Osteonecrosis (ON) around the knee joint can be spontaneous (primary), without comorbid medical conditions, or secondary. ON is characterized by sudden pain, usually self resolving after six months to a year, night pain, and involving mainly women older than 55 years of age.

Twenty seven patients with primary ON were retrospectively reviewed. The patients were evaluated clinically (physical examination and H.S.S. scoring) and radiographically (plain radiographs were done at the beginning and at the end of follow-up and MRI studies). The MRI study included calculations of the area and the volume of the lesion, osteoarthritic changes, bone collapse and meniscal pathologic findings.

The majority of the patients, 96% (26/27), complained of sudden pain. Night pain was observed in 46% of the patients. Osteoarthritic changes observed in 60% of the patients and further deterioration (37%) was noted at follow up radiographs. Osteoarthritic changes were more predominant in patients with tibial condyle involvement rather than those with femoral involvement, 61% vs. 43% (p=0.04), respectively. Diagnosis of osteonecrosis by the initial radiographs was more common in patients with femoral involvement (7 femurs vs. 1 tibia).

All ON lesions were diagnosed by MRI, from these, 17 meniscal tears (63%) were detected.

Several observations were noted in the volumetric analysis: Tibial lesions were larger than femoral lesions (8.1cm³ vs. 3.1cm³, respectively, p=0.026). Women had significantly larger lesions in volume and area than men (area – 5.09cm² vs 2.05cm², p=0.01; and volume – 6.6cm³ vs. 1.2cm³, p=0.001). Significant correlation was found between bone collapse and the need for total knee replacement (TKR): 40% collapse in the TKR patients versus 13.6% collapse without TKR (p=0.028). Patients that were diagnosed with collapse at the end of the follow up had a larger area (5.8cm² vs 3.7cm²) and a lower HSS score (68.5 points vs. 83 points, p=0.02) than those without. A significant correlation was noted between a larger lesion area and lower HSS scoring (p=0.037) at the end of the follow-up. Patients with tibial lesions had more meniscal pathologies involvement, 56.3% vs. 43.8%, however it was not statistically significant (p=0.68).

We conclude that MRI study for patients with ON and calculating the area and the volumetric data particularly, is an accurate predictive tool for the natural history of knee ON. The findings presented here may guide us in determining the best possible treatment.