Looking for Source of Knee Implant Failure? Biomarkers May Help

SAN FRANCISCO — Tests for certain metals could help clinicians determine why knee replacements appear to be failing, according to preliminary studies reported here.

For patients with metallic knee implants, high serum levels of cobalt and chromium ions may indicate that the devices are wearing down excessively, an explant analysis indicated.

Alternatively, a separate study found, barium and zirconium in synovial fluid could flag aseptic loosening in cemented knee arthroplasties. Both studies were presented at the American Academy of Orthopaedic Surgeons annual meeting.

Authors of both studies noted that when knee arthroplasty patients begin to complain of increasing pain and/or instability, it can be difficult to figure out why. Advanced imaging may be needed and even then it might not provide a definitive answer. That may come only with revision surgery, when physicians can lay eyes and hands on the failing device. Thus, a blood test that can identify likely candidates would be welcome.

David Langton, PhD, director of the ExplantLab in Newcastle-upon-Tyne in England, and colleagues conducted the study on cobalt-chromium release from knee implants. The group was following up on earlier reports that these metals were showing up in the blood of knee arthroplasty patients. Those findings brought to mind the previous decade’s controversy over metal-on-metal hip replacements, which prompted the field to move away from such devices. Nevertheless, Langton’s group noted “skepticism” over whether circulating metal ions lead to hypersensitivity reactions that eventually cause a form of device rejection.

The researchers analyzed 90 cobalt-chromium knee implants, of three different design types, that had been removed from patients after an average of 3-4 years. The most common reasons for removal were increasing pain and/or loosening. Patients were mostly in their 50s to 70s.

Nearly 60% of the tibial trays showed significant pitting and, in some cases, “gross abrasive changes,” the group reported. Among 30 patients with periprosthetic tissue samples taken, 10 showed at least mild aseptic lymphocyte-dominant vasculitis-associated lesions (ALVAL), a sign of hypersensitivity reactions. Device trays in all the ALVAL cases showed pitting or were no longer fitting snugly with other components.

Forty of the patients had serum cobalt and chromium ion measurements. Medians were, respectively, 2.5 and 1.7 μg/L — but with ranges up to almost 70 μg/L for cobalt and 13 μg/L for chromium. Langton’s group calculated that no more than about 2.5 mm³ of metal had eroded from the tibial trays, which would not be enough to account for such high levels. For individuals with blood levels above 4 μg/L, component loosening or metal-on-metal contact must be contributing, the researchers surmised. “We recommend routine metal ion measurements” for patients complaining of suboptimal performance of their knee implants, they concluded.

In the second study, led by Aleksander Mika, MD, of Vanderbilt University in Nashville, Tennessee, the investigators wondered whether certain components of bone cement might indicate that a knee implant is coming loose. A device stem rattling around within the bone would presumably mobilize cement particles that could then be detected in the synovium. Among the constituents of bone cement are barium sulfate and/or zirconium oxide, used as so-called radiopacifiers to make the cement stand out on x-rays.

The researchers analyzed 20 cases in which cemented knee implants had been removed for revision. Of these, seven had definitely come loose; in the other 13 cases, the devices were found to be well-fixed, despite problems such as increasing pain, infection, implant disruption, overt instability, and patellar complications.

Although these were small numbers, dramatic differences in synovial barium and zirconium levels were seen between the two sets of cases. Mean barium levels stood at 314 μg/L when the devices had loosened, compared with just 32 μg/L when the devices remained fixed. Similarly, zirconium averaged 45 μg/L with loosened devices as opposed to exactly 0 when the implants remained firmly in place.

In terms of diagnostic performance, detection of zirconium showed sensitivity of 57% and specificity of 100%. For barium detected at any level, sensitivity and specificity were 86% and 54%, respectively.

“We continue to enroll patients and are planning future multicenter studies to confirm these preliminary results,” Mika and colleagues indicated.