Abstract

We report a case of severe reexpansion pulmonary edema that occurred immediately after reinflation of a collapsed lung by rapid negative pressure drainage of prolonged malignant pleural effusion and pneumohemothorax. Although hemodynamic stability was difficult to maintain under aggressive treatment with inhalation of nitric oxide, inotropics and prostacyclin infusion, conventional pulmonary artery catheterization was not adequate for surveillance and adjustment of fluid therapy. For balancing the preload and the extent of pulmonary edema, pulse contour cardiac output monitoring using a single transpulmonary thermal dilution technique was applied to achieve optimal cardiac preload for organ perfusion and to prevent worsening of pulmonary edema from fluid overload.

Keywords

extravascular lung water; fluid therapy; pulmonary edema;

1. Introduction

Reexpansion pulmonary edema (RPE) is a rare but well described complication that occurs when a collapsed lung expands rapidly by evacuation of air or fluid in the pleura.¹ The onset of symptomatic RPE can be rapid and dramatic after decompression and mortality is estimated to be as high as 20%.² If RPE is complicated with significant third spacing in the lung,^3,4 a rapid fluid shift could deplete the intravascular volume and lead to cardiovascular collapse.⁵ In such a condition, fluid management can be challenging for physicians attempting to achieve optimal organ perfusion without worsening pulmonary edema arising from fluid overload. We report a case of RPE complicated by hypovolemic shock after operation and highlight that pulse contour cardiac output (PiCCO) monitoring is useful in functional hemodynamic monitoring to guide fluid management in patients with pulmonary edema and hypovolemia.

2. Case Report

A 29-year-old woman with a history of excision of right thigh synovial sarcoma, followed by localized radiotherapy 7 years previously, developed progressive dyspnea during her 24^th gestational week. Chest X-ray showed massive pleural effusion all over the right lung. She repeatedly received thoracocentesis to relieve the symptoms, but no further aggressive intervention was made because of her strong desire to keep the baby. As her symptoms worsened, emergency cesarean section under general anesthesia was performed in the 29^th gestational week. As dyspnea and generalized edema were aggravated within a week after the delivery, she was transferred to our hospital. Follow-up chest X-ray showed total white-out of the right hemithorax with left mediastinal shift, and computed tomography findings were compatible with right hemopneumothorax.

Emergency thoracotomy and decortication were performed under general anesthesia with pressurecontrol positive-pressure one-lung ventilation. The peak inspiratory pressure of mechanical ventilation was set at less than 30 cmH₂O, with positive endexpiratory pressure (PEEP) of 5−10 cmH₂O to maintain adequate oxygenation. More than 6 L of serosanguinous pleural effusion and sustained blood loss were drawn out during decortication. In addition, blood transfusion and medications were needed because of persistent hypotension with tachycardia. Onelung ventilation could barely maintain pulse oximetry at 90−95% at an inspired O₂ fraction (FiO₂) of 100% (30 cmH₂O of peak inspiratory pressure; 250− 300 mL of tidal volume delivered), even with additional application of high-frequency jet ventilation at the rate of 150 per minute to the non-dependent lung. Before closure of the right hemithorax, the collapsed right lung was reexpanded with manual positive-pressure inflation with peak inspiratory pressure less than 30 cmH₂O to examine whether there was air leakage. Immediately after reexpansion of the collapsed right lung, frothy, pink and bloodstained sputa were found in the tracheal lumen of the left-sided double-lumen endotracheal tube (35 Fr, Broncho-Cath^TM Left; Mallinckrodt Medical, Athlone, Ireland). RPE of the right lung was suspected.

For monitoring the hemodynamics and fluid management, a pulmonary artery catheter was placed via the right internal jugular vein soon after the operation. In the first 12 hours in the intensive care unit (ICU), severe hypotension, diffuse pulmonary edema over the right hemithorax on the chest X-ray (Figure 1) and poor lung injury score (LIS; PaO₂/ FiO₂: 105 mmHg) were noted, but the central venous pressure (CVP; 8 mmHg) and pulmonary capillary wedge pressure (PCWP; 11 mmHg) were acceptable. Pulmonary hypertension (43/21 mmHg) persisted despite use of inhaled nitric oxide, inotropics and prostacyclin infusion. Transesophageal echocardiography was performed, which revealed low ventricular filling without systolic or diastolic dysfunction of the left ventricle.

In order to achieve a better guide for fluid management and a better method for surveying pulmonary edema, the PiCCO system (PiCCO Plus; Pulsion Medical Systems, Munich, Germany) was then established. Optimization of hemodynamics was achieved according to a protocol-guided therapeutic strategy, including global end diastolic volume index (GEDVI; 680−800 mL/m² ) and extravascular lung water index (EVLWI; < 10 mL/kg). During the stay in the ICU, the data obtained by the PiCCO system, i.e. GEDVI, EVLWI, pulmonary vascular permeability index (PVPI) and LIS were as shown in Table 1, the lowest LIS being on the 6^th and 7^th postoperative days. Under hemodynamic and volumetric monitoring by the PiCCO system, the cardiovascular circulation was optimized for adequate organ perfusion and the respiratory oxygenation improved from then on. She was extubated and discharged from the ICU on day 14 without cardiac and renal complications.

Download full-size image

Figure 1 Reexpansion pulmonary edema over the right lung field shown on chest X-ray 12 hours after operation.

Download full-size image

3. Discussion

Conventional treatments of pulmonary edema include water restriction, diuretics, increase of colloid oncotic pressure, and protective ventilatory strategy (low tidal volume, application of PEEP). Nevertheless, optimization of hemodynamics and fluid balance remain difficult in mechanically ventilated patients with pulmonary edema and hypovolemia. Adequate cardiac preload is crucial to restore and maintain organ perfusion but these patients are at high risk of aggravation of pulmonary edema due to potentially leaky capillaries.³ Hypotension often follows RPE due to rapid massive effusion, continuous surgical bleeding or fluid shift into the alveolar space;^4,5 even cardiovascular collapse may ensue in severe cases.⁶ In this case of severe progressive RPE, conventional monitoring with CVP and PCWP was incapable of assessing adequate cardiac preload because of mechanical ventilation with high PEEP support to maintain respiratory stability. On the other hand, GEDVI is reportedly a more sensitive guide for fluid management than PCWP and CVP, especially under the influence of applied PEEP.⁶ Here, we present the application of the PiCCO system for adequate fluid management of severe RPE, using parameters like GEDVI, PVPI and EVLWI to balance the cardiac preload with the pulmonary condition.

Along with LIS and the clinical picture on chest X-ray, EVLWI can serve as a surrogate for parameters determined by pulmonary arterial catheter in severe pulmonary edema.⁷ In combination with PVPI, the causes of pulmonary edema can be further differentiated between cardiogenic fluid overload, congestive heart failure, or a noncardiogenic origin, arising from a permeability change from capillary membrane disruption, such as in acute respiratory distress syndrome.^8,9 In this case, the changes in PVPI were correlated well with the changes in LIS that increased from the 2^nd day and peaked on the 6^th to 7^th day (Table 1). The delayed worsening of LIS were considered to be mainly effected by permeability change, instead of hydrostatic pressure from fluid overload, because EVLWI remained relatively high and stable throughout the whole course in the ICU, whereas the PVPI increased. It was difficult to maintain pulmonary oxygenation while the increase in permeability continued. While permeability change plays a major part in pulmonary edema in sepsis and acute respiratory distress syndrome, PVPI may be regarded as a better guide for fluid management in pulmonary edema arising from capillary membrane disruption. Prolonged higher EVLWI than average may be the result of capillary membrane disruption from prolonged pneumohemothorax or physiological change during pregnancy.

The pathophysiology underlying the development of RPE has not been well determined.¹⁰ For a permeability change from reperfusion and reoxygenation to cause damage of the endothelium,¹¹⁻¹⁴ hydrostatic factors, such as pulmonary vascular flow redistribution, have been proposed to be involved.¹⁰ Although RPE is usually described in the treatment of a chronically collapsed lung caused by pneumothorax and pleural effusion, it can also occur immediately following surgical procedures that employ one-lung ventilation in operating theaters.^15,16 It is preferable that reexpansion of the collapsed lung should be carried out gradually after the thoracic procedures while the thoracotomy tube for drainage should not be under excessive negative pressure suction. Diagnosis should be mainly based on knowledge of this potentially life-threatening complication (history of collapsed lung with rapid reinflation or evacuation, prominent unilateral increased infiltrate on chest X-ray), while at the same time excluding cardiogenic or other factors such as transfusionrelated acute lung injury, pneumonia, or fluid excess as the cause of pulmonary edema. If RPE occurs, the treatment is supportive, including positive-pressure ventilation. Diuresis and inotropic support should be considered in individual cases.

In conclusion, it is very important that physicians who take care of a patient with a collapsed lung should be aware of RPE. This will facilitate prevention of RPE and offer timely and appropriate treatment if it does occur. The balancing of fluid management on preload and extravascular lung water remains difficult in severe cases of RPE. For better management of optimal preload for organ perfusion and prevention of worsening of pulmonary edema, we suggest that PiCCO monitoring could offer a more rational guide than pulmonary artery catheterization in patient management. Its role in this aspect warrants further investigation.