Persistent air leak (PAL) is caused by an abnormal communication between the alveolar space and the pleural space lasting more than 5–7 days.1

It occurs as a complication in up to 15% of thoracotomies, in spontaneous pneumothorax in patients with underlying lung disease, chest injury, as a complication of radiofrequency ablation of lung tumors, for treatment-related reasons (thoracocentesis, pleural biopsy, transthoracic aspirations) or as a complication of empyema or necrotizing pneumonias.2 It generally develops in patients with chronic lung diseases and/or multiple pathologies, and is associated with long hospital stays, more intensive care unit (ICU) admissions, prolonged use of chest tube drainage, with the resulting increased risk of infection, greater morbidity, and risk of death, all of which are reflected in high healthcare costs.3,4

PAL has traditionally been managed with surgical intervention, however, due to the above-mentioned comorbidities, some patients are inoperable. For this reason, minimally invasive endoscopic techniques have been developed in recent years, including one-way endobronchial valves (EBV). These devices offer the best efficacy and safety available to date.5

EBV were initially designed for the endoscopic treatment of loss of volume in emphysema, and were later introduced for the treatment of PAL. The device works in a similar way to the Heimlich valve, preventing airflow toward the pulmonary areas distal to the valve, while permitting drainage of mucus and air in the proximal direction.6,7 In some cases, collapse of the affected segment and expansion of the normal adjacent lung segment is achieved,8 thus contributing to the reduction or resolution of the leakage defect and subsequent healing of the lung parenchyma.9

We report a series of 8 patients with persistent airflow treated with endobronchial valves, and compare our experience with the results from other series published to date.

Two patients were excluded from the study after the origin of the PAL could not be located with the occlusion procedure.

Nine procedures were performed in 8 patients with PAL. The series consisted of 7 men and 1 woman, with a mean age of 68.5 years. All patients has significant comorbidities, including severe pulmonary emphysema in 7 patients, respiratory failure in 5, thrombopenia <20000 platelets per microliter in 1 and ischemic heart disease in 1 patient. In 6 patients, PAL occurred after development of a spontaneous pneumothorax secondary to underlying lung disease (Fig. 4A and B), after treatment-related pneumothorax in 1 and after anatomical resection of lung cancer in 1.

Figure 4.

Patient with right pneumothorax and lung emphysema (patient no. 5). (A) Chest X-ray after chest tube placement. (B) Detail of chest X-ray.

(0.09MB).

Air leak was continuous in 5 cases and intermittent in the other 3. On average, 2 valves were placed per patient (IBV in 7 and Z-EBV in 1), with a mean pleural drainage time before insertion of 15.5 days. One patient with secondary spontaneous pneumothorax (case no. 8) did not initially respond to valve placement in the right lower lobe. The patient underwent surgery, but this was also ineffective, so a new EBV was placed in the right upper lobe and the clinical picture was finally resolved.

No complications were recorded during the procedure, mean time of pleural drainage after valve placement was 13 days and leakage was stopped in 6 patients, giving a success rate of 75% in our patients. There were no adverse events associated with the implant during the 3 months of follow-up.

The valves were left in place for a median time of 52.5 days, although in 1 case it was not removed until 2 years later, as the patient failed to attend the follow-up visit. Patient characteristics are described in Table 1.

In this paper we report the largest series of patients with PAL treated with implantation of endobronchial valves published to date in Spain. This treatment was effective in secondary pneumothorax and pneumothorax after lung resection. Complete cessation of leakage in our patients was achieved in 75% of the cases, with no need for other procedures after withdrawal of the chest tube.

Relatively few studies in the literature have addressed this issue. In general, retrospective case series with a small heterogeneous population in terms of patient characteristics and air leak etiology have been described. The first case of a PAL successfully treated with endobronchial valves in the world literature was published in 2006 by Feller-Kopman et al.12 in Spain, the only publication on this subject was that of Rosell et al.13 which appeared in 2010. The largest series are those published by Travaline et al.14 with 40 patients, reporting a success rate of 48% and by Firlinger et al.15 in 2013, reporting on 16 patients with a success rate of 76.9%. Other publications report case series of between 1 and 7 patients.8,9,12,13,16–23

These articles were reviewed, and without taking into account the heterogeneity of the sample, the use of valves has been reported in 90 patients. Complete resolution of air leak in 63% of these patients (57 patients) and partial resolution in 28% (26 cases) was achieved. Mean number of valves used in all the reports was 3.

In general terms, all these reports present EBV as a valid alterative for the treatment of PAL, coinciding with the findings of our study. However, ideal candidates for this treatment cannot be clearly identified from these results. In our series, 3 patients had to undergo surgery, despite their severe comorbidities, as the air leak was not completely resolved after EBV placement. The first of these underwent an intervention 14 days after valve placement. In this patient, the occlusion test may have been unsuccessful as the air leak, being intermittent, was difficult to identify. In the second patient, the air leak was so severe that despite EBV treatment, the decision was taken to operate on day 5. In the third case (case no. 8), surgery after valve placement was unsuccessful, so a second non-invasive procedure was performed for placement of new EBV that were effective in stopping the air leak.

When PAL is intermittent, it can be difficult to locate the air leak using the occlusion test, particularly if it is not completely stopped and only some reduction is observed. Firlinger et al.15 excluded patients with intermittent leak from their study, and this would explain the greater success rate in their series compared to the figures published in the literature. Moreover, they used a digital measurement system that may have helped in the selection of patients for this type of treatment. In our series, a similar success rate was obtained, despite the inclusion of 3 cases with intermittent air leak.

Since the series was small, it is difficult to determine reasons for response or non-response to EBV, suggesting that prospective studies and analyses of predictors for success in this type of treatment would be of interest: type and volume of PAL, age, cause of leak, previous lung function, body mass index, etc.

Extrapolating from the work of Wan et al.24 in the context of volume reduction treatment, possible complications of the procedure include pneumonia, empyema, hemoptysis, and respiratory failure. Mortality is low and associated more with the underlying disease than the procedure. Jenkins et al.25 also described valve migration as a possible complication and in the series published by Travaline et al.14 patients with incorrect valve positioning, expectoration of a valve, moderate desaturation, pneumonia, and colonization of the valve with methicillin-resistant Staphylococcus aureus were reported. Another significant complication mentioned in the literature is dyspnea and respiratory failure due to bronchial occlusion in patients with emphysema and with prior lung resection for lung cancer. In this respect, Dooms et al.26 described a case in which, after right lower lobectomy, occlusion of the bronchi of the right upper lobe produced no atelectasis of that lobe, probably due to collateral ventilation. However, collateral ventilation does not prevent leakage closure after valve insertion, since the valve itself reduces the passage of air, allowing the lung to heal. No atelectasis was found in our series, so none of the patients experienced increased dyspnea or oxygen desaturations after valve placement. Moreover, development of atelectasis over time was not expected as full lobe occlusion was not performed in any of our cases.

To sum up, in our series the procedure was safe and well-tolerated and free of complications. No complications developed during the 3 months of follow-up, although the youngest patient had a recurrence, in the form of a small apical chamber that did not require drainage and was resolved with respiratory physiotherapy. As in other studies, the limitations of a lack of control group, the small sample size and the range of causes of air leak must be mentioned.

Treatment of persistent air leak with endobronchial valves is minimally invasive, effective, and safe. It is a good alternative for patients with high surgical risk. The results of our series are similar to those published by other groups of investigators, and the study presents the same weaknesses: it is retrospective, the sample size is small, and the causes of PAL are heterogeneous. Well-designed prospective studies are needed to confirm the indications for EBVs and factors predictive of success.

Dr. Cordovilla was responsible for conducting procedures, study concept and design, data collection, analysis and interpretation, drafting the manuscript and review of its intellectual content and review of the final version submitted for evaluation.

Dr. Torracchi participated in conducting the procedures, study concept and design, data collection, analysis and interpretation, drafting of the manuscript and review of its intellectual content, and review of the final version submitted for evaluation.

Dr. Varela participated in conducting the procedures and in the review of its intellectual content and review of the final version submitted for evaluation.

Dr. Jiménez participated in conducting the procedures and in the review of its intellectual content and review of the final version submitted for evaluation.

Dr. Aranda participated in conducting the procedures and in the review of its intellectual content and review of the final version submitted for evaluation.

Dr. Novoa participated in conducting the procedures and in the review of its intellectual content and review of the final version submitted for evaluation.

Dr. Barrueco contributed to the review of the intellectual content of the manuscript and review of the final version submitted for evaluation.

The authors do not have any conflict of interests.