Abstract PD4-06: Investigation of the recurrence dynamics of breast cancer (BC) according to the body mass index (BMI)

Abstract

Abstract Background: In cancer follow-up (FU), in addition to the evaluation of disease free and overall survival probabilities, there is a fundamental need of assessing the recurrence dynamics. In BC as well as in other cancers, the hazard function for first recurrence presents multiple peaks, with a first major peak occurring before three years of FU. Although the baseline risk is modulated by known prognostic factors with possible time dependent effects such as the estrogen receptor (ER), so far no other factor proved to disentangle this multi-peak behavior. Here, we postulated that adiposity, which is closely related to a state of hyperinsulinemia and chronic inflammation, and reflected by increased patient's BMI, could influence the recurrence dynamics. Material and methods: In this study 777 patients with early node-positive BC from a phase III randomized clinical trial were considered (Piccart et al. JCO 2001). The trial compared intermediate or full doses of epirubicin–cyclophosphamide with cyclophosphamide, methotrexate and 5-fluorouracil. BMI was calculated using the WHO classification and was available for 734 patients, of whom 27(4%) were underweight, 377(51%) normal, 213(29%) overweight and 117(16%) obese. Underweight and normal patients were grouped together. Disease free survival (DFS), loco-regional and distant recurrence endpoints were considered. Median FU-time was 15.4 years. Cox regression analysis was performed, adjusting for standard clinico-pathological variables and treatment. Piecewise exponential models with cubic natural and regularized tensor product splines were carried out to estimate the hazard function according to categorical and continuous BMI, respectively. Results: Older age at diagnosis, postmenopausal status, and increased tumor size were significantly associated with increased BMI. Adjusted Cox models supported the association between overweight and disease recurrence (HR=1.39; 95%CI=1.05-1.84) as well as distant metastases (HR=1.41; 95%CI=1.01-1.97). There was no evidence of association for loco-regional recurrences. We observed a multi-peak behavior of distant recurrences for all BMI categories. Although, there was no shift of the first peak of recurrences according to BMI categories, occurring at ∼2.5 years of FU, a major increase in peak heights for the overweight and obese patients was evident. Obese patients showed a sharper first peak. When considering the three BMI categories according to the ER-status, we observed the worst prognosis for overweight ER-negative patients, as well as different recurrence patterns (Table). NormalOverweightObeseER-positiveMain recurrence peak culminating at 2 years, with multiple smaller peaks laterOne broad peak around 3 years and one smaller around 10 yearsOne narrow peak at 2 years and one smaller around 10 yearsER-negativeSimilar pattern to ER-positive BC, with however major increase in height of first peakOne broad peak around 3 years, with a major increase in height compared to ER-positive BCOne peak at 2-3 years Conclusion: This is the first study to show that both the ER-status of the tumor and the patient's BMI at diagnosis are influencing the recurrence dynamics related to BC dormancy. Citation Format: Biganzoli E, Desmedt C, Fornili M, Three-arms Investigators, de Azambuja E, Di Leo A, Sotiriou C, Piccart M, Demicheli R. Investigation of the recurrence dynamics of breast cancer (BC) according to the body mass index (BMI) [abstract]. In: Proceedings of the 2016 San Antonio Breast Cancer Symposium; 2016 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2017;77(4 Suppl):Abstract nr PD4-06.