Fractures of the humerus are relatively uncommon in companion animals, accounting for approximately 10% of all appendicular skeletal fractures in dogs and cats (Pearson et al., 2016). These injuries most frequently result from high-velocity trauma such as motor vehicle accidents, falls from height and blunt force, though bite wounds may also be implicated (Fossum, 2019). Humeral fractures are anatomically classified as proximal, mid-diaphyseal, or distal, with distal fractures-particularly those involving the condyle-being the most frequently encountered and often the most comminuted (Bardet et al., 1983; Cardona et al., 2015).
       
The humerus presents distinctive surgical challenges attributable to its S-shaped configuration, eccentric loading mechanics and intimate association with critical neurovascular structures. The radial nerve traverses the musculospiral groove of the distal humerus from medial to lateral, while the median and ulnar nerves course along the medial aspect. Consequently, fractures of the diaphysis and distal humerus carry a significant risk of partial or complete radial nerve injury (neuropraxia), which may result in temporary or permanent limb dysfunction (Langley-Hobbs and Straw, 2005).
       
Conservative management of humeral fractures using external coaptation is generally unsatisfactory and associated with complications including limb shortening, malunion, nonunion and joint ankylosis (Simpson, 2004). Consequently, surgical internal fixation is the preferred treatment approach. Reported fixation techniques include intramedullary (IM) pinning with or without cerclage wire, external skeletal fixation, plate-rod constructs, interlocking intramedullary nails and single or bilateral bone plating with or without lag screw fixation (Pearson et al., 2016). The craniolateral aspect of the humerus is recognized as the tension surface and is the preferred site for bone plate application (Harari et al., 1986).
       
While plate-rod fixation techniques are established in the veterinary orthopaedic literature, comparative clinical data from teaching hospital settings in India-reflecting locally available implants, diverse fracture morphologies and varied patient demographics-remain limited. The present study was therefore undertaken to document and evaluate the clinical and radiographic outcomes of multiple surgical fixation methods applied to naturally occurring humeral fractures in dogs, with the aim of providing regionally relevant guidance for implant selection.

Study design and case selection
 
This prospective observational clinical study was conducted between January 2021 and December 2025 at the Department of Veterinary Surgery and Radiology, P.V. Narsimha Rao Telangana Veterinary University (PVNR TVU), Rajendranagar, Hyderabad. Dogs of any age, sex, or breed presenting with lameness of the forelimb and radiographic evidence of humeral fracture were eligible for inclusion. Exclusion criteria comprised open fractures with extensive soft tissue loss, pathological fractures, concurrent spinal injuries precluding ambulation assessment and cases lost to follow-up before the 90-day evaluation.
 
Clinical and radiographic assessment
 
A standardized orthopaedic and neurological examination was performed at initial presentation. Parameters assessed included degree of lameness, pain on palpation, crepitus, angular deformity, limb swelling, proprioception, presence of elbow drop and dorsal paw placement. Concurrent thoracic, abdominal, or cranial injuries were excluded by thoracic auscultation and physical examination prior to anaesthesia.
       
Preoperative radiographs were obtained using a computed radiography system in standard mediolateral (ML) and craniocaudal (CrCd) projections.  Representative preoperative radiographs of dogs with humeral fractures are shown in Fig 1. The ML view was used to estimate medullary canal diameter and implant length, while the CrCd view determined screw length and condylar dimensions. Fractures were classified according to the AO/ASIF-Vet system as adapted for long bones in dogs and cats by Unger et al. (1990), categorized by anatomical location (proximal, mid-diaphyseal, or distal diaphyseal, including condylar, transcondylar, supracondylar and metaphyseal subtypes). The fracture classifications observed in the present study are summarized in Table 1.




 
Perioperative and anaesthetic protocol
 
Fracture stabilization pending surgery was achieved with a modified Robert Jones bandage. Preoperative analgesia was provided with meloxicam (Inj. Melonex®, Intas Pharmaceuticals, Ahmedabad) at 0.2 mg/kg subcutaneously once daily. Owners were instructed to withhold food for 12 hours and water for 6 hours prior to induction.
       
Intravenous access was established via the cephalic or saphenous vein. Fluid therapy was administered with 0.9% normal saline at 10 mL/kg/hour intraoperatively until extubation. Preanaesthetic medication comprised xylazine (1 mg/kg), ketamine hydrochloride (10 mg/kg) and butorphanol (0.2 mg/kg) administered intramuscularly, in combination with atropine sulphate (0.04 mg/kg; Inj. Atropine, Ek-Tek Pharma, India). Perioperative antibiotic prophylaxis was provided with cefotaxime (20 mg/kg IV; Inj. Taxim®, Alkem Laboratories, India) administered 30 minutes before skin incision. Anaesthesia was induced and maintained with propofol 1% (Inj. Neorof, Neon Laboratories, India) at 4 mg/kg IV to effect. The surgical site was prepared with chlorhexidine gluconate-cetrimide-isopropyl alcohol solution (Aceptik®, Raman and Weil, India) and painted with povidone-iodine (Bectodine®, Glide Chem, India). Sterile field draping was applied using towels, towel clamps and a sterile disposable glove over the distal limb.
 
Surgical technique
 
All fractures were approached via the standard craniolateral approach, which provides optimal access to the diaphysis and condyles while permitting identification and preservation of the radial, median and ulnar nerves. The skin incision extended between the greater tubercle proximally and the lateral epicondyle distally. Subcutaneous tissue, brachial fascia and the brachiocephalicus muscle were incised and separated. The axillobrachial and omobrachial veins were ligated. The radial nerve was identified in the musculospiral groove and carefully isolated and protected throughout the procedure. The brachialis and triceps muscles were retracted using Gelpi or Hohmann retractors to expose the proximal and central humeral shaft. A periosteal elevator was used to mobilize fracture fragments.
       
Implant selection was based on fracture morphology, patient size and fragment geometry, as per the principles described by DeCamp et al. (2016). Plate-rod constructs were applied using locking compression plates (LCPs) on the lateral humeral surface in combination with retrograde Steinmann intramedullary pins occupying approximately 40% of the medullary canal diameter at the isthmus (n = 11). LCP or dynamic compression plate (DCP) fixation alone was applied in dogs with short oblique and transverse mid-diaphyseal fractures (n = 4). Steinmann IM pins occupying 70-80% of the medullary canal were used in retrograde fashion for supracondylar fractures in small-breed dogs (n = 2) and one dog received IM pinning supplemented with accessory cerclage wire. For transcondylar and metaphyseal fractures, cross-pinning of the medial and lateral condyles was combined with a transcondylar lag screw and an LCP on the lateral surface. One dog with a proximal diaphyseal fracture was managed with an intramedullary interlocking nail (IILN) inserted in a normograde direction using an aiming device.
       
Wound closure was performed in layers: the superficial pectoral and brachiocephalicus muscles were apposed to the fascia with 2/0 polyglactin 910 (Vicryl®, Ethicon) in simple interrupted sutures. Brachial fascia and subcutaneous tissue were closed in simple continuous fashion. Fascia of the flexor carpi ulnaris, ulnaris lateralis and anconeus muscles were opposed with simple interrupted 2/0 polyglactin 910. Skin closure was achieved with 2/0 polyamide sutures in a cross-mattress pattern.
 
Postoperative management and outcome assessment
 
Immediate postoperative radiographs (ML and CrCd projections) were obtained to confirm implant placement and fracture reduction. A supportive bandage was maintained during the early postoperative period. Activity restriction was enforced throughout the healing phase. Postoperative analgesic and anti-inflammatory therapy were continued as clinically indicated.
       
Functional recovery was assessed using the five-point lameness grading system of Vasseur et al., (1995) at postoperative days 1, 15, 30, 60 and 90. Radiographic follow-up was conducted at days 30, 60 and 90 to evaluate implant position, callus formation and progression of bone healing. Radiographic union was defined as the presence of bridging callus across the fracture site, progressive loss of the fracture line, restoration of cortical continuity and maintenance of implant stability. Complications including implant failure, pin migration, infection and wound dehiscence were recorded throughout the follow-up period.
       
Due to the limited sample size and heterogeneity of fracture configurations and fixation methods, inferential statistical comparisons were not performed. Descriptive statistics are presented as mean±standard error of the mean (SEM) for continuous variables. Clinical and radiographic outcomes were described qualitatively and compared with findings reported in the published veterinary orthopaedic literature.

Patient demographics
 
Nineteen dogs were enrolled, comprising 11 mongrels and one each of German Shepherd, Pug, French Bulldog, Pomeranian (2 cases), Husky, Golden Retriever and Beagle. This breed distribution reflects the local hospital population and is consistent with reports from comparable Indian institutions (Singh et al., 2015). Age ranged from 2 to 36 months (mean±SEM: 7.97±2.13 months) and body weight ranged from 2.5 to 40 kg (11.23±2.20 kg). Male dogs (n = 12) outnumbered females (n = 7), a distribution attributable to their more active and exploratory behaviour, consistent with findings of Priyanka (2018).
       
The predominant aetiologies were automobile trauma (n = 8; 42%), falls from height (n = 7; 37%), dog bites (n = 3; 16%) and unknown cause (n = 1; 5%). These findings are consistent with those of Bardet et al., (1983) and Kumar et al., (2026), who similarly identified road traffic accidents as the leading cause of humeral fractures. Eleven fractures involved the left forelimb and eight the right. Pre-operative evaluation of lameness was done as per Vasseur et al., (1995).
       
Fracture morphology was heterogeneous. Details of fracture classification, fixation method and complications, grouped by fixation method, are presented in Table 1. The most frequently encountered fracture pattern was mid-diaphyseal oblique (n = 8), followed by mid-diaphyseal transverse (n = 3), transcondylar metaphyseal (n = 2), distal supracondylar (n = 2) and one each of proximal diaphyseal, mid-diaphyseal comminuted, distal diaphyseal transverse and mid-diaphyseal spiral fractures. This distribution is in keeping with the observation that the middle and distal thirds of the humerus are most frequently fractured (Langley-Hobbs, 2018).
       
Post-operative functional evaluation was performed using lameness grading, proprioceptive assessment and weight-bearing observations at predetermined follow-up intervals.
       
Functional recovery improved progressively in all dogs throughout the follow-up period. Pre-operative and post-operative lameness score trends, grouped by fixation method, are summarized in Table 1. Mean lameness scores decreased from 4.0±0.15 on postoperative day 1 to 2.31±0.10, 2.10±0.09, 1.21±0.09 and 1.0±0.0 on days 15, 30, 60 and 90, respectively. By day 90, all dogs achieved Grade I lameness, indicating normal weight bearing.
       
All dogs achieved functional weight bearing before the 15^th postoperative day except those with transcondylar metaphyseal fractures, which required approximately 60 days. Serial radiographic evaluations demonstrated progressive fracture healing with maintenance of implant stability in most cases. Representative postoperative radiographs demonstrating fracture reduction and sequential bone healing are presented in Fig 2 and Fig 3. Implant-related complications were observed in three dogs, in which intramedullary pin loosening necessitated premature implant removal; however, satisfactory fracture healing was subsequently achieved. No postoperative radial nerve deficits, implant breakage, wound dehiscence, or surgical site infections were observed during the follow-up period.




 
Comparative assessment of fixation methods
 
Based on the clinical and radiographic outcomes observed in this series, plate-rod constructs were successfully applied across a broad range of fracture configurations and yielded favourable results. The combination of an LCP and IM pin resists all principal fracture forces-bending, torsion and axial compression-by sharing load between the plate and the pin, reducing stress at the plate-bone interface and lowering the risk of plate fatigue failure. This biomechanical rationale has been well established in previous studies (Ayyappan et al., 2011; Pearson et al., 2016).
       
LCP alone provided satisfactory fixation in mid-diaphyseal short oblique and transverse fractures, consistent with the recommendation that rigid plate fixation is appropriate when adequate cortical purchase can be achieved without rotational instability. IM pinning alone was applied in small-breed dogs with supracondylar fractures where plate application was technically precluded by patient size; however, the limited resistance of IM pins to rotational forces necessitates cautious patient selection. IILN fixation in the proximal diaphyseal fracture case yielded excellent stability and early weight-bearing (Fig 4), supporting the use of this technique for proximal and long mid-diaphyseal fractures, as described by Pearson et al., (2016) and consistent with the management principles for proximal humeral fractures outlined by Turner (2005).


       
Recent regional reports lend further support to these findings. In a large retrospective series of dogs presenting to a South Indian teaching hospital, Kumar et al., (2026) documented the relative incidence and anatomical distribution of long-bone fractures, including the humerus, corroborating the predominance of road-traffic trauma and the comparatively lower proportion of humeral fractures relative to the tibia, radius-ulna and femur observed in Indian clinical settings. Pravalika et al., (2023) reported favourable clinical and radiographic outcomes with pre-contoured plate fixation for periarticular long-bone fractures in dogs, reinforcing the rationale for anatomically contoured implants at metaphyseal and condylar sites such as those encountered in the present series. Lee et al., (2025) described successful long-bone fracture reconstruction using autologous bone grafting in a case complicated by soft-tissue loss, underscoring the value of individualized implant and biological augmentation strategies in complex fracture presentations. Jain et al., (2023) further demonstrated that adjunctive biological augmentation, such as autologous bone marrow concentrate combined with a hydroxyapatite-collagen scaffold, can enhance radiographic healing in canine long-bone fractures, suggesting a potential avenue for improving outcomes in cases with delayed union or comminution.
       
It should be noted that the present study is descriptive and does not include control groups or formal statistical comparison between fixation methods. Plate-rod constructs were the most frequently employed fixation method and yielded consistent clinical and radiographic outcomes across multiple fracture configurations. Although direct comparison among fixation techniques was not possible because of case heterogeneity and limited sample size, the observed outcomes support the usefulness of plate-rod fixation in canine humeral fracture management.
 
Study limitations
 
The principal limitations of this study include the small sample size (n = 19), heterogeneity of fracture configurations and fixation methods and the absence of a formal control group. The lack of blinded radiographic assessment and the use of a descriptive lameness scoring system rather than objective gait analysis are also acknowledged. These factors limit the generalizability of the findings and preclude definitive conclusions regarding the comparative performance of individual fixation techniques.

The present clinical study demonstrated that surgical management of humeral fractures in dogs using fixation methods selected according to fracture morphology and AO/ASIF-Vet classification resulted in satisfactory functional recovery and radiographic fracture healing. Plate-rod constructs, bone plating, intramedullary pinning and intramedullary interlocking nailing were successfully applied in appropriate fracture configurations, with favourable clinical outcomes observed throughout the follow-up period. Owing to the limited sample size and heterogeneity of fracture types and fixation methods, definitive comparisons among treatment techniques could not be made. Further prospective studies involving larger and more homogeneous populations with standardized outcome measures are warranted to establish evidence-based guidelines for implant selection in canine humeral fracture repair.

The author declares that there is no conflict of interest.