Subcutaneous Injection of Adalimumab Trial compared with Control (SCIATiC): a randomised controlled trial of adalimumab injection compared with placebo for patients receiving physiotherapy treatment for sciatica

Article history

The research reported in this issue of the journal was funded by the HTA programme as project number 12/201/02. The contractual start date was in January 2015. The draft report began editorial review in February 2017 and was accepted for publication in May 2017. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The HTA editors and publisher have tried to ensure the accuracy of the authors’ report and would like to thank the reviewers for their constructive comments on the draft document. However, they do not accept liability for damages or losses arising from material published in this report.

Declared competing interests of authors

Nefyn H Williams declares membership of the Health Technology Assessment (HTA) Primary Care, Community and Preventative Interventions Panel. Dyfrig A Hughes reports other grants from National Institute for Health Research (NIHR) during the conduct of the study and membership of the HTA Clinical Trials Board from 2010 to 2016, the HTA Funding Teleconference from 2015 to 2016 and the Pharmaceuticals Panel from 2008 to 2012. Eifiona Wood reports other grants from NIHR during the conduct of the study. Nadine E Foster declares membership of the HTA Primary Care, Community and Preventative Interventions Panel. David A Walsh reports grants from Pfizer Ltd and personal fees from Novartis Pharmaceuticals UK Ltd outside the submitted work. Kika Konstantinou reports other grants from NIHR during the conduct of the study. Jaro Karppinen reports personal fees from lectures (Pfizer Ltd, Merck & Co., Inc. and ORION Pharma GmbH), personal fees from a scientific advisory board (Axsome Therapeutics Inc.) and personal fees from stocks (ORION Pharma GmbH) outside the submitted work. Stephane Genevay reports grants from AbbVie Inc., Merck & Co., Inc., Pfizer Ltd, University Hospitals of Geneva, the Rheumasearch Foundation, Fondation de bienfaisance Eugenio Litta and Centre de Recherches outside the submitted work.

Copyright statement

© Queen’s Printer and Controller of HMSO 2017. This work was produced by Williams et al. under the terms of a commissioning contract issued by the Secretary of State for Health. This issue may be freely reproduced for the purposes of private research and study and extracts (or indeed, the full report) may be included in professional journals provided that suitable acknowledgement is made and the reproduction is not associated with any form of advertising. Applications for commercial reproduction should be addressed to: NIHR Journals Library, National Institute for Health Research, Evaluation, Trials and Studies Coordinating Centre, Alpha House, University of Southampton Science Park, Southampton SO16 7NS, UK.

Trial objectives

1. To evaluate the clinical effectiveness of subcutaneous injections of adalimumab plus physiotherapy compared with a placebo injection of 0.9% sodium chloride plus physiotherapy for patients with sciatica who have failed first-line primary care treatment. Potential participants were planned to be identified during primary care consultation, after referral to musculoskeletal service or following a practice database search.

   The primary effectiveness outcome was sciatica-related health status using the ODI. 33 Secondary effectiveness outcomes included pain intensity, location, duration and anticipated trajectory; the risk of poor outcome; psychological measures, including fear-avoidance beliefs, self-efficacy, anxiety and depression; employment status; and adverse effects.

2. To evaluate the cost-effectiveness of subcutaneous injections of adalimumab plus physiotherapy compared with a placebo injection of 0.9% sodium chloride plus physiotherapy for patients with sciatica who have failed first-line primary care treatment from a health service and personal social care perspective. The primary economic outcome was the incremental cost per quality-adjusted life-year (QALY) gained. QALYs would be estimated by administering the EuroQol-5 Dimensions, 5-level version (EQ-5D-5L)34 at each follow-up visit.

Trial flow

The flow chart to show the different stages of the trial is presented in Figure 1. The flow chart to show the experience of the participant through the trial is in Figure 2.

FIGURE 1.

Consolidated Standards of Reporting Trials flow chart. MRI, magnetic resonance imaging; TB, tuberculosis.

FIGURE 2.

Participant flow chart. CXR, chest radiography; MRI, magnetic resonance imaging; TB, tuberculosis.

Design

Pragmatic, multicentre RCT with blinded participants, clinicians, outcome assessment and statistical analysis with concurrent economic evaluation and internal pilot.

Main centres

The RCT was planned to recruit from sites overseen by five collaborating centres in North Wales, London, Keele, Nottingham and Cardiff.

Selection and withdrawal of participants

Each collaborating centre oversaw one or more treatment sites that were delegated responsibility for delivering the interventions. Each treatment site had a number of patient identification centres, which consisted of general medical practices and local musculoskeletal services. The target population was adults with suspected sciatica for whom primary care treatment had failed. This was defined as troublesome symptoms (e.g. back and leg pain, pins and needles, numbness in leg, weakness), persisting for > 4 weeks and < 6 months. As the recruitment process took at least 4 weeks before participants were randomised, we did not have a lower time limit for duration of symptoms when identifying the target population and had an upper time limit of 20 weeks. These patients were identified in three ways:

1. by their GP

2. following a search of the general practice patient record database

3. after referral to local musculoskeletal services.

Telephone contact by the research physiotherapist

All those who had contacted the research team to state that they were interested in participating were sent a participant information sheet and were contacted by telephone by the research physiotherapist. The telephone call determined if they had unilateral leg pain and, if back pain was present, that leg pain intensity was worse than, or as bad as, the back pain. It also determined whether or not symptoms had persisted for > 20 weeks (to allow participants to be within the 6-month limit at randomisation). Finally, the telephone call allowed them to discuss any questions that they may have had about the study or their symptoms.

Research clinic

Those who satisfied the eligibility criteria were given an appointment slot in a research clinic run by research physiotherapists. At this research clinic all potential participants were assessed by the research physiotherapist for eligibility. Eligible participants who gave initial consent were registered and provided with a unique participant identification number. The following data were recorded on case report forms:

• demographic details such as age, sex, height and weight

• clinical findings such as pain location, pain duration, other presenting complaints, straight-leg raise test (left and right), femoral stretch test, muscle power, pinprick and light-touch sensation, quadriceps and Achilles tendon reflexes.

The research physiotherapist arranged for the participant to have the following blood tests taken by the phlebotomist to exclude haematological and biochemical abnormalities: full blood count, urea and electrolytes, estimated glomerular filtration rate, liver function test and glycosylated haemoglobin. The participant received TB screening in accordance with local practice and biological agents counselling. The research physiotherapist then arranged an appointment for MRI to exclude serious spinal pathology. All of these tests were completed within 2–3 weeks of the initial clinic visit and were recorded on the case report forms. The presence or absence of a disc prolapse on MRI was not to be used as an inclusion criterion, because the degree of disc displacement, nerve root enhancement or neural compression found on MRI does not correlate with sciatic symptoms. 14 MRI was reported by the local radiologist using a trial-specific standard operating procedure. The initial report stated whether or not the participant had serious spinal pathology that required a different treatment. A full MRI report was available only after completion of the study. Individual results were made available if a report was needed in an emergency, or if a spinal surgery referral was contemplated.

The research physiotherapist ensured that the participant received all the required tests. If there was any issue that required action, then the participant’s referring GP or musculoskeletal clinician was informed. When MRI had excluded serious spinal pathology, participants were contacted by the research physiotherapists either by telephone or post to attend the research clinic, where they received a second clinical assessment by the research physiotherapist, 2–3 weeks after their initial visit, to assess if they were still eligible. If they were still eligible, further consent was obtained for trial entry and randomisation when they were provided with a unique participant identification number. Table 1 shows the schedule of forms and procedures.

Inclusion criteria

• Aged ≥ 18 years.

• Clinical features of sciatica.

• Leg pain worse or as bad as back pain, elicited by asking the participant.

• Unilateral leg pain approximating a dermatomal distribution (contralateral buttock pain permitted if it did not succeed the inferior gluteal margin), obtained by asking the participant.

• One of the following:

  • positive neural tension test, such as straight-leg raise test restricted to < 50° by leg pain; positive femoral stretch test; muscle weakness or loss of tendon reflex affecting one myotome

  • loss of sensation in a dermatomal distribution.

• Persistent symptoms for at least 4 weeks and < 6 months despite first-line treatment in primary care, obtained by asking the participant.

• Moderate to high severity (score of ≥ 30 points on the ODI). 33

Female partners of sexually active male participants had to use adequate contraception for at least 5 months after the last injection. Female participants were required to have had a negative urine pregnancy test within 2 weeks prior to randomisation, unless they were post menopause or had been sterilised. Sexually active male partners of female participants were also required to use adequate contraceptive methods. The researcher ensured that the risks, and consequences, of not using adequate contraception were fully understood by the participants and provided information and pathways as deemed necessary.

Exclusion criteria

• Symptoms persisting for > 6 months (elicited by asking the participant).

• A previous episode of sciatica in the last 6 months.

• Unable to undergo MRI (e.g. magnetic metal implants, potential metallic intraocular foreign bodies, claustrophobia, extreme obesity), obtained from the medical records and by asking the participant.

• Serious spinal pathology (including cauda equina syndrome, malignancy, recent fracture, infection or very large disc prolapse), which might require an urgent spinal surgery opinion, identified from participants’ previous medical history in their medical records or from MRI.

• Incidental serious pathology identified by MRI (e.g. adrenal tumour).

• Neurological deficit involving muscle weakness requiring an urgent spinal surgery assessment (e.g. foot drop).

• Widespread pain throughout the body including the upper limb. 35 Pain was considered widespread when all of the following were present: pain in the left side of the body, pain in the right side of the body, pain above the waist and pain below the waist. Axial skeletal pain (cervical spine or anterior chest or thoracic spine or low back) had also to be present.

• Prior use of biological agents targeting TNF-α within the previous 6 months obtained from the medical records and by asking the participant.

• Previous lumbar spinal surgery, elicited from the medical records and by asking the participant.

• Contraindications to adalimumab injection including serious infection such as active or latent TB, transplanted organ, demyelinating disorders, malignancy, cardiac failure, low white blood cell count, pregnancy (determined from the medical records, the results of investigations and by asking the participant).

• Pregnant or breastfeeding (women must not breastfeed for at least 5 months after the last adalimumab injection).

• Unable to communicate in English or Welsh.

• Unable or unwilling to give informed consent.

Informed consent

At the initial physiotherapy clinic, the research physiotherapist determined preliminary eligibility, explained the nature of the trial and gave a repeat participant information sheet. The participant information sheet had been approved by the ethics committee and set out all of the key information, including the practicalities of the trial, the possible benefits and risks, and trial assessments. Participants were registered onto the trial and details were recorded on a database and a screening log at each of the trial treatment sites, and were assigned a unique participant identification number. Anonymised details labelled with the unique participant identification number were transferred to a separate database in the trials unit, which was used for recording all of the trial data. This ensured that the outcome measurement and statistical analysis would be performed blind to treatment allocation. All databases were password protected. The participant consent forms were stored in a locked filing cabinet in each treatment site. Clinical findings were recorded on case report forms.

Eligible participants who gave initial consent had blood tests to exclude haematological and biochemical abnormalities (full blood count, urea and electrolytes, estimated glomerular filtration rate, liver function test, glycated haemoglobin levels). They received TB screening, including plain chest radiography, biological agents counselling and MRI to exclude serious spinal pathology within 2–3 weeks of their initial visit. When MRI had excluded serious spinal pathology, and TB screening, a pregnancy test (in the case of eligible women) and biological agent counselling had been completed, participants attended a further appointment with the research physiotherapist. A second clinical assessment was performed and those who were still eligible were asked to complete a second informed consent form, approved by the ethics committee. In order to enter the RCT, the participant was randomised by the research physiotherapist using a remote web-based system. The treatment site sent a letter to each participant’s GP, informing them of their patient’s participation in the trial and requesting that the GP make a note of this in the patient’s record. In addition, GPs were asked to inform the trial team if they became aware that the participant had experienced an adverse event during the trial.

Three copies of the consent form were signed by the participant. The original was kept by the research team, one copy was kept by the participant and the third was filed in the participant’s hospital medical records. All participant information sheets, letters of invitation and consent forms were provided in Welsh and English in the two Welsh centres.

Magnetic resonance imaging

Participants who had given initial informed consent underwent MRI to exclude serious spinal pathology, but the presence or absence of a disc prolapse was not used as an inclusion criterion. The MRI scans were read and reported by a local radiologist, using a trial-specific standard operating procedure at each treatment site, who was independent of the trial team. Only results that showed serious pathology or suspected serious pathology were revealed to the research team, referring GP and the musculoskeletal clinician, who would then exclude the participant and refer for urgent assessment. Otherwise, the research team were informed that no serious spinal pathology was identified. The findings of MRI would be made available to the participant’s treating clinician only after completion of the study. Individual results were made available if a report was needed in an emergency, or if a spinal surgery or epidural injection referral was being contemplated, and were distributed to the clinical team, referring GP and the musculoskeletal clinician. The MRI findings were to be used in a planned a priori subgroup analysis. For clinical purposes, radiologists from each site provided a clinical report of the MRI. For the purpose of reporting standardised findings for research, two independent radiologists reported the MRI findings for all trial participants. The radiologists interpreted the report according to the MRI findings only.

Tuberculosis screening and biological agent counselling

The screening and counselling protocols used routinely by the rheumatology departments in each of the treatment sites were used and administered by an experienced rheumatology specialist nurse. All of the participating centres had access to either a specialist TB clinic or an infectious disease service, where any identified cases were referred and managed.

Second physiotherapy assessment

Participants attended a second appointment with the research physiotherapist 2–3 weeks after the initial appointment, after the MRI results had been reported and following TB screening and biological agent counselling. A second clinical assessment was performed and all the results of the tests performed were checked. If the participant remained eligible, a second consent form was completed. Participants completed a baseline questionnaire and were randomised using a remote web-based system. If they no longer fulfilled the criteria for trial entry, because their symptoms had improved at or below the 30-point threshold on the ODI, they were given advice about managing their remaining symptoms and discharged back to the care of their GP. Clinical findings were recorded on case report forms.

Registration

Once the first consent had been obtained, participants’ details were recorded on a database in the trial centre and each participant was assigned a unique participant identification number. Anonymised details labelled with the unique participant identification number were transferred to a separate database in the trials unit, which was used for recording all of the trial data. All databases were password protected. The participant consent forms were stored in a locked filing cabinet in each treatment site. Participants’ GPs were informed in writing about their participation in the trial.

Randomisation

After completion of the second consent form and once baseline outcome measures had been collected, participants were individually randomised. Randomisation to the Subcutaneous Injection of Adalimumab Trial compared with Control (SCIATiC) was achieved by secure web access to the remote randomisation system at the trials unit. This system was maintained and monitored independently of the trial statistician and any trial staff who needed to remain blind to the treatment allocation. In order to protect against subversion, while ensuring that the trial maintained good balance to the allocation ratio of 1 : 1 both within each stratification variable and across the trial, the randomisation was performed using a dynamic adaptive randomisation algorithm. 36 Participants were stratified by (1) treatment centre and (2) presence of neurological signs (motor weakness or sensory loss). The research physiotherapist who obtained informed consent requested the randomisation code from the web-based randomisation system, the result of which was e-mailed to the pharmacy and the rheumatology nurse, but not the research physiotherapist. The dispensing pharmacist logged and dispensed the appropriate injection in line with Medicines and Healthcare products Regulatory Agency (MHRA) guidelines. The injection was given on the day of randomisation; if this was not possible, a further appointment was arranged by the research physiotherapist so that the treatment could be given within 3 days from randomisation.

Withdrawal of participants

Withdrawal from the trial did not affect participants’ medical care, something that was emphasised in the participant information sheet. Failure to complete any one follow-up assessment did not constitute formal withdrawal from the trial, and, unless participants requested complete withdrawal of their data, data were used to impute values for the analysis. The imputation of missing values ensured that the data set was utilised to its full power. The full imputation details were prespecified as part of the statistical analysis plan.

Expected duration of trial

We planned to recruit participants over a 20-month period and to follow them up for 12 months.

Subcutaneous injections

All participants were randomised to receive two doses of subcutaneous injection, 2 weeks apart, into the posterior thigh. The intervention group received 80 mg of adalimumab followed by 40 mg,18 in order to achieve a therapeutic dose of adalimumab for a period of 4 weeks. The control group received an equivalent volume of 0.9% sodium chloride.

Injection process

The injections were prescribed by a consultant rheumatologist and administered by a rheumatology nurse experienced in the administration of these injections. The first injection was given on the same day as randomisation; if this was not possible, a further appointment was arranged by the research physiotherapist so that the treatment could be given within 3 days of randomisation. It was not possible to make the adalimumab and placebo syringes indistinguishable in appearance, nor was it possible to blind the pharmacy or the rheumatology nurse who administered the injections. Blinding of participants and the other clinicians was maintained using the following strategies. The rheumatologist wrote a prescription for ‘SCIATiC trial injection’ and was kept blind to treatment allocation. The research physiotherapist who obtained informed consent requested the randomisation code from a web-based randomisation system. The randomisation code was not sent to this physiotherapist, but was e-mailed to the pharmacy and the rheumatology nurse. The rheumatology nurse collected the injection from the pharmacy, which was transported in an undistinguishable box containing the adalimumab inside its original packaging or the 0.9% sodium chloride-containing ampoules. Communication between the participant and rheumatology nurse concerning the injection was kept to a minimum, and the rheumatology nurse administered the injections into the participant’s posterior thigh. The research physiotherapist was not present and did not communicate with the rheumatology nurse about the injection. In addition, in order to provide reassurance that other clinicians were not present, a log was kept of all people present in the room when each injection was administered. All research staff received full training on the blinding procedures. In order to assess whether or not blinding had been maintained, the participants were asked to complete a five-point Likert scale that asked if the participant considered the treatment to be:

1. definitely in the 0.9% sodium chloride injection group

2. more likely to be in the 0.9% sodium chloride injection group

3. equally likely to be in the 0.9% sodium chloride injection group or the adalimumab injection group

4. more likely to be in the adalimumab injection group

5. definitely in the adalimumab injection group.

Concurrent physiotherapy

Physiotherapy is usually considered normal practice for those participants who fail to improve with GP care alone. In this trial we aimed to investigate the clinical effectiveness of adalimumab in addition to physiotherapy. Current evidence on physiotherapy interventions for participants with sciatica indicates that specific exercise approaches (directional preference-based exercises or ‘McKenzie’ exercises based on certain spinal movements with or without manual therapy techniques) seem to relieve pain. 37 There was also evidence that exercise-based physiotherapy treatment added to GP care was beneficial. 38 Regimes including strengthening exercises of the lumbar and pelvic muscles also show some promise in terms of improvements in this group of participants. In this trial, both groups received a concurrent course of physiotherapy intervention that could be described as ‘best conservative care’. It was delivered in local physiotherapy departments by ‘treating’ physiotherapists and not by the ‘research’ physiotherapists who were carrying out the assessments of eligibility and randomisation. The physiotherapy intervention consisted of a package of directional preference (McKenzie), strengthening exercises or other exercises,37,38 and manipulative techniques that had been determined by consensus using a panel of extended scope physiotherapists. Treatments were intended to take into account and address participants’ individual needs, including clinical monitoring; appropriate advice and reassurance; assessment of psychosocial obstacles to recovery, such as excessive worrying or unhelpful beliefs about physical activity; and encouragement of appropriate, gradual return to full function, including work when applicable. The first session was expected to last approximately 45 minutes, with subsequent sessions lasting 30 minutes each. The therapy sessions were to be provided over a period of 12 weeks. The number of sessions provided would be determined by participant and therapist preference, and also response to treatment. We captured and described these aspects of physiotherapy treatment as part of the trial. The physiotherapy treatment started at the same time as the injection intervention in both arms of the trial. Participants were discouraged from receiving any other NHS-based co-intervention until this physiotherapy treatment had finished.

Clinical management of persistent symptoms

The protocol was designed such that, once the participants had completed their course of physiotherapy, and symptoms had settled or were improving, then no further intervention would be organised. They were to be discharged to the care of their GP and followed up by the research team. If troublesome symptoms persisted, then further treatment could be planned, as appropriate, by referral to musculoskeletal interface clinics or secondary care specialists according to local arrangement in each of the centres. The plans for further treatment were at the discretion of the treating clinicians and could include epidural corticosteroid injections or referral for disc surgery. The full result of MRI was to be made available if a spinal surgery referral was contemplated. All additional treatments were recorded in detail in a case report form.

Internal pilot trial

The pilot built on previous research in this participant group undertaken by team members,20,21,30,39 which had already provided information on trial administration, the characteristics of sciatica participants and the effects of biological treatments from previous studies. The internal pilot was designed to rehearse the procedures and logistics to be undertaken in the main trial. It was designed to assess the feasibility of the arrangements for delivering the interventions, recruitment rate and initial retention rate. The internal pilot would be based on the first 50 participants recruited into the trial. We planned to start recruitment in two centres (North Wales and London) and then to roll out recruitment in the other three centres over the following 3 months. We expected the recruitment rate to build up over the first 3 months to the target rate of four participants per collaborating centre per month. We anticipated that this would take 7 months. The indicative stopping criteria at the end of this internal pilot were recruitment, which failed to reach 80% of the planned recruitment rate target, dropouts up until the 6-week postal questionnaire assessment exceeding 20% or more than one centre failing to commence recruiting. Any procedural changes identified in the pilot would be implemented across all trial sites subject to ethics approval of the appropriate major amendment. Data from participants in the internal pilot would be automatically rolled into the main trial data unless the Trial Management Group (TMG) believed that data were incompatible with the remaining data. No interim analysis at the primary end point (1 year) was proposed and, therefore, this internal pilot would not affect the overall power of the trial. Wittes and Brittain’s40 method would be used for sample size recalculation if required.

Primary outcome

The primary clinical outcome was back pain-specific disability measured using the ODI33 at 12 months. The primary economic outcome was the incremental cost per QALY gained, estimated by administering the EQ-5D-5L34 at each follow-up visit.

Outcome measures

Follow-up

The questionnaires followed best practice in their design, to maximise response rate. The baseline questionnaire was administered by the research physiotherapists and completed by the participant. We planned to send follow-up postal questionnaires at 6 weeks, and at 6 and 12 months. Non-responders were to be sent an additional copy of the questionnaire. Persistent non-responders were to be contacted by telephone in order to collect a minimum data set. We planned to contact all participants by telephone 2 weeks after the 12-month questionnaire was sent. This would allow us to collect a minimum data set from non-responders, and to conduct a a brief semistructured interview with all participants asking about their overall experience of the trial and subsequent follow-up treatment. Blinding to treatment allocation would be maintained during these telephone interviews. Once again, in order to assess if blinding had been maintained, participants would be asked to complete a five-point Likert scale about which treatment group they believed that they were in.

Assessment of safety

As part of site initiation, training included an overview of possible side effects/potential adverse reactions associated with adalimumab.

Statistics

Trial management

Direct access to source data and documents

Source data were the hospital-written and NHS electronic medical records. Access to these data was through the participant’s clinicians, physiotherapist and research nurse. Trial-related monitoring, audits, REC reviews and regulatory compliance inspections were permitted, allowing access to data and documents when required.

Quality assurance

This trial was conducted in line with the trial protocol and followed the principles of good clinical practice outlined by the International Committee of Harmonisation Good Clinical Practice E6 (R1) Current Step 4 Version57 and complied with the European Union directive 2001/20/EC. 58

A monitoring plan was developed based on a trial risk assessment, which provided details of day-to-day quality control, audits, etc., and was delegated to members of the trial team to ensure that collected data adhered to the requirements of the protocol; only authorised persons completed case report forms; the potential for missing data was minimised; data were valid through validation checks (e.g. range and consistency checks); and recruitment rates, withdrawals and losses to follow-up were reviewed overall and by hospital site.

Data handling

The sources of data for the trial were as follows: recruitment details; baseline outcome measures captured electronically onto password-protected and encrypted computers by the research physiotherapists or research nurses; postal questionnaires at 6 weeks’ and at 6 and 12 months’ follow-up entered into the MACRO system (version 4; InferMed, London, UK); and telephone minimum data collection from non-responders captured on computers by researchers. Additional health service use data obtained from primary and secondary care records, with participants’ consent, would be recorded electronically on the computers. Each centre would input data into the MACRO data management program, which is a web-based system allowing controlled access to data by all centres and allows a full audit trail.

Trial sponsor

Bangor University (reference number 12/201/02; contact Dr Huw Roberts).

Ethics approval

Wales REC-3 granted approval on 27 May 2015 (15/WA/105). Clinical trial authorisation was approved from the MHRA on 15 April 2015 (21996/0002/001-0001).

Trial progress

Funding for the trial was approved by the HTA programme on 11 August 2014; the intention at that time was for the trial to open in January 2015 and close in June 2018. It had been planned that all the trial documentation for the regulatory approval for the trial would be completed between July 2014 and December 2014 so that regulatory approval could be obtained by January 2015.

Five collaborating sites planned to participate in the trial – North Wales, Cardiff, London, Keele and Nottingham – with training at the five sites taking place between February 2015 and April 2015. The initial plan was that the trial would be set up and recruiting participants at North Wales and London by April 2015, with the other sites opening to recruitment in June to October 2015. The trial documentation for the regulatory approval was not in place until December 2014. Regulatory approval was obtained from the MHRA on 15 April 2015 and from the REC on 27 May 2015. During this time the three English sites submitted requests for excess treatment costs (ETCs) to the NHS England for Clinical Commissioning Groups, ETCs for the two Welsh sites were agreed, and contracts were sent to both lead and collaborating sites from the sponsor, Bangor University. There were delays and unforeseen complexities in obtaining the ETCs for the English sites. Contracts also proved an issue and caused major delays because of difficulties with the delegation of the roles and responsibilities and what was required within the different contracts between university and university, and university and NHS sites. One of the sites withdrew from the trial in February 2016 because of concerns about the dosage of the biologic used and its patient population, which was higher than the standard dose used for patients with rheumatoid arthritis but similar to dosage in other conditions. This withdrawal led to a risk review for the other sites, which felt that, as no new data were available, the risk of infection for participants was acceptable, and they all agreed to continue participating in the trial.

The trial opened to recruitment on 8 December 2015 at North Wales and Nottingham, with Keele opening to recruitment on 11 August 2016. At the time of trial closure, contractual discussions were still ongoing between Bangor University, Cardiff University and Cardiff and Vale University Health Board.

During this time all sites dealt with a number of challenges. In North Wales, a research physiotherapist was seconded to the trial in September 2015 but, because of a shortage of physiotherapists in the department, was required by the health board to return to their previous employment and then left the post in February 2016 for personal reasons. As a result of physiotherapy staffing shortages within Betsi Cadwaladr University Health Board (BCUHB), the site was unable to employ a replacement research physiotherapist for the required research physiotherapist time. Participants were identified and recruited via the musculoskeletal clinic and physiotherapy clinics. The PI at the site contacted fellow consultants throughout the health board to ensure that all potential participants were identified. Only three participants were recruited at this site between March and July 2016.

Nottingham had intended to recruit its participants for the trial from the Sherwood Forest Hospitals NHS Foundation Trust Back Pain Unit diagnostics clinics only, based on pre-study clinic data. During the first 3 months after the start of recruitment at that site, no eligible participants were identified. The site PI extended screening to include orthopaedic clinics in addition to back pain unit clinics, but without any significant increase in recruitment. The PI at this site investigated whether or not there had been a change in referral pathways for people with sciatica in the region that might have affected referrals into these clinics. The PI confirmed, in February 2016, that this had been the case. GPs had been given direct access to MRI scans for sciatica, and were requesting and reviewing the results of MRI before requesting opinion or treatment from the back pain service. As well as introducing delay to referrals, the local GPs were tending to refer those with sciatica and a congruent disc prolapse on MRI to an alternative spinal surgical unit (e.g. Nottingham University Hospitals NHS Trust). A potential participant referred to the PI had already received MRI, which made them ineligible. Trial progress was an agenda item at the TMG. Owing to changes in pathways at Sherwood Forest Hospitals NHS Foundation Trust Back Pain Unit, it was agreed by the TMG that the protocol should be amended to include participants who had already undergone MRI. This was approved by the REC on 15 April 2016 and by the MHRA on 27 May 2016. During this time, Sherwood Forest Hospitals NHS Foundation Trust had also been in discussions with local primary care colleagues to arrange identification of potentially eligible participants from local GP practices through database searches or opportunistic referral. This was agreed and implemented in June 2016. Sherwood Forest Hospitals NHS Foundation Trust recruited five participants to the trial between February and September 2016.

During this period, Keele had obtained agreements for its ETCs and contract negotiations were finalised between Bangor University and Keele University, and between Bangor University, Keele University and Royal Wolverhampton NHS Trust, and also between Bangor University, Keele University and Staffordshire and Stoke-on-Trent Partnership Trust. The Royal Wolverhampton NHS Trust was opened to recruitment on 11 August 2016; one potential participant was identified prior to the trial closure on 20 September 2016. Randomisation of this participant was not permitted as a result of trial closure on 26 September 2016.

The TSC met on 11 January 2016. The problems with recruitment were discussed at the meeting, and the members were very sympathetic to the trial team’s frustration of the poor recruitment to the study. The TSC recommended the following steps if the funders allowed the trial to continue:

1. optimise recruitment for the two sites that were open

2. work harder to reduce bottlenecks (e.g. more MRI slots in North Wales)

3. increase the number of general practices searching for potential participants

4. approach the musculoskeletal triage clinics to join the study as sites

5. inform Keele and Cardiff at the next management group meeting that if they were not about to start recruiting then they would no longer be part of the trial

6. contact other possible sites that would be able to recruit to the study within the next 6 months.

The chief investigator contacted four sites one each in north, south and mid-Wales, and also the Royal Free Hospital in London.

Although the additional North Wales site was eager to participate and there was sufficient staff, there was a lack of clinical space to accommodate the trial within the rheumatology department. The rheumatology consultant submitted a case to the hospital managers with plans to increase the space available to carry out clinical trials. Unfortunately, this request was turned down because of competing demands on clinical space in the hospital, so it was not possible for the rheumatology department to be involved with SCIATiC and other clinical trials.

The hospital site in south Wales had expressed interest in participating in the trial in April 2016, and the site was arranging to accommodate the trial when we had to notify them that the trial was terminating early.

The hospital site in mid-Wales had also expressed interest in participating and had notified us on 17 August 2016 that its physiotherapy team had agreed to accommodate the trial. Unfortunately, we informed the site on 18 August 2016 about the discussions that we had with the funders and the expected termination of the trial.

The Royal Free Hospital had been discussing increasing their portfolio of clinical trials within the physiotherapy department and in April 2016 had expressed interest in participating. Unfortunately, it later informed us that it was doubtful if it would receive funding for the intervention arm of the trial. The hospital also felt that it would have difficulty recruiting participants with a symptom duration of < 6 months. In addition, its rheumatology department did not have spare capacity as it was already busy with existing research activities.

Owing to concerns with the slow recruitment, the chief investigator and Bangor trial team met with the funders on 28 January 2016, who informed the trial team that they should:

• finalise existing contracts immediately

• open sites that had not yet opened within the month

• start recruiting from all sites

• complete recruitment to the internal pilot of 50 participants by June 2016.

Another meeting with the funders was held on 16 August 2016. At the meeting the funders requested that the project team should submit closedown proposals as soon as possible and, ideally, by no later than 31 August 2016. Two different scenarios were proposed:

1. an immediate closedown of recruitment with submission of the project report by the end of December 2016

2. closure of recruitment in 6 months, until which time the study team should seek to establish the most effective recruitment routes for any future study, with submission of the project report by the end of June 2017.

From 16 August to 26 September 2016 only one participant was randomised to the trial by the Sherwood Forest Hospitals NHS Foundation Trust site. This information was relayed back to the funder, and on 26 September 2016 the chief investigator confirmed that the funders had asked for the trial to be closed immediately and the project report to be completed by the end of December 2016.

Trial timetable

The trial timetable is outline in Table 2.

Trial recruitment

Recruitment data for the trial are presented in Figure 3 and reasons for withdrawal or exclusion in Table 3. Sherwood Forest Hospitals NHS Foundation Trust and BCUHB recruited from December 2015 to September 2016. The Royal Wolverhampton NHS Trust recruited from August 2016 to September 2016. During this time, eight participants were randomised. No adverse events or adverse reactions were recorded for any of the participants.

FIGURE 3.

Participant flow diagram for Sherwood Forest Hospitals NHS Foundation Trust, BCUHB and the Royal Wolverhampton NHS Trust.

Excess treatment costs

Applications for ETCs were agreed by the Welsh Government by the two Welsh sites, but agreement for the three English sites was more problematic. On 4 February 2014, the trial chief investigator asked all participating sites to submit an application for ETCs. This was submitted by Sherwood Forest Hospitals NHS Foundation Trust in June 2014, and the ETCs were approved on 11 March 2015. Discussion had also taken place in Wolverhampton and an application submitted, but ETCs had been declined because of insufficient funds. This site subsequently explored other potential sources of funding (e.g. from primary and secondary care, or the pharmaceutical company Abbott UK, if they could provide a discounted or free drug). On 1 June 2015, the chief investigator requested details of how to obtain a subvention from the National Institute for Health Research. The chief investigator then contacted the Department of Health and requested assistance with the matter and was told that the subvention budget is not used to fund specific study sites and that, as sites for the trial are already signed up for ETCs, they would expect the trust to cover the costs. Keele led negotiations with both the trust and the Clinical Commissioning Groups in the West Midlands (Wolverhampton). Owing to anti-TNF-α falling outside payments by results tariff, both parties argued that they were not funded to support the ETCs associated with this trial. Following negotiation it was agreed that the local Clinical Commissioning Groups and the trust (charitable funds) would fund an equal split of the ETCs. ETCs were approved for Royal Wolverhampton NHS Trust on 19 August 2015 and provisional ETCs agreed for Bart’s Health NHS Trust; the trust was told that it would be finalised once R&D approval was given.

Outcome measure results

The original intention was to use a linear mixed-model approach to assess the effects of time, group and time × group. This was not possible as data were only available for eight randomised participants on trial closure. The data for the eight participants are presented. The demographic information is presented in Appendix 1. Results from all the measures that were collected at different time points are reported in Appendix 2, with the pain manikin drawings presented separately in Appendix 3. Resource use results are presented in Appendix 4 and concomitant medications in Appendix 5. Finally, the physiotherapy information is presented in Appendix 6 and concomitant medications in Appendix 7. It is not possible to make any conclusions from these data as no analysis was performed.

Patient and public involvement

The patient and public involvement representative contributed to trial design by commenting on the trial protocol and patient-facing documents, as well as participating in the TSC meetings. Unfortunately, we lost contact with the patient and public involvement representative when trial recruitment was closed.

Comments and feedback from the trial management team and research teams

The trial management team and all sites were asked to reflect on the trial about what worked, and what did not. The following was noted.

Summary of main findings

We have demonstrated that a RCT of adalimumab for persistent sciatica of 3–6 months’ duration is acceptable to some participants. The trial methods were feasible in terms of randomisation method and outcome measurement. However, recruitment rates were lower than expected and several other factors contributed to trial termination prior to pilot study completion. We were therefore unable to demonstrate that this trial was feasible. There was a long delay agreeing and exchanging subcontracts with participating centres and sites. The contracting discussions with one site were never concluded and another centre dropped out of the study before recruitment started. There were delays negotiating ETCs for the English centres and sites. We did attempt to recruit additional sites in Wales, but negotiations were not concluded at the time of trial closure. Delays in the contracting process also caused delays in recruiting research staff, in particular the research physiotherapists. In the two sites that did open in time to recruit participants, recruitment was slow. In one site, the sciatica management pathway changed around the time that it opened to recruitment. This site initially only relied on referrals to its secondary care musculoskeletal service, but later involved the primary care research network, which was starting to identify participants just before trial closure. In the other site there were operational issues identifying the research physiotherapist resource. This site relied on retrospective GP record review to identify potential participants and although large numbers of invitations were sent to potential participants, there was a low rate of uptake with only a small proportion seen for an initial assessment, and entered into the trial. We were in the process of increasing the number of GP practices within the North Wales area to assist with increasing recruitment; this was not concluded prior to the trial closure. Recruitment was improving just before the trial was shut down with five potential participants ready to be recruited within the following month.

Strengths and weaknesses

The trial methods, in terms of randomisation method and outcome measurement, worked smoothly but for only eight participants, so it is not possible to claim that the methods were feasible. An internal pilot study was planned, but unfortunately we were unable to recruit sufficient numbers.

We had modelled the numbers of eligible participants for our recruitment projections on the ATLAS cohort study. 39 However, we made unrealistic assumptions about the numbers of identified participants who would be willing to participate in a clinical trial of an investigational medicinal product. Although we identified large numbers of potential participants, only small numbers returned reply slips indicating a willingness to participate. It was not known why eligible participants did not wish to participate. Presumably, some found the trial procedures too burdensome, such as the complex two-stage recruitment process, whereas others did not want to participate in a RCT, especially in a clinical trial of an investigational medicinal product involving a medication with known potential side effects. The ATLAS study cohort was led by the team at Keele University and recruited 610 participants in 23 months. 60 The two clinical sites supported by Keele signed their contracts just prior to trial closure. Although potential participants had started to be identified, there was insufficient time to recruit them. Because of this, it was not possible to compare rates of recruitment into the RCT with those found in the ATLAS cohort study, nor the more recent HTA-funded Sciatica Outcomes in Primary Care trial using similar methods of identifying participants. One site in east London dropped out of the trial just before recruitment began because the PI had concerns about patient safety. The PIs in the other sites reviewed the safety risks to potential participants and felt that these risks were justified and that measures were in place to mitigate, detect and address any adverse effects. Although there were only short delays in obtaining research permissions and finalising the trial contract with the funder, there were very long delays negotiating subcontracts with three of the sites. The difficulty of achieving clarity of delegation of functions, and clarity about what needed to be in the different contracts (university to university vs. university to NHS trusts) are key learning experiences from this trial. Negotiations concerning ETCs in England were protracted and complex, which added to the delay negotiating contracts and setting up sites.

Comparison with previous literature

The previous systematic review of biological agents for sciatica found a small number of RCTs and other studies with small numbers of participants recruited. 21 Many of these studies also had poor rates of recruitment, especially in the UK NHS. 24

Implications for future research

A number of factors contributed to the lack of recruitment to the trial. There were delays in contracting, the process for identifying and recruiting participants was inefficient, there were delays in site set-up and a lack of investigator engagement (possibly because of a lack of equipoise). After the London centre withdrew, we asked the other PIs whether or not the research question was still in equipoise. They all agreed that the risk of infection from the dose of drug administered in this trial was acceptable, and that they were still in equipoise.

In order to reduce delays in the contracting process, there needs to be an early agreement about what the contracts need to include, and how the contracting process should be arranged, so that the academic partners and clinical sites have a clear understanding of their delegated roles and tasks. Early discussions about site requirements, perhaps using a site feasibility questionnaire, early dialogue with sites’ R&D departments and the early appointment of research staff in each site would facilitate trial set-up.

We are unable to make any recommendations for future practice in this area because of a lack of trial results. Without any results from the internal pilot study, it is difficult to make recommendations for future research in this area. It may be that there is insufficient equipoise around the question of adalimumab for sciatica among patients and some clinicians. However, this would need to be addressed in further qualitative research. The two-stage recruitment process was complicated and not feasible. We had modelled the number of potentially eligible patients on results from the ATLAS cohort study. However, ATLAS did not use the same two-stage process, nor did it rely on retrospective searches of GP records, but rather relied on the identification and invitation of eligible patients who were currently consulting their GP for sciatica. The same recruitment process for ATLAS, which relies on GPs entering relevant diagnostic Read codes into their computer systems, triggering a ‘pop-up’ reminder about sending eligible patients to dedicated clinics for further assessment and eligibility checking, has worked in the Sciatica Outcomes in Primary Care (SCOPiC) trial. 61 Similar recruitment methods were going to be used in the two sites supported by the Keele co-applicants. Unfortunately, because of delays in finalising contracts and setting up sites there was insufficient time to recruit any participants and to test these recruitment methods before trial closure. So, it is still not possible to say whether or not the use of the recruitment method used in the SCOPiC trial would be feasible for a similar RCT in the NHS testing an investigational medicinal product.

A trial of biological therapy in patients with sciatica still needs to be done, but would require a clearer contracting process, qualitative research to ensure that patients would be willing to participate and simpler recruitment methods.

Contributions of authors

Nefyn H Williams (Clinical Senior Lecturer in General Practice) was the chief investigator and grant holder, was responsible for study design, conduct and analysis, led the writing of Chapters 1, 2 and 4, contributed to all other chapters, led the discussions of the implications of study findings and had overall responsibility for the study and final report.

Alison Jenkins (Trial Manager) was responsible for overseeing day-to-day conduct, contributed to all chapters of the report and contributed to the discussion of the implications of study findings.

Nia Goulden (Trial Statistician) conducted the statistical analysis for the trial, led the writing of Chapter 3 and gave feedback on other chapters of the report.

Zoe Hoare (Principal Trial Statistician) gave input to study design, was responsible for the statistical analysis design, provided methodological oversight and support for the trial statistician, and contributed to the discussion of the implications of study findings.

Dyfrig A Hughes (Professor of Health Economics) was a co-investigator, contributed to the study design, was responsible for the economic evaluation, and commented on all chapters of the final report and to the discussion of the implications of study findings.

Eifiona Wood (Senior Research Fellow in Pharmacoeconomics) was the trial health economist, commented on all chapters of the final report and contributed to the discussion of the implications of study findings.

Nadine E Foster (National Institute for Health Research Professor of Musculoskeletal Health in Primary Care) was a co-investigator, contributed to methodology and study design, commented on all chapters of the final report and contributed to discussion of the implications of study findings.

David A Walsh (Professor of Rheumatology) was a co-investigator, was responsible for study design, provided methodological oversight throughout the study, commented on all chapters of the final report and contributed to the discussion of the implications of study findings.

Dawn Carnes (Senior Lecturer in Musculoskeletal Health) was a co-investigator, and contributed to the methodology and study design.

Valerie Sparkes (Reader in Arthritis Research, Director of Impact and Innovation) was a co-investigator, contributed to the methodology and study design, and provided physiotherapy expertise.

Elaine M Hay (Professor of Community Rheumatology) contributed to the methodology and study design.

John Isaacs (Professor of Clinical Rheumatology) gave input to the study design.

Kika Konstantinou (Spinal Physiotherapy Specialist) was a co-investigator, contributed to methodology and study design, provided physiotherapy expertise, commented on all chapters of the final report and contributed to the discussion of the implications of study findings.

Dylan Morrissey (Consultant Physiotherapist, Clinical Reader) was a co-investigator, contributed to the methodology and study design, and provided physiotherapy expertise.

Jaro Karppinen (Professor of Physical and Rehabilitation Medicine) contributed to the methodology and study design.

Stephane Genevay (Head of the Multidisciplinary Back Pain Clinic) contributed to the methodology and study design.

Clare Wilkinson (Deputy Head of Research, School of Healthcare Sciences) was a co-investigator, provided feedback on study protocol and contributed to the discussion of the implications of study findings.

Data sharing statement

All available data are included as appendices or can be obtained from the corresponding author.

Disclaimers

This report presents independent research funded by the National Institute for Health Research (NIHR). The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, NETSCC, the HTA programme or the Department of Health. If there are verbatim quotations included in this publication the views and opinions expressed by the interviewees are those of the interviewees and do not necessarily reflect those of the authors, those of the NHS, the NIHR, NETSCC, the HTA programme or the Department of Health.

Five sites scheduled to participate in SCIATiC

© Queen’s Printer and Controller of HMSO 2017. This work was produced by Williams et al. under the terms of a commissioning contract issued by the Secretary of State for Health. This issue may be freely reproduced for the purposes of private research and study and extracts (or indeed, the full report) may be included in professional journals provided that suitable acknowledgement is made and the reproduction is not associated with any form of advertising. Applications for commercial reproduction should be addressed to: NIHR Journals Library, National Institute for Health Research, Evaluation, Trials and Studies Coordinating Centre, Alpha House, University of Southampton Science Park, Southampton SO16 7NS, UK. UK (English) © 2009 EuroQol Group EQ-5D™ is a trade mark of the EuroQol GroupWales (Weish) © 2013 EuroQol Group EQ-5D™ is a trade mark of the EuroQol Group