Perfusion CT assessment of the colon and rectum: Feasibility of quantification of bowel wall perfusion and vascularization

Sairah Khan, Vicky Goh, Emily Tam, D. Wellsted, Steve Halligan

Research output: Contribution to journalArticlepeer-review

11 Citations (Scopus)

Abstract

The aim was to determine the feasibility of vascular quantification of the bowel wall for different anatomical segments of the colorectum. Following institutional ethical approval and informed consent, 39 patients with colorectal cancer underwent perfusion CT. Blood flow (BF), blood volume (BV), mean transit time (MTT), and permeability surface area product (PS) were assessed for different segments of the colorectum: ascending, transverse, descending colon, sigmoid, or rectum, that were distant from the tumor, and which were proven normal on contemporary colonoscopy, and subsequent imaging and clinical follow up. Mean (SD) for BF, BV, MTT and PS for the different anatomical colorectal segments were obtained and compared using a pooled t-test. Significance was at 5%. Assessment was not possible in 9 of 39 (23%) patients as the bowel wall was ≤5mm precluding quantitative analysis. Forty-four segments were evaluated in the remaining 30 patients. Mean BF was higher in the proximal than distal colon: 24.0 versus 17.8mL/min/100g tissue; p=0.009; BV, MTT and PS were not significantly different; BV: 3.46 versus 3.15mL/100g tissue, p=0.45; MTT: 15.1 versus 18.3s; p=0.10; PS: 6.84 versus 8.97mL/min/100 tissue, p=0.13, respectively. In conclusion, assessment of bowel wall perfusion may fail in 23% of patients. The colorectum demonstrates segmental differences in perfusion.
Original languageEnglish
Pages (from-to)821-824
JournalEuropean Journal of Radiology
Volume81
Issue number5
Early online date9 Mar 2011
DOIs
Publication statusPublished - 2012

Fingerprint

Dive into the research topics of 'Perfusion CT assessment of the colon and rectum: Feasibility of quantification of bowel wall perfusion and vascularization'. Together they form a unique fingerprint.

Cite this