Seizures can occur at any stage of HIV infection and were the presenting symptom in 18% of patients as described by Holtzman, Kaku, and So.¹ Seizures as the presenting symptom may reflect advanced HIV disease, and thus may be associated with poor outcome, as noted by Aronow, Brew, and Price,² who reported that 7 of 7 patients with status epilepticus died within 1 month of the presenting episode. Initial data regarding HIV-associated seizures suggested that approximately 50% of new-onset seizures were related directly to HIV and the remainder were secondary to HIV-associated complications. However, current thought reflects the more dynamic view that multiple processes may occur simultaneously and overlap significantly.³

New-onset seizures in HIV-positive patients are most commonly generalized tonic-clonic and not infrequently (13%) involve status epilepticus upon presentation.⁴ A retrospective review of all patients with new-onset generalized seizures presenting to St. Vincent’s Hospital in New York identified 26 patients who were known to be HIV-positive. Of this cohort, 31% (8 patients) were determined to have idiopathic seizures; in another 31% (8 patients), the seizure etiology was judged to be HIV encephalopathy. The remaining etiologies were determined to be central nervous system (CNS) toxoplasmosis (5 patients, 18%), alcohol withdrawal (2 patients, 8%), progressive multifocal leukoencephalopathy (2 patients, 8%), and CNS lymphoma (1 patient, 5%). A comparison group of 120 patients without HIV infection was judged to have idiopathic causes (43 patients) or alcohol-related seizures (29 patients) as the two most common diagnoses. Of interest, of the 6 HIV-positive patients with mass-occupying lesions requiring admission to the hospital, only 2 presented with focal signs on admission that would have suggested the need for hospital admission under then-existing seizure
guidelines written for patients without HIV infection, which highlights the need for a thorough evaluation for new-onset seizures in patients with known or suspected HIV disease.⁵

Electroencephalograms (EEGs) in patients with HIV infection and acquired immuno-deficiency syndrome (AIDS) frequently demonstrate epileptiform features, including sharp waves, spikes, and focal slowing associated with epileptic foci. AIDS dementia complex (ADC) frequently demonstrates low-amplitude, slow, monotonous EEG patterns that are quite indistinct from patterns identified with other types of dementing illness.⁶,⁷ Gabuzda, Levy, and Chiappa⁷ demonstrated the presence of frequent EEG abnormalities identified in a series of patients with AIDS, with the most common pattern being generalized slowing (38%), compared to unremarkable patterns (37%), focal slowing (19%), and epileptiform activity (6%). Among those patients with a clinical diagnosis of seizure disorder, half demonstrated sharp wave patterns and another 17% demonstrated focal slowing.⁷ The presence of focal slowing on an EEG in these patients is especially important, because such patterns can be derived from an underlying seizure disorder as well as the presence of focal abnormalities, including tumor and infection. Accompanying neuroimaging is almost always advisable in such instances and remains a mandatory part of the initial evaluation of most new-onset seizures.⁸

In contrast to patients with active HIV infection and AIDS, there is conflicting evidence to indicate that asymptomatic HIV infection may or may not lead to EEG abnormalities. Nuwer et al.⁹ conducted visual and quantitative EEG testing in 200 asymptomatic homosexual men, half of whom were HIV-positive, and found abnormalities or borderline slowing in 32%. However, EEG changes did not appear related to HIV serostatus, but rather to underlying impaired neuropsychological test performance. The authors used this evidence to highlight the importance of further evaluation in patients with HIV infection and an abnormal EEG, rather than cursorily attributing the EEG changes to an effect of the serostatus of the individual.⁹ This finding was replicated by Tinuper et al.,¹⁰ who noted that EEGs correlated with CNS involvement and that neurologically asymptomatic individuals demonstrated no abnormal tracings. These data clearly strengthen the approach that borderline or abnormal EEGs should be further investigated before assigning seropositivity as the primary determinant of the abnormal rhythm.

Quantitative EEG (qEEG) results in patients with HIV disease have more consistently demonstrated the evolution of electrophysiologic abnormalities across the HIV disease spectrum; however, caution should be appropriately used when examining qEEG studies, because qEEG is an investigative technique for research purposes and has not demonstrated a routine clinical role in the evaluation or management of HIV-infected patients or patients with neurologic disease. Nevertheless, the results provide interesting data and demonstrate the promise of an increasing role for these techniques in the foreseeable future.

Quantitative techniques across multiple studies appear to confirm the essential absence of clinical EEG abnormalities in asymptomatic HIV-positive patients compared to healthy controls, and they have confirmed this pattern using methods suitable for statistical analysis, which strengthens the data further.¹¹ Quantitative EEG measures have also served to strengthen the concept of HIV disease as having a disproportionate impact on subcortical processes and eventually producing a subcortical-type dementia. At least one study examined the relationship between cerebral metabolism and brain electrical activity in patients with AIDS by the use of positron emission tomography and qEEG and found that data were suggestive that coherence abnormalities may reflect the presence of subcortical AIDS-related disease, but perhaps more importantly supported the validity of coherence measures in studies of AIDS-associated neuropsychiatric dysfunction.¹²

There is preliminary evidence that qEEG may be used potentially to monitor antiviral effectiveness in the CNS. Baldeweg et al.¹³ demonstrated that in a group of asymptomatic HIV-1–positive patients administered zidovudine versus placebo, patients receiving zidovudine did not experience the progressive increase in delta and theta slow-frequency qEEG activity over the study period that was identified in the placebo group. Long-term studies examining the role of other pharmacologic agents involved in highly active antiretroviral therapy (HAART) and their potential CNS protective effects are needed, and data correlating measured qEEG abnormalities to clinical or neuropsychological deficits are required for optimal interpretation of available qEEG data.¹³

qEEG changes in patients who have already progressed to AIDS or AIDS-related complex (ARC) have high rates of clinical abnormalities, and abnormal rates have been reported to be as great as 67%. Two of the most common etiologies of abnormal EEGs include AIDS-related dementia and opportunistic infections, with AIDS dementia associated with intermittent or continuous slowing, whereas focal slowing or sharp activity was more reflective of underlying neuronal irritability associated with focal CNS processes such as toxoplasmosis and lymphoma. These data demonstrate that EEG may be especially helpful in the differential diagnosis of dementia versus other mental status changes.¹⁴ A careful investigation of all observed clinical EEG abnormalities is required to adequately detect and diagnose the myriad of complications of this disease.