Relationship between Anatomical Risk Factors, Articular Cartilage Lesions, and Patient Outcomes Following Medial Patellofemoral Ligament Reconstruction

Abstract

Objective

The purpose of this study was (1) to determine which risk factors for patellar instability were associated with the presence of patellofemoral cartilage lesions and (2) to determine how cartilage lesion presence, size, and grade affect postoperative disease-specific quality of life.

Design

Preoperative, intraoperative, and postoperative demographic, anthropometric (body mass index, Beighton score, hip rotation), radiographic (crossover sign, trochlear bump), cartilage lesion morphology (presence, size, location, grade), and outcomes data (Banff Patella Instability Instrument 2.0 [BPII 2.0]) were prospectively collected from patients undergoing isolated medial patellofemoral ligament reconstruction. For all knees (n = 264), single and multivariable logistic regression was used to determine if any patellar instability risk factors affected the odds of having a cartilage lesion. In patients with unilateral symptoms (n = 121), single variable linear regression was used to determine if the presence, size, or ICRS (International Cartilage Regeneration & Joint Preservation Society) grade of cartilage lesions could predict the 12 or 24+ month postoperative BPII 2.0 score.

Results

A total of 84.5% of knees had patellofemoral cartilage lesions (88.3% involved the distal-medial patella). Trochlear dysplasia (high grade: odds ratio = 15.7, P < 0.001; low grade: odds ratio = 2.9, P = 0.015) was associated with the presence of a cartilage lesion. The presence, size, and grade of cartilage lesions were not associated with 12 or 24+ month postoperative BPII 2.0 scores.

Conclusions

Trochlear dysplasia was a risk factor for the development of patellofemoral cartilage lesions in this patient population. Cartilage lesions most commonly involve the distal-medial patella. There was no significant relationship between patellofemoral cartilage lesion presence, size, or grade and postoperative BPII 2.0 scores in short-term follow-up.

Introduction

Cartilage lesions of the patellofemoral (PF) joint are a common pathology observed in patients with patellofemoral instability (PI),^1-11 and it is suggested that their presence is associated with decreased quality of life (QOL) and inferior surgical outcomes.^{1,4,7,8,12-14} Additionally, a history of PI has been associated with an increased risk of developing early PF osteoarthritis (OA)^11,15-17 and chronic anterior knee pain (AKP). ¹³ The risk factors for the development of PI are multifactorial and have been documented extensively.^3,7,18-24 Some of these risk factors include trochlear dysplasia,^3,19,21,25 patella alta, patellar tilt, increased tibial tubercle-trochlear groove distance, hip rotational abnormalities,^18,21,25,26 and generalized ligamentous laxity.^7,20-24 While there is evidence that these risk factors are associated with the development of PI, their contribution to the development of PF cartilage lesions remains unknown. Additionally, it is unclear how the presence, size, and grade of these lesions affect patient-reported QOL.

While previous studies have described trochlear dysplasia^14,27-29 and hip rotational abnormalities ³⁰ as potential risk factors for the development of cartilage lesions, there have been no reported comparative analyses between these or other potential risk factors in a population of patients with recurrent PI. Without a clear understanding of the individual risk factors for developing cartilage lesions, and how these lesions affect QOL, it is challenging for physicians to implement effective treatment plans for these patients. Until recently, there has been no disease-specific assessment for collecting patient-reported outcome measures in patients with PI. Several studies report the use of non–disease-specific assessments to analyze outcome measures in this population. The Kujala score, ³¹ a measure developed to assess AKP, is among the most commonly utilized. With the development and validation of the Banff Patella Instability Instrument (BPII) and subsequent BPII 2.0, a disease-specific patient-reported QOL measure is now available to evaluate the clinical influence of PI and associated cartilage lesions.^32,33

Understanding how the presence of demographic and anatomic risk factors affects the development of PF cartilage lesions and, subsequently, how these cartilage lesions influence disease-specific QOL can help clinicians direct patient counseling about short-term management options and long-term prognosis. The purpose of this study was 2-fold: (1) to determine which risk factors for PI were associated with the presence of PF cartilage lesions and (2) to determine how the presence, grade, location, and size of these cartilage lesions influenced the disease-specific QOL in patients following isolated open medial patellofemoral ligament (MPFL) reconstruction for recurrent lateral PI. The study hypotheses were the following: (1) the presence of risk factors for PI would be associated with the presence of PF cartilage lesions and (2) postoperative disease-specific QOL would be affected by the presence and extent of PF cartilage lesions.

Methods

A retrospective cohort study was performed utilizing prospectively collected registry data from the practice of a single sports medicine fellowship-trained surgeon (LAH) with a clinical focus in PF disease. The current study included male and female patients between the ages of 9 and 50 years with symptomatic recurrent lateral PI who underwent an isolated primary MPFL reconstruction (with or without previous or concurrent lateral release). The decision to operate was based on appropriate history and corroborating physical examination. AKP was not an indication for surgery for any patient in this cohort. Patients often struggle to accurately describe the differences in subluxation and dislocation; therefore, surgical decisions are made based on reports of “symptomatic lateral patellar instability” with a corroborating physical exam. Patients with tibial tubercle-trochlear groove (TT-TG) distances of >20 mm or patella alta with a Caton Deschamps ratio of >1.2 who underwent a concomitant tibial tubercle osteotomy were excluded. Patients with high-grade trochlear dysplasia (Dejour B, C or D) were considered for a concomitant trochleoplasty and were excluded if they underwent this procedure. Additional study exclusions were as follows: chronic knee pain, Worker’s compensation, current or pending associated litigation, congenital patellar dislocation, neurological disorders, systemic inflammatory disease (e.g., rheumatoid arthritis), previous patellar stabilization procedures, previous tibial tubercle osteotomy, previous knee ligament surgery (e.g., ACL [anterior cruciate ligament] reconstruction). Ethical approval was granted by the University of Michigan medical institutional review board under a data use agreement with Banff Sport Medicine.

Data were prospectively collected as a component of the primary surgeon’s standard clinical practice. Outcomes of interest included gender, age at first patellar dislocation, age at time of surgery, number of self-reported dislocations, WARPS (weak, atraumatic, risky anatomy, pain, subluxation)/STAID (strong, traumatic, anatomy normal, instability, dislocation) score, ³⁴ presence of bilateral symptoms, body mass index (BMI), involved limb (right/left), radiographic risk factors for patellar instability (supratrochlear bump, crossing sign, Dejour classification of trochlear dysplasia ³ ), presence of anatomical risk factors for patellar instability (Beighton score, excessive knee hyperextension, excessive femoral internal rotation), presence of a knee cartilage lesion at the time of surgery, International Cartilage Regeneration & Joint Preservation Society (ICRS) grade/location/size of identified knee cartilage lesions (https://cartilage.org/society/publications/icrs-score/), and pre- and postsurgery (12 and 24+ months) disease-specific QOL outcome score (BPII 2.0^32,33). Trochlear dysplasia was graded as absent, low grade (Dejour A), or high grade (Dejour B, C, or D) on a lateral radiograph. Patients with a Beighton score of ≥4 were considered hypermobile. Knee hyperextension of ≥10° was considered positive. Femoral internal rotation was measured with the patient supine with the hip and knee flexed to 90°. It was considered excessive if the internal rotation was more than 70° and 30° greater than external rotation.

Statistical Analysis

All outcomes were descriptively summarized with mean, median, standard deviation (SD), and interquartile range for continuous data and frequency tabulations for categorical and dichotomous data. The location of cartilage damage was intraoperatively mapped to 30 possible regions covering the patella (9), trochlea (3), medial (9), and lateral (9) femoral condyles as per the ICRS documentation maps. The frequency of damage to each section was tabulated to identify the distribution of damage.

Univariable and multivariable logistic regression was used to determine the ability of the variables of risky anatomy, patient anthropometrics, and disease status (time from first dislocation to surgery, number of dislocations, etc.) to predict the presence of a knee cartilage lesion. Following univariable analysis, all variables with p ≤ 0.1 were entered into a multivariable regression using backwards stepwise regression to reach the final model. While knee hyperextension is a component of the Beighton joint hypermobility score, it was not identified as an independent risk factor for cartilage lesions. These 2 variables were therefore not assessed for collinearity.

Univariable and multivariable linear regression were planned to determine the ability of the cartilage injury factors (presence, size, ICRS grade) to predict disease-specific QOL (BPII 2.0). Each independent variable was evaluated using univariable linear regression. As there were no significant variables, multivariable regression was not performed.

This was a retrospective cohort analysis and as such the study population was established prior to analysis. There were 231 patients in the complete cohort involving surgery on 264 separate knees. Of the 152 patients with unilateral symptoms, there were 102 patients with complete 24+ month follow-up at the time of the analysis (67% follow-up), and these were included in the unilateral group. Only unilateral patients were included in the evaluation of QOL outcomes as it is unclear how patients with bilateral symptoms respond when completing patient-reported outcomes. Previous epidemiological studies involving regression analyses have determined that 10 patients per variable provides acceptable power for regression analysis.^35,36

Results

The study included 231 patients (264 knees), of which 70.6% were female. Mean age at first patellar dislocation was 15.6 ± 5.0 years (7-33 years). Mean age at the time of surgery was 24.2 ± 8.4 years (10-47 years), and mean BMI was 24.1 ± 4.1 kg/m² (14.3-38.5 kg/m²; Table 1 ). Radiographic evaluation indicated that 85.2% of patients had signs of trochlear dysplasia, with 39.2% of patients demonstrating high-grade dysplasia. A positive Beighton score was present in 53.5% of patients, and hip rotation abnormalities were present in 26.6% ( Table 1 ). At the time of surgery, 84.5% of knees demonstrated cartilage lesions. According to ICRS grading, 6.8% of knees had ICRS grade I changes, 24.2% had grade II, 42.4% had grade III, and 11.0% had grade IV lesions. In patients with cartilage lesions, the mean total area of damage was measured to be 231.1 ± 143.4 mm² (range of 10-800 mm²). At final follow-up, there were 9 patellar dislocations for an overall failure rate of 3.4% (9 of 264).

Table 1.

Descriptive Statistics for Selected Patient Characteristics.

Variable	Mean	SD	Range	N
BMI (kg/m2)	24.1	4.1	14.3-38.5	264
Time from first dislocation to surgery (years)	8.7	8.1	0-35	264
WARPS-STAID score	5.4	2.7	1-10	264
WARPS	44.1%
STAID	55.9%
Bilateral symptoms	33.9%			231
Limb (% left)	55.3%			264
Crossover sign	82.9%			263
Trochlear dysplasia				263
None	14.8%
Low	46.0%
High	39.2%
Trochlear bump	43.3%			263
Trochlear bump height (mm)	2	2.41	0-9.8	258
Trochlear bump height if bump was present (mm)	4.6	1.3	2-9.8	109
Beighton score	3.7	2.7	0-9	256
Beighton score (% ≥4)	53.5%			256
Knee hyperextension	46.9%			256
Hip rotation abnormality	26.6%			248

BMI = body mass index; WARPS = weak, atraumatic, risky anatomy, pain, subluxation; STAID = strong, traumatic, anatomy normal, instability, dislocation.

The univariable regression of patient characteristics to predict the presence of cartilage lesions revealed several significant variables ( Table 2 ). The highest odds ratios resulted from risk factors related to the presence of trochlear dysplasia. Other significant univariable predictors for the presence of a cartilage lesion at the time of surgery included age at surgery, BMI, time from the first dislocation, and a Beighton score of ≥4 (categorical variable). These significant variables plus Beighton score (continuous variable, P = 0.05), age at first dislocation (P = 0.05), and knee hyperextension (P = 0.1) were entered into the multivariable regression. Backwards stepwise multivariable regression yielded a final parsimonious model that included age at surgery and trochlear dysplasia (R² = 0.17; Table 3 ).

Table 2.

Univariate Logistic Regression to Predict the Presence of Cartilage Lesions.

Variable	OR	95% CI	P
Crossover sign	3.7	1.8-7.8	0.001
Dysplasia (low)	2.4	1.1-5.3	0.03
Dysplasia (high)	13.9	4.2-45.8	<0.001
Trochlear bump	5.5	2.2-13.7	0.001
Trochlear bump height	1.4	1.2-1.7	0.001
Age at surgery	1.1	1.04-1.15	0.001
BMI	1.1	1.01-1.21	0.03
Time from fist dislocation to surgery	1.07	1.01-1.12	0.02
Beighton score ≥4	0.36	0.17-0.77	0.009
Beighton score	0.88	0.78-1.00	0.05
Knee hyperextension	0.55	0.27-1.12	0.1
Age at first dislocation	1.09	0.998-1.18	0.05
Number of dislocations	1	0.97-1.03	0.99
WARPS STAID	0.92	0.81-1.04	0.197
WARPS STAID categorized	0.61	0.30-1.23	0.17
Hip rotational abnormality	0.78	0.37-1.65	0.52
Preoperative BPII 2.0 (unilateral patients)	1.03	0.99-1.07	0.19

OR = odds ratio; CI = confidence interval; BMI = body mass index; WARPS = weak, atraumatic, risky anatomy, pain, subluxation; STAID = strong, traumatic, anatomy normal, instability, dislocation; BPII = Banff Patella Instability Instrument 2.0.

Table 3.

Multiple Logistic Regression to Predict the Presence of Cartilage Lesions.

Variable	OR	95% CI	P
Age at surgery	1.1	1.04-1.17	0.001
Dysplasia (low)	2.9	1.2-6.7	0.015
Dysplasia (high)	15.7	4.6-53.5	<0.001

OR = odds ratio; CI = confidence interval.

Of the knees identified to have cartilage lesions (n = 223), 99.1% were found to have lesions involving the patella, 22.4% had trochlear lesions, 4.0% had lateral femoral condyle lesions, and 3.6% had medial femoral condyle lesions. Prevalence maps of lesion location for these regions are displayed in Figure 1A and B. For the patella, the distal-medial aspect was affected most often, but in fact lesions in the central and medial 4 quadrants were common (82.5% to 88.3%). Trochlear lesions were most commonly located on the lateral aspect (17.5%). Femoral condyle lesions were uncommon.

Figure 1.

(A) Patellar cartilage lesion prevalence map. (B) Femoral cartilage lesion prevalence map.

In patients with unilateral symptoms, BPII 2.0 scores increased by a mean of 45.2 points between preoperative and 24+ month postoperative evaluations (n = 90). The greatest increase in mean scores occurred between baseline and 12 months, with a smaller incremental increase between 12 and 24+ months ( Table 4 ). There was no significant relationship between cartilage lesion presence, grade, or size and BPII 2.0 scores presurgery, 12 (n = 121), or 24+ (n = 102) months postsurgery ( Table 5 ).

Table 4.

BPII 2.0 Scores and Progression for Patients with Unilateral Symptoms.

Variable	Mean	SD	Range	N
Preoperative	26.1	13.2	1.9 to 74.2	100
12 Months	66.1	21.6	14.0 to 98.9	121
24+ Months	71.8	21.6	9.2 to 99.3	102
Change from preoperative to 24+ months	45.2	22.8	−13.7 to 90.4	90

BPII = Banff Patella Instability Instrument 2.0.

Table 5.

Linear Regression for the Prediction of BPII 2.0 by Cartilage Lesion Parameters (12 Months, n = 121; 24+ Months, n = 102).

Variable	Coefficient	95% CI	p
BPII 2.0: 12 months postoperative
Presence of cartilage lesion	5.26	−5.2 to 15.7	0.32
Cartilage grade = 1	8.9	−7.1 to 24.9	0.28
Cartilage grade = 2	10.5	−1.7 to 22.8	0.09
Cartilage grade = 3	1.9	−9.4 to 13.3	0.74
Cartilage grade = 4	1.3	−14.7 to 17.3	0.88
Area of cartilage damage	−0.017	−0.04 to 0.007	0.17
BPII 2.0: 24+ months postoperative
Presence of cartilage lesion	3.2	−9.2 to 15.6	0.61
Cartilage grade = 1	9.2	−10.8 to 29.2	0.36
Cartilage grade = 2	4	−10.8 to 18.8	0.60
Cartilage grade = 3	3.3	−9.7 to 16.4	0.61
Cartilage grade = 4	−5.1	−24.3 to 14.0	0.60
Cartilage area	−0.009	−0.03 to 0.017	0.49

CI = confidence interval; BPII = Banff Patella Instability Instrument 2.0.

Discussion

Predictors of Cartilage Lesions

The results of this study indicate that the presence of trochlear dysplasia was a major risk factor for the development of PF cartilage lesions in patients with recurrent patellar instability. While many factors were predictive of lesion presence in the univariable regression, the multivariable regression reduced these to age at the time of surgery, and trochlear dysplasia. The odds ratio associated with high-grade dysplasia was more than 5 times greater than the odds ratio for low-grade dysplasia (15.7 vs. 2.9), indicating that the grade of dysplasia was a strong predictor of the presence of cartilage lesions. The odds ratio for age at surgery was just 1.1, and as such, this variable may be of questionable clinical relevance.

The association of trochlear dysplasia with PF cartilage lesions has been supported by previous literature; however, its impact in patients with PI has not been described.^14,27,29,37 Dejour and Allain found evidence of trochlear dysplasia in 78% of patients with isolated PF OA. ³⁷ Duran et al. reported that, in patients with AKP, patellar cartilage defects were associated with certain measures of trochlear dysplasia, including lower sulcus angle, lateral trochlear inclination, and higher trochlear angle. ¹⁴ Additionally, Mehl et al. have shown that trochlear dysplasia was found at higher rates in patients with patellar cartilage defects than in patients without. ²⁹ More recently, Ambra et al. have demonstrated trochlear dysplasia as being the most common anatomical factor associated with operative PF cartilage lesions in a group of patients without patellar instability. ³⁸ Despite not being studied specifically in patients with PI, the results of these studies indicate that trochlear dysplasia is a risk factor for developing PF cartilage defects, supporting the findings of the present study.

One component of trochlear dysplasia, the proximo-medial trochlear bump, has been independently associated with the presence of cartilage lesions of the patella. ²⁷ Ali et al. determined that patients with severe PF cartilage defects had trochlear bumps 2 mm larger, on average, than patients without PF cartilage defects. ²⁷ In the current study, univariable analysis demonstrated that the presence of a trochlear bump resulted in a 5.5 times higher odds of having a cartilage lesion. This relationship did not remain in the multivariate model, likely as it is a component of trochlear dysplasia, which was a stronger predictor. The association between the trochlear bump and resulting cartilage lesions may relate to how the patella interacts with this bump as it tracks through flexion. As the knee begins to flex, the distal patella interacts initially with the trochlear bump, which may result in increased contact stress in this area.^27,28,39,40 This theory is supported by the cartilage mapping results of the current study, highlighting one of the reasons why the distal-medial patella was the most common site of cartilage injury. Further study on the contact mechanics between the patella and trochlear bump is required to confirm this association.

Hip rotational abnormalities and generalized ligamentous laxity are two extra-articular factors that have previously been associated with the development of PI.^30,41 A biomechanical study conducted by Liao et al. revealed that internal femoral rotation was associated with elevated hydrostatic pressure and shear stress on the articular surfaces of the PFJ. ³⁰ The current study, however, did not find a significant association with hip rotational abnormalities and the presence of PF cartilage lesions. Regarding ligamentous laxity, Stanitski reported that articular hypermobility was associated with a 2.5 times decreased risk of cartilage injury compared to nonhypermobile controls following acute patellar dislocation. ⁴¹ The present study results aligned with these findings, whereby a Beighton score of ≥4 demonstrated a protective effect (odds ratio = 0.36; p = 0.009) in the univariable logistic regression. This relationship, however, was not sustained in the multivariable regression model.

Many factors are associated with PI, and as such, it is important to discuss how instability may inherently predispose to the development of cartilage lesions. Episodes of instability result in abnormal contact pressures within the PF joint and may contribute to the formation of cartilage defects. With recurrent PI, it has been speculated that the repeated dislocations lead to increased wear on the articular surfaces, and greater prevalence of cartilage lesions.^1,5,11 The present study found no significant association between the duration of symptoms or number of dislocations and the presence of PF cartilage lesions. This finding aligns with results of a study reported by Franzone et al. ⁵ While the current study analysis was limited to evaluating predictors of the presence of lesions, previous studies have reported that, while cartilage lesion prevalence does not increase over time, the severity of the lesion does worsen.^5,9,11 This finding was detailed in 2 studies that revealed that 95% to 97% of patients with acute patellar dislocations had patellar cartilage lesions.^9,10 This high incidence of lesion development at the time of first dislocation may explain why duration of symptoms does not predict the presence of lesions and why such a high proportion of patients present with cartilage lesions at the time of surgery. Nomura et al. additionally reported that lesions from initial dislocation events worsened, morphologically, over an average of 16.5 months. ⁹ These results lend to the discussion of pathogenesis and prognosis of cartilage lesions associated with recurrent patellar dislocation.

Cartilage Lesion Location

The location of PF cartilage lesions observed in this study are consistent with the findings of several previous studies, and provide additional specific mapping of the injury locations. The current study aligns with previous research that indicated the medial patellar facet and lateral femoral trochlea as the locations most frequently injured.^{4,7,10,11,40,42} There is some variation in the literature, however, with some studies identifying the central patellar dome as the most commonly injured region, or lateral and medial patellar facets showing equal involvement.^4,11,43 The current study clearly shows that the central and distal-medial areas of the patellar cartilage are more frequently affected (Fig. 1A and B). In this light, the use of medializing osteotomies, or those with a component of medialization, to treat patellar instability must be evaluated carefully to avoid inadvertently overloading already damaged cartilage. This is in line with the recommendations of Ambra et al., who suggested consideration of the changes in PF contact pressures and patellar tracking that can occur with these osteotomies and the potential future complications that may arise from overloading previously damaged cartilage. ³⁸

It is possible that the variations in reported injury patterns might be due to differences in patient characteristics. For example, patients with more severe trochlear dysplasia might be more likely to sustain distal-medial patellar lesions due to trochlear bump presence. In contrast, patients without trochlear dysplasia may demonstrate more frequent involvement of the lateral trochlea, damaged during dislocation, as well as the medial patellar facet, damaged during relocation. Further work is therefore needed to explore these relationships.

Clinical Outcomes of Patellofemoral Cartilage Lesions

The results of this study showed that the presence, size, or grade of PF cartilage lesions were not related to BPII 2.0 scores presurgery, or at 12 and 24+ months following surgery in this patient population. The available literature on the impact of cartilage lesions on QOL measures for this patient population is limited. Previous studies are limited by the use of differing outcome measures, follow-up periods, patient populations, and conflicting results.^44-48

While the current study utilized the BPII 2.0 to assess disease-specific QOL, previous studies have used Kujala scores to study outcomes in this population. ³¹ The Kujala scoring method was developed to assess patients with AKP, and while AKP may be a component of PI, it does not accurately reflect the complete range of outcomes relevant to this patient population. A study conducted by Astur et al. found that restoration of severe (grade III or IV) patellar chondral defects improved Kujala scores 2 years after autologous osteochondral transplantation. ⁴⁹ Additionally, Christiansen et al. demonstrated that patients with grade III or IV PF chondral lesions were more likely than those with no or minor lesions to have chronic pain following isolated MPFL-R (medial patellofemoral ligament reconstruction). ¹³ This finding, however, was based on only 4 patients with chronic postoperative pain and ICRS grade III or IV lesions, compared with 8 of 40 patients without chronic pain who had grade III-IV chondral lesions. These findings may suggest that the presence of severe patellar lesions may be associated with poorer postoperative pain outcomes, but this is based on a limited sample size. The current study did not focus on AKP as an outcome given that the main indication for patellar stabilization surgery is recurrent unacceptable patellar instability and not AKP. This study has demonstrated that the broader outcome of patient-reported patellar instability-related QOL was not affected by the presence, severity, or size of concomitant knee chondral lesions.

In young patients with joint injury and identified cartilage lesions, the progression toward posttraumatic OA and its sequelae is a serious concern. Several studies observed the relationship between PI and PF OA.^11,15-17 It is important to recognize that 77.6% of the knees in the current study demonstrated grade II-IV cartilage lesions at the time of surgery. In such a young patient population, the high incidence of these cartilage lesions raises concern for the long-term risk of these patients developing OA.

Limitations

The high prevalence of cartilage lesions within the study population may have limited the ability to accurately identify risk factors for the presence of cartilage injury. With over 84% of patients demonstrating the primary outcome measure (and almost 80% of those with ICRS grade II or III lesions), most risk factors for PI were found to be predictive for cartilage lesions in the univariable model. Second, as the cartilage lesions were identified during surgical stabilization, which occurred an average of 8.7 years after their first dislocation, it was not possible to determine when the cartilage lesion(s) actually occurred and how they have progressed over time. Bilateral symptoms were reported in 33.9% of the study population. Validly collecting patient-reported QOL scores in patients with bilateral symptoms is challenging. As such, only patients with unilateral symptoms were included in the component of the study relating cartilage lesions to the BPII 2.0 scores. While this decision reduced the sample size and may have reduced the power of the analysis, having in excess of 100 patients for this analysis should have been sufficient for the number of outcomes included in the model(s). There are a multitude of risk factors that have been associated with, or suggested to be associated with, patellar instability. It was not possible to include all of these factors in the current study, which prevents comment on their potential effect. Another important limitation is the short-term postsurgery follow-up. Previous research investigating cartilage lesions has indicated that symptoms and function diminish over time, and as such the length of follow-up in the present study may not be sufficient to demonstrate outcome variability based on cartilage lesion size, location, or grade. As a disease-specific QOL outcome, the BPII 2.0 is designed to evaluate how the symptoms of patellar instability affect many facets of a patient’s QOL including their emotional state and their ability to participate in activities of daily life, occupational demands, and sports activity. While the breath of questions in the validated BPII 2.0 questionnaire should provide sensitivity to changes in a patient’s patellar instability-related QOL, it is possible that a single factor, such as a chondral injury, may not be sufficient, especially considering the aforementioned limitations, to produce a measureable difference in outcome. Finally, the results of this study cannot be generalized to all patients with PI but only to those patients with symptomatic PI requiring surgical intervention.

Future Considerations

Considering the young age of the study population and the high prevalence of grade II-III PF cartilage lesions, future studies should evaluate the long-term outcomes of these lesions in regard to QOL and progression to PF OA. While many of the variables in this study were independently associated with the presence of cartilage lesions, an evaluation of how these variables may affect the size, location, or grade of cartilage lesions was not performed. In addition, while there was no significant relationship between cartilage lesion presence, size, or grade and BPII 2.0 scores, the relationship between cartilage lesion location and BPII 2.0 scores remains unknown. Future investigations should consider (1) identifying how patient demographic and anatomic characteristics relate to the location, grade, and size of PF cartilage lesions; (2) how cartilage lesion location affects clinical outcomes; and (3) if alterations in intervention timing or strategies can result in improved long-term QOL, and/or decreased rates of progression to PF OA. Understanding how these factors contribute to the array of clinical outcomes will enable clinicians to develop individualized and effective treatment plans for this patient population.

Conclusion

Grade II/III PF cartilage lesions were present at high rates in patients with symptomatic PI undergoing MPFL-R. Trochlear dysplasia is a significant risk factor for the presence of PF cartilage lesions in these patients. PF cartilage lesions are most frequently located on the patella, specifically the central and distal-medial quadrants, followed in prevalence by lateral trochlear lesions. These regions of patellar damage may have consequences when considering medializing osteotomies as part of the approach to correcting instability. There was no significant association between PF cartilage lesion presence, size, or grade and BPII 2.0 QOL scores presurgery, or at 12, or 24+ months following surgery.