Five-year follow-up of socket movements and loosening after revision with impacted morselized allograft bone and cement: a radiostereometric and radiographic analysis.

In 1999, we reported on the 2-year results of a series of 21 first-time socket revisions using impacted morselized allograft bone. Seven still migrated between 1.5 and 2 years. Seventeen remained for the current 5-year follow-up. No socket had been rerevised. Five sockets showed signs of radiographic loosening. These 5 cases also exhibited radiographic signs of allograft resorption as well as high rates of socket migration and rotation as measured by radiostereometric analysis. Of the 6 remaining sockets that had migrated between 1.5 and 2 years, 3 stabilized and 3 were among those with signs of radiographic loosening. Fifteen patients (15 hips) revealed pain reduction at the 5-year follow-up. Three had slight pain on walking that disappeared immediately at rest (pain score 4). All the others revealed either no pain (pain score 6) or slight pain that disappeared with activity (pain score 5). No pattern of early socket migration according to radiostereometric analysis could be identified, predicting later socket migration or loosening. The rate of cases with signs of radiographic loosening (29%, 5/17) was comparable to that reported by the Nijmegen group but the follow-up was shorter in the current study. The rate of cases with signs of radiographic loosening was comparable to both conventionally cemented socket revisions and cementless revisions. The future will show if further sockets loosen and if the loose sockets up to date will end up in rerevisions.