Abstract
Background:
The integrity of articular cartilage in patients with osteochondral lesions of the talus (OLTs) guides treatment. The ability to predict cartilage integrity in OLTs, as previously published for OCD of the knee1, would be beneficial. The purpose of this study is to evaluate the association of radiographic and MRI findings and articular cartilage integrity at the time of ankle arthroscopy for OLTs.
Methods:
A single-institution retrospective review identified patients 19 years of age and younger with operative treatment of OLTs from 2010 – 2017. Demographics and intra-operative findings at the time of ankle arthroscopy were identified by chart review. Radiographs were assessed for physeal status, OLT location, and Berndt and Hardy grade. MRIs were reviewed for OLT size and location, modified Kramer grade, and cartilage status.
Results:
53 patients with 54 OLTs and a mean age of 13.6 years (range 7 – 19 years) were included. OLTs were located in the posteromedial talus in 39 patients (72%). Physeal status was closed/closing in 32 patients (59%) and open in 22 patients (41%). On MRI, the cartilage was predicted to be disrupted in 40 patients (74%) and intact in 14 patients (26%). At the time of ankle arthroscopy, the cartilage was found to be disrupted in 38 OLTs (70%) and intact in 16 OLTs (30%). Table 1 lists variables predictive of cartilage integrity.
Summary of Cohort: Overall and by Cartilage Integrity Determined by Arthroscopy
| Variable | Level | N | Overall | Disrupted | Intact | P-Value |
|---|---|---|---|---|---|---|
| N=54 | N=38 | N=16 | ||||
| Physeal Status | closed | 54 | 27 (50%) | 22 (57.89%) | 5 (31.25%) |
|
| closing | 5 (9.26%) | 5 (13.16%) | 0 (0.00%) | |||
| open | 22 (40.74%) | 11 (28.95%) | 11 (68.75%) | |||
| Radiographic Grade | 1 | 54 | 14 (25.93%) | 4 (10.53%) | 10 (62.5%) |
|
| 2 | 16 (29.63%) | 11 (28.95%) | 5 (31.25%) | |||
| 3 | 20 (37.04%) | 19 (50%) | 1 (6.25%) | |||
| 4 | 4 (7.41%) | 4 (10.53%) | 0 (0.00%) | |||
| MRI Grade | 1 | 54 | 14 (25.93%) | 2 (5.26%) | 12 (75%) |
|
| 2 | 25 (46.3%) | 22 (57.89%) | 3 (18.75%) | |||
| 3 | 14 (25.93%) | 13 (34.21%) | 1 (6.25%) | |||
| 4 | 1 (1.85%) | 1 (2.63%) | 0 (0.00%) | |||
| Cartilage on MRI | Disrupted | 54 | 40 (74.07%) | 36 (94.74%) | 4 (25%) |
|
| Intact | 14 (25.93%) | 2 (5.26%) | 12 (75%) |
MRI classification of cartilage integrity was 95% sensitive and 75% specific for arthroscopic integrity, with 11% misclassification. In the 16 patients less than 13 years, MRI perfectly predicted arthroscopic cartilage integrity. In patients with open physes, MRI sensitivity was 92% and specificity 100%. In patients with closing/closed physes, MRI sensitivity was 89% and specificity 50%.
Receiver operator curve characteristics of a model to predict arthroscopic cartilage integrity combining MRI cartilage integrity, physeal status, and radiographic grade has an AUC of 0.955.
Conclusions:
Physeal status, radiographic grade, MRI grade, and cartilage integrity on MRI are independent predictors of cartilage integrity at the time of ankle arthroscopy for patients with OLTs. Overall, MRI has 95% sensitivity and 75% specificity for cartilage integrity at the time of arthroscopy, which improves to near 100% sensitivity and specificity in patients with open physes. A model combining MRI cartilage integrity, physeal status, and radiographic grade has the highest predictability of intra-operative cartilage integrity. SIegaSll E, Faust JR, Herzog MM, Marshall KW, Willimon SC, Busch MT. Age predicts disruption of the articular surface of femoral condyles in knee OCD: can we reduce usage of arthroscopy and MRI? J Pediatr Orthop 2018;38:176-180.