Share this article on:

Why Antiretroviral Drugs May Be Causing Dementia in HIV-Positive Patients

Valeo, Tom

doi: 10.1097/01.NT.0000452271.33610.5e
Back to Top | Article Outline


Investigators reported that HIV patients taking antiretroviral drugs that achieved a high penetrance of the central nervous system developed dementia 74 percent more often than patients taking drugs that achieved a low penetrance.

HIV patients taking antiretroviral drugs that achieved a high penetrance of the CNS developed dementia 74 percent more often than patients taking drugs that achieved a low penetrance, according to a study published in the June 6 online edition of Neurology.

The study included 61,938 antiretroviral therapy naïve individuals in the HIV-CAUSAL Collaboration, an effort by six European countries and the United States to answer three questions: When should antiretroviral therapy begin? Which regimen should be used first? When should drugs be switched?

The penetrance of each patient's drug regimen was determined by the Central Nervous System Penetration Effectiveness (CPE) Score, which ranks drugs on a scale of 1–4, with 4 having the highest penetration of the CNS. The CPE scores of the drugs in each patient's regimen were added and categorized as low if less than 8, medium if 8 or 9, and high if greater than 9.

The participants were followed for a median of 37 months. During that time participants developed 235 cases of HIV dementia, 169 cases of toxoplasmosis, 128 cases of cryptococcal meningitis, and 141 cases of progressive multifocal leukoencephalopathy.

“We found that initiating an antiretroviral regimen with a high CPE score increases the risk of HIV dementia compared with a low CPE score, but we did not find this for other AIDS-defining neurologic conditions,” said lead author Ellen Caniglia, a student in the department of epidemiology at the Harvard School of Public Health. “Our findings were unexpected and they should be interpreted cautiously. The underlying mechanisms by which antiretroviral penetration could cause HIV dementia remain unknown.”

How might antiretroviral drugs cause an increase in HIV dementia? One mechanism proposed by Caniglia and her colleagues in their paper is that antiretroviral penetration might stimulate the deposition in the brain of beta-amyloid, believed to trigger Alzheimer's disease (AD), “but this hypothesis requires further research,” Caniglia said.

Back to Top | Article Outline


The results create a dilemma for clinicians, according to Christina M. Marra, MD, professor of neurology at the University of Washington School of Medicine.

“I don't know which way to turn,” she said. “I saw a patient recently — a 55-year-old man who is profoundly demented — who has had HIV for 20 years. He is on a high CPE regimen, and has really well controlled virus in his blood, but is his high CPE regimen helping or hurting him? Should I change him to a regimen with a lower CPE? What data can I use to help me make that decision?”

Although the study considered age, which may intensify the neurotoxic effects of antiretrovirals, Dr. Marra suggested that other factors could have been considered as well, including previous head injury, vascular disease, drug abuse, or co-infections such as hepatitis C that might have increased the risk of dementia. Knowing that “would be so helpful to me as a clinician,” she said.

She also wondered if the study might suffer from “confounding by indication,” with some patients receiving a regimen based on their cognitive symptoms. “For example, if patients starting an antiretroviral regimen complain of being forgetful, even if they don't meet the criteria for dementia, the doctor may prescribe a regimen that gets into the CNS better, thinking that will help more,” she said. “These patients would get a high CPE regimen and then go on to develop dementia, which would suggest that the cause of the dementia was the CPE regimen when it fact they were destined to get dementia anyway. That would be confounding by indication. It's a huge problem with an observational trial, and one that's not insurmountable.”

The study authors attempted to get around this problem by analyzing the subset of individuals who were enrolled before 2008, when the first CPE scoring system was published, “but information on drug penetration has been known earlier than 2008,” Dr. Marra said.

In addition, failure to comply with a high-CPE drug regimen due to headaches, nausea, bad dreams, and other side effects also could be a confounding factor, according to Dr. Marra.

“Maybe that's why those with a higher CPE score got demented — they weren't taking their drugs,” she said. “The authors included plasma viral load in their analysis, but restricting the analysis to those with well-controlled peripheral virus would address this potential confounder, and would have made the results more clinically relevant to the provider.”

What's needed now, according to Dr. Marra, is a large controlled trial in which HIV patients are randomized to a high or a low CPE regimen, and are followed for at least a year.

“This study adds to a growing body of data showing that CNS drug penetration is a two-edged sword,” Dr. Marra said. “It can be good or bad. What I really need is help in figuring out when they're good and when they're bad for individual patients,” she said.

Back to Top | Article Outline


Previous research has hinted at a link between high CPE scores and neurocognitive problems in HIV patients. In a 2009 paper in the journal AIDS, Dr. Marra and her colleagues found that antiretrovirals with high penetration of the CNS were associated with poorer neurocognitive outcomes.

“Our paper was one of the first to talk about high CPEs as being neurotoxic, and everybody got so angry,” she said. “Our paper was still observational in that we didn't assign the antiretroviral regimens, but we selected a group of people who were highly likely to be neurocognitively impaired and were either switching their medicines or starting a new regimen, and we followed them for a year. Those on a high CPE regimen did worse neurocognitively than people who got a low CPE regimen.”

Also, a 2010 study in Neurology found that “neurocognition improved significantly following antiretroviral treatment (ART) discontinuation.”

However, the absolute rate of HIV dementia reported in the Neurology paper — 235 cases out of 61,938 patients — is extremely low, said Igor Grant, MD, director of the HIV Neurobehavioral Research Program at the University of California, San Diego (UCSD).

“In the literature generally, people have rates of at least mild HIV neurocognitive disorders up to 40 percent,” said Dr. Grant, distinguished professor and chair of the department of psychiatry at UCSD. “Even if you use narrow criteria it's still 10–15 percent. A difficulty in interpreting the report is that the criteria used to ascertain the main outcome — HIV dementia — are unclear. One way they could get the low prevalence they report would be if they used the criterion of such severe cognitive impairment that patients were almost incapacitated. How likely is it really that an antiviral regimen would produce profound dementia? We've seen no evidence in the literature, and I cannot imagine what the mechanism for producing this fairly profound neurological change might be.”

While the brains of these HIV patients may not handle beta-amyloid normally, the accumulation of the protein probably does not fully explain the increased dementia, said David B. Clifford, MD, Melba and Forest Seay professor of clinical neuropharmacology in neurology at Washington University in St. Louis. Brain scans of HIV patients with cognitive impairment do not show the type of amyloid buildup found in AD patients, said Dr. Clifford, “but changes in the immune system with aging may be exacerbated by some of the therapies we use for HIV. One of most consistent findings at present concerning cognitive impairment in HIV patients is that it has a tighter association with vascular risk factors such as diabetes, hypertension, cholesterol, and so on than it does with CD4 counts and viral loads. There does seem to be a trend to move toward those risk factors and away from the risk factors more directly related to HIV.”

Also, HIV patients seem to display accelerated aging, Dr. Clifford added. “So the degenerative conditions that plague old age might come on earlier in HIV-treated population,” he said. “To be honest, I'm not convinced that's happening, but there are reasons to be concerned about it, so we're examining that. It still looks like there's a disproportionate impairment at the ages at which we see HIV patients compared to the rest of the population.”

The correlation between high CPE scores and HIV dementia surprised Ned Sacktor, MD, Director of the Johns Hopkins NIMH Center for Novel Therapeutics of HIV-associated Cognitive Disorders Clinical Outcomes Core, and professor of neurology at Johns Hopkins Memory and Alzheimer's Disease Treatment Center. He suspects that the older antiretroviral drugs may in fact be neurotoxic.

“The majority of patients who started treatment after 2004 were given newer antiretroviral drug combinations including tenofovir,” he said. “Tenofovir-based regimens may be better at suppressing viral replication than some of the older regimens that were used at the higher penetrating combinations, which were predominately zidovudine-based (Retrovir) regimens. That would make it more effective and better, and less likely to cause dementia. The high-penetrating arm included predominantly patients from 1999–2003 who were using older combinations of drugs that may be less effective than the newer drugs. That could account for why we're seeing worse cognitive outcomes with higher penetrating combo.”

The incidence of HIV dementia has plummeted since the advent of antiretroviral drugs, Dr. Sacktor noted. “It occurred in up to two-thirds of advanced AIDS patients who lived long enough and didn't die of one of the opportunistic problems,” he said. “It was extremely common. It's vanishingly rare in the current era of effective anti-retroviral therapy.”

And as Dr. Marra observed, “clearly we help people with antiretrovirals more than we hurt them. Even if they're more likely to become demented, the risk of death from HIV and from opportunistic infections is much higher in patients with advanced and untreated HIV. We need to finely tune this and figure out who's susceptible and who's not.”

Back to Top | Article Outline






Back to Top | Article Outline


•. Caniglia EC, Cain LE, Hernán MA, et al. Antiretroviral penetration into the CNS and incidence of AIDS-defining neurological conditions: The HIV-CAUSAL Collaboration. Neurology 2013; E-pub 2014 Jun 6.
    •. Marra CM, Zhao Y, Clifford DB, et al. AIDS Clinical Trials Group 736 Study Team. Impact of combination antiretroviral therapy on cerebrospinal fluid HIV RNA and neurocognitive performance. AIDS 2009;23(11):1359–66.
      •. Robertson KR, Su Z, Margolis DM, Krambrink A, et al. A5170 Study Team. Neurocognitive effects of treatment interruption in stable HIV-positive patients in an observational cohort. Neurology 2010;74(16):1260–6.
        •. Previous research on the link between HIV and cognitive decline in Neurology Today:
          •. More research on the link between HIV and cognitive decline in Neurology:
            © 2014 American Academy of Neurology