ABC | Volume 110, Nº2, February 2018

Original Article Gripp et al Global Longitudinal Strain Accuracy for Cardiotoxicity Arq Bras Cardiol. 2018; 110(2):140-150 Table 1 – General characteristics of the population included in the study and excluded from it. Variable population included n= 49 population excluded n = 9 p Age (years) * ll 49.7 ± 12.2 51.0 ± 12.9 0.78 Weight (kg) * ll 67.6 ± 12.6 90.5 ± 12.5 < 0.05 Height (m) * ll 1.5 ± 0.06 1.5 ± 0.09 0.75 BSA (m 2 ) * ll 1.65 ± 0.2 1.9 ± 0.2 < 0.05 BMI (kg/m 2 ) † § 26.1 (23.6 - 30.4) 37.9 (31.6 - 40.9) < 0.001 SBP (mm Hg) * ll 125.1 ± 17.4 132.2 ± 12.0 0.25 DBP (mm Hg) *§ 74.7 ± 12.0 84.4 ± 5.3 0.02 HR (bpm) * ll 77.2 ± 10.1 83.4 ± 13.7 0.12 EF (Teicholz - %) * ll 69.0 ± 0.7 67.7 ± 9.3 0.59 Total AnthracyclineDose (Equivalence) (mg/m 2 ) †§ 600 (534-760) 600 (507-590) 0.68 Total Traztuzumab Dose (mg/m 2 ) * ll 6823.3 ± 2395.6 7079 ± 2207.6 0.88 SAH ‡ 16 (32.7) 4 (44.4) 0.37 Type II DM ‡ 2 (4.1) 0 0.71 Beta-blocker ‡ 0 1 (11.1) 0.15 ACEI / ARB ‡ 0 1 (11.1) 0.15 ASA ‡ 2 (4.1) 0 0.71 HCTZ ‡ 14 (28.6) 3 (33.3) 0.52 Statin ‡ 3 (6.1) 0 0.59 Right breast CA ‡ 24 (49.0) 8 (88.9) 0.03 Left breast CA ‡ 25 (51.0) 2 (22.2) 0.11 Invasive ductal carcinoma ‡ 34 (69.4) 9 (100) Lobular carcinoma ‡ 7 (14.3) 0 0.16 Other subtypes ‡ 8 (16.3) 0 Pre-chemo surgery ‡ 20 (40.8) 3 (33.3) 0.49 Radiotherapy ‡ 26 (53.1) 7 (77.8) 0.16 Doxorubicin ‡ 20 (40.8) 2 (22.2) 0.25 Epirubicin ‡ 29 (59.2) 7 (77.8) 0.25 Trastuzumab ‡ 8 (16.3) 2 (22.2) 0.48 * Mean (standard deviation); † Median (25 th - 75 th Percentile); ‡ N (%); BSA: Body Surface Area; BMI: Body Mass Index; SBP: Systolic Blood Pressure; DBP: Diastolic Blood Pressure; HR: Heart Rate; EF: Ejection Fraction; SAH: Systemic Arterial Hypertension; DM: Diabetes Mellitus; ACEI: Angiotensin-Converting-Enzyme Inhibitor; ARB - Angiotensin Receptor Blocker; ASA: Acetylsalicylic Acid; HCTZ: Hydrochlorothiazide; CA: Cancer; Chemo: Chemotherapy; + Median (25 th – 75 th Percentile); bpm: beats per minute. Categorical variables compared by use of chi-square test ‡ , p value ≤ 0.05. Continuous variables compared by use of Mann Whitney U test § or Student t test ll, p value ≤ 0.05. When assessing the percentage reduction in the LV GLS, from baseline to the third month, between patients with and without cardiotoxicity, a clear difference is observed between the two groups (Figure 2A). That same behavior was not observed when assessing the percentage reduction in the EF in the same period, confirming that EF is not as sensitive as GLS to diagnose cardiotoxicity (Figure 2B). The RV strain and RV free wall strain were acquired by using the same software developed for the analysis of the left ventricle, and showed mild non-significant changes on the third and sixth months, with subsequent normalization. However, TAPSE and tissue Doppler of the tricuspid annulus, measures related to the right ventricle, did not change during the follow-up. Predictors of cardiotoxicity Aiming at assessing the association of each echocardiographic variable with cardiotoxicity (outcome), Cox regression analysis was performed (Table 5). The variables with p ≤0.05 on Cox regression univariate analysis went to multivariate analysis of independent predictors of cardiotoxicity: EF (Simpson’s method), LV GLS on the thirdmonth, left atrial volume, and diastolic function. Two models were created, separating the information of left atrial volume and diastolic function, because both variables express similar information, and can be interpreted in the concept of collinearity. Only LV GLS on the third month remained an independent predictor of cardiotoxicity, 143