Canadian Optically-guided approach for Oral Lesions Surgical (COOLS) trial: study protocol for a randomized controlled trial

Oral cancer is a major health problem worldwide, accounting for 274,000 new cases and 145,000 deaths annually [1]. Although it occurs at a site that is easily accessible for examination it is often diagnosed at an advanced stage, with 5-year survival rates ranging from 30-60%, depending on the global locale. Treatment of early stage squamous cell carcinoma is an essential component of effective oral cancer management; a recent large (~190,000) randomized trial of a screening program showed the mortality rate ratio between intervention and control groups was 0.79 [95% CI 0.51 - 1.22] [2]. However, even with treatment there are high rates of second oral malignancies, with up to a third of these patients suffering a recurrence or a second primary [3, 4]. There has been extensive research in the importance of examining the field surrounding oral cancers for risk assessment and management of this disease [5–7]. Using molecular technology, it is becoming increasingly apparent that genetically altered cells are often widespread across the mucosa of patients with oral cancer, extending into clinically and histologically normal tissue, and that these cells can drive the process of field cancerization. Recognizing this, surgeons try to remove oral squamous cell carcinomas (SCC) with a significant margin of surrounding normal-looking oral mucosa. However, the occult disease varies in size and a wealth of evidence suggests that it frequently extends beyond the tumor clearance area [5–8]. This extension may be responsible for the high rate of cancer recurrence at the primary site (10-30% of SCC cases) [9–11]. Taking margins that are too large can result in over-cutting (causing severe cosmetic and functional morbidity) and margins that are too small may leave cancerous tissue behind, as evidenced by frequent positive surgical margins and high local and regional recurrence - a failure of the 'best practice'.

Our research team has developed an approach using fluorescence visualization (FV) technology, which improves the recognition of the field at risk surrounding a visible oral cancer [12, 13]. Our preliminary results show that using FV to define the field at risk for surgical resection can result in a marked reduction in recurrence rates at 3 years (0/38 vs. 7/22) [14]. Our goal in this study is to rigorously evaluate and validate the effectiveness of FV-guided surgeries compared to white light in a multicentre pan-Canadian randomized control study.

The trial has 4 objectives. The primary objective is to determine the effectiveness of FV-guided surgery in reducing local recurrence after surgical excision of severe dysplasia or cancer of the oral cavity. Secondary objectives include 1) To collect molecular and phenotypic evidence in margins to test if FV produces a shift in surgical field, sparing normal tissue while catching high-risk occult tissue; 2) To collect relative cost-effective evidence of the two arms in both the cost per avoided recurrence and the cost per quality-adjusted life years gained; and 3) To develop a knowledge translation (KT) strategy that will foster the dissemination of FV-guided surgery across Canada and globally.

The trial data will be analysed by the biostatistician (PMB) who will be blinded to the treatment allocation (masked as Treatment A or Treatment B) during the study period.

A Data Safety and Monitoring Board (DSMB) will meet yearly to monitor accrual and adverse events. The DSMB will also review the results of the interim analyses and provide a recommendation as to whether or not the trial should be stopped early for efficacy. The DSMB will be comprised of a medical oncologist, a head & neck surgeon, and a biostatistician.

The COOLS trial examines the efficacy of using a device. How to transfer the technique with uniformity to the participating sites across Canada is the key to success.

It will be necessary to establish uniform standard operating procedures (SOPs) for all key activities; however, each site will have its own unique environment and personnel. Prior to activation, each site will be visited with the delivery of study devices. During these visits the multidisciplinary team of surgeons, pathologists, project coordinators, and FVS at each site will be trained in all study procedures. These site visits will also provide a face-to-face opportunity to build strong relationships with site team members that should facilitate problem solving throughout the study.

Along with the site visits, a comprehensive training manual has been created and is made accessible to each site through study's web-based database. This training manual includes study protocols, operation details for each key step, how to complete and upload the scannable CRFs, sections for clinicians/surgeons with an atlas of clinical lesions in both WL and FV images and instructions for the VELscope, VELscope Vx and camera, sections for the pathologists with details in tissue blocking and the definition of key fields in the pathology synoptic report form. This manual will serve as a prototype that will be evaluated and fine-tuned throughout the project for future knowledge translation. The study website also includes a library of all study CRFs that can be downloaded.

For clinical images and the use of VELscope, all FV operators (FVS and SS) will need to pass a 2-step control process. Step 1: as part of the training process, two sets of 8 images will be provided to the site surgeons and FVS, as a first step in calibrating their judgment on clinically visible tumor boundaries and FV positive (FV loss) boundaries. The criteria for passing the step is that both clinically visible tumor boundaries and FV boundaries need to be within ± 5 mm of those drawn by the experienced FV operators from the British Columbia (BC) site. Step 2 : to be certified, all FV operators require hands-on experience in outlining the tumor and FV boundaries on 2 real patients. The same criteria will be used to assess this activity, with a requirement that both clinically visible tumor and FV boundaries be within ± 5 mm of those drawn by the experienced FV operators from BC site. This time, the BC specialists will draw on the images received from the study sites. This can be used to review the entire operation from new patient assessment to the completion of surgery (see Key Steps above) to ensure the high-quality operation.

After passing this 2-step control process, a site will be ready to recruit the first patient to the study. The PI (CFP) will review the first 10 patients from each site to ensure the quality of images is maintained and that mapping of clinical and fluorescence boundaries proceeds as per initial training and to provide suggestions to the site team members for any problems experienced during these activities. Subsequently, about 5% random samples from each site will be selected from time to time to check for the accuracy of boundary drawings. This can also mitigate 'margin creep' over time.

One risk is the potential contamination of endpoint by the surgeon, i.e., margin creep (expanding margins) in the control arm. We are requiring that both the FV and white light margins be justified just prior to surgery in the operating room with the presence of both surgeon and an independent party for FV assessment. In this fashion, the surgical margin will be placed in a consented and unbiased fashion. In order to keep the study double blinded and avoid potential bias, the FVS will be a different person from the SS. The periodic review of sites by central management team will also mitigate this problem.

The acquisition of high quality data is key to the success of any clinical trial. We have planned a systematic strategy for control of data and image capture with a special focus on ensuring that the involved processes are efficient and of high quality. For clinical data, we use a set of scannable teleforms. These forms can be scanned and quickly uploaded from the participating sites to the central database. The use of scannable forms will significantly increase the efficiency of the knowledge capture and avoid transcription errors during the data collection process. The central database manager will review uploaded CRFs for completeness and review any problems with the local site coordinators who will have access to the uploaded forms through a user interface. In order to maintain the high quality operation and integrity of the study, the CPM will periodically perform a random check of all processes of selected cases, with at least 1 case from each site every 3 months. The database will be programmed to alert the Centre and Site if a patient is overdue for their scheduled follow-up appointment. A monthly teleconference with site coordinators and site PIs will share problems among sites. We monitor the trial activity on a day-to-day basis and a common email account has been set up to alert urgent issues from the site. Any operational questions in the project will be brought forward to the Steering Committee. All problems and solutions will be logged for future reference.

A major risk to the success of this project is patient enrolment. This risk will be mitigated by requiring each site to develop patient pipeline strategies. This is a mile-stone driven project. With the consideration of allowing sufficient start-up time for each site to build its infrastructure, we project to recruit 400 patients over the first 36 months of the trial starting date.

The COOLS trial is a multicenter, phase III randomized controlled trial comparing 2 surgery approaches, one guided with a optical device (VELscope; experimental arm) and one without (control arm), which is different from the conventional drug trial. This is a large scale pan Canadian study to validate an FDA- and Health Canada-approved optical device to better define the surgical margin. We acknowledge the support and the shared vision from the Terry Fox Research Institute.

The uniqueness of this trial is multi-fold. Firstly, this is the first-ever pan Canadian surgical study of such calibre, involving 9 institutions across 6 provinces, more than 70 health professionals, and 400 patients. The 9 institutions are drawn from academic centres with a residency program and affiliated to provincial cancer agencies, with patient populations of interest, linkages to the dental system (referral pathway for patients), and onsite pathology services. The catchment areas of these participating sites cover over the locations from which 70% of oral cancer patients come from. Every year, there are ~3,400 patients diagnosed with oral cancer in Canada and ~60% are early stage oral cancer, which are the target population of our trial. Hence, if the results are validated, it can make enormous impact at the patient level in increasing patients' survival, morbidity, and quality of life, at the institutional level in changing the standard of care, and at the population level in improving the cure rate and reducing costs to the health system.

Secondly, this is an exceptional example of translational research from bench to bedside, from discovery to clinical implication. With the strong support of the ongoing Oral Cancer Predictive Longitudinal (OCPL) Study, BC researchers have built a strong foundation to lead this prospective, multicenter study. The OCPL study is an NIH/NIDCR-funded study which involves ~900 patients with oral precancer and oral cancer that have been monitored regularly since 1999 for progression/recurrence of disease. The OCPL study has served as a pipeline for referral of such patients to the trial in BC [21].

Thirdly, in order to change clinical practice we need to ascertain clinical efficacy through assessment of local recurrence at the clinical, histological and molecular level, assess the impact on quality of life and cost effectiveness, and construct a plan for KT. The molecular goal will examine the margin to see if the tool can produces a shift in surgical field, sparing normal tissue while catching high-risk occult tissue. The health economics goal will collect relative cost-effective evidence of the two treatments in both the cost per avoided recurrence and the cost per quality-adjusted life years gained and last but not the least, in the knowledge translation (KT) goal, we plan to develop a KT strategy that will facilitate rapid scale up of FV-guided surgery in Canada and beyond. With such a comprehensive package, the COOLS trial promises to collect evidence and information necessary for bridging the gap from discovery to clinical application of FV-guided surgery into an oral cancer solution for patients worldwide.

The COOLS trial is a large-scale study - large number of patients, types of health professional specialties, and geographically widespread institutions. The site establishment, personal training to standardize the protocol including the use of study device, and quality assurance are critical components to the success of the trial. To ensure the quality of the trial, a multidisciplinary trial coordinating centre has been developed to monitor the trial operation and a web-based database application has been created to facilitate the data, images, and sample flow. A data safety monitoring board and a steering committee have been formed to oversee the trial. Additionally, a central pathology review committee (chaired by KWB) has been formed to document and review all the margin pathology of the surgical samples and the adjudication of the endpoint based on the pathology of the follow-up biopsy.

At the time of writing, September 2011, 69 patients have been enrolled from 4 sites. It is anticipated that recruitment will be completed in August, 2013. Follow-up will be completed two years later with results available by 2015.

This trial represents a pan-Canadian partnership of sites and investigators/surgeons and is the first randomized trial of the pan-Canadian Network for Oral Cancer Control or PanCanNOC. The network, serving as the machinery for future research idea/projects, is committed to rigorously evaluating interventions in the prevention and treatment of oral cancer.

Publication and reporting date

Expect late 2015 or early 2016.

CFP, Trial Principle Investigator, is an Oral Pathologist, a Clinician Scientist at the BC Cancer Research Centre (Departments of Integrative Oncology and Cancer Control Research) and an Associate Professor at the University of British Columbia (Faculty of Dentistry). She is the winner of the Clinician Scientist Award from the Canadian Institute of Health Research (2007-2010) and the Scholar Award of the Michael Smith Foundation for the Health Research (2007-2013). Her primary research focus involves application of molecular and imaging tools for screening, diagnosis, and management of cancerous and precancerous oral tissues. Her investigations also involve the impact of oral cancer screening in medically underserved communities. JSD, Trial Principle Investigator, is Clinical Professor and Acting Head, Division of Otolaryngology, Faculty of Medicine, the University of British Columbia. He is also Head and active staff of the Department of Otolaryngology, Vancouver Hospitals & Health Sciences Centre (Vancouver General Hospital and UBC Hospital), and a Consulting Surgical Oncologist, Head and Neck Tumor Group of the BC Cancer Agency. His research interests include Head and neck oncology and oromaxillofacial reconstructive surgery. PMB, Trial Biostatistician, is a Senior Research Scientist at UBC affiliated-Centre of Clinical Epidemiology and Evaluation (C2E2) and an associate member of the Department of Statistics at UBC. She is the Statistical Editor for the Canadian Journal of Anesthesia and has served on several grant review committees and ethics boards. Her research interests include clinical research methodology, secondary use of administrative data and statistical education. DWA is Clinical Professor, Division of Otolaryngology, Faculty of Medicine, the University of British Columbia. He is the Divisional Head of Otolaryngology for the Fraser Health Region Hospitals, an active staff of the Department of Otolaryngology, Vancouver Hospitals & Health Sciences Centre (Vancouver General Hospital and UBC Hospital), and a Consulting Surgical Oncologist, Head and Neck Tumor Group of the BC Cancer Agency. His research interests include Head and neck oncology and quality of life post head and neck surgery. KWB, Chair, Trial Pathology Committee, is an experienced head and neck pathologist. He is a Consultant Pathologist at Vancouver General Hospital. He is also a Clinical Professor with the University of British Columbia (Department of Pathology and Laboratory Medicine). He did his undergraduate training and medical school at the University of Alberta and Residency in Anatomic Pathology at the University of British Columbia. He will be the lead pathologist in the Histology Centre Review Committee and member of the Outcome Jurisdiction Committee. CEM, Trial Imaging Consultant, is the Head of the Integrative Oncology Department at the BC Cancer Research Centre. He is also a faculty member of both the Department of Pathology and Laboratory Medicine and the Department of Physics and Astronomy at the University of British Columbia. His research is focused on automated image analysis of cell preparations, in vivo tissue imaging, quantitative microscopy, and the development of imaging technologies for the early detection of epithelial cancer types and one of the inventors of the VELScope. He is also the co-Lead of the substudy to look for phenotypic evidence using Quantitative Pathology in margins for surgical field shift. JJL, Trial Design Consultant, is a Professor and Kenedy Foundation Chair in Cancer Research in the Department of Biostatistics, Division of Quantitative Sciences at the University of Texas M. D. Anderson Cancer Center in Houston, Texas. He also serves as an Adjunct Professor at both Rice University (Statistics) and the University of Texas (School of Public Health, Biostatistics). His expertise and research interests include design and analysis of clinical trials, survival analysis, longitudinal data analysis, statistical computation/graphics, statistical methods for determining drug interaction in combination studies, and cancer chemoprevention. MPR, Co-PI and Scientific Director of the COOLS study, is the Director of the British Columbia Oral Cancer Prevention Program (BCOCPP), a structure created within the BC Cancer Agency that has brought together ~40 clinicians and scientists of very diverse disciplines experience who are engaged in leading the evolution of system-wide change in oral cancer control in BC. MPR is a Senior Scientist in the Department of Cancer Control Research at the BC Cancer Research Centre. She also serves as a Professor at both Simon Fraser University (Biomedical Physiology and Kinesiology) and the University of British Columbia (Pathology and Laboratory Medicine). She is the winner of the Oral Health Promotion Award of the Canadian Dental Association (2010). She has extensive experience with the development and management of research projects that involve large multidisciplinary teams both globally and locally in Vancouver. She leads the NIH-funded longitudinal study, the Oral Cancer Predictive and Longitudinal study, that is developing and evaluating new technology and molecular markers for prediction of cancer progression and recurrence. She is the co-Lead of the substudy to look for molecular evidence using Loss of Heterozygosity in margins for surgical field shift.