1 Introduction

The use of intra-articular antibiotics is an area of increasing interest for arthroplasty surgeons as periprosthetic joint infection (PJI) continues to be a devastating complication. As demand for lower limb arthroplasty increases in aging populations, and the overall numbers of prosthetic joint infections also increase. PJI rates are estimate to range from 1% to 2% in primary joint arthroplasties, with revision surgeries being even higher.¹ The combined annual USA hospital costs related to PJI of the hip and knee are estimated to be $1.85 billion by 2030.² The ramifications of this both for the individual patients and for health systems are significant. Studies have shown 5-year mortality for patients with PJI of 21 %,^3,4 with costs of treatment of being several times that of uncomplicated primary surgery.² Clearly interventions that can reduce PJI will have significant benefits for individuals and society alike.

Intra-articular (IA) antibiotic use has potential benefit of very high doses around the prosthesis, which is orders of magnitude greater than possible with intravenous administration. This also has the benefit of reduced systemic effects and initial less renal excretion, ease of administration, and possible cost effectiveness in prevention of PJI.⁵ Direct application of antibiotic has demonstrated reduced infections in spinal,⁶ anterior cruciate,⁷ trauma surgery,⁸ and in rat models⁹indicating its therapeutic value. Furthermore, there is a growing body of evidence in total joint arthroplasty with a number of reviews suggesting intra-articular antibiotic importance.^10–12 A recent report by Lawrie and colleagues examined a series of total knee replacement who had IA and found it did not reach sustained toxic level up to 24 h after surgery. The effect of IA antibiotic over a longer period and on other joints is not yet clear.¹³

Our study examines the safety of IA vancomycin in TJA, with respect to renal and ototoxicity and intra-articular levels of the medication within the joint at 24- and 48-h post-surgery in a series of primary hip and knee replacements.

2 Methods

From August to October 2021, 68 patients undergoing primary total joint arthroplasty (THA & TKA) were given 1g vancomycin intra-articularly. While implant make was not standardised, components were cemented using antibiotic enriched Simplex bone cement (Tobramycin). 18 of those THR patients had the vancomycin powder placed into the acetabular immediately prior to cup impaction (retro-acetabular group), the remainder (50) had the injection into the joint. 1g vancomycin was drawn up in a 10 ml syringe with 10 ml sterile normal saline and injected with an 18-gauge spinal needed after deep fascial closure. All patients received 2g cefazolin intravenously (IV) 30 min prior to the procedure as is our standard prophylaxis, and 21 of the 50 patients were also administered an additional 1 gm IV vancomycin, as is usual prophylaxis of one author. Post-operative blood vancomycin level and creatinine monitoring was performed day 1 and day 3 and compared to pre-operative values, and patients were questioned prior to discharge regarding hearing changes. To determine the post-operative joint fluid vancomycin levels in THR patients, surgical drain fluid was sampled at day 1 and 2 in a subgroup of 10 patients (5 from the retro-acetabular vancomycin group, and 5 from the intra-articular group). The first 1 ml was discarded and 2–4 ml sampled and sent for analysis, and drains removed following sampling on day 2.

Continuous data were summarized using means and SD or confidence intervals. Differences in blood vancomycin, creatinine and eGFR levels over time were estimated using multivariate-linear models, adjusted for type of arthroplasty, location of antibiotic and Age, Sex, and BMI. Fisher's exact tests were used to compare categorical variables. A p-value of <0.05 was deemed statistically significant. Statistics were performed using R Studio (R Core Team, 2022). Institutional ethical approval was gained for this retrospective study.

3 Results

The average age of the total group 65 ± 10.5 years (IA after closure - 62.7 ± 11.8 years; Retro-acetabular group – 68/2 ± 7.7 years). There were no differences between the groups in terms of age, sex, and BMI, (Table 1). While there were more TKR's in the intra-articular group, operation type was not a significant factor in blood vancomycin, eGFR or creatine levels. As expected, increased BMI was significantly associated with higher creatine and lower eGFR levels, but not vancomycin levels. In the group where vancomycin was injected after fascial closure, the average blood vancomycin level day 1 was 4.2 μg/ml (CI 5.942 to 7.117) and significantly dropped on day 3 to <1.4 μg/ml (below the threshold for the test or effectively unmeasurable) (p < 0.001). Serum vancomycin levels for patients with IA + IV administration also dropped from 6.52 μg/ml on day 1 to <1.4 μg/ml by day 3 (Fig. XXX) (p < 0.001). The retro-acetabular group blood vancomycin level was <1.4 μg/ml on both days 1 and 3. There was no difference between any of the groups.

The average pre-op creatinine levels ranged between 69.9 μmol/L (CI 58.0 to 81.8) and 79.7 μmol/L (CI 68.0 to 90.3) and between 69.9 μmol/L (CI 58.0 to 81.8) and 78.4 (CI 66.8 to 89.9) on day 3, (p = 0.619). The average pre-op eGFR levels (ml/min/1.73 m²) were 80.8 (CI 74.3 to 88.1) compared to 82.8 (75.6–89.9) on day 3 and 83.0 (76.8–89.2) (p = 0.387). There was no difference between any of the groups for either Creatine or eGFR levels (p = 0.668, and 0.737, respectively). Furthermore, no patient suffered acute kidney injury or noted hearing loss from either intra-articular (1g) or Intra-articular and IV (2g) vancomycin in a single surgery. Finally, there were a higher level of joint fluid vancomycin in the intra-articular subgroup compared to the Retro-acetabular subgroup (Table 2).

4 Discussion

Intra-articular antibiotic administration shows promise to reduce infection rates in total joint arthroplasty and other types of surgery, with low risks of complication. Currently infection rates for TJA are in the order of 12 % for primary hip and knee arthroplasty and 2–5% for revision surgery.¹⁴ Our study demonstrates the safely of a single dose of 1g vancomycin intra-articularly in a series of primary hip and knee replacements, measuring the plasma levels of vancomycin post-operatively, joint fluid levels and renal and otic complications.

Whiteside looked at intra-articular levels of vancomycin, and demonstrated safety and efficacy in its use for treatment of infected joint replacement in several studies.^15–18 One study¹⁸ showed synovial fluid vancomycin concentrations at the time of incision after IV use to average 6.8 μg/ml (range 3.4–13.6 μg/ml) and commented that levels might be subtherapeutic for hours (MIC 2–5 μg/ml for Staph aureus) before the next IV dose. In our study we aspirated fluid at 24 and 48 h from the joints of 10 patients, five who had had injection into the joint, and five with powder place behind the acetabulum component and found IAA provide antibiotic levels orders of magnitude greater than with synovial levels with IV administration alone for up to 48 h. We noted levels within the blood and within the joint itself are significantly reduced by placing the antibiotic powder behind the acetabular component, and as a result we have discontinued this practice.

Our study showed no increase in creatinine and eGFR post-operatively, with either IA or IA + IV administration demonstrating systemic safety. This is not unexpected and makes intuitive sense as antibiotic from the joint has to enter the blood stream via the lymphatics for excretion, and while the exact mechanisms of vancomycin induced renal toxicity are not fully understood,¹⁹ a lower concentration released more slowly is likely protective of the kidney and safer compared with the IV route.

Determining the best antibiotics, dosage and route for prophylaxis in TJA has still not been settled. Intraosseous regional administration (IORA) has recently demonstrated vancomycin levels in both bone and periarticular soft tissue 5–15 times higher than available from IV administration.²⁰ Perhaps a combination of IORA and IA vancomycin may provide the best soft tissue, bone and synovial fluid levels, with higher concentrations for a lower dose for longer duration than are bioavailable with IV use and with a lower systemic side effect profile.

The results of this study need to be interpreted in light of its limitations. This is a retrospective study and had a relatively small sample size. It does nevertheless demonstrate that IAA provide low serum vancomycin levels, most of which failed to reach half the upper limit of safety of approximately 20 μg/ml,²¹ whilst having highly therapeutic IA levels for at least 48 h. This study illustrates the safety and efficacy of the use of 1g IA vancomycin or 1g IA + IV (2g) in the perioperative management of primary hip and knee joint arthroplasty. Ototoxicity was only formally questioned as to whether patients noticed any hearing change, but no audiometry was undertaken, however with the low serum levels that we reported, the likelihood of hearing damage is low.

5 Conclusion

The use of IA vancomycin in TJA is appears safe, providing high antibiotic levels within the joint itself, with no renal or auditory complications in our investigation. Whilst role of IA antibiotics is evolving and its position in total joint replacement has not yet been fully determined, these finding may stimulate further study in this area of clearly significant importance in the prevention of PJI.

Ethical committee approval

Not required.

Funding

NO funding was accepted for this study.

If the study is accepted for publication, particularly for open access, an application to the Canberra Orthopaedic Research and Education (CORE) foundation will be made.

Patient consent

All patients were consented for the routine pre and post-op blood tests as part of total joint replacement consent, and for the use of intra-articular antibiotics with our standard practice consent process, using the Calvary John James Hospital or the Australian Capital Territory public hospital consent forms.

JOO website suggest individual consent is only required for case presentations and photographic images.

CRediT authorship contribution statement

Alexander W.R. Burns: Conceptualization, Original draft preparation Methodology, Writing,

Tat Chao: Investigation, data collection, patient recruitment

Nicholas Tsai: Reviewing, data collection, patient recruitment

Joseph T. Lynch: Data curation and statistics, Reviewing

Paul N. Smith: Reviewing, data collection, patient recruitment,