Abstract IA002: DCIS: Pathological heterogeneity and prognosis definition

Abstract

Abstract Ductal carcinoma in situ represents 15 to 20% of all breast carcinomas and is a non-obligate precursor of invasive carcinomas. Clinically DCIS presents most frequently as microcalcifications detected on screening mammography. DCIS is a proliferation of neoplastic cells confined to the ductulo-lobular tree. Its morphology can recapitulate the diversity of that of invasive carcinomas. DCIS lesions are classified according to the nuclear grade (low, intermediate, high), the presence of necrosis and microinvasion. In resection specimens, the extent of the disease and the quality of the margins are key element for patients management as well as the association of DCIS to the microcalcifications. The reproducibility of the DCIS grading system is relatively poor and promising help emerges from deep learning approaches. The four molecular classes of invasive carcinomas exist but with different proportion in DCIS. Triple negative represent 5-8% of DCIS. Conversely, HER2 overexpression is more frequently observed in particular in high grade DCIS (up to 60% of the cases). The analysis of Estrogens and progesterone receptors is not systematic in clinical practice unless an endocrine therapy may be decided after the patient has completed the loco-regional treatment. At initial diagnosis, it remains very challenging to predict the clinical outcome of the lesion. However, after complete surgical excision followed by whole breast irradiation therapy, the prognosis is excellent with a mortality of 3.7%. The rate of relapse is 7% at 5 years and 16% at 10 years of follow-up. Concerns about overdiagnosis and overtreatment exist worldwide. Retrospective studies identified older age, mammographic detection, small size, non-high grade histology and negative margins at excision associated with favorable outcome. Unfortunately, de-escalation clinical prospective trials based on these criteria, failed to identify the patients presenting DCIS of good prognosis. Biological parameters have been retrospectively associated with a poor prognosis (HER2 amplification, proliferation, RB loss, COX2 expression…). The molecular signature Oncotype DCIS®, has been validated retrospectively in a subgroup of patients from the ECOG E5194, and determined DCIS associated with different risk of relapse. Its prospective validation is on-going in a non randomized trial (DUCHESS). Three on-going trials based in part on clinical criteria, one adding estrogen receptors expression, the other comedonecrosis, propose after a diagnostic biopsy, a simple surveillance. Another trial, leaded by UNITRAD/UNICANCER (France), Romance is a prospective randomized study of omission of whole-breast radiotherapy following breast-conserving surgery in patients with very low risk luminal A ductal carcinoma in situ and enrolment is on. Research efforts is moving into the integration the ecosystem of DCIS (myoepithelial cells, stroma, immune cells and carcinomatous cells) to identify more accurately the DCIS associated with a higher aggressiveness that would allow secure de-escalation clinical trials. Citation Format: Anne Vincent-Salomon. DCIS: Pathological heterogeneity and prognosis definition [abstract]. In: Proceedings of the AACR Special Conference on Rethinking DCIS: An Opportunity for Prevention?; 2022 Sep 8-11; Philadelphia, PA. Philadelphia (PA): AACR; Can Prev Res 2022;15(12 Suppl_1): Abstract nr IA002.