Editor's Note: Breast cancer is a malignant tumor that seriously threatens women's health. Neoadjuvant therapy is a crucial means to provide more surgical opportunities and pave the way for a cure. However, the pathological complete response (pCR) varies significantly with different molecular subtypes and treatment regimens, making it challenging to predict treatment efficacy early in neoadjuvant therapy for individual patients. From May 31 to June 4, 2024, the American Society of Clinical Oncology (ASCO) Annual Meeting was grandly held. At the meeting, Professor Kun Wang's team from Guangdong Provincial People's Hospital presented a study on a model called NeoMDSS, which predicts the effectiveness of neoadjuvant therapy for breast cancer based on MRI and other clinical pathological parameters. Oncology Frontier interviewed Professor Kun Wang to share the main results and clinical significance of this study.Study Overview
Poster #184 (Abs #592)
Title: An MRI-based multi-parameter clinical decision support system (NeoMDSS) for early prediction of pathological complete response after the first cycle of neoadjuvant therapy in breast cancer: A multi-center prospective observational cohort study.
Background
Currently, there is no reliable and convenient method to predict the efficacy of neoadjuvant therapy (NAT) for breast cancer in its early stages. This lack of early prediction may adversely affect patients who are unlikely to achieve pathological complete response (pCR) from NAT.
Methods
This prospective, multicenter, observational study primarily enrolled breast cancer patients undergoing radical surgery after neoadjuvant therapy. All patients underwent MRI scans before and after the first cycle of NAT (1st-NAT). The study analyzed the impact of clinical pathological features, including histological grade, hormone receptor (HR) and HER2 status, Ki-67 expression, blood cell analysis, and MRI parameters on NAT outcomes. By integrating independent influencing factors, a multi-parameter clinical decision support system (NeoMDSS) model was developed based on retrospective cohort data and validated in a prospective single-center internal cohort and a prospective multicenter external cohort. (Clinical trial information: NCT04909554)
Results
From January 2019 to December 2023, 301 breast cancer patients were enrolled, including 140 in the training cohort, 120 in the internal validation cohort, and 41 in the external validation cohort. The NeoMDSS model showed excellent performance in predicting pCR after 1st-NAT, with an AUC of 0.874 (95% CI 0.813-0.935) in the training cohort, 0.845 (95% CI 0.771-0.919) in the internal validation cohort, and 0.867 (95% CI 0.742-0.992) in the external validation cohort. In the internal validation cohort, the sensitivity, specificity, and accuracy of the NeoMDSS model were 80.0%, 81.3%, and 80.5%, respectively, and 87.0%, 83.3%, and 85.4% in the external validation cohort. Calibration curves and decision curve analysis (DCA) further supported the clinical value of the NeoMDSS model. To facilitate the clinical application of the NeoMDSS model, researchers developed a corresponding nomogram and an online website (www.gdphneomdss.com) to calculate the probability of pCR.
Conclusion
The NeoMDSS model is a convenient clinical practice tool for calculating and accurately predicting pCR rates after 1st-NAT. This model helps clinicians decide whether to modify treatment plans and optimize personalized treatment strategies for patients with poor responses to NAT.
Researcher’s Insight
Oncology Frontier: As the main participant in this study, could you briefly introduce the main functions and significance of the NeoMDSS model?
Professor Kun Wang: In clinical practice, neoadjuvant therapy is one of the main treatment methods for various molecular subtypes of breast cancer. However, the pCR rate varies across different subtypes and treatment regimens. For example, the pCR rate for HER2-positive breast cancer using the TCbHP regimen is about 60%, meaning that about 40% of patients cannot achieve pCR even after six cycles of treatment. This study is a multicenter, prospective, observational clinical study led by Guangdong Provincial People’s Hospital, aiming to evaluate patients’ pCR rates using the NeoMDSS model, which integrates clinical presentation parameters, MRI parameters, and blood test-related parameters. For instance, if the model evaluates that the pCR rate exceeds 80% after one cycle of neoadjuvant therapy, the treatment can continue; if the pCR probability is low, the treatment regimen can be changed, or surgery can be considered. Therefore, the greatest significance of this study lies in predicting the efficacy of neoadjuvant therapy based on patients’ dynamic changes at a very early stage.
Oncology Frontier: How do you view the application of the NeoMDSS model in clinical practice? How does it help doctors formulate more personalized treatment plans, especially for patients with poor responses to neoadjuvant therapy?
Professor Kun Wang: The biggest advantage of the NeoMDSS model is that it allows doctors to adjust neoadjuvant therapy plans early. When we predict that a patient can achieve pCR, the original regimen can continue; when the model predicts a poor outcome, the treatment plan can be changed after 2-4 cycles of neoadjuvant therapy. For patients with particularly poor efficacy, such as those with stable disease (SD) or progressive disease (PD), surgery may be considered immediately. Therefore, the NeoMDSS model enables doctors to provide individualized and precise treatment methods for patients, bringing more benefits to them.
Oncology Frontier: What challenges did you and your team encounter during the development of the NeoMDSS model, and how did you overcome them?
Professor Kun Wang: Previous similar evaluation models often focused on baseline status, with an accuracy of 60-70%. Due to the heterogeneity of breast cancer tumors, the development of each patient varies after 1-2 cycles of neoadjuvant therapy, posing difficulties for model design and application. We included comprehensive indicators such as tumor trend, MRI parameters, tumor typing, TNM staging, and blood-related factors. Through retrospective clinical research, we developed the NeoMDSS model with clinical value. Subsequently, prospective, multicenter clinical research verified that the model’s accuracy remained above 82%, showing good performance across different breast cancer molecular subtypes. Therefore, I believe the NeoMDSS model is very worthy of promotion.
Oncology Frontier: What do you think about the application prospects of the NeoMDSS model in early prediction of pCR after neoadjuvant therapy for breast cancer?
Professor Kun Wang: The NeoMDSS model has undergone multicenter validation, with stable results, and I hope it can be widely used. Currently, we have developed a corresponding website for NeoMDSS and made it publicly available for interested doctors to try. By entering relevant parameters on the website, the model automatically evaluates a relatively accurate pCR prediction value. I hope that after predicting the results, users can provide feedback to help us further refine and optimize the model, making it even better.
