2.1 basic clinical characteristics of two groups
A total of 59 patients (42 females, 17 males) were included, with 18 cases (30.5%) in the non-PAH and 41 cases (69.5%) in the PAH group. In the PAH group, there was a significant increase in the respiratory rate, as well as higher occurrences of dyspnea and dizziness compared to the non-PAH group (P < 0.05). Furthermore, patients in the non-PAH group exhibited a higher frequency of accompanying congenital heart disease, while a history of acute pulmonary embolism was more prevalent in the PAH group (P < 0.05), see Table 1.
Table 1 Clinical characteristics of the patients in two groups
|
non-PAH (n = 18)
|
PAH (n = 41)
|
T/Z/χ2
|
P
|
Age (year)
|
48.22 ± 17.96
|
47.65 ± 14.91
|
0.128†
|
0.898
|
Sex (male/female)
|
7 / 11
|
10 / 31
|
1.282§
|
0.258
|
Signs
|
|
|
|
|
aortic diastolic blood pressure (mmHg)
|
75.78 ± 11.09
|
77.67 ± 10.97
|
-0.614†
|
0.542
|
aortic systolic blood pressure (mmHg)
|
126.28 ± 21.43
|
116.98 ± 16.42
|
1.840†
|
0.071
|
respiratory rate (times/min)
|
18 (18~20)
|
20.00 (19~20)
|
-2.083‡
|
0.037
|
pulse (times/min)
|
80 (71.5~84)
|
82.5 (72.75~91.75)
|
-1.401‡
|
0.161
|
Symptoms
|
|
|
|
|
dyspnea (%)
|
4 (22.2%)
|
36 (83.6%)
|
24.642§
|
<0.001
|
fever (%)
|
1 (5.6%)
|
0 (0%)
|
-
|
0.295
|
dizzy (%)
|
8 (44.4%)
|
0 (0%)
|
-
|
<0.001
|
edema of both lower limbs (%)
|
1 (5.6%)
|
1 (2.3%)
|
-
|
0.507
|
chest pain (%)
|
1 (5.6%)
|
1 (2.3%)
|
-
|
0.507
|
palpitation (%)
|
0 (0%)
|
2 (4.7%)
|
-
|
1.000
|
syncope (%)
|
1 (5.6%)
|
2 (4.7%)
|
-
|
1.000
|
Baseline diseases
|
|
|
|
|
congenital heart disease (%)
|
12 (66.7%)
|
11 (25.6%)
|
9.119§
|
0.003
|
acute pulmonary embolism (%)
|
2 (11.1%)
|
18 (41.9%)
|
5.444§
|
0.020
|
chronic obstructive lung disease (%)
|
0 (0%)
|
3 (7.0%)
|
-
|
0.548
|
coronary heart disease (%)
|
3 (16.7%)
|
4 (9.3%)
|
-
|
0.411
|
rheumatic disease (%)
|
0 (0%)
|
1 (2.3)
|
-
|
1.000
|
Note: †, t-test; ‡, Z test; §, χ2 test; --, Fisher exact probability test.
Abbreviations: PAH, pulmonary arterial hypertension.
2.2 comparison of RHC and CTPA parameters in two groups
The parameters of the RHC (mPAP, SPAP, and DPAP) and CTPA (DMPA, DRPA, DLPA, AMPA, ARPA, ALPA, VMPA, VRPA, VLPA) were significantly higher in the PAH group than in the non-PAH group (P < 0.05), see Table 2.
Table 2 the pulmonary arterial pressures and measurements of the pulmonary arteries on CTPA images of the patients in two groups
|
non-PAH (n = 18)
|
PAH (n = 41)
|
T/Z
|
P
|
RHC
|
|
|
|
|
SPAP (mmHg)
|
26.00 (24~29)
|
72.22 ± 19.73
|
-5.940‡
|
<0.001
|
mPAP (mmHg)
|
18 (16~19)
|
26 (20~35)
|
-5.711‡
|
<0.001
|
DPAP (mmHg)
|
10 (8.5~11)
|
27.85 ± 10.77
|
-6.079‡
|
<0.001
|
CTPA
|
|
|
|
|
DMPA (mm)
|
25.93 ± 4.68
|
36.01 ± 6.36
|
-5.901†
|
<0.001
|
AMPA (mm2)
|
601.57 ± 184.37
|
1046.88 ± 278.94
|
-6.285†
|
<0.001
|
AMPA (mm3)
|
25751.64 ± 7154.24
|
55557.10 ± 18449.48
|
-8.511†
|
<0.001
|
DRPA (mm)
|
19.37 ± 3.32
|
24.37 ± 4.15
|
-4.498†
|
<0.001
|
ARPA (mm2)
|
348.58 ± 114.38
|
536.47 ± 176.59
|
-4.041†
|
<0.001
|
VRPA (mm3)
|
15594.91(12662.13~29474.24)
|
48454.73 ± 19118.22
|
-4.893‡
|
<0.001
|
DLPA (mm)
|
17.51 ± 3.47
|
21.68 ± 4.36
|
-3.603†
|
0.001
|
ALPA (mm2)
|
314.36 ± 75.18
|
459.79 ± 182.64
|
-4.075 †
|
<0.001
|
VLPA (mm3)
|
23118.99 ± 13552.82
|
37963.42 ± 16972.93
|
-3.073†
|
0.003
|
Note: †, t-test; ‡, Z test;
Abbreviations: PAH, pulmonary arterial hypertension; RHC, right heart catheterization; SPAP, systolic pulmonary artery pressure; mPAP, mean pulmonary arterial pressure; DPAP, diastolic pulmonary arterial pressure; CTPA, computed tomography pulmonary angiography; DMPA, diameter of main pulmonary artery; AMPA, cross-sectional area of main pulmonary artery; VMPA, volume of main pulmonary artery; DRPA, diameter of right pulmonary artery; ARPA, cross-sectional area of right pulmonary artery; VRPA, volume of right pulmonary artery; DLPA, diameter of left pulmonary artery; ALPA, cross-sectional area of left pulmonary artery; VLPA, volume of left pulmonary artery.
2.3 Correlation analysis
The one-dimensional (DMPA, DRPA, DLPA), two-dimensional (AMPA, ARPA, ALPA), and three-dimensional (VMPA, VRPA, VLPA) measurements of CTPA images exhibited a significant positive correlation with the results obtained from RHC (mPAP, DPAP, SPAP), all P < 0.05. From the heatmap of the correlation coefficient between the RHC and the CTPA (Figure 3), we can read that, for the MPA and RPA, the correlation coefficients of the three-dimensional CTPA parameters were higher than the one-dimensional and two-dimensional parameters. However, for the LPA, the correlation coefficients of the three-dimensional parameters were lower than the one-dimensional and two-dimensional parameters.
2.4 The ROC analysis of the diagnostic efficacy of CTPA parameters in predicting PAH
For the MPA, the volume showed better accuracy than the diameter and sectional area (see Table 3 and Figure 4A), but the Delong tests showed no significant difference among the three ROC curves (DMPA vs AMPA, Z= 0.092, P = 0.927; DMPA vs. VMPA, Z = 0.788, P = 0.431; AMPA vs. VMPA, Z = 0.887, P = 0.375).
For the RPA, the volume showed higher accuracy than the diameter and cross-sectional area (see Table 3 and Figure 4B), meantime, the Delong tests showed that volume had significantly a greater area under the curve in predicting PAH than the counterparts of diameter and the sectional area(DRPA vs. VRPA, Z = 2.029, P = 0.042; ARPA vs. VRPA, Z = 2.119, P = 0.034).
For the LPA, the volume showed lower accuracy than the diameter and sectional area (see Table 3 and Figure 4C). And the Delong test showed no significant difference among the three curves (DLPA vs ALPA, Z= 0.286, P = 0.775; DLPA vs VLPA, Z = 0.379, P=0.705; ALPA vs. VLPA, Z = 0.206, P=0.837).
Table 3 ROC curve for the identification of PAH based on CTPA measurements
|
cutoff value
|
AUC
|
95% CI
|
sensitivity
|
specificity
|
PPV
|
NPV
|
Accuracy
|
DMPA (mm)
|
29.300
|
0.906
|
0.832~0.990
|
0.853
|
0.833
|
0.904
|
0.764
|
0.687
|
AMPA (mm2)
|
803.005
|
0.908
|
0.827~0.984
|
0.805
|
0.889
|
0.904
|
0.764
|
0.694
|
VMPA (mm3)
|
38853.430
|
0.934
|
0.887~0.999
|
0.805
|
1.000
|
0.881
|
0.722
|
0.805
|
DRPA (mm)
|
22.900
|
0.822
|
0.713~0.932
|
0.667
|
0.889
|
0.761
|
0.583
|
0.556
|
ARPA (mm2)
|
499.998
|
0.812
|
0.705~0.928
|
0.564
|
1.000
|
0.795
|
0.642
|
0.564
|
VRPA (mm3)
|
39564.981
|
0.906
|
0.839~0.985
|
0.718
|
1.000
|
0.860
|
0.706
|
0.718
|
DLPA (mm)
|
20.750
|
0.802
|
0.681~0.921
|
0.657
|
0.889
|
0.734
|
0.555
|
0.546
|
ALPA (mm2)
|
417.252
|
0.776
|
0.668~0.909
|
0.552
|
0.944
|
0.705
|
0.428
|
0.517
|
VLPA (mm3)
|
27277.279
|
0.753
|
0.629~0.898
|
0.684
|
0.778
|
0.782
|
0.667
|
0.462
|
Abbreviations: ROC, receiver operator characteristic; AUC, areas under the curve; PPV, positive predictive value; NPV, negative predictive value; DMPA, diameter of main pulmonary artery; AMPA, cross-sectional area of main pulmonary artery; VMPA, volume of main pulmonary artery; DRPA, diameter of right pulmonary artery; ARPA, cross-sectional area of right pulmonary artery; VRPA, volume of right pulmonary artery; DLPA, diameter of left pulmonary artery; ALPA, cross-sectional area of left pulmonary artery; VLPA, volume of left pulmonary artery; CI, confidential interval.
2.5 Linear regression model for the prediction of PAH
In the multiple linear regression model, the predictive equation of mPAP, DPAP, and SPAP were as follows:
predicted mPAP = 8.178 + 0.0006 * VMPA
predicted DPAP = 1.418 + 0.0005 * VMPA
predicted SPAP = -11.137 + 0.0006*VRPA + 1.259 * DMPA.
The prediction of the mPAP, DPAP, and SPAP showed good Goodness of fit with its true values, see Figure 5.