1. Synovial tuberculosis was much commoner than osseous tuberculosis in the knee joint (in this series the proportion was sixty-eight to ten), and tended to occur at a later age.
2. In no proven case of synovial disease did the patient recover sufficiently to bend the knee to a right angle.
3. In no fewer than eight of forty-seven cases of unproven synovial disease was full range of the knee recovered.
4. Of twenty cases of synovial disease treated by immobilization for an average period of five years, in seventeen ankylosis was unsound enough to necessitate arthrodesis.
5. Arthrodesis produced bony union without deformity in all except two patients in whom it was undertaken: one of these died and the leg of the other had to be amputated. The average length of time for which immobilization was necessary after arthrodesis was one and one-fourth years in patients between the ages of nine and thirty. The arthrodeses were performed so as to fix the joint in full extension. Adequate protection of the limb was successful, even in children, in preventing subsequent deformity at the site of operation or at the epiphysis.
6. A careful examination of the skiagrams taken shortly after arthrodesis showed the epiphyseal lines (if present) to be intact in all cases. The distance of the epiphyseal cartilage from the end of the bone leaves the surgeon so much margin that such a result may reasonably be hoped for in synovial infection when the disease has not extended to the epiphyseal cartilage. Thus, following arthrodesis, there is no reason to expect shortening of the limb beyond that which is due to interference with growth by the tuberculous process.
7. The time taken to achieve bony union was independent of any variation in operative procedure. Furthermore, in patients of fifteen or over, it was unaffected by previous immobilization. In younger patients immobilization preceded arthrodesis in every case but one.
8. A comparison of the synovial and osseous groups with respect to mortality brings out the fact that, in the former group, two deaths occurred in sixty-eight cases; whereas, in the latter, three deaths occurred in ten cases.
9. In half the cases of osseous disease complete recovery resulted. It is probable, therefore, that, in cases in which the disease is not too advanced, the knee joint suffers only a so called sympathetic inflammation from which recovery may reasonably be expected whenever radical treatment can be directed to the focus in the bone without opening the knee joint itself. Moreover, only thus can the spread of tuberculous infection into the joint be prevented with any certainty.
10. In the cases proved to be non-tuberculous, the disease of the synovial membrane appeared either to recover completely, or to pass into a condition resembling the osteo-arthritis of middle age, or to get definitely worse with the development of sepsis.
These facts suggest the existence of a distinct disease of the synovial membrane of the knee joint, which lasts for years without affecting other joints. It is not syphilitic. It tends with treatment (or even without) in some cases to get completely well, in some to recover partly, with the development of a premature osteo-arthritis, and in some to become septic. In all these cases the synovial membrane showed the microscopic appearance of a non-specific chronic inflammation.
11. The fact that no proven case of synovial disease recovered suggests that the eight cases of unproven disease that did so may be regarded as: (a) wrongly diagnosed, or (b) primarily osseous cases in which the bony focus was too small to show on the skiagram.
Against the latter view is the fact that none of these cases proved fatal. And in favor of the former is the fact that, of twenty-six cases subjected to synovectomy, five showed no evidence of tuberculous infection. There is, therefore, nothing improbable in the hypothesis that there was an error of diagnosis in eight of the forty-two unproven cases.
(C) 1932 All Rights Reserved.The Journal of Bone and Joint Surgery, Inc.