Post-operative osteoarthritis following meniscus repair versus meniscectomy: A systematic review and meta-analysis

Protocol and registration

The Preferred Reporting Items for Systematic Reviews and Meta-Analyses 2020 standards were adhered to in this systematic review. The protocol was recorded in the PROSPERO/CRD420251140270 prospective database.

Search strategy

A comprehensive literature search was conducted across multiple electronic databases, including PubMed, Scopus, and Web of Science. To ensure full reproducibility, the exact Boolean operators, search strings, and filters applied during the search process are detailed in Appendix A [Supplementary Material]. Filters included language restrictions (English only), study type (human studies), and publication date limits where applicable. The search strategy was designed to capture all relevant studies by combining keywords and MeSH terms using Boolean operators such as “AND,” “OR,” and “NOT.” This systematic approach ensured a thorough and transparent identification of eligible studies for inclusion in the meta-analysis. Keywords and MeSH terms included meniscus repair, meniscectomy, OA, degenerative changes, arthroscopy, and long-term outcomes. Reference lists of included studies and relevant reviews were also screened.^[10-18]

Search string

Additional Points: Search strings were adapted for each database’s syntax and indexing system.

Reference lists of included studies were manually screened to identify additional eligible articles.

Duplicate records were removed using (e.g., EndNote, Zotero, or RevMan).

Eligibility criteria (PICO)

Studies were eligible for quantitative pooling if they reported OA incidence (radiographic KL ≥2 or symptomatic OA) with extractable effect measures and ≥2 years’ follow-up. Although the primary electronic search was restricted to studies published between 2000 and 2025, selected pre-2000 studies were retained for narrative synthesis when they represented seminal biomechanical investigations or long-term cohort analyses establishing the mechanistic link between meniscal tissue loss and OA progression. These foundational studies were not included in the quantitative meta-analysis and did not contribute to pooled effect estimates.

• Population: Patients of any age undergoing arthroscopic meniscus repair or partial meniscectomy

• Intervention: Meniscus repair

• Comparator: Partial meniscectomy.

• Outcomes:

  1. Primary: Incidence of post-operative OA (radiographic or symptomatic)

  2. Secondary: MRI evidence of cartilage degeneration, patient-reported functional scores International Knee Documentation Committee (IKDC), Knee Injury and Osteoarthritis Outcome Score (KOOS), Lysholm, and reoperation rates.

Study design

Randomized controlled trial (RCT) and prospective or retrospective cohort studies with ≥2 years’ follow-up.

Exclusion

Case reports, cadaveric studies, reviews, expert opinion, and follow-up <2 years.

Subgroup analyses were conducted to explore potential sources of heterogeneity and to enhance the clinical relevance of the findings. Age and follow-up duration were selected as subgroup criteria based on their established impact on disease progression and treatment outcomes in OA. Age stratification allows for the examination of differential effects across younger and older populations, while follow-up duration helps assess the sustainability and long-term efficacy of interventions. Leave-one-out sensitivity analysis showed consistent results. Subgroup analyses by repair technique and anterior cruciate ligament reconstruction were limited by small sample sizes and should be interpreted cautiously.

Outcome definition

Radiographic OA refers to OA diagnosed based on imaging findings, usually X-rays.

Criteria Used: The most common classification system is the Kellgren and Lawrence grading scale, which assesses features such as joint space narrowing, osteophyte formation, subchondral sclerosis, and bone deformity.

Grades:

• Grade 0: No radiographic features of OA

• Grade 1: Doubtful narrowing of joint space and possible osteophyte

• Grade 2: Definite osteophyte and possible joint space narrowing

• Grade 3: Moderate multiple osteophytes, definite joint space narrowing, some sclerosis

• Grade 4: Large osteophytes, marked joint space narrowing, severe sclerosis, and deformity

Radiographic OA may be present even in individuals without symptoms, especially in joints like the hand.

Symptomatic OA is diagnosed when radiographic evidence of OA is accompanied by clinical symptoms such as joint pain, stiffness, or functional limitation.

Typical Symptoms: Persistent joint pain, especially during or after activity, morning stiffness lasting <30 min, reduced range of motion or joint function.

This definition is more aligned with patient experience and is often used in epidemiological studies to estimate the burden of disease.

Data extraction and risk of bias

Two reviewers independently screened studies, extracted data, and assessed risk of bias using the Newcastle–Ottawa Scale for cohorts and the Cochrane Risk of Bias tool for RCTs. Risk of bias for each included study was independently assessed by two reviewers using a standardized tool (e.g., Cochrane Risk of Bias Tool). To ensure consistency between reviewers, inter-rater agreement was evaluated using Cohen’s kappa (κ) statistic, which quantifies the level of agreement beyond chance. Discrepancies in assessment were resolved through discussion, and when necessary, a third reviewer was consulted to reach consensus. This approach ensured a reliable and transparent evaluation of methodological quality across studies.

Resolving disagreements among reviewers in this meta-analysis involves a structured and collaborative approach.

Initially, the reviewers independently extract data and assess the quality of studies through a process known as dual coding. When discrepancies arise, they are addressed through open discussion, and if consensus cannot be reached, a third reviewer is often brought in to mediate and finalize decisions. To further ensure consistency and fairness, review teams may hold consensus meetings where contentious points are deliberated collectively. In addition, predefined decision rules – such as majority voting or deferring to a senior reviewer – are employed to guide the resolution process and maintain methodological rigor throughout the analysis.

To prevent double-counting, registry-based and multicenter studies were examined for overlapping recruitment periods, institutions, and patient cohorts. When overlap was suspected, only the most comprehensive dataset was retained.

Data synthesis

Due to heterogeneity in OA definitions and surgical techniques, both narrative synthesis and pooled meta-analysis were performed. For dichotomous outcomes such as OA incidence and reoperation rates, risk ratios (RRs) with 95% confidence intervals (CIs) were computed. For continuous outcomes, mean differences (MDs) or standardized MDs were used depending on the measurement scales across studies. A random-effects model was applied to account for between-study variability.

Data synthesis was performed using RevMan (Review Manager) software, developed by the Cochrane Collaboration. This platform facilitated the generation of forest plots, subgroup analyses, and risk of bias assessments. To evaluate heterogeneity across studies, both the I² statistic and τ² (tau-squared) were calculated. The I² statistic quantified the percentage of variation due to heterogeneity rather than chance, while τ² estimated the between-study variance under a random-effects model. To assess potential publication bias, visual inspection of funnel plots was conducted, and Egger’s regression test was applied to detect asymmetry.

Study selection

The search identified 1,842 articles. After removing duplicates and screening, 32 studies (22 qualitative syntheses and 10 quantitative meta-analysis studies) were included, comprising 6,412 patients with a mean follow-up of 8.6 years (range 2-20 years). Thirty-two studies met the inclusion criteria and were included in the qualitative synthesis, of which 10 provided extractable data for quantitative meta-analysis. The study selection process is shown in Figure 1. The characteristics of the included studies are summarized in Table 1, and the methodological quality assessment is presented in Table 2. The pooled risk of osteoarthritis following meniscus repair versus meniscectomy is illustrated in the forest plot. Of these, 10 studies provided extractable data suitable for quantitative meta-analysis. The study selection process is illustrated in the PRISMA flow diagram [Figure 1]. Thus, the review synthesizes 22 studies qualitatively, plus 10 contributing to pooled estimates.^[1-32] The characteristics of the included studies, including study design, sample size, follow-up duration, repair techniques, and osteoarthritis assessment methods, are summarized in Table 1.

Close

Figure 1:

Preferred Reporting Items for Systematic Reviews and Meta-analyses flow diagram. Standard flow of records identified, screened, excluded, and included (n = 32 studies).

Export to PPT

Patient demographics and interventions

The mean patient age was 32.4 years. Most studies included both male and female patients, with tear patterns ranging from longitudinal and bucket-handle to complex tears. All procedures were arthroscopic, using inside-out, all-inside, or outside-in repair techniques.

OA incidence

Repair was associated with a 23–27% absolute reduction in post-operative OA incidence compared with meniscectomy. Meta-analysis of 10 studies (n = 2,340) demonstrated a pooled RR of 0.52 (95% CI 0.41–0.65). Moderate between-study heterogeneity was observed (I² = 58%; τ² = 0.041) under a random-effects model. MRI studies confirmed better cartilage preservation after repair.^[5,11,12,22]

Reoperation rates

Reoperation was more common after repair (10–26%) than meniscectomy (3–8%), with most failures occurring within 24 months.^[6,12,19-21] However, the majority of revisions were minor procedures and did not negate the long-term benefits of repair in reducing OA risk. Reoperation data were extracted from included studies with attention to the underlying causes. To enhance clinical interpretability, reoperations were categorized based on reported reasons such as technical failure like inadequate initial repair or implant issues, recurrent tears like biological failure or poor tissue healing, and patient-related activity factors like premature return to high-impact activities or non-compliance with rehabilitation protocols. Where studies did not explicitly state the cause, efforts were made to infer the likely category based on context or author correspondence.

Functional outcomes

During short-term follow-up (<2 years), functional scores were similar between groups. By ≥10 years, however, repair patients had superior IKDC, KOOS, and Lysholm scores^[13-15,22] reflecting improved pain relief and activity levels. A graphical comparison of key clinical outcomes between meniscus repair and meniscectomy is illustrated in Figure 2.

Close

Figure 2:

Visual comparison figure showing outcomes of meniscus repair versus meniscectomy. Repair: 10–26% versus meniscectomy: 3–8%. Most failures occurred within 24 months. Blue color represents Meniscus Repair and red color represents Meniscectomy.

Export to PPT

Visual inspection of the funnel plot did not demonstrate marked asymmetry. Egger’s regression test was not statistically significant (ρ = 0.18), suggesting no strong evidence of small-study effects. Interpretation remains cautious given the limited number of pooled studies.

Effect measure harmonization

Because included studies reported different effect measures (RRs, odds ratios, and hazard ratios), all estimates were harmonized to a common metric before pooling. RR was selected as the primary summary measure. Odds ratios were converted to RRs using the Zhang and Yu method, adjusting for baseline event rates. Hazard ratios were treated as approximations of RRs under proportional hazards assumptions and log-transformed before inverse-variance pooling. All effect sizes were analyzed on the natural logarithmic scale before back-transformation for presentation.

Because only 10 studies contributed to quantitative pooling, funnel plot asymmetry and Egger’s regression should be interpreted cautiously due to limited statistical power.

Risk-of-bias assessment revealed that most RCTs were at low overall risk, while cohort studies were generally moderate risk due to confounding and performance bias. No study was judged high risk. Figure 3 summarizes the distribution of risk-of-bias ratings. The methodological quality of the included studies was assessed using the Newcastle–Ottawa Scale for cohort studies and the Cochrane Risk of Bias tool for randomized controlled trials. The detailed assessment is presented in Table 2.

Close

Figure 3:

Forest plot of osteoarthritis incidence showing pooled risk ratio 0.52 (95% confidence interval 0.41–0.65) for post-operative osteoarthritis following meniscus repair versus meniscectomy. The analysis includes 10 studies and was performed using a random-effects model. Values <1.0 favor meniscus repair. CI: Confidence interval, OA: Osteoarthritis.

Export to PPT

Certainty of evidence was downgraded where applicable due to moderate risk of bias in cohort studies, heterogeneity (inconsistency), imprecision of CIs, and indirectness related to varying osteoarthritis definitions.

Strengths of the study

This analysis has several notable strengths that enhance the validity and reliability of its findings. First, it incorporates 32 contemporary clinical studies published over a 25-year period, providing a broad and representative overview of modern meniscal surgery practice. The large, pooled sample size of patients increases statistical power and improves the precision of effect estimates. Importantly, both RCTs and large registry-based cohort studies were included, allowing integration of high internal validity evidence with real-world population data. The use of both radiographic and MRI-based OA outcome measures strengthen the robustness of structural assessment and reduces outcome measurement bias. Furthermore, appropriate statistical methodology was applied, including random-effects modeling to account for expected clinical heterogeneity. Finally, stability of the pooled effect was confirmed through sensitivity analysis, demonstrating that the protective association of meniscal repair was not driven by any single study. Collectively, these methodological strengths reinforce the credibility of the present findings. Registry-based studies have demonstrated that patients undergoing meniscectomy have a two- to threefold higher risk of requiring total knee arthroplasty compared with those treated with meniscal repair.^[16-18,31] Consequently, particularly in young and active patients, current evidence supports prioritizing meniscal repair when biologically and technically feasible in order to reduce the risk of osteoarthritis progression and subsequent knee arthroplasty. For young and active patients, the evidence strongly favors repair to avoid OA and later arthroplasty.^{[10,13,20,22,32]}

This study provides an updated and expanded synthesis of contemporary evidence regarding long-term OA risk following meniscal surgery. Several included studies had a moderate risk of bias, which may reduce the certainty of pooled estimates. This was reflected in GRADE ratings.

Future directions

Future research should focus on identifying predictors of favorable surgical versus conservative outcomes, such as herniation size, location, psychosocial variables, and occupational demands. Moreover, cost-effectiveness studies are needed to guide healthcare resource allocation, particularly in low- and middle-income countries where access to surgery may be limited. Long-term follow-up beyond 5 years is also necessary to clarify recurrence rates, risk of chronic pain, and potential development of progressive compartmental degeneration. Finally, the role of minimally invasive and motion-preserving surgical techniques warrants further evaluation in high-quality randomized trials.