Introduction

The treatment of articular cartilage defects of the lower extremities is challenging given the poor recruitment of regenerative cells into the defect, which results in a limited self-healing capacity of native cartilage tissue [1, 2]. Patients with full-thickness cartilage defects may experience significant pain, decrease in joint function and quality of life [3]. Cartilage defects result from trauma or repetitive shear and torsional forces applied to the cartilage surface [4]. If left untreated, cartilage lesions can progress to early osteoarthritis, joint pain, and mobility impairment [5, 6]. The previous studies estimated that the incidence of full-thickness cartilage lesions varies from 5% to 10% in knees of patients undergoing arthroscopy [7,8,9]. The prevalence of cartilage lesions in the ankle joint varies from 40% to 95% in patients with persistent pain after an ankle sprain with and without chronic ankle instability [10].

Microfracture (MFX) is considered the first-line treatment for cartilage defects, in both the knee and ankle joints, given the simplicity of the procedure, its low cost, minimal invasiveness, and satisfactory short-term clinical outcome [11, 12]. However, there have been concerns regarding the durability of MFX over time, since the clinical outcomes may worsen over time, in particular in larger lesions and more active patients [13,14,15]. To overcome limitations of the MFX technique, various forms of autologous chondrocyte implantation (ACI) and osteoarticular transfer system (OATS) have been introduced [16]. Although multiple techniques have been suggested to be effective in the management of articular cartilage defects, international recommendations and clear guidelines are missing. Moreover, it is also uncertain in which patient-dependent factors predict success after any of the above-mentioned cartilage treatment options.

Therefore, a systematic review was conducted to investigate whether patient characteristics at baseline exert an influence on surgical outcome in terms of patient-reported outcome reports (PROMs) and complications.

Materials and methods

Search strategy and data source

This systematic review is according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses: the PRISMA statement [17]. The literature search was conducted independently by two authors (F.M. & J.E.). In May 2022, the following databases were accessed: PubMed, Google scholar, Embase, and Scopus. The following keywords were used in combination: chondral, cartilage, articular, damage, defect, injury, chondropathy, knee, pain, matrix-induced, periosteal, periosteum, collagen, autologous, chondrocyte, transplantation, implantation, MFX, microfractures, mosaicplasty, mACI, cACI, pACI, AMIC, OAT, osteochondral transplantation, allograft, autograft, membrane, therapy, management, surgery, outcomes, revision, hypertrophy, failure. Articles resulting from the literature search were screened by the same authors. The full text of the articles of interest was accessed. The bibliographies were also screened by hand for inclusion. Disagreements were solved by a third author (NM).

Eligibility criteria

All the studies investigating the outcomes of surgical management for knee and/or talus chondral defects were accessed. Only studies performing OAT, MFX, ACI or AMIC were considered. Given the authors language abilities, articles in English, German, Italian, French and Spanish were eligible. Studies with level I to IV of evidence, according to Oxford Centre of Evidence-Based Medicine [18], were considered. Abstracts, reviews, comments, editorial and opinion were not considered. Animals, biomechanics or in vitro studies were not considered. Studies enhancing the surgical procedures with stem cells were also eligible. Only studies that clearly stated the nature of the surgical intervention were included. Studies which included patients with large chondral and/or osteochondral lesions (>5 cm2) and obese (BMI > 30 kg/m2) were not included. Studies that reported data on patients with end-stage joint degeneration were not eligible. Only articles reporting quantitative data under the outcomes of interest were considered for inclusion. Missing data under the outcomes of interest warranted the exclusion from this study.

Outcomes of interest

Data extraction was performed separately by two authors (F.M. & J.E.). Data concerning author, year, journal, type of study and length of the follow-up were collected. The following data at baseline were collected: the number of patients, age, sex, mean BMI (kg/m2), size of the defect (cm2), and duration of symptoms (months). Data concerning the following scores were extracted at baseline and at last follow-up: visual analogic scale (VAS), American Orthopedic Foot and Ankle Score (AOFAS) [19], Tegner Activity Scale [20], Lysholm Knee Scoring Scale [21], and International Knee Documentation Committee (IKDC) [22] scores. Furthermore, rate of hypertrophy, failures, and revisions were also retrieved. The primary outcome was to investigate the association of baseline patient-specific characteristics on surgical outcomes following restorative cartilage procedures for the knee and ankle.

Methodology quality assessment

The methodological quality assessment was performed by two authors independently (F.M. & J.E.). The risk of bias graph tool of the Review Manager Software (The Nordic Cochrane Collaboration, Copenhagen) was used. The following risk of bias was evaluated: selection, detection, attrition, and other source of bias.

Statistical analysis

All statistical analyses were performed by one author (F.M.) using the software STATA/MP 14.1 (StataCorp, College Station, TX). The Shapiro–Wilk test was performed to investigate data distribution. For normal data, mean and standard deviation were calculated. For nonparametric data, median and interquartile range were calculated. The Student’s T test was used to assess significance for parametric data, while the Mann–Whitney U-test was used for nonparametric variables. Values of P < 0.05 considered statistically significant. A multivariate analysis was performed to assess associations between data of patients at baseline with the clinical scores at last follow-up and complications. A multiple linear model regression analysis through the Pearson Product–Moment Correlation Coefficient (\(r\)) was used. The Cauchy–Schwarz formula was used for inequality: + 1 is considered as positive linear association, while − 1 a negative one. Values of 0.1 <|\(r\) |< 0.3, 0.3 <|\(r\) |< 0.5, and |\(r\) |> 0.5 were considered to have weak, moderate, and strong association, respectively. The overall assessment of significance was performed using the χ2 test, with values of P < 0.05 considered statistically significant.

Results

Search result

The literature search resulted in 795 articles. Of them, 309 were duplicates. A further 302 articles were not eligible: surgical technique (N = 74), not focusing on knee or ankle (N = 41) study design (N = 140), not reporting quantitative data under the outcomes of interest (N = 19), other (N = 24), language limitations (N = 4). This left 184 articles for the present study. The literature search results are shown in Fig. 1.

Fig. 1
figure 1

Flowchart of the literature search

Full size image

Methodological quality assessment

The risk of selection bias was judged as moderate. Indeed, 46.7% of studies (86 of 184) performed a retrospective analysis, while 38.0% (70 of 184) were prospective, and 15.2% (28 of 184) were randomized. Only few studies (35 of 184) performed assessor blinding; thus, the risk of detection bias was high. The risk of attrition and reporting bias were moderate, as was the risk of other bias. In conclusion, the overall review authors' judgements about each risk of bias item scored moderate, attesting to this study acceptable methodological assessment. The risk of bias graph is shown in Fig. 2.

Fig. 2
figure 2

Methodological quality assessment

Full size image

Patient demographics

Data from 8905 procedures were retrieved. The median duration of symptoms before the index surgery was 36 (23.6–50.8) months. 41.7% (3713 of 8905) of patients were women. The mean age of the patients was 33.9 ± 6.9 years, while the mean BMI 25.6 ± 1.5 kg/m2. The mean defect size was 3.3 ± 2.3 cm2. The median follow-up time was 41.8 (24 to 60) months. Generalities and demographic data of the study are shown in Table 1.

Table 1 Generalities and patient baseline of the included studies
Full size table

Outcomes of interest

Female sex evidenced moderate association with greater VAS at last follow-up (r = 0.3; P = 0.02). Patient’s age evidenced negative association with the AOFAS score (r =  − 0.2; P = 0.04) and Lysholm Knee Scoring Scale (r =  − 0.4; P = 0.03). Greater BMI was moderately associated with the rate of graft hypertrophy (r = 0.6; P = 0.009). VAS, IKDC, AOFAS, and Tegner Activity Scale at baseline were positively associated with themselves at last follow-up: VAS (r = 0.9; P < 0.0001), IKDC (r = 0.5; P = 0.007), AOFAS (r = 0.6; P = 0.0002), Tegner Activity Scale (r = 0.4; P = 0.009). The VAS score at baseline was inversely associated with the Tegner Activity Scale (r =  − 0.8; P < 0.0001) at last follow-up. No further statically significant associations were evidenced. The results of each of the pairwise correlation is shown in greater detail in Table 2.

Table 2 Overall results of the multivariate analyses
Full size table

Discussion

The management of articular cartilage defects still presents a major challenge. Therefore, identification of prognostic factors would allow to predict the outcome of various surgical techniques in multiple joints, and it would help to educate patients on the success (or not) of their surgical intervention. Older age was associated with lower values of the AOFAS and Lysholm scores at last follow-up, while women evidenced a positive association with VAS. Given the weak associations between these endpoints, the role of sex and age still remain not fully defined. BMI evidenced a moderate positive association with the rate of graft hypertrophy. VAS, IKDC, AOFAS, and Tegner scores at baseline were associated among themselves at last follow-up, demonstrating that the final outcome is influenced by the pre-operative performance status of the patients. Interestingly, symptom duration prior to the surgical intervention and cartilage defect size did not show any significant association with the surgical outcome.

Neri et al. analyzed 48 patients who underwent microfractures of knee cartilage defects at a mean follow-up of 5.7 years [23]. Patients’ age, BMI, time from diagnosis to surgery, and size of the cartilage lesion were negatively associated with the functional outcome. Differences between these findings and our results may be explained by the fact that Neri et al. only included 48 patients treated for knee articular cartilage defects with a longer follow-up, while we included 8905 procedures including various treatment options with a variable follow-up. Similar findings were reported by Andriolo et al. in 113 patients with knee cartilage defects treated with matrix-assisted autologous chondrocyte implantation [24]. Older age, female sex, degenerative lesions, longer symptoms duration, and previous surgery were negatively associated with outcome.

Age has been identified as one of the most important factors for success in the treatment of cartilage defects [25, 26]. A study comparing microfracture to ACI or OATS showed better clinical outcomes for patients younger than 30 years compared to those older than 30 years, unrelated to treatment type [27, 28]. Robb et al. were able to identify age as a prognostic factor for lower clinical outcomes after treatment [29]. The structure and composition of the matrix molecules as well as the synthetic function of chondrocyte change with age, may explain the lower functional outcomes after cartilage defect management in older patients [30, 31]. These previous findings confirm our results of age being negatively associated with AOFAS and Lysholm scores.

Previously, it was also shown, in particular for the ankle joint, that patients’ BMI is a negative prognostic factor [32], with worse clinical outcomes for patients with a BMI > 30 kg/m2 [33, 34]. Jaiswal et al. found similar results showing an influence of BMI on the Modified Cincinnati Score after anterior cruciate ligament reconstruction and matrix-assisted autologous chondrocyte implantation [35]. The poorer results following cartilage repair in obese patients may be explained by an increase in mechanical forces across the joint leading to cartilage breakdown.

There was evidence of weak association between female sex and VAS. Females have lower femoral and retropatellar cartilage volumes than males, and this decreases with age [36]. These findings might explain the higher risk for knee osteoarthritis in women compared to men. Kreutz et al. studied 52 patients after ACI showing worse outcomes in women compared to men [37]. Furthermore, higher complication rates in cartilage repair surgery were found in women 24 months after surgery, which might be related to lower satisfaction levels in women, possibly resulting in more postoperative complaints [38,39,40].

Interestingly, cartilage defect size, in both the knee and ankle joints, is not associated with negative outcome. This was previously confirmed in studies highlighting that defect size does not predict the clinical outcome after treatment of cartilage defects, confirming that functional outcome seems to be independent of cartilage defect size [25, 41].

This study identified prognostic factors for successful cartilage repair management in the knee and ankle joints, regardless of the surgical procedure. However, there are also some limitations that need to be addressed. Although we followed established guidelines for the preparation of systematic reviews, the risk of bias of the included studies was only moderate, with acceptable methodological assessment. Given the lack of quantitative data, primary and revision settings could not be analyzed separately. To increase the pooling data, the sex of the patients, mean age and BMI, defect size, and mean length of prior symptoms duration were not analyzed separately according to the body location (knee and ankle). Furthermore, we considered only the most common surgeries strategy for chondral repair, potentially increasing the risk of selection bias. Given their uncertain results, less common or more innovative procedures were not considered. Given the lack of data, surgical indications were not considered separately for analysis. Patients with larger chondral and/ or osteochondral lesions (>5 cm2) and obese (BMI > 30 kg/m2) were not considered, as the surgical outcomes are strongly negatively influenced by these variables [33,34,35]. Large lesions require challenging surgery, with transplants and unpredictable outcome. Similarly, in obese patients, the articular cartilage is subjected to high loads, and lesions may not heal properly. Further clinical investigations are required to establish the proper management of chondral defects. Results from the present study should be considered in the light of these limitations. Further high-quality investigations should validate the results of the present study in a clinical setting.

Conclusion

Our results suggest that the clinical outcomes were mostly related to the patients’ performance status prior surgery and that greater BMI could be associated with greater rate of hypertrophy. Female sex and older age evidenced fair influence on outcome, while symptom duration prior to the surgical intervention and cartilage defect size evidenced no association with the surgical outcome. These results should be interpreted in the light of the limitations of the present study, and further investigations are needed to validate them in a clinical setting.