Author + information
- Received December 15, 2000
- Revision received May 31, 2001
- Accepted June 19, 2001
- Published online October 1, 2001.
- Ralph Wiedemann, MDa,
- H.Ardeschir Ghofrani, MDa,
- Norbert Weissmann, PhDa,
- Ralph Schermuly, PhDa,
- Karin Quanz, MDa,
- Friedrich Grimminger, MD, PhDa,
- Werner Seeger, MDa and
- Horst Olschewski, MD, FESCa,* ()
- ↵*Reprint requests and correspondence:
Dr. Horst Olschewski, Department of Internal Medicine II, Klinikstr. 36, 35392 Giessen, Germany
The goal of this study was to assess atrial natriuretic peptide (ANP) levels during inhalation of iloprost in severe primary (PPH) and nonprimary pulmonary hypertension (NPPH).
The ANP system is activated in pulmonary hypertension and may help protect from right ventricular (RV) decompensation. It is unknown if ANP regulation is the same in severe PPH and NPPH and if the dynamic regulation is intact in a highly activated ANP system.
In 11 patients with PPH and seven patients with NPPH, right heart catheter investigations were performed. Pulmonary and systemic artery ANP and cyclic guanosine monophosphate (cGMP) levels as well as hemodynamics were measured before and after iloprost inhalation.
The baseline hemodynamics of patients with PPH and patients with NPPH were comparable (mean pulmonary artery pressure [mPAP]: 61 ± 5 mm Hg vs. 52 ± 5 mm Hg, pulmonary vascular resistance [PVR]: 1,504 ± 153 dyne·s·cm−5vs. 1,219 ± 270 dyne·s·cm−5. Atrial natriuretic peptide and cGMP levels were increased about tenfold and fivefold compared with controls in both PPH and NPPH. Iloprost inhalation significantly decreased mPAP (−9.1 ± 2.5 mm Hg vs. −7.9 ± 1.5 mm Hg), PVR (−453 ± 103 dyne·s·cm−5vs. −381 ± 114 dyne·s·cm−5), ANP (−99 ± 63 pg/ml vs. −108 ± 47 pg/ml) and cGMP (−4.6 ± 0.9 nM vs. −4.2 ± 1.6 nM). Baseline ANP including all patients significantly correlated with PVR, right atrial pressure, cardiac index, RV ejection fraction, mixed venous oxygen saturation and cGMP.
The ANP system is highly activated in patients with severe PPH and NPPH. Atrial natriuretic peptide levels are significantly correlated with parameters of RV function and pre- and afterload. Iloprost inhalation causes a rapid decrease in ANP and cGMP in parallel with pulmonary vasodilation and hemodynamic improvement.
As therapies of left heart dysfunction have been altered based on investigations into neurohormonal profiles, there may be potential therapies of right heart failure that may be discovered by investigation of the neurohumoral system. Atrial natriuretic peptide (ANP) is a potent pulmonary vasodilator (1)and possesses considerable antiproliferative (2–4)and diuretic properties. These features are the opposite of the pathophysiologic abnormalities encountered in primary pulmonary hypertension (PPH) and nonprimary pulmonary hypertension (NPPH), such as pulmonary vasoconstriction, pulmonary artery remodeling with proliferation of various cell types and edema formation. The effect of ANP is directly antagonistic to the renin-angiotensin-aldosterone system (5)and to endothelin, the latter of which has been shown to be increased in pulmonary hypertension (6). In addition, ANP has successfully been employed to ameliorate vasoconstriction (7)and remodeling (3)in hypoxic pulmonary hypertension in animal models. All this is consistent with the concept that ANP may serve as a negative feedback mechanism in pulmonary hypertension (7), slowing down the progression of disease. In pulmonary hypertension due to congestive heart failure, it was demonstrated that the ANP level is a marker of left ventricular failure. In contrast, in patients suffering from lung diseases like chronic obstructive pulmonary disease (COPD), interstitial fibrosis and pulmonary hypertension, ANP levels were found to be correlated with parameters of right, rather than left, ventricular function (8,9).
Primary pulmonary hypertension is a disease of unknown origin and poor prognosis (10,11). Previous studies have shown increased ANP levels in patients with PPH (12,13), but it is unknown whether the ANP system is activated to the same extent as in NPPH. Moreover, there is a lack of information whether the chronically activated ANP system still dynamically responds to acute changes in pulmonary hemodynamics. To induce such changes, we employed iloprost, a stable analogue of prostacyclin that was administered after aerosolization via the inhalative route to minimize the systemic side effects of the drug (14). Previous studies with this approach have demonstrated preferential pulmonary, compared with systemic, vasodilation, with the vasorelaxant response leveling off within approximately 1 h. The aim of this study was to characterize baseline ANP levels and their responses to short-term iloprost inhalation in patients with severe primary, compared with non-primary, precapillary pulmonary hypertension.
As depicted in Table 1, a total of 18 patients was enrolled in the study. In 11 patients, PPH was diagnosed based on the criteria of the National Institute of Health registry on PPH (10), and seven patients were classified as suffering from pulmonary hypertension with associated or underlying chronic diseases (Table 2). The study was approved by the local ethics committee, and each patient gave written informed consent for catheterization and short-term vasodilator testing. Four of the patients with PPH were tested three times, five patients twice and two patients once. At least 12 weeks separated repeat tests. One of the patients with NPPH was tested three times and one twice, with the other patients with NPPH being tested once. All patients were admitted to our hospital for testing of pulmonary vasoreactivity and therapy options after having received a diagnosis of severe pulmonary hypertension. All patients were investigated by electrocardiogram; pulmonary function test; echocardiography; perfusion and ventilation scintillation scan and spiral computed tomography (CT) and/or pulmonary angiography; high resolution CT of the lung; and laboratory screening for liver disease, chronic viral infection, renal disease, parasitic disease and collagen vascular disease. Patients were assigned to the PPH group if all of the above investigations gave normal results and/or no underlying or associated disease could be identified. Patients with NPPH were eligible for the study if their mean pulmonary arterial pressure (mPAP) was >40 mm Hg. Exclusion criteria were congestive heart disease as suggested by echocardiography and/or a pulmonary artery wedge pressure (PAWP) >12 mm Hg. During the hospital stay, the patients remained on their maintenance therapy, which included low-dose calcium antagonists and diuretics. Twenty-four hours before vasodilator testing, therapy with calcium antagonists was withheld. None of the patients was treated with beta-adrenergic blocking agent medication for at least six weeks before inclusion.
Hemodynamic studies and inhalation procedure
A total of 34 diagnostic right heart catheterizations were performed using a fiberoptic thermodilution pulmonary artery catheter (Baxter Edwards 93A, 754H, 7.5F) for measurement of right atrial pressure (RAP), mPAP and PAWP, cardiac output (CO), right ventricular ejection fraction (RVEF) (15)and mixed-venous oxygen saturation (SVO2) and for drawing blood samples. A femoral artery catheter was employed for measuring mean systemic arterial pressure (mSAP) and arterial oxygen saturation and for drawing blood samples. Iloprost (Ilomedin, Schering AG, Berlin, Germany, 100 μg/ml) was diluted 1:10 in NaCl 0.9% and jet-nebulized with room air at a pressure of 80 kPa (fluid flux 0.9 ml/min, mass median aerodynamic diameter of particles 2.8 μM, geometric standard deviation 2.6 as determined by impactor technique) and delivered to a spacer connected to the afferent limb of a y-valve mouthpiece for 12 min (14,16–19), resulting in an inhaled dose of about 3 μg (20). Before (baseline) and immediately after aerosolization of iloprost, hemodynamic measurements were performed, and blood samples for blood gas analysis were drawn.
Blood samples for plasma ANP and cyclic guanosine monophosphate (cGMP) were taken from the pulmonary and systemic artery in parallel with performance of hemodynamic measurements. Blood was drawn into chilled tubes containing ethylenediaminetetraacetic acid and 1,000 KIU/ml aprotinin (Trasylol, Bayer AG, Leverkusen, Germany) and placed immediately on ice. Within 2 h, the tubes were centrifuged at 3,000 rpm for 15 min at 4°C, and plasma was transferred to polypropylene tubes and stored at −80°C. Measurements of plasma ANP and cGMP concentrations were done in duplicate by radioimmunoassay (RIA) (Nichols, Bad Nauheim, Germany; Coulter-Immunotech, Hamburg, Germany). All ANP measurements in one catheter investigation were done with one assay kit. The mean interassay variation of the ANP RIA was about 34% and of the cGMP RIA about 5%, as determined by repetitive measurements of control probes. These differences were compensated by correction factors derived from a standard probe.
If there were repeated tests in individuals, means of all these tests were calculated for subsequent analysis. Data are presented as means ± SEM of all patients in the two groups. The Wilcoxon matched-pair signed-rank test (StatXact software, Cytel Software) was used for between-group comparison, and the Pearson correlation was used for estimation of linear regression. Regression analysis for baseline data (Fig. 1)was calculated based on semilogarithmic data.
As depicted in Table 2, five of the patients with NPPH suffered from pulmonary fibrosis, one from chronic pulmonary embolism and one from the CREST (calcinosis, raynaud, esophageal hypomobility, sclerodactyly and teleangiectasia) syndrome without pulmonary fibrosis (as assessed by HR-CT) as underlying diseases. The patients with PPH and NPPH did not significantly differ with respect to their physical and hemodynamic characteristics (Table 1). Patients with PPH showed somewhat higher pulmonary artery pressure and pulmonary vascular resistance (PVR) as well as slightly more severe impairment of right ventricular (RV) function than patients with NPPH, but the differences in these parameters were not statistically significant. Pulmonary function values were in the normal range in all of the patients with PPH, except for the lowered CO diffusion capacity (Table 1). In NPPH, signs of moderate obstruction and restriction were found. The statistical comparison of the pulmonary function data between the groups, however, did not reveal significant differences, except for the CO diffusion capacity (p < 0.001), which was markedly more reduced in the patients with NPPH than in the patients with PPH.
Atrial natriuretic peptide levels were elevated by more than one order of magnitude in patients with PPH and NPPH (Table 3), compared with a range of 22 pg/ml to 64 pg/ml measured in 18 healthy volunteers in a supine position. Moreover, consistent elevation of cGMP levels was noted, being increased approximately fivefold above control data (Table 3, upper normal level 5 nM for the used test). The pulmonary artery to systemic artery difference of ANP levels was significantly different from zero in patients with PPH but not in patients with NPPH. No significant transpulmonary differences were noted for cGMP. None of the ANP and cGMP concentrations displayed any statistically significant difference when comparing patients with PPH with patients with NPPH.
The correlations between baseline hemodynamics and pulmonary artery ANP levels are shown in Figure 1. The analysis included both PPH and NPPH patients because there were no significant differences between the two groups in baseline parameters and hemodynamic responses. Atrial natriuretic peptide was significantly correlated with RAP (r = 0.76) and SvO2(r = −0.67) as well as PVR (r = 0.58) and cGMP (r = 0.45). There was no significant correlation between ANP levels and mPAP (r = 0.34) and PAWP (r = 0.12). The correlation of ANP levels with RVEF and cardiac index (CI) showed a threshold below which the ANP levels increased drastically. For CI, this threshold was approximately 2 l/min/m2, and for RVEF, it was approximately 20%. Linear regression analysis resulted in r = −0.71 and r = −0.79 between CI and RVEF and ANP, respectively. Cyclic GMP levels showed no significant correlation with any of the hemodynamic parameters, except with SvO2(r = −0.4) (data not shown).
Iloprost inhalation had a significant effect on hemodynamics, ANP levels and cGMP levels in both patient groups, as shown in Table 4. A marked pulmonary vasodilation was noted, associated with a minor systemic artery pressure drop. Right atrial pressure decreased, while CI and RVEF increased significantly (Table 4), suggesting some acute recompensation of the RV function. The hemodynamic responses did not significantly differ between the PPH and the NPPH group.
Parallel with the hemodynamic changes induced by iloprost inhalation, the pulmonary artery ANP concentrations decreased significantly in both PPH and NPPH patients, associated with a smaller and more variable decrease in the systemic arteries. Cyclic GMP levels decreased in both groups in the pulmonary and systemic artery. The transpulmonary gradient of cGMP decreased significantly during iloprost inhalation in the patients with PPH (p < 0.05), caused by a stronger systemic (−35%) than pulmonary artery (−25%) cGMP decrease during iloprost inhalation. In the patients with NPPH, systemic and pulmonary cGMP changes were similar (−34% vs. −18%) but did not lead to statistically significant changes in the transpulmonary gradient (p = 0.19).
The correlations between the acute changes in hemodynamics and the corresponding changes in ANP levels are shown in Figure 2. The closest correlation was found with respect to changes in RAP (r = 0.73) and CI (r = −0.73). There was also a significant correlation of ANP changes with the changes in PVR (r = 0.61), SvO2(r = −0.61), and mPAP (r = 0.51), although in PVR and mPAP this could also be due to leverage points. There was no significant correlation with the changes in mSAP (r = 0.21) and cGMP (r = 0.23). The correlation between changes in hemodynamics and changes in cGMP was not significant (data not shown).
It is generally assumed that the ANP secretion depends mainly on the transmural stretch of the left and right atrium (21). In heart failure, however, the ventricles are also known to be a major source of ANP production (22). In experimental pulmonary hypertension, stimuli such as hypoxia, sympathetic nerve activity and RV hypertrophy were noted to be further strong triggers for ANP secretion (23). Concerning the signal transduction pathways involved in the release reaction of stretched cardiac myocytes, endothelin, angiotensin II and nitric oxide are believed to be important mediators (24). Thus, several mechanisms involved in the pathophysiology of pulmonary hypertension may contribute to severely increased circulating ANP levels.
Increased baseline ANP levels
In line with this background in this study, patients with PPH and NPPH revealed considerably elevated ANP levels in both mixed venous and arterial blood samples, with maximum values surpassing 2,000 pg/ml. None of the patients was in the normal range for ANP (<65 pg/ml), and the mean values of PPH and NPPH patients were increased about tenfold compared with the upper normal value. These results agree with data of Morice et al. (13). No significant difference was noted between PPH and NPPH patients, and analysis of the single data showed broad overlapping of the two groups. This suggests that the ANP system is activated in response to the hemodynamic impairment in severe pulmonary hypertension irrespective of the underlying disease.
Baseline hemodynamic parameters were significantly correlated with the pulmonary artery ANP levels (Fig. 1), corresponding to previous studies (12,25). One patient with PPH, who showed near normal ANP levels despite considerably elevated PVR and mPAP, presented with a RAP of zero due to high-dose diuretic therapy. This might explain the large deviation from the other patients concerning the relationship between ANP and hemodynamics. Compared with other hemodynamic parameters, mPAP was not significantly correlated with ANP. This is in contrast to data on patients with COPD (8)and mild pulmonary hypertension (mean PVR approximately 300 dyne·s·cm−5) and a slightly reduced RV function, whereas our population consisted of patients with severe pulmonary hypertension, characterized by a mean PVR of approximately 1,350 dyne·s·cm−5and a severely compromised RV function. This condition results in a plateauing or even a decrease in pulmonary artery pressure upon increasing severity of disease and may explain the poor correlation between ANP and PAP values in our study.
Apparently, there was a “threshold” for CI and RVEF, below which ANP levels increased dramatically (Fig. 1). This threshold was in the range of a CI of 2 l/min/m2and an RVEF of 20%. This is consistent with the concept that additional control mechanisms for ANP are activated when overt RV decompensation occurs.
Transpulmonary ANP gradient
In the PPH group, the pulmonary arterial ANP level was significantly higher than the systemic arterial level (approximately 20% difference), while such transpulmonary difference was less prominent in NPPH (approximately 10% difference). In accord with previous studies (26), we suggest that a changed pulmonary ANP clearance rate may contribute to the highly elevated circulating ANP levels in severe pulmonary hypertension.
The circulating cGMP values in the blood of the PPH and NPPH patients were considerably increased compared with controls. This is in line with data of Bogdan et al. (27). No significant difference was noted between patients with PPH and patients with NPPH. This finding and the presence of very high ANP levels in both groups supports the notion that this peptide may represent the predominant activator of guanylate cyclase(s) in patients with pulmonary hypertension. In agreement, Muramatsu et al. (28)demonstrated that augmented circulating cGMP levels in experimental pulmonary hypertension are not predominantly caused by NO but are largely attributable to ANP. The current cGMP data show significant correlation (r = 0.45) between the baseline values of cGMP and ANP, although the scattering of data also indicates that additional factors may contribute to the elevated cGMP levels. This could be due to increased nonpulmonary nitric oxide synthases or to pulmonary sources of endothelial nitric oxide synthase (eNOS) or other NOS isoforms, although the pulmonary eNOS was reported to be reduced in patients with PPH (29). Other circulating activators of the guanylate cyclase, such as brain natriuretic peptide, may add to the appearance of cGMP, and in addition, this is influenced by phosphodiesterase activity. Finally, it remains unclear if the hemodynamic alterations in the patients with PPH and NPPH affect the cellular escape of this intracellular mediator. Thus, any interpretation of the plasma cGMP data is much more uncertain than that of the plasma ANP levels.
Dynamic control of ANP and cGMP levels
Iloprost inhalation caused an immediate preferential pulmonary vasodilation in both PPH and NPPH patients, with all hemodynamic effects leveling off within approximately 1 h (14,16–19). During aerosolization, PVR decreased by approximately 30%, and PAP by approximately 20% (Table 4). Concomitantly, ANP levels in the pulmonary artery decreased by 19% in patients with PPH and 15% in patients with NPPH, with values being significantly different from pre-inhalation data for both groups. In contrast, during intravenous prostacyclin infusion, pulmonary artery ANP levels did not change, despite a significant hemodynamic improvement (13). This discrepancy may be explained by the lower basal ANP levels in that study (mean values approximately 350 pg/ml) or by the different hemodynamic profile of intravenous prostacyclin in comparison with inhaled iloprost, with a more systemic than pulmonary pressure decrease occurring in response to prostanoid infusion (14).
The correlation of ANP changes with hemodynamic changes (Fig. 2)scarcely showed any differences compared with the baseline correlations (Fig. 1). The best correlations were found between ANP and parameters of RV preload and function (RAP and CI). This might suggest that the dynamic control of the ANP secretion is dependent on right atrial and ventricular conditions.
Cyclic GMP levels in the pulmonary artery decreased by 30% during iloprost inhalation in PPH and NPPH patients. This might be explained by the decreasing ANP levels; however, there was no significant correlation between the changes in cGMP levels and ANP levels. As discussed above, the plasma appearance of cGMP is influenced by a large number of variables, many of which may be altered by iloprost inhalation. This renders any interpretation of the rapid cGMP decline in response to the prostanoid aerosolization uncertain.
We conclude that the ANP (and the cGMP) system is, to the same extent, massively activated in severe PPH and NPPH. Baseline ANP levels are significantly correlated with parameters of RV function as well as pre- and afterload. As a rapid decline of ANP occurs in parallel with hemodynamic improvement in response to iloprost inhalation in PPH and NPPH, it seems that the regulation of the ANP secretion is intact even in states of strong, long-term stimulation. Therefore, ANP levels may be considered as a non-invasive marker of RV failure. Further investigations into the neurohumoral system in right heart failure are needed.
The authors thank Mary Kay Steen-Mueller for carefully reviewing the manuscript.
☆ Supported by grants of the “Deutsche Forschungsgesellschaft (DFG).”
- atrial natriuretic peptide
- cyclic guanosine monophosphate
- cardiac index
- cardiac output
- chronic obstructive pulmonary disease
- computed tomography
- endothelial nitric oxide synthase
- mean pulmonary arterial pressure
- mean systemic arterial pressure
- nonprimary pulmonary hypertension
- pulmonary artery wedge pressure
- primary pulmonary hypertension
- pulmonary vascular resistance
- right atrial pressure
- right ventricle, right ventricular
- right ventricular ejection fraction
- mixed venous oxygen saturation
- Received December 15, 2000.
- Revision received May 31, 2001.
- Accepted June 19, 2001.
- American College of Cardiology
- Cargill R.I,
- Lipworth B.J
- Pedram A,
- Razandi M,
- Hu R.M,
- Levin E.R
- Ou L.C,
- Sardella G.L,
- Hill N.S,
- Thron C.D
- Nootens M,
- Kaufmann E,
- Rector T,
- et al.
- Morice A.H,
- Pepke-Zaba J,
- Brown M.J,
- Thomas P.S,
- Higenbottam T.W
- Hoeper M.M,
- Olschewski H,
- Ghofrani H.A,
- et al.
- Gessler T,
- Schmehl T,
- Hoeper M.M,
- et al.
- Edwards B.S,
- Zimmerman R.S,
- Schwab T.R,
- Heublein D.N,
- Burnett J.C Jr.
- Baertschi A.J,
- Jiao J.H,
- Carlson D.E,
- et al.
- Nagaya N,
- Nishikimi T,
- Okano Y,
- et al.
- Klinger J.R,
- Arnal F,
- Warburton R.R,
- Ou L.C,
- Hill N.S
- Bogdan M,
- Humbert M,
- Francoual J,
- et al.
- ↵Muramatsu M, Tyler RC, Gutkowska J, et al. Atrial natriuretic peptide accounts for increased cGMP in hypoxia-induced hypertensive rat lungs. Am J Physiol 1997;272:L1126–32.