|Beitragstitel||Radiographic signs or hip pain 5 years after THA: useful in predicting future revision?|
Substantial effort from industry and the research community is directed at improving implant longevity. To identify early and reliable predictors of late failure is essential for both implant safety and innovation.
Our objective was to assess the relation between radiographic signs or hip pain at 5yrs postoperative and future revision for aseptic loosening up to 20yrs postoperative.
Methods: We included all primary THAs operated between 1996-2011 (same uncemented cup, PE-ceramic bearing, 28mm head, cemented stem) for whom baseline and follow-up radiographs were available. At the 5-year clinical visit radiographic signs assessed were: femoral linear/focal osteolysis and/or stem migration ≥2mm. Pain was evaluated with Harris Hip score pain sub-scale (no/occasional vs. mild-severe). Sensitivity, specificity and relative risk (RR, 95%CI) were calculated.
Results: 1054 THAs were included, mean age 70.9yrs, 62% women. Median follow-up was 14.4yrs (range 5.1-20.9yrs). Over that period 287 patients died (27.2%). Radiographic analysis at 5 years revealed 55 cases with femoral osteolysis (5.2%), 112 with stem migration ≥2mm (10.6%), and 158 (15%) had at least one radiographic sign. Twenty-one (2.0%) revisions for aseptic loosening (stem only or total) were performed between 5.2 and 18.8yrs postoperative (median 13yrs). At 5yrs, 14 of the 21 patients later revised had osteolysis/stem migration (Sensitivity 66%, Specificity 88%). RR for future revision in patients with vs. without radiographic signs was 11.3 (95%CI 4.8-29.5). In patients under 65 years of age 12 of the 15 patients later revised in this age group had osteolysis/stem migration (Sensitivity 80%, Specificity 88%). RR for future revision was 19.2 (95%CI 6.5-81.5).
At the 5-year clinical visit, 120 of the 1054 THA patients (11.3%) reported mild to severe hip pain. Among the 21 patients later revised 7 reported hip pain (Sensitivity 33%, Specificity 88%) at 5 years. RR for future revision in patients with vs. without pain was 3.7 (95%CI 1.4-8.7).
Conclusion: In a prosthesis with known excellent long-term survival radiographic signs were already present 5 years postoperative in two-third of the patients revised for aseptic loosening over the following 14 years. This proportion increased to 80% in patients under 65 years of age. Pain was not a useful predictor of long-term failure.