Although the accuracy of preoperative MRI staging has been established on follow-up histopathologic examination, the reproducibility of MRI staging has been evaluated in studies with expert radiologists reading a large sample of MRI images and therefore is not generalizable to the real-world setting.
The purpose of this study was to evaluate the interrater reliability of MRI for distance to the mesorectal fascia, T category, mesorectal lymph node status, and extramural depth of invasion for preoperative staging of primary rectal cancer.
This was a prospective, cross-sectional survey.
The study was conducted in Ontario, Canada.
Participants included GI radiologists.
Participants read 5 preselected staging MRIs using a synoptic report and participated in an educational Webinar.
Distance to the mesorectal fascia, T category, extramural depth of invasion, and mesorectal lymph node status for each MRI were abstracted. Data were analyzed in aggregate using percentage of agreement, Fleiss κ, and interclass correlation coefficients to assess interrater reliability.
Reliability was highest for distance to the mesorectal fascia with an intraclass correlation of 0.58 (95% CI, 0.27–0.80). Kappa scores for T category, mesorectal lymph node status, and extramural depth of invasion were 0.38 (95% CI, 0.23–0.46), 0.41 (95% CI, 0.32–0.49), and 0.37 (95% CI, 0.16–0.82). There was no difference when radiologists were stratified by experience or volume.
Scores may have been affected by MRI selection, because they were chosen to demonstrate diagnostic challenges for the Webinar and did not reflect a representative sample.
Interrater reliability was highest for distance to mesorectal fascia, and therefore, it may be a more reliable criterion than T category, extramural depth of invasion, or mesorectal lymph node status. Combined with the fact that an uninvolved mesorectal fascia is more consistent with the overall goal of rectal cancer surgery, it should be considered as an important MRI criterion for preoperative treatment decision making in the real-world setting. See Video Abstract at http://links.lww.com/DCR/A763.
1 Department of Surgery, Mount Sinai Hospital, Toronto, Ontario, Canada
2 Department of Surgery, University of Toronto, Toronto, Ontario, Canada
3 Zane Cohen Centre for Digestive Diseases, Toronto, Ontario, Canada
4 Department of Diagnostic Imaging, Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada
5 Department of Radiology, St. Joseph’s Health Centre, Toronto, Ontario, Canada
6 Institute for Clinical Evaluative Sciences, Toronto, Ontario, Canada
7 Department of Radiology, Royal Marsden National Health Service Foundation Trust, Sutton, Surrey, United Kingdom
Funding/Support: Funding for this study was provided by a grant awarded through the Cancer Services Innovation Partnership, a joint initiative between the Canadian Cancer Society (Ontario Division) and Cancer Care Ontario.
Financial Disclosures: None reported.
Correspondence: Erin D. Kennedy, M.D., Ph.D., Mount Sinai Hospital, 600 University Ave, Room 449, Toronto, Ontario M5G 1 X5, Canada. E-mail: email@example.com