This is a retrospective study. The patients who met the inclusions were taken into account. The study had ethical approval of ethics committee of local hospital (No. G2016-020-1) and all patients gave informed consents.
The inclusion criteria were: unilateral recurrent patella dislocation (more than there dislocations or dislocation still occurs after three-month conservative treatments) and the contralateral knee with no symptoms of patella instability; the patients were under 12 years old and with open growth plate (the ossification of patella completes by age 13-16 years)[19]; patella with type II and III as defined by Wiberg in the dislocated knee[20]; medial patellar retinacular plasty combined with lateral retinaculum release or not was performed unilaterally.
The exclusion criteria were: cases with closed physes; concomitant cruciate ligament or collateral ligament injury; rheumatoid arthritis or osteonecrosis with cartilage damage of greater than grade II[21]; Other surgical procedures such as medial patellofemoral ligament reconstruction, tibial tubercle transfer and femoral trochlear osteotomy; Patellar redislocation in the follow-up.
At last, a total of 22 patients (age 7-12) who were treated between October 2014 and August 2018 were enrolled. Ten patients had traumatic patellar dislocation. And the mean time from injury to surgery was 2.2 ± 0.7 months. The period from pre-op CT scans and surgery was 0.9 ± 0.6 months. The knees with recurrent dislocation were treated with medial patellar retinacular plasty and were defined as the affected side. The contralateral knees were considered as the unaffected side.
Surgical techniques
The surgeries were completed by one experienced senior surgeon. Before performing medial patellar retinacular plasty, arthoscopic assessment was performed and chondral lesions and concomitant pathology were treated. A force-directed medial shift of the patella was performed and the movement was less than one quarter the width of the patella indicated overtension of the lateral retinaculum. The lateral retinacular release was undertaken in these patients[22].
The plasty procedures had been illustrated by Fu and Ji[18, 23]. The procedures vary from the position of the injury. For the injury of the medial patellar retinaculum near the patellar attachment or midportion, the junction of the vastus medialis obliquus and the medial patellar retinaculum were incised transversely. At 30 degrees of knee flexion, the medial patellar retinaculum was pulled proximally and sutured to the medial proximal border of the patella with PDS-I whipstitch, then the vastus medialis oblique muscle fibre was pulled distally and sutured to the medial distal border of the patella over the retinaculum.
For the injury of the medial patellar retinaculum near the femoral attachment, a transverse incision was made at the injured inferior border of medial patellar retinaculum and the two parts were clamped separately. The distal part was pulled to the proximal border of the adductor tubercle while the proximal part was pulled to the distal border of the adductor. Then the two parts were sutured with PDS-I whipstitch. Figure. 1 is for the patients with injury both near patellar site and femoral attachment of patellar retinaculum.
Postoperative management
On the second day after surgery, leg raising and quadriceps isometric strength exercises were performed. One week after surgery, a gradual range of motion was initiated with a hinged brace. Four weeks after operation, knee flexion with 90 degrees was achieved. At six weeks after surgery, the brace was removed and full weight-bearing was allowed without a crutch. Three months postoperatively, the patients are able to perform normal daily activities. Contact sports were permitted six months after the surgery.
Assessments
Knee function was evaluated using Kujala score before surgery and at the follow-up[24]. Physical examinations were assessed using the apprehension test for lateral stability of the patella. Knee joint CT scans in the axial plane were performed to measure patella tilt angle, patella congruence angle, trochlea sulcus angle and TT-TG distance preoperatively and at the final follow-up on both sides. In addition, patella tilt angle had be evaluated preoperatively and postoperatively on both sides.
For patellar morphology measurements (Figure. 2), the slice image with the widest diameter of the patella was used for the measuring in the axial plane. The line between the most medial point A and the most lateral point B of the patella was defined as the transverse diameter (Line AB)[25]. The posterior patellar point farthest from the Line AB was defined as point D. The thickness of the patella was measured by the length of line CD which is vertical to the Line AB. The insertion between line CD and line AB was defined as point E. The Wiberg angle (∠D) was measured, and Wiberg index (length of BE/length of AB) was calculated as description by Fucentese et al. [11]. The measurements were performed in RadiAnt-DICOM software (Medixant Ltd., Poznan, Poland), which provides an accuracy of 0.01 mm for length and 0.1° for angle. Intra-observer and interobserver consistency were performed.
Statistical analysis
Before the investigation, the sample size was estimated using Wiberg-index as the primary variable. On the basis of previous study [11], the standard deviation was assumed at 0.06 in both affected and unaffected sides. An estimated difference between the groups was set as 0.06. A power calculation was performed by G-power Software (3.1.9.6) with a confidence level of 95% (α = 0.05) and power (1-β) of 80%. An estimated sample size of 17 knees per group was yielded.
Statistical analysis was performed using the SPSS version 21.0 (SPSS, IL, USA). The normality and homogeneity of the data were tested by the Kolmogorov–Smirnov test and Levene’s test respectively. All numerical variables showed a normal distribution and equal variance. Student’s t test was used to test the difference of the numerical variables. A P value < 0.05 was determined as statistically significant. The results are expressed as mean ± standard deviation.
To determine the intra-observer variation, the observer A repeated the measurements two weeks after first observing. To determine the inter-observer variation, the measurements were measured by observer A, observer B and observer C. Intra-observer and interobserver consistency were analysed using the intra-class correlation coefficient (ICC). ICC > 0.75 was regarded as excellent, ICC 0.40 – 0.75 was fair to good, and ICC < 0.40 was poor.