Background

Periprosthetic joint infection (PJI) is a complication of arthroplasty. It is characterized by bacterial biofilms on implants, which render systemic antimicrobial therapy ineffective^1,2,3,4. The minimum biofilm eradication concentration (MBEC) required is often 100 times higher than that required to inhibit bacterial growth alone^5,6. Surgical interventions, such as two-stage revision surgery for removing infected implants, frequently result in considerable muscle weakness and reduced mobility⁷. PJI treatment includes intravenous and local antibacterial treatments, with the latter being crucial for implant preservation by delivering antimicrobials at concentrations above the MBEC⁸.

Antimicrobial-containing polymethylmethacrylate is utilized for targeted, high-concentration antibiotic delivery⁹. However, maintaining a consistent dose with local antibacterial treatments is challenging, and postoperative dose adjustments are not feasible¹⁰. Continuous local antimicrobial perfusion (CLAP), used alongside conventional surgery, provides an effective infection management strategy by ensuring high-concentration, low-flow antimicrobial delivery^{11,12,13,14,15}. CLAP implementation, facilitated by specific devices, allows for postsurgical adjustment of concentration and flow rate, maintaining consistent local antimicrobial dosing. However, the antibiotic release from polymethylmethacrylate may decrease over time¹⁰.

Rationale

CLAP may enhance biofilm destruction and implant preservation in chronic PJI. However, no studies have evaluated its efficacy and safety in PJI. This study aimed to evaluate implant survival and complication rates in patients with PJI who underwent debridement, antibiotics, and implant retention (DAIR) surgery supplemented with CLAP.

Patients and Methods

Study design and setting

We conducted a retrospective chart review to identify and enroll patients diagnosed with PJI of the hip who underwent DAIR with CLAP between December 2010 and July 2023 at our institutions and without CLAP between March 2013 and February 2021 as the historical control group, and the participants were classified into the CLAP and non-CLAP groups. The institutional review board of Yokohama City University (No: 230900002) approved this retrospective study. All the procedures were performed in accordance with relevant guidelines. Written informed consent regarding potential complications was obtained at the time of CLAP treatment, and an online opt-out form was used for data collection. Approval for the off-label use of the CLAP tubes was obtained from the Ethics Committee. All procedures were conducted in accordance with the Declaration of Helsinki.

Participants

DAIR with CLAP was performed in all patients with PJI without implant loosening, provided the CLAP device was ready on the day of surgery. Patients with PJI whose implants had septic loosening were indicated for one or two-stage revision surgery. Patients with chronic PJI infection without the implants loosening also underwent DAIR upon request rather than a two-stage revision if the patient’s condition, range of infection, and causative organisms were identified. This study included only patients who had undergone surgery at least 1 year postoperatively. Patients who died over 1 year postoperatively due to causes unrelated to PJI or CLAP complications were included, whereas those who died within 1 year postoperatively due to factors unrelated to surgery were excluded.

We utilized the diagnostic criteria from the 2018 International Consensus Meeting on PJI, which defines PJI based on either major criteria or a score of ≥ 6 points¹⁶. For this purpose, we investigated the following: the presence of a fistula, number of specimens positive for bacterial culture, phase of infection (acute or chronic), preoperative serum C-reactive protein level, D-dimer level, erythrocyte sedimentation rate, neutrophilic infiltration on postoperative pathology, and intraoperative abscess formation.

Description of treatment or surgery

During DAIR surgery, bone and soft tissue were debrided using 0.35% saline-diluted povidone-isodine. Easily replaceable implants, such as sliding surfaces, were replaced as feasible. CLAP was performed according to the following procedure in the CLAP group, and drains were inserted only in cases considered necessary in the non-CLAP group.

Intraoperative administration route for CLAP

Intra-soft tissue antibiotic perfusion (iSAP): A Salem Sump tube (Cardinal Health K. K., Tokyo, Japan) > 20 Fr (Fig. 1a–c) was used. This tube has a double-lumen structure, allowing antimicrobial administration from one side and exudate collection under negative pressure from the other. After debridement, Salem Sump tubes were placed intraoperatively within the infected soft tissue, such as in the upper or lower layers of the fascia. This tube was originally designed for use as a gastric tube. Its application in this treatment constitutes an off-label use. This method was employed only after obtaining comprehensive informed consent from all patients.

Device components of the continuous local antibiotic perfusion treatment system are shown. (a, b). Salem Sump tube (Cardinal Health K.K., Tokyo, Japan). This double-lumen tube allows for antimicrobial administration on one side and negative pressure application on the other. It features multiple holes for suction under continuous negative pressure to prevent blockage and is used for intra-soft tissue or intra-joint antibiotic perfusion. (c). Connector VRF606 (ISIS Co., Ltd., Osaka, Japan). The nasogastric tube is connected to the Salem Sump tube, constructing a route for antimicrobial administration. (d). An intramedullary antibiotic perfusion pin (Cubex Medical, Tokyo, Japan). These pins are fixed to the cortical bone, enabling antimicrobial administration into the bone marrow, as indicated by the white arrow. (e). Intramedullary antibiotic perfusion pin insertion. After insertion into the ala of the ilium, saline is injected intraoperatively (white arrow), while saline dispersal around the acetabulum is verified. (f). RENASYS TOUCH (Smith & Nephew, Watford, United Kingdom) system. It applies a continuous negative pressure of 60–80 mmHg to perfuse and collect local antimicrobials. The drainage fluid will be discarded. (g). Application of foam filler (Smith & Nephew, Watford, United Kingdom). Placed immediately above the wound, it applies continuous negative pressure to perfuse the wound, administering antimicrobials through the right tube while simultaneously collecting drainage fluid. (h). Y-connector branch tip (Smith and Nephew, Watford, United Kingdom). The tip is cut and connected to a Salem Sump tube for continuous negative pressure suction.

Intra-joint antibiotic perfusion (iJAP): Salem Sump tubes were placed in the hip joint to perfuse the joint with antibacterial agents and eradicate the biofilm around the implants, guided by preoperative computed tomography imaging or intraoperative findings.

Intramedullary antibiotic perfusion (iMAP): An iMAP pin (Cubex Medical, Tokyo, Japan) was used. This hollow pin allows the injection of antimicrobials from one end and discharge on the tap side for cortical bone fixation and antimicrobial administration into the bone marrow (Fig. 1d). In cases of osteomyelitis identified preoperatively, an iMAP pin was inserted into the infected bone marrow to facilitate antibacterial perfusion (Fig. 1e). Bone holes were created using a Kirschner’s wire to establish an effective perfusion system to enhance the delivery of antimicrobials from the infection site in the suction direction.

Wound dressing

The wound was sparsely sutured during surgery to ensure complete dermal penetration of the antimicrobials administered using the CLAP device and to apply negative pressure to the wound effectively. Immediately after wound closure, the RENASYS TOUCH (Smith & Nephew, Watford, United Kingdom) was used as a negative pressure wound therapy (NPWT) device (Fig. 1f, g). A cotton foam filler (Smith and Nephew, Watford, United Kingdom) was placed over the wound to prevent subcutaneous hematoma formation. Subsequently, a waterproof, airtight film was applied to cover the area (Fig. 1g). A round hole several centimeters in diameter was made in the center of this film, directly above the foam, to attach a soft port linked to the RENASYS TOUCH through a Y-connector, and the setup depended on the number of connected Salem Sump tubes (Fig. 1h). The cotton foam filler was replaced once weekly, and the wound was examined.

Postoperative antibiotics

Gentamicin sulfate eradicates biofilms effectively in susceptible general bacteria and in methicillin-resistant bacteria^2,17. Thus, for bacterial infections, regardless of the causative bacteria or resistance profile, dilutions of gentamicin sulfate, amikacine sulfate or arbekacin sulfate were prepared in saline at 1.2 or 2.0 mg/mL, respectively (Fig. 2a–d). These solutions were continuously administered at a rate of 2.0 mL/h/route for 24 h daily using a continuous precision pump through a Salem Sump tube. For fungal infections, micafungin sodium, which is effective for biofilms, was diluted in saline to 50 µg/mL and administered at the same rate and duration¹⁸. Notably, each route was flushed daily with 3.0–10.0 mL of gentamicin-diluted solution or saline solution to prevent clogging. The gentamicin diluent was stopped one day before the scheduled removal of the CLAP device, allowing the antibacterial agent or fluid to be collected in the tissue over a day. If CLAP administration was obstructed by tube clogging, the CLAP was removed. Tubes, pins, or devices could be easily removed from a patient’s bedside.

Continuous local antibiotic perfusion treatments. (a). Treatment for periprosthetic joint infection after revision total hip arthroplasty. It involves debridement with implant retention. One intra-soft tissue antibiotic perfusion device is placed distally and one proximally around the femur and acetabulum. An intra-joint antibiotic perfusion device is also placed in the hip joint, all perfused with local antimicrobials. (b). Antimicrobial pathways. White arrows indicate the routes of antimicrobial administration; white dotted lines indicate the collection of drainage fluid after perfusion. (c). Treatment following hip-resurfacing arthroplasty. The treatment is performed for periprosthetic joint infection in combination with debridement and implant retention. One intra-soft tissue antibiotic perfusion device is placed distally around the femur and one intra-joint antibiotic perfusion device is placed in the hip joint. In addition, one intramedullary antibiotic perfusion device is placed from the ala of the ilium, all perfused with local antimicrobials. (d). Administration and drainage details. White arrows indicate the route of antimicrobial administration, and white dotted lines indicate the collection of drainage fluid after antimicrobial perfusion. Drilling through the acetabulum allows for antimicrobial perfusion.

Postoperatively, a broad-spectrum systemic antimicrobial was used intravenously at first, and de-escalation was performed based on bacterial culture results. Anti-methicillin-resistant Staphylococci (MRS) antimicrobials were used in combination for the patients with the molecular diagnosis positive for MRS. Systemic antimicrobials were discontinued at the surgeon’s discretion once the serum inflammation improved. If the blood tests indicated persistent inflammation in chronic suppression owing to the patient’s condition, antibacterial medication was continued. All the complications due to CLAP use that required treatment adjustment or discontinuaction were documented.

Variables, outcome measures, data sources, and bias

The following characteristics were investigated: age at the time of surgery, sex, body mass index (kg/m²), follow-up duration since surgery, implant type, the host’s compromised medical history, causative organisms identified in cultures, and sensitivity to gentamicin sulfate. Patients who died over 1 year post-surgery were considered to have completed the follow-up period. Treatment details assessed included operative time, intraoperative blood loss, CLAP type (iSAP, iJAP, or iMAP), treatment duration until CLAP device removal, number of DAIR surgeries involving CLAP, types of antibacterial or antifungal agents used, hospitalization duration, length of postoperative systemic antibacterial treatment (duration of intravenous and oral administration), and activities of daily living 1 year postoperatively.

The primary endpoint was the implant survival rate post-surgery, with failure indicated by either the need for implant removal or replacement due to recurrent infection. The secondary endpoint was the presence or absence of adverse events including gentamicin or arbekacin levels in the blood at 3 days and 1 week postoperatively and immediately before the removal of the CLAP device. Adverse events assessed were renal impairment (increased blood creatinine level; renal function deterioration was defined as an increase in blood creatinine concentration of ≥ 0.30 mg/dL¹⁹), hepatic impairment (increased aspartate transaminase or alanine aminotransferase levels), pancytopenia, hearing impairment, visual impairment, skin impairment, and allergic symptoms. Serum assays for white blood cell counts and C-reactive protein levels were conducted immediately before surgery, and at 3 days, 1 week (± 1 day), 2 weeks (± 1 day), 3 weeks (± 1 day), 1 month (± 2 days), and 3 months (± 2 weeks) postoperatively.

Statistical analysis and Study size

All the PJI cases with or without CLAP were retrospectively analyzed. The collected items were compared between the CLAP group and the non-CLAP group, which was used as the historical control group. All the statistical analyses were performed using JMP Pro version 17.0 (SAS Institute, Inc., Cary, NC, USA). Statistical significance was determined using Student’s t-tests. Non-parametric Mann–Whitney tests were applied to data sets that were not normally distributed or were not of equal variance. Fisher’s exact test was used to analyze associations between categorical variables. Statistical significance was set at p < 0.05. Graphs were created in Prism 9 (MDF Co., Ltd., Tokyo, Japan), showing the interquartile range and the mean as a broken line.

Demographics and Description of the study population

The screening dataset included 24 patients (15 women and nine men) in the CLAP group and 10 patients (six women and four men) in the non-CLAP group. Of these patients in the CLAP group, two (a male and a female) died within a year post-surgery owing to factors unrelated to CLAP. We analyzed the data from the remaining 22 patients (14 women and eight men); two of them (nos. 8 and 17) who died of causes unrelated to the surgery over 1 year postoperatively without infection recurrence and were included in the study. In contrast, no cases in the non-CLAP group were lost to follow-up, and all 10 cases were included in the analysis.