Neurococognitive and Functioal Assessment of Patients With Brain Metastases

This research protocol is designed to perform neuropsychological testing and fMRI scanning in patients with brain metastases that are to undergo either WBRT or SRS treatment at Thomas Jefferson University Hospital. Eligible patients will be referred through the department of Neurosurgery and Radiation-oncology. The patients will complete the informed consent for both the pre- and post-treatment (4 month after WBRT or SRS is completed) fMRI scanning and pre- and post-treatment (4 and 12 months after WBRT or SRS is completed) neuropsychological testing sessions and be told that participation in either the fMRI scans and the neuropsychological testing is voluntary. Patients will also be told that the data will be collected for observational research purposes only and will not directly affect their clinical care. Patients and their insurers will not be additionally billed as a consequence of this study aside from the clinical billing that is consistent with current clinical practice. The fMRI study will be added on to the pre-existing treatment planning and follow up MRI scan that would be obtained out of clinical necessity without the study. In order to be able to identify and distinguish normal versus abnormal brain activity patterns associated with the various fMRI tasks and resting state imaging procedures that are part of this study, it will be necessary to scan a subset of healthy (age, education, gender matched) normal controls on the various cognitive and behavioral tasks. This is the only way to determine the validity of the tasks used in terms of their activation properties and the normal brain areas involved in the task. Specifically, the healthy age matched controls are critical for this study as there is no other way to validate the fMRI findings with the hippocampal and executive function activations tasks. Normal matched controls are needed for proper analysis of both the neurocognitive and the neuroimaging data. For the neurocognitive data, the normal control data are needed to construct the Reliable Change Index (RCI) referenced below in the Statistics Section. Also, a comparison to normal is important for estimating the magnitude of any change observed in the patients, to determine if it has clinical relevance. For the neuroimaging data, the normal controls are needed to provide a reference or comparison point to quantify the degree to which the brain-imaging map (referred to as a Statistical Parametric Map, SPM) at each point in time (baseline, post-treatment) deviates from normal. By comparing the patient groups to normals, fMRI activation related to ancillary or idiosyncratic factors during the scan cancel out, leaving only the activation directly related to the cognitive task. Also, determining the presence of new, atypical brain activation sites, lost activation sites, or deactivations, can only be achieved by have a normal SPM brain map for comparison. This applies equally for the task-driven fMRI and the resting state functional connectivity data. To analyze within-subject analyses with the neuroimaging data, we will examine the neuroimaging data on a within-subject but that will not be the primary quantitative approach to the imaging data, as such within-subject comparisons can be prone to methodologic problems. There is a phenomenon known in the neuroimaging literature as "double dipping" which concerns the ability to distinguish real versus random variation over time. Double dipping refers to the use of test findings at one point in time to constrain or bias the data outcomes and changes registered at a second point in time. More concretely, differences seen in the imaging data may reflect changes in the error, artifactual activations, not change in true-task related activation. To avoid this problem we operationalize change by comparing each patient's imaging data with an index of their deviation from healthy normals at each point in time (pre-surgery/baseline, and post-surgery). This procedure provides a systematic means of quantifying and then comparing the typicality of both pre and postsurgical neurocognitive activations and network organization, a procedure that alleviates concern about "double dipping". This "double dipping" phenomenon and the remedy we propose applies equally for the task-driven fMRI, the resting state functional connectivity data, and the DTI data. In essence, the methodology proposed minimizes reliance on random, session-specific fluctuations as part of the reference or comparison point for determining change. Our method quantifies the deviation of our patients from normal, examining the relative difference in those deviations as our index of change. In so doing, we avoid the use of our pre-treatment neuroimaging results as the template for comparing and determining change, minimizing concerns of double dipping. There is no risk associated with the scanning procedure so long as there is no metal in the body. The study of normal volunteers also allows for determination of the stability of the activation findings over time; this allows for verification of the reliability of the fMRI tasks and other procedures over time. Without an understanding of the potential normal change that can occur in brain activation and structure over time, it is impossible to determine if a pathologic change has occurred. This group of normal subjects will be studied in advance in order to validate the fMRI tasks. Furthermore, normal control subjects will be yoked or linked to matched patients and return for additional scanning sessions at the same time points as their matched patient.