Negative airway pressure pulmonary oedema is rare yet a few cases have been described. This complication is likely to occur more often than it is diagnosed. The condition is life threatening and requires prompt institution of proper management. Pulmonary oedema due to negative airway pressure during anaesthesia and intensive therapy may have various causes, e.g. acute epiglottitis, bronchial stenosis after stenting, strangulation, goitre-related compression, obstructed endotracheal tubes or laryngeal masks, laryngospasm [1, 2, 3, 4].

We report a case of pulmonary haemorrhage caused by negative airway pressure during general anaesthesia.

CASE REPORT

A 25-year-old man weighing 71 kg had an accident during a motor race. Immediately after the accident, he was hospitalized in the local hospital and diagnosed with a right clavicle fracture with an intermediate fragment, brain concussion and head contusion. The patient left hospital at his request, however, 3 days later, he came back and was scheduled for surgery of a fractured clavicle on the following day.

The history taken did not reveal any additional diseases or risks. The patient did not report dyspnoea; no symptoms of respiratory failure were observed. The physical examination during the anaesthetic consultation showed symmetric respiratory murmur over the lung fields, HR 70 min-1, no abnormalities except for pain in the region of fracture and nape. During the 3-day period at home, the patients had no dyspnoea; his activities were only limited by local pain. A neurological consultation after admission revealed brain concussion; head CT was normal. The additional tests showed slight anaemia (Hb 7.3 mmol L-1, Ht 0.33), which was associated with trauma.

The patient was qualified for surgical reposition and stabilization of the fractured clavicle under general anaesthesia. Chest X-ray was not ordered as the physical examination carried out by a surgeon and anaesthesiologist was not suggestive of any respiratory problems. Chest X-ray taken during the first stage of treatment was considered sufficient for surgical diagnosis.

General anaesthesia was induced with 0.5 mg of atropine, 2 mg of midazolam, 0.2 mg of fentanyl, 250 mg of thiopentone and 90 mg of suxamethonium. Laryngoscopy and endotracheal intubation with the 7.5 tube were performed without complications; after intubation vecuronium was administered 4 mg. Anaesthesia was maintained with the mixture of N2O and O2, with the addition of isoflurane 1% followed by 0.6% and fractionated doses of fentanyl; controlled ventilation was carried out. During the procedure, 2000 mL of crystalloids were transfused. The 75-minute anaesthesia was uneventful, haemodynamic condition of the patient was stable, and airway pressure normal, symmetric alveolar murmur was heard over the lungs. Since the induction until the removal of the endotracheal tube, SpO2 maintained within 98-100%, pCO2 – 37-39 mm Hg (4.0-5.2 kPa). After the procedure and intravenous administration of 0.5 mg of atropine and 1.5 mg of neostigmine, the patient, spontaneously breathing, with efficient respiration and logical c
ontact, was extubated.

Several minutes later, respiratory deterioration, cyanosis, decreased SpO2 to 40% and inspiratory wheezing were observed. The anaesthesiologist assessed that this condition resulted from residual effects of muscle relaxants and administered passive, followed by active oxygen therapy and 0.5 mg of atropine and 0.5 mg of neostigmine. SpO2 remained decreased (40%), cyanosis increased, oral bloody discharge and bradycardia were detected. Atropine was administered, the patient was immediately intubated and artificial lung ventilation with oxygen initiated. The foamy, bloody content was sucked out of the endotracheal tube; numerous moist rales were heard over the lungs. Despite the therapy used, cyanosis persisted and arterial blood pressure was 90/50 mm Hg.

The patient was transferred to ITU, chest X-ray was taken; arterial pressure was directly monitored and additional peripheral venous accesses were provided. The patient received colloids and crystalloids. Arterial blood gasometry revealed: pH 7.21; pCO2 81.2 mm Hg (10.8 kPa); pO2 32 mm
Hg (4.3 kPa); BE 0.9 mmol L-1; SaO2 47.3%.

During the first postoperative hours, once the general condition was stabilized, the haemorrhage was considered to be caused by re-intubation and subsequent laryngeal/ bronchial trauma or tracheal puncture. However, CT scan was suggestive of pulmonary oedema or aspiration during bronchial tree haemorrhage and developing circulatory-respiratory failure.

Chest CT disclosed numerous densities of the pulmonary parenchyma in the upper lobes of both lungs, mainly in the perihilar and posterior parts. Similar, yet less numerous lesions were visible in the lower lobes next to the poles of both hili. No fluid or air was found in pleural cavities. The values of clotting parameters and electrolytes were within normal limits, blood test results were similar to the preoperative values. In ITU, CMV was continued at FIO2 1.0. The urinary catheter and gastric tube were placed. Intensive pharmacotherapy (vitamin K, cyclonamine, aprotynine, furosemide, aminophylline, ambroxol, pipecuronium, fentanyl and midazolam in continuous infusions, cefuroxime and amikacin) and fluid therapy (crystalloids, 2 units of fresh frozen plasma, 1 unit of red blood cells) were initiated. Due to an increase in SAP/DAP to 170/80 mm Hg, a single dose of urapidil was administered followed by the infusion of nitroglycerine, which led to blood pressure normalization. The heart action was regular, firstly about 100 min-1, then – 70 min-1. Auscultatory changes were still heard over the lung fields, the foamy content and bright blood from the airway were sucked out. The jugular veins were not dilated. Due to oedema of the affected region, the central line was placed into the right femoral vein: the first venous pressure measurements were at the level of 0 cm H2O, on the next day – from 1 to 3 cm H2O.

After several hours, airway haemorrhage stopped, gasometric parameters gradually normalized, and auscultatory changes subsided. The anti-haemorrhagic therapy was continued which allowed to decrease gradually the O2 content in the respiratory mixture. After 24 h, the pipecuronium infusion was discontinued, after further 10 h mechanical lung ventilation was also withdrawn and the patient was extubated. During the ITU therapy, no clotting disturbances or symptoms of heart failure were observed.

On day 3, the nitroglycerine infusion was discontinued and the patient was transferred to the trauma ward with fully efficient respiratory and circulatory functions; on day 9, he was discharged home. Shortly after, he returned to sports. Several months later, the patient had the osteosuture removed under general anaesthesia - without complications.

DISCUSSION

Airway haemorrhage may be caused by damage to the airway or pulmonary parenchyma, systemic disease or haemorrhagic diathesis.

In the case described, lung contusion and post-traumatic lesions may be excluded as the causes of pulmonary oedema and haemorrhage since before anaesthesia, no pathological respiratory symptoms were observed and during anaesthesia, the monitored respiratory parameters (SpO2, ETCO2 and airway pressure) were within normal limits. CT performed directly after the procedure excluded features of lung contusion and atelectatic lesions.

Considering the patient`s young age, good general condition and negative history, no systemic diseases or Goodpasture`s syndrome were suspected; moreover, fast recovery and lack of symptoms during the several-year observation were not suggestive of any such diseases.

Clotting abnormalities did not cause haemorrhage as the history, laboratory tests and the course of surgery evidenced normal clotting parameters.

Bronchoscopy was not performed during the episode described because in the first hours the patient`s condition was too severe and bronchoscopy could have deepened already dramatic hypoxaemia. Delayed bronchoscopy did not seem necessary as symptoms were quickly subsiding, imaging and gasometric follow-up examinations were normal. In general, bronchoscopy is not recommended for short-term haemoptysis (less than a week) in a patient below 40 years of age [5].

On awakening from anaesthesia, after extubation, the patient developed laryngospasm inducing acute respiratory failure. Initially, the anaesthetist wrongly interpreted this situation as persisting muscle relaxation and administered the additional dose of neostigmine, which did not cause any improvement. The spasm of respiratory muscles with the glottis closed most likely resulted in the development of this rare complication, i.e. pulmonary oedema and haemorrhage caused by high negative airway pressure.

During forced inspiration, negative intrapleural pressure may reach 30 mm Hg (4 kPa); in muscular patients with the glottis closed – even 100 mm Hg (13.3 kPa) [2]. High negative thoracic pressure changes haemodynamic conditions. The venous return increases both due to decreased pressure in the right atrium – associated with transfer of negative pressure inside the thoracic cavity  and increased mean pressure in the systemic circulation – related to constriction of veins caused by catecholamines secreted in response to anxiety, hypoxia and hypercarbia. Negative pressure inside the thoracic cavity and catecholamine-induced increased vascular tone in the systemic circulation increase the transmural pressure within the left ventricle thus increasing the tonus of its walls. This increased afterload reduces the left ventricular ejection. Due to all these phenomena, the blood transfers from the systemic to pulmonary circulation.

Pulmonary oedema develops once transmural pressure in pulmonary capillaries increases. Hydrostatic pressure in pulmonary capillaries increases due to increased blood volume in the pulmonary circulation and enhanced tone of pulmonary vessels (caused by hypoxia and acidosis). During the Mueller manoeuvre (inspiration test at closed glottis) hydrostatic pressure may be highly negative contributing to high transmural pressure in the pulmonary capillaries. Such pressure changes account for the development of oedema with the transudate without erythrocytes [6].

The phenomena described may result in the transudate but not in alveaolar haemorrhaging. Although the exact aetiology of haemorrhage resulting from negative airway pressure is unclear, it seems that the disruption of pulmonary capillary continuity is likely to be essential [6].

The walls of pulmonary capillaries are extremely thin. It has been demonstrated that increased pressure in these vessels generates their stress failure and causes pulmonary oedema with excessive permeability of walls or even pulmonary haemorrhage. The experiments on rabbits showed that once transmural pressure was 38.6 mm Hg (5.1 kPa), discontinuity of alveolar epithelium or even discontinuity of basilar membranes were observed under electron and scanning electron microscope.  The capillary wall was damaged due to their tension (i.e. tonus divided by the wall thickness) rather than tonus. For a capillary of 5 μm radius and 0.3 μm wall thickness at transmural pressure 52.5 cm H2O (5.1 kPa), the calculated wall tension is 90 kPa. Such a tension is extremely high, similar to that in the aortal wall, which is strengthened by the layers of collagen and elastin. Although the pulmonary circulation is considered the low-pressure one, its thin-walled capillaries are exposed in such cases to very strong forces. It is n
ot surprising than that pulmonary capillaries are damaged but that is happens so rarely [7].

The cause of haemorrhage may also be increased pressure in the systemic circulation induced by hypoxia and adrenergic activation, which results in discontinuity of bronchial walls belonging to the systemic circulation [8].

A few cases of transient airway haemorrhage in healthy patients diving without scuba devices were reported. Haemoptysis developed after immersions during which they contracted their diaphragm at closed glottis causing negative airway pressure. The symptoms quickly subsided [9].

General anaesthesia for routine surgery in a healthy and young patient may cause severe and unpredictable complications. Thanks to proper management according to generally accepted guidelines, the patient, whose causes of the general anaesthesia-related complication cannot be rapidly recodnized, is likely to survive.

..............................................................................................................................................................

REFERENCES

1.    Koh MS, Hsu AAL, Eng P: Negative pressure pulmonary oedema in the medical intensive care unit. Intensive Care Med 2003; 29: 1601-1604.

2.    Mirshab C: A 19-year-old man with postoperative respiratory insufficiency; in: Anesthesia Pearls (Ed.: Duke J), Hanley and Belfus Inc, Philadelphia, 2003: 183-187.

3.    Sow Nam Y, Garewal D: Pulmonary hemorrhage in association with negative pressure edema in an intubated patient. Acta Anaesthesiol Scand 2001; 45: 911-913.

4.    Dolinsky SY, MacGregor D, Scuderi PE: Pulmonary hemorrhage associated with negative-pressure pulmonary edema. Anesthesiology 2000; 93: 888-890.

5.    Bhavani-Shankar K, Hart NS, Mushlin PS: Negative pressure induced airway and pulmonary injury. Can J Anaesth 1997; 44: 78-81.

6.    Schwartz DR, Maroo A, Malhotra A, Kesselman H: Negative pressure pulmonary hemorrhage. Chest 1999; 115: 1194-1197.

7.    West JB, Mathieu-Costello O: Stress failure of pulmonary capillaries: role in lung and heart disease. Lancet 1992; 340: 762-767.

8.    Koch SM, Abramson DC, Ford M, Peterson D, Katz J: Bronchoscopic findings in post-obstructive pulmonary oedema. Can J Anaesth 1996; 43: 73-76.

9.    Kiyan E, Aktas S, Toklu AS: Hemoptysis provoked by voluntary diafragmatic contractions in breath-hold divers. Chest 2001; 120: 2098-2100.

..............................................................................................................................................................

address:

*Anna Smędra-Kaźmirska
Katedra i Zakład Medycyny Sądowej
Uniwersytetu Medycznego w Łodzi
ul. Sędziowska 18a, 91-304 Łódź
tel. 0 - 42-654-45-36
fax 0 - 42-654-42-93
e-mail: karolanka@wp.pl

received: 18.10.2009
accepted: 11.01.2010