Abstract
The surgical repair of acetabular physeal fractures in four kittens using a screw and tension band technique is reported. This was an appropriate method for restoring articular congruency and improving pelvic alignment. All cases had an excellent outcome and full limb use following fracture repair. In kittens younger than 12 weeks, there is a possibility of premature fusion of the acetabular bone resulting in development of a deformed, shallow acetabulum and hip subluxation. However, surgery is still justified when there is pelvic canal narrowing to decrease the risk of future defecatory problems. Early implant removal in such young kittens may decrease the severity of deformity caused by premature physeal closure. In kittens of 16 weeks or older, the prognosis is good for normal acetabular development and implant removal is not necessary.
Physeal fractures are common in immature animals. These injuries heal rapidly but the prognosis for healing without deformity is determined by the remaining growth potential, site and type of fracture, method and time of reduction, and whether the fracture is open or closed (Marretta and Schrader 1983). No specific reports on the management, occurrence or outcome of canine or feline acetabular physeal fractures could be found in the literature. This paper describes the surgical management and clinical and radiographic outcomes of four kittens with acetabular physeal fractures treated with tension band wire techniques.
Materials and Methods
The case records and radiographs of all kittens with fracture through the acetabular physes treated at the Queen's Veterinary School Hospital, Cambridge between January 2001 and December 2004 were retrieved. Information recorded from the files included signalment, cause of trauma, duration of lameness prior to presentation, degree of pre- and postoperative lameness, surgical approach, implants, complications and outcome. Follow-up information was obtained by telephone or e-mail contact with all owners; owners were specifically asked if their cats had ever suffered lameness, stiffness or episodes of constipation.
Radiographs were reviewed for pre-operative fractures and fragment displacement, postoperative fracture alignment, evidence of pelvic canal narrowing, development of degenerative joint disease and implant loosening.
A separate search was made through the radiology archives of the Queen's Veterinary School Hospital, Cambridge for normal pelvic radiographs of kittens aged between 8 and 24 weeks. Radiographs were evaluated to assess the status of the acetabular physes. The physes were considered closed if there was complete cortical continuity and no radiographic evidence of a physis (Houlton and McGlennon 1992). The age of kittens with closed acetabular physes was recorded.
Results
Four kittens with fracture of the acetabulum through the physes were identified. Further details are presented in Table 1.
Case details of the four kittens with acetabular fractures
DSH=domestic shorthaired cat, M=male, E=entire, F=female, R=right, L=left; SHI=Salter and Harris type I, TBW=tension band wire, RTA=road traffic accident, DJD=degenerative joint disease, PMMA=polymethylmethacrylate (PMMA Palacos R-20 with Gentamicin: Schering-Plough Europe), post-op=postoperative, SI=sacroiliac.
Precise size of orthopaedic wire unknown – smaller than 0.8 mm.
All cats presented with unilateral pelvic limb lameness varying from 3/10 to 8/10 in severity. The duration of lameness prior to presentation varied from 1 to 10 days with a mean of 6 days. All kittens had treatment before referral for injuries and shock relating to road traffic accidents or for lameness. Ventrodorsal and laterolateral pelvic radiographs were taken of all four kittens, if not already available from referring veterinarians (Figs 1a, 2a, 3a and 4a). All four kittens had surgical stabilisation of their acetabular physeal fractures performed within 24 h of referral, and a mean of 7 days after trauma (range 2–11 days).
(a) Laterolateral and ventrodorsal radiographs of case 1 with a mildly displaced acetabular fracture after a fall. (b) Immediate postoperative radiographs showing good fracture alignment after stabilisation with two screws, washers and a tension band wire. (c) Laterolateral and ventrodorsal (frog leg) radiographs taken 7 months postoperatively (1 year of age) showing the healed acetabular fracture with no evidence of degenerative joint disease.
(a) Laterolateral and ventrodorsal radiographs of case 2 with an acetabular physeal fracture with displacement of the cranial fragment laterally and a right sacroiliac subluxation. (b) Immediate postoperative radiographs showing good fracture alignment after stabilisation with two screws, crossed K wires, polymethylmethacrylate and a tension band wire. (c) Laterolateral and ventrodorsal (frog leg) radiographs taken 2 months postoperatively showing the healed acetabular fracture.
(a) Laterolateral and ventrodorsal radiographs of case 3 with a markedly displaced acetabular fracture and pubic fractures that occurred after a road traffic accident. (b) Immediate postoperative radiographs showing good acetabular alignment, but there was still marked pelvic canal narrowing, this may have been caused by plastic deformation of the bone. (c) One year postoperatively radiographs show a shallow deformed acetabulum and hip subluxation
(a) Laterolateral and ventrodorsal radiographs of case 4 with a displaced acetabular fracture and pubic fractures that occurred after a road traffic accident. (b) Immediate postoperative radiographs showing good fracture alignment, but there is still some pelvic canal narrowing. (c) Implants were removed after 3 weeks in this kitten; these are the 1-year postoperative radiographs showing a slightly shallow acetabulum but otherwise minimal changes and a wide pelvic canal.
The kittens were premedicated with acepromazine (ACP Injection 2 mg/ml; Novartis Animal Health, UK) and methadone (Physeptone; Martindale Pharmaceuticals). Potentiated amoxycillin (Augmentin; Smith Kline Beecham) was injected pre-operatively. Anaesthesia was induced with diazepam (Diazemuls; Dumex) and an intravenous bolus of thiopentone (Thiovet 2.5 g; Novartis Animal Health, UK) and the kittens were maintained on a gaseous mixture of isoflurane (Isoflurane Vet; Merial Animal Health), nitrous oxide and oxygen. Carprofen (Rimadyl Small Animal Injection; Pfizer) was given pre-operatively. All drugs were given according to the manufacturers' recommendations. The kittens were positioned in lateral recumbency with the affected hemi-pelvis uppermost. The pelvic limb was clipped from hock to midline, prepared for surgery in a sterile fashion and draped in a quadrant fashion, with a sterile foot bag to allow limb manipulation intra-operatively.
In three cats (cases 1, 2 and 3) a dorsal approach was made to the acetabulum via tenotomy of the gluteal muscles (Brown 1953, Piermattei and Johnson 2004a). In one cat (case 4), a dorsal intergluteal approach was made (Piermattei and Johnson 2004b). Blood clot and fibrous tissue at the fracture site were removed. Pointed reduction forceps applied to the ischial tuberosity were used to manipulate the ischial fragment and achieve fracture reduction. When possible, reduction was maintained using bone-holding forceps or by placement of crossed K wires. Two screws were placed in the dorsal acetabulum, one cranial and one caudal to the physeal fracture. A single piece of orthopaedic wire was applied in a figure of eight fashion and tightened by twisting. The surgical repair was supplemented with bone cement in two cases. Full details of the implants placed are presented in Table 1. When possible, the joint capsule was repaired with simple interrupted sutures of size 2M polydioxanone (PDS II; Ethicon UK). Tenotomies were repaired with mattress sutures using 2M polydioxanone. A simple continuous suture pattern was used for subcutaneous closure using 2M poliglecaprone (Monocryl; Ethicon UK), and monofilament polyamide (Ethilon; Ethicon UK) was used for simple interrupted skin sutures. Ventrodorsal and laterolateral pelvic postoperative radiographs were taken (Figs 1b, 2b, 3b and 4b).
Postoperative analgesia was provided by injections of methadone for the initial 12–24 h followed by buprenorphine (Vetergesic; Arnolds Veterinary Products). The cats were discharged home with a 3–5 day course of clavulanate-potentiated amoxycillin (Synulox; Pfizer). Exercise restriction of 4 weeks cage rest followed by 4 weeks room restriction was recommended at discharge. Immediate postoperative lameness varied from 3 to 5/10 in severity. This gradually improved and all cats were sound within 3–8 weeks postoperatively.
Postoperative complications included a mild serous discharge from the operation site for 24 h in one cat (case 2). Two cats (cases 2 and 4) had mild swelling at the surgical site consistent with postoperative fibrosis noted at the first recheck examination. No other complications were noted.
Follow-up radiographs were taken between 2 and 6 weeks postoperatively and again between 2 and 4 weeks later. Final radiographs were taken in three cats (cases 1, 3 and 4) when aged 7 months or older. Mean time of radiographic follow-up was 7 months after surgery. There was no loss of fracture alignment, implant loosening or breakage identified. In two cats (cases 1 and 2), the acetabular fractures healed with no evidence of degenerative joint disease (DJD) (Figs 1c and 2c). In the other two cats (cases 3 and 4), there was some flattening and sclerosis of the acetabulae and mild hip subluxation (Figs 3c and 4c). Changes were more severe in case 3 in comparison with case 4; case 4 was the only kitten that had its implants removed; this was performed at 3 weeks postoperatively (Fig 4).
Mean time of follow-up was 27 months after surgery. All owners described their cats as normal, with no lameness, stiffness, constipation or problems relating to defecation. Without prompting, all four owners commented that their cats were very active.
The radiographs of 26 kittens aged between 8 weeks and 6 months with normal pelves were evaluated to determine the time of closure of the ilioischial and iliopubic physes. The results are shown in Fig 5. The iliopubic physis was closed in all kittens by 20 weeks (5 months) of age and the ilioischial physis was closed in all kittens by 24 weeks (6 months) of age (Fig 5).
Time of closure of the acetabular physes measured from the radiographs of 26 kittens.
Discussion
Despite the young age of the four kittens presented in this paper and the reported success of conservative management of pelvic fractures (Denny and Butterworth 2000), this was not felt to be the best method of management in these cases. There was significant fracture displacement in three cases, two of which had narrowing of the pelvic canal (Figs 3 and 4) and one had medial displacement of the caudal fragment in combination with a sacroiliac luxation (Fig 2). The authors aimed to restore acetabular articular congruency and pelvic alignment and thus minimise potential complications that can occur with malunion such as constipation and obstipation (Ward 1967, Leighton 1969, Evans 1980, Matthiesen et al 1991, Schrader 1992, McKee and Wong 1994, Ferguson 1996). In case 1, with a minimally displaced fracture (Fig 1a), surgical stabilisation of the articular fracture was chosen as the development of DJD may be reduced by surgical fixation of acetabular fractures (Hulse and Root 1980); the fracture in this cat healed with no radiographic evidence of DJD.
In cases 3 and 4, a degree of pelvic canal narrowing remained after surgery (Figs 3 and 4), but to a lesser extent than was present pre-operatively. This occurred despite the appearance of accurate realignment of the articular surfaces both at surgery and radiographically. These two kittens were estimated to be 8–9 weeks old and both had suffered road traffic accidents. Immature animals can suffer plastic deformation of bones after trauma (Mabrey and Fitch 1989). The plastic deformation is largely due to paediatric bone having a lower mineral content than adult bone (Mabrey and Fitch 1989). Although plasticity allows children's long bones to absorb more energy before fracturing, a significant deformity may persist after injury (Mabrey and Fitch 1989). Both cases also had pubic fractures; this combined with the flexibility at the sacroiliac joint, may have contributed to medial displacement of the hemi-pelvis. Despite the retention of some pelvic canal narrowing the authors believe that improvement in pelvic canal diameter and articular congruency justified surgical stabilisation in these two kittens.
In one kitten (4) the hemispherical articular cartilage lining the acetabulum was found to be intact at surgery; it had remained attached to the caudal fragment but was detached from the cranial fragment. With fracture displacement this cartilage was interposed between the two main bone fragments. This is known as a Salter and Harris type 7 separation, these occur between cartilage and bone at any chondro-osseous epiphyseal interface and they are impossible to diagnose radiographically (Ogden et al 1993). This is an unusual injury that the authors have not seen previously. Acetabular depth increases during development as a result of interstitial growth in the hemispherical cartilage and appositional growth of the cartilage periphery (Heeg et al 1988). Detachment at the chondro-osseous physis is likely to have caused interference of growth at this interface. This may have contributed to the resultant mild decrease in acetabular depth in this kitten (Fig 4c).
The dorsal acetabulum is a tension surface and, therefore, suitable for tension band fixation but reportedly only for interlocking, stable two-piece fractures (Brinker et al 1997a). The acetabular fractures in cases 1, 3, 4 were all simple two-piece fractures and all healed with no loss of alignment or implant loosening indicating that fixation using two screws and a tension band were strong enough implant combinations for these three kittens. In one of the two older kittens, that had a displaced and mildly comminuted fracture, crossed K wires were used in addition to the screws and TBW (Fig 2). In a review of acetabular fracture repair in dogs, Wheaton et al (1973) stated that if a figure of eight wire and screws were used without additional crossed pins, shear motion could occur with loss of fracture alignment. An in vitro study using dog pelves described the additional use of one or more ‘K’ wires as an aid to maintaining reduction while the screws and wire were placed, and as a method of improving the overall stiffness and yield load of the repair (Beaver et al 2000). However, placement of the wires can be technically demanding (Lewis et al 1997). In young kittens, with faster healing rates, the addition of crossed K wires is probably not essential for repair stiffness and yield strength, but they may aid maintenance of reduction particularly in the older kitten, or if comminution is present, where additional stability may be beneficial.
The addition of PMMA to screws and wire in the management of acetabular fractures has been found to improve the stiffness, yield load and maximum load sustained by the repair (Lanz et al 1999) and it protects the wires from the cyclic stresses of weight bearing (Renegar and Griffiths 1977). The negative aspects of using PMMA are cost, closure can be more difficult due to its bulk, increased difficulty of implant removal and interference with radiographic visualisation of the fracture (Fig 2b and c). Smooth stainless steel washers were used in two kittens (cases 1 and 4) to prevent the figure of eight tension band from slipping off the screw heads and one rationale for PMMA use in the other two kittens (cases 2 and 3), where washers were not used, was to reduce the chance of wire displacement and screw loosening. PMMA was not used in case 4 as it was intended to remove the implants.
One of the principles of articular fracture repair is perfect anatomic reduction. Conventional plate and screw fixation of acetabular fractures relies on having a perfectly contoured plate so that fracture alignment is maintained when the plate screws are tightened. Compared to 2.0 mm veterinary acetabular plates, the screw, wire, pin and PMMA (SWP) technique has been shown to allow superior maintenance of accurate alignment during fixation and superior strength and stiffness of repair (Stubbs et al 1998). The two screws should be placed so that the axis of tension created between the implants is perpendicular to the fracture line, allowing the tension band wire to create optimum compression without translation (Lewis et al 1997) and thus prevent loss of alignment during wire tightening. Given the small size of the patients in this series it was felt that the SWP combination was preferable to plate and screw fixation as it required a smaller surgical approach, therefore causing less bone exposure, and achieved superior alignment.
Each os coxae is composed of three principal parts; the ilium, ischium and pubis. The acetabular bone is a fourth and significantly smaller part (Fig 6). These parts are distinct in the kitten and are separated by cartilage. This becomes ossified in the adult to form a continuum of bone. The walls of the acetabulum comprise the ilium, ischium and acetabular bones; the pubis bone does not form part of the acetabular wall; it is separated from it by the acetabular bone (Crouch and Lackey 1969). The only report on radiographic closure times for the acetabular physes are for the ilioischial physis in the cat, with a reported closure time of between 20 and 25 weeks (Houlton and McGlennon 1992) which is similar to the findings of this study. As the ischioacetabular, ilioacetabular and puboacetabular physes are not visible on radiographs because of superimposition of the femoral head, knowledge of ilioischial and iliopubic physeal closure times does not allow accurate prediction of when the other acetabular physes fuse.
The immature cat pelvis and the acetabular physes.
In humans, the acetabular triradiate cartilage is the composite growth plate of the iliac, ischial and pubic bones (Heeg et al 1988). It is equivalent to the acetabular bone in the dog and cat. Premature fusion of the triradiate cartilage can occur after pelvic trauma in children resulting in acetabular deformity and subluxation of the hip (Heeg et al 1988). The age at the time of injury is important; severe deformity has not been described with injury in children more than 11 years of age (Bucholz et al 1989). The two younger kittens in this study, estimated to be less than 12 weeks old, developed varying degrees of acetabular deformity with decreased acetabular depth and sclerosis and increased hip subluxation (Figs 3c and 4c). This may have been due to premature fusion of the acetabular bone. In the two older kittens (16 weeks), the acetabulum continued to develop normally with no evidence of hip dysplasia. Thus the prognosis for continued normal acetabular development in kittens seems to be age related; those suffering trauma at less than 12 weeks of age had an increased likelihood of premature physeal closure and those of 16 weeks or older had a good prognosis for continued normal development of the acetabulum.
The fractures in this series were both articular and physeal. The ideal stabilisation devices for physeal fractures are smooth implants that cross perpendicular to the physis (Brinker et al 1997b). However, articular fractures should be stabilised with devices that compress fracture fragments thus preventing micro-motion and promoting direct bone union (Mitchell and Shepard 1980). Tension band fixation on the dorsal surface of the acetabulum provides ideal interfragmentary compression for an articular fracture, but compression of the physis is not recommended as it may increase the chance of premature closure; if such a form of fixation is unavoidable in order to achieve a secure fixation then it should be used, but early removal of implants is recommended (Brinker et al 1997b). Thus in one case (case 4), it was elected to remove the implants 3 weeks after fracture repair in order to allow potential for further growth; the outcome was good (Fig 4c) as the kitten only developed mild acetabular abnormalities. Although based on the outcome of one case, the authors recommend considering implant or tension band removal in the very immature kitten.
Footnotes
Acknowledgements
Acknowledgements are due to the owners of the four kittens Lorna Kirk, Professor and Mrs Bateson, Olwen Lloyd, and John Burford; to Gareth Arthurs for reading the manuscript and to the surgery residents, radiologists, nurses and radiographers at Cambridge University who assisted with management of these kittens.