Purpose: We reviewed retrospectively 40 ligamentoplasties of the anterior cruciate ligament in patients who had undergone several prior reconstructions (1 to 6). The initial operation had used a synthetic ligament in 13 cases, patellar tendon in 23 and divers implants in four.

Material and methods: Forty patients, 12 women and 28 men, with recurrent ligament tears where reoperated by the same surgeon. Mean age was 28.5 years (16–51). Mean follow-up after the last operation was 27 months. All patients were reviewed (history, physical examinatin and KT 1000); 20 of them responded to a self-administered questionnaire using the IKDC chart later after the clinical review, and 23 of the 40 had radiographs to measure laxity. Arthroscopic reconstruction was used in 33 cases. A surgical procedure was necessary on a peripheral ligament in six cases and osteotomy for tibial valgisation in one. Reconstruction was achieved with the quadriceps tendon in 11 cases, the patellar tendon in 18, the hamstrings in eight and Mac in Jones in three. The IKDC score was used for clinical assessment with manual arthrometric measurement of laxity. Radiographs were obtained to measure the anterior drawer of the medial and lateral compartments of both knees for right-left and pre-postoperative comparisons.

Results: The initial handicap was marked (IKDC: 18 D, 21 C, 1 B). Reconstruction was good or very good in 72.5% of the knees (IKCD: 2 A, 27 B, 10 C, 1 D). Self-evaluation revealed 25% painful knees for intense activities. Laxity was improved with a mean differential gain of 5.35 mm of the maximal pre- and postoperative KT 100 (7.24 versus 1.89). Among the 24 knees with radiographic assessment, the mean differential preoperative was 9.14 mm preoperatively and 4.69 mm at last follow-up giving a gain of 4.45 mm. One athlete was able to resume sports activities at the same or higher level and others at a lower level.

Discussion, conclusion: This study confirms that the results obtained after revision repair are less satisfactory than after primary repair. This homogeneous (one operator) and large series with a sufficient follow-up can be compared with the rare published series. Each ligament reconstruction is specific and warrants a specific surgical approach adapted to each individual case.

Purpose: There are three main causes of failure after valgisation osteotomy of the tibia requiring repeated osteotomies: insufficient valgus, excessive valgus, or loss of the valisation correction after a variable delay. The purpose of this study was to evaluate outcome after repeated oseotomies performed in relatively young patients or too active to propose arthroplasty. The technical problems were different for each aetiology.

Material and methods: The series included 47 knees operated on between 1974 and 1998 after a first osteotomy performed at a mean age of 46 years. Mean delay between the two operations was five years (1 to 12). A medial closure osteotomy had been performed at the first operation in 34 cases and a lateral opening osteotomy in 13. For the 19 knees with valgus, the second osteotomy was a medial closure in 14 and a lateral opening in five. A repeat valgisation was performed in 28 cases, 18 by lateral closure, one by medial opening and nine by curviplanar osteotomy. The IKS score was determined to assess function. The femoraotibial axes (HKA angle) were determined on full stance views. The Ahl-back osteoarthritis grading was used. For 17 patients who had undergone operations in other institutions, exact measurements were not always available concerning the preoperative status and the initial correction.

Results: The overall IKS score for function improved in 87% of the cases with a mean follow-up of five years. The IKS knee score improved from 73 to 89 points and the IKS function score from 65 to 81 points. For the 19 over-corrections, the mean HKA angle was changed from 190° to 184°. For the 28 under-corrections, the mean HKA angle was changed from 173° to 182°. The tibial tilt remained unchanged at 7° as did lateral gapping at 3°. Delay to consolidation was a mean 96 days.

Discussion: Revision osteotomies performed for correction defects should be distinguished. For these procedures, it would be logical to expect a good result if a 3 to 5 degree valgus is achieved. Revisions after a long period (33 cases) are different; required for wear, these cases correspond to progressive loss of the initial osteotomy effect. These patients are often candidates for prosthesis if seen after 70 years. Good results can however be obtained with a second osteotomy irrespective of the initial technique. We prefer reoperating with medial opening after initial lateral closure.