O

Na

Aa

b

c

ARA

KTPCM

1

ibMpa[aupTc

(

1

Orthopaedics & Traumatology: Surgery & Research 101 (2015) 461–468

Available online at

ScienceDirectwww.sciencedirect.com

riginal article

o demonstrable benefit for coronal alignment outcomes in PSI kneerthroplasty: A systematic review and meta-analysis

. Mannana,∗, T.O. Smithb, C. Sagara, N.J. Londonc, P.J.A. Molitora

Scunthorpe General Hospital, Cliff Gardens, Scunthorpe, DN15 7BH North Lincolnshire, UKUniversity of East Anglia, Norwich, UKHarrogate District Hospital, Lancaster Park Road, HG2 7SX Harrogate, UK

a r t i c l e i n f o

rticle history:eceived 22 July 2014ccepted 15 December 2014

eywords:otal knee arthroplastyatient-specific instrumentationomputer-assisted surgeryeta-analysis

a b s t r a c t

Aim: A systematic review and meta-analysis of clinical studies assessing alignment outcomes in patient-specific instrumented (PSI) knee arthroplasty was conducted.Materials and methods: PRISMA compliant data was extracted from literature databases up to January2014.Results: Twenty-six studies met the inclusion criteria, reporting a total of 1792 knees. Twenty-threestudies reported alignment outcomes in the coronal plane, 11 in the sagittal plane. In all but three series,MRI was the preoperative imaging modality. Range of mean postoperative alignment (hip–knee–ankle[HKA] angle) was 176.5 to 181.70. The proportion of three degrees of outliers showed an overall mean of18.6%. In total, fifteen studies compared alignment outcomes between standard and PSI. From these, fourstudies showed significantly higher accuracy of coronal plane alignment with PSI (HKA angle). Meta-analysis of seven high-quality comparative studies demonstrated no significant increased accuracy inpostoperative mechanical axis (HKA angle) with PSI. Subgroup meta-analysis of both femoral and tibial

rotation was not feasible due to a low number of inclusive high-quality series.Conclusions: PSI knee arthroplasty is shown not to confer increased accuracy in reconstituting the post-operative mechanical axis. Further studies are required to demonstrate both clinical and radiologicalalignment outcomes in PSI knee arthroplasty with focus upon tibial and femoral rotation.Level of evidence: Level 2 – meta-analysis.

© 2015 Elsevier Masson SAS. All rights reserved.

. Introduction

The requirement of necessary restoration of a correct mechan-cal axis in total knee arthroplasty has been supported iniomechanical [1], finite element [2] and clinical [3] studies.alalignment typically is present in either the coronal or sagittal

lanes or is rotational in nature, and may lead to wear, looseningnd patellar instability often necessitating revision procedures4,5]. Coronal malalignment specifically has been shown to bessociated with altered pressure distribution and load bearingpon the medial and lateral compartments of the tibial com-

onent [6], higher failure rates and lower functional scores [7].he reconstitution of a coronal plane postoperative mechani-al axis within a target range of: zero degrees ± three degrees

∗ Corresponding author. Tel.: +44 7713333488.E-mail addresses: ashim mannan@hotmail.com, ashim.mannan@nhs.net

A. Mannan).

http://dx.doi.org/10.1016/j.otsr.2014.12.018877-0568/© 2015 Elsevier Masson SAS. All rights reserved.

(valgus/varus) has been shown to reduce the occurrence of neg-ative outcomes including implant failure [8]. The limitations ofconventional instrumentation in effecting full precision in bonecuts and accuracy of alignment has previously been reported[9,10]. Computer-assisted navigation has been shown to improvemechanical axis alignment over conventional instrumentation[11], but has been associated with longer operative duration andno significant enhancement in short-term clinical outcomes [12].Additionally, in the absence of a preoperative computed tomog-raphy scan, navigated knee arthroplasty has not been shown tosignificantly improve rotational alignment outcomes as comparedto conventional instrumentation [13]. These factors have promotedefforts to pursue a ‘precise’ surgical technology model with thekey goal of improving postoperative alignment and positioning.

Patient-specific instrumentation (PSI) is a recent technology

that presents an alternative to conventional instrumentation andcomputer navigation, with the primary aim of achieving accuratealignment and positioning of implants [14]. PSI exploits preoper-ative anatomical data obtained by either computer tomography

462 A. Mannan et al. / Orthopaedics & Traumatology

Box 1: MEDLINE search strategy.

• Total knee replacement.tw.• Exp. Total knee arthroplasty• OR/1,2• Conventional.ti,ab• Patient-specific.ti,ab• Custom-fit.ti,ab• Alignment.ti,ab• OR/4-7• Exp. clinical trial/• Randomised controlled trial.ti,ab

(prmbtrkoiwti

2

udLtmlaGtPIP

arcss

••••

•••

•

ga

• Exp.randomized controlled trial• OR/9-11

CT) imaging or magnetic resonance imaging (MRI) to create dis-osable patient-specific bone resection guides, individualised toeflect the patient’s anatomy [14]. Although PSI is principally aodern system, several (prospective and retrospective) series have

een published documenting alignment outcomes. The aim ofhis systematic review was to document and describe the cur-ent literature and evidence in alignment outcomes relating to PSInee arthroplasty, and additionally to apply meta-analysis meth-ds to compare alignment outcomes in PSI versus conventionalnstrumentation or navigation techniques studies. Through this,

e aimed to answer the following research question: what arehe alignment outcomes in patients who undergo patient-specificnstrumented knee arthroplasty procedures.

. Methods

A detailed search of relevant literature within the period Jan-ary 2006–January 2014 was performed, including the followingatabases: Pubmed, the Cochrane Collaboration Trial Registry andibrary, MEDLINE and EMBASE. The specific search terms used forhe MEDLINE search strategy are presented in Box 1. These were

odified for each of the individual databases. A search of the greyiterature/unpublished literature databases and trial registries waslso undertaken. These included: ISI Web of Knowledge and Openrey (System for Information on Grey Literature in Europe), and

he trial registries: the WHO International Clinical Trials Registrylatform, Current Controlled Trials, the United States Nationalnstitute of Health Trials Registry, and NIHR Clinical Researchortfolio Database.

Two investigators independently reviewed all search terms,bstracts and full text of articles potentially eligible for abstracteview. Further examination of reference lists of retrieved arti-les was conducted to cross-reference any further studies that metelection criteria. All abstracts and references were selected in con-ensus with the PRISMA flowchart (Fig. 1).

Selection criteria included studies that documented:

primary PSI TKA;prospective/retrospective/case controlled and cohort studies;postoperative implant position and limb alignment outcomes;minimum of 10 patients in either PSI (or conventional series whendocumented in the same study);exclusion with fracture deformity and tumour;exclusion of animal and cadaver studies;exclusion of ‘kinematic restorative’ PSI knee implants (as thesedo not specifically aim to correct mechanical axis);studies specifically examining unicondylar knee implants.

Eligible studies were evaluated independently by two investi-ators. Data was obtained from each included study, entered into

data extraction form, verified for accuracy and then analysed.

: Surgery & Research 101 (2015) 461–468

Further studies were sought via examination of reference lists. Theprimary outcome measure was deviation from the postoperativehip–knee–ankle (HKA) angle (i.e. the angle between the mechanicalaxis of the femur and the tibia, with both lines crossing at the knee).This endpoint was documented as the number/proportion of three-degree outliers (i.e. 0 ± three degrees varus/valgus). Secondaryoutcomes included: femoral coronal alignment (i.e. the anglebetween the articular margin of the femoral component and thefemoral axis); tibial component alignment (angle between tibialcomponent surface and the tibial axis); and the zone of mechanicalaxis (where the tibial base plate is divided into three equal zonesi.e. medial, central and lateral, and the proportion/number in whichthe mechanical axis bisected the central zone). Finally, sagittaloutcome measures were documented, these included tibial slopeangle (FTC) and femoral component flexion (FFC), the endpointbeing the number and proportion of three-degree outliers.

Methodological quality of comparative studies (comparing PSIversus conventional instrumentation) was assessed by the DetskyQuality Assessment scale [15]. The Detsky scale includes 14 itemsthat assess reporting quality among five categories. These include:

• randomisation;• outcome measures;• eligibility criteria and patient exclusion;• interventions;• statistical analysis.

This validated and established scoring criteria assesses the qual-ity of study design, methodology and also outcome reporting.

Where appropriate, a random-effects model meta-analysiswas performed to evaluate the clinical primary outcomes. Meta-analysis was deemed appropriate in the absence of clinicaland statistical heterogeneity. Clinical heterogeneity was assessedthrough observation of data extraction tables whilst statisticalheterogeneity between studies was evaluated with �2 with demon-stration of values for P and ʃ2, P < 0.1 and ʃ2 > 50% indicatingheterogeneity. Deviance from a postoperative neutral mechani-cal axis was measured as the proportion of three-degree outliers.RevMan 5.0 (The Cochrane Collaboration) was used for purposes ofstatistical analysis. For dichotomous variables, odds ratio (OR) and95% confidence intervals (CI) were calculated and graphical outputwas documented by forest plots. A funnel plot was constructed toassess small sample size publication bias for the primary outcomemeasure. A sensitivity analysis was undertaken of higher qualitystudies (Detskey score ≥ 16) on meta-analysis.

3. Results

3.1. Search results

The results of the search strategy are presented in Fig. 1’sPRISMA flowchart. A total of 259 studies were identified from thekeywords search. Two hundred and twenty-eight studies wereexcluded from review of abstracts. In addition, two studies exam-ined animal models [16,17], three studies focused on ‘kinematic’knee arthroplasty [18–20], two studies reported on outcomes inunicondylar knee replacement [21,22], three studies reported onseries less than 10 in number [23–25], three studies examinedaccuracy of templating with computer navigation and one studyexamined 3D implant position with exclusion of reference to coro-nal and sagittal plane outcomes [24,26–28].

One study examined the role of extra-articular deformity in

alignment outcomes [29]. Overall, 26 studies met the criteriareporting on a total of 1792 knees (Table 1). Twenty-one studieswere prospective and five retrospective. Twenty-three studiesexamined outcomes in the coronal plane and 11 in sagittal plane

A. Mannan et al. / Orthopaedics & Traumatology: Surgery & Research 101 (2015) 461–468 463

for selection of included studies.

(ap

3

11ts

3

oT

Fig. 1. PRISMA flow diagram

Table 2). Six studies reported femoral component rotation asn outcome (Table 3). In all but three studies, MRI was thereoperative imaging modality.

.2. Level of evidence and Detsky Quality Assessment scores

In 15 studies, the level of evidence was either III or IV (Appendix). Eleven studies were demonstrably stronger (Level I–II). Of the5 included comparative studies, eight had a Detskey score ≥ 16herefore classified as of high methodological quality. The Detskeycore mean was 15.6 (Appendix 2).

.3. Publication bias

Fig. 2 illustrates an asymmetry to the funnel plot for the primaryutcome (proportion of three-degree outliers from HTA measure).his suggests evidence of small sample size publication bias.

Fig. 2. Funnel plot for studies that included number of 3-degree outliers as primaryoutcome measure. SE(log[OR]): standard error (log [odds ratio]); OR: odds ratio.

464 A. Mannan et al. / Orthopaedics & Traumatology: Surgery & Research 101 (2015) 461–468

Table 1Patient-specific knee instrumentation studies: coronal plane alignment outcomes.

Study Location/centre Implant Preoperativeimaging

TotalPSI

HKA mean (range) HKA (outliers)%

ZMA: (M zone%)

FFC: mean(range)

FTC: mean(range)

Howell et al.2008 [30]

Sacramento,USA

Vanguard,Biomet

MRI 48 178.6 ± 2.8 – – – –

DaniilidisandTibesku,2013 [31]

Hannover,Germany

Genesis II,Smith &Nephew

MRI 124 178.5 ± 1.7(174.0–183.4)

11% 75% – –

Lustig et al.,2013 [28]

Sydney,Australia

VISIONAIRE,Smith &Nephew

MRI 60 180.6 ± 2.9(173.0–189.5)

20.7% – 89.9 (±1.8) 90.6 (±1.9)

Nunley et al.,2012 [32]

Washington,USA

Signature/OtisMed,Smith &Nephew

MRI 100 Signature:179.35 ± −1.35 to0.04OtisMed:−177.24 ± −3.46 to−2.06

Signature: 18%OtisMed: 44%

Signature; C:60%OtisMed; C:36%

90.7 89.9

Ng et al.,2012 [33]

Ohio, USA Signature,Biomet

MRI 105 180.6 9% C: 88% – –

Boonen et al.,2012 [34]

Sittard-Geleen,Netherlands

Vanguard,Biomet

MRI 40 181 (171–188) 29% – 90 (84–93) 91 (87–96)

Spencer et al.,2009 [12]

PA, USA OtisMed,Hayward

MRI 21 178.8 ± (−4 to +6) 9.5% – 91 (88–94) 87.1 (86–96)

Barrack et al.,2012 [35]

Missouri, USA Vanguard,Biomet

MRI 100 178.3 (±2.5) 31% 57% – –

Lombardiet al., 2008[36]

Ohio, USA Vanguard,Biomet

MRI 40 – 100% (4–8degrees valgus)

– –

Conteducaet al., 2012[27]

Rome, Italy VISIONAIRE,Smith &Nephew

MRI 15 – – – 91.2 (±0.6) 91.2 (±1.2)

Heyse et al.,2012 [37]

Marburg,Germany

Genesis II,Smith &Nephew

MRI 46 – – – – –

Bali et al.,2012 [38]

Victoria,Australia

Genesis II,Smith &Nephew

MRI 32 179.9 (±2) 12.5% – – –

Victor et al.,2013 [39]

Ghent, Belgium Signature,Biomet

MRI 61 179.4 24.6% – 89.9 (85.5–95) 89.7 (83–95)

Daniilidiset al., 2013[40]

Hannover,Germany

Genesis II,Smith &Nephew

MRI 170 178.4 9.3% – – –

Roh et al.,2013 [41]

Seoul, SouthKorea

Signature,Biomet

CT 42 179.5 12% – Varus 1.0(±1.4)

Valgus 0.2(±1.4)

Boonen et al.,2013 [42]

Netherlands Vanguard,Biomet

MRI 86 179 30% – 89 (82–93) 90 (84–94)

Chareancholvicet al., 2013[43]

Bangkok,Thailand

Nexgen,Zimmer

MRI 40 179.7 2.5% – 90.1 (87–93) 89.8 (87–93)

Paratte et al.,2013 [44]

Marseille,France

Nexgen,Zimmer

MRI 20 179 20% – 90.1 (84–93) 89.1 (85–96)

Noble et al.,2013 [45]

Louisiana, USA VISIONAIRE,Smith &Nephew

MRI 15 181.7 (180–186) – – – –

Chen et al.,2013 [46]

Singapore Nexgen,Zimmer

MRI 29 179.2 (±3.4) 31% – 89.9 (±2.1) 89.8 (±1.9)

Barrett et al.,2013 [47]

Seattle, USA PFC, Depuy CT 66 182.23 (±1.9) 19% – 89.8 (±1.3) 90.0 (±1.1)

Macdessiet al., 2013[48]

Sydney,Australia

Nexgen,Zimmer

MRI 115 179.9 (−5 to +7) 8.7% – – –

Yaffe et al.,2013 [49]

Chicago, USA Nexgen,Zimmer

– 44 180.98 (±2.3) 22.7% – – –

Koch et al., Zurich, GMK, CT 301 180.1 (±2.0) 12.4% – – –

±2.9)

3

dttF

2013 [50] Switzerland MyKneeNam et al.,

2013 [51]New York, USA Signature,

BiometMRI 41 180.8 (

.4. Clinical outcomes

Fifteen studies compared alignment outcomes between stan-

ard instrumentation and PSI knees (Table 1). The ‘OtisMed’ arm ofhe study by Nunley et al. [32] was excluded, as it is a kinematic sys-em that operates upon an anatomical and not mechanical model.or purposes of meta-analysis calculations, Noble and Vundelinckx

29.3% – 89.9 (±1.5) 89.6 (±1.6)

et al.’s series [45,52] were excluded as the proportion of three-degree outliers from a neutral postoperative axis was not listed.

3.5. Primary outcome measure

In total, from 15 studies reporting a total of 1847 knees (PSI andstandard instrumentation), a postoperative HKA malalignment of

A. Mannan et al. / Orthopaedics & Traumatology: Surgery & Research 101 (2015) 461–468 465

Table 2Patient-specific knee instrumentation studies: sagittal plane alignment outcomes.

Study Mean sagittal LFC (range) 3-degree outliers Mean sagittal LTC (range) 3-degree outliers

Conteduca et al., 2012 [27] 86.3 ± 2.0 – 93.8 ± 2.4 –Boonen et al., 2012 [34] 85 (74–94) 41% 94 (87–102) 40.00%Lustig et al., 2012 [28] 87.9 (±2.8) 34.6% 90.1 (±2.6) 19.3%Roh et al., 2013 [41] 86.8 (±2.3) 10% 93 (±2.0) 4%Paratte et al., 2013 [44] 81.85 (76–88) – 84.1 (91–97) –Boonen et al., 2013 [42] 96 (81–112) 49% 92 (86–101) 33%Victor et al., 2013 [39] 86.8 (±2.3) 52.5% 93 (±2.0) 21.3%Chen et al., 2013 [46] 87.7 (±2.6) 24% 84.2 (±3.4) 24%Yaffe et al., 2013 [49] 88.6 (±4.8) 87.2 (±5.0) –Macdessi et al., 2013 [48] – 5.2% 83 8.7%Koch et al., 2013 [50] – 9.0% – 12.3%

Table 3Patient-specific knee instrumentation studies: femoral rotation alignment outcomes.

Study Mean femoral rotationconventional

3-degreeoutliers

Mean femoralrotation PSI

3-degreeoutliers

Accuracy conclusion

Heyse and Tibesku, 2012 [37] 2.6 (±2.6) 11/48 (22.9%) 2.1 (±1.5) 1/46 (2.2%) PSI more accurateLustig et al., 2013 [28] 0.6 (±2.5) 10/45 (22.6%) –Magdessi et al., 2013 [48] 7/92 (8.7%) 11/115 (8.7%) No significant difference

ligFrotiF

3

3

hoTt

F

Paratte et al., 2013 [44] 0.2 0/20 (0%)

Roh et al., 2013 [41] 6/48 (13%)

Victor et al., 2013 [39] 1.6 11/64 (17%)

ess than three degrees occurred in 20.4% (184/901) of the standardnstrumentation group and 18.4% (175/946) of the patient-specificroup with a meta-analysis odds ratio of 0.90 (95% CI: 0.72–1.14;ig. 3). The results did not show significantly increased accu-acy with PSI based on an incidence of less than three degreesf malalignment (P > 0.05). The sensitivity analysis, where onlyhose studies with a Detskey score greater or equal to 16 werencluded, reported a similar finding (OR: 1.03; 95% CI: 0.72–1.49;ig. 4).

.6. Secondary outcome measures

.6.1. Coronal plane alignment: HKA/knee/ankle angle (FKA)In total, 16 studies reported the HKA. Overall, 17% of patients

ad deviation in the mechanical axis greater than three degreesf neutral. Range of mean alignment was 178.3 to 181.7 degrees.here was a wide (mean) range of three-degree outliers from 2.5o 31%.

ig. 3. Relative risk of producing deviation greater than 3 degrees from neutral in mechani

0.4 0/20 (0%) No significant difference4/42 (10%) No significant difference

0.5 15/64 (23%) PSI more accurate

3.6.2. Coronal plane alignment: zone of mechanical alignment(ZMA)

Four studies examined the ZMA as an alignment outcome in atotal of 429 patients. Central zone alignment was achieved in 36 to88% (mean 71.9%) of knees.

Coronal plane of alignment: frontal femoral component (FFC)angle and frontal tibial component (FTC) angles. Thirteen studiesreported upon deviation of femoral and tibial components fromthe mechanical axis in the coronal plane in a total of 484 knees(Table 4). In these studies, range of mean alignment was: FFC (89.9to 91.3 degrees) and FTC (87.1 to 91 degrees).

3.6.3. Femoral rotationSix studies documented femoral implant rotation as an out-

come measure in 342 knees with a 12% proportion of three-degreeoutliers. Five studies compared PSI to conventional instrumenta-tion [37,39,41,44,54], which showed 12.9% three-degree outlierson postoperative CT scan (Table 3). These two studies showed PSI

cal axis; patient-specific vs. conventional instrumentation (all comparative studies).

466 A. Mannan et al. / Orthopaedics & Traumatology: Surgery & Research 101 (2015) 461–468

F echa

tol

3

nmd5

4

sPdIbmamsoaltfaa

TS

ig. 4. Relative risk of producing deviation greater than 3 degrees from neutral in m

o be significantly more accurate [37,44]. Subgroup meta-analysisf quality studies (Detskey score ≥ 16) was not possible due to theow number of qualifiable studies [39,41,44].

.6.4. Sagittal plane alignment: LTC/LFCSagittal alignment of both femoral (LFC) and tibial (LTC) compo-

ents was described in 11 studies (Table 2). In these series, range ofean alignment was: LFC (81.85 to 96 degrees) and LTC (83 to 94

egrees). The proportion of three-degree outliers ranged between.2 and 52.5% for LFC and 4 to 40% for LTC.

. Discussion

Findings of this systematic review and meta-analysis demon-trate a lack of increased accuracy in coronal alignment outcomes inSI total knee arthroplasty, demonstrated by the number of three-egree outliers from a neutral HKA angle in the coronal plane.

ncluded studies were of all demonstrably high-quality as definedy the Detskey scoring system however with a variable level ofethodological rigor, in particular with regards to randomisation

nd allocation concealment. There was variability in the method ofeasurement of postoperative alignment i.e. a combination of CT

canograms, long leg radiographs, and computer-assisted technol-gy. Within the literature, disparity has previously been shown inccuracy of mechanical axis measurements when comparing longeg radiographs and coronal scout CT scans [55]. Recommenda-

ions should therefore be viewed with consideration. The outcomesrom this study are supported by findings from a recent meta-nalysis showing no increased risk of malalignment in mechanicalxis with PSI versus conventional TKR. However, eligibility criteria

able 4tudies comparing standard instrumentation vs. PSI: HKA angle outliers.

Study Standard: HKA mean (outliers)

Barrack et al., 2012 [53] 179.5 (23%)

Nunley et al., 2012 [32] (Signature) 180.08 (16%)

Paratte et al., 2013 [44] 178.3 (10%)

Boonen et al., 2012 [34] 179 (46%)

Boonen et al., 2013 [42] 178 (18%)

Victor et al., 2013 [39] 179.1 (28.1%)

Roh et al., 2013 [41] 179.7 (10%)

Chareancholvanich et al., 2013 [43] 179.7 (7.5%)

Ng et al., 2012 [33] 181.1 (22%)

Daniilidis and Tibesku, 2013 [31] 178.7 (21.2%)

Noble et al., 2013 [45] 181.7 (180–186)

Chen et al., 2013 [46] 180.4 (10%)

Barrett et al., 2013 [47] 182.09 (23%)

Yaffe et al., 2013 [49] 179.76 (32.5%)

Magdessi et al., 2013 [48] 179.4 (19.6%)

nical axis; patient-specific vs. conventional instrumentation (Detskey scale = 16).

and methodological scoring differed when compared to our study[54].

Implant malalignment has been directly associated with earlyfailure as evidenced from previous studies [14]; a varus or valgusmalalignment specifically predicting an increased risk of failure[56]. Jeffrey et al. documented rates of aseptic loosening in implantswith less than ± three degrees malalignment at 24% as compared tojust 3% in TKAs falling within the ± three-degree range [57]. How-ever, restoration of the mechanical axis has not clearly been shownto predict a more positive functional outcome [58], with regardto the computer-navigated model. In fact, computer-assisted navi-gation improves precision and accuracy of alignment over manualinstrumentation [59], but disadvantages include: lengthy landmarkregistration, longer operative duration, greater cost, risk of pin loos-ening and a significant learning curve. With specific regard to varusmalalignment, Collier et al. [60] demonstrated three factors predic-tive of increased medial compartment wear: shelf age of the liner,age of the patient and postoperative varus alignment on HKA mea-surement. It has more recently been postulated that three degreesis an arbitrary figure and that any deviation from a neutral axis ofalignment will reduce longevity by a degree proportional to themalalignment.

Within the small number of comparative studies examiningaccuracy of femoral implant rotation, subgroup meta-analysis ofconventional versus patient-specific instrumentation was not fea-sible given the current paucity of quality studies reporting upon thisoutcome. The role of patient-specific instrumentation in achiev-

ing accurate rotational placement of implants is further verifiedby findings that computer navigation has limited control on axialpositioning of implants [61,62]. To date, computer-assisted naviga-tion has been shown to be effective in reducing outliers in both the

PSI: HKA mean (outliers) P value Accuracy conclusion

178.3 (31%) 0.203 No significant difference179.35 (18%) 1.0 No significant difference

179 (20%) n.s No significant difference181 (29%) 0.1 No significant difference179 (18%) n.s No significant difference

179.6 (24.6%) 0.69 No significant difference179.5 (12%) 0.542 No significant difference179.7 (2.5%) 0.615 No significant difference180.6 (9%) 0.018 PSI more accurate178.4 (9.3%) 0.0031 PSI more accurate182.8 (180–185) 0.03 PSI more accurate179.2 (31%) 0.045 PSI less accurate

182.23 (19%) 0.646 No significant difference180.98 (22.7%) 0.164 No significant difference

179.9 (8.7%) 0.039 PSI more accurate

tology

catoi

acd

5

wsrs

D

i

A

wd

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

A. Mannan et al. / Orthopaedics & Trauma

oronal and sagittal planes [33], but has not been shown to increaseccuracy of rotational alignment [63–65]. It has been proposedhat reproducible and accurate registration of bony landmarks forptimal rotational alignment is more difficult to achieve with nav-gation [66].

Malposition of tibial and femoral implants in the axial plane isn important predictor of postoperative pain and patello-femoralomplications [53,67]. Further determination of the role of PSI inetermining rotational accuracy is therefore advocated.

. Conclusion

This systematic review describes alignment outcomes in PSIithin the current literature. Meta-analysis confirms no demon-

trable overall benefit in coronal plane outcomes with particularegard to the HKA angle (three-degree outliers). Further qualitytudies are required to determine rotational alignment outcomes.

isclosure of interest

A.M., T.S., C.S. and P.J.A.M. declare that they have no conflicts ofnterest concerning this article.

N.L.: consultant for Zimmer and received personal fees.

ppendix A. Supplementary data

Supplementary data (Appendixes 1 and 2) associatedith this article can be found, in the online version, atoi:10.1016/j.otsr.2014.12.018.

eferences

[1] Green GV, Berend KR, Berend ME, Glisson RR, Vail TP. The effects of varus tibialalignment on proximal tibial surface strain in total knee arthroplasty: the postermedial hot spot. J Arthroplasty 2002;17:1033–9.

[2] Liau JJ, Cheng CK, Huang CH, Lo WH. The effect of malalignment on stressesin polyethylene component of total knee prostheses: a finite element analysis.Clin Biomech (Bristol Avon) 2002;17:140–6.

[3] Choong PF, Dowsey MM, Stoney JD. Does accurate anatomical alignment resultin better function and quality of life? Comparing conventional and computerassisted total knee arthroplasty. J Arthroplasty 2009;24:560–9.

[4] Rand JA, Trousdale RT, Ilstrup DM, Harmsen WS. Factors affecting the durabilityof primary total knee prostheses. J Bone Joint Surg 2003;85:259–65.

[5] Sierra RJ, Cooney WP, Pagnano MW, Trousdale RT, Rand JA. Reoperations after3200 revision TKAs: rates, etiology and lessons learned. Clin Orthop Relat Res2004;425:200–6.

[6] Werner FW, Ayers DC, Maletsky LP, Rullkoetter PJ. The effect of valgus/varusmalalignment on load distribution in total knee replacements. J Biomech2000;38:349–55.

[7] Huang NF, Dowsey MM, Ee E, Stoney JD, Babazadeh S, Choong PF. Coronal align-ment correlates with outcome after total knee arthroplasty: five-year follow-upof a randomised controlled trial. J Arthroplasty 2012;27:1737–41.

[8] Hernandez-Vaquero D, Suarez-Vasquez A, Iglesias-Fernandez S. Can computerassistance improve the clinical and functional scores in total knee arthroplasty?Clin Orthop Relat Res 2011;469:3436–42.

[9] Chauhan SK, Scott RG, Breidahl W, Beaver RJ. Computer assisted knee arthro-plasty versus a conventional jig-based technique: a randomised prospectivetrial. J Bone Joint Surg 2004;86:372–7.

10] Laskin RS. Instrumentation pitfalls: you just can’t go on autopilot. J Arthroplasty2003;18(D):18–22.

11] Hetaimish BM, Khan MM, Simunovic N, Al-Habri HH, Bhandhari M, Zalzal P.Meta-analysis of navigation vs. conventional total knee arthroplasty. J Arthro-plasty 2012;27:1177–82.

12] Spencer JM, Chauhan SK, Sloan K, Taylor A, Beaver RJ. Computer navigationversus conventional total knee replacement: no difference in functional resultsat two years. J Bone Joint Surg Br 2007;89:477–80.

13] Michaut M, Beaufils P, Galaud B, Abadie P, Boisrenoult P, Fallet L. Rotationalalignment of femoral component with computed-assisted surgery (CAS) duringtotal knee arthroplasty. Rev Chir Orthop Reparatrice Appar Mot 2008;94:580–4.

14] Lewallen DG, Bryan RS, Peterson LF. Polycentric total knee arthroplasty. A ten-year follow-up study. J Bone Joint Surg Am 1984;66:1211–8.

15] Detsky AS, Naylor CD, O’Rourke K, McGeer AJ, L’Abbé KA. Incorporating vari-ations in the quality of individual randomized trials into meta-analysis. J ClinEpidemiol 1992;45:255–65.

[

: Surgery & Research 101 (2015) 461–468 467

16] Hafez MA, Cheulule KL, Seedhom BB, Sherman KP. Computer-assisted totalknee arthroplasty using patient-specific templating. Clin Orthop Relat Res2006;444:181–4.

17] Gan Y, Xu D, Lu S, Ding J. Novel patient-specific navigational template for totalknee arthroplasty. Comput Aided Surg 2011;16:288–97.

18] Howell SM, Howell SJ, Kuznik KT, Cohen J, Hull ML. Does a kinematically alignedtotal knee arthroplasty restore function without failure regardless of alignmentcategory? Clin Orthop Relat Res 2013;471:1000–7.

19] Dossett HG, Swartz GJ, Estrada NA, LeFevre GW, Kwasman BG. Kine-matically versus mechanically aligned total knee arthroplasty. Orthopedics2012;35:e160–9.

20] Howell SM, Papadopoulos S, Kuznik KT, Hull ML. Accurate alignment and highfunction after kinematically aligned TKA performed with generic instruments.Knee Surg Sports Traumatol Arthrosc 2013;21:2271–80.

21] Koeck FX, Beckmann J, Luring C, Rath B, Grifka J, Basad E. Evaluation of implantposition and knee alignment after patient-specific unicompartmental kneearthroplasty. Knee 2011;18:294–9.

22] Volpi P, Prospero E, Bait C, Cervellin M, Quaglia A, Redaelli A, et al. High accuracyin knee alignment and implant placement in unicompartmental medial kneereplacement when using patient-specific instrumentation. Knee Surg SportsTraumatol Arthrosc 2013, http://dx.doi.org/10.1007/s00167-013-2794-3.

23] Klatt BA, Goyal N, Austin MS, Hozack WJ. Custom-fit total knee arthroplasty(OtisKnee) results in malalignment. J Arthroplasty 2008;23:26–9.

24] Chan KY, Teo YH. Patient-specific instrumentation for total knee replacementverified by computer navigation: a case report. J Orthop Surg (Hong Kong)2012;20:111–4.

25] Mayer SW, Hug KT, Hansen BJ, Bolognesi MP. Total knee arthroplastyin osteopetrosis using patient-specific instrumentation. J Arthroplasty2012;27:1580 [e1–4].

26] Issa K, Rifai A, McGrath MS, Callaghan JJ, Wright C, Malkani AL, et al. Reliabilityof templating with patient-specific instrumentation in total knee arthroplasty.J Knee Surg 2013;26:429–33.

27] Conteduca F, Iorio R, Mazza D, Caperna L, Bolle G, Argento G, et al. Evaluationof the accuracy of a patient-specific instrumentation by navigation. Knee SurgSports Traumatol Arthrosc 2013;21:2194–9.

28] Lustig S, Scholes CJ, Oussedik SI, Kinzel V, Coolican MR, Parker DA. Unsat-isfactory accuracy as determined by computer navigation of VISIONAIREpatient-specific instrumentation for total knee arthroplasty. J Arthroplasty2013;28:469–73.

29] Thienpont E, Paternostre F, Pietsch M, Hafez M, Howell S. Total knee arthro-plasty with patient-specific instruments improves function and restores limbalignment in patients with extra-articular deformity. Knee 2013;20:407–11.

30] Howell SM, Kuznik K, Hull ML, Siston R. Results of an initial experiencewith custom-fit positioning total knee arthroplasty in a series of 48 patients.Orthopaedics 2008;31:857–63.

31] Daniilidis K, Tibesku C. Frontal plane alignment after total knee arthroplastyusing patient-specific instruments. Int Orthop 2013;37:45–50.

32] Nunley RM, Zhu J, Howell SM. Do patient-specific guides improve coronal align-ment in total knee arthroplasty? Clin Orthop Relat Res 2012;470:895–902.

33] Ng VY, DeClaire JH, Berend KR, Gulick BC, Lombardi Jr AV. Improved accuracyof alignment with patient-specific positioning guides compared with manualinstrumentation in TKA. Clin Orthop Relat Res 2012;470:99–107.

34] Boonen B, Schotanus MG, Kort NP. Preliminary experience with the patient-specific templating total knee arthroplasty. Acta Orthop 2012;83:387–93.

35] Barrack RL, Ruh EL, Williams BM, Ford AD, Foreman K, Nunley RM. Patientspecific cutting blocks are currently of no proven value. J Bone Joint Surg Br2012;94(Suppl. A):95–9. [60].

36] Lombardi Jr AV, Berend KR, Adams JB. Patient-specific approach in total kneearthroplasty. Orthopedics 2008;31:927–30.

37] Heyse TJ, Tibesku CO. Improved femoral component rotation in TKA usingpatient-specific instrumentation. Knee 2012;21:268–71.

38] Bali K, Walker P, Bruce W. Custom-fit total knee arthroplasty: our initial expe-rience in 32 knees. J Arthroplasty 2012;27:1149–53.

39] Victor J, Dujardin J, Vandenneucker H, Arnout N, Bellemans J. Patient specificguides do not increase accuracy in total knee arthroplasty. A prospective ran-domised controlled trial. Clin Orthop Relat Res 2014;472:263–71.

40] Daniilidis K, Tibesku CO. A comparison of conventional and patient-specificinstruments in total knee arthroplasty. Int Orthop 2014;33:503–8.

41] Roh YW, Kim TW, Lee S, Seong SC, Lee MC. Is TKA using patient-specific instru-ments comparable to conventional TKA? A randomized controlled study of onesystem. Clin Orthop Relat Res 2013;471:3988–95.

42] Boonen B, Schotanus GM, Kerens B, van der Weegen W, van Drumpt RAM,Kort NP. Intra-operative results and radiological outcome of conventionaland patient-specific surgery in total knee arthroplasty: a multicentre, ran-domised controlled trial. Knee Surg Sports Traumatol Arthrosc 2013;21:2206–12.

43] Chareancholvanich K, Narkbunnam R, Pornrattanamaneewong C. A prospectiverandomised controlled study of patient-specific cutting guides compared withconventional instrumentation in total knee replacement. Bone Joint J 2013;95-B(3):354–9.

44] Parratte S, Blanc G, Boussemart T, Ollivier M, Le Corroller T, Argenson JN.Rotation in total knee arthroplasty: no difference between patient-specific

and conventional instrumentation. Knee Surg Sports Traumatol Arthrosc2013;21:2213–9.

45] Noble Jr JW, Moore CA, Liu N. The value of patient-matched instrumentation intotal knee arthroplasty. J Arthroplasty 2012;27:153–5.

4 tology

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

68 A. Mannan et al. / Orthopaedics & Trauma

46] Chen JY, Yeo SJ, Soon Yew AK, Tay DK, Chia SL, Lo NN, et al. The radiologicaloutcomes of patient-specific instrumentation versus conventional total kneearthroplasty. Knee Surg Sports Traumatol Arthrosc 2014;22:630–5.

47] Barrett W, Hoeffel D, Dalury D, Mason JB, Murphy J, Himden S. In vivo align-ment comparing patient specific instrumentation with both conventional andcomputer assisted surgery (CAS) instrumentation in total knee arthroplasty. JArthroplasty 2014;29:343–7.

48] Macdessi SJ, Jang B, Harris IA, Wheatley E, Bryant C, Chen DB. A comparison ofalignment using patient specific guides, computer navigation and conventionalinstrumentation in total knee arthroplasty. Knee 2014;21:406–9.

49] Yaffe M, Luo M, Goyal N, Chan P, Patel A, Cayo M, et al. Clinical, functional, andradiographic outcomes following total knee arthroplasty with patient-specificinstrumentation, computer-assisted surgery, and manual instrumentation: ashort-term follow-up study. Int J Comput Assist Radiol Surg 2014;9:837–44.

50] Koch PP, Müller D, Pisan M, Fucentese SF. Radiographic accuracy in TKA with aCT-based patient-specific cutting block technique. Knee Surg Sports TraumatolArthrosc 2013;21:2200–5.

51] Nam D, Maher PA, Rebolledo BJ, Nawabi DH, McLawhorn AS, Pearle AD. Patientspecific cutting guides versus an imageless, computer-assisted surgery systemin total knee arthroplasty. Knee 2013;20:263–7.

52] Vundelinckx BJ, Bruckers L, De Mulder K, De Schepper J, Van Esbroeck G. Func-tional and radiographic short-term outcome evaluation of the VISIONAIREsystem, a patient-matched instrumentation system for total knee arthroplasty.J Arthroplasty 2013;28:964–70.

53] Barrack RL, Schrader T, Bertot AJ, Wolfe MW, Myers L. Component rotationand anterior knee pain after total knee arthroplasty. Clin Orthop Relat Res2001;392:46–55.

54] Thienpont E, Schwab PE, Fennema P. A systematic review and meta-analysis ofpatient-specific instrumentation for improving alignment of the componentsin total knee replacement. Bone Joint J 2014;96:1052–61.

55] Burghardt RD, Hinterwinner S, Burklein D, Baumgart D. Lower limb alignment

in the frontal plane: analysis from long standing radiographs and computertomography scout views: an experimental study. Arch Orthop Trauma Surg2013;133:29–36.

56] Ritter MA, Faris PM, Keating EM, et al. Postoperative alignment of total kneereplacement. Its effect on survival. Clin Orthop Relat Res 1994;299:153–6.

[

: Surgery & Research 101 (2015) 461–468

57] Jeffery RS, Morris RW, Denham RA. Coronal alignment after total knee replace-ment. J Bone Joint Surg Br 1991;73:709–14.

58] Kim YH, Park JW, Kim JS. Computer navigated vs. conventional total kneearthroplasty: a prospective randomized controlled trial. J Bone Joint Surg Am2012;94:2017–24.

59] Mason JB, Fehring TK, Estok R, Banel D, Fahrbach K. Meta-analysis of alignmentoutcomes in computer-assisted total knee arthroplasty surgery. J Arthroplasty2007;22:1097–106.

60] Collier MB, Engh CA, McAuley JP, Engh GA. Factors associated with the loss ofthickness of polyethylene tibial bearings after total knee arthroplasty. J BoneJoint Surg Am 2007;89:1306–14.

61] Bedard M, Vince KG, Redfern J, Collen SR. Internal rotation of the tibial com-ponent is frequent in stiff total knee arthroplasty. Clin Orthop Related Res2011;469:2346–55.

62] Huddleston JI, Scott RD, Wimberley DW. Determination of neutral tib-ial rotation alignment in rotating platform TKA. Clin Orthop Related Res2005;440:101–6.

63] Matziolis G, Krocker D, Weiss U, Tohtz S, Perka C. A prospective, randomizedstudy of computer-assisted and conventional total knee arthroplasty. Three-dimensional evaluation of implant alignment and rotation. J Bone Joint SurgAm 2007;89:236–43.

64] Lutzner J, Krummenauer F, Wolf C, Gunther KP, Kirschner S. Computer-assistedand conventional total knee replacement: a comparative, prospective, ran-domised study with radiological and CT evaluation. J Bone Joint Surg Br2008;90:1039–44.

65] Schmitt J, Hauk C, Kienapfel H, Pfeiffer M, Efe T, Fuchs-Winkelmann S, et al. Nav-igation of total knee arthroplasty: rotation of components and clinical results ina prospectively randomized study. BMC Musculoskelet Disord 2011;15:12–6.

66] Jerosch J, Peuker E, Philipps B, Filler T. Interindividual reproducibility in peri-operative rotational alignment of femoral components in knee prostheticsurgery using the transepicondylar axis. Knee Surg Sports Traumatol Arthrosc

2002;10:194–7.

67] Matsuda S, Miura H, Nagamine R, Urabe K, Hirata G, Iwamoto Y. Effect offemoral and tibial component position on patellar tracking following totalknee arthroplasty: 10-year follow-up of Miller-Galante I knees. Am J Knee Surg2001;14:152–6.