Page 80 - Assessing right ventricular function and the pulmonary circulation in pulmonary hypertension Onno Anthonius Spruijt

5
Cardiac magnetic resonance imaging (CMRI)
All CMRI scans were performed on a Siemens 1.5T Sonato or Avanto scanner (Siemens Medical Solutions, Erlangen, Germany). Image acquisition and post-processing was performed as described elsewhere [3]. Short-axis images were analyzed using the MASS software package (MEDIS Medical Imaging Systems, Leiden, the Netherlands). Endocardial borders of the RV were manually traced on end-diastolic images (defined as the first cine after the R-wave trigger) and on end-systolic images (defined as the cine with visually the smallest cavity). Trabeculae as well as papillary muscles were excluded from the cavity and RV end-diastolic and end-systolic volumes were assessed using the Simpson rule. RVEF was calculated as:
Statistical analysis
Normal distribution of the data was checked and presented as mean ± SD, unless stated otherwise. Simple linear regression was performed to test the relation between CMRI-derived-RVEF and echo- derived parameters of RV systolic function. Receiver operating characteristic (ROC) analyses were used to test which echo parameter best predicted a RVEF<35% or ≥35% *3+. The areas under the curves (AUCs) of the different parameters were tested using the DeLong-method.
Intra-observer and inter-observer variability of all four echo parameters were assessed in 15 patients and tested using Bland and Altman analysis and intraclass correlation coefficients. Intra- observer and inter-observer variability was also assessed of CMRI-derived RVEF in 10 patients using Bland Altman analysis. For echo and CMRI intra- and interobserver variability measurements the moment of end-systole was independently chosen [16].
The comparison between the change in echo-derived parameters of RV systolic function and the change in CMRI-derived-RVEF during follow-up was firstly analysed by simple linear regression. Subsequently, sensitivity, specificity, negative predictive values (NPV) and positive predictive values (PPV) of the ability of the four echo-parameters to predict a deterioration in RV function were analysed. A decrease of >3% in CMRI-derived-RVEF was considered as deterioration in RV function