AI-Assisted System for Accurate Diagnosis and Prognosis of Breast Phyllodes Tumors

Official title

Development of an Artificial Intelligence-Based System for Precise Diagnosis and Prognosis of Breast Phyllodes Tumors

Overview

Breast phyllodes tumor (PT) is a rare fibroepithelial tumor, accounting for 1% to 3% of all breast tumors, categorized by the WHO into benign, borderline, and malignant, based on histopathology features such as tumor border, stromal cellularity, stromal atypia, mitotic activity and stromal overgrowth. Malignant PTs account for 18%-25%, with high local recurrence (up to 65%) and distant metastasis rates (16%-25%). Benign PT could progress to malignancy after multiple recurrences. Therefore, Early, accurate diagnosis and identification of therapeutic targets are crucial for improving outcomes and survival rates. In recent years, there has been growing interest in the application of artificial intelligence (AI) in medical diagnostics. AI can integrate clinical information, histopathological images, and multi-omics data to assist in pathological and clinical diagnosis, prognosis prediction, and molecular profiling.AI has shown promising results in various areas, including the diagnosis of different cancers such as colorectal cancer, breast cancer, and prostate cancer. However, PT differs from breast cancer in diagnosis and treatment approach. Therefore, establishing an AI-based system for the precise diagnosis and prognosis assessment of PT is crucial for personalized medicine. The research team, led by Dr. Nie Yan, is one of the few in Guangdong Province and even nationally, specializing in PT research. Their team has been conducting research on the malignant progression, metastasis mechanisms, and molecular markers for PT. The team has identified key mechanisms, such as fibroblast-to-myofibroblast differentiation, and the role of tumor-associated macrophages in promoting this differentiation. They have also identified molecular markers, including miR-21, α-SMA, CCL18, and CCL5, which are more accurate in predicting tumor recurrence risk compared to traditional histopathological grading. The project has collected high-quality data from nearly a thousand breast PT patients, including imaging, histopathology, and survival data, and has performed transcriptome gene sequencing on tissue samples. They aim to build a comprehensive multi-omics database for breast PT and create an AI-based model for early diagnosis and prognosis prediction. This research has the potential to improve the diagnosis and treatment of breast PT, address the disparities in breast PT care across different regions in China, and contribute to the development of new therapeutic targets.

Interventions

• Diagnostic test imaging
  Patient medical imaging materials including ultrasound, mammography, CT, MRI

Primary outcome measures

• Sensitivity [Time frame: Five years]
• False-negative Rate [Time frame: Five years]
• Specificity [Time frame: Five years]
• False-positive Rate [Time frame: Five years]
• Receiver Operating Characteristic Curve [Time frame: Five years]
• Area under roc Curve [Time frame: Five years]

Eligibility criteria

Inclusion criteria

• Patients diagnosed with a phyllodes tumor of the breast

Exclusion criteria

• Blurred images, imaging artifacts

Criteria are shown verbatim from the registry (in English). Final eligibility is always assessed by the study center.

Healthy volunteers: No

Study design

Observational model

Cohort

Study locations

China · 4 centers

• Sun Yat-sen University Cancer Center — Guangzhou
• Sun Yat-Sen Memorial Hospital, Sun Yat-Sen University — Guangzhou
• The Third Affiliated Hospital of Guangzhou Medical University — Guangzhou
• Guangdong Maternal and Child Health Hospital — Guangzhou

Publications

• Mishra SP, Tiwary SK, Mishra M, Khanna AK. Phyllodes tumor of breast: a review article. ISRN Surg. 2013;2013:361469. doi: 10.1155/2013/361469. Epub 2013 Mar 20. PMID 23577269
• Belkacemi Y, Bousquet G, Marsiglia H, Ray-Coquard I, Magne N, Malard Y, Lacroix M, Gutierrez C, Senkus E, Christie D, Drumea K, Lagneau E, Kadish SP, Scandolaro L, Azria D, Ozsahin M. Phyllodes tumor of the breast. Int J Radiat Oncol Biol Phys. 2008 Feb 1;70(2):492-500. doi: 10.1016/j.ijrobp.2007.06.059. Epub 2007 Oct 10. PMID 17931796
• Bera K, Schalper KA, Rimm DL, Velcheti V, Madabhushi A. Artificial intelligence in digital pathology - new tools for diagnosis and precision oncology. Nat Rev Clin Oncol. 2019 Nov;16(11):703-715. doi: 10.1038/s41571-019-0252-y. Epub 2019 Aug 9. PMID 31399699
• van der Laak J, Litjens G, Ciompi F. Deep learning in histopathology: the path to the clinic. Nat Med. 2021 May;27(5):775-784. doi: 10.1038/s41591-021-01343-4. Epub 2021 May 14. PMID 33990804
• Wang Y, Acs B, Robertson S, Liu B, Solorzano L, Wahlby C, Hartman J, Rantalainen M. Improved breast cancer histological grading using deep learning. Ann Oncol. 2022 Jan;33(1):89-98. doi: 10.1016/j.annonc.2021.09.007. Epub 2021 Sep 29. PMID 34756513
• Chow ZL, Thike AA, Li HH, Nasir NDM, Yeong JPS, Tan PH. Counting Mitoses With Digital Pathology in Breast Phyllodes Tumors. Arch Pathol Lab Med. 2020 Nov 1;144(11):1397-1400. doi: 10.5858/arpa.2019-0435-OA. PMID 32150458
• Cheng CL, Md Nasir ND, Ng GJZ, Chua KWJ, Li Y, Rodrigues J, Thike AA, Heng SY, Koh VCY, Lim JX, Hiew VJN, Shi R, Tan BY, Tay TKY, Ravi S, Ng KH, Oh KSL, Tan PH. Artificial intelligence modelling in differentiating core biopsies of fibroadenoma from phyllodes tumor. Lab Invest. 2022 Mar;102(3):245-252. doi: 10.1038/s41374-021-00689-0. Epub 2021 Nov 24. PMID 34819630
• Kates-Harbeck J, Svyatkovskiy A, Tang W. Predicting disruptive instabilities in controlled fusion plasmas through deep learning. Nature. 2019 Apr;568(7753):526-531. doi: 10.1038/s41586-019-1116-4. Epub 2019 Apr 17. PMID 30996348

Identifiers

NCT: NCT06286267 · SYSKY-2023-351-02

Primary sources (government registries)

• ClinicalTrials.gov (US NIH)

View this study on ClinicalTrials.gov ↗