INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Primary (PPH) and secondary pulmonary hypertension (SPH) remain an unresolved clinical challenge, linked with high mortality (1). Imbalances of vasodilatory and vasoconstrictive forces have been implicated in both the predominance of increased vasomotor tone and the chronic remodeling of resistance vessels, including vascular smooth muscle cell (VSMC) growth. In patients with PPH, a reduced excretion of prostaglandin I₂ (PGI₂) and an enhanced excretion of thromboxane metabolites were noted (4). Similarly, expression of the constitutive nitric oxide (NO) synthase was found to be less in the pulmonary resistance vessels of patients with PPH (5). Moreover, enhanced activities of phosphodiesterases (PDEs), which hydrolyze the PGI₂- and NO-induced second messengers cAMP and cGMP, were observed in experimental conditions of pulmonary hypertension (6).

Systemic vasodilator therapies, in particular intravenous PGI₂ and oral calcium channel blockers, are used for treatment of severe pulmonary hypertension (7). Possible disadvantages of this approach include systemic vasodilation and ventilation-perfusion mismatch with impairment of arterial oxygenation. Attempts to circumvent these disadvantages employed inhalative techniques of vasodilator application, aiming to achieve selective pulmonary vasodilation and supraselective vasodilation in well-ventilated (i.e., inhaled vasodilator-accessible) areas within the lung parenchyma. Continuous inhalation of NO (11, 12) or aerosolized PGI₂ (13) was indeed shown to selectively reduce the pulmonary artery pressure and to improve arterial oxygenation in patients with adult respiratory distress syndrome (ARDS) and moderate pulmonary hypertension. Moreover, repetitive nebulization of the long- acting prostacyclin analog iloprost has been demonstrated to result in substantial lowering of pulmonary artery pressure and resistance and improvement of hemodynamics in severe pulmonary hypertension (17).

The long-term clinical use of inhaled prostanoids, administered during repetitive brief aerosolization periods, is hampered by the fact that owing to the short biological half-life of prostacyclin (2-3 min at physiological pH), its pulmonary vasodilatory effect levels off within < 30 min after termination of nebulization (18). Nebulized iloprost is effective for 60 to 120 min, but still needs multiple daily nebulization periods to result in sustained relief of pulmonary hypertension. One approach to prolong and possibly increase the vasorelaxant properties of these agents might be the concomitant use of PDE inhibitors. The cyclic nucleotide levels within the cells are regulated by a group of nucleotide phosphodiesterases that includes several distinct isoenzymes (20). The role of several "families" of PDE has been characterized mostly by employment of isoenzyme-selective (monoselective) PDE inhibitors in biochemical and functional studies. In human pulmonary arteries, originating from patients undergoing thoracotomy for lung cancer, the presence of the PDE isoenzymes 1, 3, 4, and 5 in the cytosolic and particulate phases (of homogenized tissue) has been demonstrated (24). PDE 3 does possess high affinity for both cAMP and cGMP, with a Vmax for cAMP usually greater than that for cGMP (20). PDE 4 enzymes are characterized by their high affinity for cAMP, with cGMP representing a poor substrate. In human airway smooth muscle cells, PDE 3 and PDE 4 are the essential players coregulating the cAMP content, which may also hold true for human pulmonary vascular SMCs (23). In contrast, PDE 5 has a relatively high affinity for cGMP and hydrolyzes cAMP poorly. Interestingly, the type 3 PDE has the distinctive property that is significantly inhibited in the presence of elevated cellular levels of cGMP ("cGMP-inhibited PDE"), thus representing an important site of crosstalk between the cAMP- and the cGMP-centered signal transduction pathways.

We investigated the efficacy of monoselective and dual- selective PDE inhibitors in a model of stable pulmonary hypertension in intact rabbits. By employing subthreshold doses of inhibitors of PDE 3, PDE 4, and PDE 5, which per se did not affect pulmonary and systemic hemodynamics, a pronounced prolongation and partially enhancement of the pulmonary vasodilatory effect of short-term PGI₂ aerosolization was noted, in the absence of systemic arterial pressure drop. Dual inhibition of PDE 3 and PDE 5 was found to be most effective in this respect. Low-dose systemic PDE inhibitors may thus amplify the selective pulmonary vasodilatory effect of inhaled prostanoids.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Materials

The thromboxane A₂ mimetic U46619 and the PDE inhibitor zaprinast was supplied by Sigma (Deishofen, Germany), and PGI₂ (Epostenol) was provided by Wellcome (London, UK). Motapizone was from Nattermann (Cologne, Germany), rolipram was from Schering AG (Berlin, Germany), and the dual-selective PDE inhibitors zardaverine and tolafentrine were from Byk-Gulden Pharmaceuticals (Konstanz, Germany). All other chemicals and drug supplies were from standard commercial sources.

Surgical Preparation

Rabbits were initially anesthetized with a mixture of xylazine (2.1 mg/ kg) and ketamine (7 mg/kg), followed by a constant intravenous infusion of xylazine [25 mg/(kg/h)] and ketamine [80 mg/(kg/h)] (Injectomat S; Fresenius, Bad Hamburg, Germany) through the right peripheral ear vein. They were anticoagulated with heparin (200 U/kg). Tracheostomy was performed and the animals were ventilated with an FI_O2 of 0.5, using a volume-controlled respirator (cat ventilator; Hugo Sachs Elektronik, March Hugstetten, Germany). Frequency was set at 40 breaths/min and tidal volume at 8 ml/kg, resulting in a Pa_CO2 ranging between 35 and 45 mm Hg. A positive end-expiratory pressure of 0.5 mm Hg was used throughout. A catheter was inserted into the left carotid artery and connected to a pressure transducer for arterial pressure monitoring, and the right femoral vein was cannulated for infusion of saline and PDE inhibitors. A balloon-tipped pulmonary artery catheter (Berman angiographic balloon catheter AI-07134, 4F; Arrow, Reading, PA) was inserted into the pulmonary artery through the right external jugular vein.

Hemodynamics and Blood Gases

Mean pulmonary artery pressure () and mean aortic pressure () were continuously recorded using fluid-filled pressure transducers (Braun; Combitrans, Melsungen, Germany); the level of the left atrium was the zero reference for the measurements. Pulmonary vascular resistance (RL) was computed using the standard formula. Cardiac output () was calculated by the Fick principle, employing the mixed venous oxygen content, arterial oxygen content, and oxygen uptake. Oxygen uptake of the animals was measured online (O₂ controller; Labotect, Goettingen, Germany). Arterial and mixed venous samples were collected (1 ml) and maintained on ice until analyzed for PO₂, pH, and PCO₂ (ABL330; Radiometer, Copenhagen, Denmark). Hemoglobin and oxygen saturation was measured using an OSM2 hemoximeter (Radiometer).

Aerosolization of PGI₂

For aerosol delivery of prostacyclin, an ultrasonic nebulizer (Pulmo Sonic 5500; DeVilbiss Medizinische Produkte GmbH, Langen, Germany) was positioned in the inspiratory limb of the ventilator tubing. The aerosolized mass was calculated from the weight loss of the nebulizer during aerosolization. Prostacyclin was dissolved in glycine buffer (10 µg/ml) and diluted in isotonic saline to a final concentration of 2 µg/ml. The ultrasonic nebulizer produced an aerosol with a mass median aerodynamic diameter of 4.5 µm and a geometric standard deviation of 2.5, as measured with a laser diffractometer (Helos; Sympatec, Clausthal-Zellerfeld, Germany).

Experimental Protocols

Immediately after tracheotomy, all animals received a baseline infusion of 20 ml of sterile Krebs-Henseleit buffer per hour. In pilot studies, dose-effect curves for intravenous U46619 were established. There was some variation in the responsiveness from animal to animal, and a dosage in the range between 0.5 and 2 µg/kg · min was found to be required for augmentation of from ~ 13 to ~ 30 mm Hg within 20 min.

For establishing dose-effect curves of the PDE inhibitors in animals with U46619-elicited pulmonary hypertension, short-term infusions (10 min) of the various agents at different doses were undertaken, aiming at RL reduction of at least 20%. In addition, combinations of different PDE inhibitors (rolipram/motapizone rolipram/zaprinast, zaprinast/motapizone) were tested in the same mode. Hemodynamics and blood gases were measured at the end of the 10-min infusion period.

Prostacyclin nebulization was performed after a preceding period of stable U46619-induced pulmonary hypertension of at least 20 min, without interruption of the ongoing U46619 infusion. Aerosolization periods were 10 min throughout. In pilot experiments addressing the dose-effect relationship of this agent, a dosage of 56 ± 10 ng of nebulized PGI₂/kg · min was found to be appropriate to achieve selective pulmonary vasodilation in all animals, with rapid decline of the PGI₂ efficacy after termination of inhalation (data not shown in detail). For assessing possible augmentation of the prostacyclin effect by subthreshold doses of intravenous PDE inhibitors, the following sequence was performed. First, stable pulmonary hypertension was established by individually titrating U46619 infusion. Second, nebulization of PGI₂ (10 min) was performed, followed by a second baseline period in the absence of this agent. Third, loading with a PDE inhibitor followed by continuous infusion of this agent was started in a subthreshold concentration range, known to leave cardiac output, , and unaffected. The dosage was taken from the dose-effect curves established in the preceding experiments. It was (loading dose [steady state infusion]) 5.5 µg/kg (15 µg/kg · h) for rolipram, 2.2 µg/kg (8 µg/ kg · h) for motapizone, 50 µg/kg (300 µg/kg · h) for zardaverine, 100 µg/kg (3.3 mg/kg · h) for zaprinast, and 100 µg/kg (2 mg/kg · h) for tolafentrine. Fourth, after 15 min of PDE infusion, while observing a stable plateau, a second PGI₂ nebulization period was performed and values of the groups in the presence of PDE inhibition (Roli/ PGI₂, Mota/PGI₂, Zarda/PGI₂, Tola/PGI₂) were compared with values of the first PGI₂ nebulization period (without PDE inhibitor). Hemodynamics and blood gases were measured at the end points of each 10-min aerosolization period.

Rolipram, motapizone, zaprinast, and zardaverine were dissolved in 50% dimethyl sulfoxide (DMSO)-H₂O. In separate control experiments, the final DMSO concentration of 0.4% was shown not to influence hemodynamics in control animals or in those undergoing U46619-elicited pulmonary hypertension.

Data Analysis

Data are shown as means ± SEM. Differences between the various groups were assessed by use of analysis of variance and the Student- Newman-Keuls test for multiple comparisons, with a p value < 0.05 regarded to be significant. Differences between means of two groups were analyzed by unpaired t test.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Baseline and U46619-induced Pulmonary Hypertension

As detailed in Table 1, baseline hemodynamic data and blood gases of the animals were in physiological ranges. After a stable steady state period, the infusion of U46619 (a mean of 1.3 ± 0.5 µg/kg · min) resulted in significant pulmonary hypertension: when averaging for all groups, increased from 14.4 ± 3.9 to 28.3 ± 3.1 mm Hg (p < 0.01). Mean aortic pressure () did not change significantly, while cardiac output decreased from 471 ± 13 to 411 ± 13 ml/min (all groups, p < 0.05). There were no significant changes in blood gases as compared with control animals.

Dose-Effect Curves of PDE Inhibitors

As shown in Figure 1, all PDE inhibitors effected dose-dependent reduction of the elevated pulmonary vascular resistance in animals with U46619-elicited pulmonary hypertension. Motapizone showed the highest efficacy (dose range, 2.5-25 µg/kg body weight [bw]), followed by rolipram (range 5-250 µg/kg bw) and zardaverine (range, 50-3,000 µg/kg bw). Higher concentrations of the PDE 5 inhibitor zaprinast (range, 100-10,000 µg/kg bw) and the PDE 3/4 inhibitor tolafentrine (range, 500- 5,000 µg/kg bw) were necessary to effect a significant decrease in RL values.

View larger version (15K): [in this window] [in a new window]   Figure 1.   Dose-effect curves of intravenous PDE inhibitors on U46619-elicited pulmonary hypertension. Pulmonary vascular resistance (RL [PVR] as a percentage of U46619-induced increase) is given (mean ± SEM of six independent experiments each). PDE inhibitors were applied in various doses as short-term infusions (10 min) in randomized order.

The entire profile of changes in hemodynamics and blood gases in response to the different PDE inhibitors was assessed in separate experiments, in which each inhibitor was chosen at a concentration effecting an ~ 20% reduction of RL values (Figure 2). This was 25 µg/kg for rolipram, 5 µg/kg for motapizone, 500 µg/kg for zardaverine (not shown in detail), 1 mg/kg for zaprinast, and 1 mg/kg for tolafentrine. As detailed in Table 2, the application of all PDE inhibitors resulted in a moderate decrease in , which was significant for zardaverine and tolafentrine. In all groups a significant increase in cardiac output was noted (p < 0.05, all groups). In addition, combinations of the selective PDE inhibitors at the given concentrations were administered, aiming to suppress phosphodiesterase types 3 plus 4, 3 plus 5, or 4 plus 5. As evident from Figure 2, all combinations resulted in an additive effect on RL reduction.

View larger version (17K): [in this window] [in a new window]   Figure 2.   Influence of fixed doses of intravenous PDE inhibitors and combinations of these agents on U46619-elicited pulmonary hypertension. Pulmonary vascular resistance (RL [PVR], as a percentage of U46619-induced increase) is given (mean ± SEM of six independent experiments each). PDE inhibitors were applied as short-term infusions (10 min). The doses of the single compounds were chosen to decrease the RL by ~ 20% (all PDE inhibitor- induced RL changes differed significantly from the U46619 control).

Nebulization of Prostacyclin (PGI₂)

Inhalation of aerosolized PGI₂ at a dose of 56 ng/kg · min for 10 min resulted in a significant decrease in , reducing the U46619-induced pressor response by nearly 30% (Figure 3). The vasodilatory effect started within 2 min after onset of nebulization. Immediately after stopping the aerosol application, started to rise again, and prenebulization values of were reached within 8 min. The pulmonary vasodilatory effect of nebulized PGI₂ was not accompanied by significant peripheral vasodilation, as the values did not decrease, but rather increased slightly (Table 1). The prostacyclin-elicited decrease was accompanied by a significant increase of cardiac output, from 393 ± 40 to 458 ± 67 ml/min (p < 0.01). No significant changes in blood gases occurred (Table 1).

View larger version (25K): [in this window] [in a new window]   Figure 3.   Influence of PGI2 nebulization and its combination with subthreshold doses of intravenous PDE inhibitors on U46619-elicited pulmonary hypertension. Pulmonary artery pressure (Ppa, as a percentage of U46619-induced increase) is given (mean ± SEM of eight independent experiments each; SEM bars are missing when falling into symbol). PGI2 nebulization (56 ng/kg · min) is indicated by the horizontal bar. All PDE inhibitors were preapplied as a bolus (with maintenance infusion) in the following concentrations: rolipram at 5.5 µg/kg (15 µg/kg · h), motapizone at 2.2 µg/ kg (8 µg/kg · h), and zaprinast at 100 µg/kg (3.3 µg/kg · h). Compared with sole PGI2 inhalation, all time points from 10 to 30 min were significantly different for PGI2/motapizone (p < 0.05), as were all time points from 10 to 34 min for PGI2/rolipram, and all time points from 12 to 40 min for PGI2/zaprinast.

Combined Subthreshold Administration of PDE Inhibitors and PGI₂ Nebulization

As detailed under METHODS, loading and maintenance infusion doses of the different PDE inhibitors in these studies were derived from the dose-inhibition curves, targeting a subthreshold dosage with no effect of the PDE inhibition on the pulmonary vascular tone per se. Indeed, no changes in hemodynamics and gas exchange were provoked by these regimens of PDE inhibitor infusion in animals with U46619-elicited pulmonary hypertension. All modes of subthreshold systemic PDE inhibitor administration did, however, significantly affect the responsiveness to the standardized PGI₂ nebulization. In the presence of rolipram (5.5 µg/kg), the PGI₂-induced decrease was not substantially augmented (73.0 ± 2.6 versus 74.0 ± 2.9%), but after stop of nebulization the vasodilatory effect, defined by a value below 95% of the U46619-induced pressure plateau, was significantly prolonged, from 8 to 22 min (Figure 3). Comparable efficacy was noted for zaprinast (100 µg/kg), with prolongation of the post-PGI₂ nebulization vasodilatory effect to 30 min. Motapizone (2.2 µg/kg) both enhanced the maximum decrease in response to PGI₂ aerosolization (lowering to 52.5 ± 6.2% instead of 74.0 ± 2.9% of the U46619-induced pressure elevation; p < 0.01) and it prolonged the postnebulization vasodilatory effect to 20 min. These vasomotor effects in the pulmonary vasculature were accompanied by an enhancement of cardiac output, whereas systemic arterial pressures and gas exchange variables remained unchanged (Table 1).

Combination of inhaled PGI₂ with the dual-selective PDE inhibitors zardaverine (50 µg/kg) and tolafentrine (100 µg/kg) provoked both amplification of the maximum PGI₂-induced vasodilatory response (to 64.8 ± 7.0% [zardaverine; NS] and 56.6 ± 3.8% [tolafentrine; p < 0.01] instead of 74.0 ± 2.9% [absence of PDE inhibitor] of the U46619-induced pressure elevation) and significant prolongation of the postnebulization vasodilatory efficacy of PGI₂ (28 min in the presence of both agents) (Figure 4).

View larger version (21K): [in this window] [in a new window]   Figure 4.   Influence of PGI2 nebulization and its combination with subthreshold doses of dual-selective intravenous PDE inhibitors on U46619-elicited pulmonary hypertension. Pulmonary artery pressure (Ppa, as a percentage of U46619-induced increase) is given (mean ± SEM of eight independent experiments each; SEM bars are missing when falling into symbol). PGI2 nebulization (56 ng/kg · min) is indicated by the horizontal bar. The PDE inhibitors were preapplied as a bolus (with maintenance infusion) in the following concentrations: zardaverine at 50 µg/kg (300 µg/kg · h), tolafentrine at 100 µg/kg (2 mg/kg · h). Compared with sole PGI2 inhalation, all time points from 14 to 34 min were significantly different for PGI2/zardaverine (p < 0.05), as were all time points from 8 to 40 min for PGI2/tolafentrine.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The model of pulmonary hypertension evoked by infusion of U46619 is commonly used for testing vasodilatory agents (25- 30). The currently employed dose range of the stable thromboxane analog resulted in approximate doubling of the levels and a moderate decrease in cardiac output, without affecting systemic arterial pressure or blood gases. The pulmonary hypertensive response was stable and fully reversible. In accordance with previous studies suggesting that the precapillary resistance vessels are the predominant site of the vasoconstrictor effect of U46619 (25), no evidence of lung edema formation was obtained by performing postmortem wet-to-dry ratio measurements of the lungs previously undergoing U46619 infusion (data not shown).

In the absence of PGI₂, systemic administration of the PDE 3 and PDE 4 inhibitors motapizone and rolipram caused dose-dependent pulmonary vasodilation in rabbits with U46619-elicited hypertension. This approach did not, however, possess pulmonary selectivity, as it was accompanied by a decrease in . When combined, motapizone and rolipram were additive in effecting pulmonary vasodilation. This corresponds well with preceding studies in PGF₂-constricted human pulmonary artery rings, in which the combination of motapizone and rolipram turned out to be most effective in inducing relaxation (24). In human airway smooth muscle cells, PDE 3 and PDE 4 are the essential players coregulating the cAMP content, which may also hold true for human pulmonary vascular SMCs (23). Synergistic interaction of PDE 3 and PDE 4 inhibition was also demonstrated in cultured vascular SMCs, in an assessment of cell proliferation (31). These data suggested that selective inhibition of PDE 3 or PDE 4 might be partly compensated for by increased hydrolysis of cAMP through the alternate catabolism pathway, and full efficacy in elevating cAMP levels is forwarded by dual inhibition of both PDE 3 and PDE 4. Similar to motapizone and rolipram, intravenous administration of the PDE 5 blocker zaprinast effected dose-dependent pulmonary vasodilation, however, again accompanied by a drop in values. This finding is in line with the previous observation that PDE 5 inhibition lowers RL values in newborn and fetal lambs with acute pulmonary hypertension (32). These effects are most probably ascribed to stabilization of cGMP generated continuously in response to baseline NO synthesis in the pulmonary circulation. Zaprinast also demonstrated additive effects with both motapizone and rolipram to lower the U46619-evoked increase. This finding corresponds to the previous observation that the PDE 3 and PDE 5 inhibitors milrinone and dipyridamole reduce elevated pulmonary vascular resistance in in situ-perfused lungs in a synergistic fashion (36).

Aerosolized PGI₂ reversibly decreased the pulmonary artery pressure, while the values remained unaffected and the cardiac index even increased. Calculated RL data were thus markedly reduced, Although only one-point RL measurements were performed, the PGI₂ effect on clearly rules out the possibility that the RL decrease in response to the prostanoid might reflect only flow-related resistance changes, but active pulmonary vasorelaxation is unequivocally the predominant mechanism. The pronounced efficacy of PGI₂ to cause selective pulmonary vasodilation corresponds well to preceding studies of anesthetized sheep with U46619-induced pulmonary hypertension (30).

Wagner and coworkers (6) previously demonstrated an enhancement of isoproterenol- and forskolin-induced relaxation of isolated pulmonary artery rings, originating from rats with hypoxia-induced pulmonary hypertension, by inhibition of PDEs with milrinone (PDE 3) and rolipram (PDE 4). In the current study, subthreshold doses of motapizone and rolipram, effecting per se no hemodynamic effects, substantially prolonged (both agents) and augmented (motapizone) the pulmonary vasodilatory effect of nebulized PGI₂. Interestingly, this effect, most probably ascribed to a stabilization of PGI₂-induced cAMP, was not accompanied by a decrease in values, i.e., pulmonary selectivity was maintained. The most prominent amplification of the pulmonary vasodilatory effect of nebulized PGI₂ was achieved on engagement of the dual-selective (PDE 3 and 4) inhibitors zardaverine and tolafentrine, again in doses effecting per se no hemodynamic effects. Both agents significantly augmented the maximum vasodilatory response to this prostanoid and prolonged the post-PGI₂ reduction to > 30 min, still limiting the vasorelaxant effect to the pulmonary circulation (unchanged levels).

Phosphodiesterase type 3 apparently plays a predominant role in regulating intracellular cAMP content and smooth muscle tone. As indicated in Figure 1 (dose-response curves of intravenous PDE inhibitors), motapizone was the most effective vasorelaxant agent in the rabbit pulmonary vasculature. All agents with PDE 3 inhibitory potency amplified the prostanoid-induced vasodilation. This is in line with the important role of PDE 3 in regulating cAMP breakdown in human smooth muscle cells in vitro (24). When comparing the IC₅₀ values of the various PDE inhibitors for PDE 4, rolipram (0.1 µM), zardaverine (0.16 µM), and tolafentrine (0.1 µM) are known to possess comparable values, which contrasts with the large differences in the currently noted threshold doses and may suggest a minor role for PDE 4 in regulating cAMP breakdown in the rabbit vasculature. Motapizone and tolafentrine are strong inhibitors of PDE 3, and IC₅₀ values of 0.05 and 0.06 µM, while zardaverine inhibits PDE 3 with an IC₅₀ of 0.58 µM. Against this background the present data suggests that in particular the PDE 3-inhibitory capacities of the dual-selective inhibitors are relevant for amplification of the PGI₂-induced vasodilatory response, and that suppression of both PDE 3 and PDE 4 forwards most effective prolongation of this response.

Interestingly, subthreshold doses of the PDE 5 inhibitor zaprinast were as effective as rolipram in prolonging the pulmonary vasodilatory effect of aerosolized PGI₂. In previous studies, similar efficacy of zaprinast was demonstrated for coapplication with inhaled NO, easily attributable to stabilization of cGMP arising due to NO-related stimulation of guanylate cyclase (28, 29). It is in the same line that the cGMP- specific PDE inhibitor dipyridamole was noted to potentiate pulmonary vasodilation induced by NO in the ovine fetus and to attenuate the rebound pulmonary hypertension after inhaled NO withdrawal in postoperative congenital heart disease (35, 37). The current observation of enhanced potency of the PGI₂-cAMP axis in the presence of zaprinast does, however, suggest close interaction between cAMP- and cGMP-mediated vasorelaxant effects. This observation is in line with studies of isolated vascular strips, in which the NO/cGMP pathway was noted to amplify cAMP-mediated vasorelaxation, most probably via inhibition of the cGMP-sensitive PDE 3 (38). Inhalation of zaprinast in awake sheep with U46619-induced pulmonary hypertension increased transpulmonary plasma cGMP differences (29) four-to-fivefold, and McMahon and co-workers reported increased plasma cGMP levels in anesthetized rats after infusion of zaprinast at 1 mg/ kg · min (39). Intracellular PDE 3, which seems to play an essential role in regulating cAMP content, as suggested by the experimental data from Figure 1, is inhibited by cGMP with an IC₅₀ of 0.1-1 µM (40): an intracellular increase in cGMP by zaprinast may thus well result in a significant inhibition of PDE 3 and therefore a prolongation of the PGI₂-induced vasodilation, as previously suggested (38). Although not directly addressed in an experimental fashion in the present study, such efficacy via the cGMP-PDE 3 axis might more easily explain the zaprinast effect than some direct effect of this PDE inhibitor on the cAMP catabolism or some (unknown) efficacy of aerosolized PGI₂ via elevated cGMP levels, although these alternative explanations may not be excluded.

None of the combinations of intravenous PDE inhibitors and nebulized PGI₂ caused any deterioration of arterial oxygenation in the present study. Although ventilation-perfusion matching was not directly addressed, these findings strongly support the view that the pulmonary vasodilatory effect was not accompanied by enhanced mismatch, as often observed in response to systemic vasodilator therapy. Preferred vasodilation in well ventilated lung areas, directly accessible to the nebulized prostanoid, may thus hold true even in the presence of circulating levels of low-dose PDE inhibitors.

In conclusion, the current study established dose-effect curves of mono- and dual-selective inhibitors of PDE 3, 4, and 5 in intact rabbits with pulmonary hypertension, demonstrating synergism when combining PDE 3 plus 4, 3 plus 5, or 4 plus 5 inhibition. Although species differences may not be neglected, the currently available data on human lung smooth muscle PDE composition suggest that such synergism may also hold true for the pulmonary vasculature. Most prominently, subthreshold systemic doses of the monoselective and in particular the dual-selective PDE inhibitors all significantly amplified the pulmonary vasodilatory response to inhaled PGI₂, while fully maintaining the pulmonary selectivity of the prostanoid-elicited vasorelaxation. Combining low-dose systemic PDE administration may thus represent a therapeutic strategy for enhancement and prolongation of the efficacy of nebulized prostanoids to alleviate pulmonary hypertension while fully maintaining systemic arterial pressure and ventilation-perfusion matching.