Abstract P4-15-02: Clinical, biochemical, and genomic predictors of trastuzumab-related cardiotoxicity: Results of CATS

Abstract

Abstract Background: Adjuvant therapy for HER2-positive breast cancer (HBC) often comprises anthracyclines (A) followed by trastuzumab (T). Trastuzumab-related cardiotoxicity (TRC) is T’s primary toxicity, and led 15% of patients in NSABP-B31 to stop T early. Earlier detection of TRC might enable timely therapeutic intervention, lowering the number of patients ceasing T prematurely. Small, retrospective, unvalidated studies have proposed various markers as predictors for TRC. The Cardiotoxicity of Adjuvant Trastuzumab Study (CATS) aims to assess a panel of clinical, biochemical, and genomic/immunologic markers as predictors for TRC. Methods: HBC patients with a left ventricular ejection fraction (LVEF) ≥ 50% scheduled to receive adjuvant A followed by 12 months’ T were eligible. Serum N-terminal pro-B type natriuretic peptide (NT pro-BNP) and troponin I were measured centrally at baseline (T1), at completion of A (T2), and after 3 and 6 months of T. LVEF was measured at T1, T2, and then 3-monthly until completion of T. Germline SNPs in ERBB2 (V655I), FCGR2A (H131R), and FCGR3A (V158F) were assessed. TRC was defined as: cardiac death, NYHA class 3/4 heart failure, grade 3/4 arrhythmia/ischemia, drop in LVEF >15% from baseline, or drop in LVEF of >10% to <55%. 1 patient with LVEF <50% at T2 was excluded. Uni- and multivariate logistic regression were used to determine features associated with TRC. Serum markers were added to the model using forward selection. Results: 17 centers enrolled 222 patients (217 evaluable). Patient characteristics are shown. Baseline characteristicsAge [median (range), yrs]52 (28-77)Baseline LVEF [median (range)]64% (50-82%)Left-sided tumors [n(%)]105 (48%)History of hypertension [n(%)]48 (22%)Doxorubicin [n(%), median total dose]80 (37%), 242mg/m2Epirubicin [n(%), median total dose]137 (63%), 300mg/m2Concurrent taxane with T [n(%)]217 (100%) TRC occurred in 33 patients (15.2%). In multivariate models, lower baseline LVEF (OR 3.23 for a 5% lower LVEF, [95% CI 1.96-5.33], p<0.0001) and greater absolute decline in LVEF from T1 to T2 (OR 3.25 for each 5% drop in LVEF, [95%CI 1.83-5.79], p<0.0001] were each independent predictors of TRC. NT pro-BNP increased from T1 to T2 (median Δ=1.31pmol/L, p<0.0001) and then fell to baseline levels. Troponin I increased from T1 to T2 in 64.5% of patients (p<0.0001) and remained above baseline 6 months post-A (p<0.0001). NT pro-BNP and troponin I at T1 were not predictive for TRC. There were trends for the absolute increase in NT-pro BNP and troponin I from T1 (pre-A) to T2 (post-A) to be associated with risk of TRC (p=0.08 and 0.09) in univariate analyses. Germline SNPs were not predictive of TRC. Conclusion: This is the largest prospective study analyzing predictors for TRC, and the first studying immune-related SNPs in this context. Consistent with the literature, lower LVEF at baseline increases risk of TRC. Previously unreported, the absolute decrease in LVEF after A is also predictive for TRC. The persistent elevation of troponin I 6 months after completing A demonstrates chronic cardiac stress during therapy. Ongoing modeling analyses (to be presented at the meeting) will determine whether longitudinal changes in serum biomarkers can predict subsequent development of TRC. Citation Format: Shom Goel, Hao Guo, William Barry, Bronwyn Murray, Jodi Lynch, Patricia Bastick, Lorraine Chantrill, Belinda Kiely, Richard Bell, Ehtesham Abdi, Josie Rutovitz, Ray Asghari, Anne Sullivan, Michelle Harrison, Maija Kohonen-Corish, Jane Beith. Clinical, biochemical, and genomic predictors of trastuzumab-related cardiotoxicity: Results of CATS [abstract]. In: Proceedings of the Thirty-Seventh Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2014 Dec 9-13; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2015;75(9 Suppl):Abstract nr P4-15-02.