Knee Osteoarthritis on MRI in Symptomatic Young Adults After ACL Reconstruction From the SUPER-Knee Trial: Feature Prevalence and Association With Early-Stage Clinical Osteoarthritis Criteria

Abstract

Background:

One of the most common long-term consequences of a traumatic knee injury is the development of knee osteoarthritis (OA), yet the characteristics of OA structural change and their relationship to early-stage OA clinical criteria in those with persistent knee symptoms following traumatic knee injury has not been described.

Purpose:

To describe the prevalence of OA features on magnetic resonance imaging (MRI) and explore the association between MRI-defined OA and early-stage knee OA clinical criteria in young adults with a symptomatic knee at short- to medium-term follow-up after anterior cruciate ligament reconstruction (ACLR).

Study Design:

Cross-sectional study; Level of evidence, 3.

Methods:

Knee MRIs from 184 SUPER-Knee trial participants with a symptomatic knee at 9 to 36 months post-ACLR (68 female; mean ± SD age, 30 ± 6 years; months post-ACLR, 27.6 ± 8.4) were assessed for tibiofemoral and patellofemoral OA features using the MRI Osteoarthritis Knee Score. The prevalence of MRI-defined OA was described based on published criteria by Hunter and colleagues. Logistic regression was used to explore the association between MRI-defined OA and the Luyten classification criteria for early-stage clinical knee OA.

Results:

Across participants, cartilage loss was the most prevalent MRI-defined OA feature in both the tibiofemoral (61%) and the patellofemoral (51%) joint. Osteophytes were more common in the tibiofemoral (52%) than patellofemoral (28%) joint of participants. Bone marrow lesions (tibiofemoral, 26%; patellofemoral, 13%), meniscal tears (42%), and meniscal extrusion (18%) were also observed. Based on MRI-defined criteria, 71 (39%) participants were characterized with OA (tibiofemoral, n = 58 [32%]; patellofemoral, n = 34 [18%]). No association was observed between MRI-defined OA and the Luyten classification criteria for early-stage clinical knee OA.

Conclusion:

Keywords

anterior cruciate ligament early osteoarthritis magnetic resonance imaging cartilage

Anterior cruciate ligament (ACL) injury is an increasingly common and serious knee injury occurring predominately in young active adults.^28,38 One of the most common long-term consequences of an ACL injury, irrespective of surgical ACL reconstruction (ACLR),³³ is the development of knee osteoarthritis (OA).^6,32 Approximately 50% of individuals who experience an ACL injury will develop knee OA and associated disability within 10 years.^{6,12,32,41,53} While established OA represents an advanced and irreversible disease state, preliminary studies suggest that the early stages of disease may be modifiable.^35,42 Characterizing the early stages of knee OA is a primary focus of international initiatives,³⁴ aimed at shifting the paradigm from palliation of established disease to preventing disease onset/progression.

Knee OA disease is traditionally characterized by conventional radiography and evaluated based on the severity of joint space narrowing and presence of osteophytes.³⁰ However, radiographs are insensitive to early disease stages and incapable of detecting OA-related changes in several relevant tissues (eg, cartilage loss, subchondral bone marrow lesions [BMLs], meniscal tears).⁴² It has been suggested that by the time the first radiographic signs of OA become apparent, >10% of knee cartilage volume is already lost.²⁷ Magnetic resonance imaging (MRI) is a more sensitive modality for identifying early disease features through its multiplanar visualization of relevant joint tissues.^43,44 Advances in tissue imaging made possible through MRI have led to widespread adoption of this method of OA characterization in research settings.^20,43,44

Several studies have utilized MRI to assess the early stages of OA disease after ACL injury and ACLR.^5,11,20,37 Following ACLR, MRI OA features have been identified within the first year postoperatively^11,18 and progress rapidly over the subsequent 5 years.^39,40 In contrast to asymptomatic young adults without a history of knee injury,¹³ hallmark OA features such as cartilage loss are exhibited at much higher rates in those having undergone ACLR (48% vs 11%) in the past year.¹¹ Many of these preradiographic structural changes have been proposed as prognostic indicators for the development of established OA^29,45 and have been linked to persistent knee symptoms and poor patient-reported outcomes.^48,52,55 Despite their proposed significance, the characteristics of OA structural change in those with persistent knee symptoms following traumatic knee injury have yet to be described. Investigation of this symptomatic group will help to understand the relationship between traumatic knee injury and the development of established OA disease in young adults.⁵⁴

Early-stage OA clinical criteria (such as that proposed by Luyten et al³⁶) have been used to identify individuals with persistent symptoms and functional limitations, consistent with an early OA diagnosis, following traumatic knee injury.²² These persistent impairments have been shown to negatively affect engagement in exercise/physical activity and health-related quality of life, while also increasing the risk of further functional decline and the development of established OA.^6,12,32,41 Given the known discordance between structural change and symptoms in primary (nontraumatic) OA,^14,21,31 determining the association between MRI-defined OA and early-stage OA clinical criteria may help to inform future efforts to develop early-stage OA diagnostic/classification criteria following traumatic knee injury.

The aim of this study is to describe the prevalence of MRI OA features and MRI-defined knee OA in young adults with a symptomatic knee 9 to 36 months post-ACLR. We also explored the association between MRI-defined OA and the Luyten classification criteria for early-stage clinical knee OA.

Methods

This is a cross-sectional analysis of data from a consecutive sample of 184 participants enrolled in a post-traumatic OA trial (SUPER-Knee) in Melbourne, Australia. Eligible participants were aged 18 to 40 years and 9 to 36 months post-ACLR with a symptomatic knee (defined as an aggregate mean score of <80/100 from 4 Knee injury and Osteoarthritis Outcome Score [KOOS] subscales; including Pain, Symptoms, Sport and Recreation, and knee-related Quality of Life). The trial protocol¹⁶ and baseline sociodemographic characteristics¹⁵ have been detailed elsewhere. Briefly, participants were recruited via direct invitation following ACLR from 12 collaborating private orthopaedic surgeons and 9 public hospitals in Victoria, Australia, from February 2021 to April 2023. Additional participants were recruited from the community via print advertisement and social media. Exclusion criteria were as follows: (1) concomitant intra-articular knee fracture; (2) synthetic ACLR graft; (3) knee reinjury, surgery, or injection in the past 3 months; (3) rehabilitation undertaken for the ACLR knee in the past 6 weeks; (4) planned surgery to ACLR knee in the next 12 months (eg, due to graft rupture, presence of cyclops lesion); (5) presence of other health conditions affecting physical function; and (6) inability to commit to study assessments (including contraindication to MRI). Participants attended La Trobe University, where participant demographics, anthropometrics, self-reported outcomes, and physical examination findings were collected. Unilateral MRI of the affected knee was acquired at a partner radiology clinic (Lake Imaging, Melbourne, Australia). The trial was prospectively registered on the Australian New Zealand Clinical Trials Registry (ACTRN12620001164987). Ethics approval was granted through the La Trobe University and Alfred Hospital. All participants provided informed consent prior to participation.

MRI Acquisition and Interpretation

Unilateral MRI of the affected knee was obtained using a single 3.0-T MRI scanner (Signa Pioneer; General Electric Healthcare) and 18-channel knee coil. Proton density–weighted fat-suppressed fast spin-echo sequences were acquired in the sagittal, axial, and coronal planes. An additional 3-dimensional fast spoiled gradient echo and a multiecho spin-echo T2 sequence were acquired in the sagittal plane (Appendix 1 Table A1). Participants were advised to refrain from strenuous weightbearing exercise/activity for 24 hours prior to MRI acquisition to reduce the risk of acute tissue responses to loading influencing MRI outcomes.⁷

All MRI scans were independently dual rated by 2 trained readers (J.L.C. and T.J.W.) using a modified version of the MRI Osteoarthritis Knee Score (MOAKS).²⁵ To ensure accurate implementation of the MOAKS, a musculoskeletal radiologist (E.O.) with 15 years’ experience evaluating musculoskeletal MRIs in clinical and research settings provided extensive training to the 2 MRI readers over an 18-month period (>35 hours training). Both readers were registered physical therapists with >10 years clinical experience. Each MRI was evaluated independently by both readers, who were blinded to clinical and surgical information (apart from presence of ACLR). Discrepancies between raters were resolved by consensus, with the experienced musculoskeletal radiologist consulted as necessary.

We separated the knee into 6 subregions (4 from the tibiofemoral joint, 2 from the patellofemoral joint) to score articular cartilage loss and BMLs. Osteophytes were scored in 12 subregions (6 from the tibiofemoral joint, 6 from the patellofemoral joint). Meniscal pathology (tear, extrusion, maceration) were scored separately for the medial and lateral menisci (Appendix 2 Table A1). Articular cartilage loss was graded from 0 to 3 based on size (percentage of surface area relative to the subregion; 0, no cartilage loss; 1, <10%; 2, 10%-75%; 3, >75%) and depth (percentage of full-thickness loss relative to the subregion; 0, no full-thickness loss; 1, <10%; 2, 10%-75%; 3, >75%). A score of ≥1 was used to define the presence of articular cartilage loss. BMLs (percentage of volume relative to subregion: 0, none; 1, <33%; 2, 33%-66%; 3, >66%) and osteophytes (how far growth extends from underlying bone: 0, none; 1, small; 2, medium; 3, large) were also graded based on size. A score of ≥1 was used to define the presence of a BML. As the definition of a definitive osteophyte has yet to be established,²⁵ we considered an osteophyte to be present with a score ≥2.^40,46 Meniscal tears and maceration were described as either present or absent. Meniscal extrusion was described by size (0, <2.0 mm; 1, 2.0-2.9 mm; 2, 3.0-4.9 mm; 3.0, ≥5 mm). A score of ≥1 was used to define the presence of meniscal extrusion (Appendix 2 Table A2).

MRI-Defined Knee OA

The presence of MRI-defined knee OA (tibiofemoral or patellofemoral) was evaluated based on an established definition by Hunter et al (Figure 1).²⁴ While these criteria were originally developed to characterize established OA, its high specificity and less optimal sensitivity (due to the detection of early disease) has seen it applied to cohorts in the early stages of OA disease.⁵ Tibiofemoral OA was defined as the presence of both a definite osteophyte and full-thickness cartilage loss or the presence of either a definite osteophyte or full-thickness cartilage loss and ≥2 BMLs, meniscal extrusion/maceration/tear, or partial-thickness cartilage loss (where full-thickness loss was not present) in relevant subregions of the tibia and/or femur. Patellofemoral OA was characterized by the presence of both a definite osteophyte and partial- or full-thickness cartilage loss involving the patella and/or femoral trochlea. Combined OA was defined as meeting the criteria for both tibiofemoral and patellofemoral OA as described above.

Figure 1.

MRI-defined knee osteoarthritis criteria. MRI, magnetic resonance imaging; OA, osteoarthritis. Criteria adapted from Hunter et al.²⁴

Clinical Criteria and Outcomes

Patient-reported outcomes were evaluated using the KOOS. The KOOS comprises 42 items across 5 subscales and has demonstrated internal consistency, test-retest reliability, responsiveness to change, and construct validity, in the assessment of patient-relevant outcomes following knee injury.⁸ Items are rated on a 4-point Likert scale (0 = no problems, 4 = extreme problems) and a total score for each subscale is summated and transformed to a 0 to 100 scale (where 100 = no knee problems).

One of 6 investigators (including J.L.C., T.J.W., M.A.G.) conducted a physical examination for knee crepitus and joint line tenderness of the index ACLR knee. All investigators were registered physical therapists with >6 years of clinical experience, who underwent initial training in physical examination methods and retraining every 6-months during the data collection period to ensure fidelity. Investigators were blinded to participant symptom history and MRI outcomes. Knee crepitus was assessed by the investigator's placing his or her palm lightly over the patella during 3 consecutive squats to a depth of approximately 60° knee flexion.⁹ A positive test was defined as continuous grinding, crunching, or crackling during all 3 squats (ie, 1 or 2 clicks, pops, or cracks were not considered crepitus). This physical examination method was selected for evaluating knee crepitus over other techniques owing to its functional relevance, clinical utility, and previously demonstrated interrater reliability.⁹ Joint line tenderness was evaluated with the participant in supine and the knee flexed to 90°. The length of the medial joint line was palpated, followed immediately by the lateral joint line, and the participant asked to report any pain, tenderness, or discomfort.

Early-stage clinical knee OA was defined using a modification of the classification criteria proposed by Luyten et al.³⁶ Although not originally developed to characterize early-stage OA following traumatic knee injury, these criteria have been applied to post-ACLR cohorts.²² Participants meeting these criteria were required to demonstrate (1) a “positive” score (≤85%) on 2 of 4 possible KOOS subscales (Pain, Symptoms [including stiffness], Activities of Daily Living [short form], or knee-related Quality of Life) and (2) exhibit either joint line tenderness or knee crepitus on physical examination (Table 1).

Table 1

Early-Stage Clinical Knee Osteoarthritis Definition ^a

Criteria
Knee injury and Osteoarthritis Outcome Score ^b
Pain – 9 items
Symptoms (including stiffness) – 7 items
Activities of Daily Living (short form) – 7 items
Knee-related Quality of Life – 4 items
Physical examination ^c
Joint line tenderness
Knee crepitus

Definition adapted from Luyten et al.³⁶

Two out of 4 subscales must be positive (≤85%).

At least 1 physical examination criterion must be present.

Statistical Analysis

Descriptive statistics (mean, standard deviation, median, interquartile range, count, percentage) were used to summarize participant characteristics and the frequency of MRI OA features, MRI-defined knee OA, and early-stage clinical knee OA. The distribution of continuous data was assessed with Shapiro-Wilk tests and visual inspection of histograms and was reported as appropriate. Binary logistic regression was used to explore the association between MRI-defined OA and the Luyten classification criteria for early-stage clinical knee OA. Odds ratios and associated 95% CIs were calculated. All analyses were adjusted for age, sex, body mass index (BMI), and time from injury to ACLR, because of their potential influence on MRI-defined OA features and symptoms.^39,40 This approach aligns with the recommendations outlined in the recent OPTIKNEE consensus statement.⁵⁴ Model assumptions were assessed prior to analysis, including evaluation of multicollinearity (using variance inflation factors) and model fit. Statistical analyses were performed using Stata 18.0 (StataCorp) with a significance threshold of P < .05.

To determine the reliability of the MOAKS in this investigation, the prevalence-adjusted bias-adjusted kappa (PABAK) for each OA feature was calculated. This method of evaluating reliability provided a more realistic estimate of agreement than a traditional kappa, especially when the prevalence of a given feature was relatively low.⁴ Interrater reliability was established from the independently evaluated MRI scans (ie, before consensus) of all participants. Intrarater reliability was established by having 1 rater (J.L.C.) reevaluate a series of randomly selected participant MRIs (n = 20) approximately 9 months later (mean ± SD, 9.5 ± 3.1 months).²⁵ Interrater reliability of physical examination methods was evaluated by calculating Fleiss kappa and has been reported previously.⁹

Results

A total of 184 participants were included (68 female; mean age 30 ± 6 years; mean BMI, 27.3 ± 5.2 kg/m²) (Figure 2). Participants were, on average, 27.6 ± 8.4 months post-ACLR (range, 9.0-36.0 months), and most underwent reconstruction using a hamstring tendon autograft (82%). Knee crepitus was present in 113 (61%) participants on physical examination, and 78 (42%) participants exhibited joint line tenderness (Table 2).

Figure 2.

Flowchart of the recruitment process in the SUPER-Knee trial. ACL, anterior cruciate ligament; ACLR, anterior cruciate ligament reconstruction; KOOS, Knee injury and Osteoarthritis Outcome Score; KOOS₄, mean score from 4 of the 5 KOOS subscales covering Pain, Symptoms, Sport and Recreation and Quality of Life; MRI, magnetic resonance imaging.

Table 2

Participant Characteristics ^a

Variable	N = 184
Female sex	68 (37)
Age, y	30 ± 6
Height, m	1.7 ± 0.1
Mass, kg	82.4 ± 16.5
Body mass index, kg/m²	27.3 ± 5.2
Time post-ACLR, mo ^b	27.6 ± 8.4
Time between ACL injury and ACLR, mo, median (IQR)	4.6 (2.1-9.8)
KOOS₄	64.2 ± 13.0
KOOS subscale score
Pain	78.9 ± 11.6
Symptoms	71.9 ± 12.6
Sport and Recreation	62.3 ± 19.5
Quality of Life	43.7 ± 19.1
Activities of Daily Living (short form)	77.1 ± 13.6
Knee crepitus	113 (61)
Joint line tenderness	78 (42)
ACLR index knee
Primary Revision	158 (86) 26 (14)
ACLR graft type ^b
Hamstring tendon Quadriceps tendon Bone–patellar tendon–bone Donor, allograft	150 (82) 17 (9) 12 (7) 5 (3)
Type of injury ^c
Isolated ACL injury ACL injury with injury/surgery to meniscus and/or cartilage ^d	63 (34) 120 (65)

Data are presented as mean ± SD or n (%) unless otherwise indicated. ACL, anterior cruciate ligament; ACLR, anterior cruciate ligament reconstruction; KOOS, Knee injury and Osteoarthritis Outcome Score; KOOS₄, mean score from 4 of the 5 KOOS subscales covering Pain, Symptoms, Sport and Recreation and Quality of Life.

Data obtained from surgical records or self-report when surgical records unavailable.

Data obtained from surgical records of n = 183.

Defined as the presence of a concomitant meniscal tear (either untreated or treated surgically) and/or a cartilage defect of grade ≥2 and/or treated surgically (eg, debridement).

Interrater reliability for all reported features across both tibiofemoral and patellofemoral joints was high (PABAK, 0.73-0.96), with the exception of partial-thickness cartilage loss (PABAK, 0.58-0.68) (Appendix 1 Table A2). Intrarater reliability was also high (PABAK, 0.73-1.00) for all reported features (Appendix 1 Table A3).

Prevalence of Knee OA Features on MRI

Cartilage loss was demonstrated in the tibiofemoral joint of 113 (61%) and the patellofemoral joint of 93 (51%) participants, with tibiofemoral and patellofemoral full-thickness loss evident in 14 (8%) and 4 (2%) participants, respectively. BMLs were present in the tibiofemoral joint of 48 (26%) participants and the patellofemoral joint of 24 (13%) participants. Osteophytes were observed in the tibiofemoral and patellofemoral joint of 95 (52%) and 51 (28%) participants, respectively. Meniscal tears were demonstrated in 77 (42%) participants, and meniscal extrusion in 33 (18%) participants. No participants exhibited meniscal maceration. The prevalence of OA features separated by subregion and MOAKS grade are described in Figure 3.

Figure 3.

Prevalence of knee osteoarthritis features on magnetic resonance imaging by subregion using the MRI Osteoarthritis Knee Score. Numbers above each bar represent feature prevalence (%) by subregion. Numbers within each bar represent the prevalence of each feature by grade. Grade 1 osteophytes were excluded from analysis.

Association Between MRI-Defined OA and Early-Stage Clinical Knee OA Criteria

Based on the criteria for MRI-defined knee OA, 58 (32%) participants were characterized as having tibiofemoral OA, 34 (18%) with patellofemoral OA, and 21 (11%) with combined OA. Tibiofemoral OA was more common in the medial compartment (n = 34; 18%) than the lateral (n = 25; 14%). Overall, 71 (39%) participants were characterized as having MRI-defined knee OA in ≥1 joint. Using the Luyten classification criteria, 123 (67%) participants were classified as having clinical early-stage knee OA. Characteristics of participants meeting (and not meeting) the criteria for MRI-defined OA and early-stage clinical knee OA have been provided (Appendix 1 Table A4). There was no association between the criteria for MRI-defined OA and early-stage clinical knee OA (Table 3).

Table 3

Distribution of MRI-Defined OA and Association With Early-Stage Clinical Knee OA ^a

	Frequency of Outcome, n (%)	Unadjusted	Adjusted ^b
Tibiofemoral OA
Present	40/58 (69)	1.24 (0.64-2.41)	1.08 (0.55-2.14)
Absent	83/126 (66)	1.00 (Ref)	1.00 (Ref)
Patellofemoral OA
Present	26/34 (76)	1.76 (0.75-4.20)	1.52 (0.61-3.78)
Absent	97/150 (65)	1.00 (Ref)	1.00 (Ref)
Combined OA
Present	17/21 (81)	2.29 (0.73-7.12)	2.11 (0.64-6.92)
Absent	106/163 (65)	1.00 (Ref)	1.00 (Ref)

Data are presented as odds ratio (95% CI) unless otherwise indicated. MRI, magnetic resonance imaging; OA, osteoarthritis; Ref, reference.

Adjusted for age, sex, body mass index, and time from injury to anterior cruciate ligament reconstruction.

Discussion

In our study of symptomatic young adults 9 to 36 months post-ACLR, more than half exhibited features of knee OA on MRI, with the tibiofemoral joint more commonly affected than the patellofemoral joint. Over one-third of participants met the criteria for MRI-defined knee OA. Despite the high prevalence of MRI OA features and persistent knee symptoms demonstrated in this cohort, we did not find a meaningful association between the criteria for MRI-defined OA and early-stage clinical knee OA.

SUPER-Knee trial participants demonstrated comparable rates of patellofemoral cartilage loss (51%) to previous cohorts (45%) at similar time points post-ACLR, but slightly higher prevalence of tibiofemoral cartilage loss (61% vs 48%).^10,11 These findings contrast arthroscopic investigations, typically reporting a greater frequency of cartilage loss at the patellofemoral (compared with tibiofemoral) joint following ACLR.⁵¹ The observed rates of cartilage loss in our cohort are greater than those reported in asymptomatic individuals (aged <40 years) without a history of traumatic knee injury (11%),¹³ suggesting that the changes we observed are likely attributed to an accelerated onset of OA disease as opposed to incidental findings. Interestingly, the prevalence of cartilage loss is similar to our observed rates of meniscal pathology (ie, tear or extrusion; n = 92 [50%]), supporting the belief that meniscal lesions may be the trigger that initiates cartilage loss, or possibly, that meniscal pathology and cartilage loss share a similar disease trajectory in the early stages of knee OA.¹⁷

The prevalence of osteophytes among our participants was notably higher (tibiofemoral, 52%; patellofemoral, 28%) than other cohorts at similar time points post-ACLR (tibiofemoral, 7%; patellofemoral, 3%) evaluated with the MOAKS.¹¹ Our participants were older, demonstrated a higher BMI, and presented with a greater frequency of concomitant knee injuries at the time of ACLR compared with the previous cohort, all of which have been suggested to accelerate the progression of OA disease following traumatic knee injury.⁵⁴ While the osteophytes described in our cohort may have been preexisting (ie, present prior to ACL injury/surgery), the rapid bone area changes inherent to the OA disease process³ and relatively low rates of osteophytes demonstrated in young asymptomatic adults (8%)¹³ suggests the observed prevalence is likely driven by incident bone change following traumatic knee injury. Given that cartilage loss and osteophytes are both hallmark features of OA,⁴² the elevated prevalence of these findings in our sample is indicative of greater progression along the disease continuum relative to previously evaluated cohorts.^11,37,50

Tibiofemoral BMLs were present at a comparable rate in our cohort (26%) to that of other ACLR samples at similar time points postsurgery (30%-31%),^11,50 although patellofemoral BMLs were less common (13% vs 24%).¹¹ While the presence of BMLs following ACL injury (particularly of the lateral femoral condyle and posterolateral tibial plateau) may represent a footprint of the initial injury mechanism,⁵⁰ median traumatic BML resolution time is suggested to be approximately 10 months.² In contrast, BMLs associated with knee OA are unlikely to resolve.²⁶ While some BMLs identified in this cohort may be sequelae of the initial injury or ACLR procedure, their persistent nature (or incident development) suggests they are more likely attributable to early-stage OA disease.

Over one-third (39%) of participants in our cohort met the criteria for MRI-defined knee OA. The prevalence of patellofemoral joint OA (18%) was similar to that reported in other samples at comparable time points post-ACLR (17%).¹¹ Consistent with the observed rates of tibiofemoral OA features, the number of participants meeting the criteria for tibiofemoral OA was greater in our study (32%) compared with previously evaluated cohorts (19%).¹¹ Despite the high prevalence of MRI-defined knee OA and persistent knee symptoms demonstrated in this cohort, we did not find a meaningful association between the criteria for MRI-defined OA and early-stage clinical knee OA. These findings may underscore a disparate structure-symptoms relationship in this early-stage (post-traumatic) OA population (similar to that observed in established OA)^21,31 or the inability of current diagnostic constructs to capture both structural and symptomatic components of early-stage OA following traumatic knee injury.

In the absence of validated and established early-stage knee OA criteria, a definitive relationship between OA-related symptoms and structure in young adults following traumatic knee injury is difficult to ascertain. Despite a lack of association between MRI-defined OA and early-stage clinical knee OA criteria, the wide confidence intervals of our results suggest that OA-related structural change may drive symptomatic burden in some of these participants. This is further supported by lower mean KOOS subscale scores among those meeting MRI-defined OA criteria relative to those who did not (Appendix 1 Table A4). It should be noted that statistical comparison of KOOS subscales scores between these groups was beyond the aims of the current investigation. Without appropriate statistical comparison, it is unclear whether the differences between these groups are statistically significant. Furthermore, the observed differences between groups are below the minimal clinically important difference for KOOS subscales (ie, 8-10 points) and therefore may not be clinically meaningful.⁸ Future investigation of the association between individual knee OA features on MRI and KOOS subscale scores (or appropriate Patient Acceptable Symptom State thresholds) in this cohort will help to elucidate the symptom-structure relationship in young adults following traumatic knee injury.

Limitations

Several limitations should be considered when interpreting the findings of our study. First, the MRI definition of knee OA applied in this study was originally developed using the Whole Organ MRI Score.²⁴ Despite this, we do not consider the application of MOAKS to affect the characterization of MRI-defined knee OA. The MOAKS provides a more comprehensive assessment of knee OA features²⁵ and has previously been applied alongside the selected MRI criteria.⁵ We did not evaluate MRI-assessed bone attrition in our cohort, limiting the full consideration of all proposed tissue pathology under the applied MRI criteria. We were also unable to determine the chronicity of cartilage loss and meniscal tears in our cohort. As both features are a common occurrence at the time of ACL injury,¹¹ it is possible that some participants were categorized as meeting the MRI criteria based on acute (rather than degenerative) presentations of these features. Second, our results are likely subject to selection bias owing to the symptomatic eligibility criteria (aggregate mean score of <80/100 from 4 KOOS subscales) of the overarching clinical trial from which our cross-sectional data were drawn. This is perhaps most evident in the observed prevalence of early-stage clinical knee OA in our cohort, which is likely inflated relative to other post-ACLR samples at similar time points postsurgery.^23,41 There is also ongoing debate over whether symptoms experienced in the years following ACLR are a residual consequence of ACL injury/surgery, rather than reflective of early-stage knee OA.⁵³ Longitudinal studies post-ACLR suggest that improvements in KOOS subscale scores plateau between 6 and 12 months following surgery⁴⁹ and that KOOS subscale scores at 1 year post-ACLR remain mostly stable up to 5 and 10 years.^1,19,47 From this, it is reasonable to suggest that the symptoms evidenced in our cohort (mean, 2.3 years post-ACLR) are reflective of a chronic illness state (such as early-stage knee OA) rather than a transient consequence of the initial injury/surgery. Finally, the Luyten classification criteria for early-stage clinical knee OA includes a radiographic component (Kellgren-Lawrence grade 0-1 on standing/weightbearing) to exclude the presence of established OA.³⁶ While these criteria have been applied to ACLR cohorts in the absence of radiographic evaluation,²² we cannot be certain that participants within our study did not meet the radiographic criteria for established OA.³⁰ Based on the relatively young age of participants and comparison with previous ACLR cohorts of similar time points postsurgery,¹¹ we believe that the number of participants in our cohort with established radiographic OA is negligible (ie, <5%). We are also unable to comment on the impact of intrarater reliability on physical examination procedures in the Luyten classification criteria for early-stage clinical knee OA, as these were not evaluated.

Conclusion

Over half of symptomatic young adults 9 to 36 months post-ACLR demonstrated OA features on MRI, with the tibiofemoral joint more commonly affected than the patellofemoral joint. More than one-third of participants met the criteria for MRI-defined knee OA. There was no meaningful association between the criteria for MRI-defined OA and early-stage clinical knee OA. Sports medicine practitioners should be aware that OA features on MRI are common among young adults in the short to medium term after ACLR. Many of these young adults will also meet criteria for MRI-defined OA. It is likely that a discordant symptom-structure relationship exists in the early stages of OA following traumatic knee injury.

Footnotes

Appendix 1

Table A4

Characteristics of Participants Meeting MRI or Early-Stage Clinical OA Definitions ^a

	MRI-defined OA (+)	MRI-defined OA (–)	Early-stage Clinical OA (+)	Early-stage Clinical OA (–)
Participants, n	71	113	123	61
Female sex	24 (34)	44 (39)	44 (36)	24 (39)
Age, y	31 ± 6	29 ± 6	30 ± 6	29 ± 5
Body mass index, kg/m²	27.3 ± 5.5	27.3 ± 5.0	27.6 ± 5.2	26.7 ± 5.2
Post-ACLR, y ^b	2.4 ± 0.7	2.2 ± 0.7	2.2 ± 0.7	2.2 ± 0.7
KOOS₄	61.9 ± 13.2	65.6 ± 12.7	60.5 ± 12.7	71.7 ± 10.0
KOOS subscale score
Pain	78.1 ± 12.1	79.5 ± 11.2	76.3 ± 11.5	84.3 ± 9.7
Symptoms	69.1 ± 12.7	73.6 ± 12.3	67.5 ± 11.8	80.6 ± 9.4
Sport and Recreation	59.8 ± 18.8	63.8 ± 19.8	58.0 ± 19.9	70.8 ± 15.4
Quality of Life	40.8 ± 19.3	45.5 ± 18.8	40.0 ± 18.7	51.0 ± 17.8
Activities of Daily Living	75.4 ± 14.1	78.0 ± 13.2	74.0 ± 14.0	83.1 ± 10.2
Knee crepitus	51 (72)	63 (56)	99 (80)	15 (25)
Joint line tenderness	28 (39)	50 (44)	71 (58)	7 (11)
ACLR index knee
Primary Revision	56 (79) 15 (21)	102 (90) 11 (10)	105 (85) 18 (15)	53 (87) 8 (13)
Type of injury ^b
Isolated ACL injury ACL injury with injury/surgery to meniscus and/or cartilage^c	14 (20) 57 (80)	49 (43) 63 (56)	42 (34) 81 (66)	21 (34) 39 (64)

Data are presented as mean ± SD unless otherwise indicated. ACL, anterior cruciate ligament; ACLR, ACL reconstruction; KOOS, Knee injury and Osteoarthritis Outcome Score; KOOS₄, mean score from 4 of the 5 KOOS subscales covering Pain, Symptoms, Sport and Recreation, and Quality of Life; MRI, magnetic resonance imaging; OA, osteoarthritis.

Data obtained from surgical records of n = 183.

Defined as the presence of a concomitant meniscal tear (either untreated or treated surgically) and/or a cartilage defect of grade ≥2 and/or treated surgically (eg, debridement).

Appendix 2

Table A2

MOAKS Scoring ^a

	Grade 0	Grade 1	Grade 2	Grade 3
Cartilage loss, size ^b	None	<10	10-75	>75
Cartilage loss, depth ^b	None	<10	10-75	>75
Bone marrow lesion ^c	None	<33	33-66	>66
Osteophyte	None	Small	Medium	Large
Meniscal extrusion, mm	<2	2.0-2.9	3.0-4.9	≥5

Data are presented as percentages unless otherwise indicated. A score of ≥1 was used to define the presence of articular cartilage loss, bone marrow lesion, and meniscal extrusion. A score of ≥2 was used to define the presence of an osteophyte. Meniscal tears and maceration were described as either present or absent. MOAKS, Magnetic Resonance Imaging Osteoarthritis Knee Score.

Percentage of surface area relative to the subregion.

Percentage of volume relative to the subregion.

Acknowledgements

La Trobe University's SUPER-Knee program is not to be confused with the University of Newcastle's SuPeR Knee. Support. Predict. Recover program ().

Final revision submitted April 23, 2026; accepted May 4, 2026.

One or more of the authors has declared the following potential conflict of interest or source of funding: The SUPER-Knee study was funded by a National Health and Medical Research Council (NHMRC) of Australia Project Grant (No. GNT1158500). J.L.C. is supported by an Australian Government Research Training Program scholarship. D.O.S. is supported by an NHMRC of Australia Investigator Grant (No. GNT2033417). J.L.W. is supported by a Michael Smith Health Research British Columbia Scholar Award (No. SCH-2020-0403). M.A.G. was supported by an NHMRC of Australia postgraduate scholarship at the time of the study (No. GNT1190882). B.P. is supported by an Australian Research Council Early Career Industry Fellowship (No. IE230100135). A.G.C. is supported by an NHMRC of Australia Investigator Grant (No. GNT2008523).

Ethics approval was granted through the La Trobe University (No. HEC-19447) and Alfred Hospital (No. HREC 537/19) human research ethics committees.

ORCID iDs

Jamon L. Couch

Thomas J. West

Matthew G. King

Danilo De Oliveira Silva

Jackie L. Whittaker

Andrea M. Bruder

Mark Scholes

Ewa M. Roos

References

Ahldén

Samuelsson

Sernert

Forssblad

Karlsson

Kartus

The Swedish National Anterior Cruciate Ligament Register: a report on baseline variables and outcomes of surgery for almost 18,000 patients. Am J Sports Med. 2012;40(10):2230-2235. doi:10.1177/0363546512457348

Boks

Vroegindeweij

Koes

Bernsen

Hunink

Bierma-Zeinstra

SM.

MRI follow-up of posttraumatic bone bruises of the knee in general practice. AJR Am J Roentgenol. 2007;189(3):556-562. doi:10.2214/ajr.07.2276

Bowes

Kacena

Alabas

, et al. Machine-learning, MRI bone shape and important clinical outcomes in osteoarthritis: data from the Osteoarthritis Initiative. Ann Rheum Dis. 2021;80(4):502-508. doi:10.1136/annrheumdis-2020-217160

Byrt

Bishop

Carlin

JB.

Bias, prevalence and kappa. J Clin Epidemiol. 1993;46(5):423-429. doi:10.1016/0895-4356(93)90018-v

Chang

Roemer

Guermazi

, et al. Do existing magnetic resonance imaging definitions of knee osteoarthritis identify knees that will develop clinically significant disease over up to 11 years of follow-up? Arthritis Rheumatol. 2025;77(2):140-150. doi:10.1002/art.42982

Chen

Wang

Jiang

Chen

Radiographic osteoarthritis prevalence over ten years after anterior cruciate ligament reconstruction. Int J Sports Med. 2019;40(11):683-695. doi:10.1055/a-0902-8539

Coburn

Crossley

Kemp

, et al. Is running good or bad for your knees? A systematic review and meta-analysis of cartilage morphology and composition changes in the tibiofemoral and patellofemoral joints. Osteoarthritis Cartilage. 2023;31(2):144-157. doi:10.1016/j.joca.2022.09.013

Collins

NJD

Prinsen

CAC

Christensen

Bartels

Terwee

Roos

. Knee Injury and Osteoarthritis Outcome Score (KOOS): systematic review and meta-analysis of measurement properties. Osteoarthritis Cartilage. 2016;24(8):1317-1329. doi:10.1016/j.joca.2016.03.010

Couch

King

De Oliveira Silva

, et al. Diagnostic performance of self-reported knee crepitus using a knee injury and osteoarthritis outcome score item. Osteoarthritis Cartilage. 2025;33(6):764-769.

10.

Couch

Patterson

Crossley

, et al. Knee crepitus and osteoarthritis features in young adults following traumatic knee injury. Arthritis Care Res (Hoboken). 2026;78(3):398-406. doi:10.1002/acr.25637

11.

Culvenor

Collins

Guermazi

, et al. Early knee osteoarthritis is evident one year following anterior cruciate ligament reconstruction: a magnetic resonance imaging evaluation. Arthritis Rheumatol. 2015;67(4):946-955. doi:10.1002/art.39005

12.

Culvenor

Eckstein

Wirth

Lohmander

Frobell

Loss of patellofemoral cartilage thickness over 5 years following ACL injury depends on the initial treatment strategy: results from the KANON trial. Br J Sports Med. 2019;53(18):1168-1173. doi:10.1136/bjsports-2018-100167

13.

Culvenor

Øiestad

Hart

Stefanik

Guermazi

Crossley

KM.

Prevalence of knee osteoarthritis features on magnetic resonance imaging in asymptomatic uninjured adults: a systematic review and meta-analysis. Br J Sports Med. 2019;53(20):1268-1278. doi:10.1136/bjsports-2018-099257

14.

Culvenor

Segal

Guermazi

, et al. Sex-specific influence of quadriceps weakness on worsening patellofemoral and tibiofemoral cartilage damage: a prospective cohort study. Arthritis Care Res (Hoboken). 2019;71(10):1360-1365. doi:10.1002/acr.23773

15.

Culvenor

West

Bruder

, et al. Recruitment and baseline characteristics of young adults at risk of early-onset knee osteoarthritis after ACL reconstruction in the SUPER-Knee trial. BMJ Open Sport Exerc Med. 2024;10(2):e001909. doi:10.1136/bmjsem-2024-001909

16.

Culvenor

West

Bruder

, et al. Supervised exercise-therapy and patient education rehabilitation (SUPER) versus minimal intervention for young adults at risk of knee osteoarthritis after ACL reconstruction: SUPER-Knee randomised controlled trial protocol. BMJ Open. 2023;13(1):e068279. doi:10.1136/bmjopen-2022-068279

17.

Ding

Jones

Wluka

Cicuttini

What can we learn about osteoarthritis by studying a healthy person against a person with early onset of disease?

Curr Opin Rheumatol. 2010;22(5):520-527. doi:10.1097/BOR.0b013e32833b90e9

18.

Frobell

Le Graverand

M-P

Buck

, et al. The acutely ACL injured knee assessed by MRI: changes in joint fluid, bone marrow lesions, and cartilage during the first year. Osteoarthritis Cartilage. 2009;17(2):161-167. doi:10.1016/j.joca.2008.06.020

19.

Girdwood

Crossley

Patterson

, et al. People are more variable than their hop test would suggest: hop performance and self-reported outcomes over 11 years following ACL reconstruction. Scand J Med Sci Sports. 2024;34(9):e14727. doi:10.1111/sms.14727

20.

Girdwood

Patterson

Crossley

, et al. Hip rotation muscle strength is implicated in the progression of early post-traumatic osteoarthritis: a longitudinal evaluation up to 5 years following ACL reconstruction. Phys Ther Sport. 2023;63:17-23. doi:10.1016/j.ptsp.2023.06.001

21.

Hannan

Felson

Pincus

Analysis of the discordance between radiographic changes and knee pain in osteoarthritis of the knee. J Rheumatol. 2000;27(6):1513-1517.

22.

Harkey

Baez

Lewis

, et al. Prevalence of early knee osteoarthritis illness among various patient-reported classification criteria after anterior cruciate ligament reconstruction. Arthritis Care Res (Hoboken). 2022;74(3):377-385. doi:10.1002/acr.24809

23.

Harkey

Driban

Todem

, et al. Evaluating criteria for symptoms suggestive of early osteoarthritis over two years post–anterior cruciate ligament reconstruction: data from the New Zealand Anterior Cruciate Ligament Registry. Arthritis Care Res (Hoboken). 2025;77(4):475-483. doi:10.1002/acr.25453

24.

Hunter

Arden

Conaghan

, et al. Definition of osteoarthritis on MRI: results of a delphi exercise. Osteoarthritis Cartilage. 2011;19(8):963-969. doi:10.1016/j.joca.2011.04.017

25.

Hunter

Guermazi

, et al. Evolution of semi-quantitative whole joint assessment of knee OA: MOAKS (MRI Osteoarthritis Knee Score). Osteoarthritis Cartilage. 2011;19(8):990-1002. doi:10.1016/j.joca.2011.05.004

26.

Hunter

Zhang

Niu

, et al. Increase in bone marrow lesions associated with cartilage loss: a longitudinal magnetic resonance imaging study of knee osteoarthritis. Arthritis Rheum. 2006;54(5):1529-1535. doi:10.1002/art.21789

27.

Jones

Ding

Scott

Glisson

Cicuttini

Early radiographic osteoarthritis is associated with substantial changes in cartilage volume and tibial bone surface area in both males and females. Osteoarthritis Cartilage. 2004;12(2):169-174. doi:10.1016/j.joca.2003.08.010

28.

Kaplan

Patterson

Bruder

Ezzat

Bunzil

Culvenor

AG.

ACL—wow this is bad”: patients’ perspectives on their anterior cruciate ligament injury and its care—a systematic review and qualitative evidence synthesis. Br J Sports Med. 2026. doi:10.1136/bjsports-2025-109746

29.

Katsuragi

Sasho

Yamaguchi

, et al. Hidden osteophyte formation on plain x-ray is the predictive factor for development of knee osteoarthritis after 48 months—data from the Osteoarthritis Initiative. Osteoarthritis Cartilage. 2015;23(3):383-390. doi:10.1016/j.joca.2014.11.026

30.

Kellgren

Lawrence

JS.

Radiological assessment of osteo-arthrosis. Ann Rheum Dis. 1957;16(4):494-502. doi:10.1136/ard.16.4.494

31.

Lethbridge-Cejku

Scott

Jr Reichle

, et al. Association of radiographic features of osteoarthritis of the knee with knee pain: data from the Baltimore Longitudinal Study of Aging. Arthritis Care Res. 1995;8(3):18218-8. doi:10.1002/art.1790080311

32.

Lie

Risberg

Storheim

Engebretsen

Øiestad

BE.

What’s the rate of knee osteoarthritis 10 years after anterior cruciate ligament injury? An updated systematic review. Br J Sports Med. 2019;53(18):1162-1167. doi:10.1136/bjsports-2018-099751

33.

Lien-Iversen

Morgan

Jensen

Risberg

Engebretsen

Viberg

Does surgery reduce knee osteoarthritis, meniscal injury and subsequent complications compared with non-surgery after ACL rupture with at least 10 years follow-up? A systematic review and meta-analysis. Br J Sports Med. 2020;54(10):592-598. doi:10.1136/bjsports-2019-100765

34.

Liew

King

Mahmoudian

, et al. A scoping review of how early-stage knee osteoarthritis has been defined. Osteoarthritis Cartilage. 2023;31(9):1234-1241. doi:10.1016/j.joca.2023.04.015

35.

Lohmander

Hellot

Dreher

, et al. Intraarticular sprifermin (recombinant human fibroblast growth factor 18) in knee osteoarthritis: a randomized, double-blind, placebo-controlled trial. Arthritis Rheumatol. 2014;66(7):1820-1831. doi:10.1002/art.38614

36.

Luyten

Bierma-Zeinstra

Dell’Accio

, et al. Toward classification criteria for early osteoarthritis of the knee. Semin Arthritis Rheum. 2018;47(4):457-463. doi:10.1016/j.semarthrit.2017.08.006

37.

Macri

Whittaker

Toomey

, et al. Patellofemoral joint geometry and osteoarthritis features 3-10 years after knee injury compared with uninjured knees. J Orthop Res. 2024;42(1):78-89. doi:10.1002/jor.25640

38.

Maniar

Verhagen

Bryant

Opar

DA.

Trends in Australian knee injury rates: an epidemiological analysis of 228,344 knee injuries over 20 years. Lancet Reg Health West Pac. 2022;21. doi:10.1016/j.lanwpc.2022.100409

39.

Patterson

Culvenor

Barton

, et al. Worsening knee osteoarthritis features on magnetic resonance imaging 1 to 5 years after anterior cruciate ligament reconstruction. Am J Sports Med. 2018;46(12):2873-2883. doi:10.1177/0363546518789685

40.

Patterson

Culvenor

Barton

Guermazi

Stefanik

Crossley

KM.

Patient-reported outcomes one to five years after anterior cruciate ligament reconstruction: the effect of combined injury and associations with osteoarthritis features defined on magnetic resonance imaging. Arthritis Care Res (Hoboken). 2020;72(3):412-422. doi:10.1002/acr.23854

41.

Patterson

Emery

Crossley

, et al. Knee- and overall health-related quality of life following anterior cruciate ligament injury: a cross-sectional analysis of Australian and Canadian cohorts. J Orthop Sports Phys. 2023;53(7):402-413. doi:10.2519/jospt.2023.11838

42.

Pollard

TCB

Gwilym

Carr

. The assessment of early osteoarthritis. J Bone Joint Surg Br. 2008;90(4):411-421. doi:10.1302/0301-620X.90B4.20284

43.

Quatman

Hettrich

Schmitt

Spindler

KP.

The clinical utility and diagnostic performance of magnetic resonance imaging for identification of early and advanced knee osteoarthritis: a systematic review. Am J Sports Med. 2011;39(7):1557-1568. doi:10.1177/0363546511407612

44.

Roemer

Guermazi

Demehri

Wirth

Kijowski

Imaging in osteoarthritis. Osteoarthritis Cartilage. 2022;30(7):913-934. doi:10.1016/j.joca.2021.04.018

45.

Roemer

Kwoh

Hannon

, et al. What comes first? Multitissue involvement leading to radiographic osteoarthritis: magnetic resonance imaging–based trajectory analysis over four years in the osteoarthritis initiative. Arthritis Rheumatol. 2015;67(8):2085-2096. doi:10.1002/art.39176

46.

Runhaar

Schiphof

van Meer

Reijman

Bierma-Zeinstra

SMA

Oei

EHG

. How to define subregional osteoarthritis progression using semi-quantitative MRI osteoarthritis knee score (MOAKS). Osteoarthritis Cartilage. 2014;22(10):1533-1536. doi:10.1016/j.joca.2014.06.022

47.

Spindler

Huston

Chagin

, et al. Ten-year outcomes and risk factors after anterior cruciate ligament reconstruction: a MOON longitudinal prospective cohort study. Am J Sports Med. 2018;46(4):815-825. doi:10.1177/0363546517749850

48.

Stefanik

Gross

Guermazi

, et al. The relation of MRI-detected structural damage in the medial and lateral patellofemoral joint to knee pain: the multicenter and Framingham osteoarthritis studies. Osteoarthritis Cartilage. 2015;23(4):565-570. doi:10.1016/j.joca.2014.12.023

49.

Teitsma

van der Hoeven

Tamminga

de Bie

RA.

Impact of patient sex on clinical outcomes: data from an anterior cruciate ligament reconstruction registry, 2008-2013. Orthop J Sports Med. 2014;2(9):2325967114550638. doi:10.1177/2325967114550638

50.

van Meer

Oei

EHG

Meuffels

, et al. Degenerative changes in the knee 2 years after anterior cruciate ligament rupture and related risk factors: a prospective observational follow-up study. Am J Sports Med. 2016;44(6):1524-1533. doi:10.1177/0363546516631936

51.

Wang

Chen

, et al. Second-look arthroscopic evaluation of the articular cartilage after primary single-bundle and double-bundle anterior cruciate ligament reconstructions. Chin Med J (Engl). 2011;124(21):3551-3555.

52.

Wang

Antony

Zhu

, et al. Association of patellar bone marrow lesions with knee pain, patellar cartilage defect and patellar cartilage volume loss in older adults: a cohort study. Osteoarthritis Cartilage. 2015;23(8):1330-1336. doi:10.1016/j.joca.2015.02.018

53.

Whittaker

JL.

"Every new beginning comes from some other beginning's end": anterior cruciate ligament injury and posttraumatic knee osteoarthritis. Arthritis Care Res (Hoboken). 2022;74(3):337-339. doi:10.1002/acr.24810

54.

Whittaker

Losciale

Juhl

, et al. Risk factors for knee osteoarthritis after traumatic knee injury: a systematic review and meta-analysis of randomised controlled trials and cohort studies for the OPTIKNEE Consensus. Br J Sports Med. 2022;56(24):1406-1421. doi:10.1136/bjsports-2022-105496

55.

Yusuf

Kortekaas

Watt

Huizinga

TWJ

Kloppenburg

Do knee abnormalities visualised on MRI explain knee pain in knee osteoarthritis? A systematic review. Ann Rheum Dis. 2011;70(1):60-67. doi:10.1136/ard.2010.131904