These cases therefore have significant implications for the prognosis and management of patients with JCV CNS infection. Although PML often improves with HAART, our findings suggest this population, and others with PML, may be at risk of PML recurrence. Serial monitoring clinically and by MRI may allow early identification. There is no proven PML therapy besides immune reconstitution (Tan et al. 2009; Koralnik 2006), which is of particular concern given that our patients were not immunosuppressed anymore. However, should PML recurrence be identified in other groups such as natalizumab or fumarates-treated patients, current immunomodulating and immunosuppressive therapies may need to be reconsidered.