Periprosthetic fractures around the knee (TKAPF) often occur in old patients with concomitant diseases. Like any other fragility fracture, the aim of TKAPF management should be a patient-specific strategy considering clinical factors and associated comorbidities. The aim of the present study was to analyse if, in selected groups of unfit patients, osteosynthesis alone without revision even in cases of implant loosening (SOFCOT type 2 and 3) could be a good compromise to avoid invasive surgery and to reduce the mortality rate with reasonable results. The hypothesis was that for TKAPF SOFCOT type 2 and 3 there was no difference in survival rate between a group of patients treated by open reduction and internal fixation only compared to a group treated by revision of existing implant.

A total of 62 patients with a TKAPF and a loose implant classified as SOFCOT type 2 or 3 who have been surgical treated were included. The mean age of the population was 81.1 ± 10.3 years. Female rate was 92% (n = 57). The majority of patients had an ASA 2 or 3 score (n = 52, 83.9%). The pre-operative Parker score was 6.1 ± 2.5. Two groups were created to be compared in terms of results and complications: a group of 18 patients operated by open reduction and internal fixation (ORIF group) with a plate and a group of 44 patients operated by total knee arthroplasty revision (revision group) with or without plate osteosynthesis associated. Mortality and re-operations for mechanical complications were examined as the primary endpoint.

Time to full weight-bearing (in weeks), time to bone union (in weeks) other complications such as infection and haematoma rate were collected as secondary endpoint.

The median duration of follow-up was 2 years. The re-revision rates were respectively 15,9% in revision group, 22,2% in ORIF group. No statistically significant differences were found between the groups for mechanical complications or re-revision. The survival endpoint at 2 years follow up was 88.9% (95% CI: 78.1−94.5) without differences between the groups (p = 0.64). Time to bone union was significantly longer (15.9 weeks) in the ORIF group compared to 8.6 weeks for the revision group but with a higher bone union rate in the ORIF group (p = 0.003) as well as the time to full weight-bearing (ORIF group 8.9 weeks compared to 1.7 weeks for the revision group P < 0.001). More infections were reported in the ORIF group and a higher risk of haematoma and bleeding in the revision group.

Despite longer time to bone union and time to full weight-bearing, open reduction and internal fixation with plate for TKAPF with a loose component in selected older and unfit patients does not increase the risk of re-operation or mortality compared to a more invasive revision arthroplasty strategy.

III, comparative retrospective study.