Breast Cancer Index and prediction of benefit from extended endocrine therapy based on treatment compliance: An IDEAL study.

Abstract

531 Background: The IDEAL trial randomized hormone receptor-positive (HR+) breast cancer patients to 5 vs 2.5y of extended letrozole after completion of 5y of adjuvant endocrine therapy. In the parent trial, approximately 60% of patients overall were compliant with endocrine treatment (59% vs 74% compliance in 5y and 2.5y arms, respectively), with patient experiences as the most significant factors leading to treatment discontinuation. Breast Cancer Index is a gene expression-based signature that predicts which HR+ patients are likely to benefit from extended endocrine therapy (EET) vs those unlikely to benefit [BCI (H/I)-High and -Low, respectively]. The current study examined EET compliance and outcome by BCI (H/I) status in patients treated in the IDEAL trial. Methods: Patients with available primary tumor specimens were eligible for this blinded study. Primary endpoint was recurrence-free interval (RFI), including locoregional and distant recurrences. Kaplan-Meier and Cox proportional hazards regression analysis were used to analyze the benefit of EET. Non-compliance was defined as completion of ≤60% of treatment duration (≤3y for 5y arm; ≤1.5y for 2.5y arm) for any reason other than disease events. Overtreatment was defined as % compliant BCI (H/I)-Low patients in the 5y arm; undertreatment was % noncompliant BCI (H/I)-High patients in the 5y arm. Results: 908 HR+ patients (73% pN+, 59y, 45% pT1, 48% pT2) were included. 78% (n = 708) of patients were compliant, of which 48% (n = 338) were BCI (H/I)-High and 52% (n = 370) were BCI (H/I)-Low. In non-compliant patients (22%, n = 200), 45% (n = 91) were BCI (H/I)-High and 55% (n = 109) were BCI (H/I)-Low. BCI (H/I)-Low patients irrespective of compliance status did not derive significant benefit from EET (P = 0.922, compliant subset; 0.894 non-compliant subset). Compliant BCI (H/I)-High patients showed significant benefit from EET (HR 0.35, 95% CI 0.16-0.79; absolute benefit 11.7%; P = 0.008) whereas non-compliant BCI (H/I)-High patients did not (HR 0.75, 95% CI 0.17-3.35; absolute benefit 2.1%, P = 0.704). In this study, 38% of patients in the 5y EET arm were BCI (H/I)-Low and compliant and thus were overtreated, while 13% of patients in the 5y EET arm were BCI (H/I)-High and non-compliant and thus were undertreated. Conclusions: Patients that were BCI (H/I)-Low did not derive significant benefit from 5 vs 2.5y of EET even when compliant, and thus may be considered for treatment de-escalation. Importantly, BCI (H/I)-High patients with endocrine responsive disease showed significant improvements in outcome when compliant and should be guided to continue treatment. BCI may serve as an important genomic tool to increase EET compliance and identify patients that may be candidates for increased side effect management and support to potentially improve outcomes. Clinical trial information: NTR3077; BOOG 2006-05; Eudra-CT 2006-003958-16.