ARTIFICIAL PNEUMOTHORAX¹: A Survey of a Personal Series

A survey is given of the present situation of artificial pneumothorax as a treatment for pulmonary tuberculosis. With few exceptions the recorded series are devoid of uniform assessment and long continuity of supervision, because they are compiled from the work of numerous observers. Moreover, they deal mostly with patients whose pleural adhesions were not divided, so that many had unsatisfactory pneumothoraces.

The present paper concerns the treatment by pneumothorax of 50 patients who were under the same supervision continuously, and whose pleural adhesions were divided at an early stage of treatment whenever possible. Fifty-four pneumothoraces (four bilateral) were induced between 1935 and 1944, and the patients were observed until 1949, that is for periods of five to 15 years. Eighteen pneumothoraces were done for active infiltration, and 36 for infiltration with cavitation. Severe toxaemia, pneumonic disease, predominant fibrosis, atelectasis, and very large cavities were all regarded as contra-indications to this form of treatment. A few patients with one or other of these features were included early in the series, but they failed to benefit, or were harmed.

Atelectasis of the lung, or of a lobe, was a common early complication (22 in 54 pneumothoraces), chiefly after thoracoscopic division of adhesions (15 in 22). With only two exceptions it was unimportant. The lung was aerated quickly again in all but four cases. Pleural effusion, the main complication (30 cases in 54), did not chiefly affect patients who had little preliminary rest. Small effusions, not rising above the dome of the diaphragm, were benign apart from a little risk (four cases in 15) that obliteration of the pleural space might follow. Large effusions were equally common, were often persistent, and in eight out of 15 cases followed division of extensive adhesions. Tuberculous empyema and premature loss of the pneumothorax were the chief risks. A pneumothorax with extensive adhesions, whether these were divisible or not, had a greater risk of effusion and of an effusion becoming purulent.

The pneumothoraces were classified as (1) ‘complete’, when free from pleural adhesions; (2) ‘adequate partial’, when adhesions were thought not to prevent relaxation of the diseased area; and (3) ‘inadequate partial’, when adhesions interfered with relaxation of the diseased area. Seventeen patients with pulmonary cavitation had pneumothoraces of the first two types, and in all of them the cavities closed. Twenty others with cavitation had pneumothoraces of the third type; in 14 of these the cavities closed, aided by phrenic nerve paralysis in three.

Nine of the 54 pneumothoraces were abandoned for various reasons, or obliterated, within three years. Only one of these patients remained well without further treatment. Undesired obliteration occurred in seven others after more than three years; five of these patients remained well. Twenty-six pneumothoraces were deliberately abandoned after an average duration of 5.3 years. Twenty-two of these patients remained well. Twelve patients still have pneumothoraces of more than five years' duration; all have quiescent disease in the treated lung. Thirty-eight of the original 50 patients are well and at work. Of the remainder six have died from their disease, one has died from an unrelated illness, and five are in relapse.

Our experience with this series of patients during 15 years favours the use of artificial pneumothorax: We think the chief conditions for success have been: recognition of certain contra-indications; early division of pleural adhesions; maintenance of those pneumothoraces only which were complete, or had adhesions remote from the diseased area or limited to the superior mediastinum; and a minimum duration of three years, and preferably five years in the presence of initial cavitation. Less tangible, but equally important, is a single control of policy throughout treatment.