Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

The acute respiratory distress syndrome (ARDS) is characterized by pulmonary hypertension, lung edema formation, and severe ventilation-perfusion mismatch with a predominance in shunt flow (Rossaint et al., 1993; Walmath et al., 1993). Intravascularly administered vasodilators such as prostacyclin (PGI₂) and prostaglandin (PG) E₁, used to decrease lung (micro-)vascular pressures, are hampered by their nonselective mode of action, with vasorelaxant properties in both the pulmonary and systemic vessels and in both well-ventilated and nonventilated (shunt) areas within the lung vasculature, thereby disadvantageously increasing shunt flow (Mélot et al., 1989; Radermacher et al., 1990). Selective pulmonary vasodilation, in contrast, was first described for inhaled nitric oxide (NO) in both acute respiratory failure and chronic pulmonary hypertension (Frostell et al., 1991; Pepke-Zaba et al., 1991; Rossaint et al., 1993). Due to the transbronchial route of application, the vasodilatory property of this gas-borne agent is restricted to the lung vasculature. In addition, a regional decrease in pulmonary vascular resistance in well ventilated lung regions effects redistribution of blood flow with reduced perfusion of shunt areas. Considering possible disadvantages of NO because of its free radical features with impact on inflammatory events (Troncy et al., 1997), several groups used aerosolization technology via the transbronchial route of application for the vasodilatory prostanoid PGI₂, an agent with a well known pharmacological profile that has been used for many years (Higgenbottam et al., 1987; Hardy et al., 1988). With this agent, a decrease in pulmonary arterial pressure (pPA), with maintenance of systemic arterial pressure, concomitant with a decrease rather than an increase in shunt flow was demonstrated in patients suffering from severe ARDS (Walmrath et al., 1993, 1995, 1996; Zwissler et al., 1996). Nebulization of PGI₂ and its longer-acting analog iloprost were also demonstrated to result in substantial lowering of pPA as well as resistance and improvement in hemodynamics in nonventilated patients with severe primary and secondary pulmonary hypertension (Olschewski et al., 1996, 1998a,b). The clinical use of inhaled PGI₂ is, however, hampered by the fact that, because of the short biological half-life of this agent (2-3 min at physiological pH; Moncada and Vane, 1980), its vasodilatory effect in the lung vasculature levels off within <30 min after termination of nebulization (Olschewski et al., 1996). This window of efficacy is prolonged to 60 to 90 min on aerosolization of the more stable iloprost. Coapplication of phosphodiesterase (PDE) inhibitors for stabilization of the PGI₂-induced second-messenger cAMP may offers the possibility of prolonging and possibly increasing the vasodilatory effect of nebulized PGI₂ in the lung vasculature. In vitro studies in human pulmonary arteries, originating from patients undergoing surgery for lung cancer, demonstrated the presence of the PDE isoenzymes 1, 3, 4, and 5 (Rabe et al., 1994). PDE types 3 and 4 are relevant for cAMP catabolism in many tissues (Beavo, 1995; Conti et al., 1995; Manganiello et al., 1995; Torphy, 1998) and were shown to coregulate the cAMP content in human bronchial smooth muscle cells (Torphy, 1998). PDE5 has little direct impact on cAMP hydrolysis but is the predominant regulator of the cGMP content, thereby indirectly influencing cAMP via inhibition of PDE3. The current study was performed in perfused rabbit lungs with continuous infusion of the stable thromboxane A₂ mimetic U46619 for establishing stable pulmonary hypertension with severe disturbances in gas exchange. Using short-term aerosolization of PGI₂, we investigated the effects of prior administration of subthreshold doses of selective PDE inhibitors on the PGI₂-elicited vasodilator and gas exchange response.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Materials. Sterile Krebs-Henseleit hydroxyethylamylopectine buffer was obtained from Serag-Wiessner (Naila, Germany). The thromboxane A₂ mimetic U46619 was supplied by Paesel-Lorei (Frankfurt, Germany) and PGI₂ (Epostenol) by Wellcome (London, England). Motapizone was obtained from Nattermann (Köln, Germany), rolipram from Schering A.G. (Berlin, Germany), and the dual-selective PDE inhibitor tolafentrine from Byk Gulden Pharmaceuticals (Konstanz, Germany). All other chemicals were purchased from Merck (Darmstadt, Germany). The ultrasonic nebulizer Pulmo Sonic 5500 was obtained from DeVilbiss Medizinische Produkte GmbH (Langen, Germany).

Isolated-Lung Model. The perfused lung model has been described in detail (Seeger et al., 1994). Briefly, rabbits of either sex weighing 2.6 to 2.9 kg were anticoagulated with heparin (1000 U/kg) and anesthetized with i.v. ketamine plus xylazine. Tracheostomy was performed, and the animals were ventilated with room air via a Harvard respirator (tidal volume, 9-13 ml/kg; frequency, 10 breaths/min; positive end expiratory pressure, 1 mm Hg). After midsternal thoracotomy, catheters were placed into the pulmonary artery and the left atrium, and perfusion with Krebs-Henseleit buffer was started. The lungs were perfused with a constant flow of 120 ml/min. Left atrial pressure was set at 1.2 mm Hg in all experiments. In parallel with the onset of artificial perfusion, room air supplemented with 4% CO₂ was used for ventilation. Lungs were freely suspended from a force transducer for monitoring of organ weight. Pressures in the pulmonary artery, left atrium, and trachea were registered (zero referenced at the hilum). Perfusate samples (total perfusate volume, 500 ml) were taken from both the arterial and venous part of the system. Gas samples were taken from the outlet of an expiration gas mixing box. The whole system was heated to 37°C.

Aerosolization. PGI₂ and the PDE inhibitors were aerosolized with an ultrasonic device (Pulmo Sonic 5500). This nebulizer produces an aerosol with a mass median aerodynamic diameter of 4.5 µm and a geometric S.D. of 2.6, as measured with a laser diffractometer (HELOS; Sympatec, Clausthal-Zellerfeld, Germany). The nebulizer was located between the ventilator and the lung to be passed by the inspiration gas. The nebulization system was described previously by Schermuly et al. (1997). For a given ventilator setting, an absolute deposition fraction of 0.25 ± 0.02 was determined by a laser photometric technique (Schmehl et al., 1996).

Ventilation-Perfusion (A/) Determination in Isolated Lungs. The continuous A/ distributions were determined by the multiple inert gas elimination technique as described by Wagner et al. (1974), which has been adapted to assess gas exchange of blood-free perfused rabbit lungs (Walmrath et al., 1993). Briefly, six inert gases (sulfur hexafluoride, ethane, cyclopropane, halothane, diethyl ether, and acetone) were dissolved in isotonic saline and continuously infused at a rate of 0.3 ml/min. After an equilibration period of at least 30 min, 10-ml perfusate samples were simultaneously collected from the pulmonary artery and the left atrium. A corresponding 30-ml gas sample was drawn from the heated expiration gas mixing chamber. Each sampling was performed in duplicate. Extraction of the gases dissolved in the buffer fluid was carried out by equilibration (40 min) with nitrogen in a shaking water bath (37°C). The gas phases after equilibration of the buffer samples and the exhaled gases were analyzed by gas chromatography, as described (Walmrath et al., 1993). The ratios of arterial to mixed venous partial pressures (retention) and of expired to mixed venous partial pressures (excretion) were calculated for each gas and were plotted against buffer-gas partition coefficients (retention and excretion solubility curves). For each gas, the retention and the excretion were used for estimation of the A/ distribution by least-squares analysis with enforced smoothing with a computer program. The position of the distribution was also described by the mean A/ ratio for perfusion and ventilation and their dispersion by the log S.D. of both perfusion and ventilation (log S.D. , log S.D. A). These parameters of dispersion do not take into account either shunt or dead space. The residual sum of squares (RSS) was the result of testing the compatibility of the inert-gas data with the derived A/ distribution by the least-squares method. An indication of acceptable quality of the A/ distributions is an RSS of 5.348 in 50% of the experimental runs (50th percentile) or 10.645 in 90% of the experimental runs (90th percentile) (Wagner and West, 1984). In this study, 95% of the RSSs were <5.348, and 98.8% were <10.645.

Determination of cAMP in the Recirculating Buffer. Five-hundred-microliter samples of perfusate were collected at 0, 30, 45, 60, 75, 105, and 135 min and frozen in liquid nitrogen. cAMP was measured by a radioimmunoassay kit (Immunotech, Marseilles, France). The perfusate samples or standards were incubated with ¹²⁵I-labeled cAMP in antibody-coated tubes. After incubation, the contents of the tube were aspirated, and bound radioactivity was counted in a gamma counter. Duplicate samples were performed, and the values were calculated by a standard curve. The cAMP levels are given as picomoles per milliliter.

Experimental Protocols. As described previously (Walmrath et al., 1997), a sustained increase in pPA from 6 to 32 mm Hg was achieved by continuous infusion of 70 to 160 pmol · kg¹ · min¹ of U46619; individual titration was performed. This level of pulmonary hypertension was then maintained for at least 150 min, with variations in pPA of <2 mm Hg. In pilot studies, aerosolized PGI₂ was applied via inhalation in increasing doses, and a dose of 27 pmol · kg¹ · min¹ was found to decrease pPA by ~4.0 mm Hg when administered over a 10-min aerosolization period. The efficacy of the PDE inhibitors was assessed in dose-response curves for rolipram, motapizone, and tolafentrine; these agents were either bolus injected into the recirculating buffer fluid or nebulized over a 10-min period. In separate experiments, a subthreshold dose of the PDE inhibitor, with no effect on pPA and gas exchange per se, was administered either i.v. or via the inhalative route, followed by subsequent PGI₂ inhalation. The following experimental groups were used:

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

Data Analysis. All values are given as means ± S.E. or as coefficient of variation (S.D./mean, %). For analysis of statistical difference, two-tailed Student's t test for unpaired samples was performed. After Bonferroni's correction, P < .05 was considered significant.

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

Baseline Conditions. After termination of the steady-state period, all lungs displayed pPA values between 5 and 7 mm Hg. Baseline A/ measurements revealed unimodal narrow distribution of perfusion and ventilation to midrange A/ (0.1 < A/ < 10) areas throughout (Table 1). Shunt flow (A/ < 0.005) and perfusion flow to poorly ventilated areas (0.005 < A/ < 0.1) were extremely low (<2.5%), and no perfusion to high A/ regions (10 < A/ < 100) was observed (not shown in detail). Dead space (A/ > 100) approximated 50% of ventilation in system isolated-lung ventilation this.

U46619-Elicited Pulmonary Hypertension and Gas Exchange Abnormalities. Continuous infusion of 130 ± 27 pmol · kg¹ · min¹ (all experiments) of U46619 provoked an increase in pPA to 32.5 ± 1.1 mm Hg within 25 min, with subsequent plateauing of pulmonary hypertension. The pPA rise was accompanied by a delayed-onset lung weight gain and a progressive increase in shunt flow to 58.1 ± 3.6% of total lung perfusion after 135 min (Table 1 and Fig. 1). Dead space increased by ~10% (data not shown). A marked broadening of flow dispersion and ventilation distribution in the midrange A/ regions was noted under these conditions. The lung weight progressively increased resulting in a total weight gain of 16.6 ± 1.7 g at the end of the experiments (Fig. 2).

View larger version (21K): [in this window] [in a new window]   Fig. 1.   Influence of PGI2 nebulization and its combination with subthreshold doses of PDE inhibitors on U46619-induced intrapulmonary shunt flow. The timing of the interventions is indicated. Means ± S.E. of six independent experiments are given. Shunt, percentage of perfusion on nonventilated areas (A/ < 0.005); Roli/Mota aer., aerosolization of 135 pmol · kg1 · min1 of rolipram and 131 pmol · kg1 · min1 of motapizone for 10 min before PGI2 nebulization; Roli/Mota i.v., intravascular application of rolipram (1 nM) and motapizone (1 nM) before PGI2 nebulization; * P < .05, ** P < .01, *** P < .001, compared with lungs with PGI2 inhalation alone.

View larger version (49K): [in this window] [in a new window]   Fig. 2.   Lung weight gain in response to U46619 infusion in the different experimental groups. Means ± S.E. of six independent experiments are given. Roli, rolipram; Mota, motapizone; Tola, tolafentrine. * P < .05, ** P < .01, compared with lungs with PGI2 inhalation alone. Control lungs received neither U46619 nor vasodilatory agent.

Dose-Response Curves of PDE Inhibitors. As shown in Fig. 3, all intravascularly administered PDE inhibitors effected a dose-dependent reduction in elevated pPA values in lungs with U46619-elicited pulmonary hypertension. Rolipram showed the highest efficacy (dose range, 10-10,000 nM), followed by tolafentrine (range, 0.02-200 µM) and motapizone (range, 0.01-100 µM). Combination of subthreshold doses of motapizone (100 nM) and rolipram (50 nM) resulted in a marked amplification of vasodilatory efficacy (Fig. 4). Even a 10-fold-lower dose of each agent reduced the elevated pressure by 9.2 ± 1.7% on coapplication. Inhalation of rolipram and motapizone was virtually ineffective up to a dose of 6.7 nmol · kg¹ · min¹ for each agent (Fig. 4). When combined, amplification of the pulmonary vasodilatory effect was again obvious, demonstrated even for 10-fold-lower doses of each PDE inhibitor (Fig. 4). Tolafentrine decreased the elevated pPA values in response to U46619 infusion by 23.8 ± 3.3% at the very high dose of 120 nmol · kg¹ · min¹ (data not shown in detail).

View larger version (18K): [in this window] [in a new window]   Fig. 3.   Dose-effect curves of intravascular PDE inhibitors on U46619-elicited pulmonary hypertension. The decrease in pPA in response to each PDE dose (concentrations related to the recirculating perfusate) is given (mean ± S.E.). ** P < .01; *** P < .001, compared with lungs with U46619-elicited pulmonary hypertension.

View larger version (30K): [in this window] [in a new window]   Fig. 4.   Influence of fixed doses of intravascular and inhaled PDE inhibitors and combinations of these agents on U46619-elicited pulmonary hypertension. The decrease in pPA in response to each intravascular PDE dose or combination (concentrations related to the recirculating perfusate) or nebulized PDE dose or combination (10-min aerosolization periods) is given (in percentage of the U46619-elicited pPA increase). Mean ± S.E. of six independent experiments. Roli, rolipram; Mota, motapizone; Tola, tolafentrine. * P < .05, ** P < .01, compared with U46619 lungs.

Nebulization of PGI₂. Inhalation of aerosolized PGI₂ at a dose of 27 pmol · kg¹ · min¹ for 10 min resulted in a significant reduction in U46619-induced pulmonary hypertension, with pPA values decreasing by a mean of 4 mm Hg (Fig. 5). The vasodilatory effect commenced within 2 min after onset of nebulization. Immediately after stopping the aerosol application, pPA started to rise again, and prenebulization values were reached within 15 min. The development of intrapulmonary shunt flow was moderately lowered in response to PGI₂ aerosolization but did not differ significantly from the U46619 group (maximum 48.7 ± 5.3% at the end of the experiments; Fig. 1). Broadening of flow dispersion and ventilation distribution in the midrange A/ regions was comparable to that in the U46619 group. The total weight gain of the lungs was 17.3 ± 2.9 g at the end of the experiments (Fig. 2).

View larger version (24K): [in this window] [in a new window]   Fig. 5.   Influence of PGI2 nebulization with and without prior intravascular administration of subthreshold doses of selective PDE inhibitors on U46619-elicited pulmonary hypertension. The timing of interventions is indicated; the following doses were used: 50 nM rolipram, 100 nM motapizone, and 1 nM/1 nM for the combination of rolipram and motapizone. Means ± S.E. of six independent experiments are given. Roli, rolipram; Mota, motapizone. * P < .05; ** P < .01, compared with lungs with PGI2 inhalation alone.

Combined Subthreshold Administration of PDE Inhibitors and PGI₂ Nebulization. The 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 mode of PDE inhibition on the pulmonary vascular tone per se. All subthreshold interventions for PDE inhibition did, however, clearly affect responsiveness to the subsequent standardized PGI₂ nebulization.

Rolipram. In the presence of i.v. or inhaled rolipram, the PGI₂-induced maximum pPA decrease was not augmented, but the pPA values did not fully return to the preceding U46619-elicited plateau within the subsequent 100-min observation period (Figs. 5 and 6). The increase in weight gain was significantly decreased by both routes of rolipram application (Fig. 2). There was some reduction in shunt flow in the i.v. rolipram/PGI₂ and the aerosol rolipram/PGI₂ lungs, but the difference to the lungs with PGI₂ nebulization alone was not significant (Table 1). This was also true for the broadening of the perfusion dispersion and ventilation distribution in the midrange A/ areas.

View larger version (23K): [in this window] [in a new window]   Fig. 6.   Influence of PGI2 nebulization with and without prior aerosol administration of subthreshold doses of selective PDE inhibitors on U46619-elicited pulmonary hypertension. The timing of interventions is indicated; the following doses were used: 0.9 nmol · kg1 · min1 rolipram, 6.5 nmol · kg1 · min1 motapizone, and 135 pmol · kg1 · min1 rolipram plus 131 pmol · kg1 · min1 motapizone. Means ± S.E. of six independent experiments are given. Roli, rolipram; Mota, motapizone. * P < .05 compared with lungs with PGI2 inhalation alone.

Motapizone. Subthreshold doses of inhaled motapizone, but not i.v. motapizone, significantly enhanced the maximum pPA decrease in response to the subsequent PGI₂ nebulization from 4 to 8 mm Hg. Both routes of motapizone administration moderately retarded the pPA return to pre-PGI₂ values. No significant changes in intrapulmonary shunt flow, gas exchange variables, or weight gain were noted in the lungs undergoing either intravascular or transbronchial motapizone application before PGI₂ nebulization, compared with the lungs with PGI₂ aerosolization alone.

Tolafentrine. Intravenous administration of the dual-selective PDE3/4 inhibitor tolafentrine significantly amplified (pPA 8 instead of 4 mm Hg) and prolonged (>90 versus 15 min) the vasodilatory efficacy of PGI₂ nebulization (Fig. 7). When applied via inhalation, the same trend of tolafentrine efficacy was noted but was much less obvious. Neither mode of tolafentrine administration significantly influenced lung weight gain and gas exchange variables.

View larger version (22K): [in this window] [in a new window]   Fig. 7.   Influence of PGI2 nebulization with and without prior intravascular or aerosol administration of subthreshold doses of tolafentrine on U46619-elicited pulmonary hypertension. The timing of interventions is indicated; the following doses were used: 65 nM of i.v. tolafentrine and 1.2 nmol · kg1 · min1 of aerosolized tolafentrine. Means ± S.E. of six independent experiments are given. Tola, tolafentrine. * P < .05 compared with lungs with PGI2 inhalation alone.

Combined Rolipram and Motapizone. Intravenous preapplication of subthreshold doses of rolipram plus motapizone markedly amplified the PGI₂-elicited pPA decrease (10 compared with 4 mm Hg for PGI₂ alone) (Fig. 5). Moreover, the post-PGI₂ vasorelaxation was prolonged to >60 min, compared with 10 to 15 min. Nearly the same profile was noted for preapplication of subthreshold doses of rolipram/motapizone via inhalation. In addition, both approaches reduced lung edema formation by 50% (Fig. 2). As shown in Fig. 1, intrapulmonary shunt flow was significantly decreased (maximum shunt, 19 and 28% in lungs pretreated with rolipram/motapizone via intravascular and intrabronchial routes, respectively, compared with 49% with PGI₂ alone). Most of this shunt reduction occurred in favor of higher percentages of flow being distributed to normal A/ regions, but an increase in perfusion of low A/ areas was also noted (Table 1). The broadening of flow dispersion and ventilation distribution in the midrange A/ regions was not significantly affected (Table 1).

Ventilation Pressures. Peak airway pressure during constant-volume ventilation did not significantly change in any of the experimental groups (data not given in detail).

Measurement of cAMP. The impact of aerosolized motapizone, aerosolized rolipram, and a combination of these PDE inhibitors on PGI₂-elicited cAMP liberation was investigated (Fig. 8). In control experiments with U46619 infusion, perfusate cAMP concentrations slowly increased to 4 to 5 pmol/ml. Inhalation of PGI₂ resulted in an increase in cAMP to maximum values of 9 pmol/ml. A further increase in cAMP was noted on coadministration of the PDE3 inhibitor motapizone and the PDE4 inhibitor rolipram (maximum perfusate levels, 13-14 pmol/ml). The most prominent elevation of perfusate cAMP levels occurred on combination of motapizone and rolipram with PGI₂ (maximum cAMP levels, 16 pmol/ml). These data match the efficacy of these agents and combinations in reducing U46619-elicited pulmonary hypertension.

View larger version (24K): [in this window] [in a new window]   Fig. 8.   Influence of PGI2 nebulization with and without prior aerosol administration of subthreshold doses of PDE inhibitors on cAMP levels in lung perfusate. The timing of interventions is indicated; the following doses were used: 0.9 nmol · kg1 · min1 of rolipram, 6.5 nmol · kg1 · min1 of motapizone, 135 pmol · kg1 · min1 of rolipram plus 131 pmol · kg1 · min1 of motapizone. Means ± S.E. of six independent experiments are given. Roli, rolipram; Mota, motapizone. * P < .05 compared with lungs with PGI2 inhalation alone.

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

Continuous infusion of the thromboxane A₂ mimetic U46619 has repeatedly been used to establish stable pulmonary hypertension in perfused rabbit lungs, suitable for testing the effects of pharmacological agents on pulmonary hemodynamics (Rimar and Gillis, 1993, 1995; Lindeborg et al., 1995; Walmrath et al., 1997). Similar to inhaled NO, aerosolized PGI₂ was previously shown to be an effective pulmonary vasodilator in this model, with substantial reduction in precapillary and moderate relief of postcapillary vascular resistance (Walmrath et al., 1997). When investigated in the absence of PGI₂, both intravascular and aerosol delivery of the PDE3 and PDE4 inhibitors motapizone and rolipram caused dose-dependent pulmonary vasodilation in the lungs with U46619-elicited hypertension. When calculated on a molar basis, higher efficacy and potency of rolipram over motapizone was observed for both routes of application. Pronounced synergism was obvious when combining PDE3 and PDE4 inhibition, again for both routes of drug delivery: coapplication of 1/10 of the subthreshold dose of each substance sufficed to cause a significant decrease in pPA not anticipated from the dose-response curves of either motapizone or rolipram. This is in line with studies in PGF₂-constricted human pulmonary artery rings, in which the combination of motapizone and rolipram was found to be effective in inducing relaxation (Rabe et al., 1994), and with results from intact rabbits with pulmonary hypertension, also demonstrating high vasodilatory efficacy of i.v. coapplication of motapizone and rolipram (D.W., R.T.S., and W.S., in preparation). In smooth muscle cells originating from human airways, the cAMP content was essentially regulated by the activities of PDE3/4 pathways, and a corresponding profile may hold true for pulmonary vascular smooth muscle cells (Pan et al., 1994; Torphy, 1998). Inhibition of either PDE3 or PDE4 might partly be compensated for by enhanced breakdown of cAMP through the alternate catabolic pathway, whereas dual inhibition of both PDE3 and PDE4 is fully effective in elevating cAMP levels. Correspondingly, the dual-selective PDE3/4 inhibitor tolafentrine exerted strong pulmonary vasodilation in the U46619-constricted pulmonary vasculature; however, when calculated on a molar basis, this agent was less efficient than the combination of motapizone and rolipram by >2 orders of magnitude.

Most interesting, a marked synergistic amplification of the pulmonary vasodilatory effect of nebulized PGI₂ was noted when this approach was combined with preapplication of subthreshold doses of the PDE inhibitors, which per se caused no hemodynamic effects. The pPA decrease in response to the PGI₂ aerosol, leveling off within 10 to 15 min after termination of nebulization because of the short half-life of this prostanoid (Moncada and Vane, 1980), was prolonged in the presence of subthreshold doses of all PDE inhibitors investigated. This was most prominent for the combination of motapizone and rolipram, whether admixed to the perfusion fluid or administered via the inhalative route in a preceding aerosolization maneuver: the maximum pPA decrease in response to PGI₂ nebulization more than doubled, and sustained pulmonary vasodilation for >60 min was noted. Similar efficacy was noted for subthreshold doses of the dual-selective inhibitor tolafentrine when administered intravascularly, with somewhat lower efficacy of this agent via inhalation. These data correspond with previous studies in isolated pulmonary artery rings from rats with hypoxia-induced pulmonary hypertension, in which an enhancement of isoproterenol and forskolin-induced relaxation was noted in the presence of PDE3 and PDE4 inhibitors (Wagner et al., 1997). Again, stabilization of PGI₂-induced cAMP resulting from inhibition of both PDE pathways offers the most probable explanation for the high efficacy of combined PDE3/4 inhibition. This view is well supported by the current measurements of perfusate cAMP levels, being highest on coapplication of PDE3 and PDE4 inhibitors with PGI₂.

For a detailed analysis of gas exchange, the multiple inert gas elimination technique was used, the feasibility and validity of which has been documented in isolated, blood-free perfused rabbit lungs (Walmrath et al., 1997). Under baseline conditions, physiological A/ matching was noted. In sharp contrast, the U46619-elicited pulmonary hypertension was accompanied by gas exchange abnormalities, of which the predominant abnormality was a dramatic increase in shunt flow to >50%. Such a marked shunt flow might be largely attributable to the progressive formation of lung edema in response to ongoing U46619 infusion; however, several observations challenge such a simple view. First, the total amount of edema formation was still moderate (10-15 g when shunt already approached 50% compared with >50 g under conditions of frank alveolar edema filling of the rabbit lungs). Second, shunt flow commenced before onset of significant lung weight gain, and it is known to be partly reversible on offset of U46619 infusion despite further progression of edema formation (Walmrath et al., 1997). Third, all "therapeutic" rolipram groups displayed an ~50% decrease in edema formation in this study. However, only the simultaneous application of rolipram and motapizone reduced the shunt flow to substantially lower values, with no difference in lung weight gain between the rolipram-alone and the rolipram plus motapizone groups (see below). Thus, as previously suggested (Walmrath et al., 1997), the coappearance of edema formation and marked pPA elevation in response to the vasoconstrictor agent U46619 is suggested as the predominant mechanism underlying the excessive shunt flow.

Pressure-driven redistribution of perfusion to edema-filled regions or otherwise nonventilated areas must be assumed and is spared from perfusion by hypoxic vasoconstriction at normal pPA. An alternative explanation would be recruitment of very short term constant pathways or pathways shunting exchange areas, assumed to contribute to lung perfusion under conditions of pulmonary hypertension.

The gas exchange abnormalities with high shunt flow in this model of acute pulmonary hypertension are similar to those encountered under clinical conditions of ARDS. Inhaled PGI₂ has been shown to decrease pulmonary hypertension in patients with ARDS without affecting systemic pressure (Walmrath et al., 1993). Interestingly, transbronchial and intravascular coapplication of PDE inhibitors to enhance the vasodilatory effect of PGI₂ was not accompanied by a deterioration in gas exchange, as repeatedly demonstrated for i.v. use of vasodilators in diseased lungs (Mélot et al., 1989; Radermacher et al., 1990). In contrast, A/ matching improved, again most obviously for rolipram plus motapizone. In the presence of this combination of PDE3/4 inhibitors, the shunt flow was reduced by 50% compared with the PGI₂-only group. This impact on A/ matching translates into a marked improvement in arterial oxygenation when occurring in vivo.

Two underlying mechanisms may explain the impressive beneficial effect of rolipram plus motapizone on the U46619-elicited gas exchange abnormalities. First, a reduction in lung edema formation was obvious for all lungs receiving rolipram plus aerosolized PGI₂ and was not reproduced by motapizone or tolafentrine. In addition to its effect via reducing the capillary filtration pressure, some direct effect of rolipram on microvascular permeability may be operative, as suggested from in vitro studies (Suttorp et al., 1993; Barnard et al., 1994; Miotla et al., 1998). However, the impact of rolipram on lung fluid balance may well contribute to, but may not solely explain, the improvement in gas exchange in the lungs with rolipram plus motapizone, because the total weight gain in this group did not significantly differ from the rolipram-only group. Second, A/ is improved. Via stabilization of the second-messenger cAMP, subthreshold doses of rolipram plus motapizone, causing no hemodynamic effects per se, may amplify PGI₂-induced pulmonary vasodilation, whereas this effect is still restricted to the aerosol-accessible (i.e., well ventilated) lung areas with redistribution of blood flow from the nonventilated (shunt) regions. Notably, this profile of enhanced pulmonary vasorelaxation and improvement in gas exchange was true for both aerosol and intravascular administration of subthreshold doses of rolipram plus motapizone. Moreover, the use of subthreshold doses in this study is of interest, because under clinical conditions, higher dosages of PDE3 inhibitors are hampered by several cardiovascular side effects, including life-threatening arrythmias in patients with compromised cardiac function (Naccarelli and Goldstein, 1989), which may limit the use of these agents in patients with pulmonary hypertension. PDE4 inhibitors, especially rolipram, may induce emesis, gastric secretion, and psychotropic effects in the normal dose range (for overview, see Torphy, 1998).

In conclusion, our study demonstrates that the pulmonary vasodilatory effect of nebulized PGI₂ may be significantly enhanced and prolonged by coapplication of subthreshold doses of PDE3 and PDE4 inhibitors, which exert no hemodynamic effects per se. Both the intravascular and the transbronchial routes of PDE inhibitor administration may be used for this purpose. Note that, in contrast to the common finding of deterioration of gas exchange when using i.v. vasodilators in diseased lungs, shunt flow did not increase. However, the most effective approach for PDE3/4 inhibition even significantly decreased the perfusion of shunt areas and redistributed flow to well ventilated lung regions. The profile of nebulized PGI₂relief of pulmonary hypertension and improvement of gas exchange due to targeting the vasorelaxant properties to well ventilated regions of the lungis thus further intensified by combination with suitable subthreshold PDE inhibition for prolonged half-life of the second-messenger cAMP.