Abstract Aims Patients affected by Philadelphia chromosome+ chronic myeloid leukaemia (Ph+CML) undergoing to therapy with tyrosine kinase inhibitors (TKIs) are prone to develop cardiovascular complications, which have relevant prognostic implications. Speckle-tracking echocardiography, allowing strain and myocardial work analyses, can be useful in the early detection of cardiac toxicity. Aim of our study was to assess the cardiotoxic effects of TKIs. Methods We evaluated, at baseline and during FU, 20 patients affected by Ph+ CML (59.7 ± 12.2 years, 13 males), treated Imatinib (52.6%), Nilotinib (36.8%), Ponatinib (5.3%), Dasatinib (5.3%). We measured systolic and diastolic blood pressure (SBP-DBP) and calculated corrected QT interval (QTc). In addition, we analysed echocardiographic parameters including left ventricular ejection fraction (LVEF), global longitudinal strain (GLS), global work index (GWI), global constructive work (GCW), global wasted work (GWW), global work efficiency (GWE), and peak left atrial longitudinal strain (PALS). Cardiovascular (CV) events that we considered were symptomatic or asymptomatic LV dysfunction, acute coronary syndrome (ACS), peripheral artery disease (PAD), and arrhythmias. Results Follow-up (FU) time was 3.4 ± 1 years. Most of patients (63.2%) had cardiovascular risk factors, including arterial hypertension (50%), type2 diabetes mellitus (15%), dyslipidaemia (40%) and cigarette smoking (15%). At the end of FU, SBP was unchanged (128.9 ± 19.6 mmHg vs. 129.1 ± 9.8 mmHg; P=NS) whereas DBP increased (69.4 ± 8.5 mmHg vs. 75 ± 7.7 mmHg; P = 0.004); moreover QTc was longer than baseline (404.4 ± 20.1 ms vs. 424.3 ± 29.8 ms; P < 0.001) and LVEF showed a significant decrease (62.2 ± 3.9% at baseline vs. 59.3 ± 4.8% at FU; P = 0.003); similarly, GCW (2444.1 ± 540mmHg% vs. 2234.7 ± 179.4 mmHg%; P = 0.034), GWI (2158.1 ± 589.6 mmHg% vs. 1923.1 ± 174.5 mmHg%; P = 0.022) and PALS (36.3 ± 17.1% vs. 32.8 ± 9.7%; P = 0.002) decreased during cancer therapy. On the other hand, GLS (−18.6 ± 3.1% vs. −19.4 ± 1.1%; P=NS), GWE (94.3 ± 4.1% vs. 93.6 ± 3.6%; P=NS) and GWW (120.6 ± 94.3 mmHg% vs. 106.3 ± 68.9 mmHg%; P=NS) did not change significantly. CV events were observed in 66.7% of the study population. These were mostly represented by ACS, atrial arrhythmias and symptomatic LV dysfunction (30.7% for each) and, to a lesser extent, PAD (7.6%). By comparing patients with events (group A) with those without events (group B) we found that differently from group B, group A showed during FU a significant increase of DBP (from 66 ± 5.2 mmHg to 71.2 ± 6.1 mmHg, P = 0.010; vs. group B= from 76 ± 12.mmHg to 78 ± 4.1 mmHg, P=NS) and a significant QTc prolongation (from 415.7 ± 16.1 ms to 441 ± 29.8 ms, P < 0.001; vs. group B= from 390.4 ± 19.3 ms to 405.6 ± 23.3 ms, P=NS); as to echocardiographic parameters, we found, in patients with CV events, a significant decrease of: LVEF (from 62.7 ± 4.7% to 58.8 ± 4.3%, P = 0.004; vs. group B from 61.4 ± 2.8% to 60 ± 1.7%, P=NS), GCW (from 2566.2 ± 669.6 mmHg% to 2230.1 ± 199.4 mmHg%, P = 0.021, vs. group B 2194 ± 167.5 mmHg% to 2212.6 ± 160mmHg%, P=NS) and PALS (from 36.1 ± 17% to 29.6 ± 6.6%, P = 0.022 vs. group B from 32.7 ± 8% to 35 ± 8.5%, P = 0.003). Of these parameters, only PALS was significant independent predictor of CV events on logistic regression analysis (OR 0.82 CI 95 0.69–0.98, P = 0.034). Conclusions Advanced echocardiographic parameters, including myocardial work and left atrial strain analysis, are particularly valuable in the early detection of TKI-induced cardiac toxicity. PALS could be an useful tool to predict outcome in these patients.