• cardiotoxicity
• strain imaging.

Life-expectancy for patients with cancer is steadily improving, with an increasing rate of treatment related complications. Antineoplastic therapy employed in cancer treatment is frequently complicated by the development of cardiotoxicity. Cardiovascular complications can be different, ranging from heart failure, myocardial ischemia or infarction to hypertension, arrhythmias and thromboembolism. It is therefore mandatory to early recognize and treat cardiovascular side effects related to chemotherapy drugs. Echocardiography has been and it is still the cornerstone in the diagnosis and follow-up of cardiac dysfunction due to its availability, safety and versatility. The most used parameter to evaluate cardiac dysfunction in this context is the left ventricular ejection fraction (LVEF) and current European Society of Echocardiography and American Society of Echocardiography guidelines for the evaluation of adult patients during and after cancer therapy are based on LVEF. However this parameter has several limits and may mask cardiac dysfunction. Recently developed strain imaging may provide a more sensitive and early detection of altered left ventricular function.
In the present study we evaluated 53 patients (age 53 ± 13 years, women 64%) on cancer drug therapies referred to the cardio-oncology clinic of the Royal Brompton Hospital in London. All patients had 2D echo-derived LVEF ≥55% and echo images suitable to speckle tracking analysis for global longitudinal (GLS), circumferential (GCS) and radial strain (GRS). Concomitant troponin I and BNP levels were measured and 43 (81%) patients underwent CMR imaging. The 2D strain data were compared to 25 healthy age matched controls using the student t test. Conventional echocardiographic parameters were substantially normal, including diastolic measurements All strain parameters were significantly lower in patients on cancer drug therapies compared to controls. In the cancer population (n=53) we found mean global peak systolic values of global longitudinal strain (GLS) global circumferential strain (GCS) and global radial strain (GRS) respectively of -19.8 ± 3.3%, -23.4 ± 4.8% and 29.7 ± 14%. In the 25 controls strain imaging values were respectively -22 ± 2% for GLS, -29.5 ± 5% for GCS and 42 ± 10% for GRS.
CMR showed loss of torsion and/or fibrosis in 10(23%) patients with no correlation to strain values. BNP was elevated in 34(64%) patients with no differences in strain values compared to those with normal BNP values. Troponin I was elevated in only one patient.
Intra-observer reproducibility carried out in a subgroup of 20 random selected cancer patients revealed good correlation for global longitudinal and circumferential strain (ICC 0.8; r=0.7) and moderate correlation for global radial strain (ICC 0.7; r=0.6) while inter-observer variability showed moderate correlation for all the three parameters (ICC 0.7; r=0.5).
In our study cancer patients on current chemotherapy drugs with preserved left ventricular ejection fraction (LVEF ≥ 55%) showed significantly lower GLS, GCS and GRS values when compared to our echo-lab normal reference values (p<.05).
In our study, despite having a normal LVEF, patients on cancer drug therapies had evidence of sub-clinical myocardial dysfunction affecting all myocardial layers and in particular radial function. Strain can be therefore a sensitive tool to detect early chemotherapy related cardiotoxicity and to prevent morbidity and mortality through close follow-up and appropriate cardiac therapy. CMR findings and BNP levels may provide additional complementary information. The clinical relevance of these findings requires further study.