Lamia B, Maizel J, Ochagavia A, Chemla D, Osman D, Richard C, Teboul JL. Echocardiographic diagnosis of pulmonary artery occlusion pressure elevation during weaning from mechanical ventilation. Crit Care Med. 2009 May;37(5):1696-701.
Gokay Güngör, M.D. and Franco Laghi, M.D.
Section of Pulmonary and Critical Care Medicine
Loyola University Stritch School of Medicine
Hines VA Hospital, Hines, IL
Cardiogenic pulmonary edema induced by weaning can cause/contribute to failure to wean from mechanical ventilation. The mechanisms by which weaning-induced pulmonary edema develops are complex and include:
- Reduction of intrathoracic pressure with resultant increases in systemic venous return and increases in left ventricular afterload.
- Increase in work of breathing with resultant increase in cardiac work and myocardial oxygen demand.
- Catecholamine discharge with resultant increase in venous return, left ventricular afterload, cardiac work, and myocardial oxygen demand.
- In patients with pre-existing right ventricular disease, weaning-induced increase in right ventricular afterload may occur as a result of hypoxemia or worsening of intrinsic positive end-expiratory pressure. Increased right ventricular afterload together with the simultaneous increase in systemic venous return may lead to a marked right ventricular enlargement. The latter could impede diastolic filling of the left ventricle (i.e., biventricular interdependence).
The clinical presentation of weaning-induced cardiogenic pulmonary edema/cardiac dysfunction is often nonspecific. It follows that pulmonary artery catheterization and the response of pulmonary artery occlusion pressure (PAOP) to a spontaneous breathing trial may be required to correctly identify patients who experience weaning-induced cardiogenic pulmonary edema/cardiac dysfunction.
Doppler echocardiography is a non-invasive technique to identify left ventricular filling pressure elevation.
To assess whether transthoracic echocardiography (TTE) could be used to reliably diagnose left ventricular filling pressure elevation induced by weaning from mechanical ventilation.
39 patients who had already failed two consecutive spontaneous breathing trials were enrolled in the study. All patients had a pulmonary artery caterer in place. Weaning consisted of a trial of spontaneous respiration lasting 1 hour (or less in those patients who developed distress during the trial). The investigators recorded pulmonary and systemic physiologic data and , concurrently, obtained TTE (apical 4-chamber view). From the echocardiographic recordings, the investigators calculated two estimates of left ventricular filling pressure: (a) the early-to-late transmitral diastolic (peak) blood flow or E/A ratio and (b) the early transmitral diastolic flow-to-early diastolic mitral annular velocity (Ea) or E/Ea ratio. (Ea is a load-independent measure of myocardial relaxation). Patients were excluded if they had severe mitral regurgitation, mitral stenosis, mitral prosthesis, atrial fibrillation, or myocardial ischemia.
Seventeen patients developed weaning-induced increases in PAOP defined as a PAOP greater than 18 mm Hg. All of them failed the weaning trial. Twenty-two patients did not develop weaning-induced increases in PAOP and 46% of them succeeded the trial of weaning. No physiologic variable recorded before weaning distinguished the two groups of patients. At the conclusion of the spontaneous breathing trial, however, patients with weaning-induced increases in PAOP had higher systemic and pulmonary blood pressure, were more tachypneic and had a lower mixed venous oxygen saturation than those patients who did not developed weaning-induced increases in PAOP. Similarly, the E/A and the E/Ea ratio were greater in the first than in the second group of patients.
Receiver operating characteristic curves were constructed to determine the thresholds of E/A and of E/Ea associated with the best sensitivity and specificity for predicting weaning-induced increases in PAOP. With this method, the investigators reported that a value of E/A > 0.95 was associated with 88% sensitivity and 68% specificity to correctly identify those patients with weaning-induced increases in PAOP. A value of E/Ea > 8.5 was associated with 94% sensitivity and with 73% specificity to correctly identify those patients with weaning-induced increases in PAOP. The sensitivity and specificity associated with a combination of E/A > 0.95 and E/Ea > 8.5 to predict weaning-induced increases in PAOP were 82% and 91%, respectively.
The intraobserver variability for the E/A and E/Ea ratios were 4.4 +/- 3.5% and 4.7 +/- 2.5% and the interobeserver variability were 7.0 +/- 4.1% and 6.9 +/- 6.4%, respectively.
The investigators concluded that the combination of E/A > 0.95 and E/Ea > 8.5 at the end of the trial of spontaneous breathing accurately identified patients with weaning-induced PAOP elevation. Moreover, they concluded that the same cutoff values of E/A alone or of E/Ea alone had weak specificity in detecting weaning-induced PAOP elevation.
The investigators should be commended for this very interesting and difficulty study. The study, however, has certain limitations, some which are pointed out by the investigators themselves:
1-Calculation of the E/A or E/Ea ratios cannot be accurately performed in patients with severe mitral regurgitation, mitral stenosis, mitral prosthesis, or atrial fibrillation. This means that we cannot use this non-invasive technique in patients that are probably at high risk for weaning-induced increases in PAOP.
2-Good quality imaging is necessary for the proper interpretation of TTE. Hyperinflation and COPD are common conditions which can interfere with TTE. The fact that Lamia et al were able to obtain satisfactory TTE images in all their patients, even in those 13 who had a history of COPD (severity not reported) is remarkable. At this point, however, it is unclear how applicable this technique can be in the general population of patients cared for in an ICU.
3-Nearly half (44%) of patients developed weaning-induced increases in PAOP. This raises two considerations. First, the sample of patients studied by Lamia et al is a very selected one. It would seem unusual that, in a general intensive care unit, the mechanism for weaning failure of nearly half of patients who are difficult-to-wean is myocardial dysfunction. Second, is it possible that some degree of misclassification contributed to the high rate of weaning-associated increase in pulmonary occlusion pressure? The investigators did not record esophageal or gastric pressures during spontaneous respiration, therefore, it is impossible to know how much expiratory muscle contraction and/or dynamic hyperinflation contributed to the rise in PAOP recorded during the weaning trial. Despite this limitation, the fact that mixed venous oxygen saturation in patients who experienced weaning-induced increases in PAOP was less than in those who did not experienced weaning-induced increases in PAOP is solid evidence that the cardiovascular system in the first group of patients was truly under stress.
In conclusion, Lamia et al have shown, for the first time, that it is possible to use indices derived from transthoracic echocardiography to accurately identify patients who experience elevation of PAOP during weaning. The challenge for clinical investigators will be to study how to best use this information for in the care of difficult-to-wean patients.