This study represents our initial approach to development of a novel ovine bilateral proximal phalanx fracture model. Twelve sheep were included; and the lateral proximal phalanges of both front limbs underwent fracture surgery. In 8 sheep a novel fracture repair modality was tested in one of the two phalanges, these results are reported elsewhere (31), whilst the remaining 16 phalanges were repaired using metallic implants as described below. The bilateral approach enables within-subject comparisons, reducing animal use in line with the 3Rs of animal experimentation (Replacement, Reduction, and Refinement) (32). With one sheep euthanized on day six due to systemic illness, the final sample was 15 phalanges. The fracture models investigated were transverse proximal phalanx midline ostectomy with 3 mm (n = 4 phalanges) or 6 mm (n = 4 phalanges) gaps, and midline osteotomy with central 4.5 mm drill hole (n = 7 phalanges) (Figures 1A,B). Fractures were fixated with one dorsolaterally placed stainless-steel locking plate. The sheep had either a walking cast or a custom-made offloading hoof block combined with bandages as coaptation (Figure 1C). By the end of the study period (16 weeks), the sheep were lightly sedated with xylazine (0.08 mg/kg BW, IV, Xysol Vet., ScanVet Animal Health A/S, Denmark) and butorphanol tartrate (0.04 mg/kg BW, IV, Dolorex, MSD Animal Health A/S, Denmark) and euthanised by pentobarbital (140 mg/kg BW, IV, Euthasol vet., Dechra Veterinary Products A/S, Denmark). Bone healing was evaluated by repeated radiographs throughout the 16-week study period and by post-euthanasia computed tomography (CT) scans. Animal welfare was analysed by repeated clinical examinations, lameness, pain face analysis, and by implementing a modified pain behaviour score as detailed below.

Eleven skeletally mature Texel ewes (5.5 years ± 3 months, 73 kg ± 7.6 kg) were included. The animals were kept in pairs indoors in 6 m² pens with wood shavings as bedding. Their diet was adjusted according to nutritional requirements, and straw and water were provided ad libitum. During a 3–10 weeks acclimatization period prior to surgery, clinical health assessments were conducted following the European Animal Welfare Indicators (AWIN) protocol for sheep (33). Additionally, the sheep were trained by positive reinforcement to undergo clinical procedures, including sample collection, clinical examinations, and lameness assessment when led by a leash, so that these procedures could be conducted with minimal stress. In this acclimatization period, routine blood and faecal analyses were performed to ensure enrolment of only healthy animals, and the sheep were treated with Albendazole (4 mg/kg BW, Valbazen Vet., Zoetis Animal Health ApS, Denmark). To facilitate repeated monitoring of core temperature, a temperature-sensitive chip was inserted intramuscularly in Musculus omotransversarius, by identifying the fascial border between Musculus cleidooccipitalis and the cervical part of Musculi trapezii at the dorsal third of the neck (Global-Ident Bio-Thermo, Allflex, France).

The sheep were initially fasted for 36 h preoperatively (n = 8) (34) but based on the frequent occurrence of moderate–severe diarrhoea in the immediate postoperative period, fasting duration was reduced to 24 h (n = 4) with no adverse effects during surgery and improved gastrointestinal health post-surgically (35). Water was withheld for 4 h before the scheduled surgery time. Premedication for anaesthesia included intravenous (IV) administration of xylazine hydrochloride (0.05 mg/kg, Xysol Vet., ScanVet Animal Health A/S, Denmark) and butorphanol tartrate (0.05 mg/kg, IV, Dolorex, MSD Animal Health A/S, Denmark). General anaesthesia was induced using diazepam (0.5 mg/kg, IV, Diazedor Vet., Salfarm Danmark A/S, Denmark) and ketamine hydrochloride (2.5 mg/kg, IV, Ketador Vet., Salfarm Danmark A/S, Denmark). Inhalation anaesthesia was maintained with isoflurane (1.5%, Vetflurane, Virbac Danmark A/S, Denmark). Tetanus antitoxin (1,500 units/animal SC, Tetanus antitoxin Equine Origin, Colorado Serum Company, Colorado) and antibiotics were administered prior to surgery. Antibiotic therapy included gentamicin sulphate (6.6 mg/kg BW, IV, Gentavet, ScanVet Animal Health A/S, Denmark) and benzylpenicillin sodium (22,500 IU/kg BW, IV, Benzylpenicillin, Panpharma Nordic AS, Norway), with the latter re-dosed every 80 min during surgery. Intraoperative analgesia consisted of flunixin meglumine (2.2 mg/kg BW, IV, Wellicox, Ceva Animal Health A/S, Denmark), and a SC ring block on each forelimb using bupivacaine hydrochloride (0.1 mg/kg BW, SC, Bupivacaine Baxter, Baxter A/S, Denmark) and mepivacaine hydrochloride (0.4 mg/kg BW, SC, Mepidor Vet, Salfarm Danmark A/S, Denmark).

The sheep were placed in dorsal recumbency with flexed, stabilised forelimbs. A tourniquet was applied mid-metacarpally. A dorsolateral curved incision was made, and the bone was exposed by blunt dissection and periosteal elevation. The extensor and flexor tendons were retracted and protected. In the ostectomy group removal of the bone section was conducted by first approximating the middle of the phalanges from an intraoperative radiograph. Then a conical 0.6 mm sawblade (DePuy Synthes) was applied to cut the bone halfway through, after which a custom made 3- or 6-mm cutting guide was inserted to guide the saw blade and make a standardised ostectomy size (Figure 1A; Supplementary Figures 1a,b). Each sheep in the ostectomy group received both a 3 mm and a 6 mm ostectomy, randomly assigned to either the left or right front limb proximal phalanx (Table 1).

In the osteotomy group (n = 7) a dorsal midshaft 4.5 mm unicortical defect was created using a 4.5 mm drill bit, followed by a midline osteotomy performed with a conical 0.6 mm sawblade (46/25 × 6 × 0.6/0.4 mm (519.230S), DePuy Synthes, Johnson & Johnson MedTech, Massachusetts, USA). After ostectomy or osteotomy, metal implant fixation was performed (Table 1).

The 3- and 6-mm ostectomies were fixated with a 6-hole cuttable stainless-steel locking plate (1.5 mm, Straight (02.114.005S), DePuy Synthes, Johnson & Johnson MedTech, Massachusetts, USA) using four unicortical locking screws (cortex stardrive, Ø 1.5 mm, L 10 mm (02.214.110S), DePuy Synthes, Johnson & Johnson MedTech, Massachusetts, USA), two in each fragment (Figure 1A). The osteotomies were fixated with a five-hole cuttable stainless-steel locking plate (shorter version of plate used for ostectomies) with either four unicortical screws (n = 3 phalanges, subgroup A, same screw types as described for ostectomies) or four bicortical screws (n = 4 phalanges, subgroup B, Ø 1.5 mm, L 16 mm (VS106.018), DePuy Synthes, Johnson & Johnson MedTech, Massachusetts, USA) (Figure 1B; Supplementary Figure 1c,d).

After the osteosyntheses were completed, the soft tissues were closed by simple continuous pattern (fascia) (Novosyn 2/0, B Braun Medical Inc., USA) and simple interrupted pattern (skin) (Optilene 2/0, B Braun Medical Inc., USA).

To find the most useful coaptation, two different approaches were adopted. In the osteotomy sheep (n = 4) a hoof block adjusted to fit an ovine digit was applied. This had a height of 20 mm (Hoof block, standard, Klovshoppen, Denmark), and it was glued (Mini Moo Gloo, MooGloo, Durham, USA) under the medial digit on both front limbs for offloading of the osteosynthesized digit. Additionally, a supportive and protective elastic bandage was applied over the surgical site (Figure 1C). In osteotomy subgroup A (n = 3 sheep), a half-limb walking cast was applied for approximately 4 weeks post-surgery (Figure 1C), hereafter elastic bandages and hoof blocks were applied until week 8 (Figure 1C). In subgroup B (n = 4 sheep), we opted to use hoof block offloading and bandages as described for the ostectomy group. All sheep were permitted to ambulate unrestrictedly post-surgery. In all sheep, the elastic bandages were removed at week 8 and the hoof blocks were removed at week 10 after surgery. In the osteotomy group, one sheep was euthanized on day 70 due to hoof trauma.

Radiographs were obtained throughout the 16-week study period, i.e., intra-operatively and at 1, 2, 4, 6, 8, 10, 12, 14, and 16 weeks. For the post-surgery radiography, the sheep were lightly sedated with xylazine (0.08 mg/kg BW, IV, Xysol Vet., ScanVet Animal Health A/S, Denmark) and butorphanol tartrate (0.04 mg/kg BW, IV, Dolorex, MSD Animal Health A/S, Denmark). At week 16, radiography and CT were acquired immediately after euthanasia. Each radiographic study entailed three projections: dorsopalmar, dorsomedial-palmerolateral oblique, and lateromedial. A portable Gierth TR80/20 system was used, set to 60 kV and 3 mAs, along with Fujifilm FDR D-EVO console software. For CT, a Siemens Somatom Emotion (46,213, 14o kV) was used.

From the radiographs, the following parameters were recorded to investigate fracture fixation stability and bone healing:

Callus thickness was measured from CT images, assessing the radiographically dense (mineralised) hard callus visible on the scans (Supplementary Figure 2). Callus thickness was measured in two designated areas on each fragment, yielding four measurements per sheep. The mean value was reported. The first measurement was taken at the subjectively identified thickest region of callus on transverse CT images. From this point, a line was drawn towards the centre of the bone, and a second line was drawn from the centre at a 45° angle relative to the first line, towards the cortex, directed away from the plate. Callus thickness was then measured along this second line (Supplementary Figure 2).

Following surgery, all sheep received antimicrobial and analgesic treatment. Gentamicin sulphate was administered once daily (6.6 mg/kg, IV, Gentavet, ScanVet Animal Health A/S, Denmark), and benzylpenicillin sodium was given intravenously every 6 h (22,500 IU/kg, IV, Benzylpenicillin, Panpharma Nordic AS, Denmark) – the ostectomy groups received antimicrobial treatment for a duration of 7 days, this was downregulated for the osteotomy group for a duration of 3 days.

For pain management, flunixin meglumine (2.2 mg/kg, IV; Wellicox, Ceva Animal Health A/S, Denmark) was administered twice daily for the first 3 days, followed by once-daily dosing until day eight.

The sheep were closely monitored throughout the study period. Throughout the study period they were observed by technical staff twice daily, where appetite, general well-being and intactness of the coaptation was assessed. During the first 2 weeks after surgery, clinical assessments were conducted by veterinarians two to five times daily, with frequency decreasing over time. In weeks 3–5 veterinary assessment occurred twice a week, and for the rest of the study period once a week. The clinical assessment included temperature (by scanning the temperature-sensitive chip), heart rate, respiratory rate, rumen contractions, lameness and two pain scores. The weight of the animals and their body condition score were evaluated several times during the study period.

The hoof block was continuously monitored and adjusted as needed to optimise the sheep’s ambulation, including increasing the weight-bearing surface by adding composite material on the edges of the hoof block (Supplementary Figure 3). A detached hoof block was immediately replaced. Adjustments or replacements were performed under light sedation, as described above.

Pain facial expressions were assessed using the SPFES developed by McLennan et al. (2016) (38). Five facial regions were scored based on the presence of pain-related expressions: (1) tightening of muscles surrounding eyes, closing of the eye, (2) tightening of the muscles of the cheek and in area of the zygomatic arch, (3) ear position, (4) lip and jaw profile, and (5) shape of nostrils and philtrum. The sheep were observed for 2 min from a distance to avoid interference with their potential pain-related facial expression. The SPFES has a maximum total score of ten. Three levels of treatment criteria were defined based on the SPFES (Table 2).

Initially, the clinical severity score (CSS) described by Häger et al. (21) was utilised for the ostectomy group. The CSS includes assessments of food/water consumption, lameness, activity and presence of vocalisation. Based on the CSS and experience from the ostectomy group, the novel a priori weighted composite OOPBS was developed to create a more refined and sensitive pain scoring system, tailored to the behavioural signs relevant in orthopaedic research. By incorporating specific indicators such as posture and general demeanour, the new a priori weighted composite OOPBS aimed to provide a more sensitive and accurate assessment of pain. The OOPBS included assessment of appetite, general demeanour, vocalisation/teeth grinding, general activity, lameness, kneeling, and other adverse behaviours (such as flehmen and wool picking) (Table 3). The sheep were observed for 2 min from a distance. The maximum attainable score was 23, with varying weights assigned to different parameters, reflecting their proposed significance, similar to what has been done in horses (39). Three levels of treatment criteria were defined (Table 2).

If a sheep reached a treatment criterion in either the OOPBS and/or the SPFES, increased monitoring was initiated, and actions such as adaptation of the coaptation and/or additional analgesic treatment were taken. Available options for supplementary analgesia included flunixin meglumine (2.2 mg/kg, IV, Wellicox, Ceva Animal Health A/S, Denmark), butorphanol tartrate (0.2 mg/kg, IM, Dolorex, MSD Animal Health A/S, Denmark), meloxicam (1 mg/kg, SC, Metacam, Boehringer Ingelheim Vetmedica GmbH, Germany), transdermal fentanyl patches (100 μg/h transdermal, Fentanyl Sandoz, Sandoz A/S, Denmark), buprenorphine hydrochloride (0.01 mg/kg, IM, Bupaq Multidose Vet., Salfarm Danmark A/S, Denmark), or regional nerve blocks mid-metacarpal using bupivacaine hydrochloride (0.3 mg/kg BW, SC, Bupivacaine Baxter, Baxter A/S, Denmark) and mepivacaine hydrochloride (1.0 mg/kg BW, SC, Mepidor Vet, Salfarm Danmark A/S, Denmark) (40).

Blood and saliva samples were collected preoperatively (day −1) and on days 2, 4, 6, 7, and 14, and hereafter week 3, 4, 6, 10, and 14. All samples were collected at the same time a day, to avoid diurnal influences.

Prior to saliva collection, the sheep’s mouth was rinsed with 30 mL of tap water using a syringe. Saliva samples were collected using Salivette® tubes (Sarstedt Aps, Bording, Denmark), which consists of an absorbent component and a centrifuge tube designed for non-invasive saliva collection. The original absorbent component in the Salivette® was replaced with sponges (Esponja Marina, La Griega E. Koronis, Madrid, Spain). A sponge was held with forceps inside the sheep’s mouth for 30–60 s. The saliva-saturated sponge was transferred to a Salivette® tube. This procedure was repeated with one or two additional sponges until an adequate volume of saliva was collected. The Salivette® tube was centrifuged within 30 min after collection at 3000 × g for 10 min at 4 °C and saliva transferred to a cryotube (Maxxline Cryo Tubes, non-sterile, Dacos, Hvidovre, Denmark). Collecting saliva was a non-invasive and cheap procedure to which the sheep quickly adapted.

Blood was collected by an indwelling intravenous catheter (Long-Term MILACATH® – Large Animal Kits – Seldinger Technique, Mila International, Inc., Kentucky, USA) or by venipuncture of v. jugularis. The blood samples were collected in BD Vacutainer® 4 mL tubes coated with lithium heparin (Becton Dickinson Vacutainer Systems Europe) for plasma. The samples were centrifuged within 30 min of collection at 1,800 × g for 10 min at 4 °C and plasma transferred to cryotubes (Maxxline Cryo Tubes, Dacos, Hvidovre, Denmark).

All samples were stored at −80 °C until analysis. Cortisol levels in saliva were measured by alphaLISA technology according to López-Arjona et al. (41). Serum amyloid A (SAA) and haptoglobin concentrations were obtained with commercially available kits as described by Franco-Martínez et al. (42) (VETSAA; EIKEN Chemical Co., LTD; Tokyo, Japan) and Schmidt et al. (43) (kit haptoglobin Tridelta phase range, Tridelta Development Ltd., Bray, Ireland), respectively. The serum kits showed an intra- and inter-assay imprecision lower than 15% and were linear (r > 0.91) after serial sample dilutions.

The circumference of the proximal phalanges was measured using a string, which was placed at the height of the distal aspect coronary band of the dewclaws. The length of the string was then measured with a ruler to obtain the circumference. The same investigator undertook the measurements each time. Circumference was measured regularly. The first measurement was taken on the day of surgery and at regular intervals post-surgery, except in osteotomy subgroup A, where the initial casting prohibited assessment of circumference.

Descriptive statistics was performed by applying Excel® (Microsoft Excel software version 2,504, Microsoft Corporation, Redmond, WA, USA) and GraphPad Prism (version 10.4.2 for Windows, GraphPad Software, Boston, Massachusetts USA).¹

Post-surgery, the circumference of the phalanges peaked at week 2 in the ostectomy group, followed by a decrease. From week 8 onwards, no consistent change in circumference was observed in this group (Supplementary Figures 5a,b). No obvious differences in circumference were identified between the 3- and 6- mm ostectomies. There were no consistent changes in circumference in the osteotomy group (Supplementary Figure 5).