Introduction

Total hip replacements using highly cross-linked polyethylene show excellent clinical outcomes, low wear, and minimal lysis at 5 years follow-up. A recent RSA study reports a significant increase in femoral head penetration between 5 and 7 years. This study is a multi-center radiographic analysis to determine whether the RSA observation is present in a large patient cohort.

Studies of patients having primary THR using highly cross-linked polyethylene show excellent clinical outcomes and very low radiographic wear results at a minimum of 5 years follow-up. Recently, a radiostereometric analysis (RSA) study of a small group of patients reported that after no detectable wear during years 1–5, they found a significant increase in femoral head penetration between 5 and 7 years follow-up. However, this increase in head penetration after 5 years has not been confirmed in a larger patient cohort.

The purpose of this study was to organize a multicenter radiographic study involving leading medical centers in the U.S. having the longest-term follow-up available on this type of highly cross-linked polyethylene in order to determine if the RSA observation can be confirmed in a larger study.

Six academic centers agreed to contribute radiographic data to this study. All patients received primary total hip replacements with Longevity polyethylene liners (Zimmer, Warsaw, IN) coupled with 26, 28, and 32mm cobalt chrome femoral heads. The radiographic inclusion criteria required a minimum of four radiographs per patient: one at 1 year; at least one from 2 to 4.5 years; one 4.5 to 5.5 years; and at least one from 5.5 to 9 years follow-up. The Martell Hip Analysis Suit-eTM software was used for the wear analysis. All wear values were determined by calculating head penetration between the follow-up radiograph and the 1-year radiograph to remove creep, the majority of which has been shown to occur during the first year. Separate linear regressions, representing the wear rates, were computed for the early period from 1 year to 5.5 years and the late period from 5.5 years to 9 years follow-up. The Zar test was used to determine the significance of the difference between these two linear regressions.

We present the completed analysis of 165 hips. When the early and late data points were combined into one data set, the second-order regression indicated an inflection point at 6.3 years with a slightly positive inflection. There were 402 film comparisons in the early time period, and the slope and confidence interval of the regression line was 4.9μm/yr (95% CI of −28μm/yr to 38μm/yr). There were 188 film comparisons in the late period, and slope of the regression line for the late period was 10.8 μm/yr (95% CI of −58μm/yr to 80μm/yr). The Zar test showed no significant difference between the two slopes (Figure 1, p=0.886).

No significant increase in femoral head penetration was found for the late period after 5 years compared to the early period before 5 years follow-up in either analysis. Additionally, no significant late increase in wear was seen within individuals. While we continue to enroll patients, at this time we do not observe the increase in wear seen in the RSA study after 5 years.

Introduction and Aims: Accumulating evidence suggests that bacterially derived endotoxins may contribute to aseptic loosening. This study determined whether lipopolysaccharide (LPS), the classical endotoxin from Gram-negative bacteria, can be detected in periprosthetic tissue from patients with aseptic loosening. We utilised an assay that detects all forms of LPS and is unaffected by beta-glucan-like molecules.

Method: Periprosthetic tissue from revision total hip arthroplasty and synovia from primary total joint arthroplasty were homogenised in PBS in endotoxin-free conditions. Non-specific amidases in the homogenates were inactivated at 100 degrees C. LPS was measured using the Endospecy assay (Associate of Cap Cod). Multiple dilutions of the homogenates were assayed to maximise sensitivity, while avoiding assay inhibition assessed by spike recovery determinations. Results were corrected for colour and spike recovery. Assay results were considered positive if the absorbances were higher than the lowest standard and the LPS level was significantly greater (p< 0.05) than the PBS control. Statistical analysis was by ANOVA with Bonferroni-Dunn (Control) post-hoc tests.

Results: Samples from 13 patients have been studied to date. Multiple assays of four of these samples showed no detectable LPS while nine of these samples resulted in both positive and negative assays. This inter-assay variability prevents measurement of the concentration of LPS in the samples. Nonetheless, many of the samples contain detectable amount of LPS. Thus, six out of eight samples from revision THA patients with aseptic loosening had positive assays, as did two of four primary TJA patients. LPS was also detected in a sample from a revision control. These results demonstrate that samples from THA patients with aseptic loosening and from primary TJA contain detectable amounts of LPS derived from Gram-negative bacteria.

Conclusion: This conclusion is consistent with numerous studies, showing that human serum contains LPS derived from minor infections, gut flora, or dental procedures. It is likely that many of these samples also contain molecules derived from Gram-positive bacteria that have very similar biological effects as LPS. However, detection of these Gram-positive molecules await further improvements in assay specificity and sensitivity.