Back to Search Start Over

Prospective Comparison of the Performance of MRI Versus CT in the Detection and Evaluation of Peritoneal Surface Malignancies.

Authors :
Chia CS
Wong LCK
Hennedige TP
Ong WS
Zhu HY
Tan GHC
Kwek JW
Seo CJ
Wong JSM
Ong CJ
Thng CH
Soo KC
Teo MCC
Source :
Cancers [Cancers (Basel)] 2022 Jun 29; Vol. 14 (13). Date of Electronic Publication: 2022 Jun 29.
Publication Year :
2022

Abstract

Background: The performance of MRI versus CT in the detection and evaluation of peritoneal surface malignancies (PSM) remains unclear in the current literature. Our study is the first prospective study in an Asian center comparing the two imaging modalities, validated against intra-operative findings. Methods: A total of 36 patients with PSM eligible for CRS-HIPEC underwent both MRI and CT scans up to 6 weeks before the operation. The scans were assessed for the presence and distribution of PSM and scored using the peritoneal cancer index (PCI), which were compared against PCI determined at surgery. Results: Both MRI and CT were 100% sensitive and specific in detecting the overall presence of PSM. Across all peritoneal regions, the sensitivity and specificity for PSM detection was 49.1% and 93.0% for MRI, compared to 47.8% and 95.1% for CT (p = 0.76). MRI was more sensitive than CT for small bowel disease, although the difference did not reach statistical significance. Comparing PCI on imaging with intra-operative PCI, the mean difference was found to be −3.4 ± 5.4 (p < 0.01) for MRI, and −3.9 ± 4.1 (p < 0.01) for CT. The correlation between imaging and intra-operative PCI was poor, with a concordance coefficient of 0.76 and 0.79 for MRI and CT, respectively. Within individual peritoneal regions, there was also poor agreement between imaging and intra-operative PCI for both modalities, other than in regions 1 and 3. Conclusion: MRI and CT are comparable in the detection and evaluation of PSM. While sensitive in the overall detection of PSM, they are likely to underestimate the true disease burden.

Details

Language :
English
ISSN :
2072-6694
Volume :
14
Issue :
13
Database :
MEDLINE
Journal :
Cancers
Publication Type :
Academic Journal
Accession number :
35804951
Full Text :
https://doi.org/10.3390/cancers14133179