THE JOURNAL OF BONE AND JOINT SURGERY THE DIAGNOSIS AND TREATMENT OF PIGMENTED VILLONODULAR SYNOVITIS 291 self-perpetuating process which led ultimately to pigmented villonodular synovitis. Geschickter and Copeland (1949) suggested that the lesions originate from osteoclasts in sesamoid bones, but the anatomical sites of the lesions (fingers, hips, knees, ankles) make this view untenable.

This study examined the biomechanics of anterior cruciate ligament-patellar tendon autografts for as long as 3 years after surgery. Twenty-seven adult female goats were tested; four served as controls and the others received an autograft to the right knee with each left knee serving as an additional control. The animals with grafts were tested at 0 week (n = 4), 6 weeks (n = 4), 12 weeks (n = 4), 24 weeks (n = 3), 1 year (n = 5), and 3 years (n = 3) after surgery. The anteroposterior laxity of the knee joint, load-relaxation, and structural and material properties of the graft were tested. The anteroposterior laxity was significantly greater than that of the controls for all groups except at 3 years. Load-relaxation was greater than that of the control anterior cruciate ligaments, but in the 1 and 3-year grafts load-relaxation was less than that of the patellar tendons with 5 minutes of sustained loading. Between 12 and 52 weeks, the stiffness and modulus of the grafts increased 3-fold, but the improvement was slow afterward. At 3 years, the strength and stiffness of the grafts were 44 and 49% those of the control ligaments, respectively; the modulus was 37 and 46% that of the control anterior cruciate ligaments and patellar tendons, respectively. The persistent inferior mechanical performance at 3 years suggests that anterior cruciate ligament grafts may never attain normal strength.

Summary: This study examined the change in type-I11 collagen concentration and hydroxypyridinium crosslink density of anlerior cruciate ligament-patellar tendon autografts and their correlations with Young's modulus of the anterior cruciate autografts and anterior cruciate controls for as long as 3 years after surgery. Fifteen adult female goats (two control and 13 experimental) were tested. Each experimental animal received an anterior cruciate ligament-patellar tendon autograft to the right knee. These animals were tested at 0 (n = 2), 6 (n = 2), 12 ( n = 2). and 24 (n = 1) weeks and 1 (n = 3) and 3 (n = 3) years after surgery. After mechanical testing. the anterior cruciate autograft and control tissues were analysed for type-I11 collagen concentration and hydroxypyridinium crosslink density. The results of sodium dodecyl sulfate gel electrophoresis showed a trend of initial increase in the percentage of type-I11 collagen in the anterior cruciate ligament autografts and a subsequent decrease after 24 weeks following surgery. There was a nonsignificant (p > 0.05) negative correlation between type-111 collagen concentration and Young's modulus. The hydroxypyridinium crosslink density was highest at 1 year after surgery. There was a significant (p < 0.05) positive correlation between hydroxypyridinium crosslink density and Young's modulus in the anterior cruciate autografts and controls. This suggests that hydroxypyridinium crosslink density has a good linear relationship with the material strength of the anterior cruciate ligament autograft and hence could be used as an objective guide for rehabilitation with anterior cruciate autografts.Reconstruction of the anterior cruciate ligament with the patellar tendon is a common surgical procedure for anterior cruciate ligament rupture (8.9.18). Biomechanical studies of anterior cruciate ligamentpatellar tendon autografts have generally demonstrated low mechanical strength in these grafts with various animal model? as long as 1 year after surgery (4.7.17). Recently. we reported that goat anterior cruciate ligament-patellar tendon autografts were significantly weaker than the control anterior cruciate ligaments even at 3 years after surgery (21).The mechanical performance of anterior cruciate autografts may largely be determined by their biological changes such as collagen remodelling, which happens rapidly after surgery. as shown in rats (19). Amiel et al. (2) reported histological and biochemical trans-

To test the healing of the partially torn anterior cruciate ligament, we transected the posterolateral bundle in 11 adult female goats and tested the ligaments at 12, 24, and 52 weeks and 3 years after surgery. As early as 12 weeks after surgery translucent fibrous tissue covered the wound. The differences in anteroposterior laxity between right and left knees measured at 45 degrees and 90 degrees of flexion were not significantly different at each period. Results of Instron testing of the posterolateral bundle revealed the normalized changes in load-relaxation and Young's modulus were not significantly different at each period, but the ultimate tensile strength and stiffness at 3 years were significantly higher than at 12 weeks (P < 0.05) Failure started at the repair site for the 12-week group, but at 24 and 52 weeks the failure occurred throughout the ligament. At 3 years, the specimens failed with bony avulsion, indicating the repaired tissue was not the weakest link of the bone-ligament-bone complex. This study shows that under favorable conditions, partial anterior cruciate ligament injuries are capable of repair. What is more important, the high ultimate tensile strength and stiffness of the 3-year repaired tissue indicate full structural repair of such an artificial transection injury may be possible.