MRI OCD Knee

AAOS conducted a systematic review that identified two diagnostic studies^{40, 41} addressing this recommendation. One of these studies evaluated the diagnostic performance of clinical examination with radiographs and of selective MRI in the evaluation of intra-articular knee disorders by comparing these findings with arthroscopic findings.⁴⁰ The clinical diagnosis (from the initial visit), MRI diagnosis (from the MRI report), and the arthroscopic diagnosis (from the operative report) were retrospectively reviewed and compared. Since only a subset of all patients that underwent evaluation of intra-articular knee disorders proceeded to arthroscopic evaluation, this diagnostic study does not universally apply the reference standard of arthroscopy. Consequently, this retrospective diagnostic study without a universally applied reference standard was evaluated as a Level II study.

Similarly, the second diagnostic study identified in the systematic review, prospectively evaluated all consecutive patients undergoing knee arthroscopy who had a preoperative MRI.⁴¹ Again, this study only reports on the subset of patients that required surgery; therefore, this diagnostic study does not universally apply the reference standard of arthroscopy. Consequently, this prospective diagnostic study without a universally applied reference standard is also evaluated as a Level II study.

These Level II studies, when considered together, may have supported a moderate strength of recommendation. However, these studies found that both x-ray and MRI are good rule in tests and do not address the incremental diagnostic value of an MRI in the setting of known OCD determined by x-ray. That is, these studies do not compare the diagnostic performance of clinical examination with standard radiographs to clinical examination with standard radiographs and an MRI; therefore we downgraded the strength of this recommendation to limited.

In addition to identifying the presence of OCD lesions and distinguishing OCD lesions from other intra-articular pathology, an MRI may be used as an adjunct to clinical examination with radiographs to provide additional information that will guide therapeutic decision-making. Of the 5 therapeutic studies^42-46 that were included in the development of this guideline, three studies ^42-44 report the acquisition of an MRI at enrollment and three studies^{42, 44, 45} report the acquisition of an MRI at follow-up evaluation. Further, one prognostic study¹⁹ predicts the healing potential of stable OCD lesions, utilizing a multivariable logistic regression model. Of all of the variables that were considered (including sex, side, location, symptoms, knee dimensions, and lesion dimensions), only knee symptoms as well as normalized length and normalized width of the OCD lesion as measured on MRI were found to be predictive of healing potential.

Of note, three studies^47-49 correlated MRI findings with arthroscopic findings in patients with OCD of the knee. The evidence for assessment of stability of an OCD lesion was inconsistent.

Supporting Evidence

A single study assessed the pre-operative diagnosis of a pediatric orthopaedic surgeon, which included clinical examination, radiographs, and consideration of the MRI findings.⁴¹ This study enrolled 131 patients with various knee lesions, 19 of which were diagnosed as having OCD during arthroscopic examination. Diagnostic performance estimates from this study reflect the value of a pediatric orthopaedic surgeon’s pre- operative diagnosis to correctly identify OCD from several other lesions. Analysis of likelihood ratios (LR) and associated confidence intervals indicates that diagnosis based on exam, x-rays, and MRI findings is a good rule in and a good, moderately good, or weak rule out test for OCD (XTable 7X). However, the use of a single surgeon’s pre- operative diagnosis reduces the generalizability of these results.

Two studies evaluated the ability of MRI to distinguish OCD from several other lesions.^{40, 41} The studies enrolled 256 patients with various knee lesions, 41 of which were diagnosed as having OCD during arthroscopic examination. Likelihood ratios and the associated confidence intervals indicate MRI is a good or moderately good rule in test and a good, moderately good, or weak rule out test for OCD ( able 8).

In the three remaining studies (n = 124), MRI was evaluated for the ability to diagnose instability of the osteochondritis dissecans.^47-49 Instability at MRI was based on similar criteria, including high signal rims/lines, cysts, and focal defects (XTable 9). One study reported the results of skeletally mature patients separately from skeletally immature patients.⁴⁸ Ninety-one percent (91%) of the patients in one study were skeletally immature ⁴⁹ and 81% of the patients skeletally mature in the remaining study.⁴⁷ Thus, we analyzed the likelihood ratios and the associated confidence intervals for skeletally immature patients and skeletally mature patients separately.

The analysis in skeletally immature patients indicates MRI is good, moderately good, weak, or poor as a rule in and rule out test for instability of OCD (Table 10)X   MRI is a good, moderately good, weak, or poor rule in test for OCD instability in skeletally mature patients and a good, moderately good, or weak rule out test for OCD instability in skeletally mature patients ( able 11)