Comparing Decision on Aesthetics After Breast Cancer Locoregional Treatment.

Overview

Breast cancer is the most commonly diagnosed cancer, with an estimated 2.3 million new cases per year globally. Approximately 90% of these patients will undergo breast surgery with/without radiation (locoregional treatment). Different surgical techniques can be offered to the patient, each leading to completely different aesthetic outcomes. Moreover, the aesthetic outcome could be completely different for patients undergoing the same surgery based on individual patient factors (e.g., age, body habitus). In the CINDERELLA trial, the investigators will be using the (Breast Locoregional (BreLO) AI system (an artificial intelligence-based tool for the classification of aesthetic outcomes and matching data and photographs) integrated into CANKADO (a cloud-based healthcare platform) to create an easy-to-use application that can be used on any electronic device, to simulate visually to the patient the aesthetic outcome of a certain surgery or radiation treatment. In the CINDERELLA trial, the investigators plan to compare whether the application helped fulfil the expectations and lead to a better quality of life compared with the classical approach. In the classical approach (control arm), doctors usually propose a locoregional treatment and explain theoretically how the result will be. Nurses help by explaining further details about the surgery and possible outcomes. In most centres, no photographic evaluation is done, and expectations are not measured. The CINDERELLA trial will help overcome miscommunication and potential boundaries in the patient's or physician's understanding of the potential outcomes of locoregional breast cancer treatment.

The CINDERELLA clinical trial will be an open, prospective randomized trial that Champalimaud Foundation will coordinate. Five clinical centres agreed to participate in the trial. The trial will be designed and reported according to the latest SPIRIT-AI and CONSORT-AI guidelines. The randomization will be made by adopting a dynamical approach following the Minimization Method. Assignment of the recruited patients to the study arms will take into account the stratification of the participants (younger and older than 50 / breast-conserving or mastectomy / mastectomy with or without radiotherapy), aiming to reduce bias and confounding by assuring the balance of the group. After being proposed to the trial and checked for all the eligibility criteria, patients will be given the complete patient information before signing the informed consent. For the five centres, a minimum of 515 patients should be enrolled in each arm of the study. After randomization, the patient will either follow: 1. \- The intervention arm with the CINDERELLA APProach with the introduction and access to CANKADO with the BreLO AI system. The Expectations Questionnaire, Healthcare Professionals Multidimensional Evaluation Questionnaire and standard patient-reported outcomes measures (PROMs) (EQ-5D-5L and BREAST Q ICHOM) will be filed electronically. Standard photographic capture will be taken at this point. After the photographic capture, the BreLO-AI will match the patient biometrics and images with the more identical case existing in the BreLO repository and already classified by the BCCT.core into excellent, good, fair and poor. The new patient can then visualize the results. In case of doubts, queries will be answered through the app or, if needed, by phone call or booking another appointment. Questionnaires and Photographs will be repeated after wound healing is complete, six months and one year after the end of treatment (surgery or radiotherapy if radiotherapy was done). 2. \- The control arm with the Conventional approach with a theoretical explanation by the doctor/nurse of the proposed locoregional treatment and possible outcomes. The Expectations Questionnaire, Healthcare Professionals Multidimensional Evaluation Questionnaire and standard PROMs (EQ-5D-5L and BREAST Q ICHOM) will be filed electronically. Standard photographic capture will be taken at this point. In case of doubts, the patient will book another appointment with the doctor/nurse, as usually done in routine practice. Questionnaires and Photographs will be repeated after healing is complete, six months and one year after the end of treatment (surgery or radiotherapy if radiotherapy was done). Digital Photography (same protocol for all participating centres) - a similar protocol for image capture will exist for all centres. The standalone photography with an automatic robot will be progressively implemented (www.photorobot.com). DATA COLLECTION PATIENT-RELATED FACTORS * Date of birth, Weight / Height / BMI, Thoracic perimeter, Bra size and cup * Education degree, Profession, Hobbies * Marital status, Pregnancies and offspring * Breast-feeding, Menopausal status * Smoking * Connective tissue diseases * Confirmed Pathogenic Germline Variant TUMOUR-RELATED FACTORS * Unilateral (unifocal, multifocal, multicentric) Bilateral * Histological type according to World Health Organization (WHO) classification (e.g., invasive ductal carcinoma, lobular carcinoma) (size in mm) staging ER, PR, Her2, Ki67, cTNM - pTNM / ypTNM TREATMENT-RELATED FACTORS \*Type of Surgery/ Type of Reconstruction: (data collection regarding surgery should also include acellular dermal matrice (ADM) if used - type and placement) TYPE OF SURGERY: C1 - Conservative surgery - unilateral or bilateral, C2 - Conservative surgery with bilateral reduction (uni or bilateral), C3 - Conservative surgery with LD or LICAP/TDAP, C4 - Conservative surgery with bilateral breast augmentation, M1 - Mastectomy with unilateral reconstruction with implant, M2 - Mastectomy with unilateral reconstruction with autologous flap, M3 - Mastectomy with bilateral reconstruction with implants, M4 - Mastectomy with bilateral reconstruction with autologous flaps, M5 - Mastectomy with unilateral reconstruction with implant and contralateral symmetrisation with implant (augmentation), M6 - Mastectomy with unilateral reconstruction with implant and contralateral symmetrisation with reduction, M7 - Mastectomy with unilateral reconstruction with autologous flap and contralateral symmetrisation with reduction, M8 - Mastectomy with unilateral reconstruction with autologous flap and contralateral symmetrisation with implant (augmentation). STATISTICAL ANALYSIS An extensive descriptive analysis will be performed to characterize the groups in detail and the outcomes of the study at baseline as well as in the following points of data collection. Concerning the primary objectives, models of the class of generalized linear mixed models (in particular, multinomial regression models for ordinal data) will be estimated to evaluate the effect of the training and the women's characteristics on their evaluation of the aesthetic results of the surgery at each time and along time through longitudinal analysis. The Wilcoxon signed rank test for pairs will also be used to evaluate the effect of training on the level of agreement of the expectations and the final result. Weighted Cohen's k will be calculated for both groups (train and control) and compared using a statistical test and/or bootstrap techniques to assess the improvement in the ability to classify the aesthetic result of their surgery provided by training. A measure of similarity between self-evaluation and the BCCT.core will be computed for each participant, and a beta regression model will be estimated to assess the effect of training, controlling variables that can play as confounders, such as women's and disease characteristics at each time point and in a longitudinal perspective. Concerning the secondary objectives, the patient-reported outcome measures administered will be scored according to the official guidelines provided by the developers of the instruments. Besides the descriptive statistics, the outcomes will be compared between groups using adequate statistical tests. Again, models of the class of the general linear mixed models will be used.