Maria Carmen De Santis

Objective: The aim of this study was to evaluate the accuracy of machine and deep learning approaches on radiomics features obtained by Dynamic Contrast Enhanced Magnetic Resonance Imaging (DCE-MRI) and contrast enhanced mammography (CEM) in the characterization of breast cancer and in the prediction of the tumor molecular profile. Methods: A total of 153 patients with malignant and benign lesions were analyzed and underwent MRI examinations. Considering the histological findings as the ground truth, three different types of findings were used in the analysis: (1) benign versus malignant lesions; (2) G1 + G2 vs. G3 classification; (3) the presence of human epidermal growth factor receptor 2 (HER2+ vs. HER2−). Radiomic features (n = 851) were extracted from manually segmented regions of interest using the PyRadiomics platform, following IBSI-compliant protocols. Highly correlated features were excluded, and the remaining features were standardized using z-score normalization. A feature selection process based on Elastic Net regularization (α = 0.5) was used to reduce dimensionality. Synthetic balancing of the training data was applied using the ROSE method to address class imbalance. Model performance was evaluated using repeated 10-fold cross-validation and AUC-based metrics. Results: Among the 153 patients enrolled in the studies, 113 were malignant lesions. Among the 113 malignant lesions, 32 had high grading (G3) and 66 had the HER2+ receptor. Radiomic features derived from both CEM and DCE-MRI showed strong discriminative performance for malignancy detection, with several features achieving AUCs above 0.80. Gradient Boosting Machine (GBM) achieved the highest accuracy (0.911) and AUC (0.907) in differentiating benign from malignant lesions. For tumor grading, the neural network model attained the best accuracy (0.848), while LASSO yielded the highest sensitivity (0.667) for detecting high-grade tumors. In predicting HER2+ status, the neural network also performed best (AUC = 0.669), with a sensitivity of 0.842. Conclusions: Radiomics-based machine learning models applied to multiparametric CEM and DCE-MRI images offer promising, non-invasive tools for breast cancer characterization. The models effectively distinguished benign from malignant lesions and showed potential in predicting histological grade and HER2 status. These results demonstrate that radiomic features extracted from CEM and DCE-MRI, when analyzed through machine and deep learning models, can support accurate breast cancer characterization. Such models may assist clinicians in early diagnosis, histological grading, and biomarker assessment, potentially enhancing personalized treatment planning and non-invasive decision-making in routine practice.

Background: In the last decades, the increasing number of adolescent and young adult (AYA) survivors of breast cancer (BC) has highlighted the cardiotoxic role of cancer therapies, making cardiovascular diseases (CVDs) among the most frequent, although rare, long-term sequalae. Leveraging innovative artificial intelligence (AI) tools and real-world data (RWD), we aimed to develop a causally interpretable model to identify young BC survivors at risk of developing CVDs. Methods: We designed and trained a Bayesian network (BN), an AI model, making use of expert knowledge and data from population-based (1036 patients) and clinical (339 patient) cohorts of female AYA (i.e., aged 18 to 39 years) 1-year survivors of BC, diagnosed in 2009–2019. The performance achieved by the BN model was validated against standard classification metrics, and two clinical applications were proposed. Results: The model showed a very good classification performance and a clear causal semantic. According to the predictions made by the model, focusing on the 25% of AYA BC survivors at higher risk of developing CVDs, we could identify 81% of the patients who would actually develop it. Moreover, a desktop-based app was implemented to calculate the individual patient’s risk. Conclusions: In this study, we developed the first causal model for predicting the CVD risk in AYA survivors of BC, also proposing an innovative AI approach that could be useful for all researchers dealing with RWD. The model could be pivotal for clinicians who aim to plan personalized follow-up strategies for AYA BC survivors.

Background: We assessed the current practice concerning the axillary management of breast cancer (BC) patients undergoing upfront surgery among radiation oncologists (ROs) practising in Italy. Methods: An online survey via SurveyMonkey (including 21 questions) was distributed amongst ROs in Italy through personal contacts and the Italian Association for Radiotherapy and Clinical Oncology (AIRO) network from August to September 2022. We particularly focused on the emerging omission of axillary lymph node dissection (ALND) in the presence of 1–2 sentinel node-positive patients and the consequent change in the role of regional nodal irradiation (RNI). Results: A total of 101/195 (51% response rate) Italian Radiotherapy Cancer Care Centres answered the survey. With respect to patients with 1–2 sentinel node-positive, the relative proportion of respondents that offer patients ALND a) always, b) only in selected cases, and c) never was 37.6%, 60.4%, and 2.0%, respectively, with no significant geographical (North vs. Centre–South Italy; p = 0.92) or institutional (Academic vs. non-Academic; p = 0.49) differences. Radiation therapy indications varied widely in patients who did not undergo ALND. Among these, about a third of the respondents (17/56, 30.4%) stated that RNI was constantly performed. On the other hand, half of the respondents offered RNI in selected cases, stating that an unfavourable biologic tumour profile and extracapsular nodal extension were considered drivers of their decision. Conclusions: Results of the present survey show the variability of axillary management offered in clinical practice for BC patients undergoing conserving surgery upfront in Italy. Analysis of these attitudes may trigger the modification of some clinical approaches through multidisciplinary collaboration and create the background for future clinical investigations.

Background Mammographic density (MD) is a risk factor for breast cancer (BC) development, and recurrence. However, its predictive value has been less studied. Herein, we challenged MD as a biomarker associated with response in patients treated with neoadjuvant therapy (NAT). Methods Data on all NAT treated BC patients prospectively collected in the registry of Fondazione IRCCS Istituto Nazionale dei Tumori, Milano, Italy (2009–2019) were identified. Diagnostic mammograms were used to evaluate and score MD as categorized by the Breast Imaging-Reporting and Data System (BI-RADS), which identifies 4 levels of MD in keeping with relative increase of fibro-glandular over fat tissue. Each case was classified according to the following categories a (MD < 25%), b (26–50%), c (51–75%), and d (> 75%). The association between MD and pathological complete response (pCR), i.e., absence of BC cells in surgical specimens, was analyzed in multivariable setting used logistic regression models with adjustment for clinical and pathological variables. Results A total of 442 patients were analyzed, 120 of which (27.1%) attained a pCR. BI-RADS categories a, b, c, and d accounted for 10.0%, 37.8%, 37.1% and 15.2% of cases. Corresponding pCR were 20.5%, 26.9%, 30.5%, 23.9%, respectively. At multivariable analysis, when compared to cases classified as BI-RADS a, those with denser breast showed an increased likelihood of pCR with odds ratio (OR) of 1.70, 2.79, and 1.47 for b, c and d categories, respectively (p = 0.0996), independently of age, BMI [OR underweight versus (vs) normal = 3.76], clinical nodal and tumor status (OR T1/Tx vs T4 = 3.87), molecular subtype (HER2-positive vs luminal = 10.74; triple-negative vs luminal = 8.19). In subgroup analyses, the association of MD with pCR was remarkable in triple-negative (ORs of b, c and d versus a: 1.85, 2.49 and 1.55, respectively) and HER2-positive BC cases (ORs 2.70, 3.23, and 1.16). Conclusion Patients with dense breast are more likely to attain a pCR at net of other predictive factors. The potential of MD to assist decisions on BC management and as a stratification factor in neoadjuvant clinical trials should be considered.

Aim To explore breast cancer patient’s perspective on future genetic testing for prediction of toxicity after breast radiotherapy (RT). Materials and methods The study involved patient enrolled in the Italian branch of the REQUITE project conducted at the National Cancer Institute in Milan. Semi-structured interviews were conducted within one month from the end of radiotherapy treatment by two radiation oncologists and a radiotherapy technician previously trained by a clinical psychologist with experience in the oncology field. Semi-structured interviews are characterized by a set of pre-defined questions and developed ad hoc by researchers in Leicester within the REQUITE project. The interview questions investigated interest in undergoing the genetic test and expectations on its usefulness and disadvantages. Results Eighteen interviews were conducted and analysed. Forty-five initial codes were combined into nine themes which were then clustered in two main macro-areas (i) Opportunities and (ii) Challenges. Overall, all patients understand the aim of the genetic test and considered its intrinsic opportunity to make the physician more confident with the treatment. Regarding side effects, most of patients felt prepared to RT but not without fear. Many women considered important to have the largest and reliable information, also about negative experiences. Prevailing emotions were anxiety and fear but not connected to genetic test’s result. Conclusions A genetic test could be an opportunity because generate knowledge and give patients a dynamic role in the decision-making approach. Prediction of single patient radiosensitivity before RT could prompt suggestion to entail a more and more tailored radiation treatment in the era of personalized approach.

Objectives: To determine interobserver variability in axillary nodal contouring in breast cancer (BC) radiotherapy (RT) by comparing the clinical target volume of participating single centres (SC-CTV) with a gold-standard CTV (GS-CTV). Methods: The GS-CTV of three patients (P1, P2, P3) with increasing complexity was created in DICOM format from the median contour of axillary CTVs drawn by BC experts, validated using the simultaneous truth and performance-level estimation and peer-reviewed. GS-CTVs were compared with the correspondent SC-CTVs drawn by radiation oncologists, using validated metrics and a total score (TS) integrating all of them. Results: Eighteen RT centres participated in the study. Comparative analyses revealed that, on average, the SC-CTVs were smaller than GS-CTV for P1 and P2 (by −29.25% and −27.83%, respectively) and larger for P3 (by +12.53%). The mean Jaccard index was greater for P1 and P2 compared to P3, but the overlap extent value was around 0.50 or less. Regarding nodal levels, L4 showed the highest concordance with the GS. In the intra-patient comparison, L2 and L3 achieved lower TS than L4. Nodal levels showed discrepancy with GS, which was not statistically significant for P1, and negligible for P2, while P3 had the worst agreement. DICE similarity coefficient did not exceed the minimum threshold for agreement of 0.70 in all the measurements. Conclusions: Substantial differences were observed between SC- and GS-CTV, especially for P3 with altered arm setup. L2 and L3 were the most critical levels. The study highlighted these key points to address. Advances in knowledge: The present study compares, by means of validated geometric indexes, manual segmentations of axillary lymph nodes in breast cancer from different observers and different institutions made on radiotherapy planning CT images. Assessing such variability is of paramount importance, as geometric uncertainties might lead to incorrect dosimetry and compromise oncological outcome.