Post-Neoadjuvant Treatment MRI Based AI System to Predict pCR for Rectal Cancer (MR-AI-pCR)

A Post-Neoadjuvant Treatment MRI Based AI System to Predict Pathologic Complete Response for Patients With Rectal Cancer: A Multicenter, Prospective Clinical Study

In this study, investigators seek for a better way to identify the potential pathologic complete response (pCR) patients form non-pCR patients with locally advanced rectal cancer (LARC), based on their post-neoadjuvant treatment Magnetic Resonance Imaging (MRI) data.

Previously, a post neoadjuvant treatment MRI based radiomics AI model had been constructed and trained. Here, the predictive power of this artificial intelligence system and expert radiologist to identify pCR patients from non-pCR LARC patients will be compared in this prospective, multicenter, back-to-back clinical study

Study Overview

• Status: Recruiting
• Conditions: Rectal Cancer
• Intervention / Treatment: Procedure: artificial intelligence prediction system; Procedure: the radiologists

Detailed Description

This is a multicenter, prospective, observational clinical study for seeking out a better way to predict the pathologic complete response (pCR) in patients with locally advanced rectal cancer (LARC) based on the post-neoadjuvant treatment Magnetic Resonance Imaging (MRI) data. Patients who have been pathologically diagnosed as rectal adenocarcinoma and defined as clinical II-III stage will be enrolled from the Sixth Affiliated Hospital of Sun Yat-sen University, Sir Run Run Shaw Hospital and the Third Affiliated Hospital of Kunming Medical College. All participants should follow a standard treatment protocol, including neoadjuvant treatment, total mesorectum excision (TME) surgery. Patients with LARC who received neoadjuvant treatment will be enrolled and their post-neoadjuvant treatment MRI images will be used to predict their pathologic response (pCR vs. non-pCR). The artificial intelligence prediction system and the expert radiologist will define the pathologic response as pCR or non-pCR, respectively. The pathologist will provide the final pathology report of TME surgery specimen (pCR or non-pCR) as a standard. The predictive efficacy of these two back-to-back approaches generated will be compared in this multicenter, prospective clinical study.

• Study Type: Observational
• Enrollment (Anticipated): 205

Contacts and Locations

• Study Contact: Name: Xinjuan Fan, MD, PhD; Phone Number: 020-38254037; Email: fanxjuan@mail.sysu.edu.cn

Study Locations

• China
  • Guangdong
    • Guangzhou, Guangdong, China, 510655
      • Recruiting
      • The Sixth Affiliated Hospital of Sun Yat-sen University
  • Yunnan
    • Kunming, Yunnan, China, 650000
      • Recruiting
      • The Third Affiliated Hospital of Kunming Medical College
  • Zhejiang
    • Hangzhou, Zhejiang, China, 310000
      • Recruiting
      • Sir Run Run Shaw Hospital

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: 18 years to 75 years (ADULT, OLDER_ADULT)
• Accepts Healthy Volunteers: No
• Genders Eligible for Study: All

• Sampling Method: Non-Probability Sample

Study Population

In the study, the population are the patients with LARC, who receive neoadjuvant chemoradiotherapy or chemotherapy and TME surgery. The response of neoadjuvant treatment is unknown.

Description

Inclusion Criteria:

• pathologically diagnosed as rectal adenocarcinoma
• defined as clinical II-III staging (≥T3, and/or positive nodal status) without distant metastasis
• receive neoadjuvant chemoradiotherapy or chemotherapy
• pre- and post-neoadjuvant treatment MRI data obtained
• receive total mesorectum excision (TME) surgery after neoadjuvant therapy and get the pathologic assessment of tumor response

Exclusion Criteria:

• with history of other cancer
• insufficient imaging quality of MRI to delineate tumor volume or obtain measurements (e.g., lack of sequence, motion artifacts)
• not completing neoadjuvant chemotherapy or chemoradiotherapy
• tumor recurrence or distant metastasis during neoadjuvant treatment
• not undergoing surgery resulting in lack of pathologic assessment of tumor response

Study Plan

How is the study designed?

Number of groups / cohorts

1

Cohorts and Interventions

Group / Cohort

Intervention / Treatment

patients will be evaluated by artificial intelligence system and expert radiologist; the patients with locally advanced rectal cancer (LARC) finished the neoadjuvant treatment, and not yet receive total mesorectum excision (TME) surgery will be enrolled. The post-neoadjuvant treatment MRI images features of each enrolled patients will be captured by the artificial intelligence system, and evaluated by experienced radiologists as well. Blind to the pathologic report of TME specimen, both approaches further respectively yield a predicted pathologic response to neoadjuvant treatment for each enrolled patient, shown as pCR or non-pCR.

Procedure: artificial intelligence prediction system; The tumor ROI in the post- neoadjuvant treatment MRI images will be manually delineated, and further subjected to the AI prediction system arm to verify the predictive accuracy of this AI prediction system in identifying the pCR individuals from non-pCR patients with LARC.; Procedure: the radiologists; The enrolled patients will be assigned to the trained experienced radiologists to evaluate their predictive accuracy in identifying the pCR individuals from non-pCR patients

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
The area under curve (AUC) of Receiver Operating Characteristic (ROC) curves of AI prediction system and expert radiologists in prediction tumor response	The area under curve (AUC) of Receiver Operating Characteristic (ROC) curves of AI prediction system and expert radiologists in identifying the pCR candidates from non-pCR individuals among neoadjuvant chemotherapy or chemoradiotherapy treated LARC patients will be calculated respectively.	baseline

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
The specificity of AI prediction system and expert radiologists in prediction tumor response	The specificity of AI prediction system and expert radiologists in identifying the pCR candidates from non-pCR individuals among neoadjuvant chemotherapy or chemoradiotherapy treated LARC patients will be calculated respectively.	baseline
The sensitivity of AI prediction system and expert radiologists in prediction tumor response	The sensitivity of AI prediction system and expert radiologists in identifying the pCR candidates from non-pCR individuals among neoadjuvant chemotherapy or chemoradiotherapy treated LARC patients will be calculated respectively.	baseline

Collaborators and Investigators

• Sponsor: Sixth Affiliated Hospital, Sun Yat-sen University
• Collaborators: Sir Run Run Shaw Hospital; The Third Affiliated Hospital of Kunming Medical College.
• Investigators: Study Chair: Xiangbo Wan, MD, PhD, Sixth Affiliated Hospital, Sun Yat-sen University; Principal Investigator: Weidong Han, MD, PhD, Sir Run Run Shaw Hospital; Principal Investigator: Zhenhui Li, MD, The Third Affiliated Hospital of Kunming Medical College.

Study record dates

Study Major Dates

• Study Start (ACTUAL): February 8, 2020
• Primary Completion (ANTICIPATED): December 10, 2022
• Study Completion (ANTICIPATED): March 31, 2023

Study Registration Dates

• First Submitted: February 19, 2020
• First Submitted That Met QC Criteria: February 19, 2020
• First Posted (ACTUAL): February 20, 2020

Study Record Updates

• Last Update Posted (ACTUAL): October 26, 2022
• Last Update Submitted That Met QC Criteria: October 25, 2022
• Last Verified: October 1, 2022

More Information

Terms related to this study

• Keywords: Locally advanced rectal cancer; Neoadjuvant treatment; Pathologic complete response; Radiomics features; Artificial intelligence model
• Additional Relevant MeSH Terms: Digestive System Diseases; Neoplasms; Neoplasms by Site; Gastrointestinal Neoplasms; Digestive System Neoplasms; Gastrointestinal Diseases; Intestinal Diseases; Intestinal Neoplasms; Rectal Diseases; Colorectal Neoplasms; Rectal Neoplasms
• Other Study ID Numbers: MR-AI-pCR 2020

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: NO

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No