Identification of clinical, pathological and molecular predictors of prognosis in early-stage invasive lobular breast cancer: role of proliferation and CDK4/6 pathway.

Differently from invasive ductal carcinoma (IDC) of the breast, pathological and molecular factors that guide the prognosis of invasive lobular carcinoma (ILC) are not completely known. In this regard, the prognostic and/or predictive impact of potential drivers needs to be elucidated, in order to create a global portrait for ILC patients. Among these factors, the role of proliferation, commonly evaluated by the Ki67 antigen, represents a relevant aspect in the choice of adjuvant treatment for breast cancer, in particular in the luminal setting. This retrospective multicentric analysis, including 679 patients with early-stage resected lobular histology and comparing these with 418 patients affected by IDC, showed that a value of 4%-5% represents the best Ki67 cut-off of able to significantly discriminate the prognosis of patients affected by ILC. Regarding the role of adjuvant chemotherapy, to date the main guidelines do not recommend a different treatment approach according to histology, given the absence of randomized studies conducted in the ILC setting. From retrospective analyses, the benefit of adjuvant chemotherapy in addition to hormonotherapy for luminal ILC is still unclear. In this regard, our propensity score analysis conducted on 473 patients affected by luminal ILC showed a significant difference in terms of overall survival between hormonotherapy alone and hormonotherapy plus adjuvant chemotherapy (5- and 10-year 96.3% vs. 86.0% and 92.2% vs. 67.5%, respectively). Nowadays, one of the main emerging research strategies in cancer is based on the study of the genome of exceptional responder and prognostic ‘outlier’ patients. Adopting this strategy, we retrospectively analysed a multicenter series of nearly 500 ILC patients underwent surgical resection and we built one of the first risk classification model for ILC, subsequently validated in a cohort of 282 patients. This model, based on a combination of simple and easily available clinical-pathological parameters, was able to effectively stratify ILC patients in prognostic risk classes, with a good accuracy. Once identified the ‘best’ and ‘worst’ prognostic performers, we investigated their molecular portrait, principally by next-generation sequencing (NGS), and their expression profile to identify recurrent molecular alterations and explore their association with prognosis, in a preliminary cohort of 20 patients with good prognosis and 14 with poor prognosis. Overall, the most frequent mutated gene was the CDH1 gene (38.2%), followed by PIK3CA (29.4%) and TP53 (20.6%), while the loss of CDH1 (44.1%) and ARID1A (38.2%) were the most frequent copy number variation events. The molecular alterations were distributed regardless of the prognosis, except for the gain of CDK4, exclusively present in the poor prognosis subgroup (35.0%, p=0.03; Odds Ratio 7.98, 95%CI 1.51-42.1, p=0.014). The final aim of the overall project was to evaluate the molecular profile of 'outliers' resected ILC using modern technologies in order to identify those molecular aberrations that could potentially predict the probability of recurrence. This integrated and multi-step analysis suggested that the CDK4/6 pathway may have a significant biological impact on the ILC prognosis. Certainly, this hypothesis needs to be prospectively validated in a larger series.