The trusted source for
healthcare information and
Severe Pulmonary Hypertension in the ICU
By Richard J. Wall, MD, MPH, Pulmonary Critical Care & Sleep Disorders Medicine, Southlake Clinic, Valley Medical Center, Renton, WA, is Associate Editor for Critical Care Alert.
Dr. Wall reports no financial relationship to this field of study.
Pulmonary hypertension (ph) is common among patients in the intensive care unit (ICU). In fact, many ICU clinicians simply view PH the same way they view leukocytosis — as an expected finding caused by the "bigger problems" of sepsis, respiratory failure, congestive heart failure, volume overload, and myocardial infarction. In severe cases of PH, however, the hemodynamic consequences of the PH itself can be devastating: hypotension, low cardiac output, hypoxemia, cor pulmonale, liver congestion, renal failure, and death.
In the current review, the general approach to managing ICU patients with severe life-threatening PH will be addressed. By definition, this refers to critically ill patients with (or at risk of) hemodynamic instability. After briefly discussing the physiology and diagnosis of PH, this article will review current treatments.
Definition of PH
PH is defined as right ventricular (RV) systolic pressure > 40 mm Hg, or mean resting pulmonary arterial (PA) pressure > 25 mm Hg (> 30 mm Hg with exercise).1 Additional criteria include pulmonary vascular resistance (PVR) > 3 Wood Units and pulmonary capillary wedge pressure (PCWP) ≤ 15 mm Hg. In ICU patients, the diagnosis is commonly suspected based upon typical signs and symptoms or made with echocardiography. With the growing availability of bedside echocardiography, ICU clinicians are increasingly diagnosing this entity. In many cases, the diagnosis is already established prior to ICU admission.
In healthy individuals, PA pressure is ~20% of the systemic arterial system. In fact, the pulmonary vascular system has an incredible ability to handle changes in flow and pressure through its vasodilator reserve. When these compensatory mechanisms are overwhelmed, however, the pulmonary vasculature's response can result in a cascade of events that leads to irreversible cardiogenic shock and multi-organ failure.
PA systolic pressure is determined by RV stroke volume and the compliance of the main PA and its branches. PA diastolic pressure is determined by the tone of the pulmonary arterioles, the size of the pulmonary vascular bed, and PCWP. The latter is determined by the pulmonary venous pressure, left atrial pressure, mitral valve integrity, and left ventricular diastolic pressure.
In 2003, an international symposium outlined a 5-category classification system for PH (see Table).2 In this system, patients are grouped based on the underlying pathophysiology of their condition. Although one should ideally determine the classification of every patient with PH, this is often impossible when a patient with severe PH is intubated and on vasopressors with multi-organ failure. Those interested in learning more about the pathogenesis of PH are referred to Gaine's excellent review.3
When PH is diagnosed, it is essential that the PH is adequately characterized through the measurement of mixed venous oxygen saturations, cardiac output, and assessment of RV function. The latter is especially important because the ultimate determinant of hemodynamic stability is RV function. In ICU patients, the etiology of PH is often multifactorial. Common contributors include pulmonary emboli, hypoxemia, anemia, acidosis, sepsis, and left-sided heart failure. Many ICU patients have mild-to-moderate underlying PH prior to the acute event that lands them in the ICU.
An ICU patient with acute, severe PH may be mistakenly thought to have "septic shock" because such patients have similar findings of hypotension, lactic acidosis, and renal failure. Furthermore, infections can precipitate an abrupt deterioration of underlying chronic PH. Other common causes for rapid deterioration of underlying PH include pulmonary embolism, pneumothorax, RV infarction, gastrointestinal bleeding, pancreatitis, anemia, thyroid disease, atrial arrhythmias, ischemic bowel, hyponatremia, hypokalemia, subdural hematoma, and acute renal failure.4 Thus, clinicians must maintain an index of suspicion for PH even when an alternative diagnosis seems plausible.
Initial workup should include physical examination, chest radiograph, and an electrocardiogram (ECG). The ECG should be examined for RV or left ventricular (LV) hypertrophy, atrial arrhythmias, and ischemia. This may provide important clues to the etiology. Keep in mind that automated computerized interpretations on ECG machines often erroneously label RV hypertrophy/strain as "anterior ischemia" or "inferior MI."5
If there are no signs of left-sided failure in a patient with elevated jugular venous pressure, hypoxemia, and clear lungs, then pulmonary embolism (PE) must be excluded. The differential diagnosis for such a patient also includes cardiac tamponade, RV infarction, pericardial constriction, and acute worsening of chronic PH. Because hypoxemia is common in PH regardless of the etiology, it is not useful in the diagnosis of PE.
Although multiple strategies exist for excluding PE,6 many of these traditional approaches are impractical in ICU patients. For example, D-dimer has very limited utility in the ICU due to the high prevalence of comorbidities that cause false-positives. In the hemodynamically unstable patient, bedside echo-doppler is a useful first test. In more stable patients, computer tomographic angiography (CTA), ventilation-perfusion scanning, or conventional angiography are reasonable options. Although multidetector CTA has become the most popular first-line test, there are surprisingly few prospective data on its predictive value. In fact, a study of ICU patients found that 25% of CTA examinations were non-diagnostic, usually due to poor contrast bolus or artifact from motion and hardware.7 Intensivists should also remember that critically ill patients were excluded from the recent PIOPED II study.
Treatment of PH
In patients with suspected PE, heparin therapy should be started promptly. Recent meta-analyses suggest low-molecular-weight heparin (LMWH) is preferable to unfractionated heparin because it has lesser risk of recurrence or bleeding, but these studies did not include patients with hemodynamically unstable PE.8,9 Oral anticoagulation with warfarin should be continued for at least 6 months after the event, and longer if there is a non-reversible risk factor.
In hemodynamically unstable ICU patients with RV failure and no underlying cardiopulmonary disease, empiric thrombolytic therapy may be given without confirming the diagnosis of PE.4 Surgical embolectomy and catheter-based interventions are other options, but such modalities have not been rigorously studied in large randomized trials. Further, no trials have compared thrombolytics against surgical or catheter-based approaches. In general, surgical and catheter-based interventions should only be performed in experienced centers because the reported favorable outcomes likely have as much to do with treatment protocols at these sites as with the procedures themselves.10
Hypotension in severe PH should be managed aggressively, promptly, and intelligently. In general, hypotension is rarely due to intravascular volume depletion. Rather, hypotension is usually due to deteriorating RV failure. Determining volume status in such patients is notoriously difficult. Although intravascular volume can be low in PH patients with edema and ascites, administering fluids to patient with RV pressure overload will further dilate and compromise RV function, thereby decreasing LV stroke volume and causing downstream systemic hypotension. The imprudent administration of fluids can potentially result in a fatal cardiovascular spiral. When volume status is not clear, carefully administer a small volume challenge (250 mL saline over 10 minutes), and assess the effect on systemic pressure.
Hypotension must be reversed quickly or the "PH spiral" will overwhelm your efforts. If too much time elapses, chances are slim that you will reverse the vicious sequence of events: Systemic hypotension causes low oxygen delivery, hypoxemia and decreased blood pressure compromise coronary blood flow, ventricular ischemia causes worsening pump failure, hypoxemia and acidosis exacerbate pulmonary vasoconstriction, worsening RV function leads to decreased LV filling, and so on. Eventually, the patient will deteriorate into electromechanical dissociation or ventricular fibrillation.
In patients unable to maintain a mean systemic arterial pressure of at least 60 mm Hg, a-1 adrenergic agents (e.g., phenylephrine, norepinephrine, high-dose dopamine) should be used. The a-1 receptor increases systemic pressure, coronary perfusion, systemic resistance, and LV afterload. At the same time, it reduces RV compression of the LV outflow tract, thereby improving LV stroke volume. In patients with a decent systemic arterial pressure, inotropes such as dobutamine and milrinone can improve cardiac output. Once systemic pressure is reasonable, IV epoprostenol can be used to further increase cardiac output and reduce pulmonary pressures.
Some have successfully used inhaled nitric oxide (iNO) as a selective pulmonary vasodilator to reduce pulmonary artery pressure, improve cardiac output, and reverse hypotension.4 It should be noted, however, that this indication has not been approved by the FDA. Others have used inhaled epoprostenol in a similar manner. For the hemodynamically unstable patient in the ICU, iNO has the advantage of increasing pulmonary perfusion only in ventilated areas. As a result, iNO improves gas exchange and decreases PVR without increasing intrapulmonary shunting, as might occur when IV epoprostenol vasodilates unventilated lung segments. Another advantage is that inhaled NO is rapidly inactivated in the alveolar capillaries, thereby minimizing systemic hypotensive effects. Patients stabilized with iNO can be later converted to IV epoprostenol for long-term use.
A prompt infectious workup with cultures and broad-spectrum antibiotics should be performed. Oxygen should be titrated to keep saturations above 92%. Mechanical ventilation (either non-invasive or via intubation) may be necessary to maintain adequate oxygenation, especially if the patient has pneumonia. As mentioned previously, hypoxemia causes pulmonary vasoconstriction and worsens ventricular function, which is the last thing you want. Blood transfusions can be used to improve oxygen delivery: [(Hgb × 1.34 × SaO2%) × cardiac output]. However, the potential risks of immune suppression and volume overload must be carefully considered.
Atrial fibrillation and atrial flutter are common tachyarrhythmias that lead to rapid clinical deterioration. Treatment should follow typical ACLS algorithms, with DC cardioversion of unstable patients. Avoid IV diltiazem or adenosine in hemodynamically unstable patients. Short-acting beta blockers, IV digoxin, and amiodarone are useful agents. The latter is especially useful for maintaining sinus rhythm. Ventricular arrhythmias are usually an ominous sign that things have gone too far. Cardiopulmonary resuscitation is almost never successful in PH patients with severe right heart failure.11
Other adjunctive vasodilator agents are available, but their use in the ICU is largely anecdotal. Sildenafil, a potent phosphodiesterase-5 inhibitor, has been shown to augment and prolong the hemodynamic effects of iNO. Although maximal effects are not seen for 3-6 months, some effects are noted within hours. In addition, sildenafil prolongs the effect of NO by preventing breakdown of its second messenger (cGMP), and thus may lessen the rebound seen after discontinuation of iNO. Endothelin antagonists (e.g., bosentan, sitaxsentan) do not exhibit any effects for at least 2 weeks and have no role in acute ICU management of PH. If a patient is already on an endothelin antagonist, it is reasonable to continue it. However, one must remain aware of the risk of liver toxicity and various drug interactions with this class.
When a critically ill patient experiences acute and hemodynamically unstable PH, there is a limited amount of time to intervene. Otherwise, the patient will irreversibly spiral into biventricular failure, followed by systemic shock and death. Even among patients with chronic PH, acute PE must be excluded as the precipitant for deterioration. Prompt empiric treatment with heparin or thrombolytics should be considered in such cases, even if the diagnostic workup is still underway. Fluids should be cautiously administered, and an adequate assessment of RV function is essential at the onset. Several different medications are now available to the ICU clinician for managing this serious condition.