Article focus

• This prospective cohort study assesses patient safety in relation to use of poly(methyl methacrylate) (PMMA) spacers with microsilver in prosthetic hip infections.

• The study’s hypothesis is that the use of microsilver-loaded PMMA spacers is safe.

• The study includes: an analysis of clinical and laboratory parameters, measurement of silver concentrations in blood, urine, and drainage fluids; and a histopathological evaluation of the periprosthetic membrane around the spacers.

Key messages

• Microsilver-loaded gentamicin-PMMA spacers showed good biocompatibility, with very low blood and urine silver concentrations.

• Drainage fluids showed concentrations between 16.1 ppb and 23.3 ppb at 12 hours after implantation of the silver spacers, and between 16.8 ppb and 25.1 ppb at 48 hours after implantation, which is in line with the concept of local application of antimicrobial agents.

Strengths and limitations

• This study includes prospectively collected data in the target population of patients with a prosthetic hip infection.

• Limitations of this study included low patient numbers and a limited follow-up period.

Introduction

Prosthetic joint infections (PJIs) remain a challenge and have a tremendous negative impact both on the quality of life of patients and on healthcare systems.¹ Surgical treatment for PJIs can be divided into implant-retaining strategies, such as irrigation and debridement, and surgical options with implant removal.² The latter includes one-stage or two-stage revision procedures, which differ in the timepoint of implantation of the revision implant. For one-stage revision procedures, implantation of the revision prosthesis is performed during the same surgery as the removal of the infected prosthesis. The two-stage concept consists of two interventions, starting with the removal of the prosthesis at surgery 1, followed by insertion of a new prosthesis during the second procedure (surgery 2), which is usually performed a few weeks later. Antibiotic-loaded spacers are frequently inserted during surgery 1, acting both as a drug delivery system and void filler to facilitate implantation of the revision prosthesis.³ However, after an initial burst from the poly(methyl methacrylate) (PMMA), subinhibitory levels of the released antibiotics have been described, with the risk of bacterial recolonization of the PMMA surface with antibiotic-resistant bacterial strains.^4-6

While an increasing prevalence of infections in revision arthroplasty with resistant microbes has been reported, sobering clinical results were noted by Gomez et al⁷ for two-stage PJI revisions, the ‘gold standard’ of revision arthroplasty in knee and hip PJIs.⁸ In 504 cases, the following were reported: an interstage mortality of 7.5%, an inability to reimplant the prosthesis in 20% of cases; and a failure rate of nearly 20% in revision prosthesis implantation.⁷ In this context, the need to improve treatment of PJI is obvious.

Silver has been known and used as an antimicrobial substance in medicine for many decades, and has recently gained interest for the coating of tumour endoprostheses in orthopaedic oncology with convincing clinical results.^9,10 The major benefit of silver is its antimicrobial activity, with preclinical data suggesting that PMMA bone cement loaded with microsilver particles is highly effective against a variety of bacteria, including methicillin-resistant Staphylococcus aureus (MRSA).¹¹ Furthermore, it is uncommon for bacteria to develop resistance against silver, most likely due to the multiple mechanisms of antimicrobial action exploited by silver ions.¹² The addition of silver to antibiotic-loaded PMMA spacers therefore offers the potential benefit of broadening and prolonging the antimicrobial protection of the spacer surface due to its wide-ranging antimicrobial properties and favourable release kinetics from PMMA.

However, adverse events due to silver-coated orthopaedic prostheses have also been reported, such as argyria with grey discolouration of the skin next to the implantation site or elevation of serum silver levels,^13,14 emphasizing the need for cautious clinical evaluation of silver-loaded orthopaedic implants. Interestingly, argyria was not found to be associated with elevated serum silver concentrations or significant changes in laboratory parameters in these cases versus in patients without the development of argyria who were treated with the same type of silver-coated megaprostheses.^13,14

We hypothesized that the application of microsilver-loaded PMMA spacers in PJI after total hip arthroplasty (THA) is safe. Therefore, the purpose of the current prospective case series was to assess patient safety after silver-loaded PMMA spacer implantation in patients undergoing infection revision procedures for infected THA. The study included: clinical observation of the patient; silver concentration analysis of the serum, urine, and drainage fluids; and histological analysis of the peri-implant membrane.

Materials and Methods

Spacers

For silver-free spacers, commercially available gentamicin-loaded PMMA spacers with 0.5 g of gentamicin per 40 g of PMMA (Palacos R+G; Heraeus Medical GmbH, Wehrheim, Germany) were used. The amount of PMMA and the size of the used spacer moulds (StageOne; Biomet, Berlin, Germany) were adapted to the size of the hip joint. The following quantities of PMMA were used: 80 g in two cases; 120 g in nine cases; and 160 g in one case (Table I). In terms of silver spacers, 1% w/w of microsilver (Bio-Gate) was added to the PMMA (0.8 g in one case, 1.2 g in nine cases, and 1.6 g in one case; Table I). The silver was added to the PMMA powder and stirred with a spatula to achieve homogeneous distribution within the powder before the liquid was added. All other steps for the production of the spacers were identical to those for the silver-free spacers. In all cases, vacuum-assisted mixing (Optivac; Biomet, Berlin, Germany) was used.

Table I.

Patient demographics, microbiological findings, comorbidities, poly(methyl methacrylate) (PMMA) bone cement, and microsilver details

Patient	Sex	Age, yrs	Identified bacteria	Relevant comorbidities	Weight of PMMA powder, g	Weight of microsilver added to PMMA powder, g	Surgery 1	Surgery 2		Surgery 3
							Spacer type	Days between surgeries 1 and 2	Spacer type	Days between surgeries 2 and 3	Implantation of revision THA
1	Male	61	Staphylococcus aureus	Diabetes mellitus type II, arterial hypertension, spondylodiscitis	120	1.2	Silver	21	Non-silver	21	Yes
2	Male	45	Staphylococcus capitis	None	120	1.2	Non-silver	21	Silver	21	Yes
3	Male	76	Pseudomonas aeruginosa, MRSA	AV block grade 1	120	1.2	Silver	16	Non-silver	16	Yes
4	Female	75	Culture-negative	Sacral ulcer, diarrhoea, highly elevated CRP	120	1.2	Silver	15	Non-silver	13	Yes
5	Female	80	Staph. aureus	Depression, weakness, osteoporosis, low oxygen saturation	120	1.2	Silver	15	Non-silver	N/A	No*
6	Male	72	Staphylococcus epidermidis	Arterial hypertension, CHF, moderate renal insufficiency, diabetes mellitus type II	80	0.8	Non-silver	17	Silver	15	Yes
7	Male	80	MRSE	Pacemaker, diabetes mellitus type II, hypertension	160	1.6	Silver	N/A†	N/A†	N/A†	N/A†
8	Female	75	Enterococcus faecalis	Depression, diabetes mellitus type II, arterial hypertension	80	0.8	Non-silver	15	Silver	16	Yes
9	Male	31	E. faecalis	None	120	1.2	Silver	15	Non-silver	14	Yes
10	Male	64	Staphylococcus lugdunensis	Arterial hypertension	120	1.2	Non-silver	15	Silver	14	Yes
11	Female	77	MRSA	CHF, arterial hypertension, anaemia, atrial fibrillation	120	1.2	Silver	14	Non-silver	14	Yes
12	Female	64	E. faecalis	COPD, CHF (NYHA class IV), allergies to chrome and nickel, hypercholesterolemia	120	1.2	Non-silver	15	Silver	14	Yes

1. *

   No reimplantation possible due to poor general health status

1. †

   Patient chose to withdraw from the study after surgery 1

1. THA, total hip arthroplasty; MRSA, methicillin-resistant Staphylococcus aureus; AV, atrioventricular; CRP, C-reactive protein; CHF, congestive heart failure; MRSE, methicillin-resistant Staphylococcus epidermidis; N/A, not applicable; COPD, chronic obstructive pulmonary disease; NYHA, New York Heart Association

Preformed hip spacer moulds (StageOne) were then filled with bone cement using a cement gun to create articulating spacers (Fig. 1). After the PMMA spacer hardened, the moulds were cut, the spacers were inserted into the proximal femur, and the hip joint was reduced.

Fig. 1

a) Microsilver-loaded poly(methyl methacrylate) (PMMA) spacer after hardening in a preformed hip spacer mould (StageOne; Biomet). b) Postoperative radiograph control in anteroposterior view with correct placement of the microsilver-loaded PMMA spacer.

Results

Drainage fluids

Drainage fluids showed silver concentrations from 6.0 ppb to 38.0 ppb and 5.6 ppb to 68.0 ppb after 12 hours and 48 hours, respectively, following surgical implantation of the silver spacer. In patients who received the silver-loaded spacer first, the mean silver concentration was 16.1 ppb (sd 11.5) at 12 hours after surgery and 25.1 ppb (sd 20.4) at 48 hours after surgery (Fig. 2). The mean silver concentration decreased to 3.1 ppb (sd 3.5) after surgery 2 with the silver-free spacer. The mean silver concentration in the drainage fluid was 1 ppb in all patients at 12 hours and 48 hours after surgery 1 with a silver-free spacer. Following implantation of the silver spacer in these patients, silver concentrations of 23.2 ppb (sd 4.4) at 12 hours after surgery and 16.8 ppb (sd 6.9) at 48 hours after surgery were observed.

Fig. 2

Chart showing the silver concentration in the drainage fluid.

In six of the 11 cases, the silver concentrations between 12 hours and 48 hours only changed moderately, within a range of 5 ppb. The maximum concentration of 68.0 ppb was seen in one patient, who had already shown the highest concentration of 38.1 ppb after 12 hours.

Histology

In general, no considerable systematic differences between microsilver-loaded and silver-free PMMA spacers were detected and the histomorphological features were similar compared with regular diagnostic samples tested in the same period.

In both cases, the highest inflammatory activity (with a mean of more than ten neutrophil granulocytes per HPF) could be observed in the samples of surgery 1 with explantation of the infected prosthesis (Figs 3a and 3d). Samples taken from surgeries 2 and 3 mostly showed low-grade infection (Fig. 3b) or only a chronic inflammation (Fig. 3c), with a mean of neutrophil granulocytes below the diagnostic threshold of one per HPF (Figs 3b, 3c, 3e, and 3f). No silver particles or indications of toxic effects such as necrosis were found in any of the samples.

Fig. 3

Haematoxylin and eosin (H&E) staining and pathohistological assessment of periprosthetic/perispacer membrane in two cases. a) to c) Microsilver poly(methyl methacrylate) (PMMA) spacer first followed by silver-free PMMA spacer. d) to f) Silver-free PMMA spacer first followed by microsilver PMMA spacer. Magnification: 400×.

Polyethylene wear particles were detected in three of the initial samples and two of the later samples. Since neither of the spacer variants contained polyethylene, these wear particles must have derived from the prosthesis that had already been removed. Light microscopy showed that in seven of the initial samples and five of the later samples, there was fine (< 10 µm) brown (e.g. hemosiderin or iron rich wear particles) or black (e.g. titanium wear particles) pigmented material. There was no association between the pigmented material and the type of spacer used.

Discussion

To the best of our knowledge, the present study examined for the first time the use of silver-loaded gentamicin-PMMA spacers in a clinical trial. Our group previously showed that PMMA bone cement with 1% w/w of microsilver exhibits promising in vitro effects, with high antibacterial effectiveness against multidrug-resistant bacteria.¹¹ At the same time, no toxic effects on larger eukaryotic cells were seen.¹¹ Combined, the biocompatibility with eukaryotic cells and cytotoxic effects against multidrug-resistant prokaryotic cells provide a ‘therapeutic window’ for clinical application.¹¹ In the present study, 0.8 g microsilver was used in the PMMA spacers in most cases; in one patient, a total amount of 1.6 g microsilver was used. No silver-specific adverse events, such as argyria or impairment of the femoral or sciatic nerve, were observed. Any clinical adverse events detected, such as nausea, vomiting, low oxygen saturation, or laboratory changes (decreased thrombocytes, elevated creatinine, or alterations of potassium) were related to the procedure and to existing comorbidities of the patient, and not to the additional silver loading of the spacer. Serum and urine analysis revealed concentrations around or below the detection limit of 1 ppb in all but one patient. This patient exhibited silver concentrations of 3.0 ppb at 12 hours after the implantation of the silver spacer during surgery 1, and 2.8 ppb after 48 hours. Interestingly, these concentrations even increased up to 5.3 ppb at 48 hours after surgery 2, with the change of spacer from silver-loaded to silver-free PMMA. This patient suffered from multiple comorbidities, such as sacral ulcers, diarrhoea, and highly elevated CRP during admission, whereas creatinine levels were found to be low (0.6 mg/dl at surgery 1 and 0.58 mg/dl at surgery 2). No clear explanation for the marked difference in silver serum levels between this patient and the others could be found. Data from the literature suggest that silver serum levels in silver industry workers are between 0.1 ppb and 23.0 ppb,¹⁸ and around 11.0 ppb¹⁹ without adverse clinical effects, and, therefore, this maximum value of 5.3 ppb blood can be deemed safe for the patient. A limitation of the study is that no preoperative serum silver analysis was performed, which could have helped to explain why the silver levels reached 5.3 ppb blood in this patient.

Drainage fluid levels were between 6.0 ppb and 38.1 ppb after surgery 1 and between 5.6 ppb and 68.0 ppb after surgery 2, which is considerably higher than the measured silver levels in urine and in blood. This is fully in line with the concept of local application of antimicrobial agents, such as in antibiotic-loaded PMMA spacers or beads, to achieve high local concentration and minimize systemic concentrations of the antimicrobial agents at the same time, based on the initial ideas of Buchholz and Engelbrecht²⁰ and Klemm.²¹ Our results confirm a considerable release of silver from the PMMA spacer, which is a prerequisite to fulfil its antimicrobial effects against bacteria in the wound. Due to the fact that no silver-related adverse events were detected, the observed local concentrations of up to 68.0 ppb in the drainage fluid can be considered safe for patients.

With regard to the clinical application of silver in other musculoskeletal contexts, silver poisoning was described following the use of nanoparticulate silver for topical wound dressing (Acticoat; Smith & Nephew, London, United Kingdom).²² In this case, clinical presentation showed greyish discoloration, fatigue, and lack of appetite.²² The silver concentration in the blood was determined as 107 ppb.²² A case series describing silver uptake in patients treated with the same wound dressing reported median maximum silver levels of 56.8 ppb with no haematological or biochemical indicators of toxicity.²³ The maximum level of 5.3 ppb in our study was more than ten times lower than the median maximum levels found in the silver wound dressing group of this case series.²³

Hardes et al¹³ reported on the biocompatibility of silver-coated megaendoprostheses (MUTARS; Implantcast GmbH, Buxtehude, Germany) in a clinical case series of 20 patients with bone tumours. A maximum concentration of 56.4 ppb was found in one patient’s serum 15 months after implantation of the silver-coated megaendoprosthesis.¹³ Laboratory tests excluded the kidney, liver, and haematological impairment within a follow-up period of 24 months.¹³ Again, the maximum measured concentration of 5.3 ppb blood in our study is more than ten times less than the concentration found in the study by Hardes et al.¹³ However, different coating techniques and different types of silver (metallic silver, microsilver, silver salts, other silver complexes) result in a significant variation in the release of free silver ions (Ag⁺), which are thought to be relevant for silver toxicity.²⁴ Hence, comparison of toxicological studies should be critically evaluated.

Based on the findings of animal studies, silver and its ionic forms seem to be distributed to all organs.²⁵ In argyria, the most prominent sign is discolouration of the skin. Local argyria around the microsilver-loaded cement spacers was not detected in any of our patients. The histological findings of tissue samples taken in the immediate vicinity of the spacers showed no difference between groups.

Silver excretion in general is described as urinary and faecal.²⁴ A urinary-faecal excretion ratio in monkeys is reported as 0.019% to 0.026%.²⁶ Furthermore, urinary excretion of orally applied silver nanoparticles was found to be between 0.005% and 0.057%.²⁵ Hence, it is not surprising that no quantifiable excretion of silver was detectable in the urine of patients in our study.

Accumulation of silver in patients’ organs could not be determined directly. However, laboratory tests did not indicate any functional limitation of internal organs due to the deposition of silver. In a rabbit model, Gosheger et al²⁷ found that no toxicological effects were associated with a silver-coated MUTARS tumour endoprosthesis within a 90-day follow-up period. Accumulation of silver was observed in the liver and spleen with mean values of 90.4 ppb and 27.9 ppb, whereas blood concentrations had a mean of 1.88 ppb.²⁷

Our paper has several limitations, including a low number of patients. The case series was planned to test biocompatibility in temporary spacer implantation only. The long-term antimicrobial performance of silver-loaded PMMA cement in reducing reinfection rates after reimplantation of the prosthesis was therefore not evaluated. This will be of the utmost importance for further clinical studies with microsilver-loaded PMMA spacers. An interesting microbiological finding of the current study was the lack of bacterial growth after the use of silver spacers in six of seven cases after surgery 1, compared with proof of bacterial growth in all four cases in which non-silver spacers were used during surgery 1 (Table II). The small sample size and the primary safety study design characteristics do not allow for adequate statistical evaluation in this context.

Furthermore, the short observation period of 28 to 42 days from initial surgery until reimplantation of the revision prosthesis also limits the quality of the study. However, these spacers only serve as temporary implants, and an observation period limited to the duration of the in situ stage is acceptable. In the study conducted by Glehr et al,¹⁴ in which silver-coated megaendoprostheses were implanted during bone tumour surgery, patients developed argyria after a median exposure time of 25.7 months. This is substantially longer than the relatively short in situ period of silver spacers of only 14 to 21 days.

Bacterial resistance to silver nanoparticles and strategies to overcome this, as recently reported by Panáček et al,²⁸ have to be addressed in future studies, especially given the large numbers of multidrug-resistant bacteria.²⁹

In conclusion, this prospective cohort study on microsilver-loaded gentamicin-PMMA spacers was intended to evaluate the safety of gentamicin-loaded PMMA spacers with the additional loading of microsilver in prosthetic hip infections in a prospective cohort study. The microsilver spacers did not show any adverse clinical events related to silver, such as argyria of the skin. Blood and urine analyses were far below the published silver concentrations of other orthopaedic implants or wound dressings and confirmed the good biocompatibility. Detailed histology of the peri-implant membrane did not reveal any differences between microsilver and silver-free spacers. Therefore, the microsilver PMMA spacers used can be deemed safe, and their broad antimicrobial activity warrants further clinical research to assess their effectivity in reducing infection rates in prosthetic joint infections.