Left atrial volumes evaluation by cardiac magnetic resonance - time to revise the grading severity cut off values
Abstract Funding Acknowledgements Type of funding sources: Public grant(s) – National budget only. Main funding source(s): This work was supported by a grant of Ministery of Research and Innovation, CNCS-UEFISCDI, project number grant 5/2018 and 83/2018 and PN-III-P1-1-TE-2016-0669, within PNCDI III Background Left atrial maximal volume (LAVmax) assessed by 2D echocardiography (2DE) is incorporated in new scoring system (2019) of heart failure with preserved ejection fraction (HFpEF), using well established cut-off values for grading severity. These values have strong evidence for diagnosis and prognosis. Moreover, cardiac magnetic resonance (CMR) is also included in the structural evaluation of HFpEF. However, cut-off values for LAVmax evaluated by CMR are higher. This might generate a different classification for LA dilation by CMR, potentially including patients in another risk class. We aimed to evaluate LAVi by 2DE and CMR in a group of HFpEF patients, in order to test the agreement between these two methods, regarding severity grading. Methods. We prospectively enrolled 74 HFpEF patients (68 ± 9 yrs), and evaluate them by 2DE and CMR. Conventional 2DE was used to define anatomy and function of the left ventricle (LV). NTproBNP was done in all patients. We assessed biplane LAVmax index (LAVimax) and LA minimal volume index (LAVimin), from 4C and 2C views, and reported as a mean value, by eco and CMR (area–length method). Cohen"s k of agreement was evaluated to determine if there was an agreement between eco and CMR, using the cut off values for LAVimax by CMR and eco, provided by the guidelines (Figure) Results. Feasibility of the complete analysis was 95% (70 pts). LVEF was 60 ± 5.7%. LAVimax and LAVimin by eco and CMR were highly correlated (all R > 0.7, p < 0.001) (Table). However, CMR diagnosed a high percent of patients as having normal LAVimax values, by comparison with echo (59 vs. 5.7 %, p < 0.001) (Figure). There was no agreement between these two evaluation methods (Cohen"s k of agreement = 0.000), suggesting that CMR evaluation provides completely different severity grading. NTproBNP significantly correlated only with LAVimin by CMR (R = 0.4, p = 0.04). Conclusions. We suggest that there is need for a future update of the grading cut-offs for LAV by CMR with larger reference range studies, and also with prognosis studies, as 2DE already provided. The present CMR grading severity might generate misclassification in certain pathological condition, such as HFpEF. We also suggest that LAVi min, instead of LAVi max should be used in the future study for assessment of prognosis. LA volumes evaluation by CMR and 2D echo LAVimax LAVimin ECO (ml/m2) 47 ± 11 23 ± 9 CMR (ml/m2) 51 ± 15 26 ± 12 R coefficient 0.7 0.74 P value <0.001 <0.001 Abstract Figure.