Abstract OT1-23-01: Development of a hoRizontal data intEgration classifier for NOn-invasive early diAgnosis of breasT cancEr: the RENOVATE trial

Abstract

Abstract Background: The detection of breast lesions through self-examination or during screening tests is a frequent finding. Breast biopsy is required in case of radiologically suspect lesions, bestowing a high burden on both patients and national healthcare system, since only one every four biopsies is breast cancer (BC). To date, the assessment of circulating biomarkers failed to demonstrate clinical utility in the early diagnosis of BC, for its suboptimal accuracy and difficult transferability to clinical practice. The combination of novel cutting-edge methods for the assessment of circulating analytes in an integrated multiomic classifier may overcome such limitations, possibly allowing liquid biopsy to become a novel noninvasive procedure for the differential diagnosis of BC. Design: In the RENOVATE trial (NCT04781062), women with suspect (BI-RADS-4/5) breast lesions ≤ 2 cm (cT1) are asked, before biopsy, to donate ~ 35 mL of blood collected in four dedicated tubes and ~ 50 mL of urine at the Diagnostic Senology Unit of Ospedale Policlinico San Martino (Genoa, IT). Plasma cell-free DNA methylation and copy number alterations are assessed in a cohort of patients diagnosed with early BC and a matched set of patients with histologically proven benign lesions through cell-free methylated DNA immunoprecipitation and high throughput sequencing (cfMeDIPseq), as well as ultra-low pass whole genome sequencing (ULP-WGS). Thanks to the volume and quality of our sample set, other experimental techniques will be tested as well. Results from cfMeDIP-seq and ULP-WGS, possibly in combination with other findings, will be integrated in a unique classifier for the noninvasive differential diagnosis of suspect breast lesions. Eligibility criteria: Patients with radiologically suspect breast lesions ≤ 2 cm (i.e. BIRADS 4/5) are eligible. Patients with previous history of cancer, or diagnosed with autoimmune or active allergic diseases, acute or chronic hepatic, renal, or cardiac diseases, or acute or chronic infectious diseases are excluded from the present trial. Specific aims: The primary aim of the present trial is to develop a noninvasive classifier for the differential diagnosis of suspect breast lesions detected through mammography and/or ultrasound. For such purpose we will assess the performance of plasma cfMeDIPseq, ULP-WGS, and other promising techniques for the differential diagnosis of BC. Such techniques will be integrated in a unique classifier in order to reach the maximum possible accuracy. Statistical methods: Sample size was calculated with a semi-parametric simulation-based approach from beta-distributions of PBMC datasets: assuming to test 20,000 CpG regions, with 300 differentially methylated target CpGs, a target maximal difference in DNA methylation of 0.2 between groups and an FDR of 0.05, 1 – beta ~ 0.90 would be achieved with an overall sample size of 150 samples split in a 1:2 ratio. Target accrual and present accrual: Minimum target accrual is set at 49 patients with BC and 98 patients with benign lesions. To date, we have collected plasma samples from 74 eligible patients with BC and 115 eligible patients with benign lesions. A validation cohort accounting for ~30% of our sample set will be recruited at Istituto Nazionale dei Tumori (Milan, IT). Contact information: For further information, please contact Gabriele Zoppoli at gabriele.zoppoli@unige.it. Citation Format: Francesco Ravera, Martina Dameri, Isabella Lombardo, Mario Stabile, Alberto Tagliafico, Massimo Calabrese, Alberto Ballestrero, Lorenzo Ferrando, Gabriele Zoppoli. Development of a hoRizontal data intEgration classifier for NOn-invasive early diAgnosis of breasT cancEr: the RENOVATE trial [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr OT1-23-01.