Return to sports, dependent on the type and previous level of sports activity, gradual increase of training volume and intensity
The surgical treatment is performed in a tailored manner, addressing the respecting pathologic anatomy that predisposes to LPD. MPFL reconstruction is performed in every patient. Other surgical techniques listed below will be applied in individual combinations, dependent on patient’s needs.
Reconstruction of the medial patellofemoral ligament (MPFL) is a proven technique for LPD and today’s established standard treatment. However, some authors have reported a considerable complication rate (10). Many failures were reported due to inappropriate indications. The latter means performing isolated MPFL in patients with coexisting severe osseous pathologies like high-grade trochlear dysplasia or a pathologic tuberositas-tibiae-trochlea-groove distance (TT-TG distance) (11, 12). Isolated MPFL reconstruction is regarded as inappropriate in patients with: 1) TT-TG distance > 20mm, 2) femoral anteversion ≥ 40° (n. Waidelich/Strecker), 3) high grade trochlea dysplasia, 4) severe patella alta, 5) tibiofemoral valgus > 5°. With accurate indications and surgical technique isolated MPFL reconstruction provides good outcome in patients with LPD (13, 14). MPFL reconstruction is applied in all patients of the surgical group. MPFL reconstruction was reported with a high variety of surgical techniques (graft type, single vs. double bundle, type of fixation etc). The specific surgical technique is carried out on surgeon’s preference at the respective center of the multicentric study.
Trochleoplasty
When the trochlea is flat or convex (dysplasia Dejour type B, C or D) a “deepening trochleoplasty” should be considered. The aim of the trochleoplasty is to A) reduce the too prominent anterior bone stock and B) create better conformity with the patella (concave groove) and a lateral trochlea facet as restraint against the lateralizing quadriceps pull. Many authors have reported that trochleoplasty leads to good clinical outcome in patients’ LPD due to a dysplastic femoral trochlea (15–22).
Deepening trochleoplasty will be carried out in those patients of the surgical group who suffer from high-grade trochlea dysplasia.
Tibial Tuberosity Transfer
The most popular type of osteotomy in the setting of LPD is certainly the osteotomy and transfer of the tibial tuberosity (TTT). Many articles reported good clinical success for medialising TTT in patients with LPD and high TT-TG values (23–27). Similarly, good results were found for distalizing TTT in patients with LPD and patella alta (28, 29). TTT can be tailored to the pathology of the patient by performing combined medialization and distalization.
Medializing TTT will be applied to those patients with TT-TG distances ≥ 18mm in the MRI. Distalizing TTT will be applied in patients with Caton-Deschamps Index > 1,2 (30).
Derotational Osteotomy
In patients with valgus clinical appearance a weight-bearing whole leg radiograph should be performed to precisely assess the degree of the deformity in the frontal plane (mechanical femorotibial angle).
In cases with a mechanical femorotibial angle > 5 degrees a varus osteotomy is performed at the location of the deformity.
Applying a “pragmatic” surgical approach, not each single pathology in the patient’s anatomy is addressed. Instead, a maximum of 3 surgical techniques (including the MPFL reconstruction) are performed in one patient.
Outcome parameters
Patient-reported outcome
The Banff Patellofemoral Instability-Instrument (BPII) 2.0 was reported as valid, reliable and responsive patient-reported outcome tool in the field of patellofemoral instability (38, 39) and is used in the validated German version (40). The BPII 2.0 serves as one of two major outcome instruments (Hypothesis 1).
For exploratory reasons the following further patient-reported parameters will be assessed in both groups: As second disease specific score patients accomplish the Kujala Score (41) which was quoted as reliable, valid and responsive tool for patellofemoral disorders (42, 43). In addition, the Short-Form 12 is used (version 2, German; SF-12v2) (44) to determine the general health outcome and the Marx activity scale to rate a patient’s physical activity (45). The Marx score asks for the highest activity in the last year. For postoperative monitoring of a patient’s activity a “modified version of the Marx score” will be used that refers to the last 2 months.
All above-mentioned outcome scores are self-administered and will be assessed preoperatively, 6, 12 months postoperatively and then yearly. Those scores are collected during routine visits at the hospital.
Other outcome parameters
Recurrence rate is assessed as second major outcome parameter (Hypothesis 2). To keep proper medical records on recurrent patella dislocations the patients are interviewed by telephone on a monthly basis (in addition to the above-mentioned visits at the hospital).
The apprehension test is assessed by an experienced observer during the above-mentioned routine clinical visits (Grade 0: no evasion, Grade 1: slight evasion/avoidance, Grade 2: gross evasion/avoidance, Grade 3: patient too anxious to allow the test).
Joint degeneration is assessed preoperatively and every three years postoperatively by means of MRI (PD-FSE with Fat-sat high-resolution in all three planes / T1-TSE, sagittal / T2 weighted, isotrope 3D sequence sagittal reformatted in all three planes). The semi-quantitative MRI Osteoarthritis Knee Score (MOAKS scoring) is applied to rate the degenerative changes determined by MRI (46). The MOAKS scoring is determined by always the same experienced musculoskeletal radiologist.
In addition, the Patella Instability Severity Score is assessed for exploratory reasons (47).
Statistics
Patient characteristics will be presented as means, standard deviations, and percentages. The main analysis will use linear mixed models that allow data modelling with a varying number of assessments per patient and time-varying covariates. Such a model will be used to compare the differences in changes over time between the two study groups. The following terms will be included in the model: a random baseline, a first-order autocorrelation covariance matrix, a fixed-effect patient group, a fixed effect time point, and the group-by-time interaction (reflecting the intervention effect). The BPII 2.0 will serve as the primary outcome parameters. The main analysis is “intention to treat” and will focus on the group difference in the change of BPII 2.0 scores between pre-op and 2-year follow-up. The above-mentioned secondary outcome parameters will be analysed with the same model.
Power / sample size analysis was done for the comparison of the two study arms at 2-year follow-up with regard to the BPII score. As there are no specific minimal important difference (MID) for the BPII available from the literature, we defined the MID to be 0.5 standard deviations following general recommendations from the literature (48). A sample size of N = 64 per study arm (128 overall) provides 80% power (alpha = 0.05, two-tailed) to detect a difference of 0.5 standard deviations in a t-test for independent samples. To account for 20% attrition during the study period we plan to recruit 80 patients per study group (160 overall) at baseline.