Iodine overlays to improve differentiation between peritoneal carcinomatosis and benign peritoneal lesions

• Published in: European radiology Vol. 30; no. 7; pp. 3968 - 3976
• Main Authors: Lennartz, Simon, Zopfs, David, Abdullayev, Nuran, Bratke, Grischa, Le Blanc, Markus, Slebocki, Karin, Wagner, Anton, Wybranski, Christian, Wahba, Roger, Maintz, David, Große Hokamp, Nils, Persigehl, Thorsten
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.07.2020; Springer Nature B.V
• Subjects: Attenuation; Computed Tomography; Conspicuity; Diagnostic Radiology; Diagnostic systems; Differentiation; Gastric cancer; Image quality; Imaging; Internal Medicine; Interventional Radiology; Iodine; Latency; Lesions; Medical imaging; Medicine; Medicine & Public Health; Metastases; Neuroradiology; Peritoneal diseases; Peritoneum; Quality assessment; Radiology; Subjective assessment; Ultrasound; Abdominal cavity; Peritoneal neoplasms; Tomography, X-ray computed; Image processing, computer-assisted; Iodine
• ISSN: 0938-7994; 1432-1084
• DOI: 10.1007/s00330-020-06729-5

Objectives Peritoneal carcinomatosis (PC) is a prognostically relevant metastatic disease which may be difficult to depict in postoperative patients, particularly in early stages. This study aimed to determine whether PC could be diagnosed more accurately when using a combination of spectral detector CT (SDCT)-derived conventional images (CI) and iodine overlay images (IO) compared with CI only. Methods Thirty patients with PC and 30 patients with benign peritoneal alterations (BPA) who underwent portal-venous abdominal SDCT were included. Four radiologists determined the presence/absence of PC for each patient and assessed lesion conspicuity, diagnostic certainty, and image quality using 5-point Likert scales. Subjective assessment was conducted in two sessions comprising solely CI and CI/IO between which a latency of 6 weeks was set. Iodine uptake and HU attenuation were determined ROI-based to analyze quantitative differentiation of PC/BPA. Results Specificity for PC was significantly higher when using CI/IO compared with using CI only (0.86 vs. 0.78, p  ≤ 0.05), while sensitivity was comparable (0.79 vs. 0.81, p  = 1). In postoperative patients, the increase in specificity was the highest (0.93 vs. 0.80, p  ≤ 0.05). Lesion conspicuity was rated higher in CI/IO (4 (3–5)) compared with that in CI only (3 (3–4); p  ≤ 0.05). Diagnostic certainty was comparable (both 4 (3–5); p  = 0.5). CI/IO received the highest rating for overall image quality and assessability (CI/IO 5 (4–5) vs. CI 4 (4–4) vs. IO 4 (3–4); p  ≤ 0.05). Area under the receiver operating characteristics curve (AUC) for quantitative differentiation between PC and BPA was higher for iodine (AUC Iodine  = 0.95, AUC HU  = 0.90). Conclusions Compared with CI, combination of CI/IO improves specificity in the assessment of peritoneal carcinomatosis at comparable sensitivity, particularly in postoperative patients. Key Points • Combination of iodine overlays and conventional images improves specificity when assessing patients with peritoneal carcinomatosis at comparable sensitivity. • Particularly in postsurgical patients, iodine overlays could help to avoid false-positive diagnosis of peritoneal disease. • Iodine overlays alone provided inferior image quality and assessability than conventional images, while the combination of both received the highest ratings. Iodine overlays should therefore be used in addition to and not as a substitute for conventional images.