 with ADH). At surgery, 12 invasive (all grade 1), and 43 in situ carcinomas were identified (all ABL: 55/287, 19%; ADH only: 49/195, 25%). The area under the receiving operating characteristics curve of the model was 0.64 (95% CI 0.58-0.69) for all ABL, and 0.63 for ADH only (95% CI 0.56-0.70). For the pre-specified threshold of 20% predicted probability of cancer, NPV was 82% (77-87%) for all ABL, and 77% (95% CI 71-83%) for patients with ADH. At a 10% threshold, NPV was 89% (84-94%) for all ABL, and 85% (95% CI 78--92%) for the ADH. At this threshold, 58% of the whole ABL population (and 54% of ADH patients) could have avoided surgery with only 2 missed invasive cancers.Conclusion: The NOMAT model could be useful to avoid unnecessary surgery among women with ABL, including for patients with ADH.

Resume Traiter mieux tout en traitant moins tel est l’enjeu de la cancerologie mammaire. C’est dans ce contexte que s’inscrit le concept de desescalade du traitement locoregional dont l’objectif est de diminuer la morbidite due aux traitements et de developper des strategies therapeutiques specifiques optimisees. Les performances diagnostiques de l’IRM mammaire sont aujourd’hui largement validees au vu de son excellente valeur predictive negative superieure a 95 %. Les etudes de depistage de femmes a tres haut risque de cancer du sein ont egalement valide les excellentes performances de l’IRM, que ce soit pour la detection des carcinomes infiltrants comme des carcinomes in situ, avec une sensibilite de l’IRM superieure a celle du couple mammographie et echographie. La question posee est celle de la place de l’IRM comme outil d’aide a la desescalade therapeutique ou diagnostique. Autrement dit, l’IRM peut-elle contribuer a optimiser les strategies diagnostiques ou therapeutiques ? Nous essaierons de repondre a cette question selon le contexte clinique.

Resume La reactualisation des recommandations professionnelles de l’Institut national du cancer (INCa) sur les strategies de detection precoce des cancers du sein et de l’ovaire chez les femmes porteuses d’une mutation des genes BRCA1 ou BRCA2, valide la diminution du nombre d’incidence mammographique, limite a une seule incidence mammographique oblique externe par sein pour les femmes indemnes tant que l’IRM est realisee, l’arret du depistage par IRM a 65 ans, la poursuite d’un suivi mammographique annuel au-dela de 65 ans et l’arret d’un depistage echographique du cancer des annexes.

PURPOSE We evaluated the addition of breast magnetic resonance imaging (MRI) to standard radiologic evaluation on the re-intervention rate in women with ductal carcinoma in situ (DCIS) undergoing breast-conserving surgery. PATIENTS AND METHODS Women with biopsy-proven DCIS corresponding to a unifocal microcalcification cluster or a mass less than 30 mm were randomly assigned to undergo MRI or standard evaluation. The primary end point was the re-intervention rate for positive or close margins (< 2 mm) in the 6 months after randomization ( ClinicalTrials.gov identifier: NCT01112254). RESULTS A total of 360 patients from 10 hospitals in France were included in the study. Of the 352 analyzable patients, 178 were randomly assigned to the MRI arm, and 174 were assigned to the control arm. In the intent-to-treat analysis, 82 of 345 patients with the assessable end point were reoperated for positive or close margins within 6 months, resulting in a re-intervention rate of 20% (35 of 173) in the MRI arm and 27% (47 of 172) in the control arm. The absolute difference of 7% (95% CI, −2% to 16%) corresponded to a relative reduction of 26% (stratified odds ratio, 0.68; 95% CI, 0.41 to 1.1; P = .13). When considering only the per-protocol population with an assessable end point, the difference was 9% (stratified odds ratio, 0.59; 95% CI, 0.35 to 1.0; P = .05). Total mastectomy rates were 18% (31 of 176) in the MRI arm and 17% (30 of 173) in the control arm (stratified P = .93). For 100 lesions seen on MRI, nonmass-like enhancement was more predominant (82%) than mass enhancement (20%). Nevertheless, no specific morphologic and kinetic parameters for DCIS were identified. CONCLUSION The study did not show sufficient surgical improvement with the use of preoperative MRI to be clinically relevant in DCIS staging. However, this could be reconsidered with the improvement of new MRI sequences and new modalities in magnetic resonance techniques.

Radiation recall is a rare phenomenon, defined as an acute inflammatory reaction in a previously irradiated area, after administration of anti-tumor agents, including chemotherapy. It is most commonly reported to trigger skin reactions but internal organ involvement is possible, particularly with gemcitabine. We report here a unique case of a gemcitabine-induced radiation recall myelitis following spinal irradiation. A 53-year-old patient received analgesic irradiation of the seventh thoracic vertebra (T7) in the context of metastatic non-small cell lung cancer, at conventional radiotherapy dose and fractionation. She was subsequently treated with gemcitabine and developed myelitis whose chronology is compatible with a radiation recall reaction. Spinal MRI confirmed a T6–T7 spinal cord enhancement, with an associated spinal cord oedema. Corticosteroids and supportive care did not improve myelitis symptoms. The patient died within a year of the radiation recall, due to a metastatic progression of lung cancer. This is, to our knowledge, the first reported case of gemcitabine-induced radiation recall myelitis and only the third case involving the spinal cord. Radiation recall is a rare and poorly understood phenomenon and all cases should be reported.

Resume Apres traitement conservateur du cancer du sein, le statut des marges d’exerese constitue un facteur de risque independant de recidive locale homolaterale. S’il n’existe pas de consensus sur la definition d’une marge optimale, on peut dire qu’elle correspond au compromis entre passer tres a distance de la tumeur au risque d’enlever beaucoup de tissu et preserver un bon resultat cosmetique. Les reinterventions pour marges non saines ne sont pas sans consequence, augmentant les risques de complications locales, de retard des traitements adjuvants ou d’alteration du resultat esthetique et pesent egalement sur le plan economique. L’imagerie tient une place essentielle dans la decision de la strategie therapeutique. Peut-elle contribuer a diminuer le nombre de reinterventions pour marges non saines ? Dans cette presentation, nous aborderons tour a tour la place du bilan preoperatoire, du reperage preoperatoire et de la radiographie peroperatoire dans l’optimisation de la prise en charge chirurgicale. Enfin, nous verrons dans quelles circonstances l’imagerie postoperatoire precoce peut aider a evaluer la maladie residuelle.

Resume Dans cette presentation nous nous interesserons a la surveillance apres traitement d’un cancer du sein des femmes âgees, sous deux angles differents. Quelle surveillance au long terme des femmes âgees traitees jeunes d’un cancer du sein : l’adage « mammographie annuelle a vie » est-il toujours d’actualite ? Quelle surveillance apres traitement d’un cancer du sein de la femme âgee ?