A follow-up study of 137 patients who had 143 tuberculous lesions of the spine which were treated by constitutional therapy, chemotherapy, and débridement during the years 1948 through 1962 is presented. These lesions were selected for operation because it was felt that they were well established and unlikely to respond satisfactorily to constitutional treatment and chemotherapy only. The main findings were as follows:

The number of patients having from two to three vertebral lesions was high, being 108 out of 143 (74.8 per cent). The prognosis for deformity and stability in this group was better than in the group of thirty-five lesions (25.2 per cent) involving four vertebrae or more.

Ninety-five of the 108 patients having lesions extending over two to three vertebrae (88 per cent) recovered with no or only slight deformity, although neither anterior nor posterior grafts were used.

Seven of thirty-five patients having lesions extending over four or more vertebrae (20 per cent) recovered with no or only slight deformity. Severe deformity occurred in nine of thirty-five of these patients (25.7 per cent) and a need for anterior or posterior grafting therefore does arise in this group if a considerable gap has occurred on account of bone loss.

The principle was observed that débridement should include only necrotic portions of cartilage and bone. Thirty-four patients were followed for periods of from ten to fourteen years, and five for periods of fifteen years or more, and it was possible to see that growth had continued from cartilage end plates which had escaped destruction in children; the amount of stunting which had occurred in these children had therefore been limited. It was also possible to see in adult spinal lesions that defects had been made good by re-ossification. The need for grafts after débridement may be less than is supposed.

The principle that was gradually developed of earlier operation ensured that progressive vertebral destruction, reported in some series treated without débridement, was halted early as it was seldom seen after débridement. The gradual reduction over the years of the duration of postoperative rest in plaster and the increased restriction of the use of ambulatory splints ensured the optimum restoration of circulation to the tissues engaged in repair with improvement in their nutrition and the capacity of bone for remodeling.

The duration of chemotherapy required was less than in patients treated without débridement. Streptomycin and isoniazid were given in combination for six months, and followed by isoniazid and para-aminosalicylic acid in combination for three months, usually as out-patient therapy.

The relapse rate of patients treated later in the series had fallen to 2 per cent.