Improving the Quality of Radiotherapy by Multi-Institution Knowledge-Based Planning Optimization Models (Acronym: MIKAPOCo, Multi-Institutional Knowledge-based Approach in Plan Optimization for the Community)

IRCCS Ospedale San Raffaele

Status: Invitation-only

Conditions: Breast Cancer; Prostate Cancer

Treatments: Other: treatment plan comparison

Study type: Observational

Funder types: Other

Identifiers

NCT06317948

IG23150

Details and patient eligibility

About

Investigators central hypothesis is that it is possible to create libraries of "consistent" Knowledge-Based plan-models derived from large Institutional experiences. These libraries can be used to guide automated RT planning and serve as tools to assist centers for plan quality assurance (QA) and plan prediction.

Quantifying Inter-institute variability of RT planning and building libraries of interchangeable and validated multi-Institutional KB plan prediction models is expected to impact on the quality of planning at the national level. The project has the potential of facilitating the introduction of AI approaches in plan optimization, thus reducing intra and inter-Institute planning variability. Improving plan quality is expected to translate into better outcome after RT in terms of local control and, even more, of side effects and Quality of life. Positive impact is also expected in patient selection for advanced techniques, in plan audit and plan optimization in clinical trials, in technology comparison and cost-benefit analyses as well as in the RT educational field.

Full description

Major aims

1. To create libraries of consistently generated KB models for patients treated with RT for breast and prostate cancer and for selected stereotactic-body RT (SBRT) applications based on the experience of many Italian Institutions; to quantify planning inter-institute variability in homogeneous classes of patients.
2. To group models based on their characteristics and interchangeability. To assess groups of highly interchangeable models to be considered for multi-institutional dose-volume histogram (DVH) prediction purposes.

Enrollment: 1,000 estimated patients

Sex: All

Volunteers: No Healthy Volunteers

Inclusion criteria

• real life consecutive (or randomly chosen) plan data of patients treated for prostate cancer during the last 10 years;
• real life consecutive (or randomly chosen) plan data of patients treated for breast cancer during the last 10 years;
• real life consecutive (or randomly chosen) plan data of patients treated for selected SBRT situations (spine and prostate, according to RTOG 0631 and 0938 schemes respectively) during the last 10 years.

Exclusion criteria