Abstract
Introduction
There is considerable interest in the orthopaedic community in understanding the multifactorial process of taper fretting corrosion in total hip arthroplasty (THA). Previous studies have identified some patient and device factors associated with taper damage, including length of implantation, stem flexural rigidity, and head offset. Due to the complexity of this phenomenon, we approached the topic by developing a series of matched cohort studies, each attempting to isolate a single implant design variable, while controlling for confounding factors to the extent possible. We also developed a validated method for measuring material loss in retrieved orthopaedic tapers, which contributed to the creation of a new international standard (ASTM F3129-16).
Methods
Based on our implant retrieval collection of over 3,000 THAs, we developed independent matched cohort studies to examine (1) the effect of femoral head material (metal vs. ceramic, n=50 per cohort) and (2) stem taper surface finish (smooth vs. microgrooved, n=60 per cohort). Within each individual study, we adjusted for confounding factors by balancing implantation time, stem taper flexural rigidity, offset, and, when possible, head size. We evaluated fretting and corrosion using a four-point semiquantitative score. We also used an out-of-roundness machine (Talyrond 585) to quantify the material loss from the tapers. This method was validated in a series of experiments of controlled material removal on never-implanted components.
Results
In the first study, the ceramic cohort exhibited a 92% reduction in cumulative volumetric loss from both the head and neck taper surfaces compared to the CoCr cohort (p < 0.001). In the CoCr cohort, there was greater material loss from femoral head tapers as compared with stem tapers (p < 0.0001). There was also a correlation between visual scoring and volumetric material loss (ρ = 0.67, p < 0.01). In the second study, taper damage was not different between the smooth and microgrooved taper cohorts when evaluated at the head bore (p=0.14) or the stem tapers (p=0.35). There was also no difference in material loss between the most damaged CoCr heads in the two cohorts (p>0.05).
Conclusions
Our findings suggest that fretting and corrosion damage and material loss from the stem taper are mitigated, and on the head taper, eliminated with the use of a ceramic vs. metal femoral head. We also found that fretting and corrosion damage was insensitive to differences in stem taper surface finish and the presence of microgrooves. Although visual scoring was effective for preliminary screening to separate tapers with no or mild damage from tapers with moderate to severe damage, it was not capable of discriminating within the large range of material loss observed at the taper surfaces with high fretting-corrosion scores. Thus, for moderate to severely damaged conical tapers, direct measurement is necessary. A drawback of a matched cohort approach is that a large retrieval collection is necessary to effectively match an investigational group of implants with an appropriate control cohort. Notwithstanding this limitation, the matched cohort approach has been an effective approach to study the complex multifactorial problem of taper fretting and corrosion.
