A predictive factor in osteochondritis dissecans: A radiographic analysis

Objectives: Knee alignment is thought to have some bearing on the development of osteochondritis dissecans (OCD) lesions.1 The effect of proximal tibial morphology on the risk of OCD, however, is unknown. The purpose of this study was to evaluate proximal tibial morphology and its relationship to OCD lesions. The null hypothesis was that patients with OCD lesions would have no difference in medial and posterior tibial slope when comparing the affected side to the unaffected side and age matched controls. Methods: Using CPT codes, we identified 61 patients with unilateral OCD lesions of the medial femoral condyle seen at our institution from 2005-2010. On plain radiographs, medial tibial slope and posterior tibial slope were assessed by 2 residents and 1 attending. (Figure 1) Measurements were completed on affected, contralateral normal, and control knees. The controls were height, weight, and gender matched. In addition, a comparison was made between OCD patients requiring surgery and those without surgery. Inter-observer reliability for each measurement was determined using intra-class correlation coefficients (ICCs). A student’s t-test was used to compare the results of the affected and normal sides. Results: The average patient age was 15.1 years. There were 31 right-sided lesions and 30 left- sided lesions. Medial tibial slope of the affected knee averaged 67.81 +/- 3.92 (ICC(2,1) =.771, p<.01) compared to 69.44 +/-3.63 (ICC(2,1) = .785, p<.01) for the normal side (p=.0070). The average posterior tibial slope for the affected knee was 80.03 +/- 3.91 (ICC(2,1) = .783, p<.01) and 79.62 +/- 4.37 (ICC(2,1) = .844, p<.01) for the normal side (p=.19). Matched controls had an average medial tibial slope 68.73° +/- 5.81° (ICC(2,1) =.732, p<.01) which was statistically different from affected knees (p=.043). Matched controls had an average posterior tibial slope 81.13° +/- 2.90° (ICC(2,1) = .797, p<.01), which was also statistically different from affected knees (p=.0068). Thirty-one patients with OCD lesions had undergone surgery while thirty patients were treated non-operatively. In the operative group, the average medial tibial slope was 69.17° +/- 4.64° and the average posterior slope was 76.94° +/- 3.55°. The nonoperative group had significantly different values, with a medial slope 66.66° +/- 9.04 and posterior slope 80.28° +/- 3.77. (p=.040 and .0071). Conclusion: For medial OCD lesions, the affected knee had a greater medial tibial slope than the unaffected knee. When compared to matched controls, affected knees had a greater medial and posterior slope. In addition, patients who required surgery for the OCD lesion had greater medial and posterior slopes than did patients who did not require surgery. Future studies will be necessary to determine if these differences are clinically significant.