Osteochondral lesions of the talus (OLTs) are a prevalent clinical condition that cause pain and dysfunction in the ankle, impairing quality of life and posing a substantial burden on those affected (1, 2). There are various treatment options for osteochondral lesions of the talus (OLTs) depending on the condition. For small lesions, conservative treatment may be effective. Bone marrow stimulation is recommended when the lesion area is less than 1.5 cm² (3–5). However, for lesions larger than 10 mm in diameter accompanied by subchondral cyst formation, autologous osteochondral transplantation (AOT) is recommended (6–9) its effectiveness and practicality have been clinically recognized (10–12).

The MRI-based Magnetic Resonance Observation of Cartilage Repair Tissue (MOCART) score is a widely utilized tool for assessing cartilage repair following AOT (13, 14). However, several studies have reported a limited correlation between postoperative MOCART scores and functional outcomes, suggesting that the MOCART score may not fully reflect the clinical benefits of the surgical intervention (15–17). Additionally, the International Cartilage Repair Society (ICRS) score, based on second-look arthroscopy, is regarded as the gold standard for assessing cartilage repair, as it enables direct visualization and a more comprehensive evaluation of the repaired cartilage (18–20). This scoring system classifies cartilage damage based on characteristics such as surface integrity, depth, and tissue morphology observed during direct arthroscopic visualization. It is commonly used to grade cartilage damage in arthroscopic procedures (21). The primary objective of this study was to assess the correlation between second-look arthroscopy (ICRS), functional scales, and the MOCART score.

This study included 47 patients (35 male and 12 female), with a mean age of 35.1 years (range 26–52 years). The mean follow-up duration was 32.5 ± 6.7 months (range 24–51 months), and the average body mass index (BMI) was 24.3 kg/m² (Table 1). All patients reported a history of previous ankle sprains on the affected side, 8 patients were diagnosed with anterior talofibular ligament injury through physical examination and MRI examination. Preoperative magnetic resonance imaging (MRI) measurements revealed that the mean horizontal diameter of the lesion on talus surface was 10.5 ± 0.7 mm, the mean AP diameter was 11.9 ± 1.1 mm, and the mean depth of the lesion was 10.0 ± 1.0 mm. Typical patient shows in Figure 2.

AOFAS, VAS, ICRS, and MOCART scores showed improvement during postoperative follow-up, the scores are shown in Table 2. Pearson correlation tests were used to assess the relationships between AOFAS, VAS, MOCART, and ICRS scores at various follow-up periods. The results are illustrated in Figures 3, 4, showing the correlations between MOCART scores and AOFAS, VAS, and ICRS scores at different time points. The initial ICRS score showed a negligible correlation with preoperative AOFAS (r = 0.12, p = 0.40, 95% CI: −0.69–1.71) and VAS scores (r = 0.05, p = 0.74, 95% CI: −0.21–0.30), with r < 0.30.

At 1 year postoperatively, a weak correlation was observed between MOCART and VAS scores (r = −0.43, p < 0.001, 95% CI: −0.08 to −0.02). A similarly weak correlation was found between MOCART and AOFAS scores (r = 0.27, p = 0.07, 95% CI: −0.01–0.36). At 2 years postoperatively, the highest correlation was observed between ICRS and AOFAS scores (r = 0.56, p < 0.05, 95% CI: 1.19–3.07), indicating a moderate correlation. In contrast, correlations between MOCART and AOFAS (r = 0.34, p = 0.02, 95% CI: 0.05–0.49), VAS (r = −0.46, p < 0.05, 95% CI: −0.08 to −0.02), and ICRS (r = 0.36, p < 0.05, 95% CI: 0.40–3.11) remained weak (Figure 5).

Our retrospective review showed that transplanted hyaline cartilage survived, and ICRS scores improved in all patients at the 2-year follow-up after AOT. A weak correlation was observed between MOCART and function scores at 2 years postoperatively, and low to moderate correlation between ICRS scores and functional scores.

MRI is currently the most commonly used imaging method for observing cartilage condition (22, 23). However, among the follow-up indicators after AOT, MRI often reveals “poor signals,” such as bone marrow edema around the graft, signal mismatch between the graft and surrounding tissue, and the formation of subchondral cysts (22, 24). Previous studies have shown no significant correlation between MRI performance at an early follow-up and clinical functional results (22, 24, 25). Our results are in agreement with these studies and also found weak correlation between them. The incidence of cyst formation after AOT is reportedly 65% (16), and the incidence of cyst formation after AOT concurrent with extracellular matrix enhancement is 14% (10). Cyst formation could be attributed to the joint synovial fluid flowing into the subchondral bone through the graft–host cartilage connection interface after AOT. The signal of subchondral bone edema is universal. According to Dhollander et al. (26), all cases exhibited high levels of subchondral edema following knee osteochondral transplantation. Pain in patients with OLT was attributed to subchondral synovial fluid entering the subchondral bone through microfractures, stimulating nerve endings and causing pain (27). After AOT, even if the graft appears to fit well with the surrounding tissue, synovial fluid can still enter the subchondral bone through the surgical gap, leading to postoperative bone marrow edema and the formation of subchondral cysts. Over time, the graft heals with the surrounding subchondral bone, preventing further synovial fluid infiltration. However, this does not cause significant discomfort, which may explain the lack of a clear correlation between functional and radiological scores in the long term post-surgery.

Similarly, previous studies have suggested that the correlation between the results of second-look arthroscopy and functional results is limited (28–30). In our study, at 2 years postoperatively, AOFAS and ICRS scores showed a moderate correlation, while no significant correlation was found between the VAS and ICRS scores, which aligns with current knowledge on this topic. Second-look arthroscopy can truly reflect the condition of the articular surface. For patients with good graft healing, the condition of the subchondral bone or cyst cannot be observed. Furthermore, there are many causes of postoperative pain, including local synovitis, scar adhesion, or psychological factors (27). Therefore, it is reasonable to believe that when the surface of the articular cartilage is intact, second-look ICRS score would be good with or without pain.

For MOCART and ICRS scores, there is a certain similarity between the evaluation of the consistency of the graft, surrounding cartilage and the articular surface. Goller et al. reported MOCART 2.0 parameters showed significant correlation with Delta International Knee Documentation Committee Subjective Form (IKDC) scores in the postoperative course after retropatellar matrix-associated chondrocyte transplantation (MACT) (31). This research differ with our finding. However, the talus's convex surface and load-bearing mechanics (vs. retropatellar cartilage in Goller's study) may reduce MOCART's sensitivity to functional outcomes. This may indicate that the matrix-associated chondrocyte technique was more suitable for MOCART scale instead of AOT procedure. Furthermore, the ICRS score may not change even if MRI finds abnormal subchondral bone signals or new cysts. This may be a limitation of the ICRS score because observing only the surface of the graft will indeed lead to some pathological changes being missed, even though these conditions may be routine manifestations after AOT. In our study, there was no significant correlation between ICRS and MOCART scores. Although the ICRS score is the gold standard for cartilage evaluation, it is occasionally difficult to judge the condition of the subchondral bone, even if the changes in subchondral bone may be normal. Therefore, it may be more reasonable to combine the MOCART score with the ICRS score for an integrated evaluation of the graft.

This study also has limitations. First, More functional scores can be included in the follow-up process to comprehensively analyze correlations. Previous study reported that the Foot and Ankle Outcome Score (FAOS) as patient-specific outcome measures for the surgical treatment of OLT (29). Second, we did not collect cartilage specimens for histological analysis. Finally, since the internal fixators were taken out after 2 years from the operation, the follow-up time was set to 2 years, and we did not assess the patients for new positive findings in the long-term follow-up after removing the internal fixators. In the future, longer-term functional and imaging follow-up studies are needed to verify the current conclusions. Fourth, the relatively small sample size constitutes another limitation. Although our power analysis confirmed the adequacy of the current sample size for correlation analyses (β = 0.8, α = 0.05), it remains insufficient for more complex multivariate regression modeling. This sample size restriction. Fifth, future studies should incorporate advanced cartilage-specific MRI sequences (e.g., T2 mapping, T1ρ) for more sensitive evaluation of graft integrity.

In conclusion, AOT is an effective treatment for Hepple V OLT. However, the correlation between second-look arthroscopy findings and functional scores was weak, and a low correlation was found between arthroscopy results and MRI findings after AOT.