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 Figure 2 – Boxplot illustrating the difference between the groups with and without cardiotoxicity. A, percentage reduction in left ventricular global longitudinal strain (GLS) variation; and B, percentage reduction in left ventricular ejection fraction (LVEF) variation. 40.00 40.00 60.00 30.00 20.00 20.00 10.00 .00 .00 –10.00 –20.00 –20.00 –40.00 No Yes No Yes Cardiotoxicity Cardiotoxicity 27 28 42 A B Percentage reduction inGLS Percentage reduction inLVEF Table 5 – Cox Regression Models. B SE p HR 95%CI Cox regression model (Univariate) Diastolic function 0.551 0.221 0.013 1.735 1.126-2.675 Left Atrial Volume (ml/m 2 ) - 0.354 0.154 0.022 0.702 0.519-0.950 LVEF (%) - 0.117 0.046 0.011 0.889 0.813-0.973 GLS (%) 1.020 0.353 0.004 2.773 1.389-5.536 Cox regression model (Multivariate - A) Left Atrial Volume (ml/m 2 ) - 0.218 0.249 0.382 0.804 0.494-1.311 LVEF (%) 0.108 0.084 0.198 1.115 0.945-1.314 GLS (%) 1.41 0.686 0.040 4.097 1.068-15.716 Cox regression model (Multivariate - B) LVEF (%) 0.143 0.103 0.163 1.154 0.944-1.412 GLS (%) 1.975 0.952 0.038 7.207 1.115-46.573 Diastolic function - 0.153 0.345 0.658 0.858 0.436-1.688 B: Coefficient; SE: Standard Error; HR: Hazard Ratio; CI: Confidence Interval; LVEF: Left Ventricular Ejection Fraction; GLS: Global Longitudinal Strain. Discussion The results of the present study showed that the LV GLS was an excellent predictor of cardiotoxicity in our population, with high efficacy for its early diagnosis. Profile of morbidity of the population studied Our population was considered to have a low morbidity profile. The incidence of the risk factors that could be related to cardiotoxicity was very low, and no statistically significant association could be demonstrated. That profile differs from that of other studies, which had cases of smoking, previous use of chemotherapy, radiotherapy, in addition to a higher frequency of systemic arterial hypertension and diabetes mellitus. 14,15 The low morbidity profile can be associated with the lower incidence of cardiotoxicity observed in our population. Patients with the highest BMI were excluded from our study, because they could be considered at higher risk for cardiotoxicity, limiting the incidence rate of that event. Definition of cardiotoxicity The definition of cardiotoxicity is fundamental, because it is not uniform in different studies, hindering the assessment of the real incidence of the event. The cardiotoxicity incidence in a systematic review published in 2014 ranged from 13% to 32%. 15 Studies published by Sawaya et al. 16 and Baratta et al. 17 have found an incidence of 20%, using the same criterion of the trastuzumab committee. Our study found a cardiotoxicity 146