Elderly patients with a high mortality risk for revision surgery are severely handicapped by a loosened hip prosthesis. Loosening is mainly caused by particle-induced osteolysis leading to the formation of a synovium-like interface tissue. As an alternative to revision surgery we have investigated the possibility of removing the tissue using a gene therapy approach and thereafter stabilizing the prosthesis with percutaneous cement injection.

First we demonstrated that transduction of interface cells with a gene coding for E.coli nitroreductase (NTR) resulted in a 60-fold increase in sensitivity to the prodrug CB1954 that is converted to a toxic metabolite by NTR. Given these in-vitro data, we explored if intra-articular administration of this adenoviral vector encoding NTR followed by the prodrug was able to kill sufficient tissue in-vivo to allow refixation of the prosthesis by cement.

We report the first three patients from a phase 1 study of 12 patients with a loosened hip who are experiencing debilitating pain and have significant comorbidity. On day 1 the vector is injected into the hip joint and on day 3 the prodrug is injected. On day 10 three holes are drilled in the femur and one in the acetabulum. Biopsies are taken from the periprosthetic space and low viscosity cement (Osteopal, Biomet Merck, Sjöbo, Sweden) is injected under fluoroscopic guidance.

The first three patients have been included in the study and five more are planned for treatment before June 2005. The patients are females of 86, 72, and 79 years old. There were no adverse effects from vector injection (3x10 exp 9 particles). Six hours after prodrug injection the patients experienced nausea, (WHO grade 1) a commonly reported reaction to this prodrug. There was vomiting in two patients. Hip pain increased, but this was anticipated as this therapy will increase prosthesis loosening. 16 ml of cement was subsequently injected into the periprosthetic space in the first patient and 18 ml in the second. The patients were ambulated the day after surgery.

The first two patients have a follow-up of twelve and six weeks. There was no pain in the hip. The maximum walking distance had increased from 5 to 30 meters in the first patient.

The current study is the first to use in vivo intra-articular adenoviral mediated gene transfer in a clinical setting. Our preliminary results suggest that gene therapy and cement injection for hip prosthesis refixation is clinically feasible.