Algorithmic Approach to Neuroimaging in AIDS

摘要

Since the recognition of the human immunodeficiency virus (HIV) in the 1980s, there has been increasing awareness of the neurologic complications of the disease. The increased longevity of patients with the disease, in part secondary to multi-drug regimens, as well as the increased incidence of the disease, has furthered the opportunity for radiologic assistance in the diagnosis and management of neurologic complications from HIV. It is estimated that 7-10% of patients with acquired immune deficiency syndrome (AIDS) present with neurologic symptoms as their initial manifestation of illness and nearly 75% have central nervous system (CNS) involvement on post-mortem examination. The clinical diagnosis of neurologic illness in AIDS patients can be a daunting task. Multiple illnesses affecting the CNS have overlapping clinical presentations without specific diagnostic laboratory findings. HIV encephalitis, Progressive Multifocal Leukoencephalopathy (PML), toxoplasmosis, lymphoma, and cryptococcus have clinical presentations ranging from subtle personality and mental status changes, difficulties in concentration, depression and memory loss to profound confusion which progresses to dementia. Headaches, motor and sensory neurologic deficits and new onset seizures are also common findings. Laboratory findings such as cerebrospinal fluid (CSF) leukocytosis may indicate an infectious etiology, but are non-specific, as are measurements of CSF protein and glucose counts. We propose a simple algorithmic approach to the neuroradiologic imaging findings that may assist the clinician and radiologist in the diagnosis of the neurologic complications of HIV+ patients.