Are Old Dogs Who Undergo Total Hip Replacement More Predisposed to Perioperative Femoral Fractures Than Young Dogs ? A Knowledge Summary

Clinical bottom line There is currently insufficient evidence that old dogs undergoing total hip replacement are more predisposed to perioperative femoral fractures in comparison to young dogs.

Age at THR Dogs that had femur fractures were significantly older (7.4 years) at THR than dogs that did not sustain femur fractures (4.9 years) (p=0.0063).

Predisposing factors
Osteopathy (n=5) iatrogenic fissures created during reaming (n=9) and previous hip surgery.Fracture characteristics Fracture occurrence 22 after original THR, 1 after revision, 1 after explanation:  traumatic events in 17 dogs  osteopathy present at THR in 5 dogs  cortical thinning secondary to aseptic loosening in 3 dogs

Fracture treatment:
 plate and screw fixation (10 with and 7 without cerclage wires) resulted in the most favourable outcome -healing occurred in 6-10 weeks  full cerclages wires in 3 dogs  strict confinement in 3 dogs  euthanasia in 1 dog

Outcome:
 22 fractures healed and there were no non-unions  two fractures that did not heal: 1 dog died 16 days after surgery of an unrelated illness and 1 dog was euthanised without treatment Old dogs with osteopathies, dogs that have had previous hip surgery, and dogs that have intraoperative fissures should be recognised as potentially being at greater risk of femoral fracture.

Limitations:
 retrospective study  different surgeons, different hospitals  surgeon experience was not evaluated as a risk factor  different follow-up  10 dogs (45.5%) with fractures were followed until death  client interview

Inclusion criteria
Cases with preoperative, immediate postoperative and initial followup radiographs.

Exclusion criteria
Dogs without complete surgery reports, dogs in the non-fracture group without documentation of absence of femoral fracture on recheck radiographs at least 4 weeks postoperatively.

Statistical analysis:
 preliminary univariate tests were performed and any factors with p>.30 were excluded for further consideration  continuous variables: non parametric Wilcoxon's rank sum test  binary variables: Fisher's exact test  factors included for multicollinearity were entered in to the logistic equation  previously deleted factors were singly added to the final

Dog factors:
 Age was positively associated with femoral fracture.Mean dog age was 7.3±0.69years for fracture group and 4.77±0.37years for non-fracture group (p=0.022). CFI was negatively associated with fracture.Mean CFI was 1.80±0.09for the fracture group and 1.98±0.04for the nonfracture group (p=0.045). Body weight, intraoperative fissure, cerclage use, implant size, position and canal fill did not influence the occurrence of femoral fracture.

Intraoperative fissures
Reported in 3 cases.None of these dogs had femoral fractures.All fissures were addressed with cerclage wires.

Radiographic evaluation
None of the measures of implant positioning or canal fill were associated with risk of femoral fractures.

Limitations:
 retrospective study

Intervention details: Inclusion criteria
At least 8 weeks of documented postoperative radiographic and orthopaedic evaluations.

Exclusion criteria
THR performed as the second procedure in dogs operated bilaterally, previous ipsilateral coxofemoral surgery, cases without sufficient client communication.

Medical records of dogs
Sex, breed, age, body weight, body condition score, side of arthroplasty, date of surgery, history of previous contralateral coxofemoral surgery, urinalysis results, intraoperative surgical site culture results, lameness score at presentation (0=no lameness, 1=slight lameness, 2=obvious weight-bearing lameness, 3=intermittent non-weight bearing lameness, 4=continuous nonweight-bearing lameness) size of prostheses implanted, duration of surgery.

Complications were separated into:
 intraoperative (IOC)  short-term (STC)  long-term (LTC) Bivariate and multivariate statistical analysis was used to compare complications.
Procedures were performed by 10 different surgeons.

Study design: Retrospective cohort study.
Outcome studied: To determine the prevalence of complications and identify prognostic indicators of success or failure for the Zurich cementless THR.

Increased body weight
Prior femoral head and neck ostectomy (FHO) or cemented-THR in the contralateral joint was identified as a negative prognostic indicator for successful outcome (p<0.05).

Risk and protective factors:
 after primary cTHR dogs were more likely to have a revision surgery if the femoral implant was eccentrically placed (p=0.01) presence of radiolucent lines at the femoral cement-bone interface in the long-term period was positively associated with revision surgery (p=0.02)male dogs were more likely to have revision surgery on the primary cTHR (p=0.05)

Limitations:
 retrospective (medical records incomplete, limited followup)  different surgeons with different experience  surgeon's experience not evaluated as risk factor  complications that were treated elsewhere or left untreated would not have been detected  variable radiographic technique and patient positioning (may have affected implant assessment)  some radiographic films with inadequate quality for evaluation

Clinical outcome:
 of 65 ZCTHR, 60 were considered to have an excellent outcome, 3 a good outcome, and 2 as failed  97% good or excellent outcome at an average of 2 years post-operatively (similar to previous reported rates of 91-96%)  17% needed one or more revision

Radiographic findings:
Findings compatible with bone ingrowth fixation were observed for all acetabular and femoral implants.Focal radiolucent zones were observed in the acetabular component of 23 cases and none of the THR had a complete radiolucent zone around the cup or the stem.
Femoral fracture (n=1) well-reported complication after THR in dogs, and appears to be more common in old animals because of nonuse of the leg or other pathologic conditions.This single femoral fracture compares similarly with fractures occurring with cemented THR systems.Increased femoral cortical thickening was observed along the medial cortex and distal to the stem in most cases.This bone remodeling and apposition may prevent occurrence of femoral fractures in the long-term that occur with cemented THR because of cortical thinning.

Component loosening
(acetabular component n=1) Implant stresses are higher in the ZCTHR stem compared with cemented stems, reaching a maximum in the neck region of the implant, with a second peak at the level of the most proximal screw.

Limitations:
 retrospective descriptive case series study, no control groups, variable follow-up  two ZCTHR were not evaluated radiographically

Variables evaluated:
 age, body weight, breed, indication for THR and prosthesis

Statistical analysis:
 association between each variable and the incidence of complications were assessed using logistic regression  Mann-Whitney U-Test was performed to assess the significance of total lameness scores before and after THR Owner outcomes assessment questionnaire was used additionally to collect data from owners.

Divided in 4 sections:
 Section A assessed information regarding length of ongoing mobility problem, medications received, and other concurrent medical history unrelated to hip dysplasia (HD). Section B assessed activity and willingness to exercise before THR. Section C assessed date of surgery, overall owner satisfaction and complications. Section D assessed activity and willingness to exercise after THR.
Outcome studied: To assess the variables associated with the complications of THR and report owner-assessed outcomes, through surgeon-based registration of cases via an online database, informed owner consent and prospective outcomes assessment using a client-administered clinical metrology instrument.
A total of 20% complication rate was reported statistically significant difference in owner-assessed lameness scores before and after THR (p<0.001).Participating surgeons were requested to submit all of their operated cases to the BVOA-CHR (authors unable to control it).In theory it is possible that participating surgeons may have chosen not to submit data from a case with a less successful outcome (selection bias).The complication rate in this study would therefore be higher than documented.
Limitations: Participating surgeons were requested to submit all of their operated cases to the BVOA-CHR (authors unable to control it).In theory it is possible that participating surgeons may have chosen not to submit data from a case with a less successful outcome (selection bias).The complication rate in this study would therefore be higher than documented. client assessment subjective, no controls with force platform peak vertical force for example  51% response rate to the online owner assessment questionnaire  different complications rate between owners and surgeons  THR was most frequently performed on dogs ≤ 1year (39%)  limited ability to fully evaluate the risk factors for THR complications to date, probably because most studies are single-center and have limited power  BVOA-CHR does not currently include imaging data

Appraisal, application and reflection
The aim of this Knowledge Summary was to review, summarise and critically appraise the literature regarding the question: Are old dogs undergoing total hip replacement predisposed for perioperative femoral fractures in comparison to young dogs?This reflection was investigated in two observational retrospective studies: Liska (2004) and Ganz et al. (2010).Both studies reported that elderly dogs undergoing THR may be at an increased risk of femoral fractures.According to Liska (2004) old dogs with osteopathies, previous hip surgery and iatrogenic fissures created during reaming are predisposing factors for femur fracture after THR.The overall incidence of femur fracture after THR was 2.9%.The author describes an excellent prognosis when the fractures were treated correctly.Due to the design of the study (observational and retrospective) and the limitations (retrospective, different surgeons, surgical method/surgical experience not evaluated as a risk factor, different follow-up) the results have to be interpreted with caution.Ganz et al. (2010) evaluated the risk factors for femoral fracture after canine press-fit cementless total hip arthroplasty.The conclusions were that older dogs and dogs with lower CFI may be at increased risk of femoral fracture and the incidence of femoral fracture of cases with complete statistical analysis was 13.5%.Regarding the design of this study (observational and retrospective) and its limitations (retrospective, different surgeons, surgical method/surgical experience not evaluated as a risk factor, different follow-up) the level of evidence provided by this type of study is weak.However, the results of these studies can be used to counsel clients before performing THR in old dogs.
By reviewing the veterinary literature relative to the general complications and outcomes of total hip arthroplasty, a variation in the prevalence of femoral fractures was noticed.None of the studies listed above mentioned that the age of the patient may be a potential risk factor for general complications.Again, due to the design of all of these studies the level of evidence is weak.
In the study of Hummel et al. (2010) the prevalence of femoral fractures occurring intraoperatively was 7.4%, the prevalence of femoral fractures occurring as short-term complications was 1.2% and the prevalence of femoral fractures occurring as long-term complications was 0.6%.Increased body weight and prior cemented THR or femoral head and neck ostectomy of the contralateral hip were identified as negative prognostic factors.
Berg et al. ( 2006) described a 1.3% prevalence of femoral fracture occurring during the primary THR as short term complications, and fracture of femur diaphysis represented 3.2% of the long term complications.Eccentric positioning of the femoral stem and the presence of radiolucent lines at the femoral cement-bone interface were positively associated with the occurrence of revision surgery.
The prevalence of femoral fracture post-operatively in the study of Guerrero and Montavon (2009) was 1.5%.Forster et al. (2012) identified no significant association between weight, age, sex, breed, indication for THR, surgical technique and prosthesis and the incidence of complications of total hip arthroplasty.The incidence of surgeon-reported surgical complications was 9.4%.The femoral fractures represented 18.75% of the complications.
In conclusion, there is insufficient evidence that elderly dogs undergoing THR are predisposed to femoral fractures in comparison to young dogs.Currently, there are insufficient strongly convincing studies in the veterinary literature to support the results of Liska (2004) and Ganz et al. (2010).

Bergh et al (2006) Population:
Client-owned dogs that underwent cTHR at the University of Pennsylvania School of Veterinary Medicine.
Veterinary Evidence ISSN:2396-9776 Vol 1, Issue 2 DOI: http://dx.doi.org/10.18849/ve.v1i2.28 next review date: 20 Jan 2018 p a g e | 7 total pages: 16 IOC overall IOC rate = 11%  fracture of the femoral diaphysis (n=12)=7.4% fracture of the greater trochanter (n=3)  lost screw in soft tissue (n=1)  excessive haemorrhage (n=1)  immediate revision of acetabular cup placement (n=1)  septic loosening (n=6)  coxofemoral luxation (n=6) implant failure (n=4) fracture of the femoral diaphysis (n=1)(0.6%)Limitations:  retrospective  surgery performed by 10 different surgeons  follow-up performed in different ways and by different persons  relatively short scheduled radiographic follow-up (8 weeks)  50 cases in which the 8 week follow-up was not performed by board certified specialists (surgery and radiology), body score not evaluated in all dogs  increased rate of complications higher than previous studies.Likely influenced in part by varying degrees of Z-THR experience (technical error)  surgeon's experience not evaluated as prognostic factor  no evaluation of technical errors immediately post operatively 4. Complications Veterinary Evidence ISSN:2396-9776 Vol 1, Issue 2 DOI: http://dx.doi.org/10.18849/ve.v1i2.28 next review date: 20 Jan 2018 p a g e | 8 total pages: 16  intraoperative (IOC)  short term (STC)  long term (LTC)  categorical data: chi-square or Fisher's exact test  logistic regression analysis to access the independent contribution of possible risk factors Surgery performed by various board certified surgeons Study design: Retrospective cohort study.Outcome studied: To identify the prevalence of complications and changes following cTHR and to identify factors that may predispose to a need for revision surgery. 69 dogs had appropriately sized femoral implants  60 dogs had eccentrically placed implants Veterinary Evidence ISSN:2396-9776 Vol 1, Issue 2 DOI: http://dx.doi.org/10.18849/ve.v1i2.28 next review date: 20 Jan 2018 p a g e | 9 total pages: 16

15 total pages: 16 Exclusion / Inclusion Criteria Exclusion
StrategySearch terms: dogs OR dogs OR canine AND femoral fractures OR femur OR femoral OR fracture AND total hip replacement OR total hip arthroplasty OR cemented total hip replacement OR cementless total hip replacement OR uncemented total hip replacement OR cemented total hip arthroplasty Dates searches performed: December 2015 : Non English language, conference papers, summary updates, case reports, reviews.Inclusion: Studies which were looking for risk factors and outcomes in total hip replacement, studies which described femoral fractures as complications.Experimental studies and observational studies.