Breast Cancer
Conditions
Brief summary
This study aims to evaluate and report the clinical adverse events and dosimetric parameters in breast cancer patients undergoing an "all-in-one (AIO)" one-stop, fully automated radiotherapy workflow. By systematically tracking these clinical and physical metrics, we seek to establish a standardized clinical protocol for AIO radiotherapy in breast cancer management.
Interventions
OTHERRadiotherapy procedure
The workflow relies on specialized convolutional neural networks for automated segmentation and dose-prediction auto-planning. These breast cancer models were trained on 285 historical institutional cases spanning radical mastectomy and breast-conserving surgery over five years. Auto-delineated structures include the clinical target volume, regional lymph nodes (if involved), tumor bed (identified by surgical clips), heart, bilateral lungs, unaffected breast, spinal cord, esophagus, thyroid, and affected humeral head. These contours guide dose prediction to generate deliverable tangential arc plans via clinical-goal-guided automated optimization in the treatment planning system. To adapt to the on-couch treatment scenario, models were validated on retrospective data and offline routines to maximize target delineation accuracy and the first-approval rate of auto-plans.
Sponsors
Fudan University
Study design
Allocation
NON_RANDOMIZED
Intervention model
SEQUENTIAL
Primary purpose
OTHER
Masking
NONE
Eligibility
Sex/Gender
ALL
Age
18 Years to No maximum
Healthy volunteers
No
Inclusion criteria
* Histologically or pathologically confirmed breast cancer with definitive indications for radiotherapy (preoperative, postoperative, or radical) * ECOG performance status of 0-2 * Able to remain still and supine on the treatment couch for up to 30 minutes * Provision of signed, written informed consent * Able to comply with daily follow-ups and blood sample collections
Exclusion criteria
* Palliative radiotherapy for concurrent distant metastasis * Incomplete or ongoing chemotherapy * Synchronous multiple primary tumors * Current pregnancy or lactation * Prior history of radiotherapy to the ipsilateral breast, chest wall, thorax, or regional lymph nodes * Severe non-malignant comorbidities (e.g., cardiovascular or pulmonary diseases, systemic lupus erythematosus, scleroderma) resulting in a short life expectancy or inability to tolerate radical radiotherapy * Inability or unlikelihood to comply with study follow-up * Inability or unwillingness to provide written informed consent
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Acute adverse events | 6 months | The incidence and severity of acute adverse event include radiation dermatitis, pruritus, skin pain, radiation esophagitis, and radiation pneumonitis. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Accuracy | 2 months | Auto-segmentation accuracy was assessed by comparing automatically generated contours against the final physician-approved contours |
| Success rate | 2 months | Record AIO workflow success rate: online planning one-pass optimization success rate. |
| Quality of life (QoL) | 6 months | Quality of life will be evaluated via standardized QoL scales. |
| Time efficiency | 2 months | The time efficiency of the workflow was automatically recorded by the system |
| Full-Workflow Patient Intrafraction Motion | 2 months | Evaluated based on geometric deviations between pretreatment image-guided radiotherapy (IGRT), posttreatment imaging, and the baseline simulation CT |
| Correlation of Patient Metrology with Setup Error and Dosimetric Performance | 2 months | Evaluation of how Body Mass Index (BMI) and weight fluctuations correlate with geometric setup errors and in vivo gamma pass rates |
| Correlation of Anatomical Scale with Setup Error and Dosimetric Performance | 2 months | Evaluation of how anatomical scale/breast size correlates with geometric setup errors and in vivo gamma pass rates |
Countries
China
Contacts
CONTACTXiaoli Yu, MD, PhD
CONTACTXiaofang Wang, MD, PhD
Outcome results
None listed