Limitations
There were limitations in the present study. Patients with severe cardiac diseases and noncardiac death were excluded from our study, which had some degree of selection bias. The 3D RV endocardial border identification may be of challenge as echo dropouts frequently occur. Due to the small size of our study, we did not confirm the finding in a separate independent “validation cohort”. The development and validation of a multivariable prediction model for CAE after R-CHOP regimens treatment among patients with lymphoma is required in our future studies. Separate multivariable Cox analysis may be overfitted owing to the low incidence of adverse cardiac events induced by chemotherapy. Finally, our patients didn’t undergo cardiac magnetic resonance (CMR) so our RV data sets measured by 3D transthoracic echocardiography were not tested for the correlation with CMR.