Infliximab versus ciclosporin for steroid-resistant acute severe ulcerative colitis (CONSTRUCT): a mixed methods, open-label, pragmatic randomised trial

Summary Background Infliximab and ciclosporin are of similar efficacy in treating acute severe ulcerative colitis, but there has been no comparative evaluation of their relative clinical effectiveness and cost-effectiveness. Methods In this mixed methods, open-label, pragmatic randomised trial, we recruited consenting patients aged 18 years or older at 52 district general and teaching hospitals in England, Scotland, and Wales who had been admitted, unscheduled, with severe ulcerative colitis and failed to respond to intravenous hydrocortisone within about 5 days. Patients were randomly allocated (1:1) to receive either infliximab (5 mg/kg intravenous infusion given over 2 h at baseline, and again at 2 weeks and 6 weeks after the first infusion) or ciclosporin (2 mg/kg per day by continuous infusion for up to 7 days, followed by twice-daily tablets delivering 5·5 mg/kg per day for 12 weeks). Randomisation used a web-based password-protected site, with a dynamic algorithm to generate allocations on request, thus protecting against investigator preference or other subversion, while ensuring that each trial group was balanced by centre, which was the only stratification used. Local investigators and participants were aware of the treatment allocated, but the chief investigator and analysts were masked. Analysis was by treatment allocated. The primary outcome was quality-adjusted survival—ie, the area under the curve (AUC) of scores from the Crohn's and Ulcerative Colitis Questionnaire (CUCQ) completed by participants at baseline, 3 months, and 6 months, then every 6 months from 1 year to 3 years. This trial is registered with the ISRCTN Registry, number ISRCTN22663589. Findings Between June 17, 2010, and Feb 26, 2013, 270 patients were recruited. 135 patients were allocated to the infliximab group and 135 to the ciclosporin group. 121 (90%) patients in each group were included in the analysis of the primary outcome. There was no significant difference between groups in quality-adjusted survival (mean AUC 564·0 [SD 241·9] in the infliximab group vs 587·0 [226·2] in the ciclosporin group; mean adjusted difference 7·9 [95% CI −22·0 to 37·8]; p=0·603). Likewise, there were no significant differences between groups in the secondary outcomes of CUCQ scores, EQ-5D, or SF-6D scores; frequency of colectomy (55 [41%] of 135 patients in the infliximab group vs 65 [48%] of 135 patients in the ciclosporin group; p=0·223); or mean time to colectomy (811 [95% CI 707–912] days in the infliximab group vs 744 [638–850] days in the ciclosporin group; p=0·251). There were no differences in serious adverse reactions (16 reactions in 14 participants receiving infliximab vs ten in nine patients receiving ciclosporin); serious adverse events (21 in 16 patients vs 25 in 17 patients); or deaths (three in the infliximab group vs none in the ciclosporin group). Interpretation There was no significant difference between ciclosporin and infliximab in clinical effectiveness. Funding NIHR Health Technology Assessment programme.


Views of Patients and Professionals
The following is a summary of the qualitative elements of the CONSTRUCT trial. They are reported in depth elsewhere. 1

Aim
The qualitative components of the CONSTRUCT trial used interviews to explore the views of participating patients, doctors and nurses about severe ulcerative colitis and its management. The aims were to clarify participants' feelings about their condition, their perceptions of treatment with infliximab, ciclosporin or surgery and to understand changes to these views over time. Interviews with doctors and nurses aimed to understand their views about the efficacy, safety and administration of the trial drugs, shared decision-making with patients, and their responses to their patients' ongoing illness experience.

Method
Trial participants who indicated their willingness to be interviewed when giving consent were chosen according to a purposive quota sampling framework. This identified 12 consenting patients from each arm of the trial, who were interviewed twice, two to three, and 8-12 months after randomisation. All interviews followed a semistructured format to ensure consistency of data collection whilst enabling patients to respond to prompts if they so wished, to ensure comprehensive and rich data capture. The first interviews investigated patients' priorities for their health and well-being, and the administration, side effects and response to the treatment they received. The second interviews used a similar schedule, but included additional questions examining what had happened to them following treatment, including changes over time in their opinions of treatment, their interactions with healthcare professionals, and their current health. Patient interviews were undertaken face-to-face or by telephone depending on patient preference.
Principal investigators and nurses responsible for administering and monitoring the trial drugs were sampled purposively from trial sites based on recruitment rates to the cohort and trial. They were approached by telephone or email, and gave informed consent. All interviews were semi-structured and undertaken by telephone.
All interviews were recorded and transcribed. Patient and professional data were analysed using schema and thematic analysis frameworks that were refined over time by researchers with expertise in qualitative data. Transcripts were coded to reveal major and minor themes and categories and were also schematised to disclose succinct, multi-disciplinary overviews of key issues arising. [2][3][4] Resultsinterviews with participants Thirty-five interviews were undertaken with 20 participants, 15 of whom were interviewed on two occasions. The participants were split evenly between the infliximab and ciclosporin arms of the trial, and were representative of the main study population in all baseline characteristics. Three patients in each group had undergone a colectomy since entering the trial.
The main findings from the patient interviews were that:  Participants who had received infliximab appreciated the positive outcome from this treatment.  Those who received infliximab appeared to speak more positively about their treatment than those who received ciclosporin.  The debilitating symptoms of ulcerative colitis impact not only on their own quality of life, but also on their relationship with family and friends.  Many participants expressed their desire to return to a 'normal' quality of life and many of those who had a colectomy, whist initially adamant that they did not wish to 'lose' their colon, found relief from surgery and felt they could move on with their lives.  Patients came to terms with having to live with the ongoing unpredictability of symptoms and treatments yet they recognised that this unpredictability makes it particularly difficult for them and for the healthcare professionals treating them to manage their health.  Ulcerative colitis is considered an embarrassing condition which makes it an isolating and awkward experience for patients due to its impact on life and work.  Lack of visibility of either symptoms or outcomes, also affected patients' willingness to share knowledge of the disease with others.  Surgery was feared, but once a colectomy had been undertaken, most participants experienced relief and recognised the health benefits.
 Participants wanted to know more about the cause of ulcerative colitis and its links with stress and diet, and would have welcomed greater information provision.  Ready access to an Inflammatory Bowel Disease Specialist Nurse was suggested as particularly important for members of this patient group.

Resultsinterviews with professionals
Twenty-three interviews were undertaken with 15 principal investigators, stratified by the number of patients recruited per site, and eight senior nurses from sites that recruited well to the trial.
The main findings that emerged from these interviews were that  Healthcare professionals make judgements about the two drugs largely based on their own personal experience of prescribing or giving them to patients.  The views of nurses are influenced by the drug therapy choices of their units, the method of administration, and perceptions of effectiveness and adverse side effects.  A clear preference for infliximab amongst nurses was based predominantly on the ease of administration of a short infusion of infliximab, when compared with the continuous 24 hour infusion required for ciclosporin.  Some doctors strongly favoured infliximab, wishing to see it as the drug of choice in view of its perceived ability to manage the many complex symptoms of the disease, ease of administration, fewer side effects, and greater effectiveness.  Other doctors favoured ciclosporin, perceiving it as safe, effective and cheap.  Most doctors were unsure which drug to use in the future, and were prepared to wait for further evidence of effectiveness and safety before fully making up their minds.  Doctors questioned guidelines relating to prescribing these drugs, and the restrictions this placed on personal autonomy in delivering best patient care.

Discussion
The number of subjects interviewed in this study was small, but they were representative of the trial participants and sites. The findings include a strong preference from nurses for infliximab, based largely on a dislike of the infusion requirements for ciclosporin. Although doctors were in equipoise with regard to the trial, most but not all of those interviewed wished to see infliximab as the recommended drug of choice in the future. Patients who received infliximab tended to speak more positively their treatment than those given ciclosporin, and those who required surgery were positive about colectomy, having been fearful pre-operatively. The very debilitating and embarrassing impact of the disease on health, work and social life was emphasised by many patients.