Inter-observer variability in assessing image-defined risk factors: implications for risk stratification in locoregional abdominopelvic neuroblastoma.

PURPOSE

Risk stratification for locoregional neuroblastoma partially relies on image-defined risk factors (IDRFs). This study aimed to evaluate how inter-observer variability in assessing IDRFs impacts risk stratification in locoregional abdominopelvic neuroblastoma.

METHODS

A retrospective analysis was conducted on 123 patients who underwent upfront contrast-enhanced CT scans. Two radiologists independently assessed the presence of IDRFs. Patients were staged as either L1 (IDRF-negative) or L2 (IDRF-positive) according to the International Neuroblastoma Risk Group Staging System. Based on the radiologists' evaluations, 97 cases with sufficient clinical data were classified into risk groups using the revised Children's Oncology Group neuroblastoma risk classifier. The kappa values and 95% confidence intervals (CIs) were calculated to assess inter-radiologist agreement on IDRF evaluation and risk stratification.

RESULTS

There was low agreement between radiologists in assessing L1/L2 status with a kappa value of 0.28 (95% CI: 0.14-0.42). However, agreement for evaluating the number of IDRFs was good, with an intraclass correlation coefficient of 0.73 (95% CI: 0.64-0.80). Based on the first radiologist's evaluation, 13 patients were classified as low-risk, 52 as intermediate-risk, and 32 as high-risk. Based on the second radiologist's evaluation, 37 patients were classified as low-risk, 37 as intermediate-risk, and 23 as high-risk. The kappa value for risk stratification between the two radiologists was 0.47 (95% CI: 0.33-0.62).

CONCLUSION

Inter-observer variability in assessing IDRF presence may affect risk stratification in locoregional abdominopelvic neuroblastoma.