FRETTING AND CORROSION AT THE HEAD-NECK JUNCTION OF WELL FUNCTIONING TOTAL HIP ARTHROPLASTIES RETRIEVED AT AUTOPSY

INTRODUCTION Adverse local tissue reactions (ALTR) and elevated serum metal ion levels secondary to fretting and corrosion at head-neck junctions in modular total hip arthroplasty (THA) designs have raised concern in recent years. Factors implicated in these processes include trunnion geometry, head-trunnion material couple, femoral head diameter, head length, force of head impaction at the time of surgery, and length of implantation. Our understanding of fretting and corrosion in vivo is based largely on the analysis of retrieved prostheses explanted for reasons related to clinical failure. Little is known about the natural history of head-neck tapers in well-functioning total hip replacements. We identified ten well-functioning THA prostheses retrieved at autopsy. We sought to determine the pull-off strength required for disassembly and to characterize fretting and corrosion apparent at the head-neck junctions of THAs that had been functioning appropriately in vivo . METHODS Ten cobalt-chromium femoral stems and engaged cobalt-chromium femoral heads were retrieved at autopsy from 9 patients, after a mean length of implantation (LOI) of 11.3 ± 8 years (range 1.9–28.5). Trunnion design and material, femoral head material, size, and length, LOI, and patient sex were recorded (Table 1). Femoral heads were pulled off on a uniaxial load frame according to ASTM standards under displacement control at a rate of 0.05mm/s until the femoral head was fully disengaged from the trunnion. Mating surfaces were gently cleaned with 41% isopropyl alcohol to remove any extraneous debris. Femoral trunnions and head tapers were examined under a stereomicroscope by two independent graders to assess presence and severity of fretting and corrosion (method previously established). Trunnions and tapers were divided into 8 regions: anterior, medial, posterior, and lateral in both proximal and distal zones. Minimum possible damage score per hip was 32 (indicating pristine surfaces). The total possible score per hip was 128 (2 damage modes × 2 mating surfaces × 8 regions × max score of 4 per region). RESULTS Mean pull-off force among all retrievals was 2446 ± 841 N (1655 – 4246 N). Mean pull-off force for 14/16 tapers (2998 ± 1298 N) was larger than for 12/14 tapers (2210 ± 531 N). Seven retrievals (70%) had no evidence of damage on either the stem or head component (Fig. 1). Three retrievals showed evidence of damage: (1) corrosion in one zone of the femoral head taper (score 33); (2) a circumferential ring of fretting in one zone of the stem trunnion (score 36); (3) circumferential rings of minor fretting in two regions of the stem trunnion (score 40). LOI for damaged retrievals was 16.3 ± 6 years, longer than that for undamaged retrievals (9.1 ± 9.1 years). CONCLUSION THAs that had been well-functioning in vivo showed little evidence of fretting and corrosion. The presence of minor fretting and corrosion correlated with increased LOI. Mean pull-off force was 2446 +/- 841 N among the complete sample of ten THAs. Larger tapers were associated with greater average pull-off strength. Further investigation is required in order to clarify the clinical implications of these results.