In Langenskiöld’s original description , the patella ligament on the proximal tibia is detached. Subsequently, the patella ligament is placed inside the synovium cavity and removed out of it from a hole distal to the first. The insertion is then buried in the hole in the tibial metaphysis and fixed by sutures. In this study, the outcomes were satisfactory, and no patient reported recurrent lateral dislocation after an average follow-up of > 3 years. Although no genu recurvatum was reported in the original study, in patients with open physes, surgery in the region of tibial tubercle can result in premature closure of the physis [11, 12]. In the modified Langenskiöld procedure we used, the tibial tubercle was intact, and the Q angle was decreased. In the present study, no growth disturbances occurred. Early weight bearing and early movement were possible. In contrast, bony distal re-alignment procedures bear the risk of non-union and require longer rehabilitation. In addition to this, the modified procedure left the patella and the tendon extra-synovial, which is similar to the physiological environment.
The “four in one” procedure includes lateral release, proximal “tube” realignment of the patella, semitendinosus tenodesis, and transfer of the patellar tendon [6, 7, 13]. Semitendinosus tenodesis is performed with the semitendinosus tendon harvested and routed through an osseous tunnel in the patella. In our procedure, the patella was removed from the synovium and sutured in a medialized position of the synovium. It was stabilized in the synovial membrane, and the osseous tunnel in patella was not needed. Our procedure was much simpler and the patella was intact. The synovium was remarkably strong and inelastic, and the outcomes in our series were satisfactory.
Camathias reported a re-dislocation rate of 80% after the Stanisavljevic quadriceps transposition . In the 11 patients (16 knees) in this study, no one reported recurrent dislocations and pain during the activities of daily life after an average follow-up of 3 years. The synovium is inelastic and strong enough to hold the patella. In our study, the mean postoperative Kujala score was 95 (SD 5.9; 86–100), and the Lysholm score was 94.8 (SD 5.1; 87–100). In 2012, Efe et.al. reported a 22% re-dislocation rate, the Kujala score was 85.0 (SD 14; 51–100) treated with Insall’s proximal patellar realignment procedure . The increased Q angle is a risk factor for patellar instability [15,16,17]. To decrease the Q angle, which is significantly increased after the patella is medially relocated, we medialized the patellar ligament with the Grammont procedure, which is thought to be more suitable for skeletally immature patients. Genu valgum and knee joint hyperlaxity are also risk factors for patellar instability [16, 18, 19]. In this study, there were nine patients with knee joint hyperlaxity. They all received EPCLR, and a more stable joint was achieved after the procedure. In patients #2, #3, and #6, genu valgum was observed before the operation. Physical examinations and full-length lower extremity X rays indicated that joint hyperlaxity was the main cause of genu valgum in these patients, and distal femur deformity was the main cause in patient #6. Genu valgum improved significantly after EPCLR in these patients. In patient #6, it was not improved after EPCLR because of bone deformity caused by achondroplasia, so we performed hemi-epiphysiodesis of the medial distal part of the femur. At the latest follow-up, genu valgum was significantly improved.
Sever et al.  reported the outcomes of 12 patients (15 knees) with congenital or obligatory patellar dislocation treated with the Stanisavljevic quadriceps mechanism realignment procedure, which includes a proximal extensive subperiosteal realignment of the quadriceps mechanism, medial plication using the large overstretched medial capsule as a cover to the realigned patella, and an additional distal Roux-Goldthwait patellar tendon realignment procedure. Postoperative knee active extension was improved significantly for all patients, and no significant change in the flexion range was observed. Gao et al. reported that all the patients had full range of movement without extensor lag but 12.2% of the patients experienced flexion limitations after surgery. We had three patients (four knees) with knee flexion contracture. They had a limitation of extension by 15° to 25°. Postoperative knee extension was improved significantly for three knees after releasing the posterior capsule, and one knee needed further peroneal nerve decompression, followed by release of the biceps tendon.
The flexion of the knee was limited to 123° in one patient, which was decreased by 15° postoperatively. The medialization of the patella and tendon lead to a tighter rectus femoris [16, 19]. If knee flexion was limited significantly during the operation, then the rectus femoris was released in case of flexion contracture after the operation. Two of the 16 knees in our research performed a release of the rectus femoris tendon. There were another two cases of superficial wound infection due to incision dehiscence. They were resolved after treatment with oral antibiotics and dressing changes. An extensive separation can disrupt the blood supply of the flap, and this might have been one reason for incision dehiscence.
The present study had some limitations. The follow-up of 37.8 months was relatively short compared with similar studies [2, 8, 11, 21, 22]. The small number of patients was another limitation. We will continue the follow-up and enroll more patients in the future.