The general anesthesia was obtained using inj. atropine sulphate @ 0.02 mg/kg I/M as pre-anaesthetic, followed by inj. diazepam hydrochloride @ 1 mg/kg I/V and induction of general anesthesia with inj. propofol 4 mg/kg I/V. After induction of the adequate level of anaesthesia in dog, endotracheal intubation was performed using a laryngoscope and a suitable sized cuffed endotracheal tube no. 6 to 7.5. Adequate anaesthetic depth was maintained using inhalation anesthetic agent isoflurane at minimum alveolar concentration (MAC) of 1.0-2.5%. The flow rate of the oxygen was maintained approximately @ 10 ml/ kg b.w per minute. The semi closed or closed breathing circuit was utilized during surgical procedure as per the requirement.

A standard set of orthopaedic instruments was used for internal fixation with a reconstruction or locking compression plate.

The long bones radius-ulna, tibia and femur were exposed for internal fixation of fracture as described by Piermattie et al. (2006).

In group T1, the diaphyseal fractures of the long bones were immobilized by an internal fixation technique using reconstruction plate or locking compression plate. The reconstruction plate was used for internal fixation of the fracture in animals having body weight up to 12 kg. The LCP was used for internal fixation of the fracture in animals having a higher body weight up to 30 kg (Fig. 1a).

The fracture ends were retrieved from incision site and were reduced manually to their anatomical position. After reduction, LCP of suitable length was placed over bone surface and held with surface using bone holding forceps. The position of the plate and its alignment was checked and optimized. The LCP sleeve was fixed in hole to be drilled. Then, the holes were drilled in bone from the plate holes using universal drill. The depth of the hole was measured with orthopaedic depth gauze. This depth was used to decide the length of screws. An appropriate screw was tightened in the hole using an orthopaedic screw driver. The same procedure was followed for each hole in proximal and distal fragment of the bone. The orthopeadic screws had purchase in minimum of six cortices of bone in one segment for rigid fixation. After fixation of the fracture, using a reconstruction plate or locking compression plate, a composite mesh was used in periphery of the bone at fracture site. Both the ends of the mesh were sutured with simple continuous sutures using polyglactin 910 No. 0 or 2-0, so as to fit the mesh snugly around the periphery of bone at fracture site (Fig 1a and 1b).

In group T2, the diaphyseal fractures were immobilized using an internal fixation technique using a suitable plate. B-TCP granules were incorporated at fracture site in a composite mesh cylinder after achieving haemostasis. One or more pack of 0.5 gm P-TCP was used to make up the bone loss at fracture site.

In group T3, the stable internal fixation of fracture fragments was achieved and β-TCP was filled in a cylinder of composite mesh. One milliliter of PRP was administered locally at fracture site just before the closing the first suture line.

The rectal temperature of the dogs ranged from 101 to 103°F at time of presentation of animals with fracture. The mean rectal temperature was non-significantly (p>0.05) higher at 0 day in three groups. Desouza (2012) has also reported a higher rectal temperature in dogs at day 0 pre-operatively. The increase in rectal temperature may be due to release of mediators of inflammation viz, prostaglandins, bradykinins and interleukins following trauma and fracture.

A smooth recovery from general anaesthesia was observed after 10 to 30 minutes of turning off the isoflurane vaporizer. Raghunath et al. (2012) also maintained adequate general anaesthesia using isoflurane @ 1.0 to 2.0 MAC. In present study MAC for isoflurane ranged from 1.0 to 2.5. This could be attributed to different anaesthetic protocols used for induction of general anaesthesia.

At day 90^th, a radiograph in both the views was not able to delineate the granules from osseous tissue. The resemblance in radio-opacity of β-TCP and osseous tissue may be due to its resorption and osteoconductive properties. Jang et al. (2008) has also observed that an ideal bone graft should resorb fully and also provide three dimensional matrix to support bone ingrowth and on growth during resorption. On the same lines, Yamasaki et al. (2009) observed that the radio-opacity of the TCP decreased in a time dependent manner and was available on X-ray films at 12 weeks after implantation, but the margins of the implants were not clear indicating the resorption of β-TCP.

At 90^th day, the mean radiographic score was highest (4.50 ± 0.22) for group T3 and it was followed by the score of 4.33 ± 0.88 in group T2, showing osseous callus in evolution with presence of periosteal proliferation or exuberant osseous callus in evolution. The mean radiographic score was minimum (2.86 ± 0.88) for group T1 at 90^th day showing organized periosteal proliferation.

The increase in mean radiographic score may be attributed to the progress in fracture healing at different intervals. In group T2, P-TCP implanted at fracture site probably acted as an osteoconductive material and leaded to increase in bone formation, hence, this bone substitute contributed to increase in radiographic score. The observation is well corroborated with the findings of Neamat et al. (2009) who stated that after resorption of Cerasorb M (P-TCP), the bone was regenerated completely.

In group T3, the mean radiographic score was noticed highest at 90^th day expressing that fracture healing was better than rest of the two groups. This could be attributed to the role of PRP as an osteoinductive material in fracture repair as Fortier et al. (2011) has described the PRP as a milieu of growth factors, cytokines, perhaps some RBC’s and all the proteins. A PRP clot serves as a fibrin matrix which acts as a scaffold for tissue repair and reservoir for retention and slow release of growth factors.

The mean inflammation score at fracture site was recorded to be significantly (p<0.05) higher at 0 day in all the three groups. There was a significant reduction in mean inflammation score at 7^th day in groups T2 and T3 and at 30^th day than 0 day in all the groups. This decline was indicative of decrease in inflammation at 30^th day in affected limb. Finally, at 90^th day, the mean inflammation score reduced to almost normal values, suggesting cessation of inflammatory signs. Desouza (2012) observed clinical signs of inflammation at fracture site in affected limb at time of presentation, however; no quantification of these signs was carried out and the inflammation subsided in most of the cases after 30^th day.

The increment in this parameter was observed post operatively at 0 day itself, showing that animals were touching the toe (score 1) or touching the paw (score 2) in all the groups. The weight bearing (while standing) improved early because most of the animals started to touch the toe or paw just after some hours of the internal fixation. Similarly, Rajhans (2013) reported that weight bearing on affected limb after internal fixation using a reconstruction plate was observed at 0 day in few cases; however others did not show weight bearing on affected limb.

The mean weight bearing score (while standing) increased gradually from 0 to 30^th day indicating that all animals were touching the toe or paw of the affected limb. It was observed better in group T3 followed by group T2 and group T1 at 30^th day. This improvement in score may be due to the fact that most of the animals of group T3 showed early fracture healing as also supported by the radiographic scores. Kushwaha et al. (2012) reported full weight bearing at 18^th day on affected limb after internal fixation of a radius-ulna in a non-descript dog.

In group T3, the mean weight bearing score was 2.00 ± 0.00 at 90^th day as all the animals were showing full weight bearing on affected limb. This improvement in weight bearing could be attributed to better fracture healing probably attributed to local application of PRP as also suggested by the inflammation scores and radiographic scores in the present study.

Note: ^abvalues within groups with different superscripts differ significantly (p<0.05).

The mean weight bearing score was found to be higher in groups T2 and T3 than group T1 at 14^th, 30^th, 60^th and 90^th day. The mean weight bearing score was recorded to be highest in group T3 at 30^th, 60^th and 90^th day post operatively.

A response having fear was markedly witnessed while assessing the animals for weight bearing (while walking). The improvement in this score must have been depended upon the weight bearing score (while standing) at different time intervals as the animal showing an early weight bearing (while standing) can be predicted to express an early weight bearing (while walking) at further intervals.

The mean weight bearing score was also higher in group T3 than in group T2 at 30^th, 60^th and 90^th post operative day showing that restoration of the function was better and earlier in groups T2 and T3 than in group T1. This finding is also based on the fact that animal showing an early weight bearing (while standing) will witness an early return of limb function with full range of motion. The same has been endorsed by Cruess and Dumont (1985) who stated that bone healing can be promoted by some factors like growth hormone, thyroid hormone, calcitonin, insulin, vitamin A and D, anabolic steroids, chondroitin sulphate, hyaluronidase, anticoagulants, electric currents, oxygen and physical exercise. Early weight bearing (while standing and while walking) in group T3 in comparison to remaining two groups, might have contributed in form of physical exercise to better weight bearing score (while running) from 60^th day onwards. Further, the assessment of the weight bearing (while running) was found difficult in a clinical environment.

The complications observed during the study were mostly related with implant failure. The failure of a locking compression plate might be due to the reason that the dog was allowed to run freely within a month. The implant failure observed in radius-ulna of a Pomeranian (10 kg) might be due to sub-optimal operative procedure. The high complication rate in radius-ulna fractures may be attributed to the fact that healing of radius-ulna fractures is potentially slowed by a limited blood supply, limited soft tissue around the bone (Boudrieau, 1993). Further, the bending of the plate may be attributed to the biomechanical forces applied during post operative period which vary according to temperament of animal and compliance from owner.

The authors are thankful to the Dean, College of Veterinary Science & A.H. Jabalpur for providing required facilities for the research work.