INDICATIONS FOR REVISION UNICONDYLAR KNEE REPLACEMENT: DO ALIGNMENT AND RADIOLUCENT LINES PREDICT COMPONENT LOOSENING?

Abstract

Survivorship of unicondylar knee replacement (UKR) exceeds 85% at 10 years. During long term follow-up, progressive osteoarthritis (OA) and loosening are typical of UKR failure. The decision to revise UKR is complex as radiographic findings are not always consistent with clinical symptoms. This study of revised UKR compares intraoperative assessment of component fixation and progressive OA with prerevision radiographic evaluations.

Twenty-seven UKR were retrieved from 22 female and 5 male patients. Patient age and time in situ averaged 76 (68–87) years and 79 (25–156) months, respectively. At index arthroplasty, all knees received a fixed-bearing medial UKR with cement fixation. Prior to revision, radiolucent lines and component alignment were assessed on radiographs according to Knee Society guidelines. Suspected revision reasons based on clinical and radiographic evaluation included aseptic loosening (63%), progressive OA (22%), and wear (15%). During revision surgery, component fixation was manually assessed and graded as well-fixed or loose, and progressive OA was graded using Outerbridge classification. Intraoperative and radiographic assessments were completed independently.

Average Knee Society Scores declined > 30 points to 53+18 (pain) and 43+11 (function) before revision. During revision surgery, femoral and tibial component fixation were graded as loose in 19 (70%) and 9 (33%) knees, respectively. There was Grade III or IV progressive OA in the lateral or patellofemoral compartment of 15 (56%) and 16 (59%) knees, respectively. Radiolucent lines were evident in 8 of 19 loose femoral components and 5 of 9 loose tibial components. In contrast, 3 of 8 well-fixed femoral components and 6 of 18 well-fixed tibial components had radiolucent lines. There were 11 loose femoral components and 4 loose tibial components without radiolucent lines. Radiographic limb alignment averaged 3°+3° valgus immediately after index UKR. Change in limb alignment ranged from 0° to 17° at revision. Tibial or femoral component alignment changed 5° to 9° in 12 (44%) knees and > 10° in 5 (19%) knees. Eight of these 17 knees (47%) had malaligned components graded as loose.

The prevalence of progressive OA at revision UKR was more than double occurrence suspected from radiographs. Interpreting radiographic indications for loosening was difficult. Radiolucent lines predicted loosening in 46% (13/28) of the components graded as loose and falsely predicted loosening in 35% (9/26) of the components graded as well-fixed. Radiolucent lines were absent in 15/28 (54%) of the loose components and changes in component alignment > 5° were associated with component loosening in < 50% of the knees. Rigorous attention to clinical symptoms and careful interpretation of radiographic phenomena are needed to determine indications for revision in UKR patients.