Development of an IDEAL Framework to Standardise Cytoreductive Surgery for Colorectal Peritoneal Metastases

Overview

Cytoreduction surgery (CRS) followed by hyperthermic intra-operative peritoneal chemotherapy (HIPEC) is a relatively new treatment for selected patients with peritoneal metastases of colorectal origin (PMCR). Data from outside of trials suggest that CRS and HIPEC improves survival compared with the current standard care (chemotherapy). The big challenge is to do trials in this setting - as the intervention is complex, and there are wide variations in the process and recording of outcomes. If trials can confirm the findings from non-randomised studies there are an estimated 1000 to 2000 patients who may benefit from this intervention in the UK each year. The aim of this study is to develop a framework which can be used to undertake a randomised trial in patients with PMCR suitable for CRS with or without HIPEC. The investigators will address this using the principles of the IDEAL (Idea, Development, Evaluation, Assessment \& Long term study) framework. Here, a pre-trial feasibility study will be performed between the two national peritoneal tumour treatment centres (Manchester and Basingstoke). This study is designed as such that it will take place over the following four stages: Stage 1. Comparing the treatment data from 100 operations from each of the two centres to identify which of the key components of the intervention differ as well as testing for differences in overall survival and recurrence free survival. Stage 2. Identifying sources of these differences by selecting up to 25 patients and investigating the variation in the way surgeons score key aspects of the procedure Stage 3. Development of a 'trial manual' with standardised definitions (to minimise any differences) Stage 4. Test how well people follow the manual in practice. After this study is complete, it will be possible to use the resulting trial manual to design future randomised trials to test the most suitable clinical question.

The 2010 National Institute Centre for Excellence (NICE) IGP331 guidance recognised that there is sufficient evidence (mainly from non-randomised comparative studies) that cytoreduction surgery (CRS) followed by hyperthermic intraoperative peritoneal chemotherapy (HIPEC), in selected patients with peritoneal metastases of colorectal origin (PMCR), improves survival compared with systemic chemotherapy, current standard care. The guidance recognised that this intervention is associated with considerable morbidity and mortality, and accordingly, should be delivered with special arrangements for clinical governance, consent and research, and include trials. In the UK, with 37,000 new colorectal cancers per annum, there are an estimated 1000 to 2000 patients who may benefit from this intervention. There are currently two treatment centres (Manchester and Basingstoke), with additional new centres in Birmingham and Dublin starting to roll out to meet this unmet need. There is only one published randomised trial in this field, which is now mainly historic (2003). There are a small number of on-going trials, but these are small-scale and have encountered various early problems, which can be summarised as follows: (i) Wide variation in the treatment process, such that there has been between-arm imbalances in key prognostic components (such as different case mix between arms); and (ii) Wide variation of the reported treatment-related outcomes, such that it is unclear whether results are generalizable to other centres. There are potentially two broad sources of this variability : (i) Wide variation in clinical practice (ii) Poorly-reproducible methods of recording key prognostic factors and outcomes With the 'new wave' of emerging peritoneal tumour (PT) treatment centres, across the UK and Ireland, there are opportunities to integrate phase III trials into this clinical network - hence the timing of this study. In the past 12 months, there have been two meetings (Basingstoke: April 2014; December 2014) outlining the frameworks on which to develop new PT treatment centres, benchmarked against the established reference centres (Manchester and Basingstoke). These meetings have strongly supported the parallel development of an integrated PT research network. The overarching aim of this study is to develop a robust and reproducible framework in which to undertake a phase III trial in patients with PMCR suitable for CRS with or without HIPEC. As CRS with HIPEC is a complex multi-component intervention, this agenda will be addressed using the principles of the IDEAL (Idea, Development, Evaluation, Assessment \& Long term study) framework. Stage 1 - Documenting between-centre variability For the first stage of the study, 200 eligible patients will be identified from prospective databases located at The Christie NHS Foundation Trust and Hampshire Hospitals NHS Foundation Trust. In both cases, the data extracted will be extracted from pre-existing databases and will be fully anonymised and at no point will the research team see any patient identifiable information. The anonymised data will be entered into a password protected Microsoft Excel Spreadsheet and sent using secure NHS Mail account to the study co-ordinator. To investigate the between-centre variability the investigators will collect the following 19 core variables from 100 patients from each trust: 1. Age at time of operation 2. Gender 3. WHO Performance Status 4. Primary Vs Secondary Surgery 5. Previous chemotherapy - If Yes - regimen, date and dose 6. Nodal status at diagnosis and time or surgery 7. HIPEC regimen used 8. Primary Tumour Site 9. Blood transfusion requirements 10. Duration of operation 11. Peritoneal cancer index (PCI Score) 12. Cytoreduction completeness score 13. Type of Stoma 14. Highest complication grade 15. Peri-operative 30-day mortality 16. Date Last Seen 17. Date of Recurrence 18. Date of Initial Cytoreduction \& HIPEC 19. Date of Death (if applicable) The data collected will then be analysed to establish if there are any significant differences between the variables collected from each dataset. This will be primarily carried out by the project statistician using STATA 14 software. Stage 2 - Identifying sources of variability The second stage will focus on reproducibility and explore within and between-observer sources of variability. The procedure for this stage will be partially dependant on the findings from the analysis at the end of stage 1. However, the main focus will be on the assessment of intra-operative staging \[namely Peritoneal Cancer Index (PCI) score\] - the best available instrument to stratify case-mix. This stage of this study involves intra-operative photographs \& videos of intra-abdominal tumour deposits in patients undergoing cytoreductive surgery for peritoneal metastases of colorectal origin. The photographs \& videos will be taken in a systematic manner through 13 zones, using an internationally recognised grid, to derive patient-level scores of the burden of tumour. The only risk to patients is approximately 15 minutes additional anaesthetic time. The study requires 12 patients with complete recordings of all 13 zones - it is estimated that up to 25 patients might need to be consented to derive a full complement of 12 sets. These individuals will consent to partake in the study. The scores will be linked with other patient-level clinical, pathological, and imaging data. Patient identification and recruitment process Patients will first be identified at the weekly colorectal Multi-Disciplinary Team meeting by the colorectal consultant supervising the study. The selected patients will later be approached by a core member of the clinical team during their pre-op clinic visit to participate in the study. A patient information sheet will be handed to them at least 24 hours before they are required to consent for the study. This will contain all the details of the study and state that participation is voluntary. There will also be details on how to withdraw from the study if so desired. Once all the required photographs \& videos have been collected, they will be imported into a secure database and used to create an electronic survey for colorectal surgeons based within the UK and internationally. Each surgeon will be sent a link to the survey where they will first view a welcome page informing them of the process. They will then proceed to view a set of videos taken at the start of an operation and then asked to assess how they would score the PCI. Once they have entered a score between 0 and 39, they will then go on to view the photographs/videos taken after the cytoreduction has been performed and assess how they would rate the CC score (CC0, CC1, CC2 or CC3). They will then go on to do the same for the remaining procedures. The order in which the procedures appear will be randomised in order to reduce any scoring bias. Once all required consultants have completed the survey, the scores will be extracted and analysed for inter-observer variation. Stage 3 - Development of trial manual to standardise variation The next stage of the study is to utilities the Delphi method to send out a modified version of the electronic questionnaire over a number of rounds. The results from the initial questionnaire sent out in Stage 2 will be analysed to establish which regions had agreement from less than 70% of the participants. The survey will then be redistributed with a set of explanations \& definitions which will aim to have participants converge on the same ("correct") score. In addition, each answer will display the percentage of participants who selected each score in the previous round. If there continues to be regions with an agreement level of less than 70%, additional rounds will take place with further definitions based on the feedback received until all regions have reached the required level of agreement. Stage 4 - Testing of the trial manual The final stage of the project will be to test the fidelity of the trial manual assembled in stage 3. The workshop that takes place in stage 3 will involve a discussion of exactly which components of the trial manual are to be audited. This audit will take place during the final three months of the project and will use the same databases from the first stage of the study. In addition, the scoring of PCI and CC scores will be re-assessed using the same images \& videos captured in stage 2 of the study to see if there has been any standardisation in the scoring since the manual was distributed. The results from this audit will be sent to the colorectal surgery teams at all involved centres and if required, will highlight any areas of the trial manual that do not appear to be adhered to. At the end of stage 1, 2 and 3 the data collected will be analysed by the project statistician. STATA 14 software will be used to compare baseline and matched characteristics between patients treated at The Christie NHS FT and Hampshire Hospitals NHS Foundation Trust. This will be done by using standard tests for continuous variables (Kruskal-Wallis and Wilcoxon signed-rank tests, respectively) and categorical variables (χ² and McNemar tests, respectively). The results from the staff questionnaire will also be used to investigate between and within-observer variability.