Assessment of sacral ratio in patients with anorectal malformations: Can magnetic resonance imaging replace conventional radiograph?

Wilfried Krois; Francesca Palmisani; Peter Gröpel; Patricia Feil; Martin L Metzelder; Janina M Patsch; Carlos A Reck-Burneo

doi:10.1016/j.jpedsurg.2021.01.013

Assessment of sacral ratio in patients with anorectal malformations: Can magnetic resonance imaging replace conventional radiograph?

J Pediatr Surg. 2021 Nov;56(11):1993-1997. doi: 10.1016/j.jpedsurg.2021.01.013. Epub 2021 Jan 16.

Authors

Wilfried Krois¹, Francesca Palmisani², Peter Gröpel³, Patricia Feil², Martin L Metzelder², Janina M Patsch⁴, Carlos A Reck-Burneo²

Affiliations

¹ Department of Surgery, Clinical Department of Pediatric Surgery, Medical University of Vienna, Spitalgasse 23, 1090 Vienna, Austria. Electronic address: wilfried.krois@meduniwien.ac.at.
² Department of Surgery, Clinical Department of Pediatric Surgery, Medical University of Vienna, Spitalgasse 23, 1090 Vienna, Austria.
³ University of Vienna, Department of Sport Science, Vienna, Austria.
⁴ Medical University of Vienna, Department of Biomedical Imaging and Image-Guided Therapy, Vienna, Austria.

PMID: 33485613
DOI: 10.1016/j.jpedsurg.2021.01.013

Abstract

Introduction: In patients with anorectal malformations (ARM), the sacral ratio (SR) serves as an established predictive marker for functional prognosis and is derived from conventional radiographs. More recently, magnetic resonance imaging (MRI) has emerged as a diagnostic alternative for preoperative/baseline assessment of patients with ARM. With this study we aimed to evaluate if the SR could be reliable be measured from MRI images and if it correlated to SR measurements obtained from radiographs.

Methods: Two raters analyzed MRI data and conventional radiographs from thirty-one subjects (n = 17 with ARM; n = 14 controls). We calculated intra-class correlation coefficients (ICCs) to test inter-rater reliability and applied paired t-tests to examine if SR parameters from MRI were comparable with those from pelvic radiograph. We further computed Pearson's correlation coefficients to test the linear relationship between SR calculated from MRI and conventional radiographs.

Results: The ARM and control groups did not statistically differ in their age and weight on the day of the MRI scan. Reliability analysis revealed an excellent inter-rater agreement for SR from radiograph parameters with an intraclass correlation coefficient (ICC) of 0.94, respectively of 0.86 for MRI (frontal plane), and of 0.84 for MRI (sagittal plane). The correlation coefficient between SR calculated from the sagittal MRI and SR calculated from radiograph images was significant and high (r = 0.80, P < 0.001). The SRs from MRI images did not significantly differ from SRs from radiographs, but were also not statistically equivalent.

Conclusion: Our results demonstrate that the SR can be derived from MRI images with good inter-rater reliability. The SR value is marginal higher when calculated on MRI, presumably due to inclusion of cartilaginous yet unossified structures.

Keywords: Anorectal malformations; Bladder function; MRI; Predictive value; Radiographs; Sacral ratio.

MeSH terms

Anorectal Malformations* / diagnostic imaging
Humans
Magnetic Resonance Imaging
Observer Variation
Radiography
Reproducibility of Results
Sacrum / diagnostic imaging