Portal de Periódicos da CAPES

Balancing the positives and negatives of the diastolic pulmonary gradient

2017; Elsevier BV; Volume: 19; Issue: 1 Linguagem: Inglês

10.1002/ejhf.704

1879-0844

Emmanouil Tampakakis, Ryan J. Tedford,

Congenital Heart Disease Studies

This article refers to 'Determinants and prognostic implications of the negative diastolic pulmonary pressure gradient in patients with pulmonary hypertension due to left heart disease' by A.I. Nagy et al., published in this issue on pages 88–97. Combined post- and pre-capillary pulmonary hypertension (CpcPH) caused by left heart disease (PH-LHD) has an undoubtedly worse prognosis compared with isolated post-capillary (IpcPH),1 yet how to best distinguish these two groups is still a matter of great debate. The diastolic pulmonary gradient (DPG), defined as diastolic pulmonary artery pressure (dPAP) minus pulmonary artery wedge pressure (PAWP), should be less affected by left heart failure-induced changes in vascular compliance,2 and, as a result, during the Fifth World Symposium on Pulmonary Hypertension it was proposed that a DPG ≥7 mmHg alone should define CpcPH.3 Since then, studies exploring the prognostic value of DPG have yielded mixed results4-9 and more recent guidelines have reincorporated pulmonary vascular resistance (PVR) into the CpcPH definition.10 Perhaps surprisingly these studies, as well as earlier ones,11 report a relatively high rate of negative DPG values. However, it remains unclear if a negative DPG is physiologically plausible or whether these values are simply a result of measurement error or catheter artefact.4 In this issue of the Journal, Nagy and colleagues12 attempt to answer this very intriguing question as well as define the prognostic significance of a negative DPG. Of the 316 subjects who were evaluated haemodynamically with right heart catheterization and echocardiography, a high proportion (n = 256, 84.5%) met criteria for PH-LHD and were analysed further. Of these, 162 were initially referred to the Karolinska University Hospital in Stockholm, Sweden for evaluation of heart failure and 94 were referred for percutaneous transvenous mitral commissurotomy for mitral valve stenosis at the Sri Sathya Sai Institute in Bangalore, India. Subjects with mitral valve stenosis had significantly worse haemodynamics including higher pulmonary pressures, higher PVR and lower cardiac index compared with the cohort of PH-LHD. When these cohorts were combined, a negative DPG was calculated in almost half of all subjects (48%). A negative DPG was associated with a more favourable haemodynamic profile including lower pulmonary pressures, transpulmonary gradient (TPG) and PVR as well as better right heart function with smaller right atrial and ventricular size, lower right atrial pressures and higher tricuspid annular plane systolic excursion (TAPSE). A better haemodynamic profile was also seen in the negative DPG group when only analysing those subjects with PVR between 3 and 7 Wood units (WU). Most interestingly, the authors found an inverse relationship between V-wave amplitude and DPG in those with a PVR 10 mmHg), who were primarily mitral valve stenosis patients (62%), were found to have lower and predominately negative DPG as well as higher mean PAWP. Higher V-wave amplitude was also associated with a greater difference between the mean pulmonary artery pressure (mPAP) and dPAP (and therefore TPG and DPG, or ΔPG). Lastly, a survival analysis was performed in the subset of PH-LHD subjects (n = 127). Over a median follow-up period of 15.6 months, negative DPG was associated with lower occurrence of death or heart transplantation than DPG 0–6 mmHg. Notably, owing to the low number of subjects with DPG ≥7 mmHg (n = 17), no comparisons with this group were performed. Neither elevated TPG nor PVR or combinations of DPG and TPG had prognostic value in this study. The authors should be commended for their effort to address this controversial and clinically relevant issue, and their findings offer significant insight into the determinants of DPG, namely the impact of large V-waves. One of the major strengths of the analysis—which is rare in haemodynamic studies—is that even though catheterizations were performed at different institutions, all tracings were reviewed by a single investigator, which reduces variability in interpretation. In addition, although the mitral valve stenosis and PH-LHD cohorts had significant differences in clinical characteristics and haemodynamics, the consistent impact of V-waves on the DPG improves the generalizability of this finding. The mechanism by which the V-wave leads to a lower, and sometimes negative DPG, is less clear. Mathematically, V-waves could inversely influence DPG values by one of two mechanisms: either by lowering measured dPAP or increasing measured PAWP values. The authors propose that V-waves, which occur just after the onset of left ventricular systole, contribute to an asymmetrical backward transmission of the phasic left atrial pressure. While this may be true, and may contribute to higher pulmonary arterial pressures toward the end of systole or perhaps early diastole given the delay of pressure transmission, it is difficult to comprehend why it would lead to diastolic pulmonary pressures that fall below the left atrial pressure at end-diastole. With this in mind, and as the authors importantly discuss, the technique used for PAWP measurements likely accounts for a significant portion of these findings. Although recent guidelines13 have attempted to standardize PAWP measurements by recommending end-expiration measures in the absence of breath hold, no guidance is provided on how specifically to measure the PAWP. In practice, two methods are generally employed. One, and as the authors have performed for this study, is taking the mean PAWP over the cardiac cycle. The second is measuring PAWP at end-diastole, which coincides best with the pre-C wave (mitral valve closure) on the PAWP tracing. Because the pre-C wave is rarely appreciated when using fluid-filled catheters, averaging the mean of the 'a'-wave is performed (in the presence of atrial fibrillation, the z-point or pre-V-wave point may be used). The latter more closely approximates left ventricular end-diastolic pressure (LVEDP) while the former may be a better estimate of the pressures 'felt' by the pulmonary circulation. Typically, these values are quite similar, but in the presence of large V-waves, the mean PAWP can be much higher than the mean 'a'-wave PAWP. This is illustrated nicely in Figure S2 of the manuscript where using the mean 'a'-wave instead of the mean PAWP results in a DPG of 2 mmHg rather than –2 mmHg. In the study by Haskell and French14, mean PAWP overestimated LVEDP in the presence of large V-waves secondary to mitral regurgitation and the pre-V wave (or trough of the x-descent) more closely approximated LVEDP. The same is true of mean PAWP in stiff left atrial syndrome, where DPG is calculated as negative using mean PAWP but positive if LVEDP is used instead.15 Indeed, in the current study, when using the alternative measures of PAWP in a subset of 34 subjects (the z-point in this case; Table S2), DPG values were more positive. The impact of large V-waves on the PAWP, however, would not explain the observed difference between mPAP and dPAP, which increased in the presence of large V-waves. This finding may suggest that large V-waves lower pulmonary vascular compliance,2 thereby increasing pulmonary pulse pressure, and mPAP relative to dPAP. A more direct impact on systolic pulmonary artery pressure from backward transmission of pressure is also a possibility. The findings of better transplant-free survival in the negative DPG cohort is in contrast to reports from Gerges et al.8 who found elevated risk at lower DPG. In our earlier study,4 negative DPG was associated with a trend toward better survival compared with those with normal DPG (P = 0.053; Figure 1). The discrepant findings likely reflect a combination of differences in patient populations, sample size, follow-up time, and perhaps measurement techniques. Importantly, the impact of this study goes beyond our understanding of a negative DPG and a call for further standardization of PAWP measures. It also implies that large V-waves, because they may lower the interpreted DPG value, could lead to a misclassification as IpcPH if only the DPG is considered. At a threshold of PVR >7 WU, the impact of V-waves on the DPG was no longer significant in this study. However, only ∼15% of the cohort fit into this subset and no prevalence of large V-waves was reported in this group. The use of direct left atrial pressure measurements rather than PAWP resulted in the reclassification of 7/51 (14%) to either positive or negative DPG. Because both measures should be equally impacted by V-waves, this again may speak to a measurement fidelity issue with the DPG. Taken together, and particularly in the setting of a significantly elevated PVR, defining CpcPH by the DPG alone may be misleading. In conclusion, this provocative study by Nagy and colleagues12 provides substantial insight into our understanding of the DPG and moves the field forward. However, for now, the search for a clear definition of CpcPH continues. Conflict of interest: none declared.