The Knee 21 (2014) 406–409
Contents lists available at ScienceDirect
The Knee
A comparison of alignment using patient specific guides, computer navigation and conventional instrumentation in total knee arthroplasty Samuel J. MacDessi a,⁎, Bob Jang a, Ian A. Harris b, Emma Wheatley c, Carl Bryant c, Darren B. Chen a a b c
Sydney Knee Specialists, St George Private Hospital, Sydney, NSW, Australia South Western Sydney Clinical School, University of New South Wales, Australia Bryant Radiology, St George Private Hospital, Sydney, NSW, Australia
a r t i c l e
i n f o
Article history: Received 25 July 2013 Received in revised form 23 September 2013 Accepted 6 November 2013 Keywords: Patient specific guides Alignment Conventional
a b s t r a c t Background: Patient specific guides (PSG) have been introduced as a tool in total knee arthroplasty (TKA) in an attempt to improve limb alignment and reduce operative time compared to other established surgical techniques. The purpose of this study was to compare the post-operative radiographic alignment and operative time in patients who underwent TKA surgery with PSG, conventional instrumentation or computer-assisted navigation surgery using fully cemented components. Methods: A cohort of 260 patients who underwent TKA surgery using PSG (PSG group, n = 115) was compared to patients who underwent TKA using either conventional instrumentation (CON group, n = 92) or computerassisted navigation (CAS group, n = 53). Post-operative CT imaging using the Perth CT protocol was used to compare alignment between the three groups. Results: In the PSG and CAS groups, the post-operative hip–knee angle (HKA) was within 3° of neutral alignment in 91.3% and 90.7% of patients, respectively. This compared to 80.4% of patients in the CON group (p = 0.02). There were no significant differences with respect to alignment when comparing individual component positioning between the PSG and CAS groups apart from tibial slope (Table 3). Total operative time was found to be significantly reduced in the PSG group (80.2 min) compared to both the CON group (86 min, p = 0.002) and the CAS group (110.2 min, p b 0.0001). Conclusions: The use of PSG resulted in similar alignment accuracy to CAS and superior alignment to CON with significantly shorter operative times. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Patient-specific guides (PSG) derived from pre-operative imaging have been recently introduced as a new strategy for improving component positioning in total knee arthroplasty (TKA). The other cited benefits of these instruments when compared to computer navigation are reduced operative times and inventory in the operating room; both of which may result in cost savings [1–3]. Most major orthopaedic implant manufacturers have introduced different versions of patient-specific guides. These guides rely on either magnetic resonance imaging (MRI), computed tomography (CT) scans with or without long alignment radiographs to produce moulds that conform to the patient's anatomy during surgery, using pre-defined radiographic parameters to restore a neutral mechanical alignment to the lower limb. Despite these guides being available for several years, there is a paucity of data on their accuracy. The few studies available have only looked
at plain radiographic outcomes of coronal alignment, with mixed results [4–7]. This is in contrast to the literature evaluating the accuracy of computer-assisted navigation (CAS) surgery, with significant evidence validating CAS as being more accurate than conventional instrumented (CON) techniques by reducing the number of alignment outliers [8–11]. Like other studies on this topic, only coronal plane alignment had been assessed. In addition, no published data exists on comparing the accuracy of patient-specific guides against both conventional instrumentation and computer-assisted navigation surgery. The aim of this paper is to compare the differences in alignment accuracy in patients who have undergone TKA using patient specific guides, conventional instrumentation, or computer-assisted navigation to see which method provides the best alignment accuracy. This is to be determined by examining the post-operative coronal, sagittal and axial alignment and comparing the alignment outliers between the groups. 2. Materials and methods
⁎ Corresponding author at: Sydney Knee Specialists, Suite 8, 19 Kensington St, Kogarah NSW 2217, Australia. Tel.: +61 2 8307 0333; fax: +61 2 8307 0334. E-mail address: [email protected] (S.J. MacDessi). 0968-0160/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.knee.2013.11.004
A cohort of patients who underwent TKA surgery using PSG (PSG group) was compared to patients who underwent TKA using either conventional instrumentation (CON group) or computer-assisted
S.J. MacDessi et al. / The Knee 21 (2014) 406–409
navigation (CAS group). Pre-operative hip–knee angle (HKA) was measured using 4-foot long standing radiographs for all patients. The HKA is the angle formed by the intersection of a line connecting the centres of the hip and knee joint and another line connecting the centres of the knee and ankle joint. Post-operative CT imaging using the Perth CT protocol [12] was used to compare alignment between the three groups.
407
coronal, sagittal and axial planes as well as comparing implant rotation. All scans were analysed by the chief CT radiographer at our institution. A second research assistant reviewed a subset of 20 scans. The mean absolute difference in angle measurements (pooling all angles measured) was 1.1° (SD = 1.1). The difference between the observers was 2° or less in 91.0% of cases. In the literature it has been noted that intraobserver error averages 2° on any given TKA radiograph [13].
2.1. Study group 2.4. Outcome measures The PSG group comprised of 115 patients who underwent TKA surgery during a consecutive time period from April 2010 to February 2012. Patients who were not suitable to have a PSG TKA underwent surgery using conventional instrumentation during the same period as those patients who underwent a PSG TKA. There were 92 patients in this group who had a CON TKA. Reasons for exclusion from undergoing a PSG TKA included inability to undergo an MR scan because of cardiac and cerebral metal implants and patient factors such as cost and short time to surgery. This group also included patients who failed to have a PSG manufactured because of decreased image quality of their MR scan due to motion artefact. The commonest reason for this was claustrophobia as a cause of movement. There were no cases where knee deformity or pathology resulted in failure to produce a PSG. The final cohort comprised 53 patients who had CAS TKA using the Orthosoft Navigation System in 2009 prior to commencement of use of PSI guides. These patients voluntarily agreed to undergo CT radiographic assessment to act as a comparison group and represented a subgroup of 98 patients who had a CAS TKA in the same year. 2.2. Surgical technique All operations were performed using an identical surgical technique by two orthopaedic knee surgeons. A medial parapatellar approach was used and a thigh tourniquet was inflated for the initial surgical exposure only. The one prosthesis was used in all cases, the Zimmer NexGen LPS Flex (Zimmer, Warsaw, IN) using fully cemented implants. The patella was resurfaced in all knees. All operations were performed in an identical manner using a measured resection technique. All three techniques aimed for restoration of a neutral hip–knee angle, femoral component angle and tibial component angle in the coronal plane, a neutral femoral component flexion relative to the sagittal mechanical axis, seven degrees of posterior tibial slope and femoral component rotation parallel to the surgical femoral epicondylar axis. For those patients who underwent a PSG TKA, a proprietary internetbased software planner was used to review, adjust and approve the surgical plan. Once registered, the approved plan was forwarded to Materialise (Leuven, Belgium) who used rapid-prototyping technology to create the customised patient specific guides. The patient specific guides used in this study rely on data acquisition from a pre-operative MRI scan to create a custom mould of the knee that conforms to the patient’s knee anatomy at the time of surgery. The distal femoral resection was performed first in all cases, followed by the proximal tibial resection. An extra-medullary tibial guide attached to a 10 mm spacer block was used to ensure creation of an adequate extension gap as well as a tibial resection perpendicular to the tibial mechanical axis. The remainder of the femoral preparation was then carried out. In the conventional instrumentation group, intramedullary femoral and tibial alignment guides were used to restore coronal and sagittal knee alignment. 2.3. Radiographic analysis All patients in this series underwent a radiographic analysis using the Perth CT Protocol [12]. The Perth CT Protocol is a technique whereby patients are CT scanned non weight bearing to allow a direct measurement of the alignment of the femoral and tibial components in the
The post-operative parameters that were analysed were hip–knee angle, individual component alignment in the coronal and sagittal plane, and axial rotation of the femoral component. The proportion of patients within 3° of the neutral mechanical alignment for each measurement was recorded. Operative times were recorded in minutes and defined as the time from the initial skin incision until final wound closure. 2.5. Statistical analysis Means, ranges, and standard deviations were recorded for all radiographic parameters. Student's t-tests and ANOVA were used to compare differences in means between continuous variables. ANOVA was used to statistically assess pre-operative demographic factors. The two-sample t-test was used to assess for differences in operative time between groups. Fisher's exact test was to test for significance in difference of alignment (proportions) between the groups. Statistical significance was set at p b 0.05. 3. Results There were no significant differences with regards to age, BMI, and pre-operative HKA between groups (Table 1). 3.1. Hip–knee coronal angle When assessing HKA, 91.3% and 90.7% of patients in both the PSG and CAS groups, respectively, were within 3° of a neutral mechanical alignment. This compared to 80.4% of patients in CON group, which was statistically significant (p = 0.039, PSG vs CON). See Table 2 for a summary of results. 3.2. Individual component alignment 97.4% and 96.3% of patients in the PSG and CAS group, respectively, had a femoral coronal alignment within 3° of a neutral mechanical axis. This compared to 89.1% of patients in the CON group and this difference was statistically significant (p = 0.02). In the PSG group, 94.8% were within 3° of neutral femoral sagittal alignment. This compared to 98.2% and 90.2% of cases in the CAS and CON groups respectively. The differences between the CAS and CON groups compared to the PSG group were not found to be significant (p = 0.178 and p = 0.282, respectively). Tibial coronal alignment was within 3° of the neutral mechanical axis in 97.4% of patients in the PSG group and to 100.0% in the CAS group (p = 0.552). In the CON group, 88.0% were within an acceptable range and this was statistically significant when compared to PSG (p = 0.011). Mean posterior tibial slope was found to be significantly decreased in the CAS group (Mean = 4.9°, SD = 1.7) compared to the CON (Mean = 6.9°, SD = 2.0) and PSG group (Mean = 7.0°, SD = 1.9, p = 0.009). The reduced tibial slope in the CAS group
Table 1 Group demographics.
Age BMI Pre-op HKA Range HKA a b c
PSGa (n = 115)
CONb (n = 92)
CASc (n = 53)
P-value
67 30.3 −3.3° −15° to +14°
66 31.5 −4.3° −12° to +12°
67 31.2 −3.3° −20° to +10°
0.7058 0.1718 0.4646 N/A
Patient Specific Instrumentation. Conventional Instrumentation. Computer navigated surgery.
408
S.J. MacDessi et al. / The Knee 21 (2014) 406–409
Table 2 Post-operative hip–knee angle.
Mean Range Cases within +/− 3° Standard Deviation
Table 4 Operative times. PSG (n = 115)
CON (n = 92)
CAS (n = 53)
−0.1° −5° to +7° 91.3% 2.2
−0.6° −9° to +7° 80.4% 2.9
0.1° −4° to +6° 90.7% 2.2
was believed to be due to intra-operative differences in referencing points in the sagittal plane and not due to significant variations in outliers. Mean femoral component rotation was within 3° of neutral alignment to the surgical epicondylar axis in 90.4%, 96.3% and 92.4% of cases in the PSG, CAS and CON groups respectively. The differences between the CAS and CON groups compared to the PSI group were not found to be significant (p = 0.230 and p = 0.810, respectively). Component alignments were compared across each group for outliers greater than 3°. These are shown in Table 3. 3.3. Operative times The PSG group had the shortest mean operative time of 80.2 min, compared to 86.0 min in the CON group and 110.2 min in the CAS group (Table 4). These differences were found to be significant between both groups compared to the PSG group (p b 0.05). 3.4. Effects of BMI on post-operative alignment A sub-group analysis of patients who had a BMI of more than 30 was undertaken to determine if there were differences in alignment when PSG was used. No statistical difference in BMI was noted for any of the recorded angles, when comparing those within 3° to those outside 3°.
3.5. Effects of pre-operative deformity on post-operative alignment using PSG Patients who underwent a PSG TKA were divided in to those who had a pre-operative HKA of greater than or equal to 6° and those who had lesser pre-operative deformities to determine of these guides were less effective for either group. No significant differences in the post-operative alignment were found in patients with HKA of 6° or greater, compared to those with HKA less than 6°.
4. Discussion Patient-specific guides have been introduced in total knee arthroplasty as a method to improve alignment whilst at the same time reducing operative times and technical issues associated with CAS TKA surgery. As PSG TKA uses pre-operative imaging to acquire registration points for component positioning, it has been hoped that more accuracy in definition of intra-operative landmarks such as the femoral epicondylar axis will allow greater accuracy in restoring alignment. Other proposed benefits include avoiding canal instrumentation
Table 3 Comparing groups for outliers N3°. PSG (n = 115) HKA FCA TCA FSA TSA FRA
8.7% (n 2.6% (n 2.6% (n 5.2% (n 8.7% (n 8.7% (n
= = = = = =
10) 3) 3) 6) 11) 11)
CON (n = 92)
NAV (n = 53)
P-value (PSI:CON)
P-value (PSI:CAS)
19.6% (n = 18) 10.9% (n = 10) 11.8% (n = 11) 9.8% (n = 9) 8.7% (n = 8) 8.7% (n = 7)
9.4% (n = 5) 3.8% (n = 2) 0.0% (n = 0) 0.0% (n = 0) 26.4% (n = 14) 3.8% (n = 2)
0.039⁎ 0.020⁎ 0.011⁎ 0.282 1.000 0.805
1.000 0.653 0.552 0.178 0.009⁎ 0.230
HKA: Hip–knee angle. FCA: Femoral coronal angle. TCA: Tibial coronal angle. FSA: Femoral sagittal angle (Femoral flexion/extension angle). TSA: Tibial sagittal angle (Tibial posterior slope angle). FRA: Femoral rotation angle. ⁎ Significant p-values.
Operative time
PSG (n = 115)
CON (n = 92)
NAV (n = 53)
P-value (PSI:CON)
P-value (PSI:CAS)
80.2 min
86.0 min
110.2 min
P = 0.002
P b 0.0001
that is required with CON TKA, and avoiding navigation pins used in CAS TKA. It is well recognised that computer navigation in TKA reduces alignment outliers when compared to conventional instrumentation [8–11]. However, computer navigation has not continued to grow in popularity as many surgeons feel that these benefits cannot be justified based on increased operative time, greater cost, extra instrumentation required in the operating room, and complications related to pin sites used for navigation trackers [9,14–16]. The majority of evidence still suggests that minimisation of alignment outliers in the coronal plane improves implant survival and possibly improves function [10,17]. To date, there are no published studies comparing the accuracy of patient-specific guides to both computer navigation and conventional instrumentation. In addition most studies have either been limited case series or have only compared patient-specific guides against conventional instrumentation [7,18]. There has also been variability in the methodology of assessing post-operative limb alignment. Some studies have only used plain radiography, which is known to be affected by rotation of the limb, whilst other studies have used a limited CT scanogram to assess overall limb alignment [1,7]. All published studies to date have only assessed coronal plane alignment [1,4–7]. Chareancholvanich et al., using Zimmer Patient Specific Instruments (PSI), reported similar hip–knee angles within 3° of neutral alignment in their CON and PSG groups. The authors concluded that the minimal advantage obtained in both alignment and operative times is unlikely to be of clinical importance [4]. Like the Zimmer PSI, the Biomet Signature system (Biomet, Warsaw, In) relies on MRI to create the cutting guides. Nunley et al. prospectively assessed three groups of patients who underwent surgery with conventional instrumentation, Biomet signature patient-specific instruments, and the Otismed system (Stryker, NJ, USA) [5]. There was no significant difference in alignment between the conventional and Biomet Signature group, with 84.0% and 82.0% of patients respectively having a hip–knee angle within 3° of neutral. This compared to only 56.0% of patients who underwent kinematic axis restoration using the Otismed system. Ng et al. compared the Biomet Signature guides to conventional instrumentation [7]. Long leg plain X-rays were used to assess postoperative HKA as well as independent coronal component positioning. Their findings were similar to our results with 91.0% of patients with a postoperative hip–knee angle within 3° compared to 78.0% in the CON group. Bali et al. evaluated an initial series of patients who underwent a TKA using the Visionaire custom-fit guides (Smith & Nephew, Memphis, TN) [6]. These guides rely on a combination of MRI to map joint surface topography as well as plain radiography to restore coronal alignment. Out of 32 patients, 29 had a postoperative hip–knee angle within 3° of the neutral mechanical axis based on full length X-rays only. This is the first study that has assessed alignment outcomes of PSG TKA and compared it to both CAS and CON TKA. It is also the first review to use a validated CT protocol to look at independent implant positioning in all three alignment planes. Our results indicate that PSG has similar accuracy to CAS surgery with 91.3% and 90.7% of patients in both groups, respectively, having a hip–knee angle within 3° of a neutral mechanical axis. Independent coronal component position was also found to be similar between the two groups with less outliers compared to conventional instrumentation. No differences were noted in femoral component sagittal alignment between all three groups.
S.J. MacDessi et al. / The Knee 21 (2014) 406–409
This study also found that the PSG group had a significantly shorter mean operative time when compared to both the CON and CAS groups. We believe however that the difference of six minutes in operative time between the PSG and CON group is not of clinical significance. This study has several limitations. Firstly, the study groups were not randomised, hence there was a potential for selection bias amongst the patients who underwent each different technique. Although demographics were similar between groups, such bias cannot be excluded. In addition, the CAS group was from an earlier time period, and may have only represented a cohort of patients who had a satisfactory postoperative outcome when compared to the remainder of the group who underwent surgery using this technique. In a recent meta-analysis by Hetaimish et al. comparing CAS to CON TKA, 87.0% of patients undergoing CAS had a hip–knee angle within 3° compared to 70.0% undergoing CON TKA [10]. In our series of patients, 91.0% of patients had a hip–knee angle within 3° in the CAS group and 80.0% in the CON group. Although it is important to highlight the potential limitations of this study, it appears that our results in terms of assessing CAS instrumentation are slightly more favourable than in the literature published on these techniques. In conclusion, we found that patient-specific guides more reliably restore the coronal alignment when compared to conventional instrumentation and have a similar accuracy when compared to computer navigation surgery. In addition operative times were found to be shorter in the PSG group, the surgery was found to be less complicated than computer navigation and canal instrumentation was avoided. 5. Conflict of interest The authors declare that there are no conflict of interest. References [1] Howell SM, Kuznik K, Hull ML, Siston RA. Results of an initial experience with custom-fit positioning total knee arthroplasty in a series of 48 patients. Orthopedics 2008;31:857.
409
[2] Slover JD, Rubash HE, Malchau H, Bosco JA. Cost-effectiveness analysis of custom total knee cutting blocks. J Arthroplasty 2012;27:180. [3] Nunley RM, Ellison BS, Ruh EL, Williams BM, Foreman K, Ford AD, et al. Are patientspecific cutting blocks cost-effective for total knee arthroplasty? Clin Orthop Relat Res 2012;470:889. [4] Chareancholvanich K, Narkbunnam R, Pornrattanamaneewong C. A prospective randomised controlled study of patient-specific cutting guides compared with conventional instrumentation in total knee replacement. J Bone Joint Surg Br 2013;95-B:354. [5] Ng V, DeClaire JH, Berend KR, Gulick BC, Lombardi Jr AV. Improved accuracy of alignment with patient-specific positioning guides compared with manual instrumentation in TKA. Clin Orthop Relat Res 2012;470:99. [6] Bali K, Walker P, Bruce W. Custom-fit total knee arthroplasty: our initial experience in 32 knees. J Arthroplasty 2012:27. [7] Nunley RM, Ellison BS, Zhu J, Ruh EL, Howell SM, Barrack RL. Do patient-specific guides improve coronal alignment in total knee arthroplasty? Clin Orthop Relat Res 2012;470:895. [8] Mason JB, Fehring TK, Estok R, Banel D, Fahrbach K. Meta-analysis of alignment outcomes in computer-assisted total knee arthroplasty surgery. J Arthroplasty 2007;22:1097. [9] Kim SJ, MacDonald M, Hernandez J, Wixon RL. Computer assisted navigation in total knee arthroplasty. J Arthroplasty 2005;20:123. [10] Hetaimish BM, Khan MM, Simunovic N, Al-Harbi HH, Bhandari M, Zalzal PK. Metaanalysis of navigation vs conventional total knee arthroplasty. J Arthroplasty 2002;27:1177. [11] Choong PF, Dowsey MM, Stoney JD. Does accurate anatomical alignment result in better function and quality of life? Comparing conventional and computer-assisted total knee arthroplasty. J Arthroplasty 2009;24:560. [12] Chauhan SK, Clark GW, Lloyd S, Scott RG, Breidahl W, Sikorski JM. Computer-assisted total knee replacement. A Controlled Cadaver Study using a Multi-Parameter Quantitative CT Assessment of Alignment (The Perth CT Protocol). J Bone Joint Surg Br 2004;86:818. [13] Lonner JH, Laird MT, Stuchin SA. Effect of rotation and knee flexion on radiographic alignment in total knee arthroplasties. Clin Orthop 1991;331:102. [14] Jenny JY, Miehlke RK, Giurea A. Learning curve in navigated total knee replacement. A multi-centre study comparing experienced and beginner centres. Knee 2008;15:80. [15] Li CH, Chen TH, Su YP, Shao PC, Lee KS, Chen WM. Periprosthetic femoral supracondylar fracture after total knee arthroplasty with navigation system. J Arthroplasty 2008;23:304. [16] Novak EJ, Silverstein MD, Bozic KJ. The cost-effectiveness of computer-assisted navigation in total knee arthroplasty. J Bone Joint Surg Am 2007;89:2389. [17] Klatt BA, Goyal N, Austin MS, Hozack WJ. Custom-fit total knee arthroplasty (OtisKnee) results in malalignment. J Arthroplasty 2008;23:26. [18] Lombardi Jr AV, Berend KR, Adams JB. Patient-specific approach in total knee arthroplasty. Orthopedics 2008;31:927.
Contents lists available at ScienceDirect
The Knee
A comparison of alignment using patient specific guides, computer navigation and conventional instrumentation in total knee arthroplasty Samuel J. MacDessi a,⁎, Bob Jang a, Ian A. Harris b, Emma Wheatley c, Carl Bryant c, Darren B. Chen a a b c
Sydney Knee Specialists, St George Private Hospital, Sydney, NSW, Australia South Western Sydney Clinical School, University of New South Wales, Australia Bryant Radiology, St George Private Hospital, Sydney, NSW, Australia
a r t i c l e
i n f o
Article history: Received 25 July 2013 Received in revised form 23 September 2013 Accepted 6 November 2013 Keywords: Patient specific guides Alignment Conventional
a b s t r a c t Background: Patient specific guides (PSG) have been introduced as a tool in total knee arthroplasty (TKA) in an attempt to improve limb alignment and reduce operative time compared to other established surgical techniques. The purpose of this study was to compare the post-operative radiographic alignment and operative time in patients who underwent TKA surgery with PSG, conventional instrumentation or computer-assisted navigation surgery using fully cemented components. Methods: A cohort of 260 patients who underwent TKA surgery using PSG (PSG group, n = 115) was compared to patients who underwent TKA using either conventional instrumentation (CON group, n = 92) or computerassisted navigation (CAS group, n = 53). Post-operative CT imaging using the Perth CT protocol was used to compare alignment between the three groups. Results: In the PSG and CAS groups, the post-operative hip–knee angle (HKA) was within 3° of neutral alignment in 91.3% and 90.7% of patients, respectively. This compared to 80.4% of patients in the CON group (p = 0.02). There were no significant differences with respect to alignment when comparing individual component positioning between the PSG and CAS groups apart from tibial slope (Table 3). Total operative time was found to be significantly reduced in the PSG group (80.2 min) compared to both the CON group (86 min, p = 0.002) and the CAS group (110.2 min, p b 0.0001). Conclusions: The use of PSG resulted in similar alignment accuracy to CAS and superior alignment to CON with significantly shorter operative times. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Patient-specific guides (PSG) derived from pre-operative imaging have been recently introduced as a new strategy for improving component positioning in total knee arthroplasty (TKA). The other cited benefits of these instruments when compared to computer navigation are reduced operative times and inventory in the operating room; both of which may result in cost savings [1–3]. Most major orthopaedic implant manufacturers have introduced different versions of patient-specific guides. These guides rely on either magnetic resonance imaging (MRI), computed tomography (CT) scans with or without long alignment radiographs to produce moulds that conform to the patient's anatomy during surgery, using pre-defined radiographic parameters to restore a neutral mechanical alignment to the lower limb. Despite these guides being available for several years, there is a paucity of data on their accuracy. The few studies available have only looked
at plain radiographic outcomes of coronal alignment, with mixed results [4–7]. This is in contrast to the literature evaluating the accuracy of computer-assisted navigation (CAS) surgery, with significant evidence validating CAS as being more accurate than conventional instrumented (CON) techniques by reducing the number of alignment outliers [8–11]. Like other studies on this topic, only coronal plane alignment had been assessed. In addition, no published data exists on comparing the accuracy of patient-specific guides against both conventional instrumentation and computer-assisted navigation surgery. The aim of this paper is to compare the differences in alignment accuracy in patients who have undergone TKA using patient specific guides, conventional instrumentation, or computer-assisted navigation to see which method provides the best alignment accuracy. This is to be determined by examining the post-operative coronal, sagittal and axial alignment and comparing the alignment outliers between the groups. 2. Materials and methods
⁎ Corresponding author at: Sydney Knee Specialists, Suite 8, 19 Kensington St, Kogarah NSW 2217, Australia. Tel.: +61 2 8307 0333; fax: +61 2 8307 0334. E-mail address: [email protected] (S.J. MacDessi). 0968-0160/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.knee.2013.11.004
A cohort of patients who underwent TKA surgery using PSG (PSG group) was compared to patients who underwent TKA using either conventional instrumentation (CON group) or computer-assisted
S.J. MacDessi et al. / The Knee 21 (2014) 406–409
navigation (CAS group). Pre-operative hip–knee angle (HKA) was measured using 4-foot long standing radiographs for all patients. The HKA is the angle formed by the intersection of a line connecting the centres of the hip and knee joint and another line connecting the centres of the knee and ankle joint. Post-operative CT imaging using the Perth CT protocol [12] was used to compare alignment between the three groups.
407
coronal, sagittal and axial planes as well as comparing implant rotation. All scans were analysed by the chief CT radiographer at our institution. A second research assistant reviewed a subset of 20 scans. The mean absolute difference in angle measurements (pooling all angles measured) was 1.1° (SD = 1.1). The difference between the observers was 2° or less in 91.0% of cases. In the literature it has been noted that intraobserver error averages 2° on any given TKA radiograph [13].
2.1. Study group 2.4. Outcome measures The PSG group comprised of 115 patients who underwent TKA surgery during a consecutive time period from April 2010 to February 2012. Patients who were not suitable to have a PSG TKA underwent surgery using conventional instrumentation during the same period as those patients who underwent a PSG TKA. There were 92 patients in this group who had a CON TKA. Reasons for exclusion from undergoing a PSG TKA included inability to undergo an MR scan because of cardiac and cerebral metal implants and patient factors such as cost and short time to surgery. This group also included patients who failed to have a PSG manufactured because of decreased image quality of their MR scan due to motion artefact. The commonest reason for this was claustrophobia as a cause of movement. There were no cases where knee deformity or pathology resulted in failure to produce a PSG. The final cohort comprised 53 patients who had CAS TKA using the Orthosoft Navigation System in 2009 prior to commencement of use of PSI guides. These patients voluntarily agreed to undergo CT radiographic assessment to act as a comparison group and represented a subgroup of 98 patients who had a CAS TKA in the same year. 2.2. Surgical technique All operations were performed using an identical surgical technique by two orthopaedic knee surgeons. A medial parapatellar approach was used and a thigh tourniquet was inflated for the initial surgical exposure only. The one prosthesis was used in all cases, the Zimmer NexGen LPS Flex (Zimmer, Warsaw, IN) using fully cemented implants. The patella was resurfaced in all knees. All operations were performed in an identical manner using a measured resection technique. All three techniques aimed for restoration of a neutral hip–knee angle, femoral component angle and tibial component angle in the coronal plane, a neutral femoral component flexion relative to the sagittal mechanical axis, seven degrees of posterior tibial slope and femoral component rotation parallel to the surgical femoral epicondylar axis. For those patients who underwent a PSG TKA, a proprietary internetbased software planner was used to review, adjust and approve the surgical plan. Once registered, the approved plan was forwarded to Materialise (Leuven, Belgium) who used rapid-prototyping technology to create the customised patient specific guides. The patient specific guides used in this study rely on data acquisition from a pre-operative MRI scan to create a custom mould of the knee that conforms to the patient’s knee anatomy at the time of surgery. The distal femoral resection was performed first in all cases, followed by the proximal tibial resection. An extra-medullary tibial guide attached to a 10 mm spacer block was used to ensure creation of an adequate extension gap as well as a tibial resection perpendicular to the tibial mechanical axis. The remainder of the femoral preparation was then carried out. In the conventional instrumentation group, intramedullary femoral and tibial alignment guides were used to restore coronal and sagittal knee alignment. 2.3. Radiographic analysis All patients in this series underwent a radiographic analysis using the Perth CT Protocol [12]. The Perth CT Protocol is a technique whereby patients are CT scanned non weight bearing to allow a direct measurement of the alignment of the femoral and tibial components in the
The post-operative parameters that were analysed were hip–knee angle, individual component alignment in the coronal and sagittal plane, and axial rotation of the femoral component. The proportion of patients within 3° of the neutral mechanical alignment for each measurement was recorded. Operative times were recorded in minutes and defined as the time from the initial skin incision until final wound closure. 2.5. Statistical analysis Means, ranges, and standard deviations were recorded for all radiographic parameters. Student's t-tests and ANOVA were used to compare differences in means between continuous variables. ANOVA was used to statistically assess pre-operative demographic factors. The two-sample t-test was used to assess for differences in operative time between groups. Fisher's exact test was to test for significance in difference of alignment (proportions) between the groups. Statistical significance was set at p b 0.05. 3. Results There were no significant differences with regards to age, BMI, and pre-operative HKA between groups (Table 1). 3.1. Hip–knee coronal angle When assessing HKA, 91.3% and 90.7% of patients in both the PSG and CAS groups, respectively, were within 3° of a neutral mechanical alignment. This compared to 80.4% of patients in CON group, which was statistically significant (p = 0.039, PSG vs CON). See Table 2 for a summary of results. 3.2. Individual component alignment 97.4% and 96.3% of patients in the PSG and CAS group, respectively, had a femoral coronal alignment within 3° of a neutral mechanical axis. This compared to 89.1% of patients in the CON group and this difference was statistically significant (p = 0.02). In the PSG group, 94.8% were within 3° of neutral femoral sagittal alignment. This compared to 98.2% and 90.2% of cases in the CAS and CON groups respectively. The differences between the CAS and CON groups compared to the PSG group were not found to be significant (p = 0.178 and p = 0.282, respectively). Tibial coronal alignment was within 3° of the neutral mechanical axis in 97.4% of patients in the PSG group and to 100.0% in the CAS group (p = 0.552). In the CON group, 88.0% were within an acceptable range and this was statistically significant when compared to PSG (p = 0.011). Mean posterior tibial slope was found to be significantly decreased in the CAS group (Mean = 4.9°, SD = 1.7) compared to the CON (Mean = 6.9°, SD = 2.0) and PSG group (Mean = 7.0°, SD = 1.9, p = 0.009). The reduced tibial slope in the CAS group
Table 1 Group demographics.
Age BMI Pre-op HKA Range HKA a b c
PSGa (n = 115)
CONb (n = 92)
CASc (n = 53)
P-value
67 30.3 −3.3° −15° to +14°
66 31.5 −4.3° −12° to +12°
67 31.2 −3.3° −20° to +10°
0.7058 0.1718 0.4646 N/A
Patient Specific Instrumentation. Conventional Instrumentation. Computer navigated surgery.
408
S.J. MacDessi et al. / The Knee 21 (2014) 406–409
Table 2 Post-operative hip–knee angle.
Mean Range Cases within +/− 3° Standard Deviation
Table 4 Operative times. PSG (n = 115)
CON (n = 92)
CAS (n = 53)
−0.1° −5° to +7° 91.3% 2.2
−0.6° −9° to +7° 80.4% 2.9
0.1° −4° to +6° 90.7% 2.2
was believed to be due to intra-operative differences in referencing points in the sagittal plane and not due to significant variations in outliers. Mean femoral component rotation was within 3° of neutral alignment to the surgical epicondylar axis in 90.4%, 96.3% and 92.4% of cases in the PSG, CAS and CON groups respectively. The differences between the CAS and CON groups compared to the PSI group were not found to be significant (p = 0.230 and p = 0.810, respectively). Component alignments were compared across each group for outliers greater than 3°. These are shown in Table 3. 3.3. Operative times The PSG group had the shortest mean operative time of 80.2 min, compared to 86.0 min in the CON group and 110.2 min in the CAS group (Table 4). These differences were found to be significant between both groups compared to the PSG group (p b 0.05). 3.4. Effects of BMI on post-operative alignment A sub-group analysis of patients who had a BMI of more than 30 was undertaken to determine if there were differences in alignment when PSG was used. No statistical difference in BMI was noted for any of the recorded angles, when comparing those within 3° to those outside 3°.
3.5. Effects of pre-operative deformity on post-operative alignment using PSG Patients who underwent a PSG TKA were divided in to those who had a pre-operative HKA of greater than or equal to 6° and those who had lesser pre-operative deformities to determine of these guides were less effective for either group. No significant differences in the post-operative alignment were found in patients with HKA of 6° or greater, compared to those with HKA less than 6°.
4. Discussion Patient-specific guides have been introduced in total knee arthroplasty as a method to improve alignment whilst at the same time reducing operative times and technical issues associated with CAS TKA surgery. As PSG TKA uses pre-operative imaging to acquire registration points for component positioning, it has been hoped that more accuracy in definition of intra-operative landmarks such as the femoral epicondylar axis will allow greater accuracy in restoring alignment. Other proposed benefits include avoiding canal instrumentation
Table 3 Comparing groups for outliers N3°. PSG (n = 115) HKA FCA TCA FSA TSA FRA
8.7% (n 2.6% (n 2.6% (n 5.2% (n 8.7% (n 8.7% (n
= = = = = =
10) 3) 3) 6) 11) 11)
CON (n = 92)
NAV (n = 53)
P-value (PSI:CON)
P-value (PSI:CAS)
19.6% (n = 18) 10.9% (n = 10) 11.8% (n = 11) 9.8% (n = 9) 8.7% (n = 8) 8.7% (n = 7)
9.4% (n = 5) 3.8% (n = 2) 0.0% (n = 0) 0.0% (n = 0) 26.4% (n = 14) 3.8% (n = 2)
0.039⁎ 0.020⁎ 0.011⁎ 0.282 1.000 0.805
1.000 0.653 0.552 0.178 0.009⁎ 0.230
HKA: Hip–knee angle. FCA: Femoral coronal angle. TCA: Tibial coronal angle. FSA: Femoral sagittal angle (Femoral flexion/extension angle). TSA: Tibial sagittal angle (Tibial posterior slope angle). FRA: Femoral rotation angle. ⁎ Significant p-values.
Operative time
PSG (n = 115)
CON (n = 92)
NAV (n = 53)
P-value (PSI:CON)
P-value (PSI:CAS)
80.2 min
86.0 min
110.2 min
P = 0.002
P b 0.0001
that is required with CON TKA, and avoiding navigation pins used in CAS TKA. It is well recognised that computer navigation in TKA reduces alignment outliers when compared to conventional instrumentation [8–11]. However, computer navigation has not continued to grow in popularity as many surgeons feel that these benefits cannot be justified based on increased operative time, greater cost, extra instrumentation required in the operating room, and complications related to pin sites used for navigation trackers [9,14–16]. The majority of evidence still suggests that minimisation of alignment outliers in the coronal plane improves implant survival and possibly improves function [10,17]. To date, there are no published studies comparing the accuracy of patient-specific guides to both computer navigation and conventional instrumentation. In addition most studies have either been limited case series or have only compared patient-specific guides against conventional instrumentation [7,18]. There has also been variability in the methodology of assessing post-operative limb alignment. Some studies have only used plain radiography, which is known to be affected by rotation of the limb, whilst other studies have used a limited CT scanogram to assess overall limb alignment [1,7]. All published studies to date have only assessed coronal plane alignment [1,4–7]. Chareancholvanich et al., using Zimmer Patient Specific Instruments (PSI), reported similar hip–knee angles within 3° of neutral alignment in their CON and PSG groups. The authors concluded that the minimal advantage obtained in both alignment and operative times is unlikely to be of clinical importance [4]. Like the Zimmer PSI, the Biomet Signature system (Biomet, Warsaw, In) relies on MRI to create the cutting guides. Nunley et al. prospectively assessed three groups of patients who underwent surgery with conventional instrumentation, Biomet signature patient-specific instruments, and the Otismed system (Stryker, NJ, USA) [5]. There was no significant difference in alignment between the conventional and Biomet Signature group, with 84.0% and 82.0% of patients respectively having a hip–knee angle within 3° of neutral. This compared to only 56.0% of patients who underwent kinematic axis restoration using the Otismed system. Ng et al. compared the Biomet Signature guides to conventional instrumentation [7]. Long leg plain X-rays were used to assess postoperative HKA as well as independent coronal component positioning. Their findings were similar to our results with 91.0% of patients with a postoperative hip–knee angle within 3° compared to 78.0% in the CON group. Bali et al. evaluated an initial series of patients who underwent a TKA using the Visionaire custom-fit guides (Smith & Nephew, Memphis, TN) [6]. These guides rely on a combination of MRI to map joint surface topography as well as plain radiography to restore coronal alignment. Out of 32 patients, 29 had a postoperative hip–knee angle within 3° of the neutral mechanical axis based on full length X-rays only. This is the first study that has assessed alignment outcomes of PSG TKA and compared it to both CAS and CON TKA. It is also the first review to use a validated CT protocol to look at independent implant positioning in all three alignment planes. Our results indicate that PSG has similar accuracy to CAS surgery with 91.3% and 90.7% of patients in both groups, respectively, having a hip–knee angle within 3° of a neutral mechanical axis. Independent coronal component position was also found to be similar between the two groups with less outliers compared to conventional instrumentation. No differences were noted in femoral component sagittal alignment between all three groups.
S.J. MacDessi et al. / The Knee 21 (2014) 406–409
This study also found that the PSG group had a significantly shorter mean operative time when compared to both the CON and CAS groups. We believe however that the difference of six minutes in operative time between the PSG and CON group is not of clinical significance. This study has several limitations. Firstly, the study groups were not randomised, hence there was a potential for selection bias amongst the patients who underwent each different technique. Although demographics were similar between groups, such bias cannot be excluded. In addition, the CAS group was from an earlier time period, and may have only represented a cohort of patients who had a satisfactory postoperative outcome when compared to the remainder of the group who underwent surgery using this technique. In a recent meta-analysis by Hetaimish et al. comparing CAS to CON TKA, 87.0% of patients undergoing CAS had a hip–knee angle within 3° compared to 70.0% undergoing CON TKA [10]. In our series of patients, 91.0% of patients had a hip–knee angle within 3° in the CAS group and 80.0% in the CON group. Although it is important to highlight the potential limitations of this study, it appears that our results in terms of assessing CAS instrumentation are slightly more favourable than in the literature published on these techniques. In conclusion, we found that patient-specific guides more reliably restore the coronal alignment when compared to conventional instrumentation and have a similar accuracy when compared to computer navigation surgery. In addition operative times were found to be shorter in the PSG group, the surgery was found to be less complicated than computer navigation and canal instrumentation was avoided. 5. Conflict of interest The authors declare that there are no conflict of interest. References [1] Howell SM, Kuznik K, Hull ML, Siston RA. Results of an initial experience with custom-fit positioning total knee arthroplasty in a series of 48 patients. Orthopedics 2008;31:857.
409
[2] Slover JD, Rubash HE, Malchau H, Bosco JA. Cost-effectiveness analysis of custom total knee cutting blocks. J Arthroplasty 2012;27:180. [3] Nunley RM, Ellison BS, Ruh EL, Williams BM, Foreman K, Ford AD, et al. Are patientspecific cutting blocks cost-effective for total knee arthroplasty? Clin Orthop Relat Res 2012;470:889. [4] Chareancholvanich K, Narkbunnam R, Pornrattanamaneewong C. A prospective randomised controlled study of patient-specific cutting guides compared with conventional instrumentation in total knee replacement. J Bone Joint Surg Br 2013;95-B:354. [5] Ng V, DeClaire JH, Berend KR, Gulick BC, Lombardi Jr AV. Improved accuracy of alignment with patient-specific positioning guides compared with manual instrumentation in TKA. Clin Orthop Relat Res 2012;470:99. [6] Bali K, Walker P, Bruce W. Custom-fit total knee arthroplasty: our initial experience in 32 knees. J Arthroplasty 2012:27. [7] Nunley RM, Ellison BS, Zhu J, Ruh EL, Howell SM, Barrack RL. Do patient-specific guides improve coronal alignment in total knee arthroplasty? Clin Orthop Relat Res 2012;470:895. [8] Mason JB, Fehring TK, Estok R, Banel D, Fahrbach K. Meta-analysis of alignment outcomes in computer-assisted total knee arthroplasty surgery. J Arthroplasty 2007;22:1097. [9] Kim SJ, MacDonald M, Hernandez J, Wixon RL. Computer assisted navigation in total knee arthroplasty. J Arthroplasty 2005;20:123. [10] Hetaimish BM, Khan MM, Simunovic N, Al-Harbi HH, Bhandari M, Zalzal PK. Metaanalysis of navigation vs conventional total knee arthroplasty. J Arthroplasty 2002;27:1177. [11] Choong PF, Dowsey MM, Stoney JD. Does accurate anatomical alignment result in better function and quality of life? Comparing conventional and computer-assisted total knee arthroplasty. J Arthroplasty 2009;24:560. [12] Chauhan SK, Clark GW, Lloyd S, Scott RG, Breidahl W, Sikorski JM. Computer-assisted total knee replacement. A Controlled Cadaver Study using a Multi-Parameter Quantitative CT Assessment of Alignment (The Perth CT Protocol). J Bone Joint Surg Br 2004;86:818. [13] Lonner JH, Laird MT, Stuchin SA. Effect of rotation and knee flexion on radiographic alignment in total knee arthroplasties. Clin Orthop 1991;331:102. [14] Jenny JY, Miehlke RK, Giurea A. Learning curve in navigated total knee replacement. A multi-centre study comparing experienced and beginner centres. Knee 2008;15:80. [15] Li CH, Chen TH, Su YP, Shao PC, Lee KS, Chen WM. Periprosthetic femoral supracondylar fracture after total knee arthroplasty with navigation system. J Arthroplasty 2008;23:304. [16] Novak EJ, Silverstein MD, Bozic KJ. The cost-effectiveness of computer-assisted navigation in total knee arthroplasty. J Bone Joint Surg Am 2007;89:2389. [17] Klatt BA, Goyal N, Austin MS, Hozack WJ. Custom-fit total knee arthroplasty (OtisKnee) results in malalignment. J Arthroplasty 2008;23:26. [18] Lombardi Jr AV, Berend KR, Adams JB. Patient-specific approach in total knee arthroplasty. Orthopedics 2008;31:927.
