AUTHOR=Sheng Jing , Li Ting-Ting , Zhang Huan-Huan , Xu Hua-Feng , Cai Xue-Mei , Xu Rong , Ji Qiong-Qiong , Wu Yu-Meng , Huang Ting , Yang Xiu-Jun TITLE=CT and MR imaging features of soft tissue rhabdoid tumor: compared with rhabdomyosarcoma in children JOURNAL=Frontiers in Pediatrics VOLUME=11 YEAR=2023 URL=https://www.frontiersin.org/journals/pediatrics/articles/10.3389/fped.2023.1199444 DOI=10.3389/fped.2023.1199444 ISSN=2296-2360 ABSTRACT=Objective

To assess the computed tomography (CT) and magnetic resonance (MR) imaging characteristics of soft tissue rhabdoid tumors (RT) and compare them with those of rhabdomyosarcoma (RMS).

Methods

We conducted a retrospective analysis of 49 pediatric patients from 2011 to 2022, comprising 16 patients with soft tissue RT and 33 patients with RMS who underwent CT or MRI scans. Key imaging features, as well as clinical and pathological data, were compared between the two groups. The multivariate logistic regression analysis was used to determine independent differential factors for distinguishing soft tissue RT from RMS, and the model was established. The final prediction model was visualized by nomograms and verified internally by using a bootstrapped resample 1,000 times. The diagnostic accuracy of the combined model was assessed in terms of discrimination, calibration, and clinical utility.

Results

Age, sex, number of lesions, and primary locations were similar in both groups. The imaging characteristics, including margin, calcification, surrounding blood vessels, and rim enhancement, were associated with the two groups of soft tissue tumors, as determined by univariate analysis (all pā€‰<ā€‰0.05). On multivariate logistic regression analysis, the presence of unclear margin (p-value, adjusted odds ratio [95% confidence interval]: 0.03, 7.96 [1.23, 51.67]) and calcification (0.012, 30.37 [2.09, 440.70]) were independent differential factors for predicting soft tissue RT over RMS. The presence of rim enhancement (0.007, 0.05 [0.01, 0.43]) was an independent differential factor for predicting RMS over soft tissue RT. The comprehensive model established by logistic regression analysis showed an AUC of 0.872 with 81.8% specificity and 81.3% sensitivity. The decision curve analysis (DCA) curve displayed that the model achieved a better net clinical benefit.

Conclusion

Our study revealed that the image features of calcification, indistinct margins, and a lack of rim enhancement on CT and MRI might be reliable to distinguish soft tissue RT from RMS.