E1998. Discrepancies of CT Protocols and Radiology Reports Between Initial and Tertiary Subspecialist Radiology Practices in Pancreatic Cancer
  1. Alexander Grogan; The Peter MacCallum Cancer Centre; The Walter and Eliza Hall Institute
  2. Samuel Banks; The Peter MacCallum Cancer Centre; The Walter and Eliza Hall Institute
  3. Benjamin Loveday; Melbourne Health; The Peter MacCallum Cancer Centre
  4. Michael Michael; Dentistry and Health Sciences - The University of Melbourne; The Peter MacCallum Cancer Centre
  5. Peter Gibbs; The Peter MacCallum Cancer Centre; The Walter and Eliza Hall Institute
  6. Belinda Lee; The Peter MacCallum Cancer Centre; The Walter and Eliza Hall Institute
  7. Hyun Ko; The Peter MacCallum Cancer Centre
Pancreatic ductal adenocarcinoma (PDAC) has a dismal prognosis and is on-track to become the second most common cause of cancer deaths by 2030 (1). Treatment decisions in newly diagnosed PDAC rely primarily on computed tomography (CT) imaging, preferably performed with a pancreas protocol (2). Concise and standardized radiologist reporting has been advocated by several radiological organizations in order to improve staging accuracy and enable easy interpretability by referrers (3). This study collected data from different real-world community practices and compared initial primary radiology reports with tertiary subspecialist radiology review in newly diagnosed PDAC.

Materials and Methods:
Data from the prospective, multi-national PURPLE pancreatic cancer registry [ACTRN1261700147434] was extracted for consecutive patients with newly diagnosed PDAC receiving treatment at two metropolitan tertiary centres from January 2016 – December 2019]. De-identified CT scanning data was collected and images were reported according to NCCN resectability criteria (4) by a blinded subspecialist radiologist and compared to initial primary reports. Results were calculated using Fishers exact and Wilcoxon’s signed rank tests.

148 patients were included with 87 (58.8%) having pancreatic protocol CT imaging. 85 (57.4%) CTs contained thin slices of 3 mm or less. Primary radiology reports classified the disease as resectable (R) (n=32, 21.6%), borderline resectable (BR) (n=18, 12.2%) and unresectable (UR) (n=98, 66.2%). Subspecialist review classified 24 (16.2%), 20 (13.5%) and 104 (70.3%) patients as R, BR and UR, respectively. There were 111 (75%) discrepancies on initial reports of which 52 (35.1%) resulted in a change in NCCN stage. Overall 24 (16.2%) patients were upstaged and 14 (9.5%) downstaged: trend towards upstaging on subspecialist review (p=0.091). 18 (56.3%) patients were R on both subspecialist and primary review, 7 (21.9%) were upstaged from R to BR (due to tumor-vessel involvement) and 7 (21.9%) cases upstaged from R to UR (n=3 missed metastatic disease, n=4 tumor-vessel involvement). 10 (55.6%) cases were upstaged from BR to UR (n=2 missed metastatic disease, n=8 vessel involvement). Overall 3 (3.1%) cases were downstaged from UR to R (n=2 metastases misdiagnosed, n=1 vessel involvement). 8 (8.2%) were downstaged from UR to BR (n=4 metastases misdiagnosed, n=4 vessel involvement). 3 (16.7%) cases downstaged from BR to R: all due to vessel involvement.

The relative low proportion of performed dedicated CT pancreatic protocols in newly diagnosed PDAC and the large percentage of treatment altering differences between initial primary and tertiary subspecialist reports reflects the challenges of the real-world oncology care. Continuous education in CT scanning technique and standardized reporting beyond tertiary centres should be strongly encouraged and facilitated to ensure that newly diagnosed PDAC patients undergo optimal treatment pathways.