Early Subsidence Predicts Failure of a Cemented Femoral Stem With Minor Design Changes.

BACKGROUND: Radiostereometry (RSA) measurements of early micromotion can predict later failure in hip and knee prostheses. In hip implants, RSA has been particularly helpful in the evaluation of composite-beam stem designs. The Spectron EF Primary stem (Smith & Nephew, London, UK) has shown inferior performance compared with its predecessors in both clinical studies and registry reports. Early RSA studies have shown somewhat greater subsidence for the Spectron EF Primary stem compared with the earlier Spectron EF, but still within boundaries considered to be safe. QUESTIONS/PURPOSES: Our primary research question was whether stem subsidence and rotation for this stem design measured with RSA at 2 years can predict later stem failure. A secondary question was whether high femoral stem offset and small stem sizes, both features specific to the Spectron EF Primary stem compared with its predecessors, are associated with stem failure rate. METHODS: Two hundred forty-seven hips (209 patients with median age 63 years [range, 29-80 years], 65% female, and 77% primary osteoarthritis) with a valid RSA examination at 2 years were selected from four different RSA studies (totaling 279 hips in 236 patients) in our department. The studies were primarily aimed at evaluating cup fixation, bone cement, and polyethylene types. All study patients received a cemented Spectron EF Primary stem. The selected hips had complete followup until stem failure, death, or the end of the followup period. Stem failure was defined as revision of a loose femoral stem or radiological failure with significant osteolysis in Gruen zones 2 to 6. Cox regression analyses were performed to evaluate if stem subsidence and rotation after 2 years, adjusted for age, sex, stem size, standard/high stem offset, and conventional/highly crosslinked polyethylene, could predict later clinical aseptic failure of the stem. We identified 32 stem failures (27 revisions, five radiological failures) at 14 years median followup (range, 3-18 years). Ten-year stem survival was 94% (95% confidence interval [CI], 90%-96%). RESULTS: Stem subsidence at 2 years (adjusted hazard ratio [HR], 6.0; 95% CI, 2.5-15; p < 0.001) and retrotorsion of the stem (adjusted HR, 1.7; 95% CI, 1.1-2.5; p = 0.018) were associated with later stem failure. Further risk factors were male sex (subsidence analysis HR, 6.9; p > 0.001), use of the two smallest stem sizes (HRsize 1, 8.0; p > 0.001, HRsize 2, 1 [reference], HRsize 3+, 0.06; p = 0.035), and the high offset option (HR, 3.1; p = 0.005). CONCLUSIONS: Stem subsidence and retrotorsion at 2 years can predict later failure in the Spectron EF Primary stem, consistent with earlier findings on composite-beam cemented stems. Small stem size and high-offset stems comprise the main group of underperforming stems. We recommend that premarket small-scale RSA studies be performed after any design change to a THA femoral component, because even seemingly minor design changes may unexpectedly result in inferior performance. LEVEL OF EVIDENCE: Level III, therapeutic study.