Purpose

Atypical teratoid/rhabdoid tumor (AT/RT) is a kind of central nervous system malignant tumor in children. In this study, we aimed to develop a practically clinical nomogram and risk grouping system to predict 1-year overall survival for patients with atypical teratoid/rhabdoid tumor.

Methods

The nomogram was constructed based on the pediatric tumor registry of Xinhua Hospital affiliated to Shanghai Jiaotong University School of Medicine. Fifty-four information-integrated patients with atypical teratoid/rhabdoid tumor were included from the database. Cox regression analyses were used to select independent prognostic factors. Based on the fitted multivariate Cox regression model, a nomogram of 1-year overall survival for atypical teratoid/rhabdoid tumor patients was generated. Moreover, the nomogram was validated by assessing its discrimination and calibration.

Results

In these patients, age at diagnosis, the extent of tumor resection, radiotherapy, and chemotherapy were included in the multivariate Cox regression model. Based on this multivariate Cox regression model, a nomogram of 1-year overall survival for atypical teratoid/rhabdoid tumor patients was generated. The nomogram had good discrimination (the concordance index was 0.781) and calibration curves showed no deviation from reference lines. Decision curve analysis demonstrated this nomogram was useful for clinical practice. The risk grouping system was built based on nomogram-derived risk scores, which could classify patients into 3 risk groups. Compared with the low-risk group, the risk of 1-year death was significantly higher in the intermediate-risk group (hazard ratio = 1.42, 95%, confidence intervals = 0.49-4.11) and high-risk group (hazard ratio = 9.78, 95% confidence intervals = 3.53-27.1).

Conclusion

A nomogram and risk grouping system were built to predict for the 1-year overall survival of atypical teratoid/rhabdoid tumor patients. The nomogram could facilitate a personalized prognostic evaluation for atypical teratoid/rhabdoid tumor patients and help medical practitioners make better treatment.