Early and delayed periprosthetic joint infection in robot-assisted total knee arthroplasty: a multicenter study

Total knee arthroplasty (TKA) has evolved significantly in recent years, becoming a prevalent procedure in orthopedic surgery [1‐3]. Despite reports of substantial clinical success, dissatisfaction persists in approximately 20% of patients undergoing conventional TKA (C-TKA) [4, 5]. This has led to the development of Robot-Assisted (RA) TKA, a technique growing in popularity due to its precision in implant placement [6, 7]. RA-TKA utilizes preoperative X-ray or CT-scan images to tailor planning and implant selection to individual patient anatomy, potentially enhancing joint function, balance, and longevity [6]. The technique may also allow for more precise, minimally invasive approaches, resulting in less tissue damage and quicker recovery [7]. Early follow-up studies hint at improved outcomes [6], although RA-TKA is associated with longer surgical times, potentially increasing infection risk [6, 8‐12].

Periprosthetic joint infection (PJI) remains a severe complication in TKA [8], leading to revisions, extended hospital stays, and increased costs, despite advancements in aseptic techniques and antibiotic management [13, 14]. Traditional TKA reports PJI incidences between 0.5% and 2.5%, making it a leading complication [13, 14]. In the U.S., infection is the primary cause of TKA revision failure, with an incidence of 44.1% [14]. However, the incidence of PJI in RA-TKA is not well-documented in the literature [12‐14].

Diagnosing PJI is complex, with multiple criteria developed over time [15, 16]. The ICM 2013 criteria have been historically prevalent [17], and in 2018, more sensitive criteria were introduced, though not universally accepted [18]. Both criteria have limitations in confirming or excluding infection [17, 18]. In response, the European Bone and Joint Infection Society (EBJIS) introduced new PJI criteria in 2021 [19], featuring a middle category for patients with a higher likelihood of infection and proposing a staged diagnostic approach, using less invasive tests initially, followed by more specific tests if needed. This approach has shown promise in improving diagnostic accuracy [19‐22].

The primary aim of this study was to assess the incidence of early PJI following RA-TKA using the EBJIS criteria. Additionally, the study evaluated the occurrence of delayed PJI in a select group of patients. A comprehensive analysis of postoperative complications was also conducted to provide a thorough evaluation of this patient population. The hypothesis posits that RA-TKA is not associated with an elevated risk of early or delayed PJI.

The study's main finding was that RA-TKA performed using the NAVIO surgical system did not result in any cases of PJI, as defined by the EBJIS criteria [19]. Specifically, no cases were classified as “Infection Likely” or “Infection Confirmed” for early or delayed PJI. Two patients required secondary surgical procedures related to different aseptic complications, at five months and one-year postoperatively. This suggests that, while certain complications were noted, short-term postoperative joint infection risks do not seem to be increased by RA-TKA.

While, traditional TKA has been reported to yield generally good clinical results, dissatisfaction rates as high as 20% have been documented [4, 5]. RA-TKA was developed to address this issue, offering enhanced precision and accuracy in implant placement. Smith et al. [9] observed a higher patient satisfaction rate of 94% in RA-TKA compared to 82% in manual TKA. Gao et al. [10] conducted a meta-analysis revealing that kinematic alignment in RA-TKA could lead to better clinical outcomes than mechanical alignment in the short-term. Similarly, Zhang et al. [11] demonstrated improved accuracy in component positioning and better patient-reported outcomes in RA-TKA versus manual TKA.

Our study contributes to the limited assessment of PJI risk in RA-TKA. We observed no increased risk of PJI in RA-TKA, and this evidence aligns with other research comparing PJI rates between RA-TKA and C-TKA [24‐28].

For instance, Vandenberk et al. [27] published clinical results of a single-center retrospective cohort study of 230 NAVIO RA-TKA patients and 489 C-TKA patients, with a mean follow-up of 31 months and 29.7 months, respectively, that resulted in a PJI rate of 0% for RA-TKA compared to 1% for C-TKA. Joo et al. [26] reported a PJI rate of 0.2% in 851 patients who underwent RA-TKA, with a follow-up ranging from 4 to 15 months. Mitchell et al. [28] performed a comparative study between RA-TKA and C-TKA, describing no statistically significant difference in complication rates at one-year follow-up.

The detailed planning and multiple steps involved in RA-TKA with robotic systems contribute to increased surgical time. Studies have consistently shown prolonged surgical durations in RA-TKA compared to C-TKA [12, 26‐28]. In their research, Nogalo et al. [12] concluded that RA-TKA is associated with longer operative times. Given the extended surgical duration, there has been a growing focus on related complications, particularly the risk of developing PJI. Pugely et al. [29] identified an increased risk of infection after 120 min of surgery, while Peersman et al. [30] and Ravi et al. [31] also reported increased infection risks associated with longer TKA surgeries. However, these findings are not universally consistent. Our study's average surgical time was 121 ± 20.5 min, similar to other studies [32‐34], which did not report an increased PJI rate despite extended surgical durations. Specifically, Singh et al. [32], Held et al. [33], and Scigliano et al. [34] found no significant difference in postoperative complications or revision surgery rates due to prolonged surgical times.

The main strengths of this study include its considerable cohort size, which enhances the level of evidence. The multicentric approach broadens patient diversity, ensuring more robust and generalizable data. Additionally, the meticulous application of exclusion criteria, particularly for patients with alternate sources of inflammation or immunodepression, significantly bolsters the validity of CRP measurements in the context of assessing surgical outcomes.

However, this study is not without limitations. Firstly, its retrospective design inherently carries biases typical of such methodologies, which might affect the interpretation of the results. Secondly, despite the implementation of uniform peri-operative protocols across different centers, subtle variations in surgical techniques could potentially impact clinical outcomes. The third limitation of the study lies in the absence of a control group, which is crucial for a comprehensive evaluation of the impact of independent variables. A control group enables the isolation of specific effects, necessitating caution in interpreting the obtained results. Lastly, the follow-up duration constraints limited our ability to assess late PJIs. The study also did not fully encompass the evaluation of delayed PJIs due to the inability to achieve a complete one-year follow-up for all participants. To mitigate this limitation, a rigorous and ongoing follow-up protocol has been instituted for the included patients. These continued assessments are essential for the prospective monitoring and detection of delayed and late PJIs, providing a more comprehensive understanding of the long-term infection risks associated with the surgical procedure.

This study demonstrates the absence of early and delayed PJI in RA-TKA cases, using the EBJIS criteria, within the study period, thereby contributing additional evidence to this technique’s short-term efficacy and safety. However, further research is necessary to corroborate these findings and to confirm the long-term reliability of RA-TKA, particularly in the context of late PJI.