Knee Arthrodesis with a Modular Silver-Coated Endoprosthesis for Infected Total Knee Arthroplasty with Extensive Bone Loss: A Retrospective Case-Series Study

Introduction: Knee arthrodesis is a limb salvage intervention for persistent periprosthetic joint infection (PJI) when revision total knee arthroplasty fails. Conventional arthrodesis techniques are associated with the increased rate of complications, especially in patients with extensive bone loss and extensor tendon deficiency. Methods: Eight patients with a modular silver-coated arthrodesis implant after failed exchange arthroplasty for infection, were retrospectively reviewed. All patients had significant bone loss, while 5 displayed extensor tendon deficiency. Survivorship, complications, leg length discrepancy, median Visual Analogue Scale (VAS) and Oxford Knee score (OKS) were evaluated. Results: The median follow up was 32 months (range 24–59 months). The survivorship rate of the prosthesis was 86% during the minimum time of follow up of 24 months. In one patient recurrence of the infection was observed and above-knee amputation was performed. The median postoperative leg length discrepancy was 2.07 ± 0.67 cm. Patients were able to ambulate with mild or no pain. The median VAS and OKS was 2.14 ± 0.9 and 34.7 ± 9.3, respectively. Conclusions: The results of our study demonstrated that knee arthrodesis with a silver coated arthrodesis implant, performed for persistent PJI in patients with significant bone loss and extensor tendon deficit, provided a stable construct, allowed eradication of infection and was associated with good functional outcome.


Introduction
Periprosthetic joint infection (PJI) after Total Knee Arthroplasty (TKA) demonstrates prevalence that ranges from 1% to 2% and is associated with implant failure [1]. In 9% to 12% of cases, the infection may persist despite the removal of the implant and the application of antibiotic spacers [2,3]. Moreover, the incidence of infection after revision arthroplasty for infected TKA is higher, reaching 26% [4]. The eradication of PJI is multifactorial and depends on the type of isolated organism, the chronicity of the infection and the immunologic status of the host.
Patients with persistent PJI usually have large bone defects, extensor mechanism incompetence and severe soft tissue deficiency, due to multiple implant extractions of failed TKAs and extensive debridement [5,6]. Resection arthroplasty and above knee amputation (AKA) should be considered as end-staged therapeutic options [7]. For patients

Materials and Methods
The clinical records and imaging studies of 8 consecutive patients with persistent PJI after two-stage revision for infected TKA that underwent knee arthrodesis with a silver-coated modular knee prosthesis (MUTARS ®® -Munster Arthrodesis Implant cast GmbH, Buxtehude, Germany) between 2016 and 2022 in our department were retrospectively reviewed. Primary osteoarthritis was the indication for primary TKA in all patients (Figures 1 and 2).
The microorganisms responsible for the knee infections ( Figure 3) varied widely and included 4 cases of methicillin resistant Staphylococcus aureus (MRSA), 2 cases of vancomycin-resistant Enterococcus, and 2 cases of multi-organism infections.
All total knee arthroplasties were removed, and meticulous debridement was performed, followed by insertion of antibiotic-impregnated cement spacers mixed with gentamicin and vancomycin ( Figure 4).           All total knee arthroplasties were removed, and meticulous debridement was performed, followed by insertion of antibiotic-impregnated cement spacers mixed with gentamicin and vancomycin ( Figure 4). In all patients, intravenous antibiotics were administered for 2-weeks followed by a 6-10 weeks per-os antibiotic regimen based on bacterial cultural results, microbial sensitivity tests and serial evaluation of inflammatory markers. If clinical examination and laboratory tests suggested persistence of infection, further debridement, and the exchange of the antibiotic-loaded polymethyl methacrylate (PMMA) spacer were performed. Before, knee arthrodesis patients had an average of 2.1 procedures for spacer placement and 0.9 irrigation and debridement procedures. Knee arthrodesis was performed only when clinical examination, laboratory tests and knee aspiration were indicative of infection clearance.
The study was compiled with the 1975 Declaration of Helsinki, and it has been approved by the Institution's Ethical Committee of our Hospital (Registration No: ΕBD 231/19-4-2021).

Technique
All surfaces were meticulously debrided of necrotic tissues with curettage and the intramedullary canals were reamed to remove all infected tissues. Multiple, at least five, intraoperative tissue samples were taken for cultures. After radical debridement, a modular silver-coated knee arthrodesis prosthesis (MUTARS-Munster Arthrodesis implant cast, Buxtehude, Germany) was used for reconstruction. The minimum resection length of this implant system is 145 mm (145-265 mm). The system consists of a femoral stem cemented (implavit ®® implant cast, Buxtehude, Germany; CoCrMo casting alloy), or cementless (implatan ®® implant cast, Buxtehude, Germany; TiAl6V4 with HA coating), an arthrodesis implant (implatan ®® implant cast, Buxtehude, Germany; TiAl6V4) with silver coating, a tibial plate (implavit ®® implant cast, Buxtehude, Germany; CoCrMo-casting alloy) with silver coating, an extension piece (implatan ®® implant cast, Buxtehude, Germany; TiAl6V4) with silver coating and a tibial stem cemented (implavit ®® implant cast, Buxtehude, Germany; CoCrMo casting alloy) or cementless (implatan ®® implant cast, Buxtehude, Germany; TiAl6V4 with HA-Coating) ( Figure 5). In all patients, intravenous antibiotics were administered for 2-weeks followed by a 6-10 weeks per-os antibiotic regimen based on bacterial cultural results, microbial sensitivity tests and serial evaluation of inflammatory markers. If clinical examination and laboratory tests suggested persistence of infection, further debridement, and the exchange of the antibiotic-loaded polymethyl methacrylate (PMMA) spacer were performed. Before, knee arthrodesis patients had an average of 2.1 procedures for spacer placement and 0.9 irrigation and debridement procedures. Knee arthrodesis was performed only when clinical examination, laboratory tests and knee aspiration were indicative of infection clearance.
The study was compiled with the 1975 Declaration of Helsinki, and it has been approved by the Institution's Ethical Committee of our Hospital (Registration No: EBD 231/19-4-2021).
In all eight patients cementless femoral and tibia stems were used. In two out of eight patients (28.6%), due to severely compromised soft-tissue coverage, a medial gastrocnemius flap was used for the coverage of the prosthesis. All surgeries were performed by the same surgeon (senior author P.J.P.). In all eight patients cementless femoral and tibia stems were used. In two out of eight patients (28.6%), due to severely compromised soft-tissue coverage, a medial gastrocnemius flap was used for the coverage of the prosthesis. All surgeries were performed by the same surgeon (senior author P.J.P.).
The diagnosis of PJI was based on the criteria of the International Consensus Group for Periprosthetic Infections. Specifically, two major (two positive periprosthetic cultures of the same microbe or sinus-tract communication with the joint) and five minor criteria (increased serum C-reactive protein, CPR, and erythrocyte sedimentation rate, ECR; increased white blood cells or polymorphonuclear neutrophils or positive leukocyte esterase (LE) strip test for synovial fluid and a positive histological test from periprosthetic tissue sample) should be met. The diagnosis of PJI was set when at least one major and three minor criteria were fulfilled [28].
The primary endpoint was the outcome (success vs. failure) of the surgical treatment according to prosthesis survivorship during the postoperative period ( Figure 6). The secondary endpoint was the assessment of the functional clinical results of the patients at their final follow up examination. The outcome was considered successful if no clinical or microbiological signs of infection were documented following surgical and antibiotic regime within the 24 months postoperative and patients achieved a painless, stable joint. Any other condition was considered as treatment failure [29]. The functional outcome was evaluated using the Visual Analogue Scale (VAS) for postoperative pain assessment and the Oxford Knee Score (OKS) [30,31]. All patients were ambulatory immediately at the first postoperative day with partial weight bearing, while full weight bearing began after 6 weeks. The diagnosis of PJI was based on the criteria of the International Consensus Group for Periprosthetic Infections. Specifically, two major (two positive periprosthetic cultures of the same microbe or sinus-tract communication with the joint) and five minor criteria (increased serum C-reactive protein, CPR, and erythrocyte sedimentation rate, ECR; increased white blood cells or polymorphonuclear neutrophils or positive leukocyte esterase (LE) strip test for synovial fluid and a positive histological test from periprosthetic tissue sample) should be met. The diagnosis of PJI was set when at least one major and three minor criteria were fulfilled [28].
The primary endpoint was the outcome (success vs. failure) of the surgical treatment according to prosthesis survivorship during the postoperative period ( Figure 6). The secondary endpoint was the assessment of the functional clinical results of the patients at their final follow up examination. The outcome was considered successful if no clinical or microbiological signs of infection were documented following surgical and antibiotic regime within the 24 months postoperative and patients achieved a painless, stable joint. Any other condition was considered as treatment failure [29]. The functional outcome was evaluated using the Visual Analogue Scale (VAS) for postoperative pain assessment and the Oxford Knee Score (OKS) [30,31]. All patients were ambulatory immediately at the first postoperative day with partial weight bearing, while full weight bearing began after 6 weeks.

Statistical Analysis
Categorical data are presented as numbers (percentages), and continuous presented as medians, in the text and figures.

Baseline Characteristics
The baseline characteristics of the patients are shown in Table 1. There we patients who were diagnosed with periprosthetic infection after revision of the throplasty and underwent surgical intervention with modular silver-coated endo sis. There were three males and five females, while the median age and Body Ma (BMI) were 67 years (min: 57, max: 75, SD: 6.50) and 27.14 kg/m 2 (ranging from SD: 3.24), respectively. All 8 patients had segmental bone loss, while five patient extensor mechanism deficiency ( Table 1). The median length of bone loss was (range, 145-205 mm). Moreover, clinical information recorded at the final follow cluded the absence for evidence of immunosuppression, previous radiotherapy, ical diseases, lymphoedema, vascular and pulmonary diseases, as well as previou ple surgical interventions.

Statistical Analysis
Categorical data are presented as numbers (percentages), and continuous data are presented as medians, in the text and figures.

Baseline Characteristics
The baseline characteristics of the patients are shown in Table 1. There were eight patients who were diagnosed with periprosthetic infection after revision of the knee arthroplasty and underwent surgical intervention with modular silver-coated endoprosthesis. There were three males and five females, while the median age and Body Mass Index (BMI) were 67 years (min: 57, max: 75, SD: 6.50) and 27.14 kg/m 2 (ranging from 22 to 31, SD: 3.24), respectively. All 8 patients had segmental bone loss, while five patients had an extensor mechanism deficiency ( Table 1). The median length of bone loss was 172 mm (range, 145-205 mm). Moreover, clinical information recorded at the final follow-up included the absence for evidence of immunosuppression, previous radiotherapy, oncological diseases, lymphoedema, vascular and pulmonary diseases, as well as previous multiple surgical interventions.

Clinical and Surgical Outcomes
The median follow-up period was 32 months (ranging from 24 to 59 months, SD: 11.88). The survivorship rate of the knee arthrodesis modular silver-coated prosthesis (primary endpoint) was 86% (six out of eight patients) during the minimum time of follow up of 24 months. One patient (14%) developed a periprosthetic infection recurrence and wound dehiscence. This patient underwent an above knee amputation and has remained free of complications for the last 30 months. In two of the patients (28%) one out of eight positive intraoperative cultures (Staphylococci sp.) were detected at the final implantation surgery, and intravenous antibiotic therapy was administered for two weeks. During the last follow up these patients did not show any clinical or serological symptoms of infection or implant-associated complications based on Henderson classification. The remaining five patients (71.4%) did not develop infection relapse and implant-associated complications during the follow up period. Postoperative calculation of leg length discrepancy showed that the affected leg was 2.07 ± 0.68 cm shorter. Regarding the secondary endpoints, none of the patients complained about severe pain at their last follow up (median VAS value = 2.4 ± 0.9) and their functional status was very satisfactory (median OKS value = 34.7 ± 9.3). All patients were able to perform everyday activities with occasional use of a cane (Table 1). Finally, clinical signs of argyria, peripheral neuropathy or other silver associated side effects were not observed in any individual.

Discussion
The treatment of persistent PJI after TKA is challenging and is associated with increased morbidity and financial burden [32]. Even though two-stage revision TKA is correlated with elevated risk of re-infection, morbidity and mortality, it remains the gold standard for the treatment of chronic PJI [33]. However, when revision or re-revisions in an infected TKA is not a viable option, knee arthrodesis may be considered as a limb salvage procedure [34,35]. Our study examined the functional and clinical outcomes and investigated the restoration of limb length and alignment after the application of modular silver-coated endoprosthesis for the infected knee arthroplasty with extensive bone loss and extensor mechanism deficiency. An additional objective of this study was to evaluate the rate of persistent or recurrent infection following the application of modular silver-coated knee arthrodesis prosthesis [21,27,[36][37][38][39][40][41][42][43] and to compare it with those of other modular implants. The studies and the referred outcomes of knee arthrodesis after the application of both silver-coated and not modular implants on infected total knee arthroplasties are summarized in Table 2.   Our results showed that MRSA was the primary cause for septic failure of TKA being in line with the international literature reporting that Staphylococcus spp. is the leading microorganism related with the infection of knee reconstructions, resulting in increased failure rates [21,27,[36][37][38][39][40][41]. It is worth noting that even if the risk for infection recurrence of TKA is not clearly defined, it tends to be lower (even 26%) including with Stanmore or custom-made arthrodesis modules [21,42,44]. Contrariwise, slightly greater risk (14.3%) for re-infection after the arthrodesis implantation has been detected by Rao, et al. [36] and Iacono, et al. [37]. Although the comparison between these case series and our research is difficult due to population heterogeneity and implant differences, our findings were in line with the outcomes being reported by the above studies. Moreover, Barlett et al. who studied the mid-term results after the application of Stanmore knee arthrodesis prosthesis in 10 patients (including a patient suffering from angiosarcoma) with a mean follow-up of 56.4 months (range, 15-199 months), reported eradication of infection in 90% of the patients. Angelini et al. examined the application of modular endoprosthesis (MUTARS-Munster Arthrodesis implant Implant cast, Germany) for knee arthrodesis in 32 patients (seven with a failed TKA accompanied by segmental bone loss, inadequate soft tissue coverage and extensor mechanism deficiency, and 25 with a major bone defect after tumour resection) reporting similar effectiveness rates in the eradication of the infective agents. In the above study, silver-coated prostheses were applied in only two-patients.
According to our results, infection recurrence has been reported in one patient (14%), who was finally treated with above the knee amputation and currently remains free of symptoms. Moreover, in the final implantation surgery of two patients (28%), positive intraoperative cultures (Staphylococci spp.) were observed, leading to the administration of intravenous antibiotics for 2 weeks. However, at the last follow up they remain free of infection or other implant associated complications. We must highlight the fact that, to the best of our knowledge, this is the first study that examined the clinical and functional outcomes of knee fusion after the application of silver coated modular mega-endoprostheses in the treatment of PJI, so far. Silver is a promising material for implant coating, and it seems to prevent bacterial colonization and periprosthetic infections [45][46][47][48][49][50]. Specifically silver ions can bind to bacterial wall proteins and to DNA, destroying the bacterial cellular wall or preventing microbe protein synthesis and bacterial colonisation [24,25]. Moreover, silver ions can result in intracellular production of reactive oxygen molecules that are toxic for the bacteria. Indeed, preclinical trials of animal experimental models demonstrated that the application of silver coated implants in rabbits was associated with significantly lower rates of infection compared with those with titanium implants (7% versus 47%) [51].
However, the exact role of silver coated implants used for knee arthrodesis following PJIs has not been fully elucidated yet. Wilding et al. examined the application of silver-coated arthrodesis nails in eights patients with known PJI and reported two cases of superficial and periprosthetic reinfection, respectively [27]. Moreover, the use of a custom-made device, by a Waldemar Link, based on the Mega C prostheses system (with the addition at the titanium non-articulating surfaces of an innovative silver coating composed of two layers) in 33 patients with septic complication or at high risk for infection was evaluated by Scoccianti et al. [52]. The results of the above study showed that the recurrence rate among the patients with known previous infection was only 9.5% [52]. Additionally, successful knee arthrodesis with eradication of the infection has been reported in an immunosuppressed patient with chronic Aspergillus osteomyelitis without detecting any silver-coated associated adverse events [53]. Consequently, the use of silver coating is a promising alternative method to reduce the infection rates of tumour prostheses or arthrodesis devices.
The survivorship of the modular arthrodesis implant may vary and depends on the type of complications and the infectious agent. In our study, the survivorship rates of the modular silver-coated arthrodesis system were estimated at 86% during a follow up time (minimum) of 24 months. According to Angelini et al., the survivorship of the modular endoprostheses was 50% at five years and 25% at ten years [44]. As the above study included both oncologic and non-oncologic patients, further multivariate analysis on prostheses did not demonstrate significant difference between the patients with tumour resection, the presence of infection or a gastrocnemius flap after a follow-up period of 9.5 years. Finally, even if the infection was complicated in 1/3 of the patients, the survivorship of the implant seems to be unaffected.
Knee arthrodesis can be used as a salvage procedure following complicated revision arthroplasty and presents superior functional outcomes when compared to amputation [1]. In most of the cases, these patients could ambulate independently with the assistance of a walking aid and displayed superior functional levels on pain indices compared to the patients after amputation. The patients included in the present study represented a difficult-to-treat population because they underwent prosthetic knee arthrodesis for unrevivable TKA after previous extensive unsuccessful interventions to eradicate periprosthetic infection. The MUTARS silver-coated prosthesis was selected for these patients to maintain mobility and to avoid amputation. Six out of eight patients were treated successfully with good functional results (median OKS value 34.6 ± 6.4), experienced mild or no pain symptoms (median VAS value 2.4 ± 0.9) and remained free of infection during their last follow up. Although our results agree with the functional outcomes reported by Rao [36], Iacono [37] and Scarponi et al. [39], it must be noted that many studies have described unchanged knee function, or patient dissatisfaction following this surgical procedure [21,27,38,43]. Postoperative calculation of leg length discrepancy displayed that the affected leg was 2.07 ± 0.68 cm shorter than the unaffected, but all the patients were able to perform everyday activities with the use of cane. Leg length discrepancy was directly related to patients' satisfaction, and it was similar between the studies that examined modular knee arthrodesis implants (Mean: 10 ± 10 mm, Min: 5, Max: 34 mm) as it was reported by Putman et al. [38]. After meticulous bone and soft tissue debridement, the application of silver-coating implant is linked to construct stability and early weight bearing making this technique a promising alternative for limb salvage.
Finally, in our cases the clinical symptoms of argyria or peripheral neuropathy were not observed. Our results are in consistence with the findings of Scoccianti et al. [52] who did not report symptoms of silver-associated side effects. Moreover, the above study identified low levels of serum silver ranging from 2.09 to 5.33 µg/L after two weeks and 36 months, respectively. Similarly, low circulating silver levels were also referred to by the study of Hardes et al. [49], ranging from 1.93 to 12.98 µg/L at the third and 24th post-operative month, respectively. These results may explain not only the long-lasting antimicrobial activity of silver coated prosthesis but the limited local and systemic clinical adverse effects of their application.
We acknowledge that, despite the encouraging mid-term results, the study has some limitations. It is a retrospective study, with a small number of patients, without a comparison group. However, all eight patients had persistent periprosthetic joint infection, extensive bone loss and extensor tendon deficiency and were operated by the same surgical team and their follow-up and functional results were evaluated by the same investigator at the time of the latest follow-up, providing the necessary accurate measurements, adequate analysis and correct interpretation for the increased validity of our results [54,55]. Randomized international multicentred studies with a larger number of patients are needed to establish the promising results of this knee arthrodesis technique.

•
Knee arthrodesis is an important alternative to amputation in infected and multiple revised TKA. • It provides functional improvement and eradication of the infection.

•
The application of silver-coated knee arthrodesis implants is associated with increased implant survivorship as well as with satisfactory mid-term clinical and patient-reported outcomes as the limb salvage procedure for failed infected TKA with significant bone loss and extensor tendon deficit.