Open access Protocol

Efficacy and cost-­effectiveness of Stem

BMJ Open: first published as 10.1136/bmjopen-2021-056382 on 29 November 2021. Downloaded from http://bmjopen.bmj.com/ on May 6, 2024 by guest. Protected by copyright.
Cell injections for symptomatic relief
and strUctural improvement in people
with Tibiofemoral knee OsteoaRthritis:
protocol for a randomised placebo-­
controlled trial (the SCUlpTOR trial)
Xiaoqian Liu ‍ ‍,1,2 Sarah Robbins,1,2 Xia Wang ‍ ‍,1,2 Sonika Virk,1,2
Karen Schuck,1,2 Leticia A Deveza,1,2 Win Min Oo,1,2 Kirsty Carmichael,3
Benny Antony,3 Felix Eckstein,4,5 Wolfgang Wirth,4,5 Christopher Little,6
James Linklater,7 Anthony Harris ‍ ‍,8 David Humphries,9 R O'Connell,10
Gillian Heller,10 Thomas Buttel,11 Stefan Lohmander ‍ ‍,12 Changhai Ding,3,13
David J Hunter ‍ ‍1,2,13

To cite: Liu X, Robbins S, ABSTRACT

Wang X, et al. Efficacy and Strengths and limitations of this study
Introduction Knee osteoarthritis (KOA) is a highly
cost-­effectiveness of Stem Cell prevalent disabling joint disease. Intra-­articular stem
injections for symptomatic relief ►► The Stem Cell injections for symptomatic relief and
cell therapy is increasingly being used for treating KOA strUctural improvement in people with Tibiofemoral
and strUctural improvement
in people with Tibiofemoral with little high-­quality evidence to support its use. The knee OsteoaRthritis trial is an investigator-­initiated,
knee OsteoaRthritis: aim of this study is to investigate the efficacy, safety and multi-­centre, 2-­year randomised clinical trial with a
protocol for a randomised cost-­effectiveness of allogeneic mesenchymal stem cells large sample size and robust methodology assess-
placebo-­controlled trial (the (Cymerus MSCs) for treating symptomatic tibiofemoral ing both clinical and radiographic endpoints, which
SCUlpTOR trial). BMJ Open KOA and improving knee structure over 24 months. will provide high-­quality evidence in examining the
2021;11:e056382. doi:10.1136/ Methods and analysis The Stem Cell injections for
bmjopen-2021-056382
efficacy of allogeneic mesenchymal stem cell (MSC)
symptomatic relief and strUctural improvement in people injections for knee osteoarthritis (KOA).
►► Prepublication history and with Tibiofemoral knee OsteoaRthritis study is a phase ►► The trial will use online recruitment and data cap-
additional supplemental material III, multi-­centre, parallel, superiority, randomised, double-­ ture to minimise face-­to-­face clinic visits, which will
for this paper are available blind, placebo-­controlled trial, which will be conducted increase recruitment efficiency; additionally, people
online. To view these files, in Sydney and Hobart, Australia. 440 participants (220 with KOA have been involved in the study design,
please visit the journal online
per arm) aged over 40 years with painful KOA and mild which will help identify logistical issues to facilitate
(http://​dx.​doi.​org/​10.​1136/​
to moderate structural change on X-­ray (Kellgren and adherence to the study.
bmjopen-​2021-​056382).
Lawrence grade 2 or 3) with medial minimum joint space ►► The trial will only include participants with tibiofem-
Received 15 August 2021 width between 1 and 4 mm in the study knee will be oral KOA, which may limit generalisability to those
Accepted 08 November 2021 recruited from the community and randomly allocated to with KOA in other knee compartments in clinical
receive either intra-­articular MSCs or saline at baseline, practice.
week 3 and week 52. The coprimary outcomes will ►► The trial will only include treatment with a single
be the proportion of participants achieving patient-­ specific allogeneic MSC source/type, dose and
acceptable symptom state for knee pain at 24 months treatment schedule, which may limit generalisability
and quantitative central medial femorotibial compartment to other cell types or dosing regimens.
cartilage thickness change from baseline to 24 months. ►► In order to facilitate comparison with other injec-
© Author(s) (or their
employer(s)) 2021. Re-­use Main secondary outcomes include change in knee pain, tion trials, normal saline was chosen as the placebo
permitted under CC BY-­NC. No Patient Global Assessment, physical function, quality treatment; we acknowledge that the carrier media
commercial re-­use. See rights of life and other structural changes. Additional data for of the active and placebo group is different with
and permissions. Published by cost-­effectiveness analysis will also be recorded. Adverse cryoprotectant and electrolytes for the stem cell.
BMJ.
events will be monitored throughout the study. The primary
For numbered affiliations see analysis will be conducted using modified intention-­to-
end of article.
treat. University of Tasmania (UTAS) HREC #: H0021868. All
Correspondence to Ethics and dissemination This protocol has been participants will be required to provide informed consent.
Professor David J Hunter; approved by The University of Sydney (USYD) Human Dissemination will occur through conferences, social
​david.​hunter@​sydney.​edu.a​ u Research Ethics Committee (HREC) #: 2020/119 and The media, and scientific publications.

Liu X, et al. BMJ Open 2021;11:e056382. doi:10.1136/bmjopen-2021-056382 1

 Open access

BMJ Open: first published as 10.1136/bmjopen-2021-056382 on 29 November 2021. Downloaded from http://bmjopen.bmj.com/ on May 6, 2024 by guest. Protected by copyright.
Trial registration numbers Australian New Zealand Clinical Trials additional procedure for cell harvesting and the potential
Registry (ACTRN12620000870954); U1111-­1234-­4897. presence of comorbidities in the preparation of autolo-
gous stem cells, the use of potential off-­the-­shelf commer-
cial preparations of allogeneic MSCs from healthy donors
INTRODUCTION may reduce the overall cost of cell therapies, while main-
Background and rationale taining an accurate quality control.22
Osteoarthritis (OA) is a highly prevalent and progressive Cymerus MSCs, produced from induced pluripotent
joint disease associated with inflammation and major stem cells (iPSCs) through the mesenchymoangioblast
structural changes of the affected joint.1 2 Knee osteo- pathway, have surface markers and trilineage differen-
arthritis (KOA) accounts for 85% of the burden of OA tiation, which meet International Society for Cellular
worldwide and affects 19% of people aged over 45% and Therapy criteria for defining multipotent MSCs,23 24 and
37% of people aged over 60 years of age.3–5 KOA causes have been shown to be safe and well-­tolerated in humans.25
substantial pain and physical dysfunction, ultimately The iPSCs were derived from CD34-­enriched peripheral
impairing quality-­of-­life and is ranked as the eleventh blood mononuclear cells using an episomal plasmid-­
highest contributor to global disability.6 The average total based, transgene-­ free, viral-­
free and feeder layer-­ free
expense per KOA patient per annum is over US$15 000, reprogramming procedure. Although in vivo study of
totalling over US$34 billion in healthcare expenditure.7 intra-­myocardial administration of Cymerus MSCs showed
The healthcare costs of KOA are projected to double by a pro-­angiogenic secretory profile with upregulation of
2040 given population ageing and escalation in obesity, pro-­angiogenic factors and downregulation of metallo-
driving an increase in KOA.8 proteinases,26 this is the first time that Cymerus MSCs is
As a disease of the whole joint, KOA requires long-­ being used intra-­articularly for OA. Therefore, the aim
term management with various treatment options over of this study is to investigate the efficacy, safety and cost-­
the course of the disease. However, the current treat- effectiveness of Cymerus MSCs for treating symptomatic
ment modalities (eg, exercise therapy, paracetamol, non-­ tibiofemoral KOA and improving knee joint structure in
steroidal anti-­inflammatory drugs, intra-­articular steroids) comparison to placebo over 24 months.
do not target structural pathology, and are either only
modestly effective in alleviating symptoms, have signif- Objectives
icant potential side effects or both.9 10 Therefore, there The primary objective is to determine the efficacy of
is a need for a sustainable long-­term effective treatment intra-­
articular allogeneic MSCs on the proportion of
focusing on mitigating the epidemic of OA, modifying its participants achieving patient-­acceptable symptom state
structural progression and symptomatic consequences.11 for knee pain and the reduction in loss of central medial
Most recently, stem cells have emerged as an intra-­ femorotibial cartilage thickness compared with placebo
articular option for KOA, although the quality of the (saline) over 24 months in people with symptomatic
current evidence to support its use is low due to poor trial tibiofemoral KOA. The secondary objectives are to assess
design (eg, inadequate blinding, biases, small studies).12 the efficacy on other clinical and structural outcomes as
The current stem-­ cell market for KOA is rapidly well as the safety and cost-­effectiveness of MSCs therapy.
expanding with 80% of patients claiming ‘symptomatic
improvement’.13 Given the considerable cost of treatment
with an average US$5000 per injection13 and very limited METHODS AND ANALYSIS
scientific evidence of efficacy and safety, high-­ quality Study design
evidence from randomised clinical trials is necessary to The Stem Cell injections for symptomatic relief and
define the role of stem cell therapies in the treatment of strUctural improvement in people with Tibiofemoral
KOA. knee OsteoaRthritis (SCUlpTOR) trial is designed as a
As multipotent precursor cells, mesenchymal stem parallel, superiority, randomised, double-­blind, placebo-­
cells (MSCs) have been suggested to be effective through controlled, two-­arm clinical trial with 1:1 allocation ratio.
their release of trophic factors that modulate inflamma- The protocol is described using the Standard Protocol
tion and recruit and stimulate resident cells to enhance Items: Recommendations for Interventional Trials guide-
repair.14 This effect has been assessed both preclinically lines on standard protocol items for clinical trials and the
and clinically in a pilot trial, which showed an increase in results will be reported using the Consolidated Standards
the articular cartilage volume assessed by MRI.15 Other of Reporting Trials statement (online supplemental file 1
pilot trials or phase II studies aslo showed clinically signif- full study protocol version 8, 18 June 2021).
icant pain and functional improvement without observed
serious adverse events (AEs).16–19 The studies indicated Study setting
that repeated intra-­articular injections of MSCs achieved The SCUlpTOR trial will be conducted in Sydney and
more consistent OA stabilisation than a single injection.16 Hobart and will include two research institutes and an
MSCs can be isolated from numerous tissues but those for imaging centre. The recruitment rate at each site will be
intra-­articular injection are presently derived mainly from independent of each other:
autologous adipose tissue or bone marrow.20 21 Given the 1. Sydney.

2 Liu X, et al. BMJ Open 2021;11:e056382. doi:10.1136/bmjopen-2021-056382

 Open access

BMJ Open: first published as 10.1136/bmjopen-2021-056382 on 29 November 2021. Downloaded from http://bmjopen.bmj.com/ on May 6, 2024 by guest. Protected by copyright.
i. Institute of Bone and Joint Research (IBJR), 10. Willingness to stop or minimise the use of nonsteroi-
Kolling Institute of Medical Research, The dal anti-­inflammatory drugs (NSAIDs) and other an-
University of Sydney (administrative procedures). algesics (except paracetamol for rescue pain relief)
ii. Castlereagh Imaging, St Leonards (face-­ to-­
face for the duration of the study.
study procedures). 11. Willingness to undergo a 1-­week medication wash-­
2. Hobart. out (for all pain medications) before each pain as-
i. Menzies Institute for Medical Research, University sessment survey at baseline, 3, 6, 9, 12, 15, 18, 21 and
of Tasmania (administrative and face-­to-­face study 24 months.
procedures). 12. Willingness to avoid a new treatment for KOA during
The IBJR will be the coordinating centre for this trial. the study.
13. Willingness and ability to travel to the study vis-
Participant recruitment and eligibility criteria its at either Castlereagh Imaging (St Leonards and
Four hundred and forty participants (220 per arm) with Cremorne in Sydney) or Menzies Institute (Hobart)
symptomatic KOA will be recruited from the community and Qscan Radiology (North Hobart).
and research volunteer databases. The recruitment strat- Participants will be excluded if they meet any of the
egies will include: (a) mailed/emailed advertisements following criteria:
to the research volunteer databases; (b) posters/flyers 1. Incomplete online screening surveys or non-­
placed on medical practices and community areas or responders after completing their online screening
newsletters/e-­newsletters in the universities and institutes; survey but before being enrolled in the study.
(c) advertisements/postings on social media networks 2. Women who are pregnant or breast feeding, or wom-
(eg, Facebook, Twitter) and research institute websites; en of childbearing potential not willing to use contra-
(d) local and major newspaper advertisements or news- ceptive methods for the duration of the study.
letter listings; (e) radio or TV interviews; (f) clinical trial 3. Radiographic evidence of predominant lateral ti-
recruitment companies; (g) community-­based events. biofemoral or patellofemoral disease based on the se-
Participants will be eligible for the study if they meet all verity of joint space narrowing in each compartment.
the inclusion criteria listed below: 4. Bilateral symptomatic KOA if the patient-­ reported
1. Ability and willingness to participate and complete pain intensity in the contralateral knee is ≥30 on a
the study. 0–100 VAS.
2. Functional English, internet access and an active 5. Significant injury in the study knee that led to sub-
email account. stantial loss of function or surgeries in the past 6
3. ≥40 years old, male or female, Australian citizen or months (eg, fracture, ligament rupture, joint disloca-
permanent resident having a valid medicare account tion, trauma, laceration or nerve damage).
for cost-­effectiveness analysis. 6. Surgery on the study knee in the past 12 months or
4. Presence of knee pain for at least half of the days in expected joint surgery for the study knee in the next
the previous month (the most painful knee will be 24 months.
considered as the study knee). 7. Prior knee joint replacement or high tibial osteotomy
5. Average pain intensity ≥40 and ≤90 out of 100 on a in the study knee.
Visual Analogue Scale (VAS) in the study knee over 8. History of crystalline (eg, gout, calcium pyrophos-
the last week prior to the online screening and base- phate deposition disease), autoimmune arthritis (eg,
line surveys. rheumatoid arthritis, psoriatic arthritis, systemic lu-
6. Kellgren and Lawrence grade (KLG) two or three pus erythematosus, ankylosing spondylitis), haemo-
and medial tibiofemoral minimum joint space width chromatosis or fibromyalgia. Except for the following
between 1 and 4 mm of the study knee based on conditions:
fixed-­flexion posteroanterior (PA) knee radiograph i. Participants diagnosed with gout are eligible for
using Synaflexer, which will be assessed by a trained the study as long as the condition is being appro-
rheumatologist with experience in KOA research. priately treated and they have not experienced
7. KOA defined by the American College of flare-­ups for at least 12 months.
Rheumatology (ACR) criteria using medical history, ii. Participants diagnosed with haemochromatosis
patient-­ reported symptoms/signs and radiographic but with normal iron levels for at least 12 months
findings:27 (a) knee pain on most days; (b) osteo- are eligible for the study.
phytes on X-­ray and (c) one of the three following 9. Signs of acute knee joint inflammation (ie, red, swol-
criteria: age >50 years, less than 30 min of morning len and hot) and/or abnormal synovial fluid sugges-
stiffness and crepitus on active motion. tive of crystals or infection.
8. Willingness to undergo a new knee X-­ray and MRI. 10. Any painful muscular or neurological condition of
9. Willingness to stop or maintain a routine (ie, on the the lower limb that, in the opinion of the investiga-
same dosage and frequency) of conservative treat- tor, is the main contributor to the pain and/or loss of
ments (eg, physiotherapy, exercise, knee brace, oral function in the study knee which may interfere with
supplements) for the duration of the study. the self-­reported assessment (eg, fracture, ligament

Liu X, et al. BMJ Open 2021;11:e056382. doi:10.1136/bmjopen-2021-056382 3

Open access

BMJ Open: first published as 10.1136/bmjopen-2021-056382 on 29 November 2021. Downloaded from http://bmjopen.bmj.com/ on May 6, 2024 by guest. Protected by copyright.
rupture, bursitis, tendinitis, hypermobility syndrome, bank and subjected to an extensive series of in-­process
joint infection, patellofemoral pain syndrome, dia- and release tests to optimise safety and batch to batch
betic neuropathy, pain referred from the back, pain reproducibility. The MSCs will be stored in the vapour
following hip, knee or ankle trauma or surgery). phase of liquid nitrogen, under temperature-­controlled
11. Cancer or other tumour-­like lesions, except for skin conditions at each research institute site (ie, Kolling and
cancer (non-­melanoma, removed and not active for Menzies).
at least 3 years).
12. Immunosuppression or acute infective processes. Randomisation, allocation and blinding
13. Intra-­articular hyaluronic acid injection in the study Eligible participants who consent to take part in the
knee in the past 6 months; intra-­ articular steroid study will be assigned to either active or placebo group
injections in any joint in the past 6 months; intra-­ with a 1:1 allocation rate as per computer-­generated
articular autologous blood product or stem cell injec- randomisation scheduled using random permuted
tion in the study knee in the past 12 months. block sizes and stratified by study sites (Sydney and
14. Regularly taking centrally acting analgesics (eg, opi- Hobart) and radiographic disease severity (ie, KLG 2
oid analgesics, duloxetine and pregabalin). vs 3).
15. Participation in other clinical trial and/or treatment The treatment allocation will only be available to the
received with any investigational agent within 30 days unblinded researcher using Research Electronic Data
before enrolment. Capture (REDCap) randomisation module. The injecting
16. Any unstable concurrent clinically significant acute, doctors, study coordinators, study physician, imaging
chronic medical conditions or abnormal laboratory readers and study statistician will remain blinded until
findings that, in the judgement of the investigator, the main results are analysed. The study participants will
would jeopardise the safety of the patient, interfere be blinded to group allocation until the end of the study
with the objectives of the protocol, or affect the par- after the final assessments at 24 months. There will be
ticipants’ compliance with the study requirements. immediate unblinding procedures available where there
17. Needle phobia. is a need due to medical issues. The assigned study inter-
18. Contraindication to MRI including but not limited to vention may need to be modified or discontinued in
a pacemaker, metal sutures, presence of shrapnel, or the case of AEs. The study coordinator will modify the
claustrophobia and/or inability to fit into the MRI study intervention in agreement with the Principal Inves-
knee coil. tigators. Participants undergoing modifications will be
Interventions retained in the trial.
The active group will receive three intra-­articular knee
injections of Cymerus MSCs at baseline, week 3 and week Concomitant and excluded medications and care
52. Each injection will consist of 2.5×107 cell culture-­ Participants will continue to take medications for other
expanded allogeneic MSCs, which are suspended in 5 mL health conditions as usual. Participants who are on a stable
excipient solution containing multiple electrolytes injec- dose of supplements, physiotherapy or other conserva-
tion with 10% human serum albumin and 2.5% dimethyl tive treatments for KOA will be asked to either stop or
sulfoxide as a non-­ toxic cryoprotectant. The placebo maintain their existing treatment regimens for the dura-
group will receive three intra-­articular knee injections tion of the trial. The use of oral or topical NSAIDs and
of saline containing 0.9% sodium chloride at the same other analgesics for KOA will be either discontinued or
timeframe. The rationale for choosing normal saline maintained at their lowest dosage for the duration of the
as the comparator as distinct from excipient is to facil- trial. The following medications and interventions will be
itate comparison with other injection trials where the prohibited during the study: (a) use of centrally acting
magnitude of effect from normal saline is known. The analgesics (eg, opioids, duloxetine and pregabalin); (b)
intra-­
articular knee injections will be guided by ultra- investigational products from another clinical trial; (c)
sound using one of the three methods: (a) lateral supra- intra-­articular injections of any other agents; (d) surgery
patellar approach (effusion present); (b) patellofemoral in the study knee; (e) any new treatment for KOA. Parac-
approach (effusion absent) or (c) medial infrapatellar etamol (up to 3000 mg/day) will be allowed to use as
approach (effusion absent). rescue pain relief during the study.
Cymerus MSCs (CYP-­004) will be provided by Cynata Concomitant and excluded medications mentioned
Therapeutics (Carlton, Victoria, Australia) and manu- above will be monitored using fortnightly and monthly
factured by Waisman Biomanufacturing (Madison, surveys. Participants will be asked to undergo a 1 week
Wisconsin, USA) according to the US current Good pain medication wash-­out before each pain assessment
Manufacturing Practice standards. The Cymerus MSCs survey at baseline, 3, 6, 9, 12, 15, 18, 21 and 24 months
will be produced in batches according to the recruit- to ensure the self-­reported pain intensity is accurate.
ment schedule due to the limited shelf life of 2 years from There will be at least 7 days gap between the treatment
the date of manufacture. All batches of Cymerus MSCs injections and any dose of vaccines (ie, COVID-­ 19
product are derived from the same donor/master cell vaccine).

4 Liu X, et al. BMJ Open 2021;11:e056382. doi:10.1136/bmjopen-2021-056382

 Open access

BMJ Open: first published as 10.1136/bmjopen-2021-056382 on 29 November 2021. Downloaded from http://bmjopen.bmj.com/ on May 6, 2024 by guest. Protected by copyright.
Outcome measures v. Knee-­related quality of life: four questions regard-
The descriptions of primary, secondary and exploratory ing the knee-­related quality of life over the last
outcome measures are as below: week contribute to the score which ranges from
1. Patient-­Acceptable Symptom State (PASS): PASS is 0 to 100 with lower scores indicating worse quality
defined as the value beyond which patients consider of life.
themselves well. The patient-­reported knee pain in- 6. Physical Activity Scale for the Elderly (PASE): the
tensity will be measured using VAS from 0 (no pain) PASE is a self-­report measure designed to capture
to 100 (worst pain possible) with the question ‘How and assess occupational, household and leisure activ-
much pain in your knee did you experience on aver- ities typically performed by older adults. Twelve ques-
age during knee movement while performing daily tions regarding levels of physical activity over the past
activities over the past week?’. The PASS threshold for week contributes to the score, which ranges from 0
pain intensity in people with KOA is less than 32 mm to 793 with higher scores indicating greater levels of
on the 0–100 mm VAS.28 physical activity.32
2. Central medial femorotibial compartment (cMFTC) 7. Assessment of Quality of Life-­ Eight Dimensions
cartilage thickness: the cMFTC cartilage thickness (AQoL-­8D): the AQoL-­8D is a 35-­item health utility
will be computed from segmentations of the weight-­ instrument, consisting of two super dimensions of
bearing femorotibial cartilages performed by man- physical and mental health or eight dimensions: in-
ually drawing disarticulation contours around the dependent living, pain, senses, mental health, happi-
cartilage edges, section by section in all MRI slices ness, coping, relationships and self-­worth. Thirty-­five
depicting the cartilage of the study knee.29 The MRI questions of eight dimensions regarding physical and
will use a 3.0T whole-­body system with dedicated ex- psychosocial quality of life over the past week contrib-
tremity coil and a fat suppressed, 3D double echo at ute to the score, which ranges from 0 to 100, with
steady state (DESS) sequence. higher scores indicating better quality of life.33
3. Wong-­ Baker FACES Pain Rating Scale (WBS): the 8. Cartilage thickness in other femorotibial plates and
WBS is used to measure pain on a numerical scale subregions: Cartilage thickness of total femorotibial,
(0, 2, 4, 6, 8 or 10) with six faces, where the patient lateral femorotibial, medial femorotibial, medial tib-
marked the face that better described the pain inten- ial, medial femoral, lateral tibial and lateral femoral
sity. The number 0 and a smiling face denote no pain, will be measured using the same method as the cM-
while the number 10 and a crying face denote the FTC cartilage thickness described above.
most severe pain.30 9. MRI Osteoarthritis Knee Score (MOAKS): the
4. Patient Global Assessment (PGA): PGA will be as- MOAKS instrument is an MRI semi-­quantitative scor-
sessed using the question ‘Considering all the ways ing of KOA, which was developed and tested on im-
your knee osteoarthritis affects you, how have you ages obtained on a 3.0T MRI system with a dedicated
been during the past week?’ along with a 0–100 VAS peripheral knee coil.34 The subscores of knee struc-
where 0 is very well and 100 is very poor. tural changes are detailed as below:
5. Knee injury and Osteoarthritis Outcome Score i. Change in the number of areas with worsening in
(KOOS): the KOOS is a knee-­specific instrument, cartilage thickness categorised as 0, 1, 2, or ≥3.
developed to assess the patients’ opinion about their ii. Worsening in osteophytes scored in each of the
knee and associated problems, which holds 42 items 12 locations according to size and categorised as
in five separately scored subscales,31 as detailed below: yes or no.
i. Pain: nine questions regarding knee pain in the iii. Change in bone marrow lesions (BMLs) scored
last week during various positions or movements based on the standardised regions ranging from
contribute to the score, which ranges from 0 to 0 to 3. Medial tibial and medial femoral condyle
100 with lower scores indicating worse pain. region BML scores will be added using categorical
ii. Other symptoms: seven questions regarding scoring (range 0–3 per region).
knee symptoms over the last week contribute to iv. Worsening in meniscal morphology features
the score, which ranges from 0 to 100 with lower scored on medial and lateral meniscus for the an-
scores indicating worse symptoms. terior, body and posterior horn and categorised
iii. Function in daily living (ADL): seventeen ques- as yes or no.
tions regarding the degree of difficulty perform- v. Change in whole knee effusion (effusion-­synovitis)
ing daily activities over the last week contribute to categorised as ‘worsen’, ‘no change’ or ‘improve’.
the score which ranges from 0 to 100 with lower vi. Change in infra-­patellar fat pad synovitis (Hoffa’s
scores indicating worse function. synovitis) categorised as ‘worsen’, ‘no change’ or
iv. Function in sport and recreation: five questions ‘improve’.
regarding the degree of difficulty performing 10. Global rating of change (GRC): the GRC scales are
sports and recreation activities over the last week designed to quantify a patient’s improvement or de-
contribute to the score which ranges from 0 to 100 terioration over time to determine the effect of an in-
with lower scores indicating worse function. tervention.35 The GRC for pain, function and overall

Liu X, et al. BMJ Open 2021;11:e056382. doi:10.1136/bmjopen-2021-056382 5

Open access

BMJ Open: first published as 10.1136/bmjopen-2021-056382 on 29 November 2021. Downloaded from http://bmjopen.bmj.com/ on May 6, 2024 by guest. Protected by copyright.
will be assessed using the question ‘Which option 2. Change in cMFTC cartilage thickness from baseline to
best represents the change in pain/change in func- 24 months.
tion/overall change in your knee since you began the The secondary endpoints are:
study?’, scored using a 5-­point Likert scale ranging 1. Change in knee pain intensity from baseline to 3, 6, 9,
from much better to much worse. 12, 15, 18, 21 and 24 months using VAS and WBS.
11. Treatment satisfaction: participants’ opinion of sat- 2. Change in PGA from baseline to 3, 6, 12 and 24 months.
isfaction will be assessed using a ‘yes/no’ question 3. Change in KOOS from baseline to 3, 6, 12 and 24
‘Taking into account all the activities you have during months.
your daily life, your level of pain, and also your func- 4. Change in PASE from baseline to 3, 6, 12 and 24
tional impairment, do you consider that your current months.
state is satisfactory?’. For those who answer ‘no’, their 5. Change in AQoL-­8D from baseline to 3, 6, 12 and 24
opinion of treatment failure will be assessed using a months.
‘yes/no’ question ‘Would you consider your current 6. Change in cartilage thickness (quantitative) in other
state as being so unsatisfactory that you think the femorotibial plates and subregions from baseline to 24
treatment has failed?’ months.
12. Quality-­ adjusted life year (QALY): the QALY is a 7. Change in structural knee features assessed by MOAKS
multidimensional measure of health outcome that (semi-­quantitative) from baseline to 24 months.
encompasses both quality-­of-­life and quantity-­of-­life 8. Change in cartilage T2 relaxation time estimates: the
(survival) gains, which will be calculated by multiply- T2 relaxation time will be assessed from the DESS
ing life years by the index of utility derived from the MRI for the same regions of interest as for cartilage
AQoL-­8D on the QALY scale ranging from 0 (equiva- morphometry.
lent of being dead) to 1 (full health).36 The exploratory endpoints include: (a) GRC for pain,
13. Cartilage T2 relaxation time estimates: the T2 re- function and overall at 3, 6, 12 and 24 months; (b) treat-
laxation time will be assessed from the DESS MRI ment satisfaction at 24 months; (c) QALY at 24 months;
for the same regions of interest as for cartilage (d) cost-­effective analysis; (e) consumption of rescue
morphometry. medication; (f) individual patient placebo response; (g)
14. Cost-­effectiveness: the cost-­effectiveness analysis will blinding success; (h) treatment adherence; (i) AEs.
be performed using a combination of the Medication
Benefits Scheme/Pharmaceutical Benefits Scheme Patient and public involvement
(MBS/PBS) data extracted for the study period and People with KOA have been involved in the study design
the monthly healthcare usage surveys. by participating in a focus group during study preparation
15. Consumption of rescue medication: the consump- and in a small pilot study. We conducted a prestudy patient
tion of paracetamol and other pain medications will focus group (three patient representatives involved)
be monitored by inspection of fortnightly surveys, to discuss the study design, obtain their opinions and
which will be reported by participants. experience that are relevant to the study. We have also
16. Individual patient placebo response: baseline included two pilot participants (one in the active group
Multidimensional Psychological Questionnaire and one in the placebo group) at each site (four in total)
(MPsQ) modules will be used to assess the placebo re- to assess the study procedures, the success of blinding,
sponse,37 38 which is a self-­reported questionnaire us- the burden of the intervention and time consumed in
ing a 5-­point scale ranging from 1 (strongly disagree) different processes in this research. The pilot partici-
to 5 (strongly agree) contributing to the evaluation pants received two injections (baseline and week 3) and
of individual patient response to placebo by assess- was followed up to 1 month from the baseline visit. They
ing participants’ basic personality traits, expectation reviewed participant-­ related documents and provided
traits and perception. feedback on them. They will be asked to review the lay
17. Blinding success: will be measured by asking which summary and an infographic summarising the main study
treatment participants believe they received. The results before sending them out to the study participants
injecting doctors and blinded assessors will also be during study close-­out. The data collected from the pilot
asked which treatment they believe was given to the participants will be excluded from the statistical analysis.
participants.
18. Treatment adherence will be reported as the number Study procedures
of injections administered. An outline of the study events and procedures is
19. AEs will be assessed at each study visit and by inspec- summarised in figure 1 and table 1.
tion of monthly surveys.
Screening and radiographic assessment
Trial endpoints People who are interested in the study will be required to
The primary endpoints will be: complete an online prescreening survey (http://​tinyurl.​
1. The proportion of participants reaching PASS thresh- com/​sculptor-​trial) to determine their initial eligibility.
old for pain intensity (VAS≤32) at 24 months. Potentially eligible participants will be redirected to

6 Liu X, et al. BMJ Open 2021;11:e056382. doi:10.1136/bmjopen-2021-056382

 Open access

Figure 1 Flow diagram of the study protocol. BMJ Open: first published as 10.1136/bmjopen-2021-056382 on 29 November 2021. Downloaded from http://bmjopen.bmj.com/ on May 6, 2024 by guest. Protected by copyright.

complete an electronic-­Participant Consent Form (e-­PCF) caudally,39 which will be used to assess eligibility and
after watching a video containing further details about exclude those who have predominant lateral tibiofemoral
the study (online supplemental file 2 consent form). Indi- KOA. The skyline view of the study knee will be taken with
viduals who sign the e-­PCF will be referred to have a knee 65° of knee flexion, which will be used to exclude those
X-­ray taken either at Castlereagh Imaging at St Leonards who have predominant patellofemoral KOA. All these
(Sydney, Australia) or at Qscan Radiology (North Hobart, radiographs will be assessed by a trained rheumatologist.
Australia) (visit 1). Fixed-­flexion PA knee radiographs
for both knees will be taken using the Synaflexer X-­ray Baseline survey (online)
positioning frame with feet externally rotated 10°, the After X-­ray assessment, eligible participants who are taking
knees and thighs touched the vertical platform anteriorly any pain medications or analgesics will be requested to
and the X-­ray beam angulated 10° (the angulation might undergo a 1 week wash-­out. After an appropriate wash-­out
be adjusted in order to achieve the best quality of films) if applicable, participants will be sent a pre-­ baseline

Liu X, et al. BMJ Open 2021;11:e056382. doi:10.1136/bmjopen-2021-056382 7

8
Table 1 Standard Protocol Items: Recommendations for Interventional Trials diagram of enrolment, interventions and assessments for the Stem Cell injections for
symptomatic relief and strUctural improvement in people with Tibiofemoral knee OsteoaRthritis study

Screening Baseline Follow-­up

Timepoints Week 3 Month 3 Month 6 Month 9 Month 12 Month 15 Month 18 Month 21 Month 24
  −t4 −t3 −t2 −t1 t0 t1 t3 t5 t9
Open access

  E-­survey Visit 1 E-­survey Visit 2 Visit 3 Visit 4 t2 Visit 5 t4 Visit 6 t6 t7 t8 Visit 7

Enrolment                    
 Screening                    
   E-­prescreening X                    
   E-­consent X                    
   Knee X-­ray X                    
   Wash-­out* X     X X X X X X X X
  Demographics X                    
  Medical history X                    
  Comorbidity X                    
  Knee MRI X                   X
 Face-­to-­face visit
  Temperature X X   X   X       X
  Blood pressure X X   X   X       X
  Height X                  
  Weight X     X   X       X
  Blood/urine X     X           X
  
MBS/PBS X                  
consent
  
Treatment X                  
allocation
  Synovial fluid* X X       X        
Interventions                    
 MSCs X X       X        
 Placebo X X       X        
Assessments                    
 Primary outcome
  PASS pain X                   X
  cMFTC X                   X
cartilage
thickness

Liu X, et al. BMJ Open 2021;11:e056382. doi:10.1136/bmjopen-2021-056382

Continued

BMJ Open: first published as 10.1136/bmjopen-2021-056382 on 29 November 2021. Downloaded from http://bmjopen.bmj.com/ on May 6, 2024 by guest. Protected by copyright.
 Table 1 Continued

Screening Baseline Follow-­up

Timepoints Week 3 Month 3 Month 6 Month 9 Month 12 Month 15 Month 18 Month 21 Month 24
  −t4 −t3 −t2 −t1 t0 t1 t3 t5 t9
  E-­survey Visit 1 E-­survey Visit 2 Visit 3 Visit 4 t2 Visit 5 t4 Visit 6 t6 t7 t8 Visit 7
 Secondary outcomes
  VAS pain X     X X X X X X X X
  Wong-­Baker X     X X X X X X X X
FACES pain
  PGA X     X X   X       X
  KOOS X     X X   X       X
  PASE X     X X   X       X
   AQoL-­8D X     X X   X       X
  
Total and X                   X
subregional
FT cartilage

Liu X, et al. BMJ Open 2021;11:e056382. doi:10.1136/bmjopen-2021-056382

thickness
  MOAKS X                   X
  
Cartilage T2 X                   X
relaxation time
Exploratory measures
 GRC     X X   X       X
 Treatment                   X
satisfaction
 QALY                   X
 MPsQ† X                    
 Blinding success                    
 Healthcare usage X X       X        
surveys (monthly) ‍ ‍
 Pain medication                  
‍ ‍
(fortnightly)
 Adverse events                  
‍ ‍
(monthly)
*If applicable.
†PER-­MPsQ will be collected after the first injection.
AQoL-­8D, quality of life eight dimension; cMFT, central medial femorotibial; GRC, global rating of change; KOOS, Knee injury and Osteoarthritis Outcome Score; MBS/PBS, Medication
Benefits Scheme/Pharmaceutical Benefits Scheme; MOAKS, MRI Osteoarthritis Knee Score; MPsQ, Multidimensional Psychological Questionnaire; MSCs, mesenchymal stem cells; PASE,
Physical Activity Scale for the Elderly; PASS, Patient-­Acceptable Symptom State; PGA, Patient Global Assessment; QALY, Quality-­adjusted life year; VAS, Visual Analogue Scale.
Open access

9
BMJ Open: first published as 10.1136/bmjopen-2021-056382 on 29 November 2021. Downloaded from http://bmjopen.bmj.com/ on May 6, 2024 by guest. Protected by copyright.
Open access

BMJ Open: first published as 10.1136/bmjopen-2021-056382 on 29 November 2021. Downloaded from http://bmjopen.bmj.com/ on May 6, 2024 by guest. Protected by copyright.
survey for further assessment of their knee pain intensity.

Fat sat
Table 2 MRI sequences used for structural outcome measures in the Stem Cell injections for symptomatic relief and strUctural improvement in people with Tibiofemoral
Those who are eligible (ie, knee pain ≥40 and ≤90 on a

Yes
Yes
Yes
Yes
No
0–100 VAS) will continue to complete the baseline survey
for the collection of baseline information including

Bandwidth
demographics, clinical characteristics, psychological and
comorbidity assessments, and self-­ reported outcome

234.7
284.8
235.3
282.6
measures.

241
Knee MRI (visit 2)

Averages
Participants who complete the baseline survey will
be referred to have a knee MRI at either Castlereagh

(NSA)
Imaging (Cremorne, Sydney) or Qscan Radiology (North

1
1
5.2 1
1
1
Hobart). The details of the 3.0T MRI machines, dedi-
cated knee coils and acquisition sequences at each site

TE

21
23
23
23
are specified in tables 2 and 3. Participants who have frac-
tures, infections or tumours will be excluded after the

18

3919
3025
2000
3095
MRI assessment.

TR
Baseline visit (visit 3) and enrolment

Voxel size

2 echoes 0.62×0.62

0.36×0.4
0.25×0.3
0.35×0.4
Eligible participants after MRI will be required to attend

0.3×0.4
a face-­to-­
face baseline assessment at a designated site
before their first injection. Temperature, blood pressure,
height and weight will be measured during this visit (the
injection will not proceed in case of systolic blood pres-

factor
Turbo

(TSE)
sure over 180 mm Hg). MBS/PBS consent form for data

11
12
10
11
extraction to be used in the cost-­effectiveness analysis will
be collected during this visit. Thirty millilitre of blood

Resolution

95%phase
90%slice
and 50 mL of urine will be collected for safety monitoring

532×343
260×260

363×416
640×448
388×290
purposes, part of the collected samples will be stored for

NSA, number of signal averages; TE, the echo time; TR, the repetition time; TSE, turbo spin echo.
those who consent for deposition in the biobank for future
analysis. The blood samples used for glucose (random),
chemistry, liver function, full blood count, estimated

2×1.2

MRI machine details: manufacturer: PHILIPS; model: Ingenia; software version: V.5.6.1.
Ipat

glomerular filtration rate, C reactive protein tests will

1.7
1.1
1.4
1.2

be analysed by Pathology North (Sydney) or Pathology

South (Hobart). The urinalysis will be conducted by the
knee OsteoaRthritis study in Castlereagh Imaging (Cremorne, Sydney)

Scan
time

3.32
2.43
7.05
2.12
3.57

research team using urine test strips, which will include

specific gravity, pH, leukocytes, nitrite, protein, glucose,
ketones, urobilinogen, bilirubin, and blood.
encoding

After the completion of the face-­to-­face baseline assess-

Phase

A>P

H>F

ment, participants will be considered to enter the study.

R>L
R>L
R>L

A study enrolment number (composed of REDCap ID

and enrolment order) will be assigned and recorded in
(mm)
Slice

0.25
0.25
gap

all participant-­specific study documents. Enrolled partic-

0.3
0.3

ipants will receive a Participant’s Identification Card for

the purpose of urgent contact in case of any medical
thickness

issues. A notification letter will be posted to their general

practitioner (GP) to inform of their enrolment in the
(mm)
Slice

0.75

2.5
2.5

study.
3
3
Slices

Intervention preparation and injection

200

40
32
36
40

In Sydney, the injections will be prepared and delivered

to participants at Castlereagh Imaging. A dry shipper
will be used to transport the frozen cells between the
qDESS_sagittal
pd_coronal_fs

pd_sagittal_fs
Sequence ID

Kolling Institute and Castlereagh. In Hobart, the cells

pd_axial_fs
pd_coronal

will be prepared and injected at the Menzies Institute.

The MSCs CryoBag will be thawed in a 37°C water bath.
The unblinded researcher will wear appropriate personal
protective equipment and transfer 5 mL stem cell solution

10 Liu X, et al. BMJ Open 2021;11:e056382. doi:10.1136/bmjopen-2021-056382

 Open access

BMJ Open: first published as 10.1136/bmjopen-2021-056382 on 29 November 2021. Downloaded from http://bmjopen.bmj.com/ on May 6, 2024 by guest. Protected by copyright.
to a 5 mL syringe using an aseptic technique. The syringe

Fat sat
Table 3 MRI sequences used for structural outcome measures in the Stem Cell injections for symptomatic relief and strUctural improvement in people with Tibiofemoral
will be covered with masking tape to occlude its contents

Yes

Yes

Yes
No
and an amber connector will be used to mask the tip of

–
the syringe. The same procedure will be used to prepare

Bandwidth
the placebo (5 mL normal saline). The sheathed/masked
syringe will be injected within 60 min post-­ thaw. The

283
237
233
241
233
administration procedure of MSCs and placebo will be
the same with guidance by ultrasound imaging during a
10 s injection period into the knee joint. Participants will
Averages

return to the same site for their second and third injec-

MRI machine details: manufacturer: SIEMENS; model: Skyra; software version: syngo MR E11; IDEA license: N4_VE11C_LATEST_20160120 from Imperial College London.
(NSA)

tion at week 3 (visit 4) and week 52 (visit 6) from the first

1
1
1
1
1
injection.
5.2 In case of joint effusion, synovial fluid will be aspirated
TE
26
25
19

19 before the injection. If the synovial fluid appearance is

18.16

abnormal (ie, cloudy, opaque and/or coloured), treat-

4210
3180
4440

5000

ment injection will not proceed. The collected sample

TR

will be sent to pathology to check the cell count, culture

and crystal analysis. If the results show a high white cell
0.4×0.4×3.0
0.3×0.3×2.5
0.4×0.4×2.5
0.5×0.5×1.5
0.4×0.4×2.5
factor (TSE) Voxel size

count, infection or crystal, then the participant will be

withdrawn from the study and results will be sent to their
GP for follow-­up care. If the result is not clinically signifi-
cant, the treatment injection will be rescheduled.
2 echoes

Follow-up
Turbo

Participants will be asked to complete online follow-­up

8
9
10

10

assessment surveys at 3, 6 (visit 5), 12 (visit 6) and 24

months (visit 7). Body temperature and blood pressure
Resolution
384×288
320×224
358×448
320×320
358×448

will be assessed at each visit after enrolment. Body weight

will be measured at 6, 12 and 24 months. Safety blood
and urine samples will be collected at 6 and 24 months.
NSA, number of signal averages; TE, the echo time; TR, the repetition time; TSE, turbo spin echo.

A follow-­up MRI scan will be performed at 24 months.

Ipat

Pain medication use will be monitored using a fortnightly

2
2
2
2
2

online survey. Healthcare usage and AEs will be moni-

Scan

tored using a monthly online survey. Pain intensity will be

time
2.27
1.08
2.32
3.22
2.34

monitored every 3 months using an online survey.

encoding

Adverse events
Phase

Any untoward events that occur from the time of the

knee OsteoaRthritis study in Qscan Radiology (North Hobart)

H>F
H>F
R>L
R>L
R>L

enrolment will be monitored and assessed using a

monthly survey indicating the start and end date of
0.25
0.25

0.25
(mm)
Slice

20%
0.3
gap

the event, details of the event, any actions taken and

outcome. Participants will also be monitored for AEs at
thickness

each study visit after the enrolment. The study physician

will assess the severity (ie, severe, moderate, mild) and
(mm)
Slice

2.5
2.5
1.5
2.5

causality (ie, definitely related, probably related, possibly

3

related, unlikely related, not related) of the AEs and give

advice accordingly. Any abnormal blood and urine tests
60–80 per echo

will be assessed as clinically significant or non-­clinically

significant.
An AE will be considered ‘serious’ when it causes: (a)
Slices

death, (b) life-­threatening event, (c) inpatient hospi-

40
36
36

40

talisation or prolongation of existing hospitalisation,

(d) persistent or significant disability/incapacity, (e) a
qDESS_sagittal

congenital anomaly/birth defect in the offspring of a

pd_coronal_fs

pd_sagittal_fs
Sequence ID

participant.
pd_coronal
Pd_fs_ax

Data and safety monitoring board (DSMB), composed

of two clinical researchers, one clinician and one inde-
pendent statistician, will meet every 6 months via

Liu X, et al. BMJ Open 2021;11:e056382. doi:10.1136/bmjopen-2021-056382 11

Open access

BMJ Open: first published as 10.1136/bmjopen-2021-056382 on 29 November 2021. Downloaded from http://bmjopen.bmj.com/ on May 6, 2024 by guest. Protected by copyright.
videoconference and review the AEs. All serious AEs will outcome data based on their randomised treatment allo-
be reported to DSMB within 24 hours of becoming aware cation, regardless of compliance. Safety analyses will be by
of the events regardless of causality to ensure the safety of treatment received, including participants who received
the participants. at least one dose of study treatment. For the study to be
declared a success both coprimary efficacy endpoints
Participant retention and withdrawal need to be significant at the two-­sided α level of 0.05 in
Strategies to maximise retention will be implemented (eg, favour of intra-­articular MSCs injections.
survey reminders, emails, phone calls, virtual birthday Demographic characteristics and baseline scores will be
cards, gift vouchers, monthly health tips and survey presented to assess the comparability of treatment groups
reminders via SMS). A participant will be considered as a at baseline. Participant characteristics will be summarised
drop-­out and will be excluded if he/she stops completing as mean (SD) for continuous variables or medians (quar-
the surveys and is unable to be reached. tiles) if the distribution is skewed. Counts with percent-
A participant will be able to withdraw from the study ages will be presented for categorical variables.
at any time by signing the electronic withdrawal form. In All continuous primary and secondary outcome
order to improve data retention, participants will have the measures will be summarised as means (SD) at each
option to continue to provide follow-­up data via online time point of interest by treatment group. The between-­
surveys and to attend the final MRI visit, or not. If a partic- group difference (with 95% CI) in mean change from
ipant chooses to stop their involvement completely, no baseline will be presented and compared using indepen-
further data will be collected from this participant, but dent samples t-­test. For binary outcome variables differ-
previously collected data will be included in the analysis ences in proportions will be presented with 95% CIs and
with the permission of the participant. compared using χ2 tests. The conditional binomial test
The investigator may also withdraw participants from will be used when expected cell counts are small. The
the study to protect their safety. Participants who drop out Wilcoxon rank-­sum test will be used to compare ordinal
and withdraw from the study will not be replaced. or discrete/continuous outcomes but not normally
distributed, between groups. Poisson regression will be
Statistical methods employed to model count variables.
Sample size estimation Analysis adjusted for baseline values to account for
The sample size calculation was based on the primary possible floor and ceiling effects will be performed for
objective to show statistical superiority of intra-­articular the following outcome measures: PASS, VAS, WBS, PGA,
MSCs injections compared with placebo for both copri- KOOS, PASE, AQoL-­8D and cartilage thickness (cMFTC
mary outcomes PASS for knee pain and cMFTC carti- and other femorotibial plates and subregions). For
lage thickness change at 24 months. Based on previous continuous outcomes analysis of covariance models will
reports, the average proportion of patients who have be fitted separately at each timepoint of interest with
achieved the PASS threshold was 35% in the placebo the change from baseline as the dependent variable. For
group (ranging from 33.1 to 35.5) and 48% in the inter- binary outcomes logistic regression will be used. Other
vention groups (ranging from 42.2% to 56.1%).40 41 In a covariables of interest will include age, gender and body
recent nested case-­control study enrolling KOA patients, mass index. Generalised estimating equations will be used
the change of cMFTC cartilage thickness over 24 months to explore trends in the effect of treatment over time.
was −0.32 mm (SD=0.40) for those with narrowed medial To assist with the interpretation of the results, we
tibiofemoral joint and persisted knee pain at baseline will calculate the cut-­ point for the minimal clinically
when compared with the control with neither radio- important improvement for pain using the mean change
graphic nor pain progression (−0.12 mm, SD=0.28).42 anchor-­ based approach. The pain item of GRC scale
A total number of 440 participants (220 per treatment will be used as the anchoring question with participants
arm) will provide 86% power to detect a 16% increase answering ‘slightly better’ considered to be reporting a
in the proportion of participants who achieve the PASS minimum clinically important improvement.45
threshold (35% vs 51%) based on a two-­sided significance The two-­sided 5% significance level will be used for all
level of 0.05.43 This sample size also provides 94% power hypothesis tests, with no adjustments for multiple testing.
to detect a between-­group difference of 0.15 mm in mean No interim analysis will be carried out for this study.
cartilage thickness change over 24 months.44 The power
to evaluate the joint effect of both coprimary endpoints is Cost-effectiveness analysis
at least 80%. The sample size allows for a drop-­out rate of The measure of effectiveness will be QALYs based on
up to 20% over 24 months. measures obtained from the AQoL at each time point
and transformed into a utility index using weights derived
Statistical analysis plan from the Australian population. Costs to the health-
The statistical analysis will be performed by a qualified care system will be based on MBS and PBS costs. Costs
biostatistician who will be blinded to the group allo- of hospitalisations will be valued at standard Australian
cation. Efficacy analyses will be according to modified Refined Diagnosis Related Group cost weights. Costs of
intention to treat, including participants with available the study treatments and private healthcare services will

12 Liu X, et al. BMJ Open 2021;11:e056382. doi:10.1136/bmjopen-2021-056382

 Open access

BMJ Open: first published as 10.1136/bmjopen-2021-056382 on 29 November 2021. Downloaded from http://bmjopen.bmj.com/ on May 6, 2024 by guest. Protected by copyright.
6
be valued at published standard rates, if available, or as Raymond Purves Bone and Joint Research Labs, Kolling Institute of Medical
reported by participants in their diaries. The aggregate Research, Faculty of Medicine and Health, The University of Sydney, Sydney, New
South Wales, Australia
of such costs will be used to estimate healthcare costs 7
Castlereagh Imaging, St Leonards, Sydney, New South Wales, Australia
incurred by participants in both arms of the trial. An 8
Centre for Health Economics, Monash University, Clayton, Victoria, Australia
incremental cost-­ effectiveness ratio will be estimated 9
School of Medicine, College of Health and Medicine, University of Tasmania Faculty
based on the difference in costs and QALYs over 24 of Health, Hobart, Tasmania, Australia
10
months between treatment arms with 95% CI calculated NHMRC Clinical Trials Centre, The University of Sydney, Sydney, New South Wales,
Australia
using non-­parametric bootstrapping and presented as an 11
Representative of People with Knee Osteoarthritis, Sydney, New South Wales,
acceptability (net benefits) curve for a range of willing- Australia
ness to pay for a QALY. 12
Department of Clinical Sciences Lund, Orthopaedics, Lund University, Lund,
Sweden
13
Clinical Research Centre, Zhujiang Hospital, Southern Medical University,
Data management Guangzhou, Guangdong, China
Electronic case report forms built in REDCap, a secure
web-­based application designed to support data capture Twitter Leticia A Deveza @LeticiaDeveza and David J Hunter @ProfDavidHunter
for research studies and hosted on The University of Acknowledgements We thank patient representatives and pilot participants
Sydney server, will be used to facilitate the collection of who were involved in the study design for their valuable feedback on the study
the data throughout the study. The data will be stored procedures and relevant documents. Cynata Therapeutics Ltd (Carlton, Victoria,
Australia) provided the investigational products free of cost. Cynata provided
in a re-­identifiable format to ensure confidentiality. All
feedback on the study design but they will have no role in the data analysis,
computer-­based files will be stored in OneDrive, which interpretation, preparation and submission of the manuscript.
will be accessible only to the researchers. A back-­up of Contributors DJH is the grant holder and conceived of the study. XL drafted
REDCap data will be regularly stored in the University the paper. DJH, CD, SL, RO’C, SR and XW contributed to the study design. All
of Sydney Research Data Store. Self-­monitoring of data investigators including LAD, WMO, KC, BA, FE, WW, CL, JL, AH, DH, GH, TB revised
entry will be used to maximise data quality. this protocol critically for important intellectual content. SR, KS, KC, BA and SV will
be responsible for the implementation.
Funding The Sculptor study is funded by an NHMRC Programme Grant (ID:
APP1162874).

ETHICS AND DISSEMINATION Competing interests DJH is supported by an NHMRC Investigator Grant and
provides consulting advice for Merck Serono, TLC Bio, Tissuegene and Pfizer. CL
This protocol has been approved by The University of has provided consulting advice for Merck Serono and Galapagos Pharmaceuticals,
Sydney (USYD) Human Research Ethics Committee and receives research funding from numerous pharmaceutical companies through
(HREC) #: 2020/119 and The University of Tasmania specific services/testing contract research agreements between and managed by
(UTAS) HREC #: H0021868. All participants will be The University of Sydney or the NSLHD. SL provides consulting advice for Arthro
Therapeutics, Pfizer and Synartro. LAD has received partial reimbursement of a
required to provide informed consent. Digital informed conference registration by Pfizer. FE is CEO/CMO and co-­owner of Chondrometrics
consent to participate will be obtained from all partici- GmbH, and he has provided consulting services to Merck KGaA, Abbvie, Samumed,
pants through the REDCap software. The results of this Kolon-­Tissuegene, Servier, Galapagos, Roche, Novartis, ICM and HealthLink. WW
study will be disseminated through conferences, social is co-­owner of Chondrometrics GmbH, and has provided consulting services to
Galapagos.
media and scientific publications. No information which
could lead to the identification of a participant will be Patient consent for publication Consent obtained directly from patient(s).
included in the dissemination of results. Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It
has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have
Timelines been peer-­reviewed. Any opinions or recommendations discussed are solely
The pilot study in Sydney commenced in November and those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability
completed in December 2020. The pilot study in Hobart and responsibility arising from any reliance placed on the content. Where the
content includes any translated material, BMJ does not warrant the accuracy and
commenced in May and completed in June 2021. Recruit- reliability of the translations (including but not limited to local regulations, clinical
ment has commenced since January 2021 in Sydney and guidelines, terminology, drug names and drug dosages), and is not responsible
August 2021 in Hobart. The complete data collection for any error and/or omissions arising from translation and adaptation or
is anticipated to be completed in June 2024. The study otherwise.
close-­out is anticipated in December 2024. Open access This is an open access article distributed in accordance with
the Creative Commons Attribution Non Commercial (CC BY-­NC 4.0) license,
which permits others to distribute, remix, adapt, build upon this work non-­
Author affiliations
1 commercially, and license their derivative works on different terms, provided the
Department of Rheumatology, Royal North Shore Hospital, Northern Clinical School,
original work is properly cited, appropriate credit is given, any changes made
Faculty of Medicine and Health, The University of Sydney, Sydney, New South Wales,
indicated, and the use is non-­commercial. See: http://​creativecommons.​org/​
Australia
2 licenses/​by-​nc/​4.​0/.
Institute of Bone and Joint Research, Kolling Institute of Medical Research, The
University of Sydney, Sydney, New South Wales, Australia ORCID iDs
3
Menzies Institute for Medical Research, University of Tasmania, Hobart, Tasmania, Xiaoqian Liu http://​orcid.​org/​0000-​0002-​6375-​1800
Australia Xia Wang http://​orcid.​org/​0000-​0003-​3172-​9390
4
Institute of Anatomy and Cell Biology and Ludwig Boltzmann Institute for Arthritis Anthony Harris http://​orcid.​org/​0000-​0003-​1641-​3320
and Rehabilitation, Paracelsus Medical University Salzburg, Salzburg, Austria Stefan Lohmander http://​orcid.​org/​0000-​0002-​5424-​9448
5
Chondrometrics GmbH, Freilassing, Germany David J Hunter http://​orcid.​org/​0000-​0003-​3197-​752X

Liu X, et al. BMJ Open 2021;11:e056382. doi:10.1136/bmjopen-2021-056382 13

Open access

BMJ Open: first published as 10.1136/bmjopen-2021-056382 on 29 November 2021. Downloaded from http://bmjopen.bmj.com/ on May 6, 2024 by guest. Protected by copyright.
REFERENCES 24 Dominici M, Le Blanc K, Mueller I, et al. Minimal criteria for defining
1 Glyn-­Jones S, Palmer AJR, Agricola R, et al. Osteoarthritis. Lancet multipotent mesenchymal stromal cells. The International Society for
2015;386:376–87. cellular therapy position statement. Cytotherapy 2006;8:315–7.
2 Safiri S, Kolahi A-­A, Smith E, et al. Global, regional and national 25 Bloor AJC, Patel A, Griffin JE, et al. Production, safety and efficacy
burden of osteoarthritis 1990-­2017: a systematic analysis of iPSC-­derived mesenchymal stromal cells in acute steroid-­resistant
of the global burden of disease study 2017. Ann Rheum Dis graft versus host disease: a phase I, multicenter, open-­label, dose-­
2020;79:819–28. escalation study. Nat Med 2020;26:1720–5.
3 Sharma L. Osteoarthritis of the knee. N Engl J Med 2021;384:51–9. 26 Thavapalachandran S, Le TYL, Romanazzo S, Rashid FN, et al.
4 Hunter DJ, Schofield D, Callander E. The individual and Pluripotent stem cell-­derived mesenchymal stromal cells improve
socioeconomic impact of osteoarthritis. Nat Rev Rheumatol cardiac function and vascularity after myocardial infarction.
2014;10:437–41. Cytotherapy 2021. doi:10.1016/j.jcyt.2021.07.016. [Epub ahead of
5 Wallace IJ, Worthington S, Felson DT, et al. Knee osteoarthritis has print: 26 Sep 2021].
doubled in prevalence since the mid-­20th century. Proc Natl Acad 27 Altman R, Asch E, Bloch D, et al. Development of criteria for the
Sci U S A 2017;114:9332–6. classification and reporting of osteoarthritis: classification of
6 Cross M, Smith E, Hoy D, et al. The global burden of hip and knee osteoarthritis of the knee. Arthritis & Rheumatism 1986;29:1039–49.
osteoarthritis: estimates from the global burden of disease 2010 28 Tubach F, Ravaud P, Baron G, et al. Evaluation of clinically
study. Ann Rheum Dis 2014;73:1323–30. relevant states in patient reported outcomes in knee and hip
7 Chen F, Su W, Bedenbaugh AV, et al. Health care resource osteoarthritis: the patient acceptable symptom state. Ann Rheum Dis
utilization and burden of disease in a U.S. Medicare population 2005;64:34–7.
with a principal diagnosis of osteoarthritis of the knee. J Med Econ 29 Wirth W, Hellio Le Graverand M-­P, Wyman BT, et al. Regional
2020;23:1151–8. analysis of femorotibial cartilage loss in a subsample from the
8 Wilson R, Abbott JH. The projected burden of knee osteoarthritis osteoarthritis initiative progression subcohort. Osteoarthritis Cartilage
in New Zealand: healthcare expenditure and total joint replacement 2009;17:291–7.
provision. N Z Med J 2019;132:53–65. 30 Lai Y-­F, Lin P-­C, Chen C-­H, et al. Current status and changes in pain
and activities of daily living in elderly patients with osteoarthritis
9 Bannuru RR, Schmid CH, Kent DM, et al. Comparative effectiveness
before and after unilateral total knee replacement surgery. J Clin Med
of pharmacologic interventions for knee osteoarthritis: a systematic
2019;8. doi:10.3390/jcm8020221. [Epub ahead of print: 08 02 2019].
review and network meta-­analysis. Ann Intern Med 2015;162:46–54.
31 Roos EM, Lohmander LS. The knee injury and osteoarthritis outcome
10 Machado GC, Maher CG, Ferreira PH, et al. Efficacy and safety of
score (KOOS): from joint injury to osteoarthritis. Health Qual Life
paracetamol for spinal pain and osteoarthritis: systematic review
Outcomes 2003;1:64.
and meta-­analysis of randomised placebo controlled trials. BMJ
32 Martin KA, Rejeski WJ, Miller ME, et al. Validation of the PASE in
2015;350:h1225.
older adults with knee pain and physical disability. Med Sci Sports
11 Hunter DJ, Bierma-­Zeinstra S. Osteoarthritis. Lancet
Exerc 1999;31:627–33.
2019;393:1745–59.
33 Richardson J, Iezzi A, Khan MA, et al. Validity and reliability of
12 Pas HI, Winters M, Haisma HJ, et al. Stem cell injections in knee
the assessment of quality of life (AQoL)-­8D multi-­attribute utility
osteoarthritis: a systematic review of the literature. Br J Sports Med
instrument. Patient 2014;7:85–96.
2017;51:1125–33.
34 Hunter DJ, Guermazi A, Lo GH, et al. Evolution of semi-­quantitative
13 Piuzzi NS, Ng M, Chughtai M, et al. The stem-­cell market for the whole joint assessment of knee oa: MOAKS (MRI osteoarthritis knee
treatment of knee osteoarthritis: a patient perspective. J Knee Surg score). Osteoarthritis Cartilage 2011;19:990–1002.
2018;31:551–6. 35 Kamper SJ, Maher CG, Mackay G. Global rating of change scales: a
14 Colombini A, Perucca Orfei C, Kouroupis D, et al. Mesenchymal review of strengths and weaknesses and considerations for design. J
stem cells in the treatment of articular cartilage degeneration: Man Manip Ther 2009;17:163–70.
new biological insights for an old-­timer cell. Cytotherapy 36 Cleland J, Hutchinson C, Khadka J, et al. A review of the
2019;21:1179–97. development and application of generic Preference-­Based
15 Song Y, Du H, Dai C, et al. Human adipose-­derived mesenchymal instruments with the older population. Appl Health Econ Health
stem cells for osteoarthritis: a pilot study with long-­term follow-­up Policy 2019;17:781–801.
and repeated injections. Regen Med 2018;13:295–307. 37 Pereira A, Gossuin C, Helleputte T. Method for prediction of a
16 Freitag J, Bates D, Wickham J, et al. Adipose-­derived mesenchymal placebo response in an individual, 2017. Available: Patent US
stem cell therapy in the treatment of knee osteoarthritis: a 15/308,502; US2017/0053082
randomized controlled trial. Regen Med 2019;14:213–30. 38 Pereira A, Duale C, Clermont F, et al. (171) characterization and
17 Matas J, Orrego M, Amenabar D, et al. Umbilical cord-­derived prediction of placebo responders in peripheral neuropathic patients
mesenchymal stromal cells (MscS) for knee osteoarthritis: repeated in a 4-­week analgesic clinical trial. J Pain 2016;17:S18.
MSC dosing is superior to a single MSC dose and to hyaluronic acid 39 Charles HC, Kraus VB, Ainslie M, et al. Optimization of the fixed-­
in a controlled randomized phase I/II trial. Stem Cells Transl Med flexion knee radiograph. Osteoarthritis Cartilage 2007;15:1221–4.
2019;8:215–24. 40 Hochberg MC, Wohlreich M, Gaynor P, et al. Clinically relevant
18 Lamo-­Espinosa JM, Blanco JF, Sánchez M, et al. Phase II multicenter outcomes based on analysis of pooled data from 2 trials of
randomized controlled clinical trial on the efficacy of intra-­articular duloxetine in patients with knee osteoarthritis. J Rheumatol
injection of autologous bone marrow mesenchymal stem cells with 2012;39:352–8.
platelet rich plasma for the treatment of knee osteoarthritis. J Transl 41 Dougados M, Moore A, Yu S, et al. Evaluation of the patient
Med 2020;18:356. acceptable symptom state in a pooled analysis of two multicentre,
19 Emadedin M, Labibzadeh N, Liastani MG, et al. Intra-­articular randomised, double-­blind, placebo-­controlled studies evaluating
implantation of autologous bone marrow-­derived mesenchymal lumiracoxib and celecoxib in patients with osteoarthritis. Arthritis Res
stromal cells to treat knee osteoarthritis: a randomized, triple-­ Ther 2007;9:R11.
blind, placebo-­controlled phase 1/2 clinical trial. Cytotherapy 42 Eckstein F, Collins JE, Nevitt MC, et al. Brief report: cartilage
2018;20:1238–46. thickness change as an imaging biomarker of knee osteoarthritis
20 Mushahary D, Spittler A, Kasper C, et al. Isolation, cultivation, and progression: data from the foundation for the National Institutes of
characterization of human mesenchymal stem cells. Cytometry A health osteoarthritis biomarkers Consortium. Arthritis Rheumatol
2018;93:19–31. 2015;67:3184–9.
21 Yu SP, Hunter DJ. Intra-­articular therapies for osteoarthritis. Expert 43 Tubach F, Ravaud P, Beaton D, et al. Minimal clinically important
Opin Pharmacother 2016;17:2057–71. improvement and patient acceptable symptom state for subjective
22 Lopa S, Colombini A, Moretti M, et al. Injective mesenchymal stem outcome measures in rheumatic disorders. J Rheumatol
cell-­based treatments for knee osteoarthritis: from mechanisms of 2007;34:1188–93.
action to current clinical evidences. Knee Surg Sports Traumatol 44 Hamasaki T, Evans SR, Asakura K. Design, data monitoring, and
Arthrosc 2019;27:2003–20. analysis of clinical trials with co-­primary endpoints: a review. J
23 Ozay EI, Vijayaraghavan J, Gonzalez-­Perez G, et al. Cymerus™ Biopharm Stat 2018;28:28–51.
iPSC-­MSCs significantly prolong survival in a pre-­clinical, 45 Copay AG, Subach BR, Glassman SD, et al. Understanding the
humanized mouse model of graft-­vs-­host disease. Stem Cell Res minimum clinically important difference: a review of concepts and
2019;35:101401. methods. Spine J 2007;7:541–6.

14 Liu X, et al. BMJ Open 2021;11:e056382. doi:10.1136/bmjopen-2021-056382