Comparison of an Old Versus a New Total Knee Replacement

Multiple factors may affect the outcome of total knee arthroplasty (1-5), and the patient satisfaction has been reported up to 90% (6). Also the physical behavior of a person may affect how much strain that is put on the prosthesis after surgery. Whether more strain is caused by a more physically active profile or by a more sedentary profile is uncertain. Obtaining real-life information about physical behavior is therefore important, and is made possible by using activity monitors measuring physical behavior continuously over longer time-periods.

The National Arthroplasty Registers worldwide has extensive data on knee and hip implants used over several years. Some are documented over 20 years, whereas newer implants appear with shorter follow-up. The Norwegian Arthroplasty Register has a registration completeness of 96% in the 2015 annual report (7).

NexGen is a total knee arthroplasty implant that has been used at great numbers in numerous countries and is well documented in several registries (Swedish Knee Arthroplasty Register, Australian Orthopaedic Association National Joint Replacement Registry).

Persona is a new implant from the same company that produces NexGen, with several changes and modernizations regarding instrumentation and surgical technique. There are a wider range of sizes available, both on the tibia and femur, as well as the menisceal insert. This should allow optimized component fit and less compromise on soft tissue balancing. The tibia is anatomical, with a more medialized keel stem, which should allow for a more accurate fit securing less risk of malrotation and better fixation.

The tibial component in the new Persona design, is anatomical, to achieve a better fit than in the old design. Additionally it facilitates proper rotation of the tibia. When a larger surface of bone is covered, it is expected a lesser risk of subsidence and early loosening of the implant. The new Persona also has a Vitamin E Highly Crosslinked polyethylene insert option for lower wear and improved strength than the regular polyethylene. However, the advent of newer highly crosslinked polyethylene has shown improved wear rates in hip arthroplasty, but not to the same degree in knee arthroplasty (8, 9). The investigators therefore will use conventional polyethylene in this study.

In theory, the new design will lessen the compromises experienced today with the older designs. There have been no independent studies published yet to prove this possible improvement.

Most countries with well-functioning joint registers are cautious in implementing new designs without studies comparing the new implant to older established implants. There have been several examples of unsuccessful joint replacements introduced to the market without thorough clinical investigation and trials, the hip resurfacing surgery being an example of this (10, 11). In Norway, the regional tender process usually demands either 10 years follow-up in clinical or registry studies or alternatively 2 or 5 years RSA studies.

Examples of previous randomized controlled clinical and RSA study where members of our study group have participated and been project leader (12-16).

AIMS OF THE STUDY

1. To compare the clinical outcome scores in the 4 study groups o KSS (Knee Society Score)

- KOOS (Knee Injury and Osteoarthritis Outcome Score)

- EQ 5D

- FJS12 (Forgotten Joint Score 12)

- Walking speed (4-m static walking speed test)

- Objective monitoring of free-living physical behaviour (ActivPAL activity monitor)

- Anchor questions

2. To assess the migration patterns of the components with main focus on the tibia using RSA. Our hypothesis is that early stability of the new Persona is improved with better coverage of the bone plateau. Early stability corresponds to clinical performance at medium term (17, 18).

3. To evaluate revision rates using the Norwegian Arthroplasty Register

DOCUMENTATION OF THE COMPONENTS

The NexGen CR implant has been well documented, as mentioned, in the last decades as excellent total knee implants (7, 19, 20).

The Persona knee has just recently been introduced to the market as a more personalized implant. The improvements with comprehensive sizing and side-specific kinematics have not yet been investigated independently in studies to our knowledge.

160 patients will be randomly allocated into 2 groups with 80 in each group, and 30 of the 80 in each group will be randomized for RSA.

CLINICAL EVALUATION

Clinical evaluation will be performed by independent physiotherapists preoperatively, using EQ5D, Knee Society Score (KSS) and Knee Injury and Osteoarthritis Outcome Score (KOOS). The scoring will be performed by a physiotherapist who is blinded with respect to the assigned study group, and by the patients themselves. An orthopaedic surgeon is consulted in the case of specific clinical problems.

The patients will be examined again at 1, 2, 5 and 10 years with EQ5D, KSS, KOOS, Forgotten Joint Score 12 (FJS12) and Anchor questions.

Preferred walking speed will be measured by timing a 4-meter walk with a stopwatch. Instructions will be given to "walk in your preferred speed" Preferred walking speed has been found valid, reliable, sensitive and specific, and correlates with functional ability and balance confidence (21). Walking speed has also been found to be a consistent predictor of adverse outcomes in different populations (22). Walking speed will be measured at pre-op, 3 months post-op and 1, 2, 5 and 10 years post-op.

The ActivPAL3TM (AP) (PAL technologies Ltd., Glasgow, Scotland) will be used to record free-living physical behavior. The AP is a small, light-weight, three axis accelerometer that will be attached to the anterior aspect of the participants thigh with a hydrogel adhesive (PALStickiesTM, PAL technologies Ltd., Glasgow, Scotland) and covered by waterproof band aid to allow for showering. The AP will be worn continuously 24 h/day to provide seven uninterrupted free-living days of physical behavior data. The participant will be asked to go about their usual activities. A proprietary software-package (PAL Professional) will be used to process acceleration data, and classify the free-living physical activity in sedentary (sitting and lying) and upright (standing and walking) postures, number of transitions from sitting to standing, and number of steps during walking. It has been found that the AP has good accuracy for detecting postures and counting steps during walking in non-impaired people (23-25) and community-dwelling non-impaired elderly (26). It has been found that the AP also was a highly valid measure of postures and transitions in elderly with walking impairments including patients post-stroke. Step counting however, was less accurate for this sample underestimating steps at very slow walking speeds <0.47 m/s (27). The AP monitor will be given to the participant at pre-op, 3 months post-op and 1, 2, 5 and 10 years post-op. After one week recording the participants will be asked to send the monitor back to the research team in a pre-paid envelope.

RADIOGRAPHY

• Standard x-rays

- Preoperatively and at all followup consultations, x-rays will be taken as Anteroposterior (AP) and Mediolateral (ML) views. In addition long axis view (HKA) and patella 45° skyview will be evaluated preoperatively and at 1 year followup. Later these values will be evaluated together with migration- and wear values.

• Radiostereometric analysis (RSA)

- RSA is a highly accurate and precise method of evaluating implant migration and polyethylene wear (17, 28-31). The precision usually is about 0.1mm (translational migration) and 0.1 degrees (rotation). The radiation dose is low compared to plain x-rays (10-20%). During surgery, tiny spheres of Tantalum are implanted in periprosthetic bone. On day 5-10 postoperatively, two simultaneous exposures of the knee and a co-ordinate system are taken. These films allow three-dimensional definition of the implant and bone relative to the co-ordinate system. Implant migration and rotation are measured on repeat examinations over time. The high precision allows us to detect small differences between groups, and also means that the number of subjects in each group can be restricted. The precision (repeatability) of the measurements in this study will be evaluated by double examinations at the 1-year RSA-visit. Limits for significant differences are calculated as the 99% confidence intervals of the absolute mean values of the double examinations. RSA will be done at day 5-10 post-operatively, at 3months, and at 1,2 and 5 years postoperatively.

- The tibia will have 9 tantulum beads of 1.0 mm spread from the cut tibia proximal to distal and anteroposterioly.

- The insert will have 6 beads; 3 x 1.0mm beads placed anteromedially, posterolaterally and anteriorly. 3 x 0.8mm beads placed posteomedially, anterolaterally and anteriorly.

- The holes are then covered with bonewax.

160 patients are included after oral and written information. Included patients give their written consent to participation in the study. 80 patients are included in each group by drawing envelopes assigning the patients randomly to one of the 2 groups. Block randomisation is arranged so that each participating surgeon operates an equal number of TKA in each group. The randomization is done by the statistician using a computer-generated randomization. The randomization protocol is kept at the registry office.

The study is singled-blinded; the patients will not receive information on implant specifics until the end of the study. Each patient receives a patient number (1 through 160) on inclusion. 30 patients in each of the 2 groups will be included for the RSA study.

FOLLOWUP

- Preop: KSS, KOOS, walking speed, ActivPAL activity monitor, standard x-rays,

- Postop: RSA, standard x-rays

- 6 weeks pain and function assessment

- 3 months: RSA, walking speed, ActivPAL activity monitor,

- 1, 2 and 5 years: RSA, standard x-rays, KSS, KOOS and FJS12, walking speed, ActivPAL activity monitor, 10 years: RSA, standard x-rays, KSS, KOOS, walking speed, ActivPAL activity monitor, All RSA examinations will be anonymised and done at the biomaterial laboratory at the Department of Orthopaedic Surgery, Haukeland University Hospital. The results of the RSA-examinations and clinical outcome measures will be collected in electronic files under the respective patient numbers. Data obtained on previous occasions will not be accessible when examining the patient.

Patient list A list will be made coupling patient name and personal identification number to a number on the patient list (1 through 160). The project leader is responsible for the safe handling of the list. Other forms with patient information are labelled with the patient number only for identification. The sponsors of the study will have access to the anonymised data. Application to Ethical committee (REK-Vest) has been accepted. The registry study is covered by the concession of the Norwegian Arthroplasty Registers concession from 2015.

Analysis of variance on repeated measurements (ANOVA) will be used on the RSA- and on the clinical data if the data are normally distributed. In our earlier RSA studies, the data have displayed normal distribution. If however, the data are not normally distributed, the investigators will use non-parametric tests (Mann-Whitney U-test, Wilcoxon matched-pairs test). A power analysis for the student t-test for independent samples showed that group sizes of 24 would give a power of 80 % to detect a 0.1 mm difference in subsidence of the tibial component with a two-sided significance level of 0.0125 and an assumed SD of 0.1 mm. The investigators anticipate exclusion of some patients and, therefore, include 30 patients in each group.

A difference of ten units in an aggregated and averaged subscale of KOOS was chosen as a minimum important change, as suggested by the developers of KOOS. With a standard deviation (SD) of 20, a sample size of 64 patients per group was necessary to achieve a power of 80% and a 5% level of significance. With possible exclusion of patients, the investigators will include 80 in each group.

Power=80% gives N=80 in each group. Type I error rate, alpha=5%. Kaplan-Meier survival percentages will be reported at 5 and 10 years. Cox regression model will be used to calculate risk ratios (RR).

Separate randomization lists will be created for each of the surgeons. Block randomization for each participating surgeon with randomly varying block sizes will be generated to achieve equal numbers in the treatment groups at all times. A central randomization office will perform the computer-generated allocation to the trial groups.

Adverse effects, side effects, and unfortunate effects of the surgery will immediately be reported to the project leader and to the sponsors. Abortion or abruption of the study will be considered in the case of unexpected complications.

The study will end 10 years after inclusion of all patients.

An approximate time schedule is presented below:

- Involving sponsors and funding for costs

- Application to the regional ethical committee

- Involving radiographs and physioterapists

- Learning the new procedure and instruments, Persona

- Starting patient inclusion

- RSA follow-up 2 years, 5 and 10 years

- Registry results 5 and 10 years

PUBLICATION PLAN

1. 2 YEAR FOLLOWUP; RSA study

• A randomized controlled clinical and RSA study of 60 total knee arthroplasty. A comparison of tibia subsidence and micromotion in two CR designs.

2. 2 and 5 year followup

• A randomized trial of 160 total knee arthroplasty using radiostereometry and registry data. A comparison of functional outcome, risk of revision and migration of components in 2 groups of TKA.

3. 10 year follow up

- As above