Page 86 - Assessing right ventricular function and the pulmonary circulation in pulmonary hypertension Onno Anthonius Spruijt

5
Parameters
Sensitivity (%)
Specificity (%)
NPV (%)
PPV (%)
ΔRVFAC (cut-off: <-5%)
ΔTAPSE (cut-off: <-2mm)
ΔFLWN (cut-off: <-5%)
ΔFTWN (cut-off: <-8%)
Table 4: Sensitivity, specificity, NPV and PPV for four echo-parameters for predicting a deterioration in RV function (defined as a decrease in RVEF >3% measured with CMRI (Ref: Bradlow et al.). Cut-off points of the four echo-parameters are based on the SD of the interobserver variability (table 3). RVFAC = right ventricular fractional area change; TAPSE = tricuspid annulus plane systolic excursion; FLWN: fractional longitudinal wall motion; FTWN = fractional transversal wall motion; NPV = negative predictive value; PPV = positive predictive value.
Discussion
We investigated in the largest cohort of patients with precapillary PH thus far, the relation between several simple echo-derived parameters of RV systolic function and CMRI-derived-RVEF (gold standard). Additionally, we studied whether echo-derived parameters of RV systolic function were able to detect a deterioration in RV systolic function during follow-up. We found that RVFAC best correlated with CMRI-derived-RVEF, however none of the four investigated echo-derived parameters were able to sensitively detect changes in CMRI-derived RVEF during follow-up.
Relation between echo parameters and CMRI-derived-RVEF
In two relatively small cohorts of PH patients, the relation was assessed between echo-derived parameters of RV systolic function and CMRI-derived-RVEF. These studies showed conflicting results. Sato et al. [11] compared CMRI-derived-RVEF with echo-derived RVFAC and TAPSE and found that RVEF was better related to TAPSE (r2=0.71) than to RVFAC (r2=0.23). Recently, Shiran and coworkers [12], found a stronger relation between RVEF and RVFAC (r2=0.76) compared to TAPSE (r2=0.64). These inconsistent findings prompted us to investigate the comparison between echo-derived parameters of RV systolic function and CMRI-derived RVEF in a large cohort of PH patients. Our results are in line with those of Shiran et al. [12] with a stronger relation between CMRI-derived- RVEF and echo-derived RVFAC (r2=0.57) compared to TAPSE (r2=0.24). Furthermore, RVFAC could best distinguish between a RVEF<35% and RVEF>35%. Kind et al. [15] compared 2D-CMRI-derived RV parameters with CMRI-derived-RVEF in a large cohort of PH patients and also found the best correlation between RVFAC and RVEF. Given the complex geometry of the RV cavity, it is not
56 33 33 44
79 83 83 86
85 80 80 83
55 63 63 50