Breast Arterial Calcifications as an Imaging Biomarker of Cardiovascular Risk

NCT ID: NCT07156006

Last Updated: 2025-09-04

Basic Information

• Recruitment Status: TERMINATED
• Total Enrollment: 149 participants
• Study Classification: OBSERVATIONAL
• Study Start Date: 2020-09-11
• Study Completion Date: 2024-04-29

Brief Summary

The goal of this observational study is to assess if there is an association between the presence of BAC and traditional cardiovascular risk factors and validate a Convolutional Neural Network (CNN) for the automatic segmentation of Breast Arterial Calcifications (BAC) in mammographic images. This study focuses on understanding the potential of BAC as an imaging biomarker for cardiovascular risk.

The main questions it aims to answer are:

• Is there an association between the presence of BAC and traditional cardiovascular risk factors?
• Can a CNN accurately segment BAC in mammographic images?
• What is the correlation between BAC and White Matter Hyperintensities (WMH) detected through brain MRI?

Participants in this study will be individuals who undergo mammographic screening. The main tasks participants will be asked to do include providing consent for participation and having mammographic images and a blood sample taken. The study will use a comparison group, comparing individuals with BAC to those without BAC, to assess potential effects on cardiovascular risk.

Detailed Description

Association between BAC and Cardiovascular Risk Factors

• Traditional cardiovascular risk factors will be analyzed, and statistical tests (t-test or U de Mann-Whitney) will be employed based on the data distribution.
• Multivariate analysis will be performed to determine the independent association between BAC load and cardiovascular risk factors.
• Linear regression will assess the relationship between BAC load and Framingham score, aiming for a clinically applicable model.

Development of CNN for BAC Segmentation

• Mammographic images will be acquired using a digital full-field mammography system as per clinical practice.
• Two experienced operators will manually segment the images to create a dataset for training, validation, and testing the CNN.
• About 60% of the images acquired in the first year will be used for training, and the remaining 40% will form the validation and test datasets.
• Performance evaluation of the CNN will be conducted using the Sørensen similarity index, Bland-Altman analysis, and Free Response Receiver Operating Characteristic (FROC).

Association between BAC and White Matter Hyperintensities (WMH)

• A subset of participants will undergo brain MRI to assess WMH.
• The association between BAC quantity in mammography and WMH load in MRI will be evaluated using machine learning techniques.
• Other small vessel disease markers, such as lacunar infarcts and microbleeds, will also be analyzed.

Patient Enrollment:

The study aims to enroll 600 women, considering a 1:1 ratio between cases and controls. With an estimated 50% adherence rate, it anticipates evaluating 1500 women over two years.

This comprehensive study integrates the development of advanced imaging techniques with clinical correlations to explore the potential of BAC as an imaging biomarker for cardiovascular risk assessment.

Conditions

Breast; Cardiovascular Calcification

Study Design

• Observational Model Type: CASE_CONTROL
• Study Time Perspective: PROSPECTIVE

Study Groups

BAC Group

Outpatients presenting in our department for annual mammography will be screened and selected for BAC presence.

Mammographic Imaging:

Participants will undergo mammographic imaging using a digital full-field mammography system, following standard clinical practices.

The acquired mammographic images will serve as the basis for the development and testing of the Convolutional Neural Network (CNN) for Breast Arterial Calcifications (BAC) segmentation.

Venous Blood Sample Collection:

For each participants, a venous blood sample will be collected and traditional cardiovascular risk factors (such as age, hypertension, hyperlipidemia) will be recorded.

Mammography

Intervention Type: DIAGNOSTIC_TEST

Participants will undergo mammographic imaging using a digital full-field mammography system, following standard clinical practices and blood sampling.

Control Group

Outpatients presenting in our department for annual mammography will be screened and matched for age and breast density to BAC Group.

Mammographic Imaging:

Participants will undergo mammographic imaging using a digital full-field mammography system, following standard clinical practices.

The acquired mammographic images will serve as the basis for the development and testing of the Convolutional Neural Network (CNN) for Breast Arterial Calcifications (BAC) segmentation.

Venous Blood Sample Collection:

For each participants, a venous blood sample will be collected and traditional cardiovascular risk factors (such as age, hypertension, hyperlipidemia) will be recorded.

Mammography

Intervention Type: DIAGNOSTIC_TEST

Participants will undergo mammographic imaging using a digital full-field mammography system, following standard clinical practices and blood sampling.

Interventions

Mammography

Participants will undergo mammographic imaging using a digital full-field mammography system, following standard clinical practices and blood sampling.

Intervention Type: DIAGNOSTIC_TEST

Other Intervention Names

Blood test

Eligibility Criteria

Inclusion Criteria

Female participants. Consent to undergo mammography screening. Agreement to participate in brain MRI for a subset of the study.

Exclusion Criteria

Male participants. Age below 40. Inability or unwillingness to undergo mammography screening. Contraindications for brain MRI, including the presence of pacemaker, intracranial ferromagnetic vascular clips, intraocular metallic fragments, severe claustrophobia, inability to maintain a supine position, involuntary movements, or pregnancy.

Known history of breast cancer. Previous reductive breast surgery.

• Minimum Eligible Age: 40 Years
• Eligible Sex: FEMALE
• Accepts Healthy Volunteers: No

Sponsors

IRCCS Policlinico S. Donato

OTHER

Sponsor Role: lead

Responsible Party

Francesco Sardanelli

Director

Responsibility Role: PRINCIPAL_INVESTIGATOR

Locations

IRCCS Policlinico San Donato

San Donato Milanese, MI, Italy

Countries

Italy

Other Identifiers

90/INT/2020

Identifier Type: OTHER

Identifier Source: secondary_id

BAKER

Identifier Type: -

Identifier Source: org_study_id