CORR Insights®: What Is the Mid-term Survivorship of Infected Rotating-hinge Implants Treated with One-stage-exchange?

Where Are We Now?

There are numerous studies comparing one- versus two-stage revision surgery in patients with periprosthetic joint infection (PJI). Several of these studies have reported on the high percentage of positive cultures at the time of reimplantation surgery [2, 4, 8]. The trend in practice toward using prolonged oral therapy after reimplantation [3, 10] may make it even more difficult to determine whether subclinical infections persist, particularly in studies where the follow-up duration is shorter. For that reason, I am somewhat mistrustful of shorter-term studies on this topic, particularly those in which oral antibiotic suppression was used. Unfortunately, studies reporting on long-term results (beyond 10 years) for patients with PJI are scarce [1, 7, 11].

Patients with PJI often present with bone and/or soft tissue defects, which may become especially severe after aggressive debridement. Bone defects imply severe problems for implant fixation, and some soft tissue deficits carry a high risk of joint instability or dislocation. These considerations put surgeons in a bind. Although using complex mechanical constructs at the time of the reimplantation surgery (such as impaction bone grafting for hips, tantalum augments for hips or knees, or rotating hinge devices for knee reconstructions) are viable options, surgeons know that such complex reconstructions may not last long. In fact, few papers have described the long-term mechanical dislocations, instabilities, or loosening for patients revised for PJI more than 10 years following surgery [1, 7].

In this month’s Clinical Orthopaedics and Related Research®, Ohlmeier and colleagues [5] report on the survivorship of infected rotating hinge TKA implants treated with single-stage direct exchange for PJI. Although it is not a long-term follow-up study, it is an important one. In the study, the patients’ outcomes scores were poorer than we’d hope to see, and a high proportion experienced repeat revision at a mean follow-up of only 6 years. Specifically, patients in this study achieved a mean Oxford Knee Score of only 37 points (on a scale of 12 to 60 points, with lower points representing less pain and better function), and about one-in-four underwent revision at a mean of 6 years, with many of those experiencing what appeared to be aseptic loosening. Considering the possibility that some infections and revisions might have occurred among the patients lost to follow-up, and some more may appear over time, these results make me question the long-term results obtained with this approach.

Where Do We Need To Go?

Given the concerns regarding dormant bacteria and the possibility of late reinfection or late appearance of persistent infection, it is going to be particularly important for studies like this one to continue to follow their patients beyond 10 years. Additionally, we must examine beyond the simple presence or absence of infection. Indeed, we must study aseptic loosening as well as pain and functional problems that can impair patients’ quality of life.

Similarly, when we use salvage reconstructive techniques—cementing into poor-quality host bone, rotating hinge TKA devices, or using constrained or dual-mobility cups in revision THA—it may be possible to win the battle by eradicating the infection but lose the war when patients develop late aseptic loosening or periprosthetic fractures over the longer term. Although my sense as an orthopaedic surgeon is that impaction allografting with cement and tantalum augments in aseptic bone defects likely are durable in most patients, it is less clear how those approaches will fare in the face of PJI. Will some remaining bacteria colonize slowly the trabeculae of the allograft or the pores of the tridimensional metal, compromising fixation or increasing the risk of persistent or recurrent PJI in the long term?

After reading the current study, I would not be tempted to use single-stage revision of rotating hinge TKAs. Still, we should consider following this group of patients for 10 years or more. Perhaps at that point, if surgeons still want to explore it, they could do so using longer antibiotic courses.

Finally, thoughtful, detailed assessments of pain and function are as important as radiologic results and studies that use revision as the key endpoint. Indeed, longer follow-up durations are important for these endpoints because patients reasonably wonder what life will be like years down the road after major reconstructions. Such data can help inform the decision of whether to undergo major reconstruction or to consider alternatives—perhaps including amputation. As I care for patients who have had these reconstructions, I sometimes wonder whether complex regional pain syndrome plays a larger role in the poor function and high pain levels that some patients report. Future studies should specifically look into this.

How Do We Get There?

Multicenter randomized trials seem attractive, but for this set of problems, they are unrealistic. Therefore, we need to look to other approaches.

To address the question of whether (or how often) dormant bacteria are present in the long term, it might be worthwhile to perform a universal screening study—a thorough work-up for PJI, using validated criteria, such as those of the Musculoskeletal Infection Society—for patients who seem to be doing well 5 years or even 10 years after revision arthroplasty for infection.

To ascertain whether complex reconstructions—such as those using impaction allografting or trabecular metal—are more likely to develop PJI, future researchers should consider experimental animal models that analyze the persistence and slow growth of remnant bacteria inside micropores. The sonication of all retired implants should be compulsory to quantify the number of present microorganisms.

To assess whether the limitations patients experience and the pain levels they endure are influenced by unusual diagnoses like complex regional pain syndrome, investigators could consider using one-stage exchange for infected rotating hinge and then report the reinfection rate, mechanical survivorship, and clinical results (subjective and functional) after 10 years.

Finally, a few general approaches might be considered in any studies that report on PJI. Multiple cultures should be obtained in any additional surgery looking for bacteria. Sonication or other techniques like the use of dithiothreitol might help identify bacteria on removed implants [6, 9]. But most importantly, we need to analyze the clinical status of patients whose operations occurred more than 10 years ago using internationally accepted approaches to outcomes measurement. Such information may be retrospective and nonrandomized, but if it is well reported, it can be compared across centers, which would be a step in the right direction.