In:
Journal of Clinical Oncology, American Society of Clinical Oncology (ASCO), Vol. 38, No. 15_suppl ( 2020-05-20), p. 3563-3563
Abstract:
3563 Background: The early stage breast cancer patients can vary in disease-free survival (DFS), innovative predictors evaluate the prognostic capacity are urgently needed. We aimed to develop and independently validate a radiomics signature based on MRI associated with phenotypes and DFS in patients with breast cancer and to establish a radiomics nomogram that incorporates the radiomics signature and clinicopathological findings using computational algorithms. Methods: In this multicenter, retrospective, cohort study, we analyzed preoperative contrast–enhanced MRI data from the prospective cohort study (n = 123) of patients who had been treated with neoadjuvant chemotherapy in phase 3 trials and independent cohort (n = 438) at the Sun Yat-sen Memorial Hospital as training cohort to develop the radiomic signature, and validated it in validation cohort (Foshan cohort, n = 121; Dongguan cohort, n = 89) between November 17, 2011, and September 21, 2019, and validated in TGCA cohort (n = 84). Machine-learning algorithm to identify robust imaging subtypes and evaluated their clinical and biologic relevance. A nomogram combining the radiomic signature and clinicopathological findings to predict individual survival based on Cox regression model. The primary endpoint was disease-free survival (DFS). This study is registered with ClinicalTrials.gov, number NCT04003558, and Chinese Clinical Trail Registry, number ChiCTR1900024020. Results: A total of 855 breast cancer patients were included. Radiomics signature was generated to classify patients into high-risk and low-risk groups in the Guangzhou training cohort. Patients with low-risk scores in the training cohort had higher DFS (hazard ratio [HR] 0.55, 95% CI 0.31 to 0.99; P= 0.045) than patients with high-risk scores, and validated in in validation cohort (HR 0.14, 95% CI 0.03 to 0.62; P= 0.003). The nomogram combined radiomics score with clinicopathological factors could accurately predict DFS benefits in training cohort (C-index = 0.83; AUC, 1, 2, 3-year were 0.80, 0.85, 0.82, respectively) and validated in validation cohorts. Conclusions: The radiomics signature are significantly associated with the DFS in patients with breast cancer. Combining the radiomics nomogram improved individualized DFS pretiction. Clinical trial information: NCT04003558 .
Type of Medium:
Online Resource
ISSN:
0732-183X
,
1527-7755
DOI:
10.1200/JCO.2020.38.15_suppl.3563
Language:
English
Publisher:
American Society of Clinical Oncology (ASCO)
Publication Date:
2020
detail.hit.zdb_id:
2005181-5
Permalink
