MRI-Based Prediction of Meniscal Tear Repairability Demonstrates Limited Accuracy and Reliability

Christopher T. Holland, Shannon Tse, Cyrus Bateni, Dillon Z. Chen, Cassandra A. LeeJournal of Clinical Medicine 2025 · DOI

TL;DR

This study found that a simplified MRI-based scoring system using common arthroscopic criteria is not accurate enough to reliably predict whether a meniscal tear can be repaired versus needing to be resected. The variability among reviewers highlights the need for more advanced techniques to improve preoperative decision-making for meniscal injuries.

What Problem Does This Paper Address?

Knee meniscal tears are a common injury, often treated with either repair or partial meniscectomy (removal of the damaged part). While Magnetic Resonance Imaging (MRI) is excellent at detecting these tears, the decision to repair or resect is typically made during surgery based on intraoperative assessment. This decision is influenced by various factors, including tear morphology, location, and the surgeon's experience, leading to variability in treatment. The paper aims to evaluate if a straightforward MRI-based scoring system, using criteria commonly considered during surgery, can reliably predict repairability before the operation, potentially improving preoperative planning and patient counseling.

Methods

The researchers conducted a retrospective review of 202 meniscal tears from patients who underwent either meniscectomy or repair between 2010 and 2018. Preoperative MRIs were independently reviewed by two radiologists and one orthopedic sports surgeon. These reviewers assessed each tear based on four established arthroscopic criteria, assigning one point for each:

  1. Proximity within 4 mm of the meniscosynovial junction (vascularized area).
  2. Tear length greater than 10 mm.
  3. Presence of an intact inner meniscal segment.
  4. Tear involving more than 50% of the meniscal thickness.

A total score of 4 points indicated a predicted "repairable" tear, while a score less than 4 indicated "irreparable." The study then calculated accuracy, sensitivity, positive predictive value (PPV), and negative predictive value (NPV) for each reviewer's prediction against the actual surgical outcome. Inter- and intraobserver reliability were assessed using kappa statistics to quantify agreement. The predictive performance of each individual criterion was also analyzed.

Datasets

The study utilized a dataset of 202 preoperative knee MRIs from patients who underwent meniscectomy or meniscal repair at a single academic center. These MRIs were acquired using 1.5 T or 3 T MRI scanners with proton density and fat saturation sequences. No new datasets were introduced; the study relied on existing patient data.

Key Insights and Ideas

  • MRI Alone Isn't Enough (Yet): The core finding is that even with a structured scoring system based on common surgical criteria, MRI alone struggles to reliably predict meniscal tear repairability. This suggests that the nuances of tear healing and the dynamic nature of the intraoperative assessment are not fully captured by these static imaging features.
  • Inter-Reviewer Disagreement is High: The poor to moderate interobserver reliability (agreement between different reviewers) is a significant finding. This variability means that even experienced clinicians looking at the same MRI might come to different conclusions about a tear's repairability, underscoring the subjective nature of MRI interpretation in this context.
  • Individual Criteria Fall Short: Analyzing each criterion separately revealed that most individual factors had limited predictive power, often achieving less than 50% accuracy. This implies that no single feature is a strong enough indicator of repairability.
  • ACL Injuries Don't Dramatically Improve Prediction: A subgroup analysis of tears with concomitant ACL injuries showed only a minimal improvement in predictive accuracy, suggesting that the challenges in predicting meniscal repairability persist even in more complex knee injury scenarios.

Key Results

  • Overall Accuracy is Modest: Reviewer accuracy in predicting repairability ranged from 48% to 76%, with a mean accuracy of 66%.
  • Reliability is Lacking: Interobserver reliability was poor to moderate (pairwise kappa ranging from 0.07 to 0.43, Fleiss' kappa of 0.11). Intraobserver reliability (agreement within the same reviewer over time) was better, ranging from moderate to substantial (kappa 0.42-0.66).
  • Predictive Values Vary: While negative predictive values (NPV) were relatively consistent (73-79%), positive predictive values (PPV) were lower (37-70%), indicating a tendency to overpredict repairability.
  • Individual Criteria Show Weak Performance: Most individual criteria achieved less than 50% accuracy across reviewers.

Limitations and Future Work

The authors acknowledge several limitations:

  • Retrospective Design: This introduces potential biases in selection and information, and intraoperative documentation of the scoring criteria was not standardized.
  • Surgeon Variability: Treatment decisions were left to the discretion of four different surgeons, potentially introducing inter-surgeon heterogeneity in repair thresholds.
  • MRI Scanner and Protocol Variability: The use of both 1.5 T and 3 T scanners without a standardized protocol could introduce variability.
  • Lack of Clinical Factors: The study focused solely on MRI findings, excluding patient-specific factors like age, which can influence surgical decisions.

Future work is suggested to focus on:

  • Advanced Imaging Techniques: Exploring modalities beyond standard MRI.
  • Artificial Intelligence (AI): Developing deep learning models trained on large datasets to improve accuracy and reduce interobserver variability. AI could potentially integrate multiple clinical and imaging variables for more individualized predictions.
  • Ultra-High Field MRI (7 T): Investigating if higher field strength MRIs can provide superior detail for assessing repair-relevant features.
  • Differentiating Repairability from Healing: Future research should distinguish between predicting whether a tear can be repaired versus whether it will ultimately heal, potentially by incorporating biological and vascular features.

Summary updated

Abstract

<b>Background:</b> While magnetic resonance imaging (MRI) is commonly used to identify meniscal tears, intraoperative assessment typically dictates repairability. This study evaluated whether a simplified MRI-based scoring system could reliably predict meniscal repair versus meniscectomy. <b>Methods:</b> Patients who underwent meniscectomy or meniscal repair between 2010 and 2018 were retrospectively identified. Preoperative MRIs were independently reviewed in a blinded fashion by two radiologists and one orthopedic sports surgeon. Reviewers scored images based on four arthroscopic criteria for tear repairability, with one point awarded for each of the following criteria-(1) proximity within 4 mm of the meniscosynovial junction, (2) length > 10 mm, (3) presence of intact inner meniscal segment, and (4) >50% meniscal thickness. Tears scoring four points were considered repairable. Accuracy, sensitivity, and positive and negative predictive values were calculated against the actual procedure performed. Inter- and intraobserver reliability were evaluated using kappa statistics. The predictive performance of each individual criterion was also analyzed. <b>Results:</b> A total of 202 meniscal tears were included (134 meniscectomies and 68 repairs). Reviewer accuracy in predicting repairability ranged from 48% to 76%. Intraobserver reliability was moderate to substantial (κ = 0.42-0.66), whereas interobserver reliability was poor to moderate (pairwise κ = 0.07-0.43; Fleiss' κ = 0.11). Analysis of individual MRI criteria demonstrated limited predictive value, with most criteria achieving less than 50% accuracy across reviewers. <b>Conclusions:</b> MRI-based prediction of meniscal repairability using arthroscopic criteria demonstrated limited accuracy and poor interobserver reliability. Overall predictive reliability remains insufficient for clinical decision-making. Further investigation, integrating advanced imaging techniques and artificial intelligence, may improve the preoperative assessment of meniscal repairability.
BibTeX
@article{holland2025mri,
  author = {Christopher T. Holland and Shannon Tse and Cyrus Bateni and Dillon Z. Chen and Cassandra A. Lee},
  title = {{MRI-Based Prediction of Meniscal Tear Repairability Demonstrates Limited Accuracy and Reliability}},
  journal = {Journal of Clinical Medicine},
  year = {2025},
  doi = {10.3390/jcm14124160},
}
Figure