CORR Insights®: What is the Likelihood That Tumor Endoprostheses Will Experience a Second Complication After First Revision in Patients With Primary Malignant Bone Tumors And What Are Potential Risk Factors?

Where Are We Now?

Although endoprosthetic replacement is the current reconstruction option most-frequently chosen for patients with bone sarcomas [2, 9], medium-term studies have shown rather high revision rates in primary reconstructions, mainly because of aseptic loosening and prosthetic joint infections (PJI) [11].

Apart from the current study by Theil and colleagues [10], few studies have been published on either long-term survival or complication rates following repeated revisions of tumor endoprostheses [1, 2]. A recent study found that after a mean follow-up of 29.4 years, the original prosthesis was still in place in only 18% of patients and the risk of amputation was 16% [2]. Revision-free survival was 55% at 5 years when a non-invasive expandable endoprosthesis was implanted upon revision arthroplasty [1].

PJI is among the most common and devastating reasons for primary implant revisions in orthopaedic oncology [8, 9]. Theil and colleagues [10] found that PJI was the most common cause (39%) of second complications, followed by structural complications (35%). Reported re-infection rates following one-stage revision vary between 22% and 58%, and 28% to 38% following two- stage revisions [8]. Since all of these studies were retrospective, however, they suffer from the kinds of bias we would expect to see in research of that design, in particular selection bias and loss to follow-up.

Theil and colleagues [10] were the first to show that nearly one half of the study population experienced a second complication at a median of approximately 1.5 years following primary revision. In addition, the probability of developing a second complication was independent of the anatomical site of first revision. This is in contrast to previous findings where e.g. proximal tibia replacements were a significant risk factor for primary revision due to PJI [2].

Where Do We Need To Go?

The study by Theil and colleagues [10] suggests that patients who had undergone a minor first revision had the same high probability for further revision as patients who had had a major first revision. The next step for future studies is to determine whether these results can also be confirmed with larger datasets and for other implant systems.

Standardized reporting of the reasons for repeated revision surgery is also of major importance. Since the publication of Henderson’s classification of implant failures in 2011 [3], many studies have used it. Unfortunately, there are two different versions, the original [3] and an update [4], both of which authors have used, resulting in inconsistencies that make it challenging to compare studies [11]. This is particularly problematic for orthopaedic surgeons who treat children, as complications are sometimes reported in specific terms, but other times they are automatically categorized as pediatric (Henderson Type 6 failure) [11], which seems much too vague to be useful.

PJI results in great harm and cost. An ideal treatment for PJI in patients with bone sarcoma with repeated infections has yet to be defined. Results of non-oncologic PJI are difficult to compare as complete implant removal is often difficult with megaprostheses. Accordingly, evidence-based guidelines for PJI in orthopaedic oncology and standardized reporting of validated diagnostic criteria such as the Musculoskeletal Infection Society guidelines are urgently needed [6, 8].

How Do We Get There?

Since randomized controlled trials are expensive and time-consuming, and since the diagnosis of bone sarcoma is so rare, they are unlikely to become the standard trial design to achieve higher evidence levels for studies dealing with repeated revision in patients with bone sarcoma [5]. However, prospective and cooperative (multicenter) studies seem feasible, and conducting them might really improve the quality of our knowledge, by reducing selection and assessment bias. Documentation for revision surgery should include complete and standardized datasets including a correct Henderson classification, time to complications, reasons, and time of loss to follow-up. For the future, an entirely separate pediatric classification of failure of limb salvage might be developed to avoid confusion and to better adapt to the challenges in skeletally immature patients.

Cooperation among orthopaedic oncology centers would increase patient numbers and would also include patient data from smaller centers that otherwise could be lost. Study collaborations should be fostered by orthopaedic oncology societies. Ideally, oncology societies define relevant projects, design study protocols, and invite members to participate in studies or form committees, such as International Society Of Limb Salvage for the updated version of Henderson’s classification of 2014.

Larger studies would allow us to detect less-common complications, and perhaps even stratify analyses to refine what we understand about risk factors for important complications, including aseptic loosening, PJI, and implant fracture. As a substantial number of patients with bone sarcoma die of the disease, death as a competing event should be considered in order to prevent overestimation of complications [12]. Schuh and colleagues [7] identified a 12% decrease in the cumulative risk of revision when they performed competing risk analyses instead of Kaplan Meier analyses. For these reasons, journals and editors should ensure that this analysis is performed whenever it is appropriate to do so based on the frequency of competing events and the duration of follow-up.