PS1188 AN UPDATED META‐ANALYSIS OF THE RISKS OF ARTERIAL AND VENOUS OCCLUSIVE EVENTS WITH NEW GENERATION BCR‐ABL TKIS IN PATIENTS WITH CHRONIC MYELOID LEUKAEMIA

Background: A prior meta‐analysis demonstrated that 3 of the new generation BCR‐ABL tyrosine kinase inhibitors (TKIs) (i.e. dasatinib, nilotinib and ponatinib) are associated with an increased risk of vascular occlusive events compared with imatinib. Additional analyses demonstrate that this risk is mainly driven by the occurrence of arterial events. At the time of these analyses, few clinical data were available on bosutinib (only one study included with few events in each arms), inducing a low statistical power to compute the risk of vascular occlusive events with this treatment. Recent clinical data comparing bosutinib to imatinib have been published and need to be incorporated in the meta‐analysis. Aims: To update the meta‐analysis on the risk of arterial and venous occlusive events in patients with Ph+ CML treated with new generation BCR‐ABL TKIs in randomized clinical trials (RCTs), particularly to assess the cardiovascular safety of bosutinib. Methods: The literature search (data lock point: January, 25 th , 2019) was conducted according to a registered protocol (PROSPERO 2014:CRD42014014147). All RCTs comparing a new generation BCR‐ABL TKI versus imatinib in patients with Ph+ CML were included. Two independent investigators screened the literature, reviewed and extracted the data using standard forms. The statistical analyses were performed using a random (REM) and a fixed (FEM) effect model according to the characteristics of the included studies. ORs with 95% CIs were computed using the Peto method. Statistical heterogeneity was quantified using the I 2 value, and publication bias was assessed by funnel plots. Results: Among the 385 abstracts identified, 13 studies fulfilled the established criteria and were included in the statistical analysis. Overall, 7.93% (197/2,484) of patients treated with a new generation BCR‐ABL TKIs developed a vascular occlusive event compared with 2.61% (56/2,149) when treated with imatinib (REM OR PETO : 2.81; 95%CI: 2.09 to 3.80). The update does not change the increased risk of vascular occlusive events with ponatinib (REM OR PETO : 3.47; 95%CI: 1.23 to 9.78), nilotinib (REM OR PETO : 3.55; 95%CI: 2.29 to 5.51) and dasatinib (REM OR PETO : 2.87; 95%CI: 1.48 to 5.55) compared with imatinib. The addition of the results from the BFORE trial (NCT02130557) improve the statistical power of the analysis and demonstrate a non‐significant trend towards an increased risk of vascular events with this treatment (REM OR PETO : 1.67; 95%CI: 0.91 to 3.05). In accordance with the results on ponatinib, nilotinib and dasatinib, this risk is mainly driven by the occurrence of arterial occlusive events. Arterial occlusive events occur in 7.08% (176/2,487) of patients treated with bosutinib whereas only 0.85% (21/2,484) of patients treated with bosutinib report a venous occlusive event. Funnel plots demonstrate no evidences of publication bias, and the I 2 statistic specifies no heterogeneity among studies. Limitations of this meta‐analysis include the absence of a time‐to‐event analysis and the inconsistent report of cardiovascular events in the literature. However, the use of a clinical trial register aimed to decrease this heterogeneity. Summary/Conclusion: This updated meta‐analysis confirmed the risk of arterial occlusive events with new generation BCR‐ABL TKIs. The addition of the results from the BFORE trial increases the statistical power of the bosutinib sub‐analysis and reduce the interval confidence of the analysis. Further investigations are needed to confirm the apparently safer profile of bosutinib in regard to cardiovascular occlusive events.