LIFE AND DEATH OF SPARE (SELECTIVE BLADDER PRESERVATION AGAINST RADICAL EXCISION): REFLECTIONS ON WHY THE SPARE TRIAL CLOSED
Robert A Huddart, Department of Urology, Royal Marsden Hospital, Sutton, Surrey, UK.e-mail: Robert.firstname.lastname@example.org
Trial Steering Committee
Selective bladder Preservation Against Radical Excision (cystectomy) in muscle invasive T2/T3 transitional cell carcinoma of the bladder
muscle-invasive bladder cancer
Prostate testing for cancer and Treatment (trial)
Research and Development
On 26 January 2010 the independent Trial Steering Committee (TSC) met and recommended that the SPARE [Selective bladder Preservation Against Radical Excision (cystectomy) in muscle invasive T2/T3 transitional cell carcinoma of the bladder] trial closed to recruitment due to poor accrual. The SPARE trial randomised patients with muscle-invasive bladder cancer (MIBC) to selective bladder preservation vs cystectomy. It was a multicentre feasibility pilot study set up to ascertain if a large phase III trial designed to determine the standard of care in these patients was possible. Unfortunately, recruitment rates were not sufficient for the trial to progress to a phase III study with an endpoint of overall survival. This is a huge disappointment to the UK (and international) urological oncology community.
During the set up phase of SPARE, urologists and oncologists were widely consulted, with extremely positive feedback. The pilot study was selected for the ‘Best of BAUS’ and the President of BAUS endorsed the study as one of the most important clinical questions facing UK urology. Members of the Trial Management Group visited Cancer Research Networks across the UK and a significant amount of effort was invested in the engagement of the wider urological community. The Prostate testing for cancer and Treatment (ProtecT) trial had shown that randomizing between radically differing treatment methods was possible in the UK . Patient-focus groups were supportive, and encouraging recruitment estimates were obtained from investigators. It was therefore with some optimism that the trial was launched in July 2007. Unfortunately, 30 months later, we had recruited just 45 patients and, despite intensive work with investigators, recruitment showed no clear signs of improving to levels that would suggest a large phase III trial would be feasible. In the circumstances, the TSC had no choice but to recommend closure.
Why should this be?
We believe that there are several reasons why recruitment to SPARE did not reach anticipated levels and suggest that the urological and wider oncology community may wish to learn from our experiences.
COMPLEXITY OF THE PATIENT PATHWAY
It soon became apparent that the patients’ treatment pathways were complex and varied significantly across networks. This trial involved three potential treatments: neoadjuvant chemotherapy, radiotherapy and surgery. This necessitated three and often four or more consultants being involved in the patients’ care (local urologist who makes the diagnosis, central urologist who would do the cystectomy, clinical oncologist for chemotherapy and/or radiotherapy and sometimes a medical oncologist for the chemotherapy) plus their junior teams, clinical nurse specialists and research nurses.
This adds a level of bureaucracy to obtaining trial approvals (see below) and makes opening the study vulnerable to individual prejudices. It also allows the possibility that, once the trial is open and despite the existence of a multi-disciplinary team, these individual prejudices (either overt or subliminal) for one treatment or another can have a negative impact on patient recruitment. Likewise, ignorance of the protocol, confused communication between clinical staff involved in the patient pathway (e.g. in determining when and how potential patients should be identified and who should initially approach them) and between clinicians and patients, and information overload for potential participants can all affect recruitment [2–4].
Within a few months of opening the trial the complex post-EU Clinical Trials Directive era was creating challenges. It became apparent that the time taken from expressing interest in the trial to a centre being ready to recruit patients was excessively long. In all, 36 centres opened to recruitment during SPARE and the mean (range) time taken from these centres gaining the first approval of the Site-Specific Information Form to opening was 8.6 (1.6–18.1) months. Although there was no single cause, most of the delays seem to involve delays within the Trust Research and Development (R&D) departments, even though treatments in the protocol represent standard therapy and SPARE was not a trial of an investigational medicinal product. R&D departments often took a long time to authorize the standard clinical trial agreement between the Trust and trial sponsor. Capacity issues at Radiotherapy Physics Departments delayed opening in some centres even though the protocol used standard radiotherapy techniques with a very light quality assurance exercise. In addition, ensuring that Principal Investigators had adequate Good Clinical Practice training caused delays in some cases. These problems were compounded by the necessity for all centres potentially involved in delivering trial treatment to a participant to have the necessary ethics and R&D approvals in place before that patient could be recruited. This usually meant that a network as a whole had to open; a local cancer unit to recruit patients, a centre to undertake cystectomy and often a separate centre to deliver radiotherapy . This had several effects. Networks were held back by the slowest centre to open; some centres that prepared all their paperwork quickly had to remain unopened whilst their associated treatment centres progressed paperwork far more slowly. Even more apparent was that this complexity could mean that recruitment was vulnerable to lack of support from one part of the research team. In several networks, despite enthusiastic support of other team members, a single key individual failed to support the trial because they ‘knew’ that their treatment was ‘best’. This was especially disappointing when previously support had been expressed.
It was always going to be a challenge to counsel patients into accepting randomization. The potential treatment strategies within SPARE are very different and it is perhaps not surprising that many patients were not randomized because they wished to receive specific treatment. The strategy of selective bladder preservation (radiotherapy if adequate response to neoadjuvant chemotherapy was observed on cystoscopy) and cystectomy otherwise, was a novel one for most patients, and some found it a difficult concept to grasp. One fact of interest is that despite cystectomy being considered by many clinicians, at least historically, as the ‘gold-standard’ more patients declined the study because they preferred radiotherapy. To address these issues a qualitative patient interview study was embedded into the trial from its inception. This study documented patients’ experience, and explored reasons why people said ‘yes’ or ‘no’ to being recruited. It also looked at patients’ perceptions of the trial recruitment process. It identified information overload, confused messages and issues of equipoise as potential recruitment barriers [2,3]. Additional advice for investigators and a revised Patient Information Sheet were designed to reflect these early experiences but these had no demonstrable effect on actual recruitment rates. A second qualitative study was launched 2 years into recruitment to investigate investigators’ perceptions of SPARE and clinician–patient communication. This led to a further iteration of advice, a re-worked Patient Information Sheet and investigator training. Despite rapid dissemination and positive feedback, these measures came too late to save the trial.
LACK OF PATIENTS
Over a 30-month period, screening logs collected from open centres provided details on >800 patients in the target population with T2/T3 N0M0 MIBC, but <250 potentially eligible patients were identified. About one fifth of eligible patients were not approached about the trial, and, of those that were, most opted for radiotherapy. A large minority of patients who declined participation did so because they did not wish to have their treatment assigned by randomization. Further, it became apparent over the course of the trial that, although we felt that we identified many patients with MIBC, the numbers eligible to receive neoadjuvant cisplatin-based chemotherapy and in whom cystectomy and selective bladder preservation were both valid options are not as frequent as we had thought. Common reasons for ineligibility included renal obstruction, widespread carcinoma in situ and, anecdotally, with the increasing use of MRI to stage patients, rising numbers of patients with radiologically equivocal pelvic lymph nodes. A lack of eligible patients was perhaps the fundamental reason why the proposed phase III trial was not feasible. A larger patient pool would make low rates of randomization a problem that could be overcome, allowing investigators to find their own way to successfully recruit as they gained experience with patients. However, the combination of low randomization rates and a restricted patient pool inevitably leads to problems.
We believe that SPARE was a well-designed protocol, generously supported by Cancer Research UK. Considerable efforts were undertaken before the trial opened to ensure that it was supported by urologists, oncologists and by the patients themselves. The failure to recruit patients to this protocol should lead the Urological Oncology community to ask themselves some hard questions. The research infrastructure developed by the National Cancer Research Network is a vast improvement on the structure before, and is probably amongst the best internationally, but the complexities of research governance and the way our cancer services are set up remain significant barriers to successful trial delivery when more than one clinical speciality is involved.
Do we want to do research into the hard questions? The experiences from ProSTART (comparing active surveillance with radical intervention options in localised prostate cancer), SPARE and to some extent RADICALS (a trial looking at radiotherapy and hormone therapy after surgery for prostate cancer) amongst others suggest that too many of our colleagues are involved in service delivery and show little real openness to the idea that there is uncertainty in medicine, and thus show restricted understanding of or willingness to participate in research that challenges their prejudices. However, the success of ProtecT remains an exception, this success was achieved by providing intensive training and funding key personnel in a few high-volume centres. Such a model is unlikely to be practical for most trials where patient numbers are much smaller and numbers of participating centres much larger.
To avoid future studies having a similar fate to SPARE we think that any new trials involving complex interventions in general and surgery in particular, need to
- (i) incorporate carefully built-in training and an education programme, ideally through expert-led qualitative research, at the commencement of the study 
- (ii) identify, at a network level, the process by which sites, investigators and patients will ‘interact’ with the trial
- (iii) ensure that all investigators involved are truly committed to the scientific process.
Our experiences with SPARE have highlighted the challenges faced when undertaking trials of complex interventions in the UK Oncology setting in general and in urological cancer specifically. We hope these reflections will help support the development of future research initiatives. For SPARE, despite the disappointments, we hope in the near future to report detailed outcomes from the in-built qualitative studies as well as the clinical outcomes from what remains the largest randomized trial of surgery vs conservative therapy in bladder cancer.
CONFLICT OF INTEREST