This study is a randomized control trial. All patients are diagnosed as lateral patellar compression syndrome depending on clinical features, MRI (Figure 1) and diagnostic arthroscopy. All patient had failed conservative measures of quadriceps strengthening exercise and analgesics for 6 months. All patients were followed by Lysholm knee scoring scale before surgery, 2 weeks, 6 weeks, 6 months, 12 months and 24 months after surgery with follow up of 2 years for functional outcome and other outcome measures involving duration of operation, length of hospital stay, intra-operative complications and post-operative complications of bleeding and infection, recurrence and medial patellar instability.
The study was carried out in a tertiary orthopedic center.
Period of study:
The study conducted between March, 2016 to November, 2019 with 2 years of follow up.
Study sample and randomization method:
Eighty patients were involved in this study, and it was divided into two groups randomly by entering the names of the patients into an excel file and by computer system, the patients were arranged randomly in a list, then patients with odd number sequences were regarded as group A and patients with even number sequences were regarded as group B. Group A (40 patients) were treated with open release after diagnostic arthroscopy and Group B (40 patients) were treated by arthroscopic release.
Inclusion criteria involve patients presented with lateral patellar compression syndrome that is proved by the following criteria and failed to conservative treatment of quadriceps strengthening exercise and non-steroidal anti-inflammatory analgesics for 6 months. [7,19]
A. Clinical features:
1/ Maximal pain and tenderness over the lateral margin of the patella. [20,23]
2/ Abnormal patellar tilt test: when the patella can’t be lifted from the lateral femoral condyle with extended knee by the examiner. [20,23]
3/ Abnormal medial patellar glide test: when the patella can’t be shifted by one or more quadrants medially by the examiner with knee flexion 10 degrees). [20,23]
B. MRI features:
1/ Patellar translation relative to the femur usually occurs more laterally than medially. Subluxation/translation is measured as the distance between perpendicular lines drawn on an axial image; one from the medial edge of the patella and another one through the most anterior part of the medial femoral condyle. A 2 mm distance is the upper accepted limit of normal.
2/ Abnormal patellar tilt, which may present with or without patellar translation, is the most closely related radiologically to lateral patellar compression syndrome. The patellofemoral angle is measured at the level of the patellar midpoint using the same method that was used on plain radiograph on sagittal imaging. It should measure more than 8° and opens laterally, if less than 8° or opens medially; it is considered abnormal. 
C. Arthroscopy. Before doing lateral release, all patients in both groups were assessed by arthroscopy to see how the patella touching the lateral femoral condyle more than medial femoral condyle with knee movement in flexion and extension as well as exclude other pathologies.
Exclusion criteria include:
2.Patellar instability: patient has medial or lateral glide test of 3 or more quadrants or history of patellar dislocation.
4.Ligament hyperlaxity based on Beighton’s criteria. 
5.Pathological femoral anteversion or tibial torsion by Staheli’test. [20,26]
6.Q-angle more than 20 degrees. [20,27]
7.Knee osteoarthritis or Patellofemoral osteoarthritis more than stage I. 
8.Previous knee surgery or infection.
9.Outerbridge Grade 3 and 4 chodropathy.
10.Patellat alta or trochlear dysplasia. 
We used the Lysholm Knee Scoring Scale to assess all patients pre-operatively and at 2 weeks, 6 weeks, 6 months, 12 months and 24 months post-operatively. Eight sections are assessed to produce an overall score on a scale of 0 to 100. Then an assignment is given as “excellent” for 95 to 100 points; “good” for 85 to 94 points, “fair” for 65 to 84 points, or “poor” for less than 65 points. 
The study was approved by the local ethical committee in the university. Both verbal and written informed consents were taken from each patient prior to participation in the study.
Under general or spinal anesthesia, patient was in supine position. Pneumatic tourniquet was applied on the upper thigh with leg holder. Through anterolateral and anteromedial portals, Diagnostic arthroscopy was done in all patients (both groups A and B) and checking of all compartments of the knee was done. The patellotrochlear engagement was assessed especially at 30-40 degrees of knee flexion.
In group A, open release of the lateral patellar retinaculum was done through 3 centimeters incision on the lateral side of the patella and the lateral retinaculum was cut longitudinally about 2 centimeters length and the wound was closed with drain after deflating the tourniquet and securing the hemostasis.
In group B, release of the tight lateral retinaculum was done arthroscopically by using a hook knife and electrocautery (Figure 3) with continuous monitoring of the patellotrochlear movement during knee flexion and extension to avoid excessive release.
Postoperative care and follow up
Knee exercise started as soon as pain was tolerated in the same day with gradual weight bearing as tolerated. Most of the patients were discharged home in the same day in group B while in the next day in group A after removal of the drain. Wound stitches were removed after 14 days postoperatively.
All patients were assessed by Lysholm knee scoring scale at 2 weeks, 6 weeks, 6 months, 12 months and 24 months after surgery.
We have no patient loss on follow up for 2 years.
Statistical Data analysis
Statistical analysis was carried out using SPSS version 21. Categorical variables were presented as frequencies and percentages. Continuous variables were presented as (Means ± SD). Student t-test was used to compare means between two groups. Paired t-test was used to compare means for paired reading. Pearson’s chi square (χ2) was used to find the association between categorical variables. A p-value of ≤ 0.05 was considered as significant.