Hip revisions with impacted morselized allograft bone and cement. Patient outcome, prosthetic fixation and risks

Sammanfattning: 144 consecutive hip revisions performed with morselized allograft bone and cement were followed up to 5 years after surgery. In 49 hips patient outcome was evaluated with the Nottingham Health Profile (NHP) questionnaire and Charnley hip scores in 47 hips prosthetic fixation was evaluated with radiostereometry (RSA)and in all 144 hips risks and complications were recorded using a prspective protocol. Patient outcome regarding pain, physical mobility and sleep at 4 years after revision was comparable to that of primary hip replacements. The correlations between NHP and Charnley pain and physical mobility scores were weak to moderate. RSA index examination should be done as early as possible after surgery. The Exeter stem subsides within the cement mantle and in revisions additional subsidence of the stem-cementbeam occurs relative to the femur. All stems and most sockets migrated, mostly early after revision, without correlation to the grade of bone stock deficiency; one third of the prosthetic components still migrated between 1.5 and 2 years postoperatively. 1 of 15 stems had aseptic loosening according to RSA while the other had migrated only marginally between the 2-year and 5-year follow-up examinations. 5 of 17 sockets had signs of radiographic loosening (radiolucent line >2mm; these sockets were also among those showing the most pronounced migration and rotation at the 5-year follow-up. Neither for the stems nor for the sockets did early migration pattern predict future prosthetic movements. In a pilot study free weight bearing was practised and the migration of the stem and socket was similar to that shown among patients who practised restricted weight bearing for 3 months after revision. Concomitant disease, large bone stock deficiency, intraoperative femoral cortical window or perforation increased the risk for femoral fracture but all 9 hips but could be managed without lasting morbidity.