Abstracts

Rationale: Progressive multifocal leukoencephalopathy (PML) is a demyelinating brain disease occurring in immunosuppressed individuals, caused by the polyoma virus JC (JCV). While PML predominantly affects white matter, the cortex can be involved. In previous retrospective studies, we observed a high incidence of seizures in PML patients, especially those with the immune reconstitution inflammatory syndrome (IRIS) and hyperintense cortical signal (HCS) on MRI. In this prospective observational study, we aim to determine the incidence and risk factors for seizures in PML long-term survivors. Methods: We enrolled patients with a diagnosis of PML according to consensus terminology criteria. At baseline, we collected demographics and PML history including symptoms at onset, etiology, CSF studies, JCV PCR, brain biopsy results, and CD4+ T-cell counts. At enrollment and at 3, 6, and 12 months, patients underwent neurologic exam, MRI of the brain (including arterial spin labeling (ASL) and MR spectroscopy) and 256-channel EEG. EEG was acquired on a Geodesics EEG system 400 with 256 channel Hydrocel Geodesic sensor net (Electrical Geodesics Inc., EGI). Epileptiform discharges were source-localized and co-registered to MRI MPRAGE images using BrainK and GeoSource software (EGI). Results: Twenty-nine patients were enrolled between January 2014 and May 2016. Enrollment and follow-up are ongoing. To date, 48% of patients have had only an enrollment visit, 24% of patients were followed for 3 months, and 24% for 12 months. Demographics and baseline PML characteristics are presented in Table 1. EEG showed focal or multifocal slowing in 21 patients (72%), both focal and diffuse slowing in 7 (24%), and no abnormalities in 1 (3%). Focal slowing was concordant with location of PML demyelinating lesions on MRI in 24 patients (83%). Epileptiform discharges were present on at least one EEG in 13 patients (45%), most commonly in temporal (10/13, 77%) and frontal (3/13, 23%) regions. Epileptiform discharges were concordant with PML lesion location in 12/13 patients (92%). A representative example of source localization results is shown in Figure 1. The presence of epileptiform discharges was not associated with a history of IRIS (spikes in 4 patients with and 9 without IRIS, p = .70). Conclusions: We report preliminary results of a prospective observational study of EEG findings and seizures in PML. We found focal EEG slowing in almost all patients that was typically concordant with MRI demyelinating lesions. Epileptiform activity was seen in nearly half of patients and was again concordant with lesion location on MRI. These findings suggest that the demyelinating lesions themselves are associated with cortical irritability and epileptogenesis. Additional follow-up will be performed to determine if the presence of epileptiform activity predicts the development of seizures in patients with PML. Funding: NIH R01 047029