The risk of prosthetic joint infection (PJI) following total hip or knee arthroplasty is 1 to 2% [1–4]. With an increasing number of arthroplasties performed, the number of PJI patients continuously increases [3]. PJI poses severe complications for patients, burdens healthcare systems, and challenges the global use of antimicrobials [5].
Therefore, gaining more knowledge about PJI development and finding and testing new diagnostic tools and treatment strategies is crucial. Large animal models that can imitate the clinical aspect of PJI are essential for this purpose [6, 7]. A PJI animal model should closely replicate the periprosthetic environment to ensure the highest clinical relevance. The prosthesis should be made of relevant materials, design, and geometry, separate the intraarticular space from the intramedullary space, and support load. Further, the model should be performed in animals with immunological and musculoskeletal properties comparable to humans, such as swine [6].
Since arthroplasty procedures, and consequently PJIs, are primarily associated with adult patients, clinically relevant animal models of PJI must be based on fully grown and skeletally mature animals. Therefore, Göttingen minipigs, which reach a mature weight of 30 to 50 kg and skeletal maturity at approximately 18 months [8], are a viable choice for arthroplasty and PJI studies. The current study aimed to develop a surgical procedure for hip replacement in adult Göttingen minipigs and a comprehensive animal welfare and post-mortem evaluation protocol. The study led to the first Göttingen minipigs with a hip hemiarthroplasty, paving the way for future minipig PJI models of the highest clinical relevance.
Methods
Animals and housing
The experimental protocol was approved by The Danish Animal Experiments Inspectorate (license no. 2022-15-0201-01130). The study was conducted at the University of Copenhagen, Frederiksberg, Denmark. Three adult female Göttingen Minipigs (Ellegaard Göttingen Minipigs A/S, Dalmose, Denmark) were included. The minipigs (40–48 kg body weight) were 19–26 months old at study start. All minipigs had previously been used for breeding (1–3 litters). The minipigs were acclimatized for two weeks prior to start and barrier-housed in single pens with the possibility of snout-to-snout contact. The minipigs were fed twice daily with a commercial pig diet (Brogaarden Altromin, 9069 – Extrudate), had free access to tap water, and had a 12-hour light/ dark cycle.
Anesthesia, analgesia and euthanization protocol
Intramuscular sedation was followed by intravenous infusion of Propofol, and intraoperative analgesia was achieved by intravenous infusion of Fentanyl. Intra-operative and postoperative analgesia was achieved with an epidural block consisting of Morphine and Bupivacain sterilely injected into the epidural space between the lumbosacral junction of L6 and S1 (Fig. 1) [9]. The reported postoperative analgesic time for this procedure is 24 hours in dogs [10]. Before surgery, the minipigs received an intramuscular injection of Meloxicam, providing postoperative analgesia for 24 hours. Following surgery, the minipigs received daily oral analgetic treatment with Meloxicam. In case of lameness or observed pain behavior, an intramuscular injection of Buprenorphine was provided every eighth hour. The minipigs were euthanized by intravenous injection of an overdose of pentobarbital. See supplementary materials (Supplementary file 1) for complete protocols (time, dose, supplier).
Prosthesis size, material, and design
Computed Tomography scanning and anatomical evaluations of five cadaveric adult female Gottingen minipigs revealed a uniformity of femur concerning size and geometry, allowing the same size and prosthesis design to be used in all animals. Cementless, titanium alloy (Ti-6AI-4V ELE (EBM)) Universal Hip Canine BFX® Femoral Stem (BioMedtrix, New Jersey, USA) size 5 (designed for medium-sized dogs) was used together with a Cobalt Chrome Universal Hip Canine Femoral Head (BioMedtrix, New Jersey, USA) size 17 + 0mm. A hemiarthroplasty was performed, i.e. no insertion of an acetabular cup component.
Surgical procedure by anterior approach to the hip joint
All minipigs were placed in left lateral recumbency. Anatomical landmarks of the greater trochanter, hip, and knee joint were marked. The entire thigh area, from the knee to the spine, was clipped and aseptically prepared by washing using Medi-scrub (Medi-Skrub, Meda AS, Allerød, Denmark) and disinfection with chlorhexidine-ethanol (0.5% chlorhexidine, 85% ethanol). Surgical drapes were placed around the surgical area, and the area was covered with Ioban (3M Ioban 2, 3Mdenmark) to reduce the risk of contamination [11].
The anterior approach to the hip joint was applied by a modified version of the "Approach to the Craniodorsal Aspect of the Hip Joint Through a Craniolateral Incision" described by Piermattei and Johnson, 2004 [12]. An incision was made through cutis and subcutis, starting two-thirds distally towards the knee, continuing towards the center of the greater trochanter, and ending halfway between the greater trochanter and the dorsal midline (Fig. 2A). The dense subcutaneous fat was undermined, and retractors were placed in the cutis and subcutis. A fasciae-incision was made in the cleavage along the cranial border of the biceps femoris muscle and the superficial gluteal muscle, and the caudal border of the tensor fasciae latae muscle (Fig. 2B). The biceps femoris muscle and superficial gluteal muscle were retracted caudally, whereas the tensor fasciae latae muscle was retracted cranially. The horizontal fibers of the middle gluteal muscle were now visible (Fig. 2C). By blunt dissection, the middle gluteal muscle was detached from the underlying muscles and retracted dorsally. The deep gluteal muscle and its attachments to the greater trochanter were now visible, and the joint space could be palpated underneath when rotating the femur externally. The deep gluteal muscle was cut near the greater trochanter and, by blunt dissection, detached from underlying tissue and retracted cranially (Fig. 2D). The cranial aspect of the joint capsule was now exposed, and a T-shaped incision was placed to open the joint. External femur rotation made the femoral head visible (Fig. 2E).
Caput femoris was elevated from the acetabular cup by a Hohmann retractor, and the femoral head ligament was cut by curved scissors. Femoral neck resection was performed close to the greater trochanter by an oscillating saw at a 45o angle (Fig. 2F). The proximal femur part was elevated to allow reaming. Prior to reaming, drilling (Drill 4mm and 5mm, BioMedtrix, New Jersey, USA) was used to access and enlarge the femoral medullary cavity (Fig. 2F). The femoral tapered reamer size four (BioMedtrix, New Jersey, USA) was used to widen the femoral medullary cavity prior to the impaction of broach size 4 followed by broach size 5 (BioMedtrix, New Jersey, USA). The femoral stem was inserted by press-fit attachment using the femoral impactor tool (Stem Impactor and Impactor Handle, BioMedtrix, New Jersey, USA) and gentle mallet strokes. The femoral head was attached to the femoral neck of the stem. Before the reduction of the joint, the acetabular cup was cleared up for remnants of the femoral head ligament. The surgical site was closed in layers, starting with the joint capsule, which needs to be closed thoroughly to add additional joint stability. Next, incised muscles were sutured, and subcutis and cutis were closed in separate layers. An antimicrobial ointment was applied to the surgical wound (Fucidin 2% Ointment, Leo Pharma, Ballerup, Denmark), followed by wound plast spray (Kruuse Wound Plast, Kruuse, Langeskov, Denmark).
Radiographic evaluation
X-rays were obtained in lateral and ventro-dorsal positions using an x-ray system (Shimadzu Radspeed MC, Fuji, Tokyo, Japan) set to 79 kv and 18 mAs, respectively. The radiographs were taken during general anesthesia at the following time points: immediately after surgery (day 0) to ensure correct prosthesis placement, at day 14 post-surgery to give a midway study evaluation of the prosthesis, and at planned euthanasia.
Postoperative care and clinical evaluation
The minipigs were monitored closely throughout the surgical recovery phase. Rubber mats were placed on the floor to ensure a non-slip surface. Initially, the minipigs were assisted with a blanket under the belly to minimize load on the operated legs when trying to stand. When able to stand and walk without trembling, the blankets were removed. The minipigs were monitored several times a day following surgery and clinically evaluated once daily with a scoring of gait, wound, general status, activity level, and signs of pain-related behavior. Impaired ability to stand, anorexia and systemic signs of infection, i.e., abnormal respiration or high fever, were set as humane endpoints.
Macroscopic pathology
Following euthanasia, the minipigs were necropsied. The surgical wound was inspected and opened in layers, i.e., cutis, subcutis, and the different muscle layers using aseptic techniques. Tissue samples for histology were collected from each layer. The joint capsule was cut open using sterile instruments, and a sterile synovial fluid sample was collected using a syringe. The synovial membrane was collected for histology. The position of the artificial femoral head was evaluated, and the femur was removed together with the prosthesis. The prosthesis stability within the femoral bone was evaluated and followed by aseptic removal. The femoral bone was evaluated for signs of pus/abscess formation, necrosis, granulation tissue, fibrosis, and new bone formation. The proximal half of the femoral bone, which had formed the interface with the prosthesis, was cut into transverse sections of approximately 0.5 cm numbered from proximal to distal and collected for histological examination (Fig. 3). The left hip joint and femoral bone were opened, removed, and sampled correspondingly. The acetabulum from both hip joints was evaluated for gross lesions and sampled for histological evaluation. The major left and right deep inguinal lymph nodes were sampled for histology. The thorax and abdomen were opened, and all organs were inspected in situ. Samples from the liver, right kidney, and left caudal lung lobe were collected for histology.
Histology
All tissue samples were fixed in 10% buffered formalin for five days. Following fixation, the bone samples were decalcified in a solution of 3.3% formaldehyde and 17% formic acid for six weeks. After fixation or decalcification, all samples were trimmed and processed through graded concentrations of alcohol and xylene, embedded in paraffin and sectioned (4–5 µm thickness). All sections were stained with Hematoxylin & Eosin (HE). On indication, bone sections were stained with immunohistochemistry (IHC) using antibodies towards staphylococci [13].
Microbiology and sonication of prosthesis
Prior to surgery, nostril swabs, venous blood samples, and a cutis sample from the surgical field were collected for microbiological examination. Additionally, blood samples were collected prior to euthanization. During necropsy, tissue samples from the lung, cutis, subcutis, the biceps femoris muscle, the deep gluteal muscle, the synovial membrane, and three samples from the right femur, two from the proximal part of the prosthesis interface and one from the distal part of the prosthesis interface, were collected aseptically, using new sterile instruments for each sample. Furthermore, sutures from the subcutis, the profound muscle layer, and the joint capsule were collected for microbiological examination. Soft tissue, synovial fluid, blood, and swabs were inoculated on blood agar plates. Sutures were suspended in 100 µL sterile saline, of which 10 µL was inoculated on blood agar plates. All plates were incubated at 37oC for 24 hours under aerobic conditions. Morphologically distinct colonies were selected and identified by Matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS) (Vitek MS RUO, bioMérieux, Marcy-l'Etoile, France) [14]. Bone samples were aseptically homogenized and serially diluted in sterile isotone saline before 100 µl were plated on blood agar plates. The plates were incubated at 37oC for 24 hours under aerobic conditions. Morphologically distinct colonies were counted to estimate CFU/ml and identified with MALDI-TOF MS. Prostheses, collected aseptically at necropsy, were placed in 50 ml sterile centrifuge tubes and covered with sterile isotone saline. All prostheses were sonicated in an ultrasound bath for improved CFU counts [15, 16]. All microbiological evaluations were performed blinded. Whole genome sequencing was performed as previously described [17] on selected identified isolates to investigate relationship and origin.
Synovial fluid
The color, viscosity and turbidity were evaluated, and two smears from each side were cytologically examined. If the cellularity was high enough, a differential count of leucocytes was carried out. The cytological examination of the synovial smears was carried out blinded.