Patient-matched positioning guides in total knee arthroplasty · 2017. 7. 9. · 88.5/1000 in women...

Bert Boonen

total knee arthroplasty

Patient-matchedpositioning guides in

Colofon Author: Bert Boonen Production/print: Datawyse | Universitaire Pers Maastricht ISBN: 978 94 6159 638 3 The publication of this thesis was kindly supported by:

De digitale versie van dit proefschrift is te vinden in de YourThesis-app en kan gelezen worden op een tablet of smartphone. De app kan gedownload worden in de App Store en de Google Play store of middels het scannen van onderstaande QR-code.

©Bert Boonen, 2016 All rights reserved. No part of this publication may be reproduced or transmitted in any form by any means, without permission of the author. Images for the cover design were provided by Zimmer Biomet. Other than providing permission to use the Zimmer Biomet images, this publication is not financially supported by Zimmer Biomet. Zimmer Biomet is the owner of the copyrights and all other intellectual property rights in relation to the images used.

Nederlandse Orthopaedische Vereniging Stichting Kliniek en Wetenschap Orthopedie, MUMC+ Maastricht University Spronken Orthopedie Smeets loopcomfort Sportho Hanssen footcare Footlife Nederland Anna Fonds Livit Wellspect Healthcare ZimmerBiomet Vigo Heerlen Bauerfeind Nederland Zuyderland Medisch Centrum

Patient-matched positioning guides in total knee arthroplasty

PROEFSCHRIFT

Ter verkrijging van de graad van doctor aan de Universiteit Maastricht,

op gezag van de Rector Magnificus, Prof. dr. Rianne M. Letschert, volgens het besluit van het College van Decanen

in het openbaar te verdedigen op vrijdag 17 februari 2017 om 10.00 uur

door

Bert Boonen

Geboren op 16 januari 1985 te Neerpelt, België

Promotor Prof. dr. L.W. van Rhijn Copromotoren Dr. N.P. Kort (Zuyderland medisch centrum, Sittard-Geleen) Dr. P.J. Emans Beoordelingscommissie: Prof. dr. R. A. de Bie (voorzitter) Prof. dr. R. G. T. Geesink Dr. H.M.J. van der Linden - van der Zwaag (LUMC, Leiden) Prof. dr. N. L. U. van Meteren Prof. dr. P. Verdonk (UZA, Antwerpen, België)

5

Table of content

Chapter General introduction 7 1

Chapter Preliminary experience with the patient-specific templating total 2knee arthroplasty. 40 cases compared with a matched control group 19 Boonen B., Schotanus M.G.M, Kort N.P. Acta Orthop. 2012 Aug;83(4):387-93.

Chapter Patient-specific positioning guides for total knee arthroplasty: no 3significant difference between final component alignment and preoperative digital plan except for tibial rotation 35 Boonen B., Schotanus M.G.M., Kerens B., Hulsmans F-J., Tuinebreijer W.E., Kort N.P. Knee Surg Sports Traumatol Arthrosc. 2015 Jun 9.

Chapter Intra-operative results and radiological outcome of conventional and 4 patient-specific surgery in total knee arthroplasty: a multicentre, randomized controlled trial 53 Boonen B., Kerens B., Schotanus M.G.M., van der Weegen W., van Drumpt R.A.M., Kort N.P. Knee Surg Sports Traumatol Arthrosc. 2013 Oct;21(10):2206-12.

Chapter No difference in clinical outcome between patient-matched 5 positioning guides and conventional instrumented TKA at 2 years follow-up: a multi-centre, double-blind, randomized controlled trial 67 Boonen B., Schotanus M.G.M., Kerens B., van der Weegen W., Hoekstra H., Kort N.P. Bone Joint J. 2016 Jul;98-B(7):939-44.

Chapter Patient-specific guides in total knee arthroplasty; 2 years follow up of 6 the first 200 consecutive cases performed by a single surgeon 79 Boonen B., Schrander D.E., Schotanus M.G.M., Hulsmans F-J., Kort N.P. Published online. JCRMM. 2015. http://content.yudu.com/Library/A3yjzj/JOURNALOFCLINICALRHE/resources/index.htm?referrerUrl=http%3A%2F%2Ffree.yudu.com%2Fitem%2F details%2F3682413%2FJOURNAL-OF-CLINICAL-RHEUMATOLOGY-AND-MUSCULOSKELETAL-MEDICINE

6

Chapter Inter-observer reliability of measurements performed on digital long- 7 leg standing radiographs and assessment of validity compared to 3D CT-scan 91 Boonen B., Kerens B., Schotanus M.G.M., Emans P., Jong B., Kort N.P. Knee. 2016 Jan;23(1):20-4.

Chapter General discussion 105 8

Chapter Valorisation 119 9

Chapter Summary 123 10 Nederlandse Samenvatting 129

Chapter Dankwoord 133 11Curriculum Vitae 139 List of presentations and publications 141 Sponsors 145

7

Chapter 1

General introduction


9

INCIDENCE AND PREVALENCE OF KNEE OSTEOARTHRITIS AND TOTAL KNEE ARTHROPLASTY

Osteoarthritis (OA) of the knee is a common problem. In a cohort of 3018 African Amer-icans and Caucasians aged ≥ 45 years, 28% had radiographic knee OA (Kellgren-Lawrence ≥ 2), 16% had symptomatic knee OA, and 8% had severe radiographic knee OA (graded as Kellgren-Lawrence 3 or 4)1.

In 2011 in The Netherlands, the prevalence of knee OA was 53.8/1000 in men and 88.5/1000 in women (Nijmegen RvC. Continue Mobiliteits Registratie Nijmegen, Rad-boudUMC, afdeling huisarts-geneeskunde. 2011).

Total knee arthroplasty (TKA) is indicated for end-stage knee OA. The demand for TKA has increased in the past decade and was projected to double between 2005 and 2016 in the United States2. Kurtz et al. even projected that between 2005 and 2030 the demand for primary TKA will grow by 673%, from 450.000 in 2005 to 3.48 million in 2030 in the United States3.

In The Netherlands, the number of primary knee prostheses (TKA, unicondylar knee replacement and patellofemoral arthroplasty) procedures performed rises roughly with about 1000 cases every year. In 2010, 20.558 primary knee prostheses were performed. In 2014, this number had risen to 26.754. 90% of primary knee prostheses were TKA’s (LROI rapportage 2014).

Factors contributing to the increasing use and prevalence of knee replacements in-clude growth, aging, and increasing longevity of the population; obesity; expanding indications for the procedure; and younger age at implantation4.

MODERN TKA

TKA is a successful and cost-effective surgical intervention that provides pain relief, enhanced mobility, and improved quality of life for patients with end stage knee OA5,6. Survival after TKA approaches 95 % at 15 years7,8.

Different alignment methods can be used to guide bony resections in preparation of prosthesis placement. Traditionally, bone cuts in the distal femur are made perpendicu-lar to the mechanical axis, and this is typically done using an intramedullary rod. Similar-ly, the proximal tibia cut is performed perpendicular to the mechanical axis of the tibia using either intramedullary or extramedullary alignment rods (conventional instru-ments: CI). By performing the cuts like this, the goal is to obtain a neutral mechanical axis of the leg after TKA. Literature suggests that deviation of more than 3 degrees from a neutral mechanical axis is associated with increased failure rates9-12. The decreased survival caused by malalignment is likely due to off-axis loading, polyethylene wear and subsequent implant loosening10,13.

Chapter 1

10

Computer-assisted TKA (CAS) was introduced in the late 1990s to improve align-ment, reduce the incidence of outliers with the goal of improving long-term survival. CAS consists of a computer platform, tracking system and a rigid body marker. During surgery, anatomical landmarks around the knee are registered and pins are placed on the femur and tibia which are attached to the optical reference frames that work as tracking arrays. The system creates a virtual image allowing for intraoperative recording of joint range of motion and kinematics. Furthermore, prosthetic sizing and bone resec-tion level can be analysed. Meta-analysis studies have indeed showed that percentage of outliers (more than 3 degrees from neutral alignment) in the frontal plane was 9-13% with CAS and up to 32 % for the conventional instrumentation13,14. Studies have shown that CAS allows for a more accurate and reproducible bony resection, ligament balanc-ing, component sizing and kinematics evaluation15,16.

Apart from the higher percentage of outliers associated with CI compared to CAS, there are additional drawbacks associated with conventional techniques that use in-tramedullary alignment rods. These include extra blood loss perioperatively17, emboliza-tion of medullary content18,19, and difficulty in intramedullary rod passage due to de-formity, retained hardware, or pathological bone disease20.

Drawbacks of CAS on the other hand include loosening of trackers and bone frac-tures, complexity, long set-up time, increased surgical time, high capital costs and a substantial learning curve21-24. Furthermore, no long-term result studies provide evi-dence that CAS helps to reduce the revision rate of TKA. In addition, despite the reduc-tion in outliers in mechanical axis, clinical trials fail to demonstrate any improvement in clinical outcome scores25. These factors have limited the widespread use of CAS.

PATIENT-MATCHED POSITIONING GUIDES

The manufacturing process for TKA implants has improved over the years. Implant de-signs, sizing options and bearing surface materials have all been improved to enhance the outcomes of TKA. However, the most recent advance in TKA has involved the in-strumentation and the processes for TKA implantation with the development of patient-matched positioning guides (PMPG). This technique was introduced in the early 2000s with the purpose of eliminating most disadvantages associated with CI and CAS.

It uses patient-specific guides to make bony resections of femur and tibia. Imaging techniques and a specific software program are used to create virtual models of the patient’s femur and tibia. Manufacturers use different imaging techniques, but CT-scan or MRI-scan are the basic imaging modalities. The virtual models are then used to iden-tify anatomical landmarks of the knee which are used to calculate ideal implant size and positioning. In most cases, traditional reference points and reference axes are used for this calculation. These calculations subsequently result in a virtual plan of the operation to be performed in theatre (figure 1). This virtual plan can be adjusted by the surgeon if


11

desired. This way it is possible to plan the depth and the coronal orientation of resec-tion, the rotation and the slope of cuts.

Figure 1: digital plan, showing expected implant positioning

The final step is to create alignment guides that have only one fitting position on the patient’s individual femoral and tibial anatomy. A rapid prototype technique is used to create these alignment guides. They dictate bony resections during surgery. Again, different manufactures use different types of guides. Two main types of PMPG are commonly produced. Pinning guides replace conventional alignment tools for determin-ing the correct location of pins used to secure jigs (figure 2). Cutting guides have slots that allow cutting through them.

Table 1 provides an overview of the different types of patient-matched positioning guides that are available on the orthopaedic market.

Chapter 1

12

Figure 2: pinning guide for femur

There are some theoretical advantages of this way of outlining a TKA. Because the medullary cavity is not opened, blood loss would be expected to be lower compared to CI. As the decision making process is transferred from the intra-operative period to the pre-operative time frame, surgical time is expected to be reduced. In theory, the tech-nique should enable the surgeon to calculate the size of the prosthesis pre-operatively. A reduction in the needed instrument trays could therefore be expected. In addition, PMPG could eliminate variability among different surgeons. Theoretically, this method of alignment would help improve implant positioning as aberrant anatomy and land-marks should be readily identifiable on the pre-operative imaging.

The accuracy of PMPG, however, is determined by several factors, not just landmark identification on pre-op imaging. The imaging itself (MRI or CT-scan) has to be of suffi-cient quality. Landmark registration on this imaging and the subsequent calculation algorithm are crucial steps in the process that dictates final prosthesis component placement. The production process itself of the PMPG, using a rapid prototyping tech-nique, has to be accurate in order to guarantee adequate guide fitting during surgery. Proper peroperative use of the guides is essential, e.g. with some guides all soft tissues have to be removed (while retaining the osteophytes) to unsure a stable position of the guides on the native anatomy of femur and tibia.


13

RATIONALE FOR THIS THESIS

PMPG were brought to the market, but high-quality evidence outlining the results ob-tained with this technique was virtually non-existing. Therefore, in this thesis, PMPG will be compared to CI, as most surgeons to date adhere to this type of alignment method.

As with PMPG, in conventionally instrumented TKA, the accuracy is determined by several factors as well. Decision making is on a step-by-step basis. The advantage is that this allows the surgeon to check the intended bone cut before every actual bone cut is made, which is not entirely possible with PMPG. The disadvantage is that in some cases adequate information to base alignment on is lacking. For instance, a very pronounced antecurvation in the femur or bowing in the tibia might be present but go unnoticed during surgery. This might result in inadequate outlining of the prosthesis components.

Furthermore, in both CI and PMPG, apart from the use of the alignment method it-self that dictates the position of the sawing slots (in certain types of PMPG) and/or pins for traditional cutting instruments, there are additional factors that will influence the final component alignment. These factors constitute of: the inherent slack for the saw-blades in the slots, type of sawblade used and sawing technique of the surgeon, cemen-tation technique and method of cement curing.

Concerning the analysis of this component alignment, several techniques exist. Standing long-leg radiographs (LLR) are frequently used to assess alignment in the coro-nal plane and are seen as the gold standard for determining knee joint alignment26,27. This imaging technique was therefore also used in this thesis to compare alignment results between PMPG and CI. However, while conducting the research for this thesis, concerns had risen about the adequacy of measurements performed on LLR. Most re-search that claims to analyse the accuracy of measurements on LLR use computed to-mography (CT)-scan or intra-operative navigation as gold standards and calculate corre-lation coefficients28,29. Correlation coefficients, however, are not a correct outcome measurement for assessing agreement between two measuring techniques. Additional research was therefore conducted using the Bland-Altman method30 to assess agree-ment between measurements performed on LLR and 3D CT-scan (as gold standard) and to provide results on reliability and validity of measurements on LLR.

The ability of PMPG for obtaining a correct alignment of the TKA will be a large part of the research presented in this thesis. However, as the proportion of younger, more active and more demanding patients undergoing TKA is rising, function after TKA is becoming more important31. Moreover, with the growing need for joint replacement surgery, changes will have to be made in multiple areas of the process to meet the future growing demands32. Therefore, results on clinical outcome are also presented. Furthermore, the potential of PMPG for improving perioperative logistics (reduction in instrument trays, reduced surgical time, reduced blood loss and transfusion needs, reduction in length of hospital stay, reduction in hospital stock) will be explored.

Chapter 1

14

Table 1: overview of different types of patient-specific guides available.

Trade Name Manufacturer Imaging modality used Type of guide Prosthesis used

Signature Biomet MRI (CT also possible) Pinning Vanguard

PSI Zimmer MRI Pinning Nexgen

Trumatch J&J DePuy CT Cutting PFC Sigma

Prophecy Wright Medical CT or MRI Pinning Advance/Evolution

My Knee Medacta CT or MRI Cutting GMK, Cinétique, Evolis

OTIS Knee Stryker MRI Cutting Triathlon

iJig ConforMIS CT or MRI iUni G2/iDuo G2

Visionaire Smith&Nephew MRI & long-leg radiograph Cutting Journey/Legion/Genesis II


15

AIMS OF THIS THESIS

In this thesis, PMPG of Biomet (now Zimmer Biomet) were studied as these guides are used in the centres were the research leading to this thesis was conducted. Therefore, the research questions that we aimed at answering in this thesis relate to this specific type of alignment guides. The following questions were postulated:

- Can we rely on PMPG for aligning a TKA?

- What are the potential pitfalls when using PMPG that compromise their accura-cy?

- Can PMPG reduce the percentage of outliers in TKA compared to CI?

- Will use of PMPG lead to shorter operation time and less blood loss compared to CI?

- Will use of PMPG lead to shorter duration of hospitalization compared to CI?

- Will use of PMPG lead to better functional outcome than CI?

- Are PMPG safe to use?

- Are PMPG able to accurately predict implant size preoperatively?

- What is the inter-observer reliability of measurements performed on LLR and is it a valid technique compared to measurements performed on 3D-CT?

In Chapter 1, the first experience with PMPG in 40 consecutive patients will be de-scribed. 40 consecutive patients operated on using PMPG were compared to 40 matched control subjects operated on using the traditional intramedullary alignment technique. In Chapter 2, the ability of PMPG to reproduce the pre-operative digital plan will be described. For this analysis 3D-CT scans were used to directly compare the postopera-tive alignment to the pre-operative digital plan. This research was deemed an essential step in the validation of PMPG for potential widespread use. The outcome of the cohort study presented in Chapter 1 resulted in a multicenter, ran-domized, double-blinded, controlled trial comparing PMPG with conventional instru-mentation. 180 patients were enrolled in this study and the intra-operative results and alignment results of this trial will be described in chapter 3.

Chapter 1

16

Chapter 4 will describe the clinical outcome of patients enrolled in this RCT. In Chapter 5 the results of the first 200 consecutive cases performed with PMPG will be presented. The focus of this paper will primarily be on the number of intra-operative adjustments to the pre-operative digital plan that had to be made. Chapter 6 will focus on the inter-observer reliability and the validity of measuring align-ment on long-leg radiographs.


17

REFERENCES

1. Jordan JM, Helmick CG, Renner JB, Luta G, Dragomir AD, Woodard J, Fang F, Schwartz TA, Abbate LM, Callahan LF, Kalsbeek WD, Hochberg MC. Prevalence of knee symptoms and radiographic and symptomatic knee osteoarthritis in African Americans and Caucasians: the Johnston County Osteoarthritis Project. The Journal of rheumatology 2007;34-1:172-80.

2. Iorio R, Robb WJ, Healy WL, Berry DJ, Hozack WJ, Kyle RF, Lewallen DG, Trousdale RT, Jiranek WA, Stamos VP, Parsley BS. Orthopaedic surgeon workforce and volume assessment for total hip and knee replacement in the United States: preparing for an epidemic. The Journal of bone and joint surgery. American volume 2008;90-7:1598-605.

3. Kurtz S, Ong K, Lau E, Mowat F, Halpern M. Projections of primary and revision hip and knee arthroplasty in the United States from 2005 to 2030. The Journal of bone and joint surgery. American volume 2007;89-4:780-5.

4. Losina E, Thornhill TS, Rome BN, Wright J, Katz JN. The dramatic increase in total knee replacement utilization rates in the United States cannot be fully explained by growth in population size and the obesity epidemic. The Journal of bone and joint surgery. American volume 2012;94-3:201-7.

5. Ethgen O, Bruyere O, Richy F, Dardennes C, Reginster JY. Health-related quality of life in total hip and total knee arthroplasty. A qualitative and systematic review of the literature. The Journal of bone and joint surgery. American volume 2004;86-A-5:963-74.

6. Losina E, Walensky RP, Kessler CL, Emrani PS, Reichmann WM, Wright EA, Holt HL, Solomon DH, Yelin E, Paltiel AD, Katz JN. Cost-effectiveness of total knee arthroplasty in the United States: patient risk and hospital volume. Archives of internal medicine 2009;169-12:1113-21; discussion 21-2.

7. Font-Rodriguez DE, Scuderi GR, Insall JN. Survivorship of cemented total knee arthroplasty. Clinical orthopaedics and related research 1997-345:79-86.

8. Ranawat CS, Flynn WF, Jr., Saddler S, Hansraj KK, Maynard MJ. Long-term results of the total condylar knee arthroplasty. A 15-year survivorship study. Clinical orthopaedics and related research 1993-286:94-102.

9. Bargren JH, Blaha JD, Freeman MA. Alignment in total knee arthroplasty. Correlated biomechanical and clinical observations. Clin Orthop Relat Res 1983-173:178-83.

10. Jeffery RS, Morris RW, Denham RA. Coronal alignment after total knee replacement. J Bone Joint Surg Br 1991;73-5:709-14.

11. Lotke PA, Ecker ML. Influence of positioning of prosthesis in total knee replacement. J Bone Joint Surg Am 1977;59-1:77-9.

12. Ritter MA, Faris PM, Keating EM, Meding JB. Postoperative alignment of total knee replacement. Its effect on survival. Clin Orthop Relat Res 1994-299:153-6.

13. 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;22-8:1097-106.

14. Hetaimish BM, Khan MM, Simunovic N, Al-Harbi HH, Bhandari M, Zalzal PK. Meta-analysis of navigation vs conventional total knee arthroplasty. The Journal of arthroplasty 2012;27-6:1177-82.

15. Anderson KC, Buehler KC, Markel DC. Computer assisted navigation in total knee arthroplasty: comparison with conventional methods. The Journal of arthroplasty 2005;20-7 Suppl 3:132-8.

16. Catani F, Biasca N, Ensini A, Leardini A, Bianchi L, Digennaro V, Giannini S. Alignment deviation between bone resection and final implant positioning in computer-navigated total knee arthroplasty. The Journal of bone and joint surgery. American volume 2008;90-4:765-71.

17. Raut VV, Stone MH, Wroblewski BM. Reduction of postoperative blood loss after press-fit condylar knee arthroplasty with use of a femoral intramedullary plug. J Bone Joint Surg Am 1993;75-9:1356-7.

18. Caillouette JT, Anzel SH. Fat embolism syndrome following the intramedullary alignment guide in total knee arthroplasty. Clin Orthop Relat Res 1990-251:198-9.

Chapter 1

18

19. Fahmy NR, Chandler HP, Danylchuk K, Matta EB, Sunder N, Siliski JM. Blood-gas and circulatory changes during total knee replacement. Role of the intramedullary alignment rod. J Bone Joint Surg Am 1990;72-1:19-26.

20. Dennis DA, Channer M, Susman MH, Stringer EA. Intramedullary versus extramedullary tibial alignment systems in total knee arthroplasty. J Arthroplasty 1993;8-1:43-7.

21. Lombardi AV, Jr., Berend KR, Adams JB. Patient-specific approach in total knee arthroplasty. Orthopedics 2008;31-9:927-30.

22. Radermacher K, Portheine F, Anton M, Zimolong A, Kaspers G, Rau G, Staudte HW. Computer assisted orthopaedic surgery with image based individual templates. Clin Orthop Relat Res 1998-354:28-38.

23. Blakeney WG, Khan RJ, Wall SJ. Computer-assisted techniques versus conventional guides for component alignment in total knee arthroplasty: a randomized controlled trial. The Journal of bone and joint surgery. American volume 2011;93-15:1377-84.

24. Dutton AQ, Yeo SJ, Yang KY, Lo NN, Chia KU, Chong HC. Computer-assisted minimally invasive total knee arthroplasty compared with standard total knee arthroplasty. A prospective, randomized study. The Journal of bone and joint surgery. American volume 2008;90-1:2-9.

25. Cheng T, Pan XY, Mao X, Zhang GY, Zhang XL. Little clinical advantage of computer-assisted navigation over conventional instrumentation in primary total knee arthroplasty at early follow-up. The Knee 2012;19-4:237-45.

26. Dexel J, Kirschner S, Gunther KP, Lutzner J. Agreement between radiological and computer navigation measurement of lower limb alignment. Knee Surg Sports Traumatol Arthrosc 2014;22-11:2721-7.

27. Hauschild O, Konstantinidis L, Baumann T, Niemeyer P, Suedkamp NP, Helwig P. Correlation of radiographic and navigated measurements of TKA limb alignment: a matter of time? Knee Surg Sports Traumatol Arthrosc 2010;18-10:1317-22.

28. Babazadeh S, Dowsey MM, Bingham RJ, Ek ET, Stoney JD, Choong PF. The long leg radiograph is a reliable method of assessing alignment when compared to computer-assisted navigation and computer tomography. The Knee 2013;20-4:242-9.

29. Hirschmann MT, Konala P, Amsler F, Iranpour F, Friederich NF, Cobb JP. The position and orientation of total knee replacement components: a comparison of conventional radiographs, transverse 2D-CT slices and 3D-CT reconstruction. J Bone Joint Surg Br 2011;93-5:629-33.

30. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1-8476:307-10.

31. Nilsdotter AK, Toksvig-Larsen S, Roos EM. Knee arthroplasty: are patients' expectations fulfilled? A prospective study of pain and function in 102 patients with 5-year follow-up. Acta orthopaedica 2009;80-1:55-61.

32. Fehring TK, Odum SM, Troyer JL, Iorio R, Kurtz SM, Lau EC. Joint replacement access in 2016: a supply side crisis. The Journal of arthroplasty 2010;25-8:1175-81.

19

Chapter 2

Preliminary experience with the patient-specific templating

total knee arthroplasty. 40 cases compared with a

matched control group

Boonen B., Schotanus M.G.M, Kort N.P. Acta Orthop. 2012 Aug;83(4):387-93.

Chapter 2

20

ABSTRACT

Background and purpose Patient-specific templating total knee arthroplasty (TKA) is a new method for alignment of a total knee arthroplasty that uses disposable guides. We present the results of the first 40 consecutive patients who were operated on using this technique. Methods In this case control study we compared blood loss, operation time, and alignment of 40 TKA’s using a patient-specific templating alignment technique with values from a matched control group, who were operated on by conventional intramedullary align-ment technique. Alignment of the mechanical axis of the leg and flexion/extension and varus/valgus of the individual prosthesis components were measured on digital, stand-ing, long-leg and standard lateral radiographs. Fraction of outliers, > 3˚, was deter-mined. Results Mean mechanical axis of templating TKA’s was 181°, with a fraction of outliers of 0.3 and mean mechanical axis of conventional TKA’s was 179˚, with outliers fraction 0.5. Fraction of outliers in the frontal plane for femoral components was 0.05 in the tem-plating TKA’s and 0.4 in the conventional TKA’s. For tibial components corresponding values were 0.2 and 0.2. Fraction of outliers in the sagittal plane was 0.4 and 0.9 for femoral components and 0.4 and 0.6 for tibial components in the templating TKA’s and conventional TKA’s, respectively.

Mean operation time was 10 min shorter and blood loss was 60 mL less for templat-ing TKA than for intramedullary aligned TKA’s. Interpretation Patient-specific templating showed improved accuracy of alignment and a small reduc-tion in blood loss and operating time compared to intramedullary aligned TKA but the fraction of outliers was relatively high. Larger RCT’s are needed for further evaluation of the technique and to define the future role of patient specific template alignment tech-niques for TKA.

Preliminary Experience with Patient-matched Positioning guides in TKA Surgery

21

INTRODUCTION

Nowadays there are several methods for alignment of a total knee arthroplasty (TKA). These alignment methods can be divided into conventional techniques and navigational or image-guided surgery.

Complications associated with conventional techniques that use intramedullary alignment rods include extra blood loss perioperative1, embolization of medullary con-tents2,3 and inability of intramedullary rod passage due to deformity, retained hardware or pathologic bone disease4.

An important factor influencing implant survival is the alignment of the mechanical axis; malalignment is associated with poorer survivorship5-8, substantial change in pres-sure distribution9 and change in total load in the medial and lateral compartments of the tibial component10. Computer navigation has been developed to improve implant and limb alignment and instability in conventionally placed prosthesis11. A recent meta-analysis showed that malalignment of the mechanical axis of more than 3° occurs in one third of conventional TKA patients. In contrast, only one tenth of computer assisted TKA’s resulted in mechanical axis malalignment of more than 3°12.

However, peroperative navigation has some major drawbacks. They include the need for accurate landmark registration13, increased surgical time and cost13,14, pin loosening and bone fractures13, complexity14, long set-up time14 and a substantial learn-ing curve13.

Recently, a patient specific alignment guide, SignatureTM Personalized Patient Care (SPPC) (Biomet, Inc., Warsaw, IN) was developed, based on a preoperative MRI-scan of the patient’s leg. With this alignment guide the intramedullary cavity is not opened, thus eliminating the risks associated with it. In addition, the new technique theoretically eliminates most of the disadvantages of intraoperative navigation.

We present the preliminary results of our first 40 consecutive cases operated with this new technique and compared them with results from a matched control group, operated using conventional intramedullary alignment technique. We expected opera-tion time and degree blood loss to be lower in the SPPC group. Alignment, in terms of fraction of outliers, was expected to be superior in the SPPC group, compared to the conventional intramedullary alignment technique.

PATIENTS AND METHODS

40 knees in 39 patients (25 women) with osteoarthritis were operated on by means of the SPPC procedure between December 2009 and March 2010 and were eligible for inclusion in this case control study. We excluded patients with a BMI above 35, patients with a history of osteotomy and patients with metal near the knee joint.

Chapter 2

22

Preoperative MRI-scanning of the hip, knee and ankle was performed 6 weeks be-fore surgery according to a standard scanning protocol to determine the mechanical axis of the leg.

Software (Materialise NV, Leuven, Belgium) was used to create virtual 3-D models of the femur and tibia. Then the program was used to determine appropriate implant size and optimal position of the prosthesis (Vanguard™ Complete Knee System, Biomet, Inc., Warsaw, IN). Component sizing was determined by measuring the AP dimension of the distal femur and the contour of the proximal tibia. Planned implant size was the best fitting size of a range of 10 standard femoral and tibial components of the Vanguard Knee System. Position of the prosthesis was calculated to obtain a neutral mechanical axis and a neutral position of femoral and tibial components relative to the mechanical axis of femur and tibia in the frontal plane. In the sagittal plane, posterior slope of the tibial component and flexion of the femoral component was set at 3 degrees. Femoral rotation was set parallel to the transepicondylar axis in the coronal plane. Rotation of the tibial component could not be calculated preoperatively by software. A digital, virtual plan of the operation to be performed was sent to the surgeon. For the femoral side, the plan shows the templated size, anteroposterior (AP), mediolateral and bottom views with and without the implant; a visual angle overview of the femur, summary tables of the angles and levels of resection and visuals showing areas of overhang. For the tibial side, the plan shows visuals of AP, mediolateral, lateral meniscus and top views of the tibia with planned resection and summary tables of the angle and level of resection. The surgeon was able to adjust the digital plan by changing implant size and position (rota-tion and translation), inclination/posterior slope and resection level. However, we did not make any changes to this proposed plan and only approved the calculations provided by the software.

The patient-specific, disposable guides, made of polyamide (Figure 1) were then manufactured.

SPPC guides and an overview of the surgical plan were delivered to our hospital 2-3 days prior to surgery.


23

Figure 1. SPPC alignment guide for femur and tibia

In all patients, a standard midline incision with a medial parapatellar arthrotomy was performed and standard exposure of the femur and tibia was carried out with patella eversion. The anterior cruciate ligament was sacrificed and the posterior cruciate liga-ment was preserved in all patients. The patient-specific guides were placed on the fe-mur and tibia (Figure 2 and 3) in the fitting position with the osteophytes still in place. Calculations for the fitting of the guides are made considering these osteophytes. Fur-thermore, we paid attention to whether or not there was a mismatch between the guides and the articular surface of the knee. When there was a mismatch, this was reg-istered on the operative record form.

Figure 2. Placement of guide on femur Figure 3. Placement of guide on tibia

The femoral guide allowed the surgeon to place pins for both the distal cutting block and the 4-in-1 cutting block to make distal, anterior, posterior and chamfer cuts. The tibial guide was used to place pins for the horizontal cut block of the knee system. After pin placement, levels of resection could be adjusted in steps of 2 mm and size of the

Chapter 2

24

femoral component could be changed as this mechanism is incorporated in the distal and horizontal cut blocks of the standard Vanguard Knee System. The bone cuts were made using traditional saw blades. A trial femoral and tibial component of the prosthe-sis was inserted to check whether the position was adequate. Rotational alignment of the tibial component was performed using an extramedullary rod, this was pointed towards the second metatarsal bone. Drill holes corresponding to design of the definite prosthesis were made in the distal femur and proximal tibia using these trial compo-nents.

Blood loss was registered at this moment and a tourniquet was inflated to 350 mmHg prior to extensively rinsing the knee with a pulse lavage system. A total knee system was placed. In 3 cases the patellae were resurfaced. Soft tissue releases medially or laterally were performed if necessary. A Bellovac ABT drain (Astratech, Mölndal, Sweden) was placed and the arthrotomy was closed in layers.

An operative record was completed, containing information on operation time (min from incision until the bandage was placed), blood loss (mL of blood in the suction de-vice, prior to application of a tourniquet and prior to rinsing the knee), size of compo-nents and polyethylene insert. Blood loss and operation time were recorded by an in-dependent OR-nurse.

Femoral nerve block was used in all patients and the catheter was removed on the second day after surgery. Oral analgesics were administered according to the standard pain protocol. Arixtra 5mg/mL, 0.5 mL (GlaxoSmithKline) was used as thrombo-embolic prophylaxis for 5 weeks after surgery. All patients participated in a rapid recovery pro-gram (Joint Care, Biomet, Inc., Warsaw, IN). Criteria for discharge of patients were: dry wound, flexion of the knee up to at least 90 degrees and ability to climb stairs.

40 patients who had been operated on by the same surgeon between April 2008 and July 2009, using the conventional technique, were matched to the SPPC group. The matching was done on the following variables: type of implant (Vanguard), patient treated in rapid recovery program, sex and age. Matching was done by searching the operative record used in our hospital. The operative procedure was identical to that in SPPC patients, except for pin placement for the cut blocks which was done using stand-ard intramedullary alignment guides. Femoral components were placed in 3° valgus relative to the anatomical axis of the femur. Tibial components were placed perpen-dicular to the anatomical axis of the tibia. Femoral component flexion and tibial posteri-or slope were set on 0°.The operative record that was completed was the same as in the SPPC group. Additionally, BMI was calculated for selected patients to check for rele-vant differences between the two groups.

Mean operation time and mean blood loss in the SPPC group were compared to the corresponding values in the matched control group. Data were obtained from the oper-ative record.


25

Default sizes of the femoral and tibial components and polyethylene insert as calcu-lated by software prior to surgery were compared to actually placed femoral and tibial component sizes and the polyethylene thickness in the SPPC group.

Digital standing, weight bearing, AP long-leg radiographs were taken preoperatively and 6 weeks postoperatively in the SPPC group. We asked patients in the control group to return to the hospital for digital long-leg radiographs. Standard lateral radiographs were already available for this control group. The mechanical axis was determined ac-cording to Tigani et al.15 (Figure 4A) and was measured using calibrated software. Devia-tions of more than 3° from a neutral mechanical axis were regarded as outliers and fractions were calculated.

The varus/valgus position of the femoral and tibial component (frontal femoral component (FFC) angle and frontal tibial component (FTC) angle), relative to the me-chanical axis, was measured on the same long-leg radiographs (Figure 4B). Values of more than 90 degrees indicated valgus positioning of the femoral and tibial component and values less than 90 degrees indicated varus positioning. Fractions of outliers of more than 3° varus or valgus were calculated.

Chapter 2

26

Figure 4A. long leg radiograph. Mechanical axis of the leg measured on standing long-leg radiograph as the medial angle between the mechanical axis of the femur, defined as the line between the centre of the hip andthe centre of the femoral component, and the mechanical axis of the tibia, defined as the line between thecentre of the tibial component and the centre of the ankle. 4B. long leg radiograph. FFC and FTC angles measured as the medial angles between the femoral compo-nent and the mechanical axis of the femur and between the tibial component and the mechanical axis of thetibia, respectively.

Femoral component flexion and tibial component posterior slope (lateral femoral com-ponent (LFC) angle and lateral tibial component (LTC) angle, respectively) were meas-ured on standard lateral radiographs according to Tigani et al. 15 (figure 5). Fraction of outliers of more than 3° were calculated. Lateral radiographs were taken 6 weeks post-operatively with the knee in a slight degree of flexion.


27

Figure 5. Lateral radiograph showing the LFC and LTC angles. LFC-angle defined as the anterior angle between the femoral component and the anterior cortex of the femur. LTC-angle defined as the anterior angle be-tween the tibial component and the posterior cortex of the tibia.

All radiographical measurements were performed by an independent reviewer, who was blind regarding the treatment groups. Three measurements were conducted sepa-rately to obtain intra-observer reliability by calculating the intra-class correlation coeffi-cient.

STATISTICS

Student’s t-test was performed to compare blood loss, operation time and alignment of mechanical axis and individual femoral and tibial components between the SPPC group and the matched control group. Non-parametric Mann-Whitney Test was used to com-pare fraction of outliers between groups.

P-values below 0.05 were considered significant. SPSS-software was used for statistics (SPSS 14 Inc., Chicago, Illinois).

Chapter 2

28

RESULTS

Patients in the SPPC group (N=40) were adequately matched to a control group (N=40) with respect to age, sex and operative procedure. Mean age was 68 years in both groups and there were 25 women in each group. Mean BMI indices for both groups indicated an overall overweight (non-obese) classification of patients. Mean preopera-tive mechanical axis in the SPPC group was 175 degrees (range: 162° – 188°, SD 6.4). 31/40 had varus mechanical axis (range 162° - 178°); 9/40 had valgus mechanical axis (range 180° - 188°). No preoperative long-leg radiographs were taken preoperatively in the control group.

Operative data

Mean operation time and mean blood loss were statistically significantly lower in the SPPC group. Femoral, tibial and insert size were similar between both groups (Table 1). Table 1. Prosthesis and OR data. Mean (SD)

SPPC Conventional P-value

Femoral size 66 (4) 65 (4) 0.3

Tibial size 74 (5) 74 (5) 0.5

Insert thickness 11 (1) 12 (2) 0.2

OR time (min) 51 (11) 61 (14) 0.001

Blood loss (mL) 239 (95) 299 (115) 0.01

Actual femoral component size, tibial size and insert thickness were all statistically sig-nificantly different from default size and thickness in the SPPC group (Table 2). In 8 cases resections were altered peroperatively for the femoral component and in 10 cases they were altered for the tibial component. Insert thickness was changed from standard 10 mm to 12 mm in 16 cases and to 14 mm in 6 cases.

Individual guides fitted well on the native bone and cartilage. Table 2. Default setting and OR data. Mean (SD)

Default OR data P-value

Femoral size 67 (5) 66 (4) 0.01

Tibial size 72 (6) 74 (5) 0.001

Insert thickness 10 (0) 11 (1) < 0.001


29

RADIOGRAPHICAL EVALUATION

Mechanical axis

In 2 radiographs in the SPPC group, measurements could not be performed. The ankle joint was not visible and the mechanical axis could therefore not be determined. 35 patients in the conventional group returned for long-leg radiographs.

Mean values of mechanical axis, range and number of outliers were compared be-tween groups and intra-class correlation coefficients were obtained for all measure-ments (Table 3). Fraction of outliers was not statistically significantly different in either group: 0.3 in the SPPC-group versus 0.5 in the conventional group. Table 3. Mechanical axis

SPPC (N=38) Conventional (N=35) P-value

Mean (SD) 181° (4) 179° (3) 0.02

Range 171° - 188° 175° - 185°

Number of outliers > 3° 11/38 16/35 0.1

ICC 0.99 0.99

FFC angle and FTC angle

In 2 radiographs in the SPPC-group measurements could not be performed because the ankle joint was not visible and in 1 radiograph of the conventional group measurements were impossible because of over projection of femoral and tibial component.

Mean values, range and number of outliers were calculated and compared between groups (Table 4). There was a statistically significant difference in fraction of outliers of FFC angle in favour of the SPPC group. No such significant difference could be found for the FTC angle. Table 4. FFC and FTC


FFC Mean (SD) FFC Range FFC number of outliers > 3° FFC ICC

90° (2) 84° - 93° 2/38 0.99

88° (2) 85° - 92° 12/34 0.82

< 0.001 0.001

FTC Mean (SD) FTC Range Number of outliers > 3° FTC ICC

91° (2) 87° - 96° 7/38 0.99

91° (2) 87° - 95° 7/34 0.99

0.7 0.8

Chapter 2

30

LFC angle and LTC angle

One radiograph in the SPPC group and two in the conventional group were of inac-ceptable quality to perform measurements.

Mean values, range and number of outliers were calculated and compared between groups (Table 5).

There was a statistically significant difference in fraction of outliers for both the LFC and LTC angle, both in favour of the SPPC group. Table 5. LFC and LTC.


LFC Mean (SD) LFC Range LFC Number of outliers > 3° LFC ICC

85° (4) 74° – 94° 16/39 0.99

84° (3) 79° – 89° 33/38 0.99

0.1 < 0.001

LTC Mean (SD) LTC Range LTC Number of outliers > 3° LTC ICC

94° (4) 87° – 102° 14 0.99

87° (3) 75° – 92° 21 0.99

< 0.001 0.02

DISCUSSION

Alignment in the frontal plane

Optimal alignment in the frontal plane has generally been considered to be within 3˚ varus/valgus of the mechanical axis5-7,16. More recently, however, it has been hypothe-sized that the 3° range for optimal alignment is an arbitrary figure and that it is more likely that any deviation from neutral will reduce longevity by an amount that is propor-tional to the malalignment17. Furthermore, a distinction has to be made between resto-ration of a neutral mechanical axis and the optimal position of the individual components relative to this mechanical axis. Ideally, the position of the femoral and tibial components is perpendicular to the mechanical axis of the femur and tibia, respectively17.

Mean mechanical axis in our SPPC series was 181° (± 4°). We observed outliers of more than 3˚ varus/valgus in 12/40 patients. The fraction of outliers was not statistically significantly lower in the SPPC group than in the conventional group. This was not what we had expected. Several explanations can be given for this observation. This cohort consisted of the first 40 consecutive patients who were operated on with this new technique and it is therefore likely that there was a learning curve. Particularly for the tibial component it takes a number of cases before a surgeon learns how to remove soft tissues and how to adequately position the guides on the native bone. However, the outliers were more or less evenly distributed among the SPPC cohort. In itself, a learn-


31

ing curve would therefore be an insufficient explanation. We hypothesize that the high-er than expected fraction is probably due, at least in part, to the fact that 5 patients were unable to fully extend the knee when long-leg radiographs were taken. This may result in possible inaccuracies in measurement5,18-20.

In a meta-analysis, Mason et al. reported on results of mechanical axis alignment outcomes for navigated and conventional techniques in TKA12. A malalignment of the mechanical axis of greater than 3° occurred in 9% of patients in the navigated TKA group as opposed to 32% of patients in the conventional group. We expected our re-sults to be comparable to peroperatively navigated TKA but the fraction of outliers dif-fered substantially from results reported by Mason et al.

For individual femoral and tibial components, there was a higher proportion of malalignment in the conventional group than in the SPPC group. Mason et al. reported similar observations when comparing peroperatively navigated TKA with conventional TKA. Malalignment of the femoral component was observed in 5% of navigated TKA’s and tibial component malalignment was observed in 4%. We expected our fraction of outliers to be comparable to that in peroperatively navigated TKA, but there was a higher fraction, mainly for the tibial component. As stated before, we believe that this was mainly the result of the learning curve. However, given these results, low accuracy of the planning software must also be considered as one of the causes of malalignment.

Alignment in the sagittal plane

We calculated the outliers in alignment considering an ideal femoral component flexion and a tibial component posterior slope of both 3° in the SPPC group. In the conventional group, the outliers were calculated considering an ideal femoral component flexion and tibial component posterior slope of 0°, as the intramedullary system is designed to give this alignment.

The fraction of outliers of the femoral and tibial component of more than 3° was statistically significantly lower for the SPPC group than for the conventional group.

We observed an overall high fraction of outliers in the sagittal plane. As hypothe-sized by others this could be the result of the high variability of the femoral cortex15. However, this combined with our observation of a higher than expected fraction of outliers in the mechanical axis and in the position of the tibial component in the frontal plane in our series, means that we must again raise the question of accurateness of the software and production process to create the SPPC guides. The producer of the guides claims that there is an extremely high level of accuracy, however. We have set up a study to further compare postoperatively achieved alignment with the alignment in the digital plan as calculated with software. Measurements will be performed using CT-scan, as the this is considered to be the most accurate method for measuring lower limb alignment21.

Chapter 2

32

Also of interest is our observation that default femoral and tibial size and insert thickness differed from actually placed sizes and thicknesses. The software calculates resections for placement of a 10 mm insert and we did not make changes to these cal-culations preoperatively. The difference in calculated and placed polyethylene thickness can be explained by the fact that in some cases the resection level was changed in-traoperatively. The software calculates very conservative bone cuts and in some cases this would have resulted in a resection level through very sclerotic bone with a higher chance of malalignment due to deviation of the sawing blade. In these cases adjust-ments were made for an extra resection with an accompanying need for thicker PE insert. Unfortunately we were unable to obtain data on the exact number of cases in which adjustments were made peroperatively. We aim to address this issue further in the study that is currently being set up.

Also for the sizing of the components, we did rely on the preoperative plan without making changes to it. Difference in size of the femoral component is probably due to the fact that the software overestimates the size because calculations try to avoid notching of the femoral component in any case. For the tibial component, the software tries to avoid overhang at all times and this is most likely the explanation of why a larger tibial component could be placed than was calculated by software.

Although adjustments sometimes had to be made intra-operatively reasons given above, we had to use less soft tissue balancing techniques in the SPPC-group than in the conventional group. Only in extreme varus or valgus deformities, were soft tissue re-leases deemed necessary.

It’s important to consistently check the digital pre-operative plan and to make ad-justments to it where appropriate.

Blood loss and operation time Blood loss was 60 mL less in the SPPC group. The decrease is most likely due to not having to open the intramedullary canal of the femur and tibia and the shorter opera-tion time in the SPPC group.

The difference in operation time was 10 min because of the fewer surgical steps needed to implant the knee. In our experience, additional time can be saved when one also considers the time needed to install instruments on sterile fields, because fewer instruments are needed to perform the operation.

Limitations and strengths of the study Limitations of our study are the relative low number of patients included and the align-ment measures that were performed by only 1 independent reviewer with the known pitfalls and inaccuracies of measurements performed on standing long-leg radiographs. Restrictions of retrospective data analysis naturally apply.

The strength of our study was the direct comparison of outcomes in the SPPC group with a recent well-matched control group, operated on by the same surgeon.


33

Future research should focus on investigating alignment with more reliable tech-niques (CT-scan) to make sure that the surgeon can rely on the digital preoperative plan, the clinical outcome of the system, additional expenses and costs saved with the procedure. Larger randomized controlled trials to compare conventional intramedullary alignment with pre-operative navigation in TKA are therefore needed.

Chapter 2

34

REFERENCES

1. Raut VV, Stone MH, Wroblewski BM. Reduction of postoperative blood loss after press-fit condylar knee arthroplasty with use of a femoral intramedullary plug. J Bone Joint Surg Am 1993;75-9:1356-7.

2. Fahmy NR, Chandler HP, Danylchuk K, Matta EB, Sunder N, Siliski JM. Blood-gas and circulatory changes during total knee replacement. Role of the intramedullary alignment rod. J Bone Joint Surg Am 1990;72-1:19-26.

3. Caillouette JT, Anzel SH. Fat embolism syndrome following the intramedullary alignment guide in total knee arthroplasty. Clin Orthop Relat Res 1990-251:198-9.

4. Dennis DA, Channer M, Susman MH, Stringer EA. Intramedullary versus extramedullary tibial alignment systems in total knee arthroplasty. J Arthroplasty 1993;8-1:43-7.

5. Jeffery RS, Morris RW, Denham RA. Coronal alignment after total knee replacement. J Bone Joint Surg Br 1991;73-5:709-14.

6. Lotke PA, Ecker ML. Influence of positioning of prosthesis in total knee replacement. J Bone Joint Surg Am 1977;59-1:77-9.

7. Ritter MA, Faris PM, Keating EM, Meding JB. Postoperative alignment of total knee replacement. Its effect on survival. Clin Orthop Relat Res 1994-299:153-6.

8. Bargren JH, Blaha JD, Freeman MA. Alignment in total knee arthroplasty. Correlated biomechanical and clinical observations. Clin Orthop Relat Res 1983-173:178-83.

9. Hsu RW, Himeno S, Coventry MB, Chao EY. Normal axial alignment of the lower extremity and load-bearing distribution at the knee. Clin Orthop Relat Res 1990-255:215-27.

10. Werner FW, Ayers DC, Maletsky LP, Rullkoetter PJ. The effect of valgus/varus malalignment on load distribution in total knee replacements. J Biomech 2005;38-2:349-55.

11. Beringer DC, Patel JJ, Bozic KJ. An overview of economic issues in computer-assisted total joint arthro-plasty. Clin Orthop Relat Res 2007;463:26-30.


13. Lombardi AV, Jr., Berend KR, Adams JB. Patient-specific approach in total knee arthroplasty. Orthopedics 2008;31-9:927-30.

14. Radermacher K, Portheine F, Anton M, Zimolong A, Kaspers G, Rau G, Staudte HW. Computer assisted orthopaedic surgery with image based individual templates. Clin Orthop Relat Res 1998-354:28-38.

15. Tigani D, Busacca M, Moio A, Rimondi E, Del Piccolo N, Sabbioni G. Preliminary experience with electro-magnetic navigation system in TKA. Knee 2009;16-1:33-8.

16. Archibeck MJ, White RE, Jr. What's new in adult reconstructive knee surgery. J Bone Joint Surg Am 2002;84-A-9:1719-26.

17. Sikorski JM. Alignment in total knee replacement. J Bone Joint Surg Br 2008;90-9:1121-7. 18. Krackow KA, Pepe CL, Galloway EJ. A mathematical analysis of the effect of flexion and rotation on appar-

ent varus/valgus alignment at the knee. Orthopedics 1990;13-8:861-8. 19. Swanson KE, Stocks GW, Warren PD, Hazel MR, Janssen HF. Does axial limb rotation affect the alignment

measurements in deformed limbs? Clin Orthop Relat Res 2000-371:246-52. 20. Hauschild O, Konstantinidis L, Baumann T, Niemeyer P, Suedkamp NP, Helwig P. Correlation of radiograph-

ic and navigated measurements of TKA limb alignment: a matter of time? Knee Surg Sports Traumatol Ar-throsc 2010;18:1317–1322.

21. Chauhan SK, Clark GW, Lloyd S, Scott RG, Breidahl W, Sikorski JM. Computer-assisted total knee replace-ment. A controlled cadaver study using a multi-parameter quantitative CT assessment of alignment (the Perth CT Protocol). J Bone Joint Surg Br 2004;86-6:818-23.

35

Chapter 3

Patient-specific positioning guides for total knee arthroplasty:

no significant difference between final component alignment and preoperative

digital plan except for tibial rotation

Boonen B., Schotanus M.G.M., Kerens B., Hulsmans F-J., Tuinebreijer W.E., Kort N.P. Knee Surg Sports Traumatol Arthrosc. 2015 Jun 9.

Chapter 3

36

ABSTRACT

Purpose To assess whether there is a significant difference between the alignment of the indi-vidual femoral and tibial components (in the frontal, sagittal and horizontal planes) as calculated pre-operatively (digital plan) and the actually achieved alignment in vivo obtained with the use of patient specific positioning guides (PSPG) for TKA. It was hy-pothesised that there would be no difference between post-op implant position and preop digital plan. Methods Twenty-six patients were included in this non-inferiority trial. Software permitted matching of the pre-operative MRI-scan (and therefore calculated prosthesis position) to a pre-operative CT-scan and then to a post-operative full-leg CT-scan to determine deviations from pre-op planning in all 3 anatomical planes. Results For the femoral component, mean absolute deviations from planning were 1.8° (SD 1.3), 2.5° (SD 1.6) and 1.6° (SD 1.4) in the frontal, sagittal and transverse plane, respec-tively. For the tibial component, mean absolute deviations from planning were 1.7° (SD 1.2), 1.7° (SD 1.5) and 3.2° (SD 3.6) in the frontal, sagittal and transverse plane, respec-tively. Absolute mean deviation from planned mechanical axis was 1.9 degrees. The a priori specified null hypothesis for equivalence testing: the difference from planning is greater than 3 or lower than -3 was rejected for all comparisons except for the tibial transverse plane. Conclusion PSPG was able to adequately reproduce the pre-op plan in all planes, except for the tibial rotation in the transverse plane. Possible explanations for outliers are discussed and highlight the importance for adequate training of surgeons before they start using PSPG in their day by day practise.

Patient-matched Positioning Guides in TKA: Is The Pre-operative Plan Reproducible?

37

INTRODUCTION

Total knee arthroplasty (TKA) is one of the most successful operative procedures with both excellent short-term survival rates1 and long-term survival. These rates vary be-tween 91% and 95% for a reported follow up of 15 to 23 years2-6. Besides the design of the prosthesis, the technical aspects (i.e. surgical skills) are key success elements. Out-lining of the prosthesis is one of these essential elements. Although several techniques for outlining of the prosthesis exist nowadays, there is still room for techniques that try to optimize both the efficacy of the operative procedure itself and the accuracy of pros-thesis alignment. Moreover, with the growing need for joint replacement surgery, changes will have to be made in multiple areas of the process to meet the future grow-ing demands7. A relatively new development to improve alignment and optimize the operative process is the use of patient-specific positioning guides (PSPG) to determine the appropriate 3-dimensional resections of femur and tibia in preparation of prosthesis placement. This technique is either MRI or CT based, meaning that either of these imag-ing techniques are used to create a pre-operative, 3D image of the individual patient’s knee. These images are subsequently used to calculate ideal implant position using predetermined reference axes and planes. A preoperative plan is created this way, showing expected implant positioning after surgery. The purpose of patient-specific positioning guides is to create guides that have only one fitting position on the native anatomy of the individual patient. These guides serve as peroperative guiding instru-ments to place the pins needed to make the bony resections. Most major orthopaedic companies have launched a PSPG system throughout the last years. All use the same basic principles but have specific algorithms.

Theoretically, this method of alignment would help improve implant positioning, would eliminate variability among different surgeons and optimise the efficacy of sur-gery. Numerous reports have been published addressing alignment obtained with PSPG. Most recent studies compare the final alignment of the prosthesis with results obtained by conventional instruments using standard radiographs, long-leg radiographs or 2D CT-scans. Moreover, these studies use reference axes and reference points on postopera-tive CT-scans to measure alignment of the prostheses components that are not all equal to the reference points and axes used to calculate implant position and subsequently resulting in the preoperative digital plan (different manufacturers use different calcula-tion algorithms based on different reference axes and points). However, when using such a methodology, potential bias exists and it cannot be determined to what extend the technique is capable of reproducing the pre-operative digital plan that forms the basis of this alignment technique. This step is essential when searching for potential weak spots associated with PSPG, but are not addressed in current literature. To our knowledge, no studies have yet been conducted that directly compared ultimate im-plant position in all three anatomical planes obtained with PSPG to the pre-operative

Chapter 3

38

digital plan that should have dictated this post-operative alignment using a reliable 3D technique.

This study was designed to address the following research question: to assess whether there is a significant difference between the alignment of the individual femo-ral and tibial components (in the frontal, sagittal and horizontal planes) as calculated pre-operatively (digital plan) and the actually achieved alignment in vivo of an experi-enced PSPG user. It was hypothesised that there would be no significant difference between the preoperative alignment as determined by software and the ultimate posi-tion of the prosthesis in vivo (H1 hypothesis of equivalence).

MATERIALS AND METHODS

Prior to this study, the operating surgeon had experience with over 200 TKA performed using PSPG. For this study 26 patients were included. Inclusion criteria were: painful and disabled knee joint resulting from osteoarthritis, ability and willingness to follow instruc-tions, including control of weight and activity level.

Exclusion criteria were: failure of previous joint replacement; pregnancy; previous knee surgery, except for arthroscopic meniscectomy; metal near knee joint (MRI-scan not possible); not able or willing to undergo MRI-scan and CT-scan.

The cohort consisted of 13 women and 13 men with an average age of 66 (range 52-83 years). All eligible patients were approached to participate in this study from June 2011 and onwards. Patients received oral and written information and when informed consent was signed, patients were included in this study. The first 26 consecutive pa-tients that gave informed consent were included. Last patient was included December 2011.

Preoperative MRI-scanning of the hip, knee and ankle was performed 6 weeks prior to surgery according to the standard Signature scanning protocol. Software (Mimics, Materialise NV, Leuven, Belgium) was used to create virtual 3-dimensional models of femur and tibia. The program was used to determine appropriate implant size and posi-tioning of the knee prosthesis (Vanguard™ Complete Knee System, Biomet, Inc., War-saw, IN) for each patient individually. For the purpose of this study, an additional full-leg CT-scan of the ipsilateral leg was performed pre-operatively (radiation dose for single CT-scan: 5,69 mSv = equivalent to half the dose of a CT-pelvis or CT-thorax). This scan was made according to a standardised scanning protocol.

A digital, virtual plan of the proposed peroperative positioning was sent to the sur-geon. The surgeon was able to adjust the digital plan when deemed necessary. After approving the digital plan, guides for peroperative use were manufactured using a rapid prototype engineering technique. Operative procedure using these guides was de-scribed in detail by Boonen et al8. Intraoperatively, the practical form and fit of the


39

guides and all peroperative changes from the pre-operative plan (level of resection, size of prosthesis) were registered.

Six weeks postoperative, the full leg CT-scan of the operated extremity was sched-uled according to the same standardised scanning protocol as to which the preopera-tive CT-scan was made. In order to define the difference between the preoperative digital planning and postoperatively achieved alignment results, the post-operative CT-scan should need to be compared with the digital plan based on the preoperative MRI-scan. However, direct comparison between this preoperative MRI-scan and the postop-erative CT-scan is difficult and inaccurate as the MRI-scan is a local knee scan and matching would therefore be difficult postoperatively as a great deal of referencing points have disappeared with prosthesis placement.

To resolve this problem, the preoperative full-leg CT-scan was made next to the preoperative MRI scan to serve as an intermediate step in the registration process. The MR images could be matched to the preoperative CT-images as digital 3D models of both scans were generated for the femur and tibia using the 3-matic software of Mate-rialise NV (Materialise NV, Leuven, Belgium). Advantage of this way of 3D registration is that it makes the results independent of scan orientation and leg position during scan-ning. After surgery, 3D reconstruction femur and tibia models of the post-operative CT-scan could be superimposed onto the pre-operative CT models that represented pre-planned cuts and prosthesis placement. In this way, exact comparison could be per-formed between pre-operatively planned resections and prosthesis placement and ultimate resections and placement in vivo (figure 1).

Chapter 3

40

Fig. 1A – D: The figure shows the registration process for the femur (upper row) and tibia (lower row). A:Implant STL registered on post-op femur (yellow). B: Post-op femur (yellow) on pre-op femur (green transpar-ent). C: CT-femurs registered on MRI-femur (red transparent) and plan (hip point, red). D: Pre-op planned implant (red) with post-op implant (grey transparent) position.


41

Measurements performed using this technique have been reported to be substantially more accurate compared to conventional radiographs and 2D CT-scans with intra-observer reliability (ICC) ranging from 0.73 to 0.99 and inter-observer reliability (ICC) ranging from 0.89 to 0.999. Deviations (in degrees) from pre-op planning in all 3 ana-tomical planes for femoral and tibial component were determined (figure 2). Positive values indicate varus, flexion/posterior slope, exorotation and negative values indicate valgus, extension/anterior slope, endorotation deviations relative to the pre-operatively calculated position. Outliers (defined as deviations more than 3 degrees from pre-operatively planned position) were calculated in each plane and for the individual pros-thesis components. Accuracy of measurements was to within 0.1 degree.

The local ethics committee approved this prospective cohort study (institutional re-view board Atrium-Orbis-Zuyd, number: 11-T-15, date: march 2nd, 2011).

Fig. 2: figure showing an example of the post-op CT-scan images of femur (first 2 images) and tibia (last 2images) with pre-op plan superimposed (red). This registration permits measurement of rotational alignmentin the transverse plane.

STATISTICAL ANALYSIS

For this study two one-sided tests (equivalence test) was used to obtain the sample size (H1: mean of difference = 0). According to several studies, in the frontal plane a postop-erative range for alignment of the mechanical axis of the leg of maximal 3 degrees of valgus to 3 degrees of varus is acceptable. Standard deviation of difference was esti-mated to be five. This was based on our pilot study in which we found a range of seven degrees of varus to 5.4 degrees of valgus[12]. Significance level was determined at 0.05 (alfa) and power was set at 80%. According to this calculation 26 patients should be included in this prospective cohort study.

Two one-sided tests (TOSTs) will be performed to examine whether the null hypoth-esis (H0: the difference of pre-operative planning and post-operative alignment is more than +3 or -3) can be rejected. The margin is specified for either side and both one-sided tests have to be rejected to establish equivalence.

Chapter 3

42

RESULTS

Mean pre-operative mechanical axis was 0.9° varus (range: 15.3° varus – 10.1° valgus). Eighteen patients had varus knee osteoarthritis and 8 patients valgus knee osteoarthri-tis.

3D CT-scan analysis

For the femoral component in all three planes and for the tibial component in frontal and sagittal plane, there was no statistically significant difference between pre-operative planning and achieved position postoperatively in either plane, because the two one-sided tests were rejected and the alternative hypothesis of equivalence was accepted. For the tibial component in transverse plane the null hypothesis of a priori specified difference was not rejected and equivalence could therefore not be estab-lished. Results of the measurements for the femoral component are summarised in table 1 and for the tibial component in table 2. Table 1. Measurements for femoral component

Frontal plane Sagittal plane Transverse plane

Mean absolute deviation from pre-op planning (SD) 1.8° (1.3) 2.5° (1.6) 1.6° (1.4)

Mean deviation from pre-op planning +3 Mean deviation from pre-op planning -3

-3.6 2.4

-4.0 2.0

-3.2 2.8

95% Confidence interval deviation from +3 95% Confidence interval deviation from -3

-4.5; -2.8 1.5; 3.3

-5.1; -2.8 .89; 3.2

-4.1; -2.4 1.9; 3.7

Range -5° ; 4° -6° ; 5° -6° ; 3°

% Outliers > 3 degrees 7.7 19.2 3.8

P-value equivalence testing*

deviation > +3 deviation < -3

<0.001 <0.001

<0.001 0.001

<0.001 <0.001

*P-value for equivalence tests with two one-sided tests (TOSTs). If both tests are rejected the non-equivalence hypothesis is rejected and the alternative hypothesis is concluded at the 0.05 significance level.


43

The absolute mean deviation from planned mechanical axis was 1.9 degrees (range: - 4° to 7°), the two one-sided tests were rejected and the alternative hypothesis of equiva-lence was accepted. 11.5% of the values were above the threshold set as outlier for the mechanical axis.

Percentages of outliers more than 3 degrees from intended position (pre-op plan) for the femoral component were 7.7%, 19.2%, 3.8% in the frontal, sagittal and trans-verse plane respectively. Percentages of outliers more than 3 degrees from intended position (pre-op plan) for the tibial component were 3.8%, 7.7%, 23.1% in the frontal, sagittal and transverse plane, respectively. Figure 3, figure 4 and figure 5 are scatter plots in which the deviations of the femoral and tibial components form intended posi-tion (pre-op plan) are presented for all patients individually in the frontal, sagittal and transverse plane respectively. Figure 6 is a scatter plot in which the deviations from the intended neutral mechanical axis are presented for all patients individually.

Table 2. Measurements for tibial component

Frontal plane Sagittal plane Transverse plane

Mean absolute deviation from pre-op planning (SD) 1.7° (1.2) 1.7° (1.5) 3.2° (3.6)

Mean deviation from pre-op planning +3 Mean deviation from pre-op planning -3

-2.1 3.9

-3.2 2.8

-1.1 4.9

95% Confidence interval deviation from +3 95% Confidence interval deviation from -3

-2.8; -1.3 3.2; 4.7

-4.1; -2.2 1.9; 3.8

-2.9; .69 3.2; 6.7

Range -3° ; 5° -5° ; 5° -5° ; 16°

% Outliers > 3 degrees 3.8 7.7 23.1

P-value equivalence testing*

deviation > +3 deviation < -3

<0.001 <0.001

<0.001 <0.001

0.222 <0.001

*P-value for equivalence tests with two one-sided tests (TOSTs). If both tests are rejected the non-equivalence hypothesis is rejected and the alternative hypothesis is concluded at the 0.05 significance level.

Chapter 3

44

Fig. 3: scatter plot in which the deviations in degrees (Y-axis) of the femoral (in grey) and tibial (in red) compo-nents form intended position (pre-op plan) are presented for all patients individually (X-axis) in the frontal plane (varus as positive values; valgus as negative values).

Fig. 4: scatter plot in which the deviations in degrees (Y-axis) of the femoral (in grey) and tibial (in red) compo-nents form intended position (pre-op plan) are presented for all patients individually (X-axis) in the sagittal plane (flexion and posterior slope as positive values; extension and anterior slope as negative values).


45

Fig. 5: scatter plot in which the deviations in degrees (Y-axis) of the femoral (in grey) and tibial (in red) compo-nents form intended position (pre-op plan) are presented for all patients individually (X-axis) in the transverse plane (exoration as positive values; endorotation as negative values).

Fig 6: scatter plot in which the deviations from the intended neutral mechanical axis are presented for allpatients individually. Positive values indicate varus mismatch and negative values indicate valgus mismatch.

Operative data

All guides fitted well on the native anatomy of the individual patients and no conver-sions to traditional alignment techniques were necessary. Correct fit was defined as a stable fixation of the guides on the native bone and cartilage and the absence of mis-match between the contours of the cartilage/bone and the contours of the guides. During surgery, in 1 patient, an extra 2mm had to be resected from the distal femur and

Chapter 3

46

in 3 patients an additional 2mm were resected from the tibia. The sizes of 4 femoral components were adjusted one size (2 cases downsized and 2 cases upsized) and 4 tibial components were adjusted one size (2 cases downsized and 2 cases upsized) dur-ing surgery to obtain a better fit of the components.

DISCUSSION

The most important finding of the present study was that there was no statistically significant difference between pre-operative planning and achieved position postopera-tively in either plane, except for the tibial component in the transverse plane.

Mean deviations from pre-op planning were all within 3 degrees, except for the tibi-al component in the transverse plane, in which mean absolute deviation from pre-op planning was 3.2 degrees. The ranges of deviations were small in all planes (maximal deviation from planning 6°), with the exception of tibial rotation in which the maximal deviation was 16°. Furthermore, the technique showed a reliable restoring of a neutral mechanical axis in accordance with the preoperative plan. Percentages of outliers are small in all planes (ranging from 3.8% to 7.7% outliers) except for the femoral compo-nent in the sagittal plane (19.2% outliers) and for the tibial component in the transverse plane (23.1% outliers). When comparing our results to the percentages of outliers with conventional instrumentation (CI) and computer assisted surgery (CAS) in recent litera-ture, our results are comparable to results obtained with CAS (table 3). Table 3. comparison of PSPG results in our study with results of CI and CAS in recent literature

PSPG Conventional CAS

Mechanical axis 11.5% 26.9% and 28.3% * 9.5% and 12.2% *

Femur frontal plane 7.7% 15.8% to 16.2% * 4.7% to 5.1% *

Tibia frontal plane 3.8% 8.6% to 11.6% * 4.0% to 4.2% *

Femur sagittal plane 19.2% 36.6% to 41.3% * 18.6% to 19.8% *

Tibia sagittal plane 7.7% 21.8% to 31.6% * 13.6% to 23.2% *

Femur transverse plane 3.8% 14.8% † 17.1% †

Tibia transverse plane 23.1% 33.3% † 32.7% †

Comparison of percentages of outliers obtained with PSPG in our study compared to conventional instrumen-tation and computer assisted surgery (CAS). * according to Thienpont et al.26 and Cheng et al.27 † according to Cheng et al.27

There are several steps in the process of guide fabrication to prosthesis alignment that are potential sources of error. In sequential order, these include: imaging (MRI or CT scan), landmark registration, calculation algorithm for constructing the digital plan, rapid prototyping production process of guides, peroperative fit of the guides, peroper-


47

ative handling of the guides, sawing/bone cuts, cementing technique and position of the knee during cement hardening. In this study the combined potential errors from the production process up to and including the position of the knee during cement harden-ing ware assessed.

Relatively higher percentages of outliers for the femoral component in the sagittal plane and for the tibial component in the transverse plane were observed. Several ex-planations can be given for the observed outliers in these planes. The inaccuracies in the sagittal plane for the femoral component are believed to be caused mainly by the microplasty instrumentation that was used in this study. The Signature guides are de-signed to guide positioning of traditional cutting blocks and have no sawing sleeves incorporated into their design. Therefore, for making the chamfer cuts, the traditional sliding instrument was used. This instrument, that is also frequently used with conven-tional intramedullary alignment of the Vanguard knee system (Biomet inc., Warsaw, IN, USA), can sometimes not be fixed stably on the surface of the distal femoral bone resec-tion. We believe therefore, that special attention should be given to users of microplas-ty instrumentation, a sliding 4-in-1 cutting block for the femur, as the use of these in-struments might predispose to outliers in the sagittal plane. In addition, it might be that part of the outliers in this plane was caused by the sliding of the femoral guide into flexion or extension when drilling the guide into place.

As for the rotational alignment of the tibia in the transverse plane, percentages of outliers were higher than expected with 3 cases (11.5%) in which rotation deviated more than 5 degrees from planning. There are several possible explanations for this. First, the tibial guide has a tendency to slide laterally when positioned on the tibia. This might lead to tibial component placed slightly more in external rotation. Secondly, es-pecially in osteoporotic bone, the proximally drilled pin holes (that dictate rotation) can be difficult to retrieve after having performed the horizontal cut for the tibia. In that case, there is a tendency to follow the contours of the resected proximal tibia when positioning the guiding instrument for the tibial punch, resulting in preparing the proxi-mal tibia in such a way that relative exorotation of the component arises (figure 7).

Fig. 7: figure showing an example of how the contours of the proximal tibia are guiding when drill holes that dictate rotation cannot be retrieved. On the left the position of tibial plateau (green) as calculated pre-operatively. On the right the position of tibial plateau (orange) as detected postoperatively.

Chapter 3

48

Smaller deviations from planning in the frontal and sagittal plane could also be the re-sult of the sawing itself. Bäthis et al. analyzed the resection process using an accurate CT-based navigation system and found deviations related to the sawing process be-tween 0.5 and 1.0 degrees10. These cutting errors were independent of the surgeon’s experience11. Besides the above-mentioned inaccuracies resulting from the sawing itself, is the observation by Catani et al. that cementation and impaction of the final components, can introduce a considerable error (up to 3° in the sagittal plane for the tibia) in alignment, regardless of how accurately the resection planes were made12.

Given these observations, it is clear that there are some potential pitfalls that might occur during surgery itself when using PSPG that could jeopardise the guides’ potential for achieving adequate component alignment. We believe that, when positioning of the tibial guide, it is important to aim at pressing the guide on the medial part of the tibia and thus avoiding lateralisation (and with that also external rotation) of the guide. Addi-tionally, before positioning the guiding instrument for the tibial punch, the holes in the tibial plateau that determine rotation should be visible. It is advisable to use the pulse lavage to make these pinholes better visible. Furthermore, we advise against using the sliding version from the 4-in-1 cutting block for preparation of the femur and we stress the importance of a stable fixation of the guide on the femur before drilling it in place. Furthermore, the PSPG system analysed in this study is a bone referencing technique, meaning that proper ligament balancing after bony resections is absolutely mandatory.

Several authors have published their results on the alignment obtained with PSPG. Diverse systems have been subject of study and results concerning outliers in alignment with PSPG differ greatly. Higher quality studies report both superior results from PSPG with respect to the percentage of outliers of the individual components when com-pared to conventional instrumentation13,14 as comparable results15-20. These studies, however, use standard radiographs, long-leg radiographs or 2D CT-scans to assess alignment. This is substantially less accurate than 3D CT-scans, which were used in this study9. In addition, other studies use reference axes and reference points on postopera-tive imaging that are not all equal to the reference points and axes used to calculate implant position and subsequently resulting in the preoperative digital plan. This, as already stated, makes direct comparison between post-surgery alignment and proposed alignment in the digital plane difficult. Computer navigation has been used in some studies in order to overcome these limitations. Three of 4 studies 21-23 report a higher percentage of outliers (more than 3 degrees deviation from intended bone cuts) than the percentages reported in our study. The fourth study, also investigating the Signa-ture system, reports comparable percentages of outliers in the investigated frontal and sagittal plane24. Limitation of using navigation to assess guide accuracy is the inherent error with respect to landmark registration25, a limitation that was overcome with the use of our study design. Additionally, data in this study were not influenced by the posi-tion of the leg during scanning, as could be a source of error in other studies.


49

There are weaknesses in this study. Clinical outcomes were not taken into account since the study was not adequately powered to make valid conclusions on clinical out-come. The purpose of this study was to compare final alignment outcome with the preoperative plan and assembling a control group was therefore not applicable (no preoperative digital plan in control group). We therefore chose to compare our results with the literature, however, for the purpose of framing only.

Given the conflicting results on alignment with PSPG in the literature and given the mentioned potential pitfalls when using PSPG highlighted in this study, we believe that PSPG cannot automatically be seen as a technique that enables the less experienced knee surgeons to obtain optimal alignment results. We think that there are still numer-ous crucial steps in order to achieve optimal alignment results when using PSPG sys-tems. Therefore, we recommend adequate training surgeons before starting using PSPG in a day by day practise.

CONCLUSION

The results of this study indicate that overall PSPG is a reliable technique for aligning the components of a TKA and for adequately restoring a neutral mechanical axis. The ob-served inaccuracies, mainly in rotational alignment for the tibia, are explained by the cutting and prosthesis placement but mainly illustrate possible pitfalls with this tech-nique. These potential pitfalls need attention and highlight the need for adequate sur-geon training and guidance when starting with PSPG.

Chapter 3

50

REFERENCES

1. Gothesen O, Espehaug B, Havelin L, Petursson G, Furnes O. Short-term outcome of 1,465 computer-navigated primary total knee replacements 2005-2008. Acta orthopaedica 2011;82-3:293-300.

2. Ecker ML, Lotke PA, Windsor RE, Cella JP. Long-term results after total condylar knee arthroplasty. Significance of radiolucent lines. Clinical orthopaedics and related research 1987-216:151-8.

3. Ma HM, Lu YC, Ho FY, Huang CH. Long-term results of total condylar knee arthroplasty. The Journal of arthroplasty 2005;20-5:580-4.

4. Pavone V, Boettner F, Fickert S, Sculco TP. Total condylar knee arthroplasty: a long-term followup. Clinical orthopaedics and related research 2001-388:18-25.

5. Ranawat CS, Flynn WF, Jr., Saddler S, Hansraj KK, Maynard MJ. Long-term results of the total condylar knee arthroplasty. A 15-year survivorship study. Clinical orthopaedics and related research 1993-286:94-102.

6. Rodriguez JA, Bhende H, Ranawat CS. Total condylar knee replacement: a 20-year followup study. Clinical orthopaedics and related research 2001-388:10-7.

7. Fehring TK, Odum SM, Troyer JL, Iorio R, Kurtz SM, Lau EC. Joint replacement access in 2016: a supply side crisis. The Journal of arthroplasty 2010;25-8:1175-81.

8. Boonen B, Schotanus MG, Kort NP. Preliminary experience with the patient-specific templating total knee arthroplasty. Acta orthopaedica 2012;83-4:387-93.

9. Hirschmann MT, Konala P, Amsler F, Iranpour F, Friederich NF, Cobb JP. The position and orientation of total knee replacement components: a comparison of conventional radiographs, transverse 2D-CT slices and 3D-CT reconstruction. The Journal of bone and joint surgery. British volume 2011;93-5:629-33.

10. Bathis H, Perlick L, Tingart M, Perlick C, Luring C, Grifka J. Intraoperative cutting errors in total knee arthroplasty. Archives of orthopaedic and trauma surgery 2005;125-1:16-20.

11. Mahaluxmivala J, Bankes MJ, Nicolai P, Aldam CH, Allen PW. The effect of surgeon experience on component positioning in 673 Press Fit Condylar posterior cruciate-sacrificing total knee arthroplasties. The Journal of arthroplasty 2001;16-5:635-40.

12. Catani F, Biasca N, Ensini A, Leardini A, Bianchi L, Digennaro V, Giannini S. Alignment deviation between bone resection and final implant positioning in computer-navigated total knee arthroplasty. The Journal of bone and joint surgery. American volume 2008;90-4:765-71.

13. Daniilidis K, Tibesku CO. A comparison of conventional and patient-specific instruments in total knee arthroplasty. International orthopaedics 2014;38-3:503-8.

14. Silva A, Sampaio R, Pinto E. Patient-specific instrumentation improves tibial component rotation in TKA. Knee surgery, sports traumatology, arthroscopy : official journal of the ESSKA 2014;22-3:636-42.

15. Barrett W, Hoeffel D, Dalury D, Mason JB, Murphy J, Himden S. In-vivo alignment comparing patient specific instrumentation with both conventional and computer assisted surgery (CAS) instrumentation in total knee arthroplasty. The Journal of arthroplasty 2014;29-2:343-7.

16. Boonen B, Schotanus MG, Kerens B, van der Weegen W, van Drumpt RA, Kort NP. Intra-operative results and radiological outcome of conventional and patient-specific surgery in total knee arthroplasty: a multicentre, randomized controlled trial. Knee surgery, sports traumatology, arthroscopy : official journal of the ESSKA 2013;21-10:2206-12.

17. Chareancholvanich K, Narkbunnam R, Pornrattanamaneewong C. A prospective randomized controlled study of patient-specific cutting guides compared with conventional instrumentation in total knee replacement. The bone & joint journal 2013;95-B-3:354-9.

18. Roh YW, Kim TW, Lee S, Seong SC, Lee MC. Is TKA using patient-specific instruments comparable to conventional TKA? A randomized controlled study of one system. Clinical orthopaedics and related research 2013;471-12:3988-95.

19. Victor J, Dujardin J, Vandenneucker H, Arnout N, Bellemans J. Patient-specific guides do not improve accuracy in total knee arthroplasty: a prospective randomized controlled trial. Clinical orthopaedics and related research 2014;472-1:263-71.


51

20. 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 surgery, sports traumatology, arthroscopy : official journal of the ESSKA 2013;21-10:2213-9.

21. Conteduca F, Iorio R, Mazza D, Caperna L, Bolle G, Argento G, Ferretti A. Evaluation of the accuracy of a patient-specific instrumentation by navigation. Knee surgery, sports traumatology, arthroscopy : official journal of the ESSKA 2013;21-10:2194-9.

22. Lustig S, Scholes CJ, Oussedik SI, Kinzel V, Coolican MR, Parker DA. Unsatisfactory accuracy as determined by computer navigation of VISIONAIRE patient-specific instrumentation for total knee arthroplasty. The Journal of arthroplasty 2013;28-3:469-73.

23. Scholes C, Sahni V, Lustig S, Parker DA, Coolican MR. Patient-specific instrumentation for total knee arthroplasty does not match the pre-operative plan as assessed by intra-operative computer-assisted navigation. Knee surgery, sports traumatology, arthroscopy : official journal of the ESSKA 2014;22-3:660-5.

24. Seon JK, Park HW, Yoo SH, Song EK. Assessing the accuracy of patient-specific guides for total knee arthroplasty. Knee surgery, sports traumatology, arthroscopy : official journal of the ESSKA 2014.

25. Lustig S, Fleury C, Servien E, Demey G, Neyret P, Donell ST. The effect of pelvic movement on the accuracy of hip centre location acquired using an imageless navigation system. International orthopaedics 2011;35-11:1605-10.

26. Thienpont E, Fennema P, Price A. Can technology improve alignment during knee arthroplasty. The Knee 2013;20 Suppl 1:S21-8.

27. Cheng T, Zhao S, Peng X, Zhang X. Does computer-assisted surgery improve postoperative leg alignment and implant positioning following total knee arthroplasty? A meta-analysis of randomized controlled trials? Knee surgery, sports traumatology, arthroscopy : official journal of the ESSKA 2012;20-7:1307-22.

53

Chapter 4

Intra-operative results and radiological outcome of conventional and patient-specific

surgery in total knee arthroplasty: a multicentre, randomized controlled trial

Boonen B., Kerens B., Schotanus M.G.M., van der Weegen W., van Drumpt R.A.M., Kort N.P. Knee Surg Sports Traumatol Arthrosc. 2013 Oct;21(10):2206-12.

Chapter 4

54

ABSTRACT

Purpose This prospective, double-blind, randomized controlled trial was designed to address the following research questions: firstly, is there a significant difference in outliers in align-ment in the frontal and sagittal plane between PSG TKA and conventional TKA. Second-ly, is there a significant difference in operation time, blood loss and length of hospital stay between the two techniques. We hypothesise that there will be fewer outliers with PSG TKA and that operation time, blood loss and length of hospital stay can be signifi-cantly reduced with PSG. Methods One hundred and eighty patients were randomized for PSG TKA (group 1) or conven-tional TKA (group 2) in two centres. Patients were stratified per hospital. Alignment of the mechanical axis of the leg and flexion/extension and varus/valgus of the individual prosthesis components were measured on digital, standing, long-leg and standard lat-eral radiographs by two independent outcome assessors in both centres. Percentages of outliers (>3˚) were determined. We compared blood loss, operation time and length of hospital stay. Results There was no statistically significant difference in mean mechanical axis or outliers in mechanical axis between groups. No statistically significant difference was found for the alignment of the individual components in the frontal plane, nor for the percentage of outliers. There was a statistically significant difference in outliers for the femoral component in the sagittal plane, with a higher percentage of outliers in the group 1 (p = 0.017). No such significant result was found for the tibial component in that plane. All interclass correlation coefficients were good. Blood loss was 100 mL less in group 1 (p < 0.001). Operation time was 5 minutes shorter in group 1 (p < 0.001). Length of hospital stay was identical with a mean of 3.6 days (n.s.). Conclusions The results in terms of obtaining a neutral mechanical axis and a correct position of the prosthesis components did not differ between groups. A small reduction in operation time and blood loss was found with the PSG system. Future research should especially focus on cost-effectiveness analysis and functional outcome of PSG TKA.

Patient-matched Positioning Guides versus Conventional Instruments: RCT Part 1

55

INTRODUCTION

A substantial rise in demand for TKA over the past decade has been observed1. An age-ing population, increase in the prevalence of obesity and more active lifestyle of the elderly are only a few factors which contribute to the increase in demand for TKA2. Therefore, with this growing need for joint-replacement, efficacy and cost-effectiveness of joint replacement surgery should be optimized. Trying to resolve the shortcomings of existing techniques while optimizing the operative procedure has relatively recently resulted in development of patient-specific guides (PSG) for TKA. These new techniques are either MRI or CT based, meaning that either of these imaging techniques are used to calculate ideal implant position using predetermined reference axes and planes. The purpose of PSG is to create jigs that have only one fitting position on the native anato-my of the individual patient. These jigs serve as peroperative guiding instruments to make bony resections. First reports suggest better results with respect to restoration of a neutral mechanical axis of the leg, measured on long-leg radiographs, using PSG com-pared to conventional instrumentation3-6 but few randomized trials have been pub-lished that report on alignment outcome. Additionally, limited reports have been pub-lished on the clinical outcome and on the peroperative results of this new technique. This prospective, double-blind, randomized controlled trial, comparing PSG with con-ventional instrumentation, was designed to address the following research questions: Firstly, is there a significant difference in outliers in alignment in the frontal and sagittal plane between the two techniques. Secondly, is there a significant difference in opera-tion time, blood loss and length of hospital stay between the two techniques.

We hypothesised that there will be fewer outliers with PSG TKA and that operation time, blood loss and length of hospital stay can be significantly reduced with PSG.


From September 2010 till March 2013, 180 patients gave written and oral consent to participate in this prospective, randomized trial. Two hospitals (hospital 1, hospital 2) participated in the study and enrolled 90 patients each. Patients with disabling osteoar-thritis of the knee that were candidates for primary unilateral TKA were eligible for inclusion. Patients with metal near the knee, ankle or hip joint, patients with contra-indications to MRI-scan, patients that had previously undergone knee surgery (except for arthroscopic meniscectomy) and patients that refused to consent were excluded.

Randomisation was done using an online random number generator (www.randomizer.org) and was stratified per hospital. Each hospital included 90 pa-tients divided over 2 groups of 45 patients: a trial group in which a TKA was placed using PSG (group 1) and a control group in which a TKA was placed with conventional instru-

Chapter 4

56

mentation (group 2). Patients were blinded to the type of alignment method used. Baseline patient characteristics are shown in table 1. Table 1. Baseline characteristics

Group 1 Group 2 P-value

Men/Women 34/56 40/50 n.s.

Mean age (SD) 69 (8.0) 65 (8.8) 0.001

BMI 30.3 29.5 n.s.

Surgical procedure

In group 1 a preoperative MRI-scan of the hip, knee and ankle was performed 6 weeks prior to surgery according to the standard Signature scanning protocol. Software (Mate-rialise NV, Leuven, Belgium) was used to create virtual 3-dimensional models of femur and tibia. The program was used to determine appropriate implant size and optimal positioning of the prosthesis (Vanguard™ Complete Knee System, Biomet, Inc., Warsaw, IN) for each patient individually. Position of the prosthesis was calculated to obtain a neutral mechanical axis and a neutral position of femoral and tibial components to the mechanical axis in the frontal plane. In the sagittal plane, posterior slope of the tibial component and flexion of the femoral component were set at 3 degrees. A digital, vir-tual plan of the proposed peroperative positioning was sent to the surgeon. The sur-geon was able to adjust the digital plan if desired. After approving the digital plan, guides for peroperative use were manufactured using a rapid prototype engineering technique.

In group 2, standard intramedullary instrumentation was used. Patients in this group also underwent MRI-scanning according to the same scanning protocol as patients in group 1 to maintain patient-blinding to the type of alignment method used. Again, the goal was to obtain a neutral mechanical axis in all patients. Femoral component flexion and tibial posterior slope were set at 0°.

A standard medial parapatellar approach was used in both groups. All procedures were performed by experienced knee surgeons in both centres. In group 1, the PSG were placed on the articular surface after removal of any soft tissues at the site of guide position. Pins were placed using these guides that determined position of the standard cutting blocks to make the bony resections. The femur was prepared first. In all cases a cemented Vanguard knee prosthesis was placed (Vanguard™ Complete Knee System, Biomet, Inc., Warsaw, IN). Patella resurfacing was performed when deemed necessary. Soft tissue balancing was performed when necessary. In one hospital all procedures were performed under bloodless field (hospital 1); in hospital 2, a tourniquet was used only during cementing. A Bellovac ABT drain (Astratech, Mölndal, Sweden) drain was used in one of 2 hospitals (hospital 2) when deemed necessary by the operating sur-geon. At hospital 1, no drains were use.


57

In group 2, the same procedure was performed except that standard intramedullary instruments were used to guide the position of the cutting blocks.

Postoperative protocol

Analgesics were administered according to the standard pain protocol. Arixtra 5mg/mL, 0.5 mL (GlaxoSmithKline) (hospital 2) or Xarelto 10mg (Bayer) (hospital 1) were used as thrombo-embolic prophylaxis for 5 weeks post-surgery. All patients participated in an enhanced recovery program (Joint Care, Biomet NL, Dordrecht NL). Criteria for dis-charge of patients were: dry wound, flexion of the knee up to at least 90 degrees and ability to climb stairs if necessary for safe mobilization at home.

Outcome measurements

Digital standing, weight bearing, AP long-leg radiographs were taken 6 weeks postoper-atively in all patients. Measurements were performed using calibrated software (PACS) and accuracy of the method was to within 1 degree. In both groups, a neutral mechani-cal alignment was pursued and absolute deviations from this neutral alignment were measured. Values lower than 180° indicating varus knee alignment, values higher than 180° indicating valgus knee alignment. Deviations of more than 3° from a neutral me-chanical axis were regarded as outliers and percentages were calculated in both groups (Fig.1a Mechanical axis of the leg measured on standing long-leg radiograph as the medial angle between the mechanical axis of the femur, defined as the line between the center of the hip and the center of the femoral component, and the mechanical axis of the tibia, defined as the line between the center of the tibial component and the center of the ankle).

The varus/valgus position of the femoral (frontal femoral component: FFC) and tibial (frontal tibial component: FTC) component, relative to the mechanical axis, was meas-ured on the same long leg radiographs. Values lower than 90° indicating varus compo-nent alignment, values higher than 90° indicating valgus alignment of components rela-tive to the mechanical axis of femur and tibia (Fig.1b Standing long-leg radiographs showing FFC and FTC angles measured as the medial angles between the femoral com-ponent and the mechanical axis of the femur and between the tibial component and the mechanical axis of the tibia on, respectively). Percentages of outliers of more than 3° varus or valgus were calculated.

Chapter 4

58

Fig. 1 a Mechanical axis of the leg measured on standing long-leg radiograph as the medial angle between themechanical axis of the femur, defined as the line between the centre of the hip and the centre of the femoral component, and the mechanical axis of the tibia, defined as the line between the centre of the tibial compo-nent and the centre of the ankle. b Standing long-leg radiographs showing FFC and FTC angles measured asthe medial angles between the femoral component and the mechanical axis of the femur and between thetibial component and the mechanical axis of the tibia, respectively.

Femoral component flexion (lateral femoral component: LFC) and tibial component posterior slope (lateral tibial component: LTC) were measured on standard lateral radi-ographs. Values lower than 90° indicating extension/anterior slope of the components relative to the anterior cortex (femur) or posterior cortex (tibia); values higher than 90° indicating flexion/posterior slope of the components (Fig.2 Lateral radiograph showing the LFC and LTC angles. LFC-angle defined as the anterior angle between the femoral component and the anterior cortex of the femur. LTC-angle defined as the anterior angle between the tibial component and the posterior cortex of the tibia). Percentages of outliers of more than 3° were calculated. We calculated the outliers in alignment considering an ideal femoral component flexion and a tibial component posterior slope


59

of both 3° in group 1. In the conventional group, the outliers were calculated consider-ing an ideal femoral component flexion and tibial component posterior slope of 0°, as the conventional system is designed to obtain this alignment. All radiographical meas-urements were performed by 2 independent reviewers in both centers, blinded to the type of alignment method used. Interclass correlation coefficients were calculated. Test-retest reliability has been proven to be high5.

Fig. 2 Lateral radiograph showing the LFC and LTC angles. LFC angle defined as the anterior angle between thefemoral component and the anterior cortex of the femur. LTC-angle defined as the anterior angle betweenthe tibial component and the posterior cortex of the tibia.

Operation time was measured by an independent OR nurse. Time from skin incision until the bandage was placed was registered.

Blood loss was measured as total volume of blood in the suction device prior to ex-tensively rinsing the knee with pulse lavage system in one hospital (hospital 2). Addi-tionally, the weight of gauze was measured to quantify additional blood loss (1g set equal to 1 ml of blood loss). Total blood loss was obtained by summing these values. This total loss was registered by an independent OR nurse. Furthermore, the need for and number of blood transfusions was registered in every patient in both centres. Whenever blood was reinfused from the Bellovac drain, the amount was registered. Hb-

Chapter 4

60

levels were obtained at 1 and 3 days (if patients were still admitted at that time) post-operatively.

Length of hospital stay was registered as total of days that patients stayed in the ward post-surgery. The day after surgery counting as the first day and the day of dis-charge was counted as a total day of inpatient care.

This study was approved by the institutional review boards of both hospitals (File nr. 10-T-21).

Statistical analysis

This study was powered on 2 primary endpoints. One being the difference in Knee Soci-ety Score (KSS) at 2 years follow-up. The other being the difference in percentage of outliers in alignment in the coronal plane as measured on standing long-leg radiographs 6 weeks postoperative. Based on the former (KSS), 156 patients should be included. To correct for loss-to-follow-up, 180 patients were included in this study. For the other outcome, 88 patients should suffice to generate enough power for the alignment end-point.

Student’s t-test was performed to compare blood loss, operation time, length of hospital stay and alignment of mechanical axis and individual femoral and tibial compo-nents between the 2 groups. Non-parametric Mann-Whitney Test was used to compare percentages of outliers between groups.

P-values below 0.05 were considered significant. SPSS-software was used for statistics (SPSS 14 Inc., Chicago, Illinois).

RESULTS

Alignment

Four radiographs in group 1 and 8 radiographs in group 2 were of inacceptable quality to perform measurements and were therefore not included in the analysis. There was no statistically significant difference in mean mechanical axis or outliers in mechanical axis between groups (table 2). Interclass correlation was good with a Cronbach’s alpha of 0.89.

In terms of individual component alignment in the frontal plane, no statistically sig-nificant difference was found for the mean FFC and FTC between the 2 groups. Percent-ages of outliers were not statistically significantly different between both groups. Inter-class correlation was excellent for both measurements: FFC: 0.83; FTC: 0.85.

As for individual component alignment in the sagittal plane, there was a statistically significant difference in outliers for the femoral component, with a higher percentage of


61

outliers in the group 2. Interclass correlation coefficients were excellent for both meas-urements: LFC: 0.94; LTC: 0.70. Table 2. Mechanical axis

Group 1 (N=86) Group 2 (N=82) P-value

Mean (SD) 179° (2.8) 178° (2.3) n.s.

Range 173° - 185° 172° - 186°

Percentage of outliers > 3° 30% 18% n.s.

Table 3. FFC and FTC


FFC Mean (SD) FFC Range FFC percentage of outliers > 3°

89° (2.1) 82° - 93° 13%

88° (1.9) 84° - 94° 13%

n.s. n.s.

FTC Mean (SD) FTC Range Percentage of outliers > 3°

90° (2.0) 84° - 94° 9%

90° (1.7) 87° - 94° 2%

n.s. n.s.

Table 4. LFC and LTC.


LFC Mean (SD) LFC Range LFC Percentage of outliers > 3°

96° (5.0) 81° – 112° 49%

95° (3.7) 88° – 104° 65%

n.s. 0.017

LTC Mean (SD) LTC Range LTC Percentage of outliers > 3°

92° (3.1) 86° – 101° 33%

88° (2.7) 81° – 94° 28%

0.000 n.s.

Surgical data

All guides fitted well peroperatively and no PSG-procedures were switched to conven-tional instrumentation. There was a statistically significant difference in mean blood loss between groups and patients in group 2 had significantly more millilitres of blood trans-fused from the retransfusion drain than patients in group 1. No packed cells were nec-essary in either group. There was no difference in number of unnecessary retransfusion drains between groups. Length of hospital stay was not statistically significant between groups. Operation time, however, was 5 minutes less in the Signature group (p < 0.001).

A subgroup analysis (patients in hospital 1 compared to patients in hospital 2) was performed. Length of hospital stay in group 1 was significantly shorter in hospital 1 than in hospital 2: 2.9 days (SD 1.3) versus 4.3 days (SD 1.2) (p<0.001).

Chapter 4

62

Operation time in group 1 was significantly shorter in hospital 1 than in hospital 2: 42.1 min (SD 6.2) versus 46.8 min (SD 6.1) (p 0.001). Length of hospital stay in group 2 was significantly shorter in hospital 1 than in hospital 2: 2.9 days (SD 1.0) versus 4.4 days (SD 1.3) (p<0.001). Operation time in group 2 was significantly shorter in hospital 1 than in hospital 2: 41.0 days (SD 8.1) versus 57.6 days (SD 5.3) (p<0.001). Table 5: Blood loss, length of hospital stay and operation time

Group 1 Group 2 P-value

Mean Hb-preop (mmol/l)a 8.4 (1.0) 8.3 (1.1) n.s.

Mean Hb day 1 (mmol/l) a 7.5 (0.7) 7.4 (1.1) n.s.

Mean Hb day 3(mmol/l)b 6.9 (0.8) 6.7 (1.3) n.s.

Mean total blood loss (mL) b 193.2 297.9 <0.001

Number of patients with bellovacb 36/45 38/45 n.s.

Reinfusion of bellovac blood (mL) b 193.9 303.3 n.s.

Length of hospital stay (SD) a 3.6 (1.5) 3.7 (1.4) n.s.

Operation time (SD) a 44.7 (6.5) 50.0 (10.6) <0.001

a Patients of both centres included in analysis (n = 180) b Patients of only hospital 2 included (n = 90)

DISCUSSION

The most important finding of the present study was that the obtained alignment in terms of restoration of a neutral mechanical axis and alignment of the individual femo-ral and tibial components in the sagittal and frontal plane was not superior in the PSI-group. This was in contrast to what we had expected and in contrast to some of the results published by others. Ng et al. also studied the accuracy of the SignatureTM sys-tem4. They found that the overall mean hip-knee-ankle angle for PSG (180.6) was similar to manual instrumentation (181.1), but they observed fewer (± 3°) hip-knee-ankle angle outliers with PSG (9%) than with manual instrumentation (22%). Daniilidis et al. report-ed 11% outliers more than three degrees in mechanical axis with the Visionnaire PSG TKA7. Noble et al. performed a prospective randomized study comparing Visionnaire PSG to standard instrumentation3. They found that the postoperative mechanical axis was significantly closer to neutral in 15 patients in the PSG-group compared to 14 pa-tients in the conventional instrumentation group.

Literature contradicting the above and presenting results more similar to the results of our study is also published. In a study of Nam et al. 70.7% of patients had an overall alignment within 3° of a neutral mechanical axis. 87.8% had a tibial component align-ment within 2° of perpendicular to the tibial mechanical and 90.2% had a femoral com-ponent alignment within 2° of perpendicular to the femoral mechanical axis8. We agree with Nam et al. that, based on the findings of our study, alignment with PSG is less accu-


63

rate than with computer assisted surgery as a meta-analysis demonstrated that me-chanical axis malalignment of more than 3° occurred in only one tenth of computer assisted TKA’s9.

Chareancholvanich et al. recently published the results of their RCT comparing con-ventional instrumentation to PSG10. They came to the same conclusion as us that there were no significant differences between the groups in terms of tibiofemoral angle or femoral component alignment. Nunley et al.11 reported on the accuracy of 2 PSI sys-tems and compared these results to results obtained with conventional instrumenta-tion. In their study, a group of 50 patients operated on with the SignatureTM procedure was compared to a group of 50 patients operated on with conventional instrumenta-tion. The mean hip-knee angle was similar between these groups. For the zone of the mechanical axis, percentages of outliers were similar. They studied a third group of 50 patients in which a PSG system was used to restore the pre-arthritic anatomy of the patient’s knee and therefore the original kinematic axis (OtisMed). In this group, there was more valgus mean alignment and the most valgus outliers in mechanical axis. For the zone of the mechanical axis, patients had a greater number of outliers (64%) that were valgus. The findings of Nunley et al. indicate that the results of our study are not automatically applicable to other PSG systems than the SignatureTM system, especially systems that are designed not to create a neutral mechanical axis but to recreate a normal kinematical axis.

We have no concise explanation as to why the alignment in the PSG-group in our study was not superior to conventional instrumentation and why results in literature are so conflicting. Possibly, the non-randomized, non-blinded nature of most reports has something to do with this. We believe a learning curve has not played a role in our se-ries as the operating surgeons all have extensive experience with this particular PSG-system.

We have reported significantly lower blood loss and shorter operation time with the PSG-system than with the conventional system. Blood loss being 100mL less and opera-tion time being 5 minutes shorter. No significant difference was found in the need for blood transfusion or amount of retranfusion from the retransfusion drain. Our results were comparable to results reported by others. Spencer et al. analysed the results of 30 TKA’s performed with the OtisMed custom-fit technique and compared them to a matched cohort operated on using conventional alignment guides. There were no dif-ferences in blood loss and there was a mean decrease in tourniquet time of 14% com-pared to a cohort of patients with conventional knee replacements (80 ± 17 min versus 93 ± 12 min)6. Chareancholvanich et al. randomized 80 patients to undergo TKR with PSG or conventional instrumentation and found that the operating time was reduced by a mean 5.1 minutes (p = 0.019), without tangible differences in post-operative blood loss (p = 0.528) or need for blood transfusion (p = 0.789)10. Noble et al. prospective randomized study comparing Visionnaire™ (15 patients) to standard instrumentation (14 patients)3. They found a small although significant reduction in the length of hospi-

Chapter 4

64

tal stay (59.2 hours vs 66.9 hours, n.s.) and operative time (121.4min vs 128.1 min, p<0.048). They found no significant difference in blood loss (71ml vs 62.5ml, n.s.). In our study, no statistically significant difference existed in terms of length of hospital stay.

Nunley et al.12 reported on the cost-effectiveness of patient-specific cutting blocks. They retrospectively reviewed 57 patients undergoing PSI TKA and matched them with 57 patients undergoing traditional TKA. On average, TKAs performed with patient-specific instrumentation had similar tourniquet times (61.0 versus 56.2 minutes) and patients were in the operating room 12.1 minutes less (137.2 versus 125.1 minutes) than those in the standard instrumentation group. They conclude that patient-specific instrumentation for TKA may not be cost-effective in its current form. We agree with Nunley et al. that given the small reductions in blood loss, not resulting in less transfu-sion requirements, and the small reduction in operation time, the technique is not cost-effective if just looked at these items. However, the true cost saving should come from theatre efficiency. With PSG less surgical instrumentation is needed, with all its associ-ated costs like sterilization, purchase and maintenance costs. Studies are currently con-ducted to specifically look into these aspects.

Limitations of this study are that results are not automatically applicable to other PSG systems and that we studied alignment in the frontal and sagittal plane only. Tradi-tionally, standing long-leg radiographs are the gold standard for assessment of coronal alignment in TKA13. However, recently there is additional focus on alignment in the sagittal and transverse plane. CT-based evaluation of alignment in these planes would be desirable to provide full insight in the accuracy of alignment with PSG, but due to costs and radiation exposure, this was not achievable in this relatively large study popu-lation. Two studies have been published using intraoperative computer navigation as a more accurate method of measuring alignment accuracy. Both conclude that PSG (Vi-sionnaire system) generates unacceptable accuracy when assessed by computer naviga-tion and they recommend an accurate control of the alignment before making the cuts14,15.

This work illustrates that currently, PSG systems offer no clear advantage over standard intramedullary alignment systems when used in the day by day clinical prac-tise. A detailed cost-effectiveness analysis might, however, lead to reconsideration of this statement and currently studies are being conducted to address this issue.

CONCLUSION

The results in terms of obtaining a neutral mechanical axis and a correct position of the prosthesis components showed comparable results for both patient specific guiding and conventional instrumentation groups. A small reduction in operation time and blood loss was found with the PSG system, however the clinical relevance is questionable.


65

ACKNOWLEDGMENTS

The authors want to thank Jorgen A. Wullems for his work in measuring the alignment on digital standing long-leg radiographs.

Chapter 4

66

REFERENCES

1. Kurtz SM, Ong KL, Lau E, Widmer M, Maravic M, Gomez-Barrena E, de Pina Mde F, Manno V, Torre M, Walter WL, de Steiger R, Geesink RG, Peltola M, Roder C. International survey of primary and revision total knee replacement. International orthopaedics 2011;35-12:1783-9.

2. Otten R, van Roermund PM, Picavet HS. [Trends in the number of knee and hip arthroplasties: considerably more knee and hip prostheses due to osteoarthritis in 2030]. Nederlands tijdschrift voor geneeskunde 2010;154:A1534.

3. Noble JW, Jr., Moore CA, Liu N. The value of patient-matched instrumentation in total knee arthroplasty. The Journal of arthroplasty 2012;27-1:153-5.

4. Ng VY, DeClaire JH, Berend KR, Gulick BC, Lombardi AV, Jr. Improved accuracy of alignment with patient-specific positioning guides compared with manual instrumentation in TKA. Clinical orthopaedics and related research 2012;470-1:99-107.


6. Spencer BA, Mont MA, McGrath MS, Boyd B, Mitrick MF. Initial experience with custom-fit total knee replacement: intra-operative events and long-leg coronal alignment. International orthopaedics 2009;33-6:1571-5.

7. Daniilidis K, Tibesku CO. Frontal plane alignment after total knee arthroplasty using patient-specific instruments. International orthopaedics 2013;37-1:45-50.

8. Nam D, Maher PA, Rebolledo BJ, Nawabi DH, McLawhorn AS, Pearle AD. Patient specific cutting guides versus an imageless, computer-assisted surgery system in total knee arthroplasty. The Knee 2013.



11. Nunley RM, Ellison BS, Zhu J, Ruh EL, Howell SM, Barrack RL. Do patient-specific guides improve coronal alignment in total knee arthroplasty? Clinical orthopaedics and related research 2012;470-3:895-902.

12. Nunley RM, Ellison BS, Ruh EL, Williams BM, Foreman K, Ford AD, Barrack RL. Are patient-specific cutting blocks cost-effective for total knee arthroplasty? Clinical orthopaedics and related research 2012;470-3:889-94.

13. Moreland JR, Bassett LW, Hanker GJ. Radiographic analysis of the axial alignment of the lower extremity. The Journal of bone and joint surgery. American volume 1987;69-5:745-9.

14. Conteduca F, Iorio R, Mazza D, Caperna L, Bolle G, Argento G, Ferretti A. Evaluation of the accuracy of a patient-specific instrumentation by navigation. Knee surgery, sports traumatology, arthroscopy : official journal of the ESSKA 2012.

15. Lustig S, Scholes CJ, Oussedik SI, Kinzel V, Coolican MR, Parker DA. Unsatisfactory accuracy as determined by computer navigation of VISIONAIRE patient-specific instrumentation for total knee arthroplasty. The Journal of arthroplasty 2013;28-3:469-73.

67

Chapter 5

No difference in clinical outcome between patient-matched positioning guides and

conventional instrumented TKA at 2 years follow-up:

a multi-centre, double-blind, randomized controlled trial

Boonen B., Schotanus M.G.M., Kerens B., van der Weegen W., Hoekstra H., Kort N.P. Bone Joint J. 2016 Jul;98-B(7):939-44.

Chapter 5

68

ABSTRACT

Aims We wished to compare the clinical outcome, as assessed by questionnaires and the rate of complications, in total knee arthroplasty (TKA) undertaken with patient-matched positioning guides (PMPGs) or conventional instruments. Patients and Methods A total of 180 patients (74 men, 106 women; mean age 67 years) were included in a multicentre, adequately powered, double-blind, randomized controlled trial. The mean follow-up was 44 months (24 to 57). Results There were no significant or clinically relevant differences between the two groups for all outcome measures (Knee society score, p = 0.807; Oxford Knee Score, p = 0.304; Western Ontario and McMaster osteoarthritis index, p = 0.753; visual analogue scale for pain, p = 0.227; EuroQol-5D-3L index score, p = 0.610; EuroQol-5D-3L VAS health, p = 0.968.) There was no difference in the rate of complications (p = 0.291). Conclusion PMPGs are already in relatively common use and their short-term clinical results are equal to conventional instrumented TKA.


69

INTRODUCTION

Patient-matched positioning guides (PMPG) have been available for use in total knee arthroplasty (TKA) for several years. The technique uses MRI or CT-scan to calculate ideal implant position using predetermined reference axes and planes. The purpose of PMPG is, based on these scans, to create jigs that have only one appropriate position on the anatomy of the individual patient. These jigs can also be used to guide the place-ment of pins onto which standard resection guides can be placed. Research has focused primarily on alignment outcomes with this relatively new technique. Most papers have shown comparable alignment results obtained with PMPG and conventional instru-ments1-6.

It is generally accepted that pain relief and improved function are the principal aims of arthroplasty. However, despite its current relatively widespread use, few authors have described the clinical outcomes and possible adverse events (AE’s) related to PMPG’s. Studies that have addressed this issue have reported short term results7,8 or suffered from a lack of adequate randomisation9,10. To our knowledge, there are no appropriately powered randomized controlled trials reporting the clinical outcome.

This double-blind, multicentre, randomized controlled trial, comparing PMPG with conventional instrumentation, was designed to address the following research ques-tions. First, is there a difference in clinical outcome between conventional instruments and PMPG as assessed with both examiner based outcome measures and patient re-ported outcome measures (PROMS).

Secondly, is there a difference in the rate of AE’s and complications between the two techniques.

It was hypothesised that there would be no difference in clinical outcome and rate of adverse events and complications between conventional TKA and PMPG TKA.

PATIENTS AND METHODS

The Consolidated Standards of Reporting Trials (CONSORT) statement was strictly fol-lowed in this trial11. Between September 2010 and March 2013, 180 patients consented to take part in this prospective, randomized, controlled trial. Two hospitals participated and enrolled 90 patients each. Patients with disabling osteoarthritis of the knee who were candidates for primary unilateral TKA, after extensive conservative treatment, were eligible for inclusion. Exclusion criteria were the presence of metal near the knee, ankle or hip joint, patients with contra-indications to MRI scans, those who had previ-ously undergone knee surgery (except for arthroscopic meniscectomy), those with indi-cations for TKA other than osteoarthritis, those with active infection near the knee or with active systemic infection and those who were unable to comply with the post-operative rehabilitation protocol, or were unwilling to participate (fig. 1).

Chapter 5

70

Randomisation involved an online random number generator (www.randomizer.org) and was stratified per hospital. The 90 patients from each hospital were divided into two groups of 45 patients – a group in which TKA was undertaken using PMPG and a control group in which conventional instruments were used. The surgeon enrolling patients in the trial was unaware of the type of treatment that patients would receive. The patients were also blinded to the method of alignment of the components that would be used. The characteristics of the patients at the time of enrolment are shown in Table I. Despite randomisation, the patients in the conventional group were signifi-cantly younger than those in the PMPG group. Table I: Baseline characteristics

PMPG conventional P-value

Men/Women 34/56 40/50 0.557†

Mean age (SD) 69 (8.0) 65 (8.8) 0.001*

BMI (range) 30.3 (22.9 – 40.7) 29.5 (21.3 to 42.7) 0.513†

* Student’s t-test † Fisher’s exact test PMPG, patient-matched positioning guides

Fig. 1: Diagram of the number of patients enrolled in the study and analysed at 2 years of follow up as rec-ommended by CONSORT.

Enrollment Assessed for eligibility (n=522)

Randomized (n=180)

Allocated to PMPG (n=90)Received allocated intervention (n=90)

Excluded (n=342)Declined to participate (n=219)Other reason (n=123)

Allocated to Conventional (n=90)Received allocated intervention (n=90)

Analysis

Lost to follow up (n=8)Withdrew (n=6)Deceased <2 year (n=2)

Lost to follow up (n=9)Withdrew (n=8)Deceased <2 year (n=1)

Follow up

Analysed (n=82) Analysed (n=81)


71

Surgical procedure and postoperative protocol

In PMPG group, MRI scans of the hip, knee and ankle were performed six weeks prior to surgery according to the standard Signature scanning protocol. Software (Materialise NV, Leuven, Belgium) was used to create virtual three-dimensional models of the femur and tibia. The program was used to determine the appropriate size and optimal posi-tioning of the components (Vanguard Complete Knee System, Biomet, Inc., Warsaw, Indiana) for each individual patient. The position was calculated to obtain a neutral mechanical (hip-knee-ankle angle of 180°) axis and a neutral position of the femoral and tibial components to the mechanical axis in the frontal plane. In the sagittal plane, the posterior slope of the tibial component and flexion of the femoral component were set at 3°. A digital, virtual plan of the proposed positioning was sent to the surgeon, who could adjust the digital plan if desired. After approving the plan, guides (Signature, Bi-omet, Inc.,) for per-operative use were manufactured using a rapid prototype engineer-ing technique. In the control group, standard intramedullary instrumentation was used. These patients also underwent MRI scanning according to the same protocol as those in the PMPG group to maintain blinding of the patients to the type of alignment that was to be used. Again, the goal was to obtain a neutral mechanical axis in all patients. Flex-ion of the femoral component and the posterior slope of the tibia were set at 0°.

A standard medial parapatellar approach was used in both groups. All procedures were performed by one of three surgeons (one at Zuyderland Medical Centre (NK) and two at St Anna Hospital (RD and HH). All had extensive experience with conventional TKA and had undertaken at least 100 TKAs using the PMPG technology. In the PMPG group, the guides were placed on the articular surface after removal of any soft tissues at this site. Pins were placed using these guides that determined the position of the standard cutting blocks. The femur was prepared first. A cemented Vanguard TKA (Van-guard Complete Knee System) was used. Patellar resurfacing was performed where necessary (i.e., in patients with rheumatoid arthritis). Soft-tissue balancing was also undertaken where necessary. In the control group, the same procedure was undertak-en, except that standard intramedullary instruments were used to guide the position of the cutting blocks.

Postoperative protocol was described in detail in a previous publication12. Arixtra 5mg/mL, 0.5 mL (GlaxoSmithKline) (Zuyderland hospital) or Xarelto 10 mg (Bayer) (St. Anna hospital) were used as thrombo-embolic prophylaxis for five weeks post-operatively. All patients participated in an enhanced recovery program (Joint Care, Biomet NL, Dordrecht, The Netherlands).

The blinding of the patients and those who assessed the outcome was maintained throughout the period of the study. They were unblinded at the completion of the study, two years post-operatively.

Chapter 5

72


Preoperatively, all patients completed the following questionnaires: Knee Society Score (KSS)13, the Dutch translated and validated version of the Oxford Knee Score (OKS)14, Western Ontario and McMaster osteoarthritis index (WOMAC)15, visual analogue scale (VAS) for pain and EuroQol (EQ-5D, 3L version)16. During the follow-up visits at 3 months, 1 year and 2 years, this same set of questionnaires was completed by the study patients (OKS, WOMAC, EQ-5D, VAS) and by an independent physician (MS at the Zuy-derland MC and WW at the St. Anna hospital), who were blinded to the alignment method used.

The KSS and WOMAC were scored from 0 to 100, 0 being the worst outcome and 100 being the best possible outcome13,17. The OKS was scored from 12 to 60, with 12 being the best possible outcome and 60 being the worst14. VAS pain was scored from 0 to 100, a score of 0 representing no pain at all, a score of 100 representing worst possi-ble pain. For the EQ-5D, a single summary index is calculated, using the value set for the Netherlands16,18,19. An index of 1 represents perfect health and no disabilities.

Scores on the questionnaires were compared between both groups at the different follow-up visits.

All complications were registered during the minimal 2 year follow-up. This study was approved by the institutional review boards of both hospitals (File nr.

10-T-21).

Statistical analysis

This study was powered on the difference in KSS at two years post-operatively. (Proba-bility of Type I error: 0.05, Probability of Type II error: 0.10, maximal clinically not rele-vant difference in KSS: 10, estimate of standard deviation of KSS at two years post-operatively: 19.657.) Based on this calculation, 158 patients would have to be included. In order to account for possible loss to follow-up, 180 patients were included. A gener-alized linear mixed model (GLMM) approach was used to take into account the repeat-ed-measures design of the study in order to cope with any missing data collected before and at three months and one and two years follow-up and to cope with the wide range of variation in relation to the time frame the data was collected.23 Student’s t-tests were performed on significant interactions. Fisher’s exact test was used to test differ-ences of proportions.

A p-value of < 0.05 was considered significant. SPSS-software was used for all statis-tical analyses (SPSS 14 Inc., Chicago, Illinois).


73

RESULTS

The mean follow-up was 44 months (24 to 57). A total of 17 patients were lost to follow-up. Two in the PMPG group and one in the conventional group died of unrelated causes. Six in the PMPG group and eight in the conventional group withdrew two years post-operatively. Thus, at this time, a total of 163 patients were analysed. All patients received the treatment that they were allocated to. No PMPG procedures had to be converted to a traditional instrumented TKA either pre- or per-operatively. The mean outcome scores at two years improved significantly within each group compared with the pre-operative values and there were no statistically significant differences in the outcome in the two groups (Table II). GLMM was adjusted for age and baseline of each outcome.

There was no statistically significant difference in the total number of complications (p = 0.291, Fisher’s exact test).A total of 14 patients in the PMPG group had a complica-tion. Two had a haematoma and eight had a poor range of movement (< 90° of knee flexion); four underwent a manipulation under anaesthesia (MUA). Two had persistent pain and one had a persistent effusion, for which a cause could not be found. One pa-tient had a pulmonary embolism.

A total of 11 patients in the conventional group had a complication. Eight had a poor range of movement; two underwent MUA. One patient had an early deep infection (within six weeks of TKA implantation) requiring debridement with retention of the components and antibiotics. Two patients had further surgery, a patellar button was introduced in one and a malaligned tibial component was revised in another.

Fig. 1: Diagram of the number of patients enrolled in the study and analysed at 2 years of follow up as rec-ommended by CONSORT.

PMPG Mean SD (95% CI)

Conventional Mean SD (95% CI)

p-value GLMM

KSS Pre 50.7 13.3 (47.4-53.9) 52.0 14.4 (48.6-55.5) 0.807

3-MTS 75.7 18.5 (71.2-80.2) 76.5 14.5 (73.0-80.0)

1-year 81.6 15.5 (77.8-85.4) 82.2 14.2 (78.8-85.7)

2-year 81.9 15.7 (78.1-85.8) 82.2 14.8 (78.6-85.8)

OKS Pre 30.0 9.3 (27.7-32.3) 29.3 7.5 (27.5-31.1) 0.304

3-MTS 18.5 9.6 (16.2-20.9) 16.4 9.8 (14.0-18.7)

1-year 16.0 8.2 (14.0-18.0) 15.0 8.2 (13.2-16.9)

2-year 15.2 8.7 (13.1-17.2) 15.1 8.5 (13.1-17.1)

WOMAC Pre 51.5 17.0 (47.9-55.1) 55.8 16.2 (52.4-59.2) 0.753

3-MTS 80.9 16.4 (77.5-84.4) 82.2 14.7 (79.1-85.3)

1-year 80.6 18.4 (76.7-84.6) 86.0 13.8 (83.1-89.0)

2-year 80.7 20.2 (76.3-85.0) 86.6 14.8 (83.4-89.8)

Chapter 5

74

PMPG Mean SD (95% CI)

Conventional Mean SD (95% CI)

p-value GLMM

VAS Pain Score

Pre 58.3 22.3 (52.9-63.8) 61.4 16.9 (57.3-65.5) 0.227

3-MTS 20.0 21.4 (14.8-25.3) 20.6 23.5 (14.9-26.2)

1-year 21.7 22.3 (16.3-27.2) 19.1 23.6 (13.5-24.8)

2-year 20.4 24.8 (14.4-26.5) 17.4 21.8 (12.2-22.6)

EQ-5D Pre 0.75 0.08 (0.73-0.77) 0.78 0.07 (0.76-0.79) 0.610

3-MTS 0.88 0.10 (0.86-0.91) 0.89 0.10 (0.86-0.91)

1-year 0.88 0.10 (0.86-0.91) 0.89 0.11 (0.87-0.92)

2-year 0.89 0.11 (0.86-0.91) 0.90 0.12 (0.87-0.93)

EQ-5D VAS

Pre 64.9 18.1 (60.4-69.3) 73.6 12.6 (70.6-76.7) 0.968

3-MTS 74.6 14.4 (71.1-78.1) 77.8 11.2 (75.1-80.5)

1-year 74.5 13.0 (71.3-77.7) 78.0 14.8 (74.5-81.6)

2-year 72.5 17.9 (68.2-76.7) 76.2 17.9 (71.9-80.5)

p-values are presented for the Knee society score (KSS), Oxford Knee score (OKS), Western Ontario McMaster Universities Osteoarthritis index (WOMAC), Visual analogue scale (VAS), EuroQol-5D (EQ 5D) using a general-ised linear mixed model (GLMM) PMPG, patient-matched positioning guides.

DISCUSSION

The most important finding of the present study was that there is no difference in clini-cal outcome between conventional and PMPG instrumentation 2 years postoperatively, as was hypothesised. The second important finding was that there was no difference in the rate of adverse effects and complications, as was also hypothesised.

Our observations concerning the clinical outcome are in line with other authors. Yan et al.8 and Abane et al.7 found no difference in clinical outcome on short-term follow-up. Anderl et al.9 and Chen et al.10 also found no difference in clinical outcome at 2 years postoperatively, however these studies were not randomized. Expectations vary greatly between patients and the mismatch of experience versus expectation after TKA is a potent cause of patient dissatisfaction20. PROMS are thought to represent the best objective measurement of clinical outcome21. Currently however, there is no single best outcome measure for TKA, and we therefore chose to use a number of scoring systems to obtain insight in clinical function of the patients participating in this RCT.

Although we used various PROMS, there are some drawbacks associated with the use of these questionnaires. PROMS are subjective measures and suffer from a ceiling effect and their pain dominance masks the functional changes22. As the proportion of younger, more active and more demanding patients undergoing TKA is rising, function after TKA is becoming more important23. It has been suggested by Bolink et al.24 that in order to characterize the changes in physical function after TKA, PROMS could be sup-


75

plemented by performance-based measures, assessing function during different activi-ties and allowing kinematic characterization with an ambulant inertial measurement unit. Concerning PROMS, The High Activity Arthroplasty Score (HAAS) could be of inter-est when assessing subtle variations in function after arthroplasty when assessing func-tion in more demanding patients. The score allows for greater differentiation of level of function between patients in assessing performance after TKA or total hip arthroplasty (THA)25. We did not we found no difference between PMPG and conventional TKA.

Given that most studies show comparable results for PMPG and conventional in-struments for restoring a neutral mechanical axis and for percentages of outliers and given our observation that there is no difference in clinical outcome or complications associated with PMPG, it seems reasonable to conclude that PMPG’s are as for aligning TKA as conventional instruments; however, no clear advantage seems to exists over conventional instruments. The cost-effectiveness of PMPG must be considered. Poten-tial cost-savings include a shortened operating time12,26, reduction in the number of instrument sets (and additional sterilisation costs in most cases)26, reduced tray pro-cessing time27, and, purely theoretically, reduction in hospital shelf stock. Additional costs include the cost of an MRI-scan or CT-scan (hospital specific), costs of the PMPG (manufacturer and hospital specific), and time needed for logistical tasks (depending on available personnel). These requirements include the scanning process, transfer of the images to the manufacturer of the PMPG, monitoring the on-time delivery of PMPG to the hospital and approval of the digital plan by the surgeon prior to fabrication of the PMPG. This last item seems to be an essential step when using PMPG in order to avoid time-consuming intra-operative changes in the proposed implant size and resection levels28,29. It has been stated that PMPG would not be cost-effective when looking only at reduction in operating time and the alignment obtained with this technique30; there-fore, the hypothesised financial savings might only be realised when the preoperative digital plan accurately predicts the size of the components.

The strengths of this study are the design and the large number of patients, based on a power calculation. Furthermore, a mixed model approach was used to analyse the data. This is considered to be more appropriate for assessing repeated measurements in clinical trials31.

A limitation of the study was the relatively short follow-up. As a result, no reliable data could be provided on the survival of the TKA in both groups. Furthermore, the target flexion and posterior slope were not identical in both groups, which might have influenced clinical outcomes. It remains unclear from the presented data to what extent this variable has influenced outcomes.

Chapter 5

76

CONCLUSION

There was no significant difference between conventional TKA and PMPG TKA regarding clinical outcome and TKA related complication at a minimal 2 years follow-up. Future research should primarily focus on the cost-effectiveness of PMPG and on the long-term clinical results.

ACKNOWLEDGMENTS

The authors would like to thank R. van Drumpt, MD, for operating on part of the pa-tients enrolled in this trial at the St. Anna Hospital.


77

REFERENCES

1. Barrett W, Hoeffel D, Dalury D, Mason JB, Murphy J, Himden S. In-vivo alignment comparing patient specific instrumentation with both conventional and computer assisted surgery (CAS) instrumentation in total knee arthroplasty. The Journal of arthroplasty 2014;29-2:343-7.

2. Boonen B, Schotanus MG, Kerens B, van der Weegen W, van Drumpt RA, Kort NP. Intra-operative results and radiological outcome of conventional and patient-specific surgery in total knee arthroplasty: a multicentre, randomized controlled trial. Knee surgery, sports traumatology, arthroscopy : official journal of the ESSKA 2013;21-10:2206-12.


4. 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 surgery, sports traumatology, arthroscopy : official journal of the ESSKA 2013;21-10:2213-9.

5. Roh YW, Kim TW, Lee S, Seong SC, Lee MC. Is TKA using patient-specific instruments comparable to conventional TKA? A randomized controlled study of one system. Clinical orthopaedics and related research 2013;471-12:3988-95.

6. Victor J, Dujardin J, Vandenneucker H, Arnout N, Bellemans J. Patient-specific guides do not improve accuracy in total knee arthroplasty: a prospective randomized controlled trial. Clinical orthopaedics and related research 2014;472-1:263-71.

7. Abane L, Anract P, Boisgard S, Descamps S, Courpied JP, Hamadouche M. A comparison of patient-specific and conventional instrumentation for total knee arthroplasty: a multicentre randomized controlled trial. The bone & joint journal 2015;97-B-1:56-63.

8. Yan CH, Chiu KY, Ng FY, Chan PK, Fang CX. Comparison between patient-specific instruments and conventional instruments and computer navigation in total knee arthroplasty: a randomized controlled trial. Knee surgery, sports traumatology, arthroscopy : official journal of the ESSKA 2014.

9. Anderl W, Pauzenberger L, Kolblinger R, Kiesselbach G, Brandl G, Laky B, Kriegleder B, Heuberer P, Schwameis E. Patient-specific instrumentation improved mechanical alignment, while early clinical outcome was comparable to conventional instrumentation in TKA. Knee surgery, sports traumatology, arthroscopy : official journal of the ESSKA 2014.

10. Chen JY, Chin PL, Tay DK, Chia SL, Lo NN, Yeo SJ. Functional Outcome and Quality of Life after Patient-Specific Instrumentation in Total Knee Arthroplasty. The Journal of arthroplasty 2015.

11. Begg C, Cho M, Eastwood S, Horton R, Moher D, Olkin I, Pitkin R, Rennie D, Schulz KF, Simel D, Stroup DF. Improving the quality of reporting of randomized controlled trials. The CONSORT statement. JAMA 1996;276-8:637-9.

12. Boonen B, Schotanus MG, Kerens B, van der Weegen W, van Drumpt RA, Kort NP. Intra-operative results and radiological outcome of conventional and patient-specific surgery in total knee arthroplasty: a multicentre, randomized controlled trial. Knee surgery, sports traumatology, arthroscopy : official journal of the ESSKA 2013.

13. Insall JN, Dorr LD, Scott RD, Scott WN. Rationale of the Knee Society clinical rating system. Clinical orthopaedics and related research 1989-248:13-4.

14. Haverkamp D, Breugem SJ, Sierevelt IN, Blankevoort L, van Dijk CN. Translation and validation of the Dutch version of the Oxford 12-item knee questionnaire for knee arthroplasty. Acta orthopaedica 2005;76-3:347-52.

15. Bellamy N, Buchanan WW, Goldsmith CH, Campbell J, Stitt LW. Validation study of WOMAC: a health status instrument for measuring clinically important patient relevant outcomes to antirheumatic drug therapy in patients with osteoarthritis of the hip or knee. The Journal of rheumatology 1988;15-12:1833-40.

Chapter 5

78

16. EuroQol--a new facility for the measurement of health-related quality of life. Health policy 1990;16-3:199-208.

17. Roorda LD, Jones CA, Waltz M, Lankhorst GJ, Bouter LM, van der Eijken JW, Willems WJ, Heyligers IC, Voaklander DC, Kelly KD, Suarez-Almazor ME. Satisfactory cross cultural equivalence of the Dutch WOMAC in patients with hip osteoarthritis waiting for arthroplasty. Annals of the rheumatic diseases 2004;63-1:36-42.

18. Brooks R. EuroQol: the current state of play. Health policy 1996;37-1:53-72. 19. Lamers LM, McDonnell J, Stalmeier PF, Krabbe PF, Busschbach JJ. The Dutch tariff: results and arguments

for an effective design for national EQ-5D valuation studies. Health economics 2006;15-10:1121-32. 20. Noble PC, Conditt MA, Cook KF, Mathis KB. The John Insall Award: Patient expectations affect satisfaction

with total knee arthroplasty. Clinical orthopaedics and related research 2006;452:35-43. 21. Rolfson O, Malchau H. The use of patient-reported outcomes after routine arthroplasty: beyond the whys

and ifs. The bone & joint journal 2015;97-B-5:578-81. 22. Terwee CB, van der Slikke RM, van Lummel RC, Benink RJ, Meijers WG, de Vet HC. Self-reported physical

functioning was more influenced by pain than performance-based physical functioning in knee-osteoarthritis patients. Journal of clinical epidemiology 2006;59-7:724-31.

23. Nilsdotter AK, Toksvig-Larsen S, Roos EM. Knee arthroplasty: are patients' expectations fulfilled? A prospective study of pain and function in 102 patients with 5-year follow-up. Acta orthopaedica 2009;80-1:55-61.

24. Bolink SA, Grimm B, Heyligers IC. Patient-reported outcome measures versus inertial performance-based outcome measures: A prospective study in patients undergoing primary total knee arthroplasty. The Knee 2015.

25. Talbot S, Hooper G, Stokes A, Zordan R. Use of a new high-activity arthroplasty score to assess function of young patients with total hip or knee arthroplasty. The Journal of arthroplasty 2010;25-2:268-73.

26. Renson L, Poilvache P, Van den Wyngaert H. Improved alignment and operating room efficiency with patient-specific instrumentation for TKA. The Knee 2014;21-6:1216-20.

27. Lachiewicz PF, Henderson RA. Patient-specific Instruments for Total Knee Arthroplasty. The Journal of the American Academy of Orthopaedic Surgeons 2013;21-9:513-8.

28. Boonen B, Schotanus MG, Kerens B, Hulsmans FJ, Tuinebreijer WE, Kort NP. Patient-specific positioning guides for total knee arthroplasty: no significant difference between final component alignment and pre-operative digital plan except for tibial rotation. Knee surgery, sports traumatology, arthroscopy : official journal of the ESSKA 2015.

29. Pietsch M, Djahani O, Hochegger M, Plattner F, Hofmann S. Patient-specific total knee arthroplasty: the importance of planning by the surgeon. Knee surgery, sports traumatology, arthroscopy : official journal of the ESSKA 2013.

30. Nunley RM, Ellison BS, Ruh EL, Williams BM, Foreman K, Ford AD, Barrack RL. Are patient-specific cutting blocks cost-effective for total knee arthroplasty? Clinical orthopaedics and related research 2012;470-3:889-94.

31. DeSouza CM, Legedza AT, Sankoh AJ. An overview of practical approaches for handling missing data in clinical trials. Journal of biopharmaceutical statistics 2009;19-6:1055-73.

79

Chapter 6

Patient-specific guides in total knee arthroplasty;

2 years follow up of the first 200 consecutive cases performed by a single surgeon

Boonen B., Schrander D.E., Schotanus M.G.M., Hulsmans F-J., Kort N.P. Published online. JCRMM. 2015. http://content.yudu.com/Library/A3yjzj/JOURNALOFCLINICALRHE/resources/index.htm? referrerUrl=http%3A%2F%2Ffree.yudu.com%2Fitem%2Fdetails%2F3682413%2FJOURNAL-OF-CLINICAL-RHEUMATOLOGY-AND-MUSCULOSKELETAL-MEDICINE

Chapter 6

80

ABSTRACT

Results of patient-matched positioning guides (PMPG) in terms of functional outcome in larger series have not yet been reported and only a few reports have presented data on the size changes of the components that were made to the preoperative plan during surgery. We present the results of our initial series of 200 consecutive cases operated on by a single surgeon and with a minimum follow-up of 2 years.

No cases in this series needed to be converted to traditional instrumentation. Aver-age operation time was 52 minutes. No learning curve on surgical time was seen. Im-planted component size prediction with an error of one size, was correct in 98.5% of femoral components and 97.0% of tibial components. Reviewing and approving the preoperative plan, reduced the number of changes that were necessary during surgery from 21.5% to 14% for the femoral component and from 41% to 29.5% for the tibial component. Average mechanical axis was 179.5° with a percentage of outliers > 3° from digital planning of 20.7%. No adverse events occurred specifically related to the use of PMPG.

This series of 200 consecutive TKA’s placed with PMPG indicates that PMPG is a safe technique. Our results stress the importance of a critical review of the digital preopera-tive plan in order to be able to rely on the component sizing capabilities of PMPG. Fu-ture research should focus on outcome of PMPG in lower volume surgeons and on the role PMPG can play in reducing hospital stocking and instrument costs.

Patient-matched Positioning Guides: Can We Rely On Their Sizing Capabilities?

81

INTRODUCTION

Total knee arthroplasty (TKA) is one of the most successful types of surgery with the majority of patients gaining rapid improvement in pain, function, and quality of life1.

Patient matched positioning guides (PMPG) is a relative novel technology in TKA with the potential for decreased operative time, less invasiveness, decreased blood loss, increased accuracy and less intraoperative surgeon decision-making compared with traditional arthroplasty2-4. Planning of alignment of the prosthesis has shifted to the preoperative timeframe and is based on imaging techniques that provide insight in the individual patient’s anatomy. Currently, most orthopaedic companies offer such a sys-tem, either CT or MRI based, and use their own algorithms. Using these imaging tech-niques PMPG can be created that dictate positioning of initial pin placement needed for traditional cutting blocks peroperatively or also have saw slots incorporated in their design to do the actual cutting. Most systems aim at restoring a neutral mechanical axis and the early research has focused primarily on comparing alignment results of this relatively new technique to alignment results of conventional TKA. Results concerning alignment differ greatly with some literature reporting superior results to conventional instrumentation2,5-7 while other literature reporting comparable8-11 or even inferior results12.

The growing number of reports and sales figures on this PMPG technique suggests that the technique is already widely used in clinical practise13. However, results in terms of functional outcome in larger series have not yet been reported and only a few re-ports have presented data on the changes that were made to the preoperative plan during surgery14,15. Further a large number of series represent the initial cases of the surgeon.

We present the results of our initial series of 200 consecutive cases operated on by a single surgeon with this new technique and with a minimum follow-up of 2 years. We focussed on alignment in the frontal plane and on the number and on the type of ad-justments that were made to the preoperative digital plan during surgery as well as on clinical outcome.


Data of the first 184 patients in whom 200 TKA’s, using PMPG, were prospectively col-lected. The patients were operated on between July 2009 and March 2011. Institutional review board (METC Atrium Orbis Zuyd, trial Nr. 12-N-139) approval was obtained for the study. Patients with disabling osteoarthritis of the knee that were candidates for primary unilateral TKA were eligible for inclusion. Patients with metal near the knee, ankle or hip joint, patients with contra-indications to MRI-scan, patients that had previ-

Chapter 6

82

ously undergone knee surgery (except for arthroscopic meniscectomy) and patients that refused to consent were excluded.

Surgical technique

The SignatureTM system (Biomet Inc., Warsaw, IN) was used in this cohort. Pre-operative preparation, operative procedure and postoperative management was done as described in a previous report4.


Operative record was completed, containing information on operation time (minutes from incision until the bandage was placed), blood loss (mL of blood in the suction de-vice, prior to application of a tourniquet and prior to rinsing the knee), size of the com-ponents and polyethylene insert.

Number of adjustments to the preoperative plan before approval, number of ad-justments to the approved plan during surgery, number of adjustments to the default plan during surgery and number of sizes predicted correctly by the software were calcu-lated.

The level of activity and patient-reported functional outcome and pain scores were evaluated preoperatively and postoperatively (1 year and 2 years) with use of the EQ5D score (Euroqol group), Oxford knee score (OKS) and the Pain Visual Analogue Score (VAS). EQ-5D scores for all 5 domains presented separately and as percentage of pa-tients per level (level 1: no problems, level 2: some problems, level 3: extreme prob-lems). Higher OKS indicate better functional outcome with a score of 48 being the best outcome. Lower scores on the VAS indicate less pain.

Standing, weight-bearing, anterior posterior long-leg digital radiographs were taken preoperatively and 6 weeks, 1 year and 2 years postoperatively. Mechanical axis was determined according to Tigani et al16 and was measured using calibrated software. Two (BB and DS) independent reviewers performed the measurements and the inter-class correlation coefficient was determined. Deviations of more than 3° from the planned mechanical axis and planned alignment of individual components were regard-ed as outliers, and percentages were calculated. Values lower than 180° indicated varus mechanical axis alignment. For calculation of the outliers, the mean outcome of both measurements (from the 2 independent assessors) was used.

All complications and adverse events are documented in the patients file. These events are divided as patient (infection, thromboembolic, wound problems) and im-plant related (loosening, revision, etc.)


83

RESULTS

All patients had a minimal follow-up of 2 years ranging from 2.0 to 4.2 years. Patient characteristics are listed in table 1.

Table 1: patient characteristics

Women (%) 117/184 (63.6%)

Mean BMI 29

Mean age (range) 68 years (48-86 years)

Bilateral TKA cases (%) 16/184 (8.7%)

Left/right 84/116

Peroperative data

The individual guides fitted well on the native bone and cartilage. No cases needed conversion to intramedullary instrumentation. As routine, the first five cases were checked with intramedullary guiding. In 159 patients a cruciate retaining TKA was placed. In six patients a posterior stabilized TKA was used due to posterior cruciate ligament insufficiency. Average blood loss was 250 ml (range 50-1000ml, SD 142.1). Average operation time was 52 minutes (range 34-80 minutes; SD 10.2). The course of OR time in the cohort is presented in Figure 1. The patella was resurfaced in 35 cases.

Figure 1: operation time in minutes of individual cases

0

10

20

30

40

50

60

70

80

90

0 50 100 150 200

OR

ti

me

Cases

OR time

OR time

Trentline OR-time

Chapter 6

84

Implant size

In Table 2 results on preoperative adjustments to the digital plan, peroperative adjust-ments to the approved plan, peroperative adjustments to the default plan and accuracy of size prediction are presented. Table 2: Number of adjustments to the preoperative plan before approval (upper 1/4th of table), number of adjustments to the approved plan during surgery (second 1/4th of table), number of adjustments to the de-fault plan during surgery (third 1/4th of table) and number of sizes predicted correctly by the software (lower 1/4th of table).

Femur Tibia

Preoperative adjustments to sizes in the default plan

One or more sizes smaller (%) 24/200 (12.0 %) 1/200 (0.5%)

One or more sizes larger (%) 4/200 (2.0%) 45/200 (22.5%)

Total adjustments 28/200 (14.0%) 46/200 (23.0%)

Peroperative adjustments to the approved plan

One or more sizes smaller (%) 20/200 (10.0%) 24/200 (12.0%)

One or more sizes larger (%) 4/200 (2.0%) 35/200 (17.5%)


Peroperative adjustments to the default plan

One or more sizes smaller (%) 40/200 11/200

One or more sizes larger (%) 3/200 71/200


Sizes predicted correctly with an error of 1 size

By the default plan (%) 195/200 (97.5%) 193/200 (96.5%)

By the approved plan (%) 197/200 (98.5%) 194/200 (97.0%)

There were 3 femoral components peroperatively changed to two sizes different as approved (2 times +2 and 1 time -2). There were 6 tibial components peroperatively changed to two sizes different as approved (4 times +2 and 2 times -2).

The cases in which adjustments were made to the approved plan during surgery were evenly distributed over time. In 6 patients both femoral and tibial components sizes were adjusted, in all other cases only one component size was adjusted during surgery.

In 4 femoral components, the size was adjusted preoperatively while the size was changed back to the original (by the software) proposed size during surgery. The same was true for 9 tibial components.

In 109 cases a 10mm insert was placed, in 78 cases a 12mm, in 12 cases a 14mm and in 1 case a 16mm. A 10mm insert is default planning.

Radiographic evaluation

Average mechanical axis was 179.5° with a percentage of outliers > 3° from digital plan-ning of 20.7% (range 167.7° – 187.1°; SD 2.7). Average femoral component alignment was 1.7 degrees of varus, relative to the mechanical axis of the femur and average tibial


85

component alignment was 1.4 degrees of valgus, relative to the mechanical axis of the tibia. Outliers > 3° from digital planning were 8.3% and 11.6% for the femoral and tibial components, respectively. Outliers were evenly distributed among the total group. Intraclass correlation coefficient for determining of the mechanical axis alignment measures was good (0.88).

There was no periprosthetic osteolysis or evidence of implant loosening after two years of follow-up visual at radiographs.

Outcome scores

The reported quality of life, functional and pain outcome scores substantially improved at the 2-year follow-up assessment compared to the preoperative assessment (Table 3). Table 3. Reported quality of life, functional and pain outcome scores.

LEVEL PRE-OP 1 YEAR 2 YEARS

EQ - 5D MOBILITY Level 1 3.7% 55.0% 73.0%

Level 2 96.3% 45.0% 26.9%

Level 3 0.0% 0.0% 0.0%

EQ - 5D SELFCARE Level 1 72.9% 89.0% 90.4%

Level 2 27.1% 11.0% 9.6%

Level 3 0.0% 0.0 % 0.0%

EQ - 5D ACTIVITY Level 1 25.2% 58.0% 69.2%

Level 2 67.2% 41.0% 30.8%

Level 3 7.5% 1.0% 0.0%

EQ - 5D PAIN

Level 1 3.7% 43.0% 63.5%

Level 2 57.0% 51.0% 34.6%

Level 3 39.3% 6.0% 1.9%

EQ - 5D ANXIETY Level 1 77.6% 97.0% 82.7%

Level 2 20.7% 3.0% 17.3%

Level 3 1.9% 0.0% 0.0%

EQ - 5D EQ VAS 63/100 78/100 76/100

OKS 9.0 25.9 27.3

VAS PAIN 6.7 2.7 1.9

General complications occurred in this cohort. One patient required fasciotomy due to compartment syndrome emerging one week after TKA placement. One knee was re-vised two years after TKA placement, due to secondary hematogenic infection after invasive treatment for colon cancer. One patient required a knee manipulation for stiff-ness. In one knee, a patella button was placed for persistent anterior knee pain after TKA. There were no deep wound infections, symptomatic deep vein thrombosis, or

Chapter 6

86

pulmonary embolisms. Two patients deceased during follow up of causes non-related to knee arthroplasty. No adverse events occurred specifically related to the use of PMPG.

DISCUSSION

We found that implanted component size prediction with an error of one size, was correct in 98.5% of femoral components and 97.0% of tibial components and that PMPG predicted the implanted component size exact in 88.0% of femurs and 70.5% of tibias. Furthermore, all guides fitted well on the native anatomy of the patient’s knee and no procedures had to be converted to traditional instrumentation during surgery. The average surgical time was low and few outliers in surgical time were seen.

Results in literature on the topic of size prediction vary with authors reporting good accuracy of the PMPG in predicting component size8 and others reporting frequent intraoperative directed changes15. Given the figures in this study that only 1.5% of fem-oral components and 3.0 % of tibial components deviated 2 or more sizes from the planned size, we believe it is fair to state that the technique is accurate in predicting component sizes. However, we observed that reviewing and approving of the preopera-tive digital plan was important in obtaining these high accuracies and this importance was previously highlighted by others as well14. Our conclusion was based on the fact that far more adjustments would have been necessary when the surgeon would rely on the default settings of the digital plan. We believe therefore, that it is essential for sur-geons using PMPG to preoperatively check and edit the default virtual planning in order to achieve best results but also to avoid medico-legal problems in case of failure as already warned for by others13. In addition, our results indicate that this PMPG system tends to overestimate femoral component size and underestimate tibial component size. This observation can be helpful for new users in order to make appropriate chang-es to the default plan as they know what to expect.

Our results on restoring a neutral mechanical axis are in line with results described in literature8,11, however some reports present results that vary considerably with these findings. Some authors are achieving fewer outliers6,10 and others are reporting higher percentages of outliers12. Number of outliers in our study (20%) are smaller than the number of outliers reported for conventional TKA (roughly 30%) but higher than the number of outliers reported for computer-assisted TKA (roughly 10%)17,18. The number of reports on alignment in the frontal plane is growing. However, limited data exist on the alignment of the components in the transverse and sagittal plane. Improved femo-ral component rotational alignment5 and tibial component rotational alignment7 using PMPG has been reported in literature. Although this rotational alignment was not the primary focus of this study, we believe that PMPG has potential for optimizing align-ment in these planes, rather than improving alignment in the frontal plane. This focus on the rotational component alignment seems mandatory in future research. In addi-


87

tion, we believe a crucial role exists for PMPG in patients with extra-articular deformi-ties that prevent traditional instrumentation of being used.

EQ-5D scores and VAS-scores improved substantially after surgery. Results contin-ued to improve up to 2 years after surgery. We did not compare these clinical data with data of a cohort of conventional TKA since the power to detect changes in clinical out-come, if at all such a difference would exist, is low, therefore needing large series. Nev-ertheless, larger, randomized controlled trials and meta-analyses are necessary in the future to access this issue according to current standards.

Average surgical time in this cohort was 52 minutes and considering a possible learning curve, no effect on surgical time over this series of the initial 200 cases was seen. Average surgical time in this series was 7 minutes more than average surgical time reported in a previous series of the same surgeon19. This difference can partly be ex-plained by the fact that in this cohort the first cases were also included in which align-ment was checked using conventional instrumentation. Additionally, our hospital is a teaching hospital and some cases were performed by orthopaedic residents under su-pervision of the senior surgeon. The senior author is a high volume surgeon and has extensive experience in TKA replacement surgery. With experienced knee surgeons, we believe PMPG adds little in terms of gain in operation time. We do not know yet if PMPG can aid in substantially reducing surgical time for the lower volume surgeons and it would be of interest to address this issue in future research.

In our experience, time saving is for sure an issue in terms of time necessary for in-stalling trays on sterile fields. With a PMPG procedure 3 trays (of which 2 are sterilised together) are used and with conventional instrumentation 6 trays are needed. This opens an interesting discussion on the cost-effectiveness of PMPG. Reports have been published on this issue20 but due to considerable local variation in both additional costs with PMPG (MRI, guides) as well as cost-savings (OR-time, sterilisation costs, shelf stocks) that come with PMPG, results of this research is not automatically applicable for every random hospital and every random user of PMPG. The efficiency of surgical de-partments is crucial to a hospital’s performance. Both manufacturers and hospitals have the same challenge of reducing costs by improving efficiency. The PMPG might be a crucial step in initiatives for improving efficiency while reducing capital investment for hospitals. This improvement is based on the idea of reducing the number of surgical trays needed during surgery and on the reduction of TKA components stock as both are huge investments for hospitals. PMPG can reduce the total set of instruments from 6 to 3 trays. Given the accuracy of PMPG to predict component size correctly with an error of one size in 98% of cases, the number of prosthesis components on the shelf could be reduced to one backup set of all prosthesis sizes in addition to the delivery of the planned component sizes plus one size larger and one size smaller for every individual case in which PMPG are used.

Although several reports on PMPG are published, the results are still not conclusive. Strengths of this study are that this is one of the first studies to report two year results

Chapter 6

88

in a large cohort, operated on by one single surgeon. In our opinion our results provide a clean insight in the results that can be expected of this PMPG technique in a day-by-day practise for a high volume surgeon.

We are aware that our results were not directly compared to results obtained in a cohort of the same surgeon in which traditional instrumentation was used. Further-more, only one PMPG system was used in this series and results are therefore not au-tomatically applicable to other PMPG systems currently on the market. In addition, this cohort represents results of a high volume TKA surgeon and results may not be applica-ble for low volume TKA surgeons. Finally, alignment was measured on conventional radiographs and in the frontal and sagittal plane only. CT-based measurements would have been desirable to draw conclusions on alignment in the transverse plane but due to radiation exposure and costs, this was not feasible.

CONCLUSION

This series of 200 consecutive TKA’s placed with PMPG indicates that PMPG is a safe technique. Our results stress the importance of a critical review of the digital preopera-tive plan in order to be able to rely on the component sizing capabilities of PMPG. Fu-ture research should focus on outcome of PMPG in lower volume surgeons and on the role PMPG can play in reducing hospital stocking and instrument costs.


89

REFERENCES

1. NIH Consensus Statement on total knee replacement December 8-10, 2003. The Journal of bone and joint surgery. American volume 2004;86-A-6:1328-35.

2. Ng VY, DeClaire JH, Berend KR, Gulick BC, Lombardi AV, Jr. Improved accuracy of alignment with patient-specific positioning guides compared with manual instrumentation in TKA. Clinical orthopaedics and related research 2012;470-1:99-107.

3. 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;31-9:857-63.


5. Heyse TJ, Tibesku CO. Improved femoral component rotation in TKA using patient-specific instrumentation. The Knee 2012.

6. Daniilidis K, Tibesku CO. A comparison of conventional and patient-specific instruments in total knee arthroplasty. International orthopaedics 2013.

7. Silva A, Sampaio R, Pinto E. Patient-specific instrumentation improves tibial component rotation in TKA. Knee surgery, sports traumatology, arthroscopy : official journal of the ESSKA 2013.

8. Barrett W, Hoeffel D, Dalury D, Mason JB, Murphy J, Himden S. In-Vivo Alignment Comparing Patient Specific Instrumentation with both Conventional and Computer Assisted Surgery (CAS) Instrumentation in Total Knee Arthroplasty. The Journal of arthroplasty 2013.

9. Vundelinckx BJ, Bruckers L, De Mulder K, De Schepper J, Van Esbroeck G. Functional and radiographic short-term outcome evaluation of the Visionaire system, a patient-matched instrumentation system for total knee arthroplasty. The Journal of arthroplasty 2013;28-6:964-70.

10. Roh YW, Kim TW, Lee S, Seong SC, Lee MC. Is TKA Using Patient-specific Instruments Comparable to Conventional TKA? A Randomized Controlled Study of One System. Clinical orthopaedics and related research 2013.

11. Victor J, Dujardin J, Vandenneucker H, Arnout N, Bellemans J. Patient-specific Guides Do Not Improve Accuracy in Total Knee Arthroplasty: A Prospective Randomized Controlled Trial. Clinical orthopaedics and related research 2013.

12. Chen JY, Yeo SJ, Yew AK, Tay DK, Chia SL, Lo NN, Chin PL. The radiological outcomes of patient-specific instrumentation versus conventional total knee arthroplasty. Knee surgery, sports traumatology, arthroscopy : official journal of the ESSKA 2013.

13. Thienpont E, Bellemans J, Delport H, Van Overschelde P, Stuyts B, Brabants K, Victor J. Patient-specific instruments: industry's innovation with a surgeon's interest. Knee surgery, sports traumatology, arthroscopy : official journal of the ESSKA 2013.


15. Stronach BM, Pelt CE, Erickson J, Peters CL. Patient-specific Total Knee Arthroplasty Required Frequent Surgeon-directed Changes. Clinical orthopaedics and related research 2012.

16. Tigani D, Busacca M, Moio A, Rimondi E, Del Piccolo N, Sabbioni G. Preliminary experience with electromagnetic navigation system in TKA. Knee 2009;16-1:33-8.


18. Cheng T, Zhao S, Peng X, Zhang X. Does computer-assisted surgery improve postoperative leg alignment and implant positioning following total knee arthroplasty? A meta-analysis of randomized controlled trials? Knee surgery, sports traumatology, arthroscopy : official journal of the ESSKA 2012;20-7:1307-22.

19. Boonen B, Schotanus MG, Kerens B, van der Weegen W, van Drumpt RA, Kort NP. Intra-operative results and radiological outcome of conventional and patient-specific surgery in total knee arthroplasty: a

Chapter 6

90

multicentre, randomized controlled trial. Knee surgery, sports traumatology, arthroscopy : official journal of the ESSKA 2013.

20. Nunley RM, Ellison BS, Ruh EL, Williams BM, Foreman K, Ford AD, Barrack RL. Are patient-specific cutting blocks cost-effective for total knee arthroplasty? Clinical orthopaedics and related research 2012;470-3:889-94

91

Chapter 7 Inter-observer reliability of measurements

performed on digital long-leg standing radiographs and assessment of validity

compared to 3D CT-scan

Boonen B., Kerens B., Schotanus M.G.M., Emans P., Jong B., Kort N.P. Knee. 2016 Jan;23(1):20-4.

Chapter 7

92

ABSTRACT

Background Long-leg radiographs (LLR) are often used in orthopaedics to assess limb alignment in patients undergoing total knee arthroplasty (TKA). However, there are still concerns about the adequacy of measurements performed on LLR. We assessed the reliability and validity of measurements on LLR using 3D CT-scan as a gold standard. Methods Six different surgeons measured the mechanical axis and position of the femoral and tibial components individually on 24 LLR. Intraclass correlation coefficients were calcu-lated to obtain reliability and bland-altman plots were constructed to assess agreement between measurements on LLR and measurements on 3D CT-scan. Results ICC agreement for the 6 observer measurements on LLR was 0.70 for the femoral com-ponent and 0.80 for the tibial component.

The mean difference between measurements performed on LLR and 3D CT-scan was 0.3 degrees for the femoral component and -1.1 degrees for the tibial component. Variation of the difference between LLR and 3D CT-scan for the femoral component was 1.1 degrees and 0.9 degrees for the tibial component. 95% of the differences between measurements performed on LLR and 3D CT-scan were between -1.9 and 2.4 degrees (femoral component) and between -2.9 and 0.7 (tibial component). Conclusion Measurements on LLR show moderate to good reliability and, when compared to 3D CT-scan, show good validity.

Inter-observer Reliability Of Measurements Performed On Digital Long-leg Radiographs

93

INTRODUCTION

Standing long-leg radiographs (LLR) are frequently used as an imaging modality to as-sess pre-operative and postoperative leg alignment after total knee arthroplasty (TKA). Literature exists that raises concerns about the adequacy of measurements on LLR. Both the rontgenographic procedure and the measurer contribute to the variability of measurements on LLR1. Foot rotation has a significant effect on the measured values of the mechanical axis2,3 and a combination of flexion and external rotation progressively alters the measured hip-knee-ankle angles4. Measurements on LLR have been com-pared to intraoperative navigation measurements and studies report no correlation between these two types of measurements5,6. However, difference between radio-graphic and navigation measurements of lower limb alignment has been reported to be low if the LLR are obtained in neutral rotation7.

Most published studies that compare measurements on LLR to a CT-scan or intra-operative navigation as gold standards calculate correlation coefficients9,10. This, how-ever, is not a correct outcome measurement for assessing agreement between two measuring techniques10. It only establishes that there is a connection between the out-comes of both techniques, as might be expected. The degree of agreement can better be expressed as precision and accuracy and calculated with the Bland-Altman meth-od10,11.

As far as we know, the agreement between LLR and 3D CT-scan (as a gold standard) has not yet been subject of study. Therefore, the aims of this study are twofold. First to assess absolute reliability of measuring alignment in the frontal plane after TKA on LLR and second to assess the agreement with measurements performed on 3D CT-scan.


We prospectively collected data of a cohort of 26 patients. The cohort consisted of 13 women and 13 men with an average age of 66 (range 52-83 years). The patients partici-pated in a previous study, approved by the local ethical committee12.

Patients were scheduled for TKA and were all operated on using the SignatureTM technique (Biomet, Inc., Warsaw, IN). This technique uses patient-matched positioning guides (PMPG), created using MRI-imaging. During surgery, these guides dictate posi-tioning of traditional cutting blocks to make bony resections. Operative procedure using these guides was described in detail in a previous publication13. The system is designed to construct a neutral mechanical axis and a right-angled position of the femoral and tibial component to the mechanical axis of the femur and tibia, respectively.

According to standard postoperative procedure in our hospital, digital LLR were tak-en at 6 weeks post-surgery. Digital LLR were taken according to the following protocol. To obtain as much of a neutral rotation of the leg as possible, patients were barefoot

Chapter 7

94

and instructed to stand upright with fully extended knees and the heels and greater toes touching. Patients were instructed to stand with their back against the detector. The X-ray beam was directed perpendicular to the detector and centered at the middle of the lower extremity. A digital ruler was projected onto the images and 3 radiographs were taken. These individual radiographs were automatically ‘stitched together’ based on the digital ruler. Brightness was set automatically.

Six independent orthopaedic surgeons were asked to measure the position of the femoral and tibial components of the total knee relative to the mechanical axis of femur and tibia on these radiographs (fig. A.1). Mechanical axis of the femur was defined as a line between the centre of the femoral head and the centre of the distal femur (mid-point between the femoral condyles at the level of the top of the intercondylar notch). Mechanical axis of the tibia was defined as a line between the centre of the tibial plat-eau and the centre of the upper surface of the talus. Measurements were done from digital images on PACS. Positive values indicate varus position (to the contemplated straight-angle) of femoral and tibial components; negative values indicate valgus posi-tion. Each surgeon performed the measurements one time. Hip, knee and ankle joint had to be fully visible on radiographs in order to perform adequate measurements.

For initial study purposes a full-leg CT-scan of the ipsilateral leg was performed 6 weeks post-surgery in this cohort of patients. This scan was made according to a stand-ardised scanning protocol. Digital 3D models were generated for the femur and tibia based on this CT-scan (Mimics, Materialise NV, Leuven, Belgium). These models were used to analyse the positions of the femoral and tibial components in the frontal plane using a coordinate system, relative to the mechanical axis of the femur and tibia, re-spectively12. The mechanical axis of the femur was defined as a line from drawn from the centre of the femoral head to the centre of the distal femur. This distal centre was defined as the intersection point of the femoral AP-axis (line through 2 points on the trochlea) and the transepicondylar axis. The mechanical axis of the tibia was defined as a line drawn from the centre of the tibial plateau (calculated at a level below the osteo-phytes and above the tuberosity) to the centre of the talus (centre point of a circle at the proximal talus at the malleolar level). Measurements on 3D CT-scan were per-formed once by an independent engineer of Materialise NV, Belgium.

The measurements on LLR were compared to corresponding values measured on 3D CT-scan (gold standard). We used the average of the measurements performed on the LLR by the 6 different orthopaedic surgeons for this comparison.


95

Fig. A.1: long-leg radiograph showing measurement of the position of the femoral and tibial components ofthe total knee relative to the mechanical axis of femur and tibia, respectively.

Chapter 7

96

STATISTICAL ANALYSIS

Descriptive statistics

On both LLR and 3D CT-scan, average deviation form planned (neutral mechanical axis and tibial and femoral component right-angled to their respective mechanical axes) position and range of deviation from planned position were obtained (fig. A.2). For measurements on LLR the average of the measured values of all 6 observers was used for this comparison.

Fig. A.2: schematic overview of measurements of the position of the femoral component in the frontal planeon 3D CT-scan using the coordinate system.

Reliability

In order to assess reliability of measurements on LLR, the ICC agreement for measure-ments of the femoral and tibial component were calculated. A two-way mixed effects model was used. ICC values of > 0.75 were considered ‘good’.

Validity

In order to assess validity of measurements on LLR, the agreement between measure-ments on LLR and 3D CT-scan was calculated. For this analysis, the average of meas-urements on LLR of the 6 different observers was used. The Bland-Altman method was used for analysing agreement. Therefore, the mean difference, range of difference and


97

limits of agreement between measurements performed on LLR and 3D CT-scan for both femoral and tibial component were obtained.

SPSS statistics, version 22 was used for all calculations and to construct Bland-Altman plots.

RESULTS

In 2 radiographs the hip joint was insufficiently visible to allow for adequate measure-ments and these radiographs were therefore excluded from the analysis, leaving 24 cases available for analysis.

In table A.1 mean deviation from goal alignment and range of deviation from goal alignment of measurements performed on LLR and 3D CT-scan are summarized. Table A.1: mean deviation from goal alignment and range of deviation from goal alignment of measurements performed on 3D CT-scan and LLR. 1 using the average of the measured values of all 6 observers. The positive numbers indicate varus alignment and negative numbers indicate valgus alignment compared to neutral alignment.

Femoral component Tibial component

LLR1 3D CT-scan LLR1 3D CT-scan

Mean deviation from planned position (SD) 1.3° (1.5) 1.8° (2.1) 1.2° (1.7) 1.8° (1.9)

Range of deviation from planned position (°) -3.9° to 3.5° -2.9° to 4.4° -3.6° to 3° -3.1° to 3.4°

Reliability

ICC agreement for the 6 observer measurements on LLR was 0.70 for the femoral com-ponent and 0.80 for the tibial component.

Validity

Mean difference between LLR and CT-scan was 0.3° for the femoral component and -1.1° for the tibial component. 95% of the differences between measurements per-formed on LLR and 3D CT-scan were between -1.9 and 2.4 degrees (femoral compo-nent) and between -2.9 and 0.7 (tibial component) as calculated using the Bland-Altman method: fig. B.1 (femoral component) and fig. B.2 (tibial component).

Chapter 7

98

Fig. B.1: Bland-Altman plot of alignment measurement by LLR and 3D CT-scan for the femoral component.

Fig. B.2: Bland-Altman plot of alignment measurement by LLR and 3D CT-scan for the tibial component.


99

The mean difference, mean absolute difference, range of difference and limits of agreement between measurements performed on LLR and 3D CT-scan for both femoral and tibial component are presented in table A.2. Table A.2: Mean difference, mean absolute difference, range of difference and limits of agreement between measurements performed on LLR and 3D CT-scan for both femoral and tibial component.

Femoral component Tibial component

Mean difference: X-ray – CT-scan (SD) 0.3° (SD 1.1) -1.1° (SD 0.9)

Mean absolute difference 0.9° 1.2°

Range of difference -1.7° – 2.2° -2.6° – 0.8°

Limits of agreement -1.9 and 2.4 -2.9 and 0.7

DISCUSSION

We analysed the inter-observer variability of measurements performed on LLR by com-paring one single measurement (for both femoral and tibial component) of 6 different orthopaedic surgeons. An ICC of 0.70 was found for measurements of the femoral com-ponent and an ICC of 0.80 was found for measurements of the tibial component. ICC for measurements for the tibial component was therefore classified as ‘good’ and for the femoral component as ‘moderate’. These observations suggest that measurements of alignment of the femoral component are less reliable than measurements of alignment of the tibial component on LLR.

Other authors published on inter-observer reliability of measuring the limb mechan-ical axis on LLR after TKA. They found better correlations than we reported for meas-urements of individual component alignment. Marx et al. analysed inter-observer relia-bility (4 observers) of measurements of mechanical axis on 42 LLR and found ICC’s rang-ing from 0.93 to 0.97 on PACS14. Babazadeh et al. also studied inter-observer correlation (3 observers) of measurements of mechanical axis on 40 LLR and found ICC’s ranging from 0.96 to 0.988. Based on our results and based on the results obtained in these studies, it seems safe to conclude that measurements on LLR are reliable, but in order to assess validity of measurements on LLR, we compared measurements performed on LLR to measurements performed on 3D CT-scan. Measurements on 3D CT-scan have been proven to be highly accurate9 and we therefore used 3D CT-scan as gold standard. To our knowledge, no studies that assess agreement between LLR and 3D CT-scan have been published. Furthermore, when comparing different imaging modalities for as-sessing the accuracy of measuring alignment on LLR, most studies calculate correlation coefficients8,9. This, however, is not a correct outcome measurement for assessing agreement between two measuring techniques11. It only establishes that there is a connection between the outcomes of both techniques, as might be expected. The de-gree of agreement can better be expressed as precision and accuracy10.

Chapter 7

100

Precision describes the scatter of a measurement with a higher variation resulting in a higher standard deviation10. Standard deviation (variation of the difference between LLR and 3D CT-scan) for the femoral component was 1.1 degrees and 0.9 degrees for the tibial component in our study. 95% of the differences between measurements per-formed on LLR and 3D CT-scan were between -1.9 and 2.4 degrees (femoral compo-nent) and between -2.9 and 0.7 (tibial component) as calculated using the Bland-Altman method. In our opinion, 3 degrees is a clinical acceptable error level. Based on these results, measuring on LLR can therefore be labelled as sufficiently precise for routine clinical practice.

Accuracy is the ability of a certain measurement method to measure the true value. The mean difference between measurements performed on LLR and 3D CT-scan was 0.3 degrees for the femoral component and -1.1 degrees for the tibial component. Based on these observations, measuring on LLR can also be labelled accurate.

Radiographs for this study were taken according to a strict protocol. Other authors have also reported on an improved quality of LLR when using a standardised technique for taking LLR15. Literature suggests that further improvements on the quality of meas-urements on LLR can be realised when using PACS software instead of hardcopy14 as was already done in our study. We analysed LLR taken 6 weeks post-surgery. However, analysing alignment at a minimum of 3 months post-surgery has been reported to im-prove quality of measurements16. Lack of extension and improper weight bearing in the early postoperative period might pose a limitation on the quality of obtained LLR16. This might have influenced our results, and purely theoretically, validity of measurements could be enhanced when taking LLR no earlier than 3 months post-surgery.

With respect to the range of deviation from the planned position, the results in this study were comparable to previous reports analysing this particular PMPG system13,17-19.

This study has some other weaknesses. We used a relatively small patient group and we did not assess repeatability of measurements on LLR as others did. According to this existing literature the mechanical axis could be repeatably measured, with test-retest measurements within +/- 2 degrees20. In day-by-day clinical practice, standard lateral radiographs are often used to assess alignment in the sagittal plane. We only studied alignment in the frontal plane on LLR and did not analyse measurements on standard lateral radiographs. Furthermore, the measurements on 3D CT-scan were done by one independent engineer, so repeatability of these measurements was not analysed.

Strengths of this study are that 6 different and independent surgeons assessed alignment on LLR, attributing to the generalizability of our results to the standard day-by-day practise. Furthermore, we used 3D CT-scan as a gold standard for assessing agreement and we used Bland-Altman plots instead of the, often misused, intra- and interclass correlation coefficients for assessing agreement.


101

CONCLUSION

We believe LLR are an adequate tool for analyzing alignment of TKA components in the frontal plane in a day-by-day practice, when a strict protocol is used for taking the LLR. This study quantifies the inherent inaccuracies of measuring on LLR and surgeons should be aware of these potential measurement errors when performing these meas-urements on LLR.

When faced with a malfunctioning TKA however, more detailed information on the position of the TKA is mandatory and we believe measurements of component position should be done using 3D CT-scan scans in these cases. This enables the surgeon to ana-lyse component position in all three anatomical planes.

ACKNOWLEDGEMENTS

We thank Materialise NV for their assistance in analysing the data of the 3D CT-scans. No funding was received for this study.

Chapter 7

102

REFERENCES

1. Huang TL, Wu HT, Liu JC, Chen WM, Chen TH. Do we get a "real" alignment of knee in the preoperative planning of high tibia osteotomy: a prospective study of reproducibility. Journal of the Chinese Medical Association : JCMA 2004;67-4:185-8.

2. Hunt MA, Fowler PJ, Birmingham TB, Jenkyn TR, Giffin JR. Foot rotational effects on radiographic measures of lower limb alignment. Can J Surg 2006;49-6:401-6.

3. Radtke K, Becher C, Noll Y, Ostermeier S. Effect of limb rotation on radiographic alignment in total knee arthroplasties. Arch Orthop Trauma Surg 2010;130-4:451-7.

4. Kannan A, Hawdon G, McMahon SJ. Effect of flexion and rotation on measures of coronal alignment after TKA. The journal of knee surgery 2012;25-5:407-10.

5. Choi WC, Lee S, An JH, Kim D, Seong SC, Lee MC. Plain radiograph fails to reflect the alignment and advantages of navigation in total knee arthroplasty. J Arthroplasty 2011;26-5:756-64.

6. Yaffe MA, Koo SS, Stulberg SD. Radiographic and navigation measurements of TKA limb alignment do not correlate. Clin Orthop Relat Res 2008;466-11:2736-44.

7. Dexel J, Kirschner S, Gunther KP, Lutzner J. Agreement between radiological and computer navigation measurement of lower limb alignment. Knee Surg Sports Traumatol Arthrosc 2013.

8. Babazadeh S, Dowsey MM, Bingham RJ, Ek ET, Stoney JD, Choong PF. The long leg radiograph is a reliable method of assessing alignment when compared to computer-assisted navigation and computer tomography. The Knee 2013;20-4:242-9.

9. Hirschmann MT, Konala P, Amsler F, Iranpour F, Friederich NF, Cobb JP. The position and orientation of total knee replacement components: a comparison of conventional radiographs, transverse 2D-CT slices and 3D-CT reconstruction. J Bone Joint Surg Br 2011;93-5:629-33.

10. Savenije OE, Brand PL. [Weighing before and after feeding: an unreliable method for estimating milk intake in infants]. Nederlands tijdschrift voor geneeskunde 2007;151-49:2718-22.

11. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1-8476:307-10.

12. Boonen B, Schotanus MG, Kerens B, Hulsmans FJ, Tuinebreijer WE, Kort NP. Patient-specific positioning guides for total knee arthroplasty: no significant difference between final component alignment and pre-operative digital plan except for tibial rotation. Knee Surg Sports Traumatol Arthrosc 2015.

13. Boonen B, Schotanus MG, Kort NP. Preliminary experience with the patient-specific templating total knee arthroplasty. Acta Orthop 2012;83-4:387-93.

14. Marx RG, Grimm P, Lillemoe KA, Robertson CM, Ayeni OR, Lyman S, Bogner EA, Pavlov H. Reliability of lower extremity alignment measurement using radiographs and PACS. Knee Surg Sports Traumatol Arthrosc 2011;19-10:1693-8.

15. Siu D, Cooke TD, Broekhoven LD, Lam M, Fisher B, Saunders G, Challis TW. A standardized technique for lower limb radiography. Practice, applications, and error analysis. Investigative radiology 1991;26-1:71-7.

16. Hauschild O, Konstantinidis L, Baumann T, Niemeyer P, Suedkamp NP, Helwig P. Correlation of radiographic and navigated measurements of TKA limb alignment: a matter of time? Knee Surg Sports Traumatol Arthrosc 2010;18-10:1317-22.

17. Roh YW, Kim TW, Lee S, Seong SC, Lee MC. Is TKA using patient-specific instruments comparable to conventional TKA? A randomized controlled study of one system. Clin Orthop Relat Res 2013;471-12:3988-95.

18. Boonen B, Schotanus MG, Kerens B, van der Weegen W, van Drumpt RA, Kort NP. Intra-operative results and radiological outcome of conventional and patient-specific surgery in total knee arthroplasty: a multicentre, randomized controlled trial. Knee Surg Sports Traumatol Arthrosc 2013;21-10:2206-12.

19. Victor J, Dujardin J, Vandenneucker H, Arnout N, Bellemans J. Patient-specific guides do not improve accuracy in total knee arthroplasty: a prospective randomized controlled trial. Clin Orthop Relat Res 2014;472-1:263-71.


103

20. Skytta ET, Haapamaki V, Koivikko M, Huhtala H, Remes V. Reliability of the hip-to-ankle radiograph in determining the knee and implant alignment after total knee arthroplasty. Acta orthopaedica Belgica 2011;77-3:329-35.

105

Chapter 8

General discussion

General Discussion

107

GENERAL DISCUSSION

In this thesis one specific knee system was investigated that was used in the hospital(s) where this research was conducted. The SignatureTM system (Biomet, Warsaw, IND), together with the Vanguard total knee prosthesis (Biomet, Warsaw, IND) was subject of study. Therefore, it must be emphasized that the results presented in this thesis do not automatically apply to other similar systems, as there are substantial differences in algorithms used to calculate implant position. Furthermore, we chose to use the name ‘patient-matched positioning guides (PMPG)’. There are numerous synonyms in current literature, however. Patient-specific instruments, patient-specific guides, patient-specific positioning guides, patient-specific alignment guides, patient-specific cutting guides, custom-fit cutting guides, custom-fit positioning guides, etc. all refer to the same technique of individualized alignment guides fabricated using a rapid prototyping technique.

This discussion section is separated into 3 parts. In the first part, answers will be postulated on the questions asked at the beginning of this thesis and the results pre-sented in this thesis will be placed in a broader context. The second part will describe which challenges will be faced when this technique would be used on a larger scale. In the final part, the implications of this thesis for clinical practise and future research will be discussed.

PART 1: ANSWERS TO QUESTIONS ASKED AND PMPG IN A BROADER CONTEXT

1. Can we rely on PMPG for aligning a TKA?

In chapter 2 we analyzed the potential of PMPG to adequately reproduce the preopera-tive surgical plan. It was hypothesized that there would be no difference between post-op implant position and pre-op digital plan. Using a very reliable 3D CT-scan technique that permitted matching of digital plan on the preoperative MRI-scan to a postoperative CT-scan we found that PMPG was able to adequately reproduce the pre-op plan in all planes, except for the tibial rotation in the transverse plane. Therefore, we state that PMPG are a reliable technique for outline a TKA, at least in the hands of an experienced, high-volume surgeon. There are, however, some potential pitfalls when using PMPG that are discussed in the following section.

When answering the question of reliability of PMPG for aligning a TKA, it is im-portant to realise that there can be significant variation in the different systems availa-ble on the market as calculation algorithms are not universal. High quality literature comparing different PMPG systems is virtually non-existing. There is literature, howev-er, comparing PMPG based on CT-scan and PMPG based on MRI-scan. Results are con-

Chapter 8

108

flicting. Some authors report superior accuracy with MRI over CT1,2, others report com-parable results3 and White et al. found inferior results with PMPG based on MRI-scan as compared to CT-scan4.

A second important factor when discussing the reliability of PMPG is the ‘surgeon factor’. As opposed to conventional instruments (CI), where decisions concerning bone cuts are made during surgery itself, all calculations in PMPG have been done in the preoperative time-frame. As a consequence, there are fewer ‘check-points’ during sur-gery. This implies that the surgeon has to learn to trust the system and in our experi-ence this takes at least a 100 cases. The surgeon has to develop a sense and a reference frame of when the guides are stable and in the right position on the native anatomy. Therefore, we feel that PMPG are not a technique that enables every surgeon to per-form a TKA without proper training (see also: Part 2: challenges when introducing PMPG on a larger scale).

Apart from this need for adequate surgeons training, PMPG are a generally applica-ble technique as there are no restrictions for performing a TKA when extra-articular deformities are present. In this respect, there are only few cases in which PMPG TKA might not be possible. Especially claustrophobic patients might not be suitable for PMPG that are based on MRI-scans. Additionally, these MRI-based systems do not allow for accurate pre-operative planning in patients with retained hardware near the knee joint. CT-based systems can be an escape in these cases.

2. What are the potential pitfalls when using PMPG that compromise their accuracy?

From chapter 2 it is clear that there are some potential sources of error when using PMPG. First, we advise against using the microplasty instruments for performing the chamfer cuts. Due to lack of stability of this sawing slot, there is an inherent risk of malalignment. Furthermore, attention is needed not to place the femoral guide in too much flexion. For the tibial guide it is mandatory to press the guide as far medial as possible to eliminate the risk of exorotation. In addition, we advise using the pulse lav-age system to visualize the drill holes in the tibial plateau that dictate rotational align-ment. We did not further analyze these sources in separate studies to quantify the role of each of these factors as sources of error.

3. Can PMPG reduce the percentage of outliers in TKA compared to conventional instruments (CI)?

To calculate the percentages of outliers, traditionally a +/- 3 degree range of deviations from goal alignment in the frontal plane (varus/valgus) is perceived as being accepta-ble5,6. In chapter 1 and chapter 3 PMPG were compared to CI with respect to their abil-

General Discussion

109

ity to obtain a neutral mechanical axis after TKA. From these investigations it wat clear that the percentages of outliers in the frontal plane cannot be reduced significantly with the use of PMPG. Most research on PMPG has focused on alignment in the frontal plane and results are conflicting, as discussed in chapter 3. More recent meta-analyses and reviews performed on the subject of restoring a neutral mechanical axis indicate that there is no difference between CI and PMPG when assessing their ability to obtain this neutral axis and the number of outliers in mechanical axis7-14.

Less research has been conducted concerning alignment in the sagittal (flex-ion/extension and anterior/posterior slope) and transverse plane (internal and external rotation). We assessed alignment in the sagittal plane in Chapter 1 and chapter 3, but used X-rays for this analysis. As discussed already in these chapters, there are some major limitations when using these X-rays. Literature that uses CT-scan to evaluate alignment in this plane in general shows no superiority of PMPG over CI15-17. We did not assess rotational alignment of the femoral and tibial component. This is a topic least studied in literature as well. Given that malrotation of the components can lead to sig-nificant patellar tracking problems, knee stiffness, poor motion and can be a source of persistent anterior knee pain18,19, it is important to address this issue as well. Again, results in literature are conflicting, with some authors reporting a reduction in outliers for the femoral component rotation17,20 and some report no difference in the number of outliers when comparing PMPG to CI8,11,15,16,21,22. Fewer authors report on the out-come of tibial component rotation; some authors report a smaller chance of internal malrotation with less dispersion and amplitude for the tibial component rotation around the neutral position23 and some authors report no difference when comparing PMPG with CI8.

A major problem with the discussion on alignment in the different anatomical planes is that there is no clear consensus about the optimal alignment of the femoral and tibial component24. Gromov et al. performed a literature search to summarise the experience in this field25. Based on their research, they made some recommendations for clinical practise. Concerning overall tibiofemoral alignment they concluded that neutral coronal alignment is still the gold standard (despite conflicting reports) and it therefore should be aimed for until there is conclusive evidence to suggest otherwise. To obtain this, the femoral component should be placed in 2-7° of valgus with respect to the anatomical axis of the femur and the tibial component should be placed in neutral alignment (90°) with respect to the anatomical axis of the tibia. In the sagittal plane it should be aimed to place the femoral component in 0-3° of flexion and the tibial component in 0-7° pos-terior slope relative to the intramedullary axes. With regard to femoral rotation, several reference axes can be used. The surgical transepicondylar axis (sTEA) is the most com-monly used. They conclude that internal rotation of the femoral component should be avoided and 2-5° of external rotation in relation to the sTEA seems the optimal range. There is no gold standard for measuring tibial rotation. As excessive internal rotation

Chapter 8

110

can lead to knee pain, this should be avoided. It is clear from these recommendations that ongoing research to clarify the optimal alignment goals is still needed.

Most knee surgeons adhere to this mechanical point of view for performing a TKA and therefore rely on restoring the patient’s hip-knee-ankle axis and placing the im-plants in relation to the body’s mechanical or anatomical axes. However, there are groups that use a kinematical approach to align the components of a TKA. In kinemati-cally aligned TKA the femoral and tibial components are positioned so that the angles and the levels of the distal and posterior femoral joint line and the tibial joint line are each restored to the patient’s natural alignment26. Few studies have compared the two alignment techniques27,28 and no clear recommendations can be made regarding the optimal alignment strategy. More studies are therefore needed to investigate whether kinematic alignment can reproduce superior survivorship and functional outcome fol-lowing TKA.

To complicate matters further, there is substantial difference in design philosophy of total knee implants: single radius versus multiradius femoral design, different or equal size medial and lateral condylar surface of femoral component, etc. No clear evidence exists for superiority of one design philosophy over the other and therefore, again, more research is needed comparing different design philosophies.

4. Will use of PMPG lead to shorter operation time and less blood loss compared to CI?

Peroperative blood loss could be reduced with 100mL on average when using PMPG, without reducing the number of needed transfusions of packed cells (chapter 3). The operation time can be reduced with 5 minutes when using PMPG (chapter 3). However, most other RCT’s and meta-analyses fail to demonstrate a significant reduction in oper-ation time and blood loss compared to conventional instruments9,13,29-31.

Although this was not subject of a specific study, time can be saved when also look-ing at time to install surgical instruments on sterile fields in the OR. A reliable reduction of instrument sets needed to perform the surgery can be obtained with PMPG9 and a decrease in operating room turnover times has been reported by some authors32. Whether the time gain associated with this reduction in trays is clinically relevant de-pends probably on the setting in which PMPG are used. In most centres it can be argued that the time needed to transfer patients is far more time consuming than surgery it-self. In high-volume centres with optimal logistics, however, this OR-time reduction per case might result in higher number of cases performed on a daily basis.

5. Will use of PMPG lead to shorter duration of hospitalization compared to CI?

The duration of hospitalization is not reduced whit PMPG (chapter 3). This is in accord-ance with other recent high-quality research29,30.

General Discussion

111

6. Will use of PMPG lead to better functional outcome than CI?

As described in the randomized controlled trial in chapter 4, the clinical outcome of PMPG TKA and CI is not different. To date, there are no long follow-up studies available. Given that there are no major differences between CI and PMPG when it comes to ob-taining certain alignment goals and given that the implanted prosthesis is identical with CI and PMPG, it seems reasonable to assume that longer clinical follow-up will neither show any differences in clinical outcome between CI and PMPG.

In recent years there has been a shift from examiner-based outcome measures to patient reported outcome measures (PROMS). Without any doubt, this has been an improvement when assessing outcomes of TKA. PROMS have some drawbacks, howev-er, that limit their use. They suffer from a ceiling effect and their pain dominance masks the functional changes33. This is of little importance when assessing an older population that underwent TKA, as their general expectation is that of pain relief. However, the number of younger, more active and more demanding patients receiving a total knee is growing. Kurtz et al have reported that patients under 55 years of age comprise the most rapidly growing subset of patients undergoing TKA in the United States. They have estimated that as many as 900,000 total knee replacements may be performed in this patient group by the year 203034. This population expects more of their joint replace-ment than just pain relief, and pain-dominant PROMS are therefore likely not the ideal outcome measures. Recent studies have suggested that between 30% and 50% of younger TKA patients experience residual symptoms during various functional activi-ties35,36. As a consequence, there is need for improved PROMS that are able to discrimi-nate successful TKA’s from unsuccessful TKA’s in this younger population. It is by focus-sing on the outcomes of TKA in this high demanding population that we have to draw the conclusion that we are still far from reconstructing biomechanics of the natural knee joint and continued research is still mandatory to improve results of TKA in these younger, more active patients.

7. Are PMPG safe to use?

PMPG are safe to use. There was no difference in TKA related complication rate be-tween PMPG and CI (chapter 4). No adverse events occurred specifically related to the use of PMPG in cohort of 200 patients (chapter 5). Literature on this topic is scarce, but the reports that have addressed this issue do not show any difference between CI and PMPG29.

8. Are PMPG able to accurately predict implant size preoperatively?

PMPG predicted the implanted component size exact in 88.0% of femurs and 70.5% of tibias. Implanted component size prediction, with an error of one size, was correct in

Chapter 8

112

98.5% of femoral components and 97.0% of tibial components. Reviewing and approv-ing the preoperative plan, reduced the number of changes that were necessary during surgery from 21.5% to 14% for the femoral component and from 41% to 29.5% for the tibial component. The accuracy of PMPG to predict implant size is therefore acceptable, potentially allowing for reduction in hospital stocks as argued in chapter 5. The im-portance of approving the pre-operative digital plan has been highlighted by others as well37. We agree that this is a critical step when introducing and performing PMPG TKA. The accuracy of PMPG to predict implant size is an important issue in the cost-effectiveness analysis of PMPG. This topic was not studied in this thesis. As discussed in chapter 4, PMPG allow for cost reduction in certain steps/logistics of the TKA procedure, but is also associated with additional costs when compared to CI. A definitive answer as to whether PMPG are cost-effective is difficult, given the strong variation in local costs of the different cost-saving and cost-increasing factors associated with PMPG TKA. Local research on a smaller scale is likely needed to answer this question for every hospi-tal/institution individually.

9. What is the inter-observer reliability of measurements performed on long-leg radiographs (LLR) and is it a valid technique compared to measurements performed on 3D-CT?

LLR are a frequently used tool to assess outcome of TKA surgery. They are especially useful for analyzing the lower limb axis post-surgery. LLR are relatively cheap and easily available in most hospitals. We used them as well to assess alignment outcome in our studies presented in this thesis. Despite their widespread use, however, there were still some concerns about the reliability and validity of measurements performed on LLR as research addressing this issue used improper statistics to allow for definite conclusions.

We assessed inter-observer reliability of measurements performed on LLR by asking six different surgeons to measure the position of the femoral and tibial components individually on 24 LLR and presented our results in chapter 6. Intraclass correlation coef-ficient (ICC) agreement for the 6 observer measurements on LLR was 0.70 for the femo-ral component and 0.80 for the tibial component.

The mean difference between measurements performed on LLR and 3D CT-scan was 0.3 degrees for the femoral component and -1.1 degrees for the tibial component. Variation of the difference between LLR and 3D CT-scan for the femoral component was 1.1 degrees and 0.9 degrees for the tibial component. 95% of the differences between measurements performed on LLR and 3D CT-scan were between -1.9 and 2.4 degrees (femoral component) and between -2.9 and 0.7 (tibial component). Therefore, we con-cluded that measurements on LLR show moderate to good reliability and, when com-pared to 3D CT-scan, show good validity.

General Discussion

113

PART 2: CHALLENGES WHEN INTRODUCING PMPG ON A LARGER SCALE

The interest in the use of PMPG is high. Thienpont et al. performed a survey among Belgian knee surgeons and 80% of surgeons indicated they had a great interest in using PMPG in TKA38. If the number of surgeons performing PMPG TKA will continue to in-crease, there are some challenges ahead.

Perhaps the most important question to be asked is how to deal with medicolegal issues. To construct the digital pre-operative plan, engineers have to identify anatomical landmarks on 3-dimentional images of femur and tibia, constructed using MRI or CT-scans. This identification is essential to obtain an optimal position of the TKA compo-nents. The digital pre-operative plan ideally has to be approved by the surgeon, howev-er, this is not mandatory for all PMPG systems although there is a tendency to make this a mandatory step in the process. In these cases, when the deadline of approval has elapsed, the default plan constructed by the engineer is automatically approved. Medi-co-legal experts point out that companies are responsible if PMPG are created and delivered without validation of the planning by the surgeon38. To avoid these legal is-sues, it is therefore mandatory that companies demand approval of the digital plan by the operating surgeon prior to guide fabrication. There is no doubt that the orthopaedic surgeon in the end is responsible for the correct placement of the TKA components. But the question is whether the orthopaedic surgeon can take this responsibility if he or she was not involved in the registration of the anatomical landmarks that lead to the digital plan, especially since research showed that there is substantial interobserver variability when identifying these landmarks on pre-operative MRI-scans39,40. This is an interesting ongoing discussion that has not been closed yet.

Second major challenge when using PMPG is the peri-operative logistics issue. The pathway of performing an MRI or CT-scan, uploading the images to the engineers server for construction of the digital plan, checking and approval of the digital plan by the operating surgeon, in time delivery of the guides and digital plan and sterilisation of the guides for peroperative use. As a first step, this implies a thorough collaboration with the radiology department of the hospital. Not only to guarantee sufficient MRI or CT capacity to scan all patients eligible for TKA, but also dedicated personnel to upload the images to the engineers server. Second, it demands dedication from the orthopaedic surgeon to review, adjust and approve the digital plan in time. From a medicolegal point of view, this is an essential step in the production process of the guides as discussed previously. Third, OR managers have to be made responsible for checking the in time delivery and sterilisation of the guides.

The third challenge is in the training of orthopaedic residents. Residents will have to learn to master both CI TKA and PMPG TKA and not just PMPG TKA. When faced with non-fitting guides during surgery for example, surgeons still have to be able to perform the procedure by use of CI. Furthermore, learning the different surgical steps and rea-soning behind the alignment of the components in different planes is probably best

Chapter 8

114

done by using CI or computer assisted surgery (CAS), as direct peroperative feedback is offered in this way. When these steps and the theoretical knowledge of performing a CI or CAS TKA are sufficiently mastered, PMPG can be safely learned. It is essential that this learning process includes the pre-operative reviewing of the digital plan as well, as this is an essential part of performing PMPG TKA as discussed previously.

Concerning training of orthopaedic surgeons that want to start using PMPG, some remarks have to be placed. As described in chapter 2, there are some potential pitfalls when starting using PMPG that might compromise its accuracy. Based on our research we believe PMPG are not a technique that enables every surgeon, also the less experi-enced ones, to adequately perform a TKA. At the very least some form of basic training addressing the potential pitfalls by experienced PMPG-users will be mandatory in order to avoid malaligned TKA’s. Multiple possible education strategies can be brought for-ward to obtain this goal. It might, for example, be of interest to explore the potential of expert opinion centres. Alternatively, this training could be made available by the or-thopaedic implant industry together with some key-opinion leaders. We have not inves-tigated in this thesis what form of educational training would work best, and it would be of interest to cover this topic in the future.

Besides this basic training in elementary principles of using PMPG, starting surgeons will have to get used to trusting the system. Most companies have the possibility to order bone-models with the guides. This means that 3-dimensional models of the distal femur and proximal tibia can be ordered to give insight in where the guides have to be positioned on the native anatomy during surgery. Based on the research presented in this thesis, we advise starting PMPG users to order these bone models in the first cases to get a better understanding of where to press on the guides to obtain the predeter-mined accurate fit to the native anatomy.

In the Netherlands, every total knee that has been implanted is registered in a cen-tral database (LROI: landelijke registratie orthopedische implantaten). Information on basic patient demographics (ASA classification, prior surgery), the type of prosthesis that has been placed (cemented or uncemented, bone graft used yes or no) and the surgical approach that has been used are collected. However, information regarding the type of alignment method used (conventional, computer-assisted, PMPG) is not gath-ered. With the introduction of more alignment techniques, it might be of interest, how-ever, to collect this information as well. This will make it easier in the long term to draw conclusions on safety and number of needed revision procedures with each alignment technique.

General Discussion

115

PART 3: IMPLICATION OF THIS THESIS FOR CLINICAL PRACTISE AND FUTURE RESEARCH

Based on the conclusions of the research presented in this thesis, it is safe to state that PMPG deserve a place in the field of TKA surgery as an alignment technique. However, the technique is not superior in any particular outcome compared to conventional in-strumented TKA and a widespread change from CAS or CI to PMPG for aligning a TKA is therefore not immediately mandatory.

There are some remaining research questions that need to be addressed. We have not studied the cost-effectiveness of PMPG compared to CI. This is still an important issue that needs attention in future research.

Furthermore, we only briefly explored the potential of PMPG to contribute to the ‘knee in a box’ philosophy. Theoretically, by using imaging techniques to calculate im-plant size prior to surgery, it should be possible to deliver the appropriate prosthesis component together with the alignment guides as ‘one box’ to the centre carrying out the procedure. We briefly explored the potential of PMPG to adequately predict im-plant size in chapter 5, but further research is still needed to provide a definitive answer as to what extend the technique enables surgeons and hospitals to reduce their implant stocks.

Last but not least, we urgently need more research investigating the optimal align-ment of TKA’s, as there is no clear consensus about this topic as discussed in part 1 of this general discussion section.

BOTTOM LINE

At present it seems PMPG are more an evolution in TKA than a revolution. It is a promis-ing technique, but there are still numerous factors that have to be analysed in order to define the future perspective.

Chapter 8

116

REFERENCES

1. Frye BM, Najim AA, Adams JB, Berend KR, Lombardi AV, Jr. MRI is more accurate than CT for patient-specific total knee arthroplasty. The Knee 2015.

2. Silva A, Pinto E, Sampaio R. Rotational alignment in patient-specific instrumentation in TKA: MRI or CT? Knee surgery, sports traumatology, arthroscopy : official journal of the ESSKA 2014.

3. Asada S, Mori S, Matsushita T, Nakagawa K, Tsukamoto I, Akagi M. Comparison of MRI- and CT-based patient-specific guides for total knee arthroplasty. The Knee 2014;21-6:1238-43.

4. White D, Chelule KL, Seedhom BB. Accuracy of MRI vs CT imaging with particular reference to patient specific templates for total knee replacement surgery. The international journal of medical robotics + computer assisted surgery : MRCAS 2008;4-3:224-31.

5. Bonner TJ, Eardley WG, Patterson P, Gregg PJ. The effect of post-operative mechanical axis alignment on the survival of primary total knee replacements after a follow-up of 15 years. The Journal of bone and joint surgery. British volume 2011;93-9:1217-22.

6. Parratte S, Pagnano MW, Trousdale RT, Berry DJ. Effect of postoperative mechanical axis alignment on the fifteen-year survival of modern, cemented total knee replacements. The Journal of bone and joint surgery. American volume 2010;92-12:2143-9.

7. Voleti PB, Hamula MJ, Baldwin KD, Lee GC. Current data do not support routine use of patient-specific instrumentation in total knee arthroplasty. The Journal of arthroplasty 2014;29-9:1709-12.

8. Thienpont E, Schwab PE, Fennema P. A systematic review and meta-analysis of patient-specific instrumentation for improving alignment of the components in total knee replacement. The bone & joint journal 2014;96-B-8:1052-61.

9. Sassoon A, Nam D, Nunley R, Barrack R. Systematic review of patient-specific instrumentation in total knee arthroplasty: new but not improved. Clinical orthopaedics and related research 2015;473-1:151-8.

10. Cavaignac E, Pailhe R, Laumond G, Murgier J, Reina N, Laffosse JM, Berard E, Chiron P. Evaluation of the accuracy of patient-specific cutting blocks for total knee arthroplasty: a meta-analysis. International orthopaedics 2015;39-8:1541-52.

11. Jiang J, Kang X, Lin Q, Teng Y, An L, Ma J, Wang J, Xia Y. Accuracy of patient-specific instrumentation compared with conventional instrumentation in total knee arthroplasty. Orthopedics 2015;38-4:e305-13.

12. Mannan A, Smith TO, Sagar C, London NJ, Molitor PJ. No demonstrable benefit for coronal alignment outcomes in PSI knee arthroplasty: A systematic review and meta-analysis. Orthopaedics & traumatology, surgery & research : OTSR 2015;101-4:461-8.

13. Sharareh B, Schwarzkopf R. Review article: Patient-specific versus standard instrumentation for total knee arthroplasty. Journal of orthopaedic surgery 2015;23-1:100-6.

14. Zhang QM, Chen JY, Li H, Chai W, Ni M, Zhang ZD, Yang F. No evidence of superiority in reducing outliers of component alignment for patient-specific instrumentation for total knee arthroplasty: a systematic review. Orthopaedic surgery 2015;7-1:19-25.

15. Marimuthu K, Chen DB, Harris IA, Wheatley E, Bryant CJ, MacDessi SJ. A multi-planar CT-based comparative analysis of patient-specific cutting guides with conventional instrumentation in total knee arthroplasty. The Journal of arthroplasty 2014;29-6:1138-42.

16. Woolson ST, Harris AH, Wagner DW, Giori NJ. Component alignment during total knee arthroplasty with use of standard or custom instrumentation: a randomized clinical trial using computed tomography for postoperative alignment measurement. The Journal of bone and joint surgery. American volume 2014;96-5:366-72.

17. Chotanaphuti T, Wangwittayakul V, Khuangsirikul S, Foojareonyos T. The accuracy of component alignment in custom cutting blocks compared with conventional total knee arthroplasty instrumentation: prospective control trial. The Knee 2014;21-1:185-8.

18. Barrack RL, Schrader T, Bertot AJ, Wolfe MW, Myers L. Component rotation and anterior knee pain after total knee arthroplasty. Clinical orthopaedics and related research 2001-392:46-55.

General Discussion

117

19. Bedard M, Vince KG, Redfern J, Collen SR. Internal rotation of the tibial component is frequent in stiff total knee arthroplasty. Clinical orthopaedics and related research 2011;469-8:2346-55.

20. Heyse TJ, Tibesku CO. Improved femoral component rotation in TKA using patient-specific instrumentation. The Knee 2012.

21. Roh YW, Kim TW, Lee S, Seong SC, Lee MC. Is TKA Using Patient-specific Instruments Comparable to Conventional TKA? A Randomized Controlled Study of One System. Clinical orthopaedics and related research 2013.

22. 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 surgery, sports traumatology, arthroscopy : official journal of the ESSKA 2013.

23. Silva A, Sampaio R, Pinto E. Patient-specific instrumentation improves tibial component rotation in TKA. Knee surgery, sports traumatology, arthroscopy : official journal of the ESSKA 2013.

24. Blumenfeld TJ. CORR Insights (R): Is TKA using patient-specific instruments comparable to conventional TKA? A randomized controlled study of one system. Clinical orthopaedics and related research 2013;471-12:3996-7.

25. Gromov K, Korchi M, Thomsen MG, Husted H, Troelsen A. What is the optimal alignment of the tibial and femoral components in knee arthroplasty? Acta orthopaedica 2014;85-5:480-7.

26. Howell SM, Howell SJ, Kuznik KT, Cohen J, Hull ML. Does a kinematically aligned total knee arthroplasty restore function without failure regardless of alignment category? Clinical orthopaedics and related research 2013;471-3:1000-7.

27. Nogler M, Hozack W, Collopy D, Mayr E, Deirmengian G, Sekyra K. Alignment for total knee replacement: a comparison of kinematic axis versus mechanical axis techniques. A cadaver study. International orthopaedics 2012;36-11:2249-53.

28. Dossett HG, Swartz GJ, Estrada NA, LeFevre GW, Kwasman BG. Kinematically versus mechanically aligned total knee arthroplasty. Orthopedics 2012;35-2:e160-9.

29. Kotela A, Lorkowski J, Kucharzewski M, Wilk-Franczuk M, Sliwinski Z, Franczuk B, Legosz P, Kotela I. Patient-Specific CT-Based Instrumentation versus Conventional Instrumentation in Total Knee Arthroplasty: A Prospective Randomized Controlled Study on Clinical Outcomes and In-Hospital Data. BioMed research international 2015;2015:165908.

30. Abane L, Anract P, Boisgard S, Descamps S, Courpied JP, Hamadouche M. A comparison of patient-specific and conventional instrumentation for total knee arthroplasty: a multicentre randomized controlled trial. The bone & joint journal 2015;97-B-1:56-63.

31. Hamilton WG, Parks NL, Saxena A. Patient-Specific Instrumentation Does Not Shorten Surgical Time: A Prospective, Randomized Trial. The Journal of arthroplasty 2013.

32. DeHaan AM, Adams JR, DeHart ML, Huff TW. Patient-specific versus conventional instrumentation for total knee arthroplasty: peri-operative and cost differences. The Journal of arthroplasty 2014;29-11:2065-9.

33. Terwee CB, van der Slikke RM, van Lummel RC, Benink RJ, Meijers WG, de Vet HC. Self-reported physical functioning was more influenced by pain than performance-based physical functioning in knee-osteoarthritis patients. Journal of clinical epidemiology 2006;59-7:724-31.

34. Kurtz SM, Lau E, Ong K, Zhao K, Kelly M, Bozic KJ. Future young patient demand for primary and revision joint replacement: national projections from 2010 to 2030. Clinical orthopaedics and related research 2009;467-10:2606-12.

35. Nam D, Nunley RM, Barrack RL. Patient dissatisfaction following total knee replacement: a growing concern? The bone & joint journal 2014;96-B-11 Supple A:96-100.

36. Parvizi J, Nunley RM, Berend KR, Lombardi AV, Jr., Ruh EL, Clohisy JC, Hamilton WG, Della Valle CJ, Barrack RL. High level of residual symptoms in young patients after total knee arthroplasty. Clinical orthopaedics and related research 2014;472-1:133-7.


Chapter 8

118

38. Thienpont E, Bellemans J, Delport H, Van Overschelde P, Stuyts B, Brabants K, Victor J. Patient-specific instruments: industry's innovation with a surgeon's interest. Knee surgery, sports traumatology, arthroscopy : official journal of the ESSKA 2013.

39. Park A, Nam D, Friedman MV, Duncan ST, Hillen TJ, Barrack RL. Inter-observer precision and physiologic variability of mri landmarks used to determine rotational alignment in conventional and patient-specific TKA. The Journal of arthroplasty 2015;30-2:290-5.

40. Goyal N, Stulberg SD. Evaluating the Precision of Preoperative Planning in Patient Specific Instrumentation: Can a Single MRI Yield Different Preoperative Plans? The Journal of arthroplasty 2015;30-7:1250-3.

119

Chapter 9

Valorisation

Chapter 9

120

This thesis primarily describes the results of one patient-matched positioning guides (PMPG) system in total knee arthroplasty. PMPG uses patient-specific guides to make bony resections of femur and tibia. Imaging techniques and a specific software program are used to create virtual models of the patient’s femur and tibia. The virtual models are then used to identify anatomical landmarks of the knee which are used to calculate ideal implant size and positioning. The final step is to create alignment guides that have only one fitting position on the patient’s individual femoral and tibial anatomy. They dictate bony resections in the preparation for prosthesis placement.

Main subjects of investigation were: alignment and clinical outcome obtained with this technique, safety and potential of the technique to adequately predict implant size. We compared PMPG to conventional instruments and concluded that they performed equally well with respect to obtaining a correct alignment, clinical outcomes, length of hospital stay and safety. A small reduction in blood loss and operation time was ob-served. Based on our observations, there is nothing against continued use of PMPG for performing a TKA. Import remaining issue is that of the cost-benefit of PMPG, com-pared to conventional instruments. This was not studied in detail in this thesis. We briefly investigated the potential of PMPG to adequately predict implant size preopera-tively and found reasonable accuracy in this respect. These sizing capabilities of PMPG are essential to bring the ‘knee in a box’ philosophy a step closer, meaning the in time delivery of adequately sized prosthesis components and resection guides for every single TKA patient. This, in turn, is essential to reduce hospital prosthesis stock and to obtain significant cost reduction. Currently, high quality studies investigating in detail the exact potential of the technique to obtain this goal are lacking. Future research will therefore be necessary to address this issue in detail.

Our results are important, primarily for patients and their treating physicians, alt-hough they also serve for orthopaedic companies to improve the concept of PMPG. Furthermore, our results could be important to health insurance companies for deter-mining their compensation policies. The issues discussed in this thesis will hopefully enable orthopaedic surgeons to make an informed decision as to whether or not to start using PMPG and when they choose to do so, what type of pitfalls are associated with their use. In any case, adequate surgeon training seems mandatory before starting to use PMPG. If PMPG would be implemented on a larger scale, this would impose some challenges as well. More specifically, the practise of training of orthopaedic sur-geons will likely have to be altered. In this respect, it remains important for trainees to master the technique of conventionally instrumented TKA. This ensures that surgeons can always fall back on traditional instruments when e.g. guides are not sterile/drop down on the theatre floor, performing revision surgery, guides to not fit adequately etc.

We used questionnaires (patient reported outcome measures, PROM’s) to study the clinical outcome of PMPG. When analysing these clinical results, we discussed the limi-tations of such questionnaires. As the proportion of younger, more active and more demanding patients undergoing TKA is rising, function after TKA is becoming more im-

Valorisation

121

portant. A search for PROM’s allowing for greater differentiation of level of function between patients in assessing performance after TKA or total hip arthroplasty seems mandatory in order to appreciate subtle variations in function.

When analysing limb alignment after TKA, long-leg radiographs (LLR) are frequently used in daily practise. Although used on a large scale, the reliability and validity of measurements performed on these LLR was still debated. We concluded in this thesis that LLR show moderate to good reliability and, when compared to 3D CT-scan, show good validity. Therefore, no objections exist to continue their use in clinical practise.

123

Chapter 10 Summary

Nederlandse samenvatting

Summary

125

SUMMARY

In this thesis the potential of patient-matched positioning guides (PMPG) as an align-ment tool for performing a total knee arthroplasty (TKA) was explored. PMPG uses pa-tient-specific guides to make bony resections of femur and tibia. Imaging techniques and a specific software program are used to create virtual models of the patient’s femur and tibia. CT-scan or MRI-scan are the basic imaging modalities. The virtual models are then used to identify anatomical landmarks of the knee which are used to calculate ideal implant size and positioning. These calculations subsequently result in a virtual plan of the operation to be performed in theatre. The final step is to create alignment guides that have only one fitting position on the patient’s individual femoral and tibial anatomy. They dictate bony resections in the preparation for prosthesis placement. In chapter 1 we discussed that the number of patients receiving a TKA has risen sub-stantially over de past years and is expected to continue to rise in the years to come. We briefly discussed conventional instruments and computer assisted surgery as align-ment methods for performing a TKA and outlined most of the disadvantages associated with these techniques. PMPG attempts to address these disadvantages of existing alignment techniques and claims to offer part of the solutions for the challenges that come with the growing demand in TKA surgery.

Most surgeons nowadays use conventional intramedullary or extramedullary align-ment guides (conventional instruments: CI) for aligning a TKA. In this thesis we there-fore compared results of PMPG to results obtained with CI. In chapter 2, we presented the results of the first 40 consecutive patients operated on by means of PMPG for TKA. We conducted a case control study and compared blood loss, operation time, and alignment of 40 TKA’s using PMPG with a matched control group, operated on by CI.

We concluded from this chapter that PMPG TKA shows improved accuracy of align-ment and a small decrease in blood loss (60 mL) and operating time (10 min) compared to CI but the fraction of outliers (more than 3 degrees deviation from neutral mechani-cal axis of the leg) was relatively high (30% versus 50% with CI).

The preliminary experience with PMPG was generally positive in comparison with CI and therefore we aimed at investigating the technique in more detail by setting up a randomized controlled trial comparing these 2 alignment methods. The results of this trial were outlined in chapter 4 and chapter 5. However, as the fraction of outliers in mechanical axis was higher than expected, we set up a separate study to assess the potential weak spot of PMPG as an alignment method. As the basic aim of PMPG is to adequately recreate the pre-operative digital plan, we investigated the potential of the technique to do so in chapter 3.

Chapter 9

126

More specifically, the goal of chapter 3 was to explore whether there was a significant difference between the alignment of the individual femoral and tibial components of a TKA (in the frontal, sagittal and horizontal planes) as calculated pre-operatively (digital plan) and the actually achieved alignment in vivo obtained with the use of PMPG for TKA.

Twenty-six patients were included in this trial. Software permitted matching of the pre-operative MRI-scan (and therefore calculated prosthesis position) to a pre-operative CT-scan and then to a post-operative full-leg CT-scan to determine deviations from pre-op planning in all 3 anatomical planes.

We concluded from this chapter that PMPG were able to adequately reproduce the pre-op plan in all planes, except for the tibial rotation in the transverse plane. We dis-cussed and highlighted the importance for adequate surgeon training before starting using PMPG. As already stated, the results outlined in chapter 2 prompted us to conduct a prospec-tive, double-blind, randomized controlled trial. Chapter 4 describes this trial and ad-dresses the following research questions: firstly, is there a significant difference in outli-ers in alignment in the frontal and sagittal plane between PMPG TKA and CI TKA. Sec-ondly, is there a significant difference in operation time, blood loss and length of hospi-tal stay between the two techniques. One hundred and eighty patients were random-ized for PMPG TKA or CI TKA in two centers.

We concluded from this chapter that the results in terms of obtaining a neutral me-chanical axis and a correct position of the prosthesis components did not differ be-tween PMPG and CI. A small reduction in operation time (5 minutes, p < 0.001) and blood loss (100mL, p < 0.001) can be realised with the PMPG system, but length of hos-pital stay was identical (mean of 3.6 days, n.s.) compared to CI. In chapter 5 we presented the clinical outcome and complication rate of PMPG as me-dium-term results of the randomized controlled trial presented already in chapter 4.

We concluded that there were no significant or clinically relevant differences be-tween CI and PMPG TKA for all questionnaires and that there was no difference in TKA related complication rate.

From chapter 2, 3, 4 and 5 it was clear that PMPG were safe to use and at least not inferior to CI in terms of obtaining a correct alignment or good clinical out-come. However based on the research presented in these chapters, no clear benefit of using PMPG over CI could be demonstrated either. Theoretically, PMPG should be able to adequately predict implant size based on the pre-operative images. Predicting this implant size could potentially reduce costs of TKA surgery and we investigated the sizing capabilities of PMPG in a separate study, presented next.

Summary

127

In this study in chapter 6, the initial series of 200 consecutive cases operated on by a single surgeon and with a minimum follow-up of 2 years was analyzed. Operation time, accuracy of component size prediction and percentages of outliers (>3 degrees) in me-chanical axis were assessed in this cohort.

No cases in this series needed to be converted to traditional instrumentation. Aver-age operation time was 52 minutes. No learning curve on surgical time was seen. Im-planted component size prediction with an error of one size, was correct in 98.5% of femoral components and 97.0% of tibial components. Reviewing and approving the preoperative plan, reduced the number of changes that were necessary during surgery from 21.5% to 14% for the femoral component and from 41% to 29.5% for the tibial component. Average mechanical axis was 179.5° with a percentage of outliers > 3° from digital planning of 20.7%.

We demonstrated in this chapter the importance of a critical review of the digital preoperative plan in order to be able to rely on the component sizing capabilities of PMPG. The question whether PMPG can reduce costs in TKA surgery is one that is diffi-cult to answer and future research will be necessary to address this issue in more detail.

While conducting the research for this thesis, concerns had risen about the ade-quacy of measurements performed on standing long-leg radiographs (LLR). We used these radiographs throughout our research presented in the former chap-ters to assess alignment outcomes. We therefore aimed at assessing the reliabil-ity and validity of measurements on LLR using 3D CT-scan as a gold standard in the next chapter of this thesis.

In chapter 7, six different surgeons measured the mechanical axis and position of the femoral and tibial components individually on 24 LLR. Intraclass correlation coefficients were calculated to obtain reliability and bland-altman plots were constructed to assess agreement between measurements on LLR and measurements on 3D CT-scan.

We concluded from this chapter that measurements on LLR show moderate to good reliability and, when compared to 3D CT-scan, show good validity. Finally, in chapter 8 we placed our findings in a broader context and answered the ques-tions postulated in the introduction of this thesis (chapter 1).

In the second part of this chapter we discussed challenges when introducing PMPG on a larger scale in TKA surgery. Perhaps the most important question to be asked was how to deal with medicolegal issues of PMPG. Medico-legal experts have pointed out that companies are responsible if PMPG are created and delivered without validation of the planning by the surgeon. There is no doubt that the orthopaedic surgeon in the end is responsible for the correct placement of the TKA components.

Second major challenge when using PMPG is the peri-operative logistics issue. The pathway of performing an MRI or CT-scan, uploading the images to the engineer’s serv-er for construction of the digital plan, checking and approval of the digital plan by the

Chapter 9

128

operating surgeon, in time delivery of the guides and digital plan and sterilisation of the guides for peroperative use poses specific demands on the organization.

The third challenge is in the training of orthopaedic residents. Residents will have to learn to master both CI TKA and PMPG TKA and not just PMPG TKA. When the steps and the theoretical knowledge of performing a CI TKA are sufficiently mastered, PMPG can be safely learned.

In the Netherlands, every total knee that has been implanted is registered in a cen-tral database (LROI: landelijke registratie orthopedische implantaten). No information is gathered on the type of alignment method used when performing a TKA. With the in-troduction of more alignment techniques, it might be of interest, however, to collect this information as well. This will make it easier in the long term to draw conclusions on safety and number of needed revision procedures with each alignment technique. We finally concluded by stating that PMPG definitively deserve a place as an alignment tool in TKA surgery. However, at this stage PMPG are more an evolution in TKA surgery than a revolution.

Nederlandse Samenvatting

129

NEDERLANDSE SAMENVATTING

In deze thesis werd het potentieel van patiënt-specifieke mallen (PSM) als een methode voor het uitlijnen van een totale knie prothese (TKP) geëxploreerd. Computer tomogra-fie scans (CT-scans) of magneetscans (MRI-scans) en een specifiek software programma worden gebruikt om virtuele modellen van femur en tibia te construeren. Deze virtuele modellen worden vervolgens gebruikt om anatomische herkenningspunten te identifi-ceren op de knie (door planningsingenieur) die vervolgens gebruikt worden om de idea-le maat en positie van de prothesecomponenten te berekenen. Deze berekeningen worden vastgelegd in een virtueel plan van de operatie. De laatste stap is het creëren van uitlijningsmallen die slechts op één manier op femur en tibia van de patiënt in kwestie passen. Deze mallen bepalen vervolgens tijdens de operatie hoe de zaagsneden in het bot verlopen en bepalen daarmee de uiteindelijke positie van de knieprothese. Bij de meeste systemen van PSM wordt gestreefd naar het construeren van een neutra-le/rechte beenas. In hoofdstuk 1 hebben we beargumenteerd dat het aantal patiënten dat een TKP on-dergaat substantieel toegenomen is in de afgelopen jaren en dat de verwachting is dat deze trend zich voortzet in de komende jaren. We bediscussieerden kort het conventio-neel instrumentarium (CI) en computernavigatie als methoden voor het plaatsen van een TKP en beschreven de meeste nadelen die geassocieerd zijn met beide technieken. PSM tracht een antwoord te bieden op deze nadelen en claimt tevens een deel van de oplossing te zijn voor de uitdagingen die de groeiende vraag naar TKP’s met zich mee-brengt.

De meeste orthopedisch chirurgen gebruiken heden ten dage conventioneel intra-medullaire of extramedullaire uitrichtapparaat (CI) om een TKP te plaatsen. Om die reden vergelijken we in deze thesis de resultaten van PSM met die van CI. In hoofdstuk 2 presenteren we de resultaten van de eerste 40 opeenvolgende patiënten die middels PSM een TKP geplaatst kregen. We voerden een case control studie uit en vergeleken bloedverlies, operatietijd en uitlijning van 40 TKP’s die geplaatst werden met PSM met een gematchte controlegroep die middels CI geopereerd werd.

We concludeerden uit dit hoofdstuk dat PSM een verbeterde accuratesse laat zien voor het uitlijnen van een prothese in termen van het verkrijgen van een neutrale beenas en bovendien zorgde voor een kleine reductie in operatietijd (10min) en bloed-verlies (60ml) in vergelijking met conventionele TKP. Echter was de fractie outliers (meer dan 3 graden deviatie van een neutrale beenas) nog hoger dan verwacht (30% versus 50% met CI).

De initiële gebruikerservaring met PSM was in zijn algemeenheid positief in ver-gelijking met CI en derhalve stelden we het doel om beide technieken in meer

Chapter 9

130

detail met elkaar te vergelijken in een gerandomiseerde trial. De resultaten van deze trial werden gepresenteerd in hoofdstuk 4 en 5. Echter, aangezien de frac-tie outliers vrij hoog was, startten we een extra studie om de potentiele achilles-hiel van PSM te achterhalen. Het basisdoel van PSM is het adequaat reconstrue-ren van het pre-operatieve plan en we analyseerden daarom in hoeverre de techniek dit kan waarmaken. Deze resultaten presenteerden we in hoofdstuk 3.

Meer specifiek was het doel van hoofdstuk 3 om te analyseren of er een significant verschil bestaat tussen de uitlijning van de individuele femur en tibia componenten van de TKP (in het frontale, sagittale en horizontale vlak) zoals die pre-operatief werd bere-kend en de uiteindelijk bereikte uitlijning na operatief ingrijpen.

Zesentwintig patiënten werden geïncludeerd in deze trial. Software werd gebruikt om de per-operatieve MRI-scan (en daarmee ook de berekende positie van het implan-taat) te matchen met een pre-operatieve CT-scan. Deze laatste werd vervolgens weer gematcht met de postoperatieve CT-scan om de afwijkingen van de pre-operatieve planning in alle 3 de vlakken te kunnen objectiveren.

We concludeerden uit dit hoofdstuk dat PSM in staat waren om adequaat het pre-operatieve plan te reproduceren, behalve voor wat betreft de rotatie van de tibiacom-ponent in het horizontale vlak. We bediscussieerden en benadrukten verder het belang van een adequate training van de orthopedisch chirurg alvorens te starten met gebruik van PSM voor TKP. Zoals reeds eerder aangegeven, brachten de resultaten uit hoofdstuk 2 ons ertoe om een prospectief, dubbelblinde, gerandomiseerde trial op te zetten. Hoofdstuk 4 be-schrijft deze trial met de volgende onderzoeksvragen: ten eerste, is er een significant verschil in outliers bij de uitlijning van de TKP in het frontale en sagittale vlak tussen TKP geplaatst m.b.v. PSM en conventionele TKP. Ten tweede, is er een significant verschil in operatietijd, bloedverlies en opnameduur tussen beide technieken. Honderdtachtig patiënten werden gerandomiseerd voor PSM TKP of conventionele TKP in twee centra.

We concludeerden uit dit hoofdstuk dat de resultaten voor het verkrijgen van een neutrale mechanische as en een correcte positie van de prothesecomponenten niet verschilden tussen PSM en CI. Een kleine reductie van de operatietijd (5 minuten, p < 0.001) en bloedverlies (100ml, p < 0.001) werd gevonden met de PSM, maar de opna-meduur in het ziekenhuis was identiek (gemiddeld 3,6 dagen, niet significant) aan die van CI. In hoofdstuk 5 presenteerden we de klinische uitkomsten en het aantal complicaties na middellange follow-up (minimaal 2 jaar) van de patiënten uit de gerandomiseerde trial beschreven in hoofdstuk 4.

Nederlandse Samenvatting

131

We concludeerden dat er geen significant of klinisch relevant verschil zat tussen PSM en CI ten aanzien van de uitkomsten van de vragenlijsten en dat er geen verschil was in TKP gerelateerde complicatie percentage.

Uit hoofdstuk 2, 3, 4 en 5 is het duidelijk dat PSM veilig waren in gebruik en dat de uitkomsten in vergelijking met CI niet inferieur waren voor wat betreft correc-te uitlijning of klinische uitkomst. Echter, de bevindingen uit ons onderzoek toonden ook geen substantiële verbetering voor deze uitkomstmaten. Theore-tisch gezien zou de techniek achter PSM in staat moeten zijn om pre-operatief de juiste implantaatgrootte vast te stellen op basis van de beeldvorming met MRI of CT-scan. Het voorspellen van de implantaatgrootte zou potentieel kunnen leiden tot een reductie in kosten die geassocieerd zijn met TKP chirurgie. We on-derzochten de potentie van PSM om dit te verwezenlijken in een aparte studie (hoofdstuk 6).

In deze studie in hoofdstuk 6, presenteerden we de resultaten van een serie van 200 opeenvolgende TKP cases die geopereerd werden door 1 enkele chirurg en die een minimale follow-up hadden van 2 jaar. Operatietijd, accuratesse van de voorspelling van de implantaatgrootte en percentage outliers (>3 graden) van de mechanische as wer-den geanalyseerd in dit cohort.

Er waren geen cases waarbij peroperatief geconverteerd moest worden naar con-ventioneel instrumentarium. Gemiddelde operatietijd was 52 minuten. Er was geen leercurve m.b.t. operatietijd in dit cohort. De voorspelling van de componentgrootte met een foutmarge van 1 maat was correct in 98,5% van de femurcomponenten en 97% van de tibiacomponenten. Het preoperatief beoordelen en accorderen van het digitale plan reduceerde het aantal maataanpassingen peroperatief van 21.5% naar 14% voor de femurcomponent en van 41% naar 29.5% voor de tibiacomponent. Gemiddelde mechanische as week 0.5 graden af van een neutrale beenas met een percentage out-liers > 3° van de digitale planning van 20.7%.

We toonden in dit hoofdstuk het belang van het kritisch reviewen van het digitale preoperatieve plan om voldoende te kunnen vertrouwen op de voorspelling van de componentgrootte. De vraag of PSM hiermee ook de kosten van TKP chirurgie kan re-duceren blijft moeilijk te beantwoorden en verder onderzoek zal nodig zijn om dit issue in meer detail te analyseren.

Lange beenas opnamen (LBO) worden vaak gebruikt in de orthopedie om de beenas (en percentage outliers van een neutrale beenas) te beoordelen bij pati-enten die een totale knie prothese (TKP) ondergaan. Tijdens het uitvoeren van het onderzoek van deze thesis ontstonden twijfels over de betrouwbaarheid van metingen op LBO. Om die reden hebben we een aparte studie opgezet om de betrouwbaarheid en de validiteit van metingen op LBO, waarbij gebruik gemaakt werd van 3D CT-scans als gouden standaard, te analyseren.

Chapter 9

132

In hoofdstuk 7 maten zes verschillende chirurgen de mechanische as en de positie van de femur en tibiacomponent op 24 LBO. Intraclass correlatiecoëfficiënten (ICC) werden berekend om de betrouwbaarheid vast te stellen en Bland-Altman plots werden gecon-strueerd om de overeenkomst tussen metingen op LBO en 3D CT-scans te analyseren.

We concludeerden uit dit hoofdstuk dat metingen op LBO een gemiddelde tot goede betrouwbaarheid tonen en, wanneer vergeleken met 3D CT-scan, een goede validiteit. Tenslotte plaatsten we in hoofdstuk 8 onze bevindingen in een bredere context en be-antwoordden we de gestelde vragen uit de introductie (hoofdstuk 1).

In het tweede deel van dit hoofdstuk bediscussieerden we de uitdagingen die zou-den ontstaan wanneer PSM op grotere schaal geïntroduceerd zouden worden voor TKP chirurgie. Waarschijnlijk is de belangrijkste vraag hierbij hoe om te gaan met het medi-colegale aspect van het gebruik van PSM. Medicolegale experts hebben aangegeven dat de industrie primair verantwoordelijk is wanneer PSM worden gefabriceerd en geleverd zonder voorafgaande validatie van de planning door de orthopedisch chirurg. Er bestaat echter geen twijfel dat de orthopedisch chirurg zelf uiteindelijk verantwoordelijk is voor de correcte plaatsing van de TKP componenten, welke methode hier ook voor gebruikt wordt.

Tweede grote uitdaging bij het gebruik van PSM draait om het logistiek verhaal. Het traject van uitvoeren van MRI of CT-scan, uploaden van de beelden naar de plannings-ingenieurs voor het construeren van het digitale pre-operatieve plan, checken en ac-corderen van dit plan door de orthopedisch chirurg, tijdig afleveren van de mallen en digitale plan en sterilisatie van de mallen voor peroperatief gebruik stelt specifieke eisen aan de hele organisatie.

Derde uitdaging ligt in het trainen van opleidingsassistenten tot orthopedisch chi-rurg. Idealiter zouden zij zowel CI als PSM moeten beheersen als technieken voor het plaatsen van een TKP. Wanneer de operatieve stappen en theoretische kennis van het plaatsen van een conventioneel geïnstrumenteerde TKP voldoende beheerst worden, kan de techniek van PSM veilig aangeleerd en gebruikt worden.

In Nederland wordt iedere geplaatste knieprothese geregistreerd in een centrale da-tabase (LROI: landelijke registratie orthopedische implantaten). Tal van informatie wordt hierin vastgelegd, maar het type methode dat gebruikt wordt voor het uitlijnen van een prothese wordt niet geregistreerd. Met het introduceren van steeds meer uitlijningsmethoden, zou het interessant zijn om deze informatie wel te verzamelen. Op de lange termijn zou het ons in staat kunnen stellen om conclusies te trekken over de veiligheid van de verschillende systemen en over het aantal revisieprocedures. Concluderend kunnen we stellen dat PSM zeker een plaats verdienen als een uitlijnme-thode voor een TKP. In dit stadium zijn PSM vooral nog te beschouwen als evolutie in TKP chirurgie eerder dan als revolutie.

133

Chapter 11

Dankwoord

Curriculum Vitae

List of presentations and publications

Sponsors

Dankwoord / Curriculum Vitae / List Of Presentations And Publications

135

DANKWOORD

Een proefschrift schrijven doe je allesbehalve alleen en dus is na het voltooien ervan een dankwoord meer dan op zijn plaats. Leden van de beoordelingscommissie: hartelijk dank dat u bereid was mijn onderzoeks-werk te beoordelen. Promotor: Prof. dr. L. van Rhijn, beste Lodewijk, graag wil ik je bedanken voor de tijd die je genomen hebt om kritisch mee te denken over de inhoud van dit proefschrift. Ik heb bewondering voor hoe je als afdelingshoofd de ‘orthopedische club’ in Maastricht leidt en voor de ambitie die je uitstraalt. Ondanks je drukke agenda heb je steeds tijd vrij gemaakt om samen met Bart en mij promotiegesprekken te plannen; hartelijk dank daarvoor! Copromotoren: dr. N. Kort, beste Nanne, een speciaal woord van dank gaat uit naar jou. Jij hebt me, samen met Martijn, in het laatste jaar van mijn geneeskunde opleiding gevraagd of ik met jullie een promotietraject wilde starten. Daarmee heb je me je volle vertrouwen gegeven, waarvoor dank. Zonder jouw bereidwilligheid om onderzoek op te starten en patiënten te includeren was van een promotietraject überhaupt geen sprake geweest.

Dr. P. Emans, beste Pieter, ook jij hartelijk bedankt voor al je moeite en tijd die je in het nalezen van de stukken hebt gestoken. Je kritische blik en opbouwende (!) kritiek heeft de inhoud van de papers op een hoger niveau gebracht. Ook voor jou bewonde-ring dat je al dit soort werk nog in je eigen, drukke schema hebt weten in te passen. Paranimfen: Martijn, Tinus, voor jou ook een speciaal woord van dank voor al je werk dat je in het bijhouden van de patiënten database gestoken hebt. Dankjewel ook voor je hulp bij de statische analyses en dank voor je kritische blik op de geschreven manuscrip-ten, maar bovenal ook jij bedankt voor het vertrouwen dat je me 6 jaar geleden gege-ven hebt, samen met Nanne. Zonder je hulp zou het niet in dit tijdsbestek gelukt zijn. Beste Bart, ook jij verdient een speciaal woord van dank. We zijn ongeveer op het-zelfde moment gestart met het traject en er is niets leuker dan het nu ook samen af te kunnen ronden. Dankjewel voor delen van lief en leed en petje af voor je doorzettings-vermogen en positieve instelling. Van harte proficiat met je mooie eindresultaat! Beste Peter, ook jij bedankt voor jouw bijdrage aan de inhoud van dit boekje. Dankjewel dat je steeds beschikbaar was om met een kritische blik naar de manuscripten te kijken en mee te filosoferen.

Chapter 10

136

Beste Walter, hartelijk dank voor jouw bijdrage aan het includeren van patiënten en verzamelen van resultaten in het St. Anna ziekenhuis in Geldrop. Dank ook aan dr. H. Hoekstra en dr. R. van Drumpt voor het includeren en opereren van patiënten voor onze gezamenlijke RCT. Een speciaal woord van dank moet uitgaan naar de patiënten die hebben deelgenomen aan onderzoek dat geresulteerd heeft in diverse publicaties. Zonder hun bereidwillig-heid om voor diverse follow-up momenten naar Sittard terug te komen, was dit proef-schrift nooit verschenen! Ook een apart woord van dank voor de dames op de polikliniek orthopedie, de financi-ele dames van de maatschap orthopedie, de zorgplanners van de orthopedie in Sittard, OK-assistenten en onze collega’s bij de radiologie-afdeling. Dankjewel voor jullie geduld als weer eens e.e.a. voor het onderzoek geregeld moest worden. Beste collega assistenten, coassistenten en semiartsen, dank ook aan jullie voor het terugzien van patiënten op de polikliniek voor het verzamelen van onderzoeksgegevens. Bedankt voor jullie motiverende woorden voor het continueren van het onderzoek.

Apart woord van dank voor jou, Stijn! We hebben het grootste deel van onze oplei-ding samen in dezelfde ziekenhuizen genoten en vaak over onderzoek gediscussieerd. Je bent meer dan een collega AIOS en stiekem hoop ik dan ook dat we straks als directe collega’s orthopeden aan de slag kunnen. Beste maatschap orthopedie, inmiddels gepensioneerde dr. A. Verburg en drs. J. van Os, dr. N. Kort, drs. P. Tilman, dr. E. van Haaren, dr. E. Jansen, drs. R. Hendrickx; beste Aart, Hans, Nanne, Pieter, Emil, Edwin en Roel: dank voor het goede opleidingsklimaat dat jullie met z’n allen verzorgen en complimenten voor de lopende onderzoekslijnen van-uit jullie maatschap. Mijn andere opleiders in de verschillende opleidingsziekenhuizen: maatschap chirurgie onder opleiderschap van dr. H. Janzing, maatschap orthopedie Viecuri Venlo onder opleiderschap van dr. W. Morrenhof, staf orthopedie MUMC+ onder opleiderschap van dr. H. Staal, maatschap orthopedie Zuyderland medisch centrum Heerlen onder oplei-derschap van prof. dr. I. Heyligers. Dank dat jullie allen op eigen manier investeerden in mijn opleiding. Het belangrijkste deel van dit dankwoord gaat natuurlijk uit naar de mensen die het dichtst bij me staan. Daarom speciaal dank aan mijn ouders, die er voor gezorgd hebben dat ik geneeskunde kon gaan studeren en steeds volledig achter mij en mijn vrouw hebben gestaan tijdens de opleidingsjaren. Het doorzettingsvermogen heb ik van jullie gekregen!

Dankwoord / Curriculum Vitae / List Of Presentations And Publications

137

Hartelijk dank ook aan de rest van mijn familie en schoonfamilie voor jullie interesse in mijn onderzoek! En ‘last but not least’: mijn enorme dankbaarheid en bewondering voor mijn vrouw, Hananja. Zonder jouw steun en opofferingen was dit boekje er nooit gekomen! Jij hebt me zoveel werk uit handen genomen zodat ik me kon concentreren op het onderzoek en de opleiding. Nu hebben we 2 prachtige kinderen samen en is de tijd gekomen om samen daar in te investeren!

139

CURRICULUM VITAE

Bert Boonen was born on the 16th of January 1985 in Neerpelt, Belgium as the first son in a family of 4 chil-dren. After graduating from high school (Salvatorcol-lege, wetenschappen-wiskunde 8uur, Hamont-Achel, België) in 2003, he studied physiotherapy for one year (Provinciale Hogeschool Limburg, Hasselt, België). In 2004 he started his medical studies at Maastricht Uni-versity where he met Hananja Stael. They married in June 2011.

During the internships at the Catharina Hospital, Eindhoven he became interested in orthopaedics, and he started with a retrospective study on the outcome of operative treatment of distal biceps tendon ruptures, together with Dr. Kees Oosterbos.

In the last year of his medical studies the foundation of this PhD thesis was laid at the Orbis Medical Centre (currently Zuyderland Medical Centre), Sittard-Geleen. In 2010 he finished his medical studies and started working as an orthopaedic resident (ANIOS) at the Orbis Medical Centre.

In 2012 he started his career as an orthopaedic resident in training (AIOS) at the Viecuri Hospital, Venlo and continued his education at Maastricht University Medical Centre; Orbis Medical Centre, Sittard-Geleen and Zuyderland Medical Centre, Heerlen. During his training he continued his research under supervision of dr. Nanne Kort, Zuy-derland Medical Centre; dr. Pieter Emans and Prof. dr. Lodewijk van Rhijn, Maastricht University Medical Centre.

During his residency, he became a father of a daughter and a son: Veerle and Jurre. He will finish his orthopaedic training in 2017.

141

LIST OF PRESENTATIONS AND PUBLICATIONS

PRESENTATIONS (RESEARCH FROM THIS THESIS)

May 2011: Preliminary experience with the patient-specific templating TKA: 40 cases compared with a matched control-group. Wetenschappelijk symposium Atrium Medisch Centrum, Heerlen, The Netherlands.

Sept 2011: Preliminary experience with the patient-specific templating TKA: 40 cases compared with a matched control-group. Oral presentation, SICOT, Prague, Czech Republic.

Oct 2013: Intra-operative results and radiological outcome of conventional and patient-specific surgery in total knee arthroplasty: a multicenter, random-ized controlled trial. Oral presentation, CORS, Venice, Italy.

Oct 2013: Accuracy of patient-specific guides for total knee arthroplasty. A prospec-tive shape matching study. Oral presentation, ISTA, Florida, USA.

Mar 2014: Intra-operative results and radiological outcome of conventional and patient-specific surgery in total knee arthroplasty: a multicenter, random-ized controlled trial. Oral presentation, AAOS, New Orleans, USA.

Jun 2014: Intra-operative results and radiological outcome of conventional and patient-specific surgery in total knee arthroplasty: a multicenter, random-ized controlled trial. Oral presentation, EFORT, Londen, UK.

Jun 2014: Accuracy of patient-specific guides for total knee arthroplasty. A prospec-tive shape matching study. Poster presentation, EFORT, Londen, UK.

Sept 2015: Overview of research leading to thesis. Presentatie voor deelnemers travelling fellowship.

Zuyderland, Sittard. Sept 2016: Critical appraisal of the literature concerning PSI. Presentatie voor Sales

verantwoordelijken Biomet. Zuyderland, Sittard.

Chapter 10

142

PUBLICATIONS

Boonen, B, Oosterbos, KJM. Resultaten van operatieve behandeling middels botankers van vijf distale bicepspees rupturen, NTvO, 2010. 17(4): p 151-156. Boonen B, Schotanus MG, Kort NP. Preliminary experience with the patient-specific templating total knee arthroplasty. Acta Orthop. 2012 Aug;83(4):387-93. (This Thesis) Kerens B, Boonen B, Schotanus MG, Kort NP. Patient-specific guide for revision of medial unicondylar knee arthroplasty to total knee arthroplasty. Beneficial first results of a new operating technique performed on 10 patients. Acta Orthop. 2013 Mar;84(2):165-9. Boonen B, Kerens B, Knee examination, Emedicine: http://emedicine.medscape.com/article/1909230-overview. Kerens B, Boonen B, Kort NP. Optimale plaatsing van de totale knieprothese. Orthopedie Actueel. 2013 maart; jaargang 5, nr1. Kerens B, Boonen B, Schotanus MG, Kort NP. Popliteal lesion due to traction during unicompartmental knee revision surgery. J Orthop. 2013 Mar 1;10(1):38-40. Kerens B, Boonen B, Schotanus MG, Lacroix H, Emans PJ, Kort NP. Revision from uni-compartmental to total knee replacement: the clinical outcome depends on reason for revision. Bone Joint J. 2013 Sep;95-B(9):1204-8 Boonen B, Schotanus MG, Kerens B, van der Weegen W, van Drumpt RA, Kort NP. Intra-operative results and radiological outcome of conventional and patient-specific surgery in total knee arthroplasty: a multicentre, randomized controlled trial. Knee Surg Sports Traumatol Arthrosc. 2013 Oct;21(10):2206-12. (This Thesis) Boonen B, Kerens B, Schotanus MGM, Kort NP. Compartment Syndrome, Femoral Nerve Lesion and Recurrent Haemarthrosis in one and the same Patient after Total Knee Ar-throplasty. International Journal of case reports in medicine, Vol. 2013. Kerens B, Schotanus MG, Boonen B, Kort NP. No radiographic difference between pa-tient-specific guiding and conventional Oxford UKA surgery. Knee Surg Sports Traumatol Arthrosc. 2015 May;23(5):1324-9.

List of presentations and publications

143

Boonen B, Schotanus MG, Kerens B, Hulsmans FJ, Tuinebreijer WE, Kort NP. Patient-specific positioning guides for total knee arthroplasty: no significant difference between final component alignment and pre-operative digital plan except for tibial rotation. Knee Surg Sports Traumatol Arthrosc. 2015 Jun 9. (This Thesis) Schotanus MG, Boonen B, Kort NP. Patient specific guides for total knee arthroplasty are ready for primetime. World J Orthop. 2016 Jan 18;7(1):61-8. Boonen B, Kerens B, Schotanus MG, Emans P, Jong B, Kort NP. Inter-observer reliability of measurements performed on digital long-leg standing radiographs and assessment of validity compared to 3D CT-scan. Knee. 2016 Jan;23(1):20-4. (This Thesis) Boonen B, Schrander D, Schotanus M. Hulsmans F. Kort N. Patient-specific guides in total knee arthroplasty: a two year follow-up of the first 200 consecutive cases per-formed by a single surgeon. Journal of clinical rheumatology and musculoskeletal medi-cine. Online: http://content.yudu.com/Library/A3yjzj/JOURNALOFCLINICALRHE/resources/index.htm? referrrUrl=http%3A%2F%2Ffree.yudu.com%2Fitem%2Fdetails%2F3682413%2FJOURNAL-OF-CLINICAL-RHEUMATOLOGY-AND-MUSCULOSKELETAL-MEDICINE. (This Thesis) Kerens B, Schotanus MG, Boonen B, Boog P, Emans PJ, Lacroix H, Kort NP. Cementless versus cemented Oxford unicompartmental knee arthroplasty: early results of a non-designer user group. Knee Surg Sports Traumatol Arthrosc. 2016 May. Boonen B, Schotanus MG, Kerens B, van der Weegen W, Hoekstra HJ, Kort NP. No dif-ference in clinical outcome between patient-matched positioning guides and conven-tional instrumented total knee arthroplasty two years post-operatively: a multicentre, double-blind, randomized controlled trial. Bone Joint J. 2016 Jul;98-B(7):939-44. (This Thesis)

145

SPONSORS

Patient-matched positioning guides in total knee arthroplasty · 2017. 7. 9. · 88.5/1000 in women...

Documents

Transcript of Patient-matched positioning guides in total knee arthroplasty · 2017. 7. 9. · 88.5/1000 in women...