Introduction: The treatment of asymptomatic osteonecrosis of the femoral head (ONFH) is controversial. The primary aim of this study was to define the optimal management of osteonecrotic lesions in patients with asymptomatic ONFH by determining the incidence of disease progression and the factors that might predict its occurrence. In order to assess the indications and timing for surgical intervention in these patients, the secondary aim was to determine whether or not pain precedes subchondral fracture in patients with asymptomatic disease.

Methods: The subjects in this study were patients with asymptomatic ONFH who were derived from two separate prospective, institutional review board-approved investigations in our institution. We determined the incidence of pain development and radiographic evidence of fracture and the temporal relationship of these events. Statistical analyses were performed to determine what factors affected either radiographic progression or the appearance of symptoms.

Results: Of the 37 hips, 12 (32%) were symptomatic at 2 years. Of these painful hips, six (50%) were associated with the simultaneous presence of a subchondral fracture. When analyzing the relationship of pain with fracture, 5 of 6 hips developed symptoms at an average of 8.1 months (1 to 28 months) prior to fracture. Three symptomatic patients had spontaneous resolution of the ONFH. Cox regression analysis revealed that an index of necrosis of > 50 and a greater extent of radiographic involvement correlate with a higher risk for developing symptoms and a subchondral fracture. If an index of necrosis of 50 is set as the lower limit for intervention, 78% of hips that fractured and 93% of hips that did not were identified.

Discussion: Asymptomatic ONFH with small lesions are amenable to observation, and intervention may be withheld until the appearance of symptoms. Asymptomatic ONFH with extensive femoral head involvement has a high probability of early progression to symptomatic ONFH and subchondral fracture. In these cases, early intervention may be beneficial in preventing fractures which may occur without any preceding symptoms. An index of necrosis of 50 is proposed as a threshold for intervention, as it is a good discriminator between those that did and did not fracture, and had a positive predictive value of 77.8%. The only independent predictor of both pain and collapse was the extent of femoral head involvement.