In this chapter, we systematically examine the evidence linking acute respiratory distress syndrome (ARDS) with pulmonary hypertension, the implications of pulmonary hypertension and consequent right ventricular failure for patient outcomes, and the data related to pulmonary vasodilator therapies in this patient group.
Pulmonary Hypertension in ARDS
Pulmonary hypertension in ARDS was described in the late 1970s and was soon accepted as a key cause of death. A consistent observation in reports at the time was that nonsurvivors of ARDS demonstrated pulmonary artery pressures that continued to increase throughout the early phase of the illness. Later, systematic studies, such as the European Collaborative ARDS study, confirmed the prognostic significance of pulmonary artery pressures for these patients. In that report, a logistic regression analysis that included multiple hemodynamic measures and other factors identified day 2 systolic pulmonary artery pressure (24.1 ± 6.7 mm Hg for eventual survivors compared with 28.4 ± 8.5 mm Hg for eventual nonsurvivors) as a potent independent predictor of mortality. More recently, a secondary analysis of results from the Fluid and Catheter Treatment Trial found that pulmonary vascular resistance was elevated early in the course of ARDS and was statistically higher in patients who died. In multivariate prediction models, pulmonary vascular resistance was a strong independent risk factor for 60-day mortality.
How common is pulmonary hypertension in ARDS? There are surprisingly few data to accurately answer this question. Zapol and Snider found that all of the 30 ARDS patients in their series had elevated pulmonary artery pressures, even after correction of hypoxemia. Clinical trials in ARDS have consistently reported baseline mean pulmonary pressures of 29 to 30 mm Hg. More recently, with a cutoff mean pulmonary artery pressure of 25 mm Hg, 92.2% of ARDS patients had pulmonary hypertension, although this was severe (defined by a mean pulmonary artery pressure of >45 mm Hg) in only 7.4%.
A combination of factors may contribute to the development of pulmonary hypertension in patients with ARDS. Correlations between lung edema and pulmonary artery pressures have been demonstrated. Intravascular thrombosis causing microvascular occlusion was an important factor in pulmonary vascular resistance in a pig model, and postmortem studies have demonstrated widespread pulmonary thromboembolism in 95% of cases of ARDS. Although marked hypoxic pulmonary vasoconstriction has been demonstrated in nonventilated areas of the lung in patients with ARDS, the effect of this phenomenon on overall pulmonary hemodynamic measures is uncertain. For example, Sibbald and colleagues reported that the severity of pulmonary hypertension occurring in ARDS correlated poorly with the degree of hypoxia. Hypoxic pulmonary vasoconstriction may be a weak contributor because it is partially or wholly inhibited by factors such as locally released nitric oxide (NO) or prostaglandin. Furthermore, pulmonary hypertension in ARDS may persist, even after the resolution of hypoxemia. One possible explanation is that pulmonary vascular smooth muscle cells proliferate over time. This results in a diminution in wall compliance.
Inflammatory mediators released in sepsis may increase vascular tone in the pulmonary circulation while decreasing it in the systemic circulation. Cytokines such as tumor necrosis factor-α have been implicated, but their exact role is unclear. Endothelin-1 (ET-1) is a potent pulmonary vasoconstrictor and activator of vascular smooth muscle proliferation. ET-1 expression is upregulated in patients with ARDS, although there is currently no evidence directly implicating ET-1 in ARDS-related pulmonary hypertension.
Pulmonary Hypertension, Right Heart Failure, and Death
The thin-walled right ventricle is accustomed to pumping into a low-pressure circuit; therefore it responds poorly to increases in afterload. In the critically ill patient, multiple factors, such as fluid overload, negative inotropy associated with sepsis, and elevated mean airway pressures, may impair right ventricular function. This is supported by data indicating that right ventricular failure both predicts and appears to cause the death of 30% of patients with ARDS. In an echocardiography-based study that evaluated the right side of the heart in 23 patients with ARDS, 9 patients were found to have normal right ventricular function, whereas 9 other patients had a slightly enlarged right ventricle with normal systolic function. The remaining five patients had a severely enlarged right ventricle with contractile dysfunction and reductions in left ventricular size. These findings suggest detrimental ventricular interdependence. Of note, all of the patients in that study had normal left ventricular systolic function by two-dimensional echocardiography. Severe right ventricular failure was strongly associated with death.
Vieillard-Baron and associates used echocardiography to evaluate the right side of the heart in ARDS. Right ventricular dysfunction was present in 19 (25%) of 75 patients on day 2. Many of these patients also had evidence of left ventricular diastolic dysfunction. Although mortality was the same as that for patients without right ventricular dysfunction, duration of respiratory support was longer. Of particular interest in this study, elevated partial pressure of carbon dioxide in arterial blood (Pa co 2 ) was identified as the sole independent predictor of acute right ventricular failure. This may reflect increased dead space, associated with high levels of positive end-expiratory pressure (PEEP) and worse outcomes with ARDS. For example, Poelaert and coworkers found incremental PEEP-induced cyclic augmentation of right ventricular outflow impedance. Jardin and Vieillard-Baron illustrated that higher plateau pressures were associated with marked increases in acute right heart failure and death. These authors recently proposed the concept of a “right ventricular protective approach” to mechanical ventilation for the patient with ARDS.
Pulmonary Vasodilator Therapies in ARDS
Inhaled Nitric Oxide
NO is a free radical gas that was identified in 1987 as the elusive endothelium-derived relaxing factor. After native generation in the endothelium, NO enters local vascular smooth cells where it activates soluble guanylate cyclase. This enzyme stimulates the conversion of guanosine 5′-triphosphate to cyclic guanosine monophosphate (cGMP), which causes hyperpolarization and attenuates calcium entry to the muscle cytoplasm. The net result is vasodilation. Deficiencies in NO production and attenuated responsiveness to NO in the pulmonary circulation have been identified and are now accepted as important factors in the pathogenesis of primary and secondary pulmonary hypertension.
Within a year of the discovery of NO, inhaled NO was confirmed as an effective pulmonary vasodilator in patients with primary pulmonary hypertension. Shortly thereafter, several small case series describing the use of NO therapy for patients with ARDS and pulmonary hypertension appeared in the literature. Studies reported not only decreases in pulmonary vascular resistance and pulmonary artery pressures but also significant improvements in oxygenation. For example, in 1993, Rossaint and colleagues gave 18 ppm inhaled NO to 10 patients with ARDS. Pulmonary artery pressures decreased by an average of 6 mm Hg, and pulmonary vascular resistance decreased by an average of 71 dyn sec cm -5 from baseline. There were no significant changes in systemic blood pressure or cardiac output. However, most compelling to clinicians at the time was an average increase in the partial pressure of oxygen in arterial blood (Pa o 2 )/fraction of inspired oxygen (F io 2 ) of 51 mm Hg. Inhaled NO was rapidly adopted for the treatment of severe ARDS. Indeed, a survey of intensive care physicians’ practices across Europe showed that, by 1998, 98.5% of respondents considered ARDS an indication for inhaled NO. Moreover, 71% considered that Pa o 2 /F io 2 ratios were sufficient criteria for initiating treatment.
In these early studies, three observations were made that would later become contentious. The first of these was that the response to NO, whether based on decreased pulmonary artery pressures or on improved oxygenation, was largely predictable and almost universal. It was later established that, at most, 40% to 60% of ARDS patients responded to inhaled NO with an improvement in one or both of these parameters. Prediction of likely responders was difficult. The second observation was that the response to inhaled NO was sustained over a prolonged period of treatment. In contrast, later data demonstrated the development of tachyphylaxis within 2 to 3 days. The final observation was that although daily interruptions of inhaled NO were noted to cause increases in pulmonary artery pressures, these changes were not thought to be particularly problematic. Rebound pulmonary hypertension after withdrawal was later appreciated as a phenomenon of real consequence, albeit one that could be overcome.
Early enthusiasm for inhaled NO was curtailed by negative Phase II and Phase III trials showing that NO did not improve overall survival in ARDS. This was supported by meta-analyses of trials of NO therapy in ARDS. In addition, NO may have an adverse effect on renal function In the United States, concerns about clinical efficacy of NO have been reinforced by the high costs associated with the delivery system. The results of a Canadian survey were likely representative of worldwide practice: By 2004, less than 40% of critical care physicians were using NO as therapy in ARDS and then only selectively.
Prostaglandins
Prostaglandins are vasodilators that act through intracellular adenylate cyclase, leading to a decrease in intracellular calcium. Various prostaglandins and their analogs have been shown to improve exercise capacity and quality of life in chronic pulmonary hypertension but with little effect on mortality.
During the late 1980s, several reports described the use of intravenous prostaglandin E 1 (PGE 1 ) for ARDS. PGE 1 appeared to exert its effects but as an anti-inflammatory and as a pulmonary vasodilator. The finding that pulmonary artery pressures were indeed decreased—by approximately 15% when given in the typical dose range —prompted two randomized controlled trials. The first, and the smaller of the two, was limited to ARDS in surgical patients and suggested a survival advantage. The subsequent, larger, and more inclusive trial failed to confirm this. Indeed, the authors reported systemic hypotension and increases in intrapulmonary shunting. As these results were emerging, reported successes with inhaled NO fueled attempts to find an inhaled prostaglandin. Iloprost, a synthetic analog of prostacyclin, emerged as a drug stable in aerosolization and suitable for inhalation. In 1993, Walmrath and coworkers first reported the use of aerosolized iloprost in three patients with ARDS. Pulmonary vascular resistance and intrapulmonary shunt decreased and oxygenation improved, all by 30% to 40%. These findings were confirmed 3 years later by 2 reports, both involving rather few patients. In a more recent report, 10 mg nebulized iloprost was administered to a series of 20 ARDS patients with pulmonary hypertension identified by echocardiography. Pa o 2 increased by a mean of 18 mm Hg without demonstrable adverse effects.
Iloprost compares well with inhaled NO for the treatment of pulmonary hypertension in ARDS. Prostacyclin and its analogs have a longer half-life (2 to 3 minutes) compared with NO (seconds). Although this could increase the risk for systemic vasodilation and hypotension, in practice this does not appear to be a significant problem. Indeed, 50 ng/kg/min, the upper end of the dose range for iloprost, caused no systemic hemodynamic effects in children with acute lung injury. Prostacyclin is also a potent inhibitor of platelet aggregation. In the absence of increased bleeding, this may be of benefit.
Nonetheless, comparisons of iloprost and NO are complicated by the limited published data. There are several small studies. Van Heerden and colleagues showed drug equivalency for iloprost at 50 ng/kg/min and NO at 10 ppm in five hypoxemic ARDS patients. Zwissler and associates compared 1, 10, and 25 ng/kg/min iloprost with NO at 1, 4, and 8 ppm, respectively, and found that both drugs produced roughly comparable effects. This also established limited dose-response curves for ARDS patients. Likewise, in 16 ARDS patients, Walmrath and coworkers found that iloprost (average dose 7.5 ± 2.5 ng/kg/min) and inhaled NO (average dose 18 ppm) were equally effective. Finally, similar comparative studies in primary pulmonary hypertension point to roughly comparable clinical effects of the two agents.
NO is degraded to nitrogen dioxide, a potential toxin. NO also requires an expensive delivery and monitoring system. Iloprost does not have this problem because it can be delivered by simple nebulizer systems. However, as with inhaled NO, rebound hypertension on drug withdrawal has been reported. What remains to be conclusively demonstrated is whether prostaglandins may succeed in NO-unresponsive patients and vice versa. Because NO and prostaglandins exert their effects by entirely different mechanisms, the hypothesis is an attractive one, but which patients will respond to either remains difficult to predict. Data from Domenighetti and colleagues suggest that patients with ARDS of pulmonary origin are less likely to respond than those with ARDS of extrapulmonary origin, but a direct comparison with NO was not performed. Of note, Brett and associates found no predictors of response to inhaled NO. Finally, as with NO, there are no data suggesting that inhaled prostacyclin alters outcome in ARDS.
Phosphodiesterase Inhibitors
Enoximone, amrinone, and milrinone are inhibitors of phosphodiesterase type 3 PDE-3, the enzyme that catalyzes the breakdown of cyclic adenosine monophosphate in myocardium and vascular smooth muscle. Inhibition of this enzyme increases myocardial contractility and causes widespread vasodilation. Although long-term survival rates are not improved for patients with chronic cardiac failure taking oral milrinone, this class of drugs is widely used in the setting of acute cardiac failure in cardiac surgical patients. Decreases in output impedance should particularly favor the failing right ventricle. In a retrospective comparison of milrinone and dobutamine in 329 patients with acutely decompensated cardiac failure, milrinone produced greater decreases in pulmonary vascular resistance with greater improvements in cardiac output. Likewise, in patients with severe pulmonary hypertension undergoing transplantation, milrinone or enoximone potently decreased pulmonary vascular resistance and increased cardiac index.
Sildenafil is an orally administered, highly selective inhibitor of PDE-5. This subtype of PDE is present in abundance in the smooth muscle cells of the pulmonary vasculature. Inhibition of PDE-5 prevents the breakdown of cGMP, thereby augmenting the vasodilating effects of native and inhaled NO.
There are some reports of sildenafil treatment for patients with new-onset, life-threatening pulmonary hypertension related to acute lung injury or ARDS. Giacomini and associates gave enteral vardenafil, a sildenafil analog, to a single patient with ARDS and pulmonary hypertension in whom weaning of inhaled NO had proved impossible. Vardenafil permitted withdrawal of the inhaled NO and was itself eventually tapered. However, a recent open-label study evaluated the effect of a single 50-mg dose of sildenafil in 10 patients with ARDS and pulmonary hypertension. Although pulmonary artery pressures decreased significantly, from means of 25 to 22 mm Hg, systemic arterial blood pressures also decreased, whereas shunt fraction increased. In the absence of further evidence, sildenafil remains an unproven therapy for pulmonary hypertension in ARDS.
Levosimendan
Levosimendan is an inodilator. The inotropic effect occurs through sensitization of troponin C in the myocardium. Contractility is improved, but this uniquely occurs without a concomitant increase in intracellular calcium or in energy consumption. Vasodilation occurs through activation of potassium–adenosine triphosphate channels in the vasculature. Activation of these channels may also account for the cardioprotective effect reported in laboratory and clinical studies. An immunomodulatory effect is also described, although the mechanism is unknown.
The LIDO (Levosimendan Infusion versus Dobutamine) study was a double-blind randomized controlled trial that compared levosimendan with dobutamine in cardiogenic shock. Not only were predetermined hemodynamic goals achieved more successfully with levosimendan, but there was also a significant survival benefit. The extreme sensitivity of the right ventricle to modest changes in afterload suggests a particular potential for levosimendan in the treatment of right ventricular failure complicating pulmonary hypertension.
There are several clinical studies describing the use of levosimendan specifically for pulmonary hypertension and right ventricular failure. In a small placebo-controlled trial, Ukkonen and associates reported marked decreases in pulmonary vascular resistance along with improvements in right ventricular mechanical efficiency and cardiac output in patients with severe right heart failure. Morelli and colleagues performed a randomized placebo-controlled trial in 35 patients with ARDS. Levosimendan decreased mean pulmonary artery pressures from 29 ± 3 to 25 ± 3 mm Hg while increasing the right ventricular ejection fraction from 45 ± 10 to 59 ± 10%. Cardiac index and mixed venous oxygen saturations also significantly increased.
Related posts:
What Lessons Have Intensivists Learned During the Evidence-Based Medicine Era?
What Is the Role of Invasive Hemodynamic Monitoring in Critical Care?
Should Fever Be Treated?
Are Computerized Algorithms Useful in Managing the Critically Ill Patient?
Are Anti-inflammatory Therapies in ARDS Effective?
Has Outcome in Sepsis Improved? What Has Worked? What Has Not Worked?
Full access? Get Clinical Tree
