Clinical trials in pulmonary tuberculosis.

Abstract

In 1982, the centennial year of the discovery of the tubercle bacillus by Robert Koch, a patient with pulmonary tuberculosis will be cured after receiving only 100 doses of antituberculosis drugs during four and one-half months. Adverse drug reactions will be manageable; the patient taking drugs can continue to work; the possibility of a recurrence is probably zero. This statement follows from the results of one of the most recent (1), certainly not the last, of a long series of clinical trials that began after the discovery of streptomycin in 1944. The modern clinical trial is a carefully designed prospective method of investigating a remedy, involving controls, often by randomization, with expert laboratory and statistical support, frequently carried out in cooperating medical centers and with unbiased centralized analysis and interpretation of results. Such a clinical trial is a recent development and dates in fact from the first Medical Research Council trial of streptomycin in 1947. This trial was the first controlled randomized clinical trial of a drug. Some of the early studies of Prontocil and penicillin were undertaken in puerperal fever and endocarditis, diseases whose predictably high fatality rates did not warrant controls (2, 3). The variable and often unpredictable course of pulmonary tuberculosis, however, required controlled studies. An intense interest in chemotherapy followed the announcement that the cause of tuberculosis was a microorganism (4-6). Many animal studies and clinical trials were undertaken. A dominant feature of the clinical trial was patient selection. The literature of the times suggests that patient selection was based on the presence of clinical features deemed appropriate to the putative action of the substance under investigation, for instance, tuberculin. Doses were adjusted in amount and frequency studies went on. The complexity of pulmonary tuberculosis, its variable pathology, and constitutional manifestations made up a very heterogeneous patient population. The necessity for controlled studies was understood; the protocols of laboratory animal studies compare favorably with present models. The heterogeneity of the clinical material found expression in the development of matched controls, either historic or contemporary. Thus, Mollgaard reported a cooperative study of gold therapy in 647 patients in 5 hospitals with comparative untreated cases control subjects (7). Amberson's clinical trial of gold in 1931 is often cited an early version, of the modern randomized clinical trial (8). Randomization in this study was indeed used. However, randomization was not between patients, but rather between two groups of patients who had been matched as close possible. This trial is of great importance historically because it shows the transition from matched to randomized controls. Amberson's carefully detailed critique of previously reported clinical studies was also a major contribution. It has been suggested that the therapy of tuberculosis in the preantibiotic era might have advanced more rapidly if controlled clinical trials had been used. One can indeed suppose that much more objective information on the indications for inducing artificial pneumothorax and for the management of pneumothorax would have resulted from such trials if they had begun before personal convictions had hardened. Only large scale randomized trials would have disclosed the value of any of certain substances that were thought to act upon the host's defenses, such tuberculin or gold. The dosage of rest and exercise in the treatment of pulmonary tuberculosis remained in dispute into the modern chemotherapy era (9). One may speculate how a protocol for an international trial of rest, exercise, altitude, and climate might have been written and what the results of that trial would have been. In any event, at the time of the discovery of streptomycin, a design of the randomized controlled clinical trial was at hand (10). The agent was so powerful that its efficacy was quickly disclosed. However, it should be emphasized that the subsequent development of chemotherapy regimens, particularly the short course regimens, required the most meticulous and carefully planned trials. In what follows I have attempted an overview of clinical trials in the modern chemotherapy of pulmonary tuberculosis. Investigators in almost every country have carried out a very large number of excellent and fruitful studies, and only in retrospect is it possible to select those that appear to have been the critical ones. The selection is, of course, not without a certain amount of personal bias.