Percutaneous Transthoracic Needle Biopsy Complicated by Air Embolism
+ Author Affiliations
Percutaneous transthoracic needle biopsy is a common procedure for evaluating pulmonary and mediastinal lesions. The most frequent complications include pneumothorax (27-49%), hemorrhage (11%), and hemoptysis (7%) [1,2,3]. Air embolism resulting from thin-needle biopsy is a rare (incidence of ≈ 0.07%) but potentially life-threatening complication . In this article we report a case of air embolism to the heart that was confirmed on CT.
A 60-year-old man with a 10-year history of corticosteroid-dependent chronic obstructive pulmonary disease presented with three episodes in the preceding 5 months of left lower lobe “pneumonia” accompanied by recurrent hemoptysis and a segmental left lower lobe opacity on chest radiography. The patient was further evaluated with chest CT that showed a focal 2.5-cm mass in the posterior segment of the left lower lobe and adjacent ground-glass opacity that was attributed to resolving pneumonia. No hilar or mediastinal adenopathy was present. Several low-density liver lesions were also seen that were too small to characterize on CT.
Because the liver lesions were small and inaccessible and because the lung mass could be approached easily, percutaneous transthoracic needle biopsy was attempted. This procedure was performed 2 weeks after the diagnostic chest CT scan was obtained. In the interval, the appearance of the left lower lobe mass and ground-glass opacity did not change, and the patient's clinical condition remained stable. His medical history was remarkable for asthma treated with steroids and a melanoma of the right cheek that was removed 4.5 years previously. The differential diagnosis of the lung lesion was metastatic melanoma or bronchogenic carcinoma.
The patient was placed in right lateral decubitus position and the lesion was localized using CT. A 19-gauge Chiba-type needle was placed into the lesion on the first pass during a single inspiratory breath-hold (Fig. 1A). This needle was not moved until the end of the procedure, and scans obtained at intervals during the procedure showed its position to be unchanged relative to the pulmonary mass. A coaxial 21-gauge needle was inserted through the 19-gauge needle, and three aspiration biopsy specimens were obtained using suction with a 10-mL syringe. The 19-gauge needle stylet was inserted between obtaining the specimens with the 21-gauge needle. With the 21-gauge needle removed, a final specimen was taken without aspiration using the 19-gauge needle. The patient was entirely cooperative during the procedure. The insertion of the needle and the aspiration biopsy were conducted during suspended inspiration. At no time did the patient cough or breathe inappropriately while the needles were in his chest.
At the end of the procedure, after the removal of the 19-gauge needle, the patient briefly coughed and expectorated a small amount of bright red blood. Immediately afterward, a 10-slice CT scan was obtained to assess whether pneumothorax was present. At the end of this scanning, the patient again coughed and expectorated a small to moderate amount of bright blood. While he was being moved to prevent aspiration, he abruptly became unresponsive, and respiratory arrest occurred. Resuscitative efforts were started immediately and the cardiac arrest team was called. Resuscitative efforts were unsuccessful and the patient died on the CT table.
Because of the acute need to prevent aspiration and the subsequent cardiac arrest, the CT scans obtained after the procedure were only briefly examined; they showed a very small left pneumothorax. After the cardiac arrest team arrived, a more careful review of the postprocedure CT scans revealed an airfluid level in the left ventricle and small bubbles of air along the presumed course of the coronary arteries (Figs. 1B and 1C).
An autopsy was performed, and subsequent microscopic analysis and immunofluorescent staining showed that the left lower lobe mass was malignant melanoma. Autopsy of the chest revealed that the portion of the lung traversed by the needle was chronically inflamed and friable, apparently because of recent pneumonia. The left lung was removed en bloc, and the bronchus was injected with water-soluble contrast material under fluoroscopic visualization. A fistula from a bronchus to a pulmonary vein could not be defined.
Percutaneous transthoracic needle biopsy is a frequently performed procedure that is accepted as a standard method for the diagnosis of pulmonary lesions. Pneumothorax is reported in 27-49% of cases and is the most frequent complication [1,2,3]. Intraparenchymal hemorrhage is also a frequent complication, occurring in about 11% of cases, and hemoptysis occurs in about 7% of cases [1,2,3]. The mortality of the procedure is unclear, although Berquist et al.  reported two deaths resulting from hemorrhage after biopsy.
With respect to air embolism, Sinner's review of the literature  found two suspected cases in 2726 patients. We have found seven additional reports of air embolization during percutaneous needle biopsy of lung lesions [4,5,6,7,8,9,10]. These cases resulted in neurologic morbidity or death. To our knowledge, the case reported here is the first to show gas in the heart and coronary arteries.
The formation of a communication between an airway and a pulmonary vein is the assumed mechanism of air embolization during percutaneous lung biopsy. It is not difficult to understand how this might occur, considering that pulmonary vein pressure is normally only 10 cm H2O. It is possible that the coughing episodes in this patient contributed to the entry of air into the pulmonary veins by increasing air pressure in the pulmonary airways. Other factors such as chronic obstructive pulmonary disease and air trapping may have also contributed to increased airway pressure, resulting in an increased pressure gradient between the airway and the pulmonary vein. An alternate mechanism for air embolization is communication between the needle and a pulmonary vein if the patient breathes while the needle hub is open to the air. This cause seems unlikely in this case, because the needle and stylet were inserted directly into the mass during a single breath-hold and the patient did not breathe while the needle hub was open to the air.
The friability of tissues and vessels is an indication that the patient's corticosteroid dependence may have contributed to the air embolization in this case. The vascular nature of this lesion was shown on autopsy, and the increased vascularity of the mass may have increased the risk of air embolism. Another contributing factor may have been postinflammatory changes in a portion of the lung traversed by the needle. At autopsy, the left lower lobe was easily torn and friable, possibly contributing to the formation of an airway—vein fistula.
This case raises several practical points. First, serious complications can occur during percutaneous lung biopsy, even when the technique is excellent and patient cooperation is perfect. Second, some practitioners perform percutaneous lung biopsies with the patient in sustained expiration, rather than inspiration, in an attempt to prevent air embolization. Such is not the prevailing practice in our department, because in our experience many patients cannot suspend respiration as well in expiration as in inspiration. Therefore, the potential benefit of expiratory biopsy must be judged against the risk of uncontrolled respiration during manipulation of the needle. Third, the use of the outer 19-gauge needle in obtaining a final biopsy specimen may result in an increased risk of an airway—vein fistula formation because of the larger diameter of the needle (compared with the usual 21-gauge biopsy needle). Although this possibility is theoretic, we have stopped using the 19-gauge needle as a biopsy device and simply use it as a conduit for the 21-gauge needle. Fourth, passage of the needle through diseased lung, as in this patient, may increase the risk of complications, including hemorrhage and air embolus. This potential risk should be considered in planning the access route to a lung lesion. In this instance, no other practical approach to the mass was present. Finally, if it had been recognized that gas was present in the left ventricle before it embolized to the brain, the patient might have been quickly placed in Trendelenburg's position and right-side down. However, to do so was not possible in this case because the patient suffered cardiac arrest immediately after the CT images were acquired and before they could be examined. The treatment for other cases of air embolism has been a decompression chamber, and that may have been an option in this case had the presence of air in the left ventricle been discovered earlier.
- Received September 7, 2001.
- Accepted December 6, 2001.
- © American Roentgen Ray Society