This chapter describes the surgical procedure for medial patellofemoral ligament (MPFL) reconstruction in skeletally immature patients using physeal-sparing femoral tunnel placement and pull-through of hamstring graft through the femoral tunnel (pull-through technique).
The importance of distal femoral physis and its relationship to the MPFL femoral attachment point is emphasized.
Fithian et al defined the highest risk population for patellar dislocation as females aged 10 to 17 years.1
Historically, the rate of patellar redislocation or persistent symptoms after first-time dislocation has been reported in the range of 17% to 44%.1,2,3,4
Lately, attempts have been made to predict recurrence after first-time dislocation to help in optimal management of patients. These prediction models are based on assessment of demographic and anatomic risk factors.5,6
The identified risk factors for recurrent instability after first-time patellar dislocation include age less than 14 years, skeletal immaturity, sports-related patellar dislocation, history of contralateral patellar dislocation, patella alta, patellar tilt, trochlear dysplasia, increased tibial tubercle-trochlear groove (TT-TG) distance, and hyperlaxity7,8,9 (Table 27.1).
Lewallen et al reported that skeletally immature patients with trochlear dysplasia had a 69% rate of recurrent instability after first-time dislocation.8
Ossification of the patella begins between age 3 and 5 years. This ossification may appear irregular but rapidly expands in a proximal to distal fashion. The ossification process is completed by the age of 16 to 17 years. Because of the delayed ossification of the distal aspect of the patella, patellar height calculations on radiographs may be spuriously elevated in children10 (Figure 27.1).
Caton-Deschamps index has been evaluated in children and has been found to be a simple and reliable measurement for patellar height, though age-based values have to be considered. The mean age-based values are 1.27 ± 0.25 (6-7 years), 1.15 ± 0.17 (8-9 years), 1.11 ± 0.19 (10-11 years), and 1.07 ± 0.18 by 12 years.11 A value greater than 1.45 (2 standard deviations above normal) was considered to be a prognostic risk factor for recurrent patellar dislocation.7
A lateralized TT has been recognized as an important factor for patellar instability. The lateralization is measured by the TT-TG distance on axial MRI or CT study.
TABLE 27.1 Predictors of Recurrent Dislocation After First-Time Patellar Dislocation
No. of Risk Factors
Predicted Risk of Recurrence (%)
History of contralateral dislocation
Can consider surgical stabilization
Can consider surgical stabilization
The average TT-TG distance in adults with and without patellar instability has been reported as 19.8 ± 1.6 mm and 12.7 ± 3.4 mm, respectively.9 Values more than 20 mm have been used as an indication for TT medializing osteotomy, with a goal to restore the TT-TG distance between 10 and 15 mm.
As TT-TG is an absolute number, it does not account for the size of the knee or age of the patient. Studies have been done to develop TT-TG ratio or index to overcome this drawback.
TT-TG distance in children shows logarithmic progression with age, with values approaching adult values as children approach skeletal maturity. In normal children (without patellar instability), the mean TT-TG distance has been reported as 8.6 ± 0.3 mm at an average age of 11 years and 11 months. In children with patellar instability, the mean TT-TG distance has been reported as 12.2 ± 1.1 mm at an average age 13 years and 5 months. Thus, younger the patient, less is the threshold value for correction of TT-TG distance.12
On a lateral radiographic view, one of the three signs for trochlear dysplasia (crossing sign, supratrochlear bump, or double contour sign) was present in all skeletally immature patients with trochlear dysplasia.13
During growth, all linear measurements of trochlear dysplasia (condylar height, trochlear height, trochlear bump) increased with age. However, the shape of the TG (sulcus angle) did not change much with increasing age.14 Progressive ossification of trochlea may give a radiographic appearance of deepening of TG (decrease in sulcus angle) with skeletal growth; this is spurious finding.15
The TG can remodel following patellar stabilization, though the remodeling capacity decreases with age. After 10 years of age, there is no significant remodeling of the trochlea.16 Thus, if possible, earlier patellar stabilization should be offered to children under the age of 10 years.
Trochlear dysplasia is one of the main risk factors for patellar instability and failure of patellar stabilization surgery. In a group of 37 children with open physis, with mean age of 14 years and failed patellar stabilization surgery (lateral release, medial reefing, Roux-Goldthwait procedure), the authors found high-grade trochlear dysplasia in 89% of patients. Patella alta and increased TT-TG distance were not significant risk factors for failure.17
The spatial and causal relationship between the anterior aspect of distal femoral physis and supratrochlear bump is less understood. The mean distance between the anterior aspect of distal femoral physis and the proximal-most aspect of the trochlea is around 4.5 mm. However, the lateral aspect of the trochlea can be proximal to the physis (13.7% knees) or at the level of the physis (17.7% knees) (Figure 27.2).
Figure 27.2 MRI sagittal sequence through lateral aspect of knee shows the trochlear articular cartilage (dashed arrow) extending proximal to the anterior distal femoral physis (solid arrow).
The distal femoral physis has an undulating pattern. The central ridge and the medial and lateral “valleys” change shape during growth (Figure 27.3). As the epiphysis starts cupping the metaphysis, the “nadir” on the medial and the lateral side deepens and the central ridge decreases in relative height. The metaphyseal-epiphyseal junctional notches are higher than the central ridge at skeletal maturity.19
Topographic and radiographic evaluation of the distal femoral physis shows it to have a cruciform pattern when viewed from above.19 The central ridge (from anterior to posterior), lateral ridge, and medial peak (from medial to lateral) divide the physis into four valleys: anteromedial, anterolateral, posteromedial, and posterolateral (Figure 27.4). The most inferior aspect of the physis on the medial side is “anteromedial” and the most inferior aspect of the physis on the lateral side is “posterolateral”20 (Figure 27.5). These positions are important to take into consideration when surgery is performed around the distal femoral physis.
Based on pediatric cadaveric study, the MPFL attached to the proximal two-thirds of the patella. On average, the center of MPFL insertion footprint was 4.7 mm above the equator of the patella and it spanned a mean 41% of the length of the patella.21
On the femur, the MPFL attached posterosuperior to the medial epicondyle and just distal to the adductor tubercle. The medial epicondyle, adductor tubercle, and MPFL femoral insertion are distal to the physis in most children and adolescents (Figure 27.6). With growth, the MPFL insertion moves closer to the distal femoral physis and may cross proximal to the physis (Table 27.2).
TABLE 27.2 Relationship Between MPFL Femoral Insertion and Distal Femoral Physis
Average Distance Distal to Physis (Minus Indicates Proximal to Physis)
Shea et al (2010)39
−2.7 mm (female)
−4.6 mm (male)
Kepler et al (2011)40
Nelitz et al (2011)41
AP, lateral x-ray
Greenrod et al (2013)25
Distance increased 0.6 mm/y
Shea et al (2014)42
1 m, 11 m, 11 m, 8 y, 10 y, 11 y
<1 y: 9 mm Others: 4 mm
Farrow et al (2014)43
Adductor tubercle and medial epicondyle distal to physis
Shea et al (2018)44
3 mm (average for all specimens)
<7 y: 4.7 mm
>7 y: −0.8 mm
MPFL origin distal and posterior for those <7 y of age; proximal and anterior for those >7 y
Abbreviations: AP, anteroposterior; MPFL, medial patellofemoral ligament.
The MPFL femoral attachment point on a lateral radiograph has been well-described for adults by Schöttle et al.22 The same technique has been found to be reproducible in the skeletally immature.23 However, the femoral attachment point is in close relationship to the distal femoral physis and there is valid concern about physeal violation and growth disturbances during pediatric MPFL reconstruction. To be at a safe distance from the physis, the attachment point could be moved slightly distal; this could lead to changes in graft isometry. To maintain graft isometry, there have been recommendations to move the attachment point more anterior (Eric J. Wall, MD, personal communication) or even proximal to the physis24 (Figure 27.7).
Because the inferior-most part “nadir” of the medial femoral physis is usually lower than the MPFL femoral attachment point, a femoral tunnel parallel to the physis would likely cause physeal injury (Figure 27.8).
Based on MRI study, a 20-mm screw placed parallel to the physis violated the physis in 64% of patients. However, if the screw was directed 45° oblique distally, it was safe in 98% of patients25 (Figure 27.8).
If the femoral tunnel was angled less than 10° in the distal direction, it would violate the distal femoral
physis. If it was angled less than 10° anteriorly, it would violate the intercondylar notch. If it was angled greater than 20° distally and anteriorly, it would violate the joint cartilage surface.26
Thus, the optimal trajectory of the femoral tunnel should be angled 15° to 20° distally and anteriorly related to the distal femoral physis26 (Figures 27.6 and 27.8).
Injury to the distal femoral physis can lead to deformity or limb length discrepancy.27 In this regard, anterior cruciate ligament (ACL) surgery is safer than MPFL surgery because the trajectory of the femoral tunnel during ACL reconstruction is relatively perpendicular to the physis, whereas the trajectory of femoral tunnel during MPFL reconstruction is relatively parallel to the physis. Thus, technical errors during femoral tunnel placement for MPFL surgery could violate significantly more area of the distal femoral physis.
Several classification systems have been described for patellar dislocations based on clinical and radiographic findings. Most of these have been reported based on patellar instability in adults.
Parikh and Lykissas described different types of patellar instability in children and adolescents.28 The main purpose to differentiate varied patterns of instability was to help guide treatment as to which patients would benefit from an isolated MPFL reconstruction. The types of patellar instability and recommended treatment options are discussed in Chapter 3.
The indications and contraindications for physeal-sparing MPFL reconstruction are listed in Table 27.3.
A thorough understanding of the patient’s history surrounding the incident leading to patellar dislocation is warranted. It is necessary to differentiate between traumatic and atraumatic causes for patellar dislocation. Most direct traumatic dislocations have normal anatomy of the patellofemoral joint and minimal anatomic risk factors.
Evaluate for associated major risk factors that could make the patient susceptible to subsequent patellar dislocation, including history of contralateral dislocation,
family history, other joint issues, sports participation, and age at first dislocation.
TABLE 27.3 Indications and Contraindications for Physeal-Sparing MPFL Reconstruction
First-time patellar dislocation with osteochondral fracture that has to be addressed in skeletally immature patients
First-time patellar dislocation with persistent symptoms of instability in spite of adequate nonoperative management in skeletally immature patients
First-time patellar dislocation with multiple risk factors (controversial) in skeletally immature patients
Recurrent lateral patellar dislocations in skeletally immature patients
Habitual dislocation in extension in skeletally immature patients
Severe malalignment (rotational or coronal plane)
Habitual dislocation in flexion
Permanent dislocation of patella
Patellofemoral pain without instability
Inability to monitor and treat growth disturbances
Elicit the number of patellar instability episodes that have occurred and any symptoms in between instability episodes.