Medical innovations to maintain the function in patients with chronic PJI for whom explantation is not desirable: a pathophysiology-, multidisciplinary-, and experience-based approach

Abstract

Introduction: PJI is the most dramatic complication after joint arthroplasty. In patients with chronic infection, prosthesis exchange is in theory the rule. However, this surgical approach is sometimes not desirable especially in elderly patients with multiple comorbidities, as it could be associated with a dramatic loss of function, reduction of the bone stock, fracture, or peroperative death. We propose here to report different approaches that can help to maintain the function in such patients based on a pathophysiology-, multidisciplinary-, and an experience-based approach.

Methods: We describe the different points that are needed to treat such patients: (i) the multidisciplinary care management; (ii) understanding the mechanism of bacterial persistence; (iii) optimization of the conservative surgical approach; (iv) use of suppressive antimicrobial therapy (SAT); (v) implementation of innovative agents that could be used locally to target the biofilm.

Results: In France, a nation-wide network called CRIOAc has been created and funded by the French Health ministry to manage complex bone and joint infection. Based on the understanding of the complex pathophysiology of PJI, it seems to be feasible to propose conservative surgical treatment such as "debridement antibiotics and implant retention" (with or without soft-tissue coverage) followed by SAT to control the disease progression. Finally, there is a rational for the use of particular agents that have the ability to target the bacteria embedded in biofilm such as bacteriophages and phage lysins.

Discussion: This multistep approach is probably a key determinant to propose innovative management in patients with complex PJI, to improve the outcome.

Conclusion: Conservative treatment has a high potential in patients with chronic PJI for whom explantation is not desirable. The next step will be to evaluate such practices in nation-wide clinical trials.

Keywords: Antibiotics; Bacteriophages; Innovations; Lysins; PJI; Phage therapy; Suppressive antimicrobial therapy.

Figure 1

The CRIOAc network in France labeled by the French Health Ministry (Direction Générale de l’Offre de Soins): reference centers appear in orange rounds, associated centers in white rounds and the address of residence of the 647 new patients managed in the CRIOAc Lyon in 2019 in yellow diamond (adapted from the maps available here: http://www.crioac-lyon.fr/origine-patients.html).

Figure 2

Pathophysiology of prosthetic joint infection with formation of pus due to planktonic bacteria into the joint and formation of bacterial biofilm at the implant surface: Plantonik bacteria are bacteria in “optimal” environmental conditions to growth in the joint liquid, leading to recruitment of polymorphonuclear cells (PMNs), and formation of pus, composed by bacterial and PMNs remnants; Bacteria embedded in biofilm at the implant’s surface, that is inseparable from the implant surface, and tolerant to the immune system.

Figure 3

Skin necrosis following knee surgery: local status of a 72-year-old male with skin necrosis 21 days after the revision of a knee arthroplasty (lateral approach). See the multiple scar and white discoloration of the skin in area healed by secondary intention relative to previous superficial necroses. From an anterior (panel A, B and C) and lateral (Panel D, E, F) views, schematic representation of scar (white dotted line), necrosis area (black area) and area healed by secondary intention (hatched area) were added to the photo (panel B and E) and after subtracting the patient’s skin.

Figure 4

Skin necrosis following DAIR: local status of a 85 year-old-male, experiencing a skin necrosis at 21 days of performance of a left knee arthroplasty (panel A); A DAIR with a medial gastrocnemius flap was performed with satisfactory early results at 14 days (panel B), notice the poor reliability of the skin flap surrounding the anterior tibial tuberosity; and favorable outcome with stable soft tissue coverage at 3 month (panel C).

Figure 5

Example of dosage individualization based on PK/PD in a patient treated with ertapenem as SAT for a relapsing PJI: 78-old woman who had a relapsing left prosthetic hip infection due to Enterobacter cloacae (only susceptible to ertapenem with a MIC of 0.064 mg/L) for whom iterative DAIR was done with persistence of the organism (panel A). Explantation was contraindicated as it was a revision prosthesis without loosening with a high risk of peroperative complication and loss of function (panel B). She received as primary antibiotics following the DAIR ertapenem at the conventional dose of 1 g/day. Subcutaneous administrations were firstly performed using a butterfly needle with injection each day in the right thigh of in the abdominal flanks (panel C), with secondary systemic diffusion of the drug in blood, and then at the site of infection. Ertapenem drug concentrations were measured, with three samples collected in red (panel D; the x-axis shows the time, the y axis represents the ertapenem plasma concentration at the steady-state; The blue marks on the x-axis show drug administrations): pre-dose, 30 min after the end of the 30 min SC infusion and 5 h post-dose. Ertapenem individual PK parameters were then estimated by a Bayesian approach based on our published population PK model of ertapenem implemented into the BestDose™ software. Panel D shows the results of the model fitting (black line for estimated concentrations during time), which was very good, and we simulated a future regimen with 1 g of ertapenem every 48 h. Our plasma concentration target for this patient was the bacterial MIC (0.064 mg/L) corrected for ertapenem protein binding (free fraction of 5%), resulting in 1.28 mg/L. We calculated that 1 g of ertapenem every 48 h would result in a trough concentration of 0.4 mg/L and 63% of time spent above the target after 48 h. While the optimal value would be 100%, this was considered acceptable, considering that 40% was reported to be sufficient to get a bactericidal effect with this agent [54].

Figure 6

A phage and its lytic circle: firstly, the phage recognizes the specific bacteria, and injects its genetic material. The phage replicates itself into the bacterial cells by hacking the bacterial replication system and produces a lysin that induces a bacterial cell wall rupture, thus freeing hundreds of new phagic components that can in their turn target other bacterial cells located in the close environment in an exponential and self-sustained reaction.

Figure 7

Activity of antibiotics alone or in combination on planktonic bacteria and on bacteria embedded in biofilm at a PJI surface: Antibiotics are active on planktonic bacteria, but not on bacteria embedded in biofilm. The phages are active on planktonic bacteria, and have the ability to replicate themselves among planktonic bacteria. The phage has the ability to disrupt the biofilm and acts synergistically with antibiotics to kill the bacteria in the biofilm and the bacteria released by the biofilm. In addition, synergistic effects are observed when phages are combined with antibiotics.