Preliminary radiological evaluation of the Vector Vision CT-free knee module for implantation of the LCS knee prosthesis
Received 25 January 2006; received in revised form 26 September 2006; accepted 2 October 2006. published online 11 December 2006.
Abstract
We have assessed the bone cuts achieved at surgery as compared to the planned cuts produced during computer assisted surgery (CAS) using a CT-free navigation system. In addition, two groups of matched patients were compared to assess the post-operative mechanical alignment achieved. Fourteen patients received a LCS total knee replacement (TKR) using the Vector Vision module and 14 received a TKR using a conventional method of extramedullary alignment jigs. The deviation in each plane (valgus–varus, flexion–extension and proximal–distal) was calculated.
For the tibia the mean deviation in coronal plane was 0.21° of varus (SD=1.37) and in the sagittal plane was 1.29° of flexion (SD=3.73) and 0.24 mm of resection distal to the anticipated cut (SD=2.14). For the femur the mean deviation in the coronal plane was 0.88° (SD=2.2) of valgus and in the sagittal plane the mean deviation was 0.3° (SD=2.91) of extension. In the transverse plane there was a mean deviation of 0.07° (SD=1.57) of external rotation. There was a mean deviation of 2.33 mm of proximal resection (SD=2.9) and 1.05 mm of anterior shift (SD=2.81).
On comparing the two groups, no statistically significant differences were found for the angles between the femoral component and the femoral mechanical axis, the tibial component and the tibial mechanical axis, the femoral and tibial mechanical axis and the femoral and tibial anatomical axis. This study has presented preliminary data regarding the efficacy of a particular navigation system with regards to improving upon the accuracy of component position with the long-term aspiration of improving upon TKR longevity. A further randomised controlled trial with greater numbers of cases and controls would improve upon our knowledge as to the efficacy of the Vector Vision system and a power analysis based upon the findings of this pilot study has suggested that at least thirty subjects be included in each group.
Successful outcome of total knee replacement (TKR) is dependent upon several factors that include implant design, surgical technique and patient-related factors [1]. Correct positioning of components is a major factor in the long-term success of TKR surgery with malalignment being associated with decreased function, accelerated wear and early failure of implants [2], [3]. Conventional mechanical extramedullary and intramedullary alignment jigs may both have limitations in their accuracy and reproducibility [4,5]. Recent attention has focussed on the use of computer navigation systems during TKR surgery in an attempt to improve upon these two variables. Various computer navigation systems have been developed, but published data is often limited to single reports regarding each system [6] and a suitable comparison group is often lacking such that valid comparisons regarding improvements in mechanical alignment cannot be made.
We have performed a prospective cohort study which has demonstrated no increased effectiveness of the BrainLAB Vector Vision image guide surgery system in improving the alignment achieved at TKR surgery as opposed to extramedullary alignment.
2. Methods
Twenty eight TKRs were entered into the study. Cases consisted of 14 consecutive primary TKRs that were implanted using the Vector Vision image guide surgery system (BrainLAB, Munich, Germany). Fully informed consent was obtained from all patients and the study had been approved by the local ethics committee. The control group consisted of 14 consecutive TKRs that were inserted using an extramedullary tibial alignment jig and an intramedullary femoral alignment jig. No patient refused to participate in the study. All operations were performed by the senior author (MLP) using an uncemented LCS (DePuy, Leeds, UK) prosthesis. The 14 patients in the test group ranged in age from 38 years to 83 years (Mean age: 63.9 years) whereas the patients in control group ranged in age from 45 to 78 years (Mean: 66.8 years). All patients except two had primary osteoarthritis. The remaining two had rheumatoid arthritis, one in each group. Right to left ratio was equal.
The Vector Vision platform is an image-free open platform i.e. information is acquired at the time of surgery and it is possible to use implants from a number of manufacturers with this system. The overall principles of such a system have previously been described in detail [8]. Briefly, during surgery a series of landmarks are collected to identify the mechanical leg axis and to generate a 3D model of a leg that is morphed to the patient's individual anatomy. Base on the intra-operatively acquired data the software recommends a treatment plan which can then be optimised or modulated by the surgeon. Movements of the surgeon's instruments are guided by, and visualised in real time, on a model displayed on a monitor. Referencing, cutting block positioning and cut verification rely upon hand-held or rigid jigs fitted with reflective balls whose position is picked up by a infrared beam produced and detected by an optical localiser connected to a computer and monitor.
In the test group data was taken from two screenshots automatically produced by the Vector Vision software for each patient. These two screen shots detail the final cut position achieved and the extent of the deviation from the planned position. Post-operatively a long leg AP radiograph was obtained for each patient in the test and the control group and analysed for the alignment of the prosthesis. Four angles were measured:
1.The angle between the femoral component and the femoral mechanical axis.
2.The angle between the tibial component and the tibial mechanical axis.
3.The angle between the femoral and tibial mechanical axis.
4.The angle between the femoral and tibial anatomical axis.
The results were analysed using the non-parametric tests (Wilcoxon signed rank test) owing to small sample sizes using the SPSS software (version 13.0). No attempts have been made here to correlate the alignment of the components or of the mechanical axes with the clinical improvement or the longevity of the implant, the primary aim of the study being whether use of navigation improves post-operative alignment of the components.
3. Results
The intra-operative tibial plane deviations in the test group as obtained by computer software generated data have been summarised as in Table 1. The deviation in each plane (i.e. valgus–varus, flexion–extension and proximal–distal) was calculated as the difference between the planned cut as on computer navigation system software and the actual cut made. The mean deviation in coronal plane was 0.21° of varus (SD=1.37) and in the sagittal plane was 1.29° of flexion (SD=3.73) and 0.24 mm of resection distal to the anticipated cut (SD=2.14).
Table 1.
Tibial plane deviations in the test group for alignment, flexion (Flex) and extension (Ext) parameters and resection distal (Dist) or proximal (Prox)
Patient ID
Tibial plane deviations
Valgus +/−Varus (°)
Slope Flex+/−Ext (°)
Resection Dist (+) or Prox (−) in mm
Actual
Deviation
Actual
Deviation
Actual
Deviation
7 01
0.9
0.9
13
3
2.4
0.4
7 02
0.3
1.8
6.5
0.3
10.6
4.7
7 03
0.5
0.5
12.3
5.3
1.7
−0.3
7 04
−0.5
−0.5
7.8
0.8
7.1
−0.9
7 05
−2.4
−2.4
1
−6
−1.3
−3.3
7 06
1
1
11.4
4.4
6.1
0.1
7 07
−0.3
−0.3
11.2
4.2
2.4
0.4
7 08
0.5
0.5
3.4
−3.6
0.7
−1.3
7 09
0
0
7.7
0.7
0.3
−1.7
7 10
0.7
0.7
8.3
1.3
2.9
0.9
7 11
−1.3
−1.3
14.4
7.4
2.3
0.3
7 12
−0.4
−0.4
10.5
3.5
6.4
4.4
7 13
−3.3
−3.3
3.7
−3.3
4.7
0.7
7 14
−0.1
−0.1
7.1
0.1
9
−1
The intra-operative femur plane deviations are summarised in Table 2, Table 3. The mean deviation in the coronal plane was 0.88° (SD=2.2) of valgus and in the sagittal plane the mean deviation was 0.3° (SD=2.91) of extension. In the transverse plane there was a mean deviation of 0.07° (SD=1.57) of external rotation. There was a mean deviation of 2.33 mm of proximal resection (SD=2.9) and 1.05 mm of anterior shift (SD=2.81).
Table 2.
Femoral plane deviations for the test group regarding alignment, flexion (Flex) and extension (Ext) parameters and resection distal (Dist) or proximal (Prox)
Patient ID
Femur plane deviations
Valgus+/−Varus (°)
Slope Flex+/−Ext (°)
Resection Dist (+) or Prox (−) in mm
Actual
Deviation
Actual
Deviation
Actual
Deviation
7 01
−1.6
−1.6
2.7
0.2
8.5
0
7 02
−1.5
−1.5
1.9
1.1
6.2
0.2
7 03
No
Screen
Shot
7 04
0.5
0.5
0.9
0.3
4.2
−4.1
7 05
−0.9
−0.9
−6.2
−6.2
9.8
−1.1
7 06
−0.3
−0.3
−0.6
−2.5
−6.7
−0.6
7 07
3.1
3.1
−1.7
−1.7
5.6
−2.3
7 08
0.4
0.4
−1.5
−1.8
3.9
−6.3
7 09
−1.8
−1.8
6.1
−3.9
−0.7
−8.7
7 10
−2.4
−2.4
7.6
2.9
1.6
−2.3
7 11
1.7
1.7
1.8
1.7
10.3
−0.6
7 12
2.9
2.9
1.8
1.7
5.2
−1.9
7 13
1.8
1.8
2.4
2.3
9.4
1.3
7 14
4.1
4.7
5.3
3.9
17.3
0.9
Table 3.
Femoral plane deviations for rotation internal (Int) or external (Ext) and anterior (Ant) or posterior (Post) shift
Patient ID
Femur plane deviations
Rotation Int (+) or Ext (−) in °
Shift Ant (+) or Post (−) in °
Actual
Deviation
Actual
Deviation
7 01
−1.5
−0.9
1.2
0.9
7 02
−1.2
−1.5
5.3
1.6
7 03
No
Screen
Shot
7 04
−0.6
−0.2
1.5
−0.2
7 05
7.8
0.3
−3
4.9
7 06
4.5
0.9
−1.1
−2
7 07
−1.5
0.4
1.7
−2.5
7 08
−3.9
−1.9
7.7
7.7
7 09
0
0.4
0.3
0.1
7 10
4.8
−1.6
5.6
2.4
7 11
−3.9
−0.4
0.6
0.5
7 12
1.9
4
−2.3
−1.9
7 13
−2.5
−1.1
−0.9
1.5
7 14
2.7
−1.6
−2.5
1.1
On analysing the post-operative X-rays, the mean angular difference between the femoral component axis and the femoral mechanical axis for the test group was 0.69° (SD=1.64) and this was measured as varus alignment. Compared to this, the same angle measured in test group averaged 1.05° (SD=1.96) again in varus alignment. The difference was not found to be statistically significant (p=0.445) using the Wilcoxon signed rank test. The mean angular difference between the tibial component axis and the tibial mechanical axis for the test group was 0.84° (SD=1.96) valgus compared to 0.65° (SD=2.14) of valgus for the control group, the difference again was not statistically significant (p=0.413). The mean difference between the femoral mechanical axis and the tibial mechanical axis for the test group was 0.68° of varus (SD=3.87) compared to 1.19° of valgus for the control group (SD=2.96), the difference again was not statistically significant (p=0.445). Lastly the mean difference between the femoral component axis and the tibial component axis in the test group was 0.1° of varus (SD=2.0) compared to 0.59° of valgus (SD=1.59) for the control group, the difference not being statistically significant (p=0.451).
A power analysis to determine an estimate of sample size was performed using a power of 80%, a level of significance of 5%, a mean estimate of standard deviation of 2.44 and an estimate of significant difference between groups of either 2° or 2 mm of intra-operative variation. This suggested that at least thirty subjects should be included in each group.
4. Discussion
To our knowledge there has been only one other published study assessing the component and mechanical axis using the Vector Vision system [7]. In this study no mention was made of differences between bone cuts achieved and the angle at which cutting blocks were set for this system. However, using a CT assessment protocol it was demonstrated that there was an overall 3.27° mean error of component malalignment over six measured parameters with 22% obtaining 0° error in all parameters.
Our preliminary results demonstrate that use of the Vector Vision system produces reliable and accurate bone cuts. However, the differences achieved between the suggested cut and those achieved by the surgeon demonstrate that variation in implantation of components is not fully resolved by the use of computers and is subject to tolerances in the instruments used and surgical technique. Given the relatively small numbers involved, this study is best regarded as a pilot for a randomised controlled trial with larger numbers as determined by power calculation which would demonstrate if the lack in difference between the two groups may be due to inadequate numbers. The power analysis performed has suggested the need for a minimum of thirty subjects in each group.
The processes of obtaining fixed reference points and especially that of the centre of rotation of the femoral head have yet to be fully assessed in the clinical setting and are possible areas of intra-operative error [9]. It is important to remember that we used radiographic analysis only for comparisons of implant positions and it is possible that this may not be sensitive enough to reliably establish all small improvements in accuracy of computer guided surgery [5], [6]. As the same observer assessed radiographs for both groups using a standard technique, we have assumed that error in measurement is equal between the two groups. It may be that CT scan protocols may be more reliable in assessing relevant differences between groups.
The aim of improving the alignment achieved in TKR surgery is driven by the aspiration that this may lead to better long-term survival and functional outcome. Recent reports, including this one, provide encouraging initial pilot data although this study failed to demonstrate a statistically significant difference in the alignment achieved as compared to using extramedullary alignment guides. It is important to keep sight of associated factors such as safety, increased operative time and cost implications. Most importantly, although there would appear to be increasing evidence that computer-assisted navigation may improve alignment in TKR, its true usefulness will only become apparent after adequately powered and controlled long-term follow-up studies of implant survival and functional outcome become available which may be difficult to achieve given the good long-term survival data already available for TKR [10].
References
[1]. [1]Knutson K, Lindstrand A, Lidgren L. Survival of knee arthroplasties: a nation-wide multicentre investigation of 8000 cases. J Bone Jt Surg Br. 1986;68B:795–803.
[2]. [2]Ranawat CS, Boachie-Adjei O. Survivorship analysis and results of total condylar knee arthroplasty. Eight–11 year follow-up period. Clin Orthop. 1988;226:6–13.
[3]. [3]Dorr LD, Boiardo RA. Technical considerations in total knee arthroplasty. Clin Orthop. 1997;205:5–11.
[4]. [4]Teter KE, Bergman D, Colwell CW. Accuracy of intramedullary versus extramedullary tibial cutting alignment systems in total knee arthroplasty. Clin Orthop. 1995;321:106.
[5]. [5]Stulberg SD, Loan P, Sarin V. Computer-assisted navigation in total knee replacement: results of an initial experience in 35 patients. J Bone Jt Surg Am. 2002;84A:90–98.
[6]. [6]Haaker RG, Stockheim M, Kamp M, Proff G, Breitenfelder J, Ottersbach A. Computer-assisted navigation increases precision of component placement in Total Knee Arthroplasty. Clin Orthop 433: 152–9
[7]. [7]Sikorski JM, Blythe MC. Learning the vagaries of computer-assisted total knee replacement. J Bone Jt Surg Br. 2005;87B:903–910.
[8]. [8]Sikorski JM, Chauhan S. Computer-assisted orthopaedic surgery: do we need CAOS?. J Bone Jt Surg Br. 2003;85B:319–323.
[9]. [9]Victor J, Hoste D. Image-based computer assisted total knee arthroplasty leads to lower variability in coronal alignment. Clin Orthop. 2004;428:131–139.
CrossRef
[10]. [10]Laskin RS. Choosing your implant. J Arthroplast. 2005;20(4Suppl II):7–9.
aArrowe Park Hospital, Wirral NHS Trust, Upton, Wirral CH49 5PE, United Kingdom
bCentre for Integrated Genomic Research, Stopford Building, The University of Manchester, Oxford Road, Manchester M13 9TP, United Kingdom
cCentre for Hip Surgery, Wrightington Hospital, Hall Lane, Appley Bridge, Wigan WN6 9EP, United Kingdom
Corresponding author. 14 The Boulevard, Didsbury Point, Manchester M20 2EU, United Kingdom. Tel.: +44 161 4489972; fax: +1 61 2751617.
FULL TEXT
