High-efficiency platforms for personalized cervix cancer treatment

Developing Clinical High Efficiency Platforms for Individualised Treatment Through Integration of Advanced Radiation Technology, Quantitative Imaging and Molecular Biology and Machine Learning for Treatment of Cervix Cancer.

This project will use past patient imaging and treatment records to build machine-learning tools that predict tumor response and normal tissue side effects for people treated for cervix cancer.

Quick facts

• Study type: Observational
• Enrollment: 1800 (estimated)
• Ages: 18 Years to 90 Years
• Sex: Female
• Sponsor: Tata Memorial Hospital (Government)
• Drugs / interventions: radiation
• Locations: 1 site (Navi Mumbai, Maharashtra)
• Trial ID: NCT05102240 on ClinicalTrials.gov

What this trial studies

This retrospective observational project uses imaging (MRI and CT), treatment plans, and outcome data from completed and ongoing cervix cancer radiotherapy and brachytherapy protocols. Machine learning will be applied to automate tumor and organ-at-risk delineation, to develop normal tissue complication probability (NTCP) and tumor control probability (TCP) models, and to extract quantitative texture features for response prediction. The team will also create a library of proton beam plans to compare achievable doses and identify patients likely to benefit from proton therapy. Models will be trained and validated on the available institutional datasets and linked clinical endpoints.

Who should consider this trial

Good fit: Patients with cervix cancer who were treated within completed or ongoing chemoradiation or brachytherapy protocols and who have MRI/CT imaging and documented outcome data in the hospital database are appropriate for inclusion.

Not a fit: Patients without available imaging or outcome records in the hospital database or those treated outside the radiotherapy/brachytherapy protocols used in the datasets are unlikely to be included or to benefit from the models.

Why it matters

Potential benefit: If successful, the tools could enable more precise targeting, better prediction of toxicities and responses, and help select patients who may benefit from intensified or proton therapy.

How similar studies have performed: Machine-learning methods for auto-contouring and toxicity prediction have shown promise in other radiotherapy settings, but comprehensive integration for cervix cancer and proton-plan selection is still relatively novel.

Eligibility criteria

Inclusion Criteria:

For Aim 1 and Aim 3:

* Patients treated within ongoing and completed clinical trials of chemoradiation and brachytherapy for cervix cancer with access to MRI/CT images at the time of diagnosis and brachytherapy For Aim 2
* Patients undergoing postoperative or definitive radiotherapy and treated within trials of postoperative or definitive RT.

Exclusion Criteria:

1. Lack of disease or toxicity outcomes.
2. Lack of images in the hospital database.

Where this trial is running

Navi Mumbai, Maharashtra

• Advanced Centre of Treatment Research and Education In Cancer,Tata Memorial Centre — Navi Mumbai, Maharashtra, India (Recruiting)

Study contacts

• Principal investigator: Supriya Sastri (nee Chopra), MD — Tata Memorial Centre, The Advanced Centre for Treatment, Research and Education in Cancer (ACTREC)
• Study coordinator: Supriya Sastri (nee Chopra), MD

Conditions Cervix Cancer; Advanced Radiation Technology; Quantitative Imaging; Molecular Biology; Machine Learning