Facing multidrug‐resistant pathogens in periprosthetic joint infections with self‐administered outpatient parenteral antimicrobial therapy—A prospective cohort study

A key factor in the successful management of periprosthetic joint infection (PJI) besides the surgical regime is a consistent antimicrobial therapy. Recently, oral versus intravenous (IV) antibiotics for bone and joint infection trial demonstrated the noninferiority of oral antimicrobial therapy compared to IV, implying that an early transition to oral administration is reasonable. It is likely that the international consensus meeting of musculoskeletal Infections and the European Bone and Joint Infection Society will consider these findings. However, rising levels of antimicrobial resistance are challenging and recommendations for dealing with multidrug‐resistant (MDR) pathogens resistant to oral antibiotics are lacking. This study focuses on establishing guidance towards their management in PJI. From December 2015 to June 2019, patients with MDR pathogens were included in a single‐center prospective cohort study and treated with self‐administered outpatient parenteral antimicrobial therapy (S‐OPAT) based on a two‐stage revision strategy. Demographics, pathogens, antimicrobial agents, and outcomes were recorded. A total of 1738 outpatient days in 26 patients were analyzed. The incidence of pathogens resistant to oral antibiotics in PJI was 4%, most frequently encountered were staphylococcus epidermidis. The Kaplan–Meier‐estimated infection‐free survival after 3 years was 90% (95% confidence interval, 84.6%–95.5%). We recorded adverse events in 6 of 54 (11%) S‐OPAT episodes (3.45/1000 S‐OPAT days). (i) S‐OPAT in two‐stage revision arthroplasty to counter increasing numbers of MDR pathogens resistant to oral agents can achieve a high infection eradication rate and (ii) should therefore be taken into account at the next society's consensus treatment updates.


| INTRODUCTION
Periprosthetic joint infection (PJI) is a major complication in total hip arthroplasty and total knee arthroplasty (THA/TKA). It occurs in 1%-2% of all primary arthroplasties and in around 4% of revision arthroplasties. [1][2][3][4] Various management concepts coexist, but until now a gold standard has not been established. In addition, guidelines for dealing with multidrug-resistant (MDR) pathogens are lacking, leading to controversial discussions due to the increasing relevance of these pathogens. Regardless of the procedure used, such as debridement, antibiotics, irrigation, and retention of the prosthesis, single-stage or multistage revision, patients with a normal postoperative course, declining inflammatory parameters, and dry wound conditions can be discharged to outpatient care after having been successfully switched from parenteral to oral antimicrobial therapy. [4][5][6][7] Even though the optimal point for switching from parenteral to oral administration remains unknown, this approach is consistent with the 2019 oral versus intravenous (IV) antibiotics for bone and joint infection (OVIVA) randomized trial, which challenged the practice of prolonged IV therapy. The noninferiority of oral antimicrobial therapy was demonstrated, implying that an early transition to oral administration is a reasonable alternative to prolonged IV therapy. 8 It is likely that the recently published results of the OVIVA trial will influence the upcoming guidelines of the different associations and societies, leading to a rethink of the role of oral antibiotic therapy. However, there may still be a mandatory indication for prolonged IV therapy in the case of MDR pathogens. More than a million THAs and TKAs are performed in the United States annually.
It is estimated that the number will almost double by 2030. 9 In addition, the incidence of revision hip and knee arthroplasty due to infections is projected to increase in the same time period. 10 Both will lead to an enlarged absolute number of mixed infections or difficult-to-treat (DTT) pathogens showing multidrug resistance, which will take up a larger proportion in specialized departments for PJI. Susceptibility-guided optimized oral antimicrobial therapy can often not be achieved for these infecting organisms. This situation occurs if the oral bioavailability is not sufficient to ensure reliable infection eradication, even at the highest doses, if the suitable agents are only available as IV medications due to antimicrobial resistance, or if oral susceptibility-guided therapy comprises only antibiotic agents with solely bacteriostatic action (Table 1).
Since the optimal duration of IV antimicrobial treatment is unclear, prolonged IV therapy was preferred in the past (IDSA guideline) and has been a driver for outpatient parenteral antimicrobial therapy (OPAT) services. However, there is a lack of published studies investigating the use of OPAT in a homogeneous PJI population and to the best of our knowledge, no prospective study has been carried out in this setting. If PJI cases are considered in OPAT studies, they are generally counted among the group of bone and joint infections, which is a heterogeneous collection of cases and encompasses widely mixed infection types. In addition to PJI spondylodiscitis, osteomyelitis and diabetic foot syndrome are frequently included, leading to different treatment concepts. 11,12 For a more targeted investigation, a well-defined study design using a homogeneous PJI cohort is required.
Currently, there are no published guidelines for the treatment of MDR organisms that cannot be controlled by oral medications. This leads to prolonged hospitalization associated with higher costs or compromises are made when selecting the antibiotic agent, which, in turn, increases the risk of persistent infection and treatment failure.
Administering OPAT represents a viable alternative and provide a positive economic benefit for the patient, the health insurance company, and the hospital. The available data do not allow any conclusions to be drawn regarding whether OPAT is useful in patients with MDR-PJI. The cure rate of MDR-PJI patients who underwent OPAT was never investigated in comparison to PJI cohorts in general. Therefore, the aims of this study were (i) to describe the outcome of treating MDR pathogens in PJI with OPAT in two-stage revision arthroplasty and (ii) to establish guidance towards the management of MDR-PJI resistant to oral antimicrobial therapy.

| Definitions
The inclusion criterion was the detection of a pathogen that could not be optimally treated with oral antibiotics according to the PRO-IMPLANT Foundation based on the management concept of Zimmerli et al. 1 and adapted by Trampuz et al. 1,6,38,39 Table 2 shows our standard antimicrobial treatment concept, modified after Izakovicova et al. 40 The pathogens were reliably identified from multiple tissue samples taken intraoperatively.
When a patient is discharged to S-OPAT, it is classified as an S-OPAT episode. Two S-OPAT episodes per patient are typically expected in two-stage revision arthroplasty. Catheter-related complications (CRC) ( Table 3) and adverse drug reactions (ADRs) are the two major adverse events that can occur and were recorded.
The definition for successfully treated PJI was based on the Delphi panel international multidisciplinary consensus. 41 Therefore, treatment was considered as successful if all of the following criteria were fulfilled at the latest follow-up examination: (i) infection eradication, characterized by a healed wound without fistula persistens, drainage, or pain, and no infection recurrence caused by the same organism; (ii) no subsequent surgical intervention for persistent or perioperative infection after revision surgery; and (iii) no occurrence of PJI-related mortality. 41 In addition, after the termination of OPAT, no oral antibiotics were continued.

| Diagnosis
A thorough physical examination with respect to the soft tissue condition was carried out and radiographs were made. Laboratory tests, including diagnostic arthrocentesis, were performed. In addition, histopathology and microbiological samples were analyzed.
After removal, the prosthesis underwent sonication to increase the accuracy of the microbiologic diagnosis. 42 PJI was diagnosed by applying the proposed EBJIS criteria (Table 4). 43

| Treatment
After confirming the diagnosis, all implants, cement remnants, and foreign material were removed, as well as infected soft tissue and bone.
Irrigation and debridement of all surrounding tissue layers were conducted. Cement spacers were used in temporary arthrodesis of the knee for stabilization, dead-space management, and local antimicrobial treatment, as part of a standardized two-stage revision concept.
Antimicrobial treatment was initiated intraoperatively after taking at least five tissue samples and synovial aspiration. would not be sufficient (Table 1), S-OPAT was initialized before discharge. A long prosthesis-free interval of ≥6 weeks due to the presence of MDR pathogens was used. All patients received antimicrobial therapy until reimplantation (i.e., without an antibiotic-free period and diagnostic aspiration). During reimplantation, one more thorough debridement procedure was performed, and multiple tissue samples were collected. After implantation of the prosthesis, a general IV treatment followed until patients showed a normal postoperative course, declining inflammatory parameters and dry wound conditions at which point they were discharged to S-OPAT, if possible with oral biofilm-active therapy. The total antimicrobial treatment duration was at least 12 weeks ( Figure 1). Empirical antimicrobial therapy Ampicillin/sulbactam 3 × 3 g IV or amoxicillin/clavulanic acid 3 × 1.

| Home care setting
To establish S-OPAT, we followed the recommendations of the British Society for Antimicrobial Chemotherapy (BSAC) 46  Patients and relatives were trained to independently apply OPAT.
Between administrations, a 200 IU/2 ml heparin block was applied.

| Outcome analysis
Patients were seen in the outpatient clinic after 3, 6, and 12 months and annually thereafter. Clinical, laboratory, and radiological evaluations were performed by a specialized orthopedic surgeon.
In addition, each case was discussed with an infectious disease specialist.

| RESULTS
Since

| Patient demographics
The mean patient age was 66 ± 15 years, with a mean American Society of Anesthesiologists score of III, Charlson comorbidity index of 5, and a body mass index of 29; 19 (70%) were female.
In one patient, both THA's were exchanged during the same stay and in another, a TKA and THA exchange on the ipsilateral side was conducted. Overall, we treated PJI in 17 TKA's and in 12 THA's. The mean total duration of OPAT was 66 ± 26 days. We recorded a total of 1738 outpatient days. One patient obtained 213 OPAT days before reimplantation, due to critical soft tissue (Table 5).
The deceased patient is included in the pathogen evaluation.
Appropriate antimicrobial therapy was selected based on the results of susceptibility testing with the first preferential of a bactericidal treatment. The most frequently used antimicrobial agent in Gram-positive pathogens was vancomycin and meropenem in Gramnegative (Table 6).
We Regarding the indication for S-OPAT (Table 1), we noticed in 14 (54%) cases the susceptibility-matched antimicrobial agents were only available for IV administration. Oral antibiotics with solely bacteriostatic activity were the reason for OPAT in 11 (42%) patients, while inadequate oral bioavailability was only noticed once.

| Adverse events
Overall, adverse events were recorded in 6 of 54 (11%) S-OPAT episodes. Five CRCs occurred (9%), giving a line-associated complication rate of 2.88 per 1000 S-OPAT days. In detail, we observed three midline catheter occlusions in two patients, leading to an exchange, as well as one midline catheter and one PICC-associated superficial thrombophlebitis. The catheters were removed and regular treatment was terminated a few days before scheduled.
An ADR was noted in one patient (2%; 0.43 per 1000 S-OPAT days). After 2 weeks of vancomycin use, acute renal impairment occurred. Antibiotic administration was paused for 3 days and then switched to daptomycin. The patient was able to remain in outpatient treatment. No critical or life-threatening complications were observed.

| Comparing the outcome
Our focus was the management of MDR pathogens resistant to all eligible oral antibiotics in PJI. Increasing numbers of primary total joint arthroplasties and revisions of hip and knee arthroplasty due to infections, 9,10 will lead to enlarged number of these microorganisms.
Between December 2015 and June 2019, those pathogens were identified in 27 out of 650 patients with PJI (4%). Therefore, their occurrence is markedly higher than fungal PJI, which has been reported in approximately 1% of all PJIs. 59 Even if these pathogens occur only rarely, there is a controversial discussion on how to treat them. The presented collective has not enough power for a subgroup analysis between knees and hips, midline versus PICC, or detailed demographic analyses and we are aware that this leads to heterogeneity within our collective.
However, we were able to present some fundamental findings, helping to treat these pathogens.
Although two-stage revision arthroplasty for the treatment of Our study has some limitations. We only investigated a short-term follow-up and 23 patients is a small sample size. Subcohort analyses, for example, between demographic parameters, THA versus TKA, and midline catheter versus PICC would be markedly underpowered and were, therefore, omitted.
However, some fundamental conclusions could be drawn. Prospective multicentre trials are needed to confirm our findings using our proposed treatment method to eradicate MDR pathogens in patients with PJI.