 TRAUMA

The Targon Femoral Neck hip screw versus cannulated screws for internal fixation of intracapsular fractures of the hip A RANDOMISED CONTROLLED TRIAL X. L. Griffin, N. Parsons, J. Achten, M. L. Costa From Warwick Medical School, University of Warwick, Coventry, United Kingdom

We compared a new fixation system, the Targon Femoral Neck (TFN) hip screw, with the current standard treatment of cannulated screw fixation. This was a single-centre, participant-blinded, randomised controlled trial. Patients aged 65 years and over with either a displaced or undisplaced intracapsular fracture of the hip were eligible. The primary outcome was the risk of revision surgery within one year of fixation. A total of 174 participants were included in the trial. The absolute reduction in risk of revision was of 4.7% (95% CI 14.2 to 22.5) in favour of the TFN hip screw (chi-squared test, p = 0.741), which was less than the pre-specified level of minimum clinically important difference. There were no significant differences in any of the secondary outcome measures. We found no evidence of a clinical difference in the risk of revision surgery between the TFN hip screw and cannulated screw fixation for patients with an intracapsular fracture of the hip. Cite this article: Bone Joint J 2014;96-B:652–7.

 X. L. Griffin, MA, FRCS (Tr & Orth), PhD, NIHR Clinical Lecturer Warwick Orthopaedics, Warwick Medical School, University of Warwick, Coventry, CV4 7AL, UK.  N. Parsons, MSc, PhD, Statistician University of Warwick, Warwick Medical School, University of Warwick, Coventry CV4 7AL, UK.  J. Achten, MSc, PhD, Research Manager University of Warwick, Warwick Orthopaedics, Warwick Medical School, University of Warwick, Coventry CV4 7AL, UK.  M. L. Costa, PhD, FRCS (Tr & Orth), Professor University of Warwick, Clinical Trials Unit, Coventry CV4 7AL, UK. Correspondence should be sent to Mr X. L. Griffin; e-mail: [email protected] ©2014 The British Editorial Society of Bone & Joint Surgery doi:10.1302/0301-620X.96B5. 33391 $2.00 Bone Joint J 2014;96-B:652–7. Received 27 October 2013; Accepted after revision 31 December 2013

652

Surgical options for the treatment of intracapsular fractures of the hip are divided into those that preserve the femoral head and various types of joint replacement. The latter are associated with an increased risk of early death and the catastrophic complications of deep infection and dislocation.1,2 However, it remains the preferred option for many of these fragility fractures due to the considerable risk of fixation failure in patients with a displaced fracture.3 The principal complication of fixation, the risk of mechanical failure, is largely related to the physical limitations of osteoporotic bone and the need for early full weight-bearing in these patients. The possibility of an impaired blood supply to the femoral head following fracture may also increase the risk of fracture healing complications. Numerous different implants have been used to treat this fracture4 including single or multiple screws or pins, some of which are associated with a plate attached to the lateral femoral cortex. To date none of these implants have been able to offer sufficiently secure fixation to reduce the risk of revision to an acceptable level.4 One novel implant, the Targon Femoral Neck (TFN) hip screw (Aesculap, Melsungen, Germany) incorporates features of both a sliding hip screw and multiple cancellous screws, the two most common implants used

for fixation. Promising early results from a consecutive case series have been reported by the developer.5 We performed a randomised trial to explore the difference in the risk of revision surgery in patients aged 65 years and older with an intracapsular hip fracture treated with the TFN hip screw and those treated with cannulated screws.

Patients and Methods This is a single-centre, parallel group, participant blinded, randomised standard-of-care controlled trial. Full details of the protocol have been published elsewhere.6 The trial was given ethical approval on 6th July 2009 by Coventry Research Ethics Committee (09/ H1210/22). The trial was registered at Current Controlled Trials, ISRCTN49197425 and UKCRN 7762. All patients aged 65 years and over with either a displaced or undisplaced intracapsular hip fracture were eligible, including those with cognitive impairment. Patients were excluded if they were managed non-operatively, presented late after their injury, had serious injuries to either lower limb which interfered with rehabilitation from their hip fracture, or had existing local disease precluding fixation, e.g. a local tumour deposit or symptomatic ipsilateral osteoarthritis of the hip. THE BONE & JOINT JOURNAL

THE TFN HIP SCREW VERSUS CANNULATED SCREWS FOR INTERNAL FIXATION OF INTRACAPSULAR FRACTURES OF THE HIP

Participants were recruited between September 2009 and October 2011 from the acute trauma admissions to University Hospitals Coventry and Warwickshire NHS Trust, Coventry, UK. This is a major trauma centre that serves a population of two million people and which treats approximately 600 patients with a hip fracture each year.7 Participants with capacity gave written consent. For those who lacked capacity, written consent was given by a consultee in accordance with the Mental Capacity Act 2005. We randomly allocated patients to one of two groups: standard-of-care fixation with cannulated screws or fixation with the TFN hip screw. The treatment allocation was determined using a computer-generated, randomised number sequence administrated by an independent Clinical Trials Unit via a secure online programme. The randomisation code was stratified by displacement of the fracture8 and split into unequal block sizes. Stratification ensured that the fractures that were undisplaced were distributed evenly between groups as undisplaced fractures are associated with a very substantially improved outcome. The degree of displacement was determined by the treating surgeon. The code was only broken at the end of the trial once the trial statistician had locked and analysed the dataset. Allocation to treatment group took place intra-operatively. If the fracture was displaced, the allocation was only performed after the operating surgeon confirmed a successful closed reduction of the fracture. Those patients in whom a reduction could not be achieved underwent hip replacement. After closed reduction of the fracture, the lower limb was supported on a fracture table. Internal fixation of the fracture was achieved through a standard lateral approach with peri-operative antibiotic cover in accordance with the NHS Trust protocol. Post-operative care was the same for both groups of patients: early active mobilisation and immediate full weight-bearing with a standardised physiotherapy rehabilitation regime. All participants received routine prophylaxis against deep vein thrombosis. Standard-of-care fixation was with parallel cannulated screws. The number and configuration of the screws were determined by the treating surgeon according to their own normal practice to ensure that the results could be generalised. For those participants allocated to the TFN group, fixation was achieved in accordance with the manufacturer’s recommended technique. Surgeons received a standardised training package from the developer before the trial started and also had an opportunity to practice the operative technique using sawbones. The operating surgeon could not be blind to the intervention. The primary outcome measure was the proportion of participants undergoing re-operation for failure of fixation within one year of sustaining the fracture. Secondary outcome measures were radiological nonunion at one year (nonunion was defined as failure of the fracture to show signs of bony union on the anteroposterior or lateral radiograph one year after surgery),9 radiological evidence of avascular necrosis at one year, the EQ-5D index VOL. 96-B, No. 5, MAY 2014

653

(York A1 value set)10 at six, 12 and 52 weeks, length of index hospital stay, mortality, and adverse events. As few data were available with which to estimate the possible size of a treatment effect of the TFN,5 the sample size was derived from an internal pilot arm within a larger comprehensive cohort.6 In the absence of an agreed method of determining sample size for a pilot study, a pragmatic decision was taken to recruit 25 participants to the TFN group for comparison with the 100 participants having cannulated screw fixation. The data from this internal pilot was then used to inform a formal sample size calculation. The literature indicates that the overall risk of failure of fixation for intracapsular fractures is in the region of 30%.9-11 Assuming a minimum clinically important absolute reduction in risk of 20% and significance at 5%, an additional 50 participants were required to provide 80% power, accounting for 20% mortality at one year. Therefore, the overall group sizes required were approximately 125 and 50 participants in the control and TFN groups respectively. Sample sizes were determined using the PS power and sample size software.12 Statistical analysis. The primary outcome measure, the proportion of patients requiring re-operation for failure of fixation (revision) within one year of sustaining the fracture, was compared between treatment groups using a chisquared test, where data from participants were analysed by treatment allocation. Treatments were considered to differ significantly if p-values were < 0.05. The primary analysis considered an available case analysis where deaths without revision were excluded. A sensitivity analysis was performed with deaths imputed as both revisions and nonrevisions to assess the effect of excluding patients who died in the year after injury. Fisher’s exact test was used to assess the significance of observed differences for the secondary proportional outcome measures. For continuous outcomes, which were approximately normally distributed, mean differences were tested using a two-tailed t-test; for non-parametric data (length of stay) differences were tested with the Mann–Whitney U test. The incidence of adverse events were reported for each treatment group, stratified by the type of event. Planned subgroup analyses were only undertaken for prespecified subgroups. Explanatory variables of gender, fracture displacement, dementia and age were entered into a logistic regression model and included associated interaction terms with the treatment arm for each. A planned subsidiary analysis used a multiple linear regression model to investigate the relationship between each participant’s EQ-5D score at one year post operation and the treatment group, after appropriate adjustment for age, gender and fracture displacement for each participant.

Results A summary of the flow of participants through the study is shown in Figure 1: 63% of eligible patients were included in the trial, largely due to recruitment taking place only

654

X. L. GRIFFIN, N. PARSONS, J. ACHTEN, M. L. COSTA

Patients admitted with fracture (n = 523) Not screened (research team unavailable) (n = 89)

Assessed for eligibility (n = 434) Excluded (n = 138) • • • • • • • •

Recruited to CCS (n = 296)

Ineligible (n = 96) Declined (n = 20) Theatre staff unavailable (n = 5) Treating surgeon decision (n = 7) Recruited contralateral hip (n = 4) Died pre-recruitment (n = 4) Arthrodesis (n = 1) Translator unavailable (n = 1)

Withdrawn (n = 21) • • • •

Randomised (n = 275)

Irreducible (n = 7) Died (n = 2) Treating surgeon decision (n = 8) Patient choice (n = 1)

Allocated to PRT test treatment (n = 101)a

Allocated to control intervention (n = 123) • •

Allocated to test intervention (n = 51) • •

Received control intervention (n = 123) Did not receive control intervention (n = 0)

Withdrawn (n = 0) Lost to follow-up (primary outcome) (n = 24) • Died (n = 26)b Secondary outcome unavailabled • 6 weeks (n = 60) • 12 weeks (n = 58) • 52 weeks (n = 73)

Received test intervention (n = 49) Did not receive test intervention (n = 2) o administrative error

Withdrawn (n = 0) Lost to follow-up (primary outcome) (n = 10) • Died (n = 14)c Secondary outcome unavailablee • 6 weeks (n = 27) • 12 weeks (n = 28) • 52 weeks (n = 29)

Analysed (n = 99)

Analysed (n = 41)

Fig. 1 CONSORT flow diagram: a) reported elsewhere (n = 13); b) two participants underwent revision prior to death; c) four participants underwent revision prior to death; d) 30 unavailable at baseline; e)18 unavailable at baseline; CCS, Comprehensive cohort study; PRT, platelet rich therapy.

during the working week. Participants in the trial were part of a larger comprehensive cohort.6 Our findings from the second arm of the cohort have been reported elsewhere.13 Of the 123 participants allocated to the cannulated screw group, 97 (79%) completed the trial protocol. Of the 51 participants allocated to the TFN group, 37 (73%) completed the protocol. In all, 40 participants died, six of whom underwent revision surgery prior to death; consequently, 140 participants were available for primary analysis. There were two patients allocated to the TFN who crossed over to have cannulated screw fixation. Missing data at each time-point (due to death, co-existing chronic confusional states at the time of recruitment,

new onset co-morbidities and participant withdrawals) are also shown in Figure 1. The baseline characteristics of the trial participants are described in Table I. There were no substantial differences between the two groups for any of the recorded baseline characteristics. Both the test and control treatments were delivered under the supervision of consultant trauma surgeons and performed by a total of 24 specialist trainees. Table II shows counts and estimated risks of revision surgery by treatment group. There was an absolute risk reduction (ARR) of 4.7% (95% CI -14.2 to 22.5) in favour of the TFN (chi-squared test, p = 0.741). Imputing all the deaths THE BONE & JOINT JOURNAL

THE TFN HIP SCREW VERSUS CANNULATED SCREWS FOR INTERNAL FIXATION OF INTRACAPSULAR FRACTURES OF THE HIP

655

Table I. Baseline characteristics for each group shown as mean (standard deviation [SD]) or percentages (%). (n/a, not applicable; n/r, not recorded) Group Characteristic

Not screened (n = 89)

Screened but not recruited (n = 138) Control (n = 123)

Test (n = 51)

Age (years) (SD) Female (%) Minimally displaced fractures (%) Demented (AMT< 8) (%) Pre-morbid EQ-5D Previously diagnosed CRF (%) Previously diagnosed diabetes mellitus (%) Previously diagnosed osteoporosis (%) Currently prescribed anti-platelet drug (%) Previously or currently prescribed systemic steroid (%) Currently prescribed NSAID (%) Currently smoking (%) Time to theatre (hours)

82 (7.7) 65 22 36 n/r n/r n/r n/r n/r n/r n/r n/r n/a

73 (18) 61 20 25 n/r n/r n/r n/r n/r n/r n/r n/r n/a

83 (7.6) 73 20 35 0.69 (0.26) 7.8 16 31 18 7.8 5.9 5.9 28 (21)

83 (7.7) 75 20 31 0.67 (0.32) 4.9 8.1 20 30 6.5 4.1 9.8 31 (30)

Table II. Revision at 12 months post index operation Group

Unrevised

Revised

Total

Risk (%)

Control Test Total

63 28 91

36 13 49

99 41 140

36.36 31.71 35.00

Table III. Between-group differences in secondary outcome measures. Values shown percentages (%); summary statistics as median and IQR Treatment group Outcome

Control (n = 99)

Test (n = 41)

Test

Significance

Radiological nonunion at one year (%) Radiological avascular necrosis at one year (%) Length of index hospital stay (days) Mortality (%)

2 1 22 (10 to 35) 26

1 1 18 (8 to 31) 14

Fisher Exact Fisher Exact Mann–Whitney Fisher Exact

1.00 0.50 0.58 0.43

as ‘revisions’ increased overall estimates of revision risks, but due to the balance across groups this had little impact on effect estimates (control risk 48.8%; ARR in favour of the TFN hip screw 3.7%, 95% CI -14.0 to 21.3; chi-squared test p = 0.783). Logistic regression analysis, with gender, fracture displacement, dementia and age added to the model, gave an adjusted odds ratio of 0.82 (95% CI 0.36 to 1.80), which was almost exactly the same as the unadjusted odds ratio of 0.81 from Table II, and provided no evidence for a significant treatment effect (z-test from logistic regression p = 0.626). Interaction terms were added to the model to test for pre-specified subgroup effects. None of the interaction terms significantly improved the model fit, providing no evidence for substantial subgroup effects. There was no significant difference in unadjusted mean EQ-5D score at one year between the control and treatment groups (mean difference (MD) = -0.055, t-test p = 0.526). This was maintained after adjusting for age, gender and fracture displacement. Similarly, there were no significant VOL. 96-B, No. 5, MAY 2014

differences in other secondary outcomes (Table III). The number and distribution of adverse events (pro rata) were similar in both treatment groups (Table IV). One participant who underwent fixation with the TFN hip screw developed a deep infection, leading to resorption of the acetabular margin and subsequent dislocation of the fixed native hip. This participant underwent revision to a total hip replacement.

Discussion We found no evidence of a difference in the risk of revision surgery between participants treated for an intracapsular fracture of the proximal femur with the TFN hip screw and those treated by cannulated screw fixation. However, as the confidence intervals include the pre-specified 20% risk reduction, we have been unable definitively to exclude a clinically important difference. A sensitivity analysis to explore the effect of decisions about the handling of the missing data and the competing risks of death and revision surgery found similar estimates of the effect size. The

656

X. L. GRIFFIN, N. PARSONS, J. ACHTEN, M. L. COSTA

Table IV. Between-group differences in adverse events. Events are not mutually exclusive Absolute number of events Expected adverse event

Control group (n = 123)

Test group (n = 51)

Wound infection Pulmonary embolus Pneumonia Urinary tract infection Blood transfusion Cerebrovascular accident Acute coronary syndrome Myocardial infarction Deep vein thrombosis Death

5 2 13 8 2 2 1 1 2 26

4 2 2 2 0 1 0 1 1 14

secondary outcomes showed effects that were concordant with the primary outcome. There was no evidence of any subgroup interaction effects. The main strength of this trial was its pragmatic nature: we included all intracapsular fractures of the hip; the treating surgeon determined the displacement or otherwise of the fracture and a large number of surgeons (including supervised trainees) performed the reduction and fixations in both groups using their preferred technique. Since each individual surgeon performed only a small number of interventions, the impact of the ‘surgeon effect’, related to both experience and technical expertise, was likely to have been small. The main limitation of the study is that the trial was conducted at a single centre. Only approximately 80% of the available population was screened for eligibility as the trial staff was often not available outside the working week. This might have produced a sampling bias, although review of the admission and screening data revealed no substantial differences in the crucial confounders of age, gender, fracture displacement and chronic cognitive impairment between the unscreened and recruited patients. Few data exist from other similar studies with which to compare these findings. This is the first report of a randomised trial of the TFN hip screw and the first data from a centre other than that of the developer. The developer reported an overall risk of revision surgery of 20% (65/ 320).5 We comparably report a risk of revision of 32% in patients randomly allocated to fixation with the Targon FN hip screw. Importantly, the developer’s series reported outcomes only in patients selected for treatment with this implant. Of note, these patients were younger (mean age 76 years) and a greater proportion had undisplaced fractures (35%). These differences in baseline characteristics may partly explain the observed differences in the absolute risk of revision surgery. It is also possible that outcome after fixation with the TFN hip screw is more favourable in the hands of the developer. Our modelling showed that fracture displacement was a significant predictor of risk of revision. This is consistent with clinical experience and previous authors’ findings.9 The cohort study reported by Parker, Raghavan and

Gurusamy9 recruited more participants than this trial and identified risk factors with smaller effect sizes. In conclusion, we have not found any evidence that the TFN hip screw reduces the risk of fixation failure in patients with an intracapsular fracture of the hip. We have been unable definitively to outline a potential benefit but, in the absence of a large reduction in the risk of failure of fixation, the standard-of-care surgery for patients with a displaced intracapsular fracture will remain joint replacement in accordance with the NICE guidance.3 Future work might investigate the effectiveness of the Targon FN hip screw in the management of fractures which may be more amenable to fixation such as those that are undisplaced, the result of high energy trauma, or occur in men. We thank B. Kearney, K. McGuinness, H. Richmond, K. Dennison, Z. Buckingham, T. Douglin, and C. Richmond for their assistance in recruitment and data collection during the trial; P. Roberts, C. Jones, P. Kimani and S. Drew for their clinical, trials, and regulatory expertise in the trial steering committee and data monitoring committee for this trial; and all the patients for their time and effort in participating in this trial. Support for salaries and consumable costs was received from The Bupa Foundation and Orthopaedic Research UK and consumables for the Targon System by B. Braun UK (Sheffield, UK). All authors were and are independent of these funders. The sponsors and funders had no role in the inception or design of the study, data collection and analysis, or preparation of this report. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. This article was primary edited by D. Rowley and first proof edited by G. Scott.

References 1. Parker MJ, Gurusamy K. Internal fixation versus arthroplasty for intracapsular proximal femoral fractures in adults. Cochrane Database Syst Rev 2006;4:CD001708. 2. Bhandari M, Devereaux PJ, Swiontkowski MF, et al. Internal fixation compared with arthroplasty for displaced fractures of the femoral neck: a meta-analysis. J Bone Joint Surg [Am] 2003;85-A:1673–1681. 3. No authors listed. The Management of Hip Fracture in Adults. National Clinical Guideline Centre, 2011. http://www.ncgc.ac.uk/ (date last accessed 15 January 2014). 4. Parker MJ, Stockton G. Internal fixation implants for intracapsular proximal femoral fractures in adults. Cochrane Database Syst Rev 2001;4:CD001467. 5. Parker M, Cawley S, Palial V. Internal fixation of intracapsular fractures of the hip using a dynamic locking plate: two-year follow-up of 320 patients. Bone Joint J 2013;95-B:1402–1405. 6. Griffin X, Parsons N, Achten J, Costa ML. Warwick Hip Trauma Study: a randomised clinical trial comparing interventions to improve outcomes in internally fixed intracapsular fractures of the proximal femur. Protocol for The WHiT Study. BMC Musculoskelet Disord 2010;11:184. THE BONE & JOINT JOURNAL

THE TFN HIP SCREW VERSUS CANNULATED SCREWS FOR INTERNAL FIXATION OF INTRACAPSULAR FRACTURES OF THE HIP

7. No authors listed. National Hip Fracture Database. NHFD Resources. http:// www.nhfd.co.uk/003/hipfractureR.nsf/resourceDisplay?openform (date last accessed 15 January 2014). 8. Parker MJ. Garden grading of intracapsular fractures: meaningful or misleading? Injury 1993;24:241–242. 9. Parker MJ, Raghavan R, Gurusamy K. Incidence of fracture-healing complications after femoral neck fractures. Clin Orthop Relat Res 2007;458:175–179. 10. Lu-Yao GL, Keller RB, Littenberg B, Wennberg JE. Outcomes after displaced fractures of the femoral neck: a meta-analysis of one hundred and six published reports. J Bone Joint Surg [Am] 1994;76-A:15–25.

VOL. 96-B, No. 5, MAY 2014

657

11. Parker MJ, Khan RJK, Crawford J, Pryor GA. Hemiarthroplasty versus internal fixation for displaced intracapsular hip fractures in the elderly: a randomised trial of 455 patients. J Bone Joint Surg [Br] 2002;84:1150–1155. 12. Dupont WD, Plummer WD Jr. Power and sample size calculations: a review and computer program. Control Clin Trials 1990;11:116–128. 13. Griffin XL, Achten J, Parsons N, Costa ML. Platelet-rich therapy in the treatment of patients with hip fractures: a single centre, parallel group, participant blinded, randomised controlled trial. BMJ Open 2013;3:e002583.