NCOG-18. IS THE RANO CRITERIA FOR LOW-GRADE GLIOMA RELIABLE IN THE CLINICAL SETTING? – A RELIABILITY STUDY

Abstract

Abstract INTRODUCTION Magnetic resonance imaging (MRI) is a fundamental component of longitudinal neuro-oncology evaluation and decision-making. The Response Assessment in Neuro-Oncology criteria for low-grade gliomas (RANOLGG) was designed as an outcome measure for clinical trials. Thus, this project intends to study the reliability of RANOLGG in the clinic setting. METHODS 21 pairs of brain MRIs, that averaged three years apart, were selected from 21 patients with tissue diagnosis of WHO grade 2 gliomas. Two neuro-oncologists and two neuro-radiologists reviewed and independently scored the MRI pairs according to RANOLGG categories of progressive disease, stable disease, minor response, partial response, and complete response. Kappa-Fleiss (KF) was used to evaluate agreement among reviewers. RESULTS Reviewers awarded identical scores in only 33% of MRI pairs and there was a complete disagreement in one MRI pair. Overall reliability of the criteria in the clinical setting is moderate (KF = 0.44). Agreement between neuroradiologists (KF = 0.51) and between neuro-oncologists (KF = 0.48) were similar. Interpretation of post-contrast T1-weighted images had a better agreement (KF = 0.31) than T2/FLAIR-weighted images which had a poor agreement (KF = -0.02). Classification of progression versus non-progression had only a moderate agreement (KF= 0.49). History of radiation therapy or chemotherapy did not influence the criteria reliability (fisher exact text, p = 0.58, p =0.27, respectively). CONCLUSION RANOLGG reliability in the clinical setting is moderate, therefore it should be used cautiously for clinical decision-making. Other tools that can support the neuro-oncologist in the follow-up of patients with low-grade glioma are additional MRI sequences other than T2/FLAIR and contrast – weighted images, computer-aided diagnosis such as volumetrics, spectroscopy, positron emission tomography, and multidisciplinary tumor boards. Likewise, when image criteria for low-grade gliomas are designed, T2/FLAIR should be used guardedly, due to low interpretation agreement.