We identified 30 patients recently diagnosed with CTEPH and 20 control patients with normal resting pulmonary artery values who had a complete data set of investigations. 10 of those with CTEPH were deemed to have distal CTEPH and 20 with proximal disease. There were no significant differences in age and sex between these groups.
Baseline characteristics
Baseline characteristics are shown in Table 1. 10 of the 20 patients with no pulmonary hypertension at rest had evidence of chronic thromboembolic disease on imaging while the other 10 had no evidence of residual thromboembolic disease. As expected, mean (SD) pulmonary arterial pressure (40.2+/- 11.4 vs 17 +/- 4.4 mmHg; p=<0.01), PVR (8.5 +/- 4.7 vs 1.9 +/- 1.2 Wu; p=<0.01) were higher in patients with CTEPH versus controls. Pulmonary artery wedge pressure was normal in both groups. NTpro-BNP was significantly higher in patients with CTEPH versus non-PH (1004+/- 1291 vs 184 +/- 155 pg/mL; p=<0.01).
CPET parameters are outlined in Table 2 , using the main indices that are used to detect and measure exercise capacity in pulmonary hypertension. There was no significant difference in peak VO2 between groups but VE/VCO2 (49.6 vs 34.3; p=<0.01) was higher in CTEPH, in keeping with increased dead-space ventilation.
Differences in CMR analysis between CTEPH and non-CTEPH groups
Differences in the novel geometric CMR indices in those with pulmonary hypertension and those without are shown in Table 3. Pulmonary artery distensibility was significantly lower in those with CTEPH (0.13+/- 0.1 vs 0.46+/- 0.23; p=<0.01) with a significantly higher EI in systole (1.3+/- 0.5 vs 1.0+/- 0.01; p=<0.01) and diastole (1.22+/- 0.2 vs 0.98+/- 0.01; p=<0.01) compared to the control group. There was no significant difference in peak pulmonary artery blood flow velocity.
Mean pulmonary artery pressure correlated with systolic LVEI (R-value 0.74, p=<0.01) and with diastolic LVEI (R-value 0.75, p=<0.01). Systolic and diastolic EI also correlated with prognostically important markers including PVR (R-value 0.75, and 0.76, respectively), mixed venous saturations (R-value -0.65, -0.68 respectively) and six-minute walk distance (R-values -0.71 and -0.78, respectively) (Table 4 and Supplementary Figure 1).
Those with distal CTEPH in our group were unable to perform cardiopulmonary exercise testing due to physical limitations, and, as such there is no data available for this group. CPET data for the other three groups are outlined in Table 7. Whilst there is a reduction in peak VO2 in the CTEPH group, this did not reach statistical significance. However, there was a progressive rise in the VE/VCO2 (p=<0.01), consistent with greater dead-space ventilation in those with CTEPH.
We have shown good interobserver agreement in our CMR determinants, lending confidence to our results using Pearson’s correlation (R=0.9)
Determining Threshold values for detection of pulmonary hypertension
Using Receiver operator curves the optimal threshold for the detection of pulmonary hypertension with the systolic EI is 1.1 (AUC=0.84), diastolic index is 1.1 (AUC=0.94) and pulmonary artery distensibility is 0.28 (AUC=0.95). We found no meaningful threshold for the peak blood flow (Figure 4).
CMR can differentiate between subgroups of thrombotic pulmonary vascular disease
Patients without pulmonary hypertension were divided into two groups – those without evidence of pulmonary vascular obstruction on CT and V/Q imaging (normal controls), and those with residual radiological proven disease. Differences in the hemodynamic indices across those who had normal pulmonary vasculature, those with CTED and those with proximal and distal CTEPH were compared (Table 6). As expected across the spectrum of thrombotic pulmonary vascular disease, there was a progressive rise in the mean pulmonary artery pressure (p=<0.01), a progressive reduction in the cardiac index (p=<0.01) and rise in the PVR (p=<0.01).
The differences between groups in the geometric novel CMR indices are shown in Figure 5. There was a significant difference between those with both distal and proximal CTEPH and CTED in the systolic LEVI (p=<0.01), diastolic LVEI (p=<0.01) and pulmonary artery distensibility (p=<0.01). No significant differences between subgroups in peak blood flow were seen.
Subgroup analysis in patients with thrombotic pulmonary vascular disease with no pulmonary hypertension
The differences in baseline variables and novel CMR indices between those without evidence of pulmonary vascular obstruction on CT and V/Q imaging (normal controls), and those with residual radiological proven disease were evaluated (Table 5). Age and sex were similar between groups. Although the mean pulmonary artery pressures and PVR were within the normal range for both groups, mPAP (19.9 (3.1) vs 14.1 (3.4) mmHg;p=<0.01) and PVR (2.5 (1.1)vs 1.2 (0.9) WU;p=0.01) were higher in the CTED group, in keeping with pulmonary vascular occlusions. In addition, there was a significantly lower pulmonary artery distensibility (0.34 (0.1) vs0.5 (0.4); p=0.03) and a significantly higher LVEI in systole (1.04 (0.03) vs 0.99 (0.08); p=<0.01) and diastole (1 (0.1) vs (0.96 (0.07); p=0.02) in the CTED group compared to normal but no significant difference in peak blood flow between these groups. Threshold analysis for identifying the presence of CTED using systolic and diastolic LVEI is 1.0 and pulmonary artery distensibility is 0.37 (Figure 2 supplemental).