healthmedicinet

Study design

The Osteoarthritis of the Knee Statin (OAKS) study is a multicentre, randomised, double-blind,
placebo-controlled trial over 2 years. The trial was registered at the Australian
New Zealand Clinical Trials Registry prior to recruitment, and trial reporting will
be guided by the CONSORT Statement 31]. A total of 350 patients with symptomatic knee OA will be recruited in equal numbers
via the OA Clinical Trial Network in Melbourne, Hobart and Adelaide, using a combined
strategy including collaboration with general practitioners, rheumatologists, and
orthopaedic surgeons, as well as advertising through local media. Ethics approval
has been obtained from The Alfred Hospital Ethics Committee (521/12), Monash University
Human Research Ethics Committee (CF13/595 – 2013000236), Tasmania Health and Medical
Human Research Ethics Committee (H0012971), and Human Research Ethics Committee (TQEH/LMH/MH)
(HREC/13/TQEHLMH/20). Written informed consent will be obtained from all participants.

Inclusion criteria

The inclusion criteria will be: males and females with symptomatic knee OA for at
least 6 months with a pain score of at least 20 mm on a 100-mm visual analogue scale
(VAS); age 40–70 years old; and meeting the American College of Rheumatology (ACR)
criteria for symptomatic knee OA 32], assessed by a rheumatologist.

Exclusion criteria

Participants with any of the following conditions will be excluded: inability to give
informed consent; severe radiographic knee OA (grade 3 according to Altman’s atlas
33]) or severe knee pain (on standing 80 mm on a 100-mm VAS); rheumatoid arthritis or
other inflammatory arthritis; significant knee injury; familial hypercholesterolaemia
for which statins are indicated, known atherosclerotic cardiovascular disease, diabetes
mellitus, taking current lipid lowering therapy, or with previous adverse reaction
to statins; absolute cardiovascular risk estimated using the Framingham Risk Equation
of more than 15 % within the next 5 years (National Heart Foundation of Australia,
2005); fasting total cholesterol level 7.5 mmol/L; clinically significant renal disease
or abnormal liver function assessed by aspartate aminotransferase and alanine aminotransferase,
creatine kinase more than twice the upper limit of laboratory normal range; patients
undergoing arthroscopy or open surgery in the index knee in the last 12 months; receiving
intra-articular therapy in the index knee in the last 12 months; concomitant use of
potent analgesics including opiates; co-morbidity that may limit participation (such
as planned index knee joint replacement or medical conditions, for example, malignancy
in the past 5 years other than non-melanoma skin cancer); relocation; any contraindication
to MRI scanning (for example, implanted pacemaker, metal sutures, presence of shrapnel
or iron filings in the eye, or claustrophobia); pregnancy, breast feeding, or women
trying to become pregnant.

Randomisation and blinding

Allocation of participants in a 1:1 ratio to either the intervention or control group
will be based on computer-generated random numbers prepared by a statistician with
no involvement in the trial. Block randomisation will be performed using a central
automated allocation procedure, stratified according to the study site. The randomised
controlled trial will be a double-blind one, with both participants and investigators
assessing outcomes blinded to treatment allocation. Allocation concealment and double
blinding will be ensured by 1) the medications being dispensed by the hospital clinical
trial pharmacy in each site; 2) the use of an identical placebo tablet; 3) objective
measures of knee structural changes being made by trained observers blinded to group
allocation and the time sequence of MRI scans; 4) subjective measures being taken
by research assistants blinded to group allocation.

Intervention

All participants will be provided usual care by their treating health practitioners.
Participants in the intervention arm will receive 40 mg of atorvastatin (Sandoz) once
daily, and those in the control arm will receive 40 mg of inactive matching placebo
(Pharmaceutical Packaging Professionals Pty Ltd, Thebarton, South Australia) once
daily. Atorvastatin was chosen as it is the most widely prescribed statin and has
potent LDL cholesterol lowering effects.

Study procedure

At screening, participants will complete questionnaires, have a knee X-ray, and undergo
biochemical testing including liver function tests, creatine kinase and renal function
tests, to ensure inclusion criteria are met. The study knee will be defined as the
one with symptomatic OA; if both are symptomatic, the one with least severe radiographic
OA (joint space narrowing) will be studied, to reduce loss to follow-up for joint
replacement. After screening, study visits will be scheduled for 0 (baseline), 6,
12 and 24 months (Table 2). The same researchers, who are blinded to treatment allocation, will measure all
clinical variables, administer questionnaires, monitor compliance, and record adverse
events at these visits. Compliance by trial medication will be assessed by pill count.
Biochemical testing will be performed at 4 weeks, and at 6, 12 and 24 months. At baseline
and 24 months the index knee will be imaged on the same whole-body MRI unit.

Table 2. Timetable and measures to be made

Quality assurance

To ensure high-quality execution of the trial in accordance with the protocol, all
trial staff will be trained by the chief investigators and provided a standard protocol
book (with details of standard operating procedures used for all trial contacts, visits,
measurements, and monitoring) and case report forms.

Outcome measures

Knee MRI

Knee MRI acquisition at the three study sites is presented in Table 3, including details of sequences and parameters being used.

Table 3. Magnetic resonance imaging sequences and parameters at three study sites

Primary outcome measure: annual percentage change in tibial cartilage volume

From sagittal T1-weighted images, medial and lateral tibial cartilage volume will
be isolated by manually drawing disarticulation contours around the cartilage boundaries
on a section-by-section basis. Measurement error and bias will be reduced by ensuring
that one reader measures participants’ paired set of images blinded to time sequence,
with a second reader performing independent consistency checks. In our previous study,
we demonstrated a coefficient of variation of 2.0–3.4 % for this method 34]. We and others have shown that change in tibial cartilage volume over 2 years can
be measured using this method and is both statistically and clinically significant
in patients with symptomatic knee OA 34]–36]. The annual percentage change in tibial cartilage volume will be calculated as (follow-up
cartilage volume – baseline cartilage volume)/baseline cartilage volume/time between
MRI scans*100.

Secondary outcome measures

Progression of cartilage defects

Cartilage defects will be graded at tibial and femoral sites (0–4) from sagittal images
37]. If multiple cartilage defects are present at one site, the highest grade will be
recorded. Our intra- and inter-observer reliability for this method is 0.89–0.94 and
0.85–0.93, respectively 37]. The progression of cartilage defects will be defined as any increase in cartilage
defect score in either tibial or femoral cartilage over 2 years.

Progression of BMLs

BMLs will be defined as discrete subchondral areas with increased signal intensity
and ill-defined margins on sagittal proton density images, and graded at tibial and
femoral sites (0–3). If multiple lesions are present at one site, the highest grade
will be recorded. The intra- and inter-reader correlation coefficient for this method
has been reported to be 0.88–0.93 13]. The progression of BMLs will be defined as any increase in their grade in either
tibial or femoral site over 2 years. The maximum areal size will also be measured
for each lesion. The lesion with the largest size will be used if more than one lesion
is present at the same site. Our intra-observer correlation coefficient for this method
is 0.97 38]. The change in areal size of BMLs over 2 years will be assessed since it is more
sensitive to change, with positive results seen over 6 months in a previous clinical
trial 39].

Change in knee symptoms and function

Knee pain, stiffness and function will be assessed using a self-administered validated
questionnaire, WOMAC (the Western Ontario and McMaster Universities Osteoarthritis
Index) 40], at each study visit. It will be used to describe the population, and change in symptoms
and function measured by change in WOMAC score will be a secondary outcome.

Other measurements

Anthropometry

Height (stadiometer), weight (electric scale) and body mass index (height/weight
²
) will be measured at baseline and 2 years.

Radiographic knee OA

A weight-bearing anteroposterior radiograph of the study knee will be scored for osteophytes
and joint space narrowing on a four-point scale (0–3) using the Altman atlas 33]. Our intra- and inter-observer reliability is 0.93 and 0.86 for osteophytes and 0.93
and 0.85 for joint space narrowing, respectively 41].

Knee angle

The knee angle will be measured from weight-bearing radiographs, as it affects cartilage
volume change 42]. Our intra-observer correlation coefficient is 0.98 42].

Tibial plateau bone area

The cross-sectional area of tibial plateau will be measured manually from the reformatted
axial MR images. The coefficient of variation of this method is 2.3–2.4 % 34]. Cartilage volume will be adjusted for bone area to account for joint size.

Meniscal tear and extrusion

The presence of meniscal tears and extent of meniscal extrusion will be determined
from sagittal images and confirmed in coronal and axial images 43], as they affect cartilage volume change 43]. The intra- and inter-reader correlation coefficient has been reported to be 0.85–0.96
43].

Physical activity

Physical activity will be measured using the International Physical Activity Questionnaire
(IPAQ) 44] short version at baseline and 2 years.

Lower limb muscle strength

Lower limb muscle strength is expected to increase with a decrease in pain, and therefore
will be measured by dynamometry at months 0, 12 and 24. The muscles measured in this
technique are mainly quadriceps and hip flexors. The devices will be calibrated by
suspending known weights at regular intervals. Repeatability estimates (Cronbach’s)
were 0.91 45].

Concomitant medication use

Concomitant medication use including glucosamine, chondroitin, corticosteroids and
non-steroidal anti-inflammatory drugs will be allowed during the trial but will be
documented. There is no definitive data showing that these treatments affect cartilage
loss. The randomisation process is the most effective method for ensuring that two
groups are as similar as possible with respect to known confounders and unknown potential
confounders including treatments. We will adjust for medication use in the analyses.

Cigarette smoking history

Cigarette smoking history will be assessed at baseline and 2 years. As it affects
cartilage loss and BMLs 46], we will adjust for smoking status in the analyses.

Knee injury

Knee injury will be assessed at baseline and 2 years. Prior knee injury is an exclusion
criterion. Any significant injury during the study will be documented, as it is a
risk factor for knee OA.

Safety assessment

Adverse events will be monitored throughout the study. Standard safety and efficacy
monitoring will be performed through regular face-to-face visits and phone calls between
visits. The participants are requested to report any adverse events to the research
staff spontaneously. Biochemical testing (liver function tests, creatine kinase and
renal function tests) will be performed at 4 weeks, and at 6, 12 and 24 months. Details
of the adverse event and its relationship with study intervention will be recorded
and reported to the Ethics Committees.

Sample size calculation

Primary outcome

Based on our previous published data, the rate of tibial cartilage volume loss in
the control group was 3.0 +/? 3.0 per annum 47]. A 33 % reduction (that is, 1 %) in this rate is clinically significant since this
indicates a significant reduction in structural progression of cartilage damage. Our
previous work has shown that a 1 % reduction in the annual rate of cartilage volume
loss over 2 years decreased the need for knee replacement surgery by 20 % over a 3-year
period based on the strong association between the rate of cartilage volume loss and
subsequent knee replacement 48]. With 140 participants in each arm, we will have 80 % power to detect a 30 % reduction
in the rate of cartilage volume loss in the intervention group (2.0 %) compared with
the control group (3.0 %), with alpha error 0.05, two-sided significance. Given our
previous experience in such studies, we expect a maximum dropout rate of 20 % over
2 years. Therefore, a total of 350 participants (175 in each arm of the study) will
be recruited.

Secondary outcome

Based on our reported progression rate of 68 % for cartilage defects 49] and 21 % 14] for BMLs over 2 years, with 140 participants per arm we will have 80 % power to detect
a 17 % difference in the progression of cartilage defects and a 12 % difference in
the progression of BMLs between the intervention group and the control group. With
140 in each arm, we will be able to detect a 30 % reduction in the WOMAC pain subscale
in the intervention group.

Statistical analysis

The primary analyses will be intention-to-treat analyses of primary and secondary
outcomes. Per protocol analyses (according to protocol adherence for patients who
have taken all the study medication and have had all the outcome measures) will be
performed as the secondary analyses. Intention-to-treat analysis will be carried out
by using multiple imputation 50], provided that patients have had a baseline MRI. This analysis only requires missing
data to be missing at random to be valid. Differences within a treatment group between
follow-up and baseline measures will be assessed using a paired samples t-test. The clinical efficacy measures, that is, changes in pain and function, will
also be analysed using the normalised area under the curve for difference in scores
from baseline to month 24. Differences between treatment groups will be assessed using
independent samples t-tests, ANCOVA (for continuous variables) or chi-squared tests (for dichotomous variables).
Multiple linear regression for continuous endpoints or logistic regression for binary
endpoints could be carried out as supplementary analyses for additional adjustment
for imbalanced baseline factors. Analyses of treatment efficacy will be done by censoring
participants at the time of any protocol deviation and developing a model for the
probability of deviation, followed by weighted analyses using only the uncensored
participants where the weights are the inverse probability of censoring. This produces
estimates of treatment effect as if there were full compliance with the protocol in
this trial and is far preferable to per protocol analyses based on (unweighted) observed
compliance 51]. Pre-specified analyses to identify subgroups which may respond better to treatment
will be examined using stratified analyses; variables include radiographic OA and
co-pathology present on MRI.

Data integrity and management

All collected data are recorded using case report forms which will be processed centrally
at the Clinical Informatics and Data Management Unit, Department of Epidemiology and
Preventive Medicine, Monash University. The hard copies of the case report forms will
be stored in a locked area at each study site with secured and restricted access.
The electronic data will be stored in a password protected database with secured and
restricted access. All data collected will be kept strictly confidential. Data transfer
will be encrypted with all data de-identified.

Withdrawal

If participants withdraw from the study before 2 years of follow-up, the reason and
date will be recorded. If the participant withdraws after a minimum of 6 months of
treatment, he/she will be requested to have a second knee MRI scan and complete the
questionnaires.

Monitoring

The principal investigators will monitor the conduct and progress of the project at
each site. The trial coordinator will visit each study site to make sure that all
trial procedures are compliant with the trial protocol. The principal investigators
and the research team will have regular teleconferences to ensure efficient study
execution and ongoing monitoring of the study progress, with summary documents circulated
after each meeting. An independent Data and Safety Monitoring Board was convened,
consisting of a clinical rheumatologist, a clinical cardiologist experienced in statin
use, and a biostatistician, all with clinical trial experience. They will monitor
adverse events. They will meet annually or more often if severe adverse events occur,
and provide a written report to the principal investigators.