IMG-24. Improving the MRI characterization of intracranial metastatic disease: a critical need

Intracranial metastases (IM) are more common than primary malignant brain tumors, yet have been under-studied, presenting an important unmet clinical need. Many clinical trials excluded patients with IM, limiting our understanding of the intracranial efficacy of emerging treatments, though fortunately this is changing. Nevertheless, IM trials commonly exclude patients with leptomeningeal metastatic disease (LMD), related to their worse prognosis and/or expected lower likelihood of treatment-response. While criteria for assessing treatment-response for patients with confirmed LMD exist, there is a surprising lack of explicit guidelines regarding how LMD should be diagnosed or excluded in broader IM trials. This includes uncertainty regarding when cerebrospinal fluid (CSF) analysis is worthwhile, especially as this can be falsely negative, which is particularly relevant when MRI (magnetic resonance imaging) abnormalities are compelling. This lack of clear guidelines conveys a substantial risk of variability and misdiagnosis, with the potential to compromise trial results. Improved characterization also has the potential to demonstrate differences in treatment-response based on lesion morphology (especially parenchymal versus leptomeningeal). Here, we propose and illustrate new MRI criteria for classifying the range of IM appearances to minimize ambiguity in distinguishing between parenchymal IM and LMD, including recommended next steps. A diagnosis of LMD implies evidence of tumor cells within the CSF, hence we have divided MRI appearances into three groups based on the expected likelihood of this: 1) consistent with LMD and not requiring confirmation by CSF analysis; 2) equivocal, warranting correlation with CSF analysis and/or MRI spine; and 3) low suspicion of LMD, not requiring further testing. Importantly, we do not consider parenchymal metastases that contact a pial surface to be diagnostic of LMD; instead, we classify these within either the equivocal or low suspicion group, depending on morphology. Additionally, we outline potential diagnostic pitfalls and technical considerations for the MRI protocol.