The Pilot

---

The project began in August 2019 and the pilot ran from 15 October 2020 to 30 September 2021, trialling the synoptic template for use by radiologists reporting CT scans in two hospitals in Melbourne. The pilot found that the use of a synoptic radiological report for the reporting of pancreatic CT scans demonstrated a clinically significant improvement in the accuracy of assessment of localised pancreatic cancer. More details of the findings from the pilot can be found here: Synoptic CT scan reporting of pancreatic adenocarcinoma to align with international consensus guidelines on surgical resectability: a Victorian pilot – PubMed (nih.gov).

TRIAL GOVERNANCE & REFERENCE GROUPS

---

The trial governance framework consists of a Trial Management Committee (TMC) and its three subcommittees, namely, Trial Operations Subcommittee, Clinical Reference group and Consumer Reference Group, which were all formed at the outset of the grant. The TMC is made up of all chief investigators, associate investigators and a consumer representative. Its role is to oversee all aspects of the study (the operational elements of which is delegated to the Operations Subcommittee). 

The SCANPatient Clinical Reference Group takes responsibility for fine-tuning any of the details of the synoptic reporting and/or logistics from a radiological and surgical perspective as well as the implementation of these reports into day-to-day practice and the satisfaction survey. This reference group consists of radiologists, surgeons and medical oncologists.

The Consumer Reference Group considers the communications strategy from a consumer perspective. The group has helped develop an appropriate ethics and governance framework and, depending on results, will consider communication around how best to mainstream practices across the country. The members of the reference group were recruited from national PC consumer advocacy groups, including PanKind (The Australian Pancreatic Cancer Foundation) and the PANCARE Foundation (a charity committed to improving outcomes for those affected by upper gastrointestinal cancers), who are both partners on the MRFF grant, as well as the Australasian Gastro-Intestinal Trials Group (AGITG) and the Cancer Research Program’s Consumer Advisory Group (CAG).

SUB-STUDIES

---

In addition to addressing the key objectives of the SCANPatient main study, researchers are also developing a number of related sub-studies which intend to provide further evidence of the reliability, consistency and utility of the synoptic reporting tool and approach.

1. ASSESSMENT OF THE INTER- AND INTRA-OBSERVER VARIABILITY OF SYNOPTIC SCAN REPORTING AMONG SEVERAL RADIOLOGISTS.

The inter- and intra-observer variability sub-study introduces a quality assurance (QA) component to the original pilot study that led to the SCANPatient trial. The pilot study tested the feasibility of synoptic reporting at two hospitals in Victoria.

The QA component has two major aims: to test the inter-observer variability of CT scan reporting using the synoptic radiology report template and to assess the intra-observer variability of scan reporting. Clinicians involved in the pilot trial will be reporting on scans previously seen, and thus there is an opportunity to draw a comparison to assess intra-observer results.

The QA component involves 60 eligible CT scans that will be identified from the pilot and an existing Pancreatic Cancer Image Biobank created by the UGICR. Six radiologists from hospitals that were involved in the pilot will be provided access to the deidentified CT scans on a secure platform (Monash University XNAT). Each radiologist will then independently assess the CT scans using the REDCAP radiological synoptic report. The responses for each radiologist will be analysed to examine reproducibility and reliability.

The project has obtained the necessary ethics and governance approvals from both Alfred Health and Austin Health, from which the internal 60 scans will be collated. The next steps involve the creation of the secure Monash University platform to upload the de-identified CT scans for the participating radiologists to report.

The concept of another proposed radiological sub-study is to review prose CT reports prior to introduction of synoptic report to analyse how many of the 63 synoptic data points are addressed in the free-form prose report.

2. CORRELATION BETWEEN IMAGING AND SURGICAL MARGIN STATUS

Approximately 20%-25% of PDAC patients have either resectable or borderline resectable disease. Standard of care for resectable stage disease is upfront surgery, and while increasing numbers of borderline resectable patients are being treated with neoadjuvant therapy prior to surgery, many still do proceed to upfront resection. Notably, a large number of upfront resections are microscopically incomplete (R1), and one quarter of patients will develop disease recurrence within 6 months. R0 resection rate is an important prognostic factor, diminishing local and distant recurrence rates, hence improving survival and is the goal for pancreatic cancer surgery.

Despite being considered resectable preoperatively, many patients return an R1 margin after resection, and this is likely more commonly the case for borderline resectable disease where the extent of disease at presentation is more advanced.

Measures to improve local control and achieve R0 margins include preoperative radiotherapy, however this is uncommonly delivered in Australia.

An understanding of the true number of R1 positive margins following resection for resectable and for borderline resectable pancreas cancer is lacking in Australia. As the SCANPatient trial will routinely be collecting pathology reports for resection specimens, and systematically documenting resectability status on imaging preoperatively, we are ideally situated to provide answers regarding true margin positive rates in Australia across up to 40 sites nationwide.

Further subtle correlations to stratify peritumoural inflammatory change on CT imaging (in distinction from tumour infiltration) may also be possible from analysis of this large dataset.

3. ARTIFICIAL INTELLIGENCE IN THE CLASSIFICATION OF PANCREATIC CANCER

Artificial Intelligence (AI) is a growing field within medicine and has the potential to greatly influence the outcomes of patients with pancreatic cancer, given its ability to integrate large amounts of data to identify the relationships between variables and answer clinical questions. The role of AI in pancreatic cancer has been explored in the preoperative, intraoperative and postoperative space.

Of particular interest for SCANPatient clinical trial, AI tools have been developed to classify pancreatic tumours using CT radiological features. In this substudy, we will conduct a review of the literature to explore and summarise the existing AI tools using CT images to classify pancreatic tumours and the algorithms developed. The barriers and limitations to the application of AI in clinical practice and how these limitations may be overcome will also be discussed.

Compiling and understanding the current AI tools in pancreas CT scans will pave the way to apply these advanced technologies to the SCANPatient trial.

4. COMPLETENESS OF CT SCAN PROSE REPORTING OF PANCREATIC CANCER IN AUSTRALIA COMPARED TO A STRUCTURED SYNOPTIC REPORTING METHOD BASED ON THE INTERNATIONAL CONSENSUS GUIDELINES FOR SURGICAL RESECTABILITY.

Pancreatic cancer is a challenging and often aggressive disease that requires careful consideration when determining the most appropriate treatment strategies for affected patients. In this regard, the findings from CT scans play a pivotal role in the decision-making process of surgeons and oncologists in determining the optimal treatment options for pancreatic cancer patients.

 In routine practice, radiologists provide narrative CT reports detailing various anatomical aspects of suspected pancreatic tumours, including their characteristics, extent of vascular involvement, the degree of local spread, as well as the presence of nodal and distant metastases.

The objective of the proposed study is to evaluate the completeness of the current CT reporting of pancreatic ductal adenocarcinoma (PDAC) in Australia. Specifically, the study aims to determine whether standard of care CT reports sufficiently address the 63 distinct fields outlined in a structured synoptic report form, which was developed and tested in two Victorian hospitals. These fields have been derived from the International Consensus guidelines on the surgical resectability of PDAC.

This study seeks to verify whether the current CT reporting approach for PDAC cases in Australia aligns with the International Consensus guidelines and encompasses all the relevant information needed for accurate assessment of surgical resectability and determination of the most suitable course of treatment.

---

We sincerely invite suggestions for more sub-study ideas and proposals. You are welcome to email us via SCANPatient@monash.edu, or using the contact form below.    

SCANPatient Information session

---

The SCANPatient Central Team are running live online information sessions to help interested sites to better understand the workflow of the trial. During these sessions, we cover the following:

• Overview of the project
• Ethics approval & Governance processes
• Site principal investigator (PI) and site liaison roles
• Workflow
• Funding support for sites
• FAQs

You can download the slides discussed during the video at this link