The univariate correlation of CT scan findings with preoperative PAP and PVR are summarized in Table 1
. The PA and the rPA showed a significant correlation with PAP, whereas the correlation with preoperative PVR was only weak or not significant. In 91.4% of patients (53 of 58 patients), central thromboembolic material was detectable. There were no significant differences of preoperative PAP or PVR between patients with and without the depiction of central thrombi (PAP, 47 ± 12 vs 45 ± 6 mm Hg, respectively; PVR, 891 vs 1,050 ± 512 dyne · s · cm−5, respectively; p > 0.05). The mean score of central thrombi was 37 ± 21% (range, 0 to 85%). The score of central thromboembolic material showed a significant correlation with neither PAP nor PVR. Dilated bronchial arteries could be demonstrated in 47.1% of patients (Fig 1
). The mean preoperative PAP and PVR were not significantly different in patients with and without dilated bronchial arteries (PAP, 44 ± 12 vs 50 ± 11 mm Hg, respectively; PVR, 817 ± 419 vs 986 ± 427 dyne · s · cm−5; p > 0.05). Peripheral densities were seen in 87.7% of patients (Fig 2
). The number of lobes with peripheral densities (scar score) demonstrated a poor correlation with PAP and a moderate correlation with PVR (mean score, 2.28 ± 1.46; range, 0 to 6). All patients showed at least two lobes with abnormal perfusion. The best parameter for correlation with preoperative hemodynamic measurements was the number of lobes with abnormal perfusion (perfusion score), which demonstrated a strong correlation with PAP and PVR (mean score, 5.17 ± 1.24; range, 2 to 6) [Fig 3]
. In multiple linear regression analysis, including all evaluated CT scan findings, the perfusion score correlated with the PAP and mean PVR (PAP, 13.82 + 6.48 × perfusion score; R = 0.60; p < 0.0001; PVR: −397.98 + 257.33 × perfusion score; R = 0.67; p < 0.0001).