Limited value of pulse wave analysis in assessing arterial wave reflection and stiffness in the pulmonary artery

Figures

• 

  FIGURE 1. Schematic of pulse wave analysis. In pulse wave analysis (a), the inflection point is identified as the first shoulder of the pressure waveform. The augmentation pressure (AP) is defined as the pressure difference between the inflection on the pressure waveform and the peak systolic pressure and augmentation index (AI) is the ratio of the augmentation pressure to pulmonary arterial pulse pressure (PAPp). The inflection time (T i) was determined as the time difference between the onset of the pressure wave upstroke (T foot) and the inflection point. The inflection point was determined using the fourth derivative (b) of the ensemble average pressure waveform (Takazawa et al., 1995). The first zero‐crossing from above to below of the fourth derivative was assigned as the onset of the pressure wave upstroke and the inflection point was taken to correspond to the second zero crossing from above to below

• 

  FIGURE 2. Pressure waveforms of all individuals. The inflection point (hollow triangle) was determined as the second zero crossing of the fourth derivative of the ensemble average pressure waveform. Pressure waveforms were classified into types (a) (augmentation index [AI] >12%), (b) (0 ≤ AI ≤ 12%) and (c) (AI < 0). CTEPH, chronic thromboembolic pulmonary hypertension; PAH, pulmonary arterial hypertension

• 

  FIGURE 3. Examples of arterial wave analyses. Arterial wave analyses of two pulmonary hypertension patients (a–f, respectively) are shown. In pulse wave analysis (a and d), the effect of the reflected waves is conventionally provided by the augmentation index (AI), i.e., the ratio of the augmentation pressure (AP) to pulmonary arterial pulse pressure. Wave intensity analysis (b and e) of the same patients revealed an initial forward compression wave (FCW) related to right ventricular contraction, a late systolic forward decompression wave (FDW) related to ventricular relaxation, and a mid‐systolic backward compression wave (BCW) due to wave reflection. The contour of the net wave intensity is highlighted in black. Finally, wave separation analysis (c and f) separates the measured pressure (P m) into a forward (P f) and backward (reflected) pressure (P b). Here, we show an example of an easily identifiable inflection point (a), where the inflection time (T i) corresponds to a decrease in velocity and coincides with the timing of BCW (b) and the increasing P b (c). In another case, a convincing inflection point was difficult to identify (d) due to the smooth contour of the pressure waveform.…

• 

  FIGURE 4. Group comparison of augmentation index and inflection time. Comparison of the augmentation index (AI) and inflection time (T i) in the main (a–d) and branch (e–h) pulmonary arteries between controls, patients with pulmonary arterial hypertension (PAH) and patients with chronic thromboembolic pulmonary hypertension (CTEPH) before and after pulmonary endarterectomy (PEA). In the left panel, all participants are included and in the right panel, only participants with 0 ≤ AI ≤ 50% are included. Differences between the groups were compared using the Kruskal Wallis test or the Wilcoxon rank‐sum test. Paired data were analyzed using the Wilcoxon signed‐rank test

• 

  FIGURE 5. Correlation between augmentation index and wave speed. The augmentation index (AI) was poorly correlated to the wave speed (c) in the main pulmonary artery, while there was a moderate correlation in the branch pulmonary artery. In the left panel (a and c), all participants are included; in the right panel (b and d), only patients with pulmonary arterial hypertension (PAH) and chronic thromboembolic pulmonary hypertension (CTEPH) that have 0 ≤ AI ≤ 50% are included

• 

  FIGURE 6. Simple correlation analyses of wave reflection magnitudes and timings. The augmentation index (AI) was poorly correlated to wave reflection index (WRI, a) and the ratio of backward to forward pressure (P b/P f, b). The inflection time (T i) was poorly correlated to the traveling time of the backward compression wave (ΔT, c) and the timing of the peak backward pressure (T Pb, d). Data are from the branch pulmonary artery of all participants. The two Pearson correlation coefficients given are for all participants (r 1) and for patients with pulmonary arterial hypertension (PAH) and chronic thromboembolic pulmonary hypertension (CTEPH) that have 0 ≤ AI ≤ 50% (r 2). Similar findings were observed in the main pulmonary artery

• 

  FIGURE 7. Bland–Altman plots of the inflection time (T i) and timing of the peak backward compression wave (T BCW). In the left panel (a, c), all participants are included. In the right panel (b, d), only patients with pulmonary arterial hypertension (PAH) and chronic thromboembolic pulmonary hypertension (CTEPH) that have 0≤ augmentation index ≤50% are included. Note that the errors in the plots are not uniform

• 

  FIGURE 8. Exploring the inflection point of type C pressure waveforms. In individuals with type C pressure waveforms, the detected inflection time (T i) was closely correlated to the timing of the peak forward decompression wave (T FDW, a and c). In pulmonary hypertension (PH) patients, T i matched T FDW (±30 ms), while T i arrived earlier than FDW in controls (b and d)