Antiretroviral Therapy Regimens with Superior CNS Penetration Did Not Reduce Risk for HIV Dementia

Abigail Zuger, MD reviewing Caniglia EC et al. Neurology 2014 Jun 6. Berger JR and Clifford DB. Neurology 2014 Jun 6.


In fact, high-penetration regimens actually increased that risk.


Some antiretrovirals penetrate the blood-brain barrier considerably better than others, and “high-penetration” regimens, as defined by a standard central nervous system (CNS) penetration-effectiveness (CPE) score, are linked to lower cerebrospinal fluid viral loads. However, whether such regimens carry clinical benefits — specifically, a reduction in risk for HIV dementia and other HIV-associated neurologic disease — remains unclear.

To explore this issue, investigators pooled results from nine prospective cohort studies in Europe and the U.S. involving >60,000 antiretroviral-naive, HIV-infected adults without clinical AIDS who began treatment after January 1, 1998. Sixty-two percent of them started a regimen considered to have low CNS penetration (most commonly efavirenz/tenofovir/FTC; CPE=7), 29% started a medium-penetration regimen (most commonly efavirenz/AZT/3TC; CPE=9), and 9% started a high-penetration regimen (most commonly nevirapine/AZT/3TC; CPE=10).

During follow-up (median duration, 37 months), diagnoses of HIV dementia were quite rare (235 cases in total). However, individuals in the high-CPE group were about 70% more likely to be diagnosed with HIV dementia than those in the low-CPE group. In contrast, common opportunistic CNS infections such as toxoplasmosis, cryptococcal meningitis, and progressive multifocal leukoencephalopathy occurred with similar frequency across all CPE groups. The analysis controlled for CD4 and viral load values, and additional analyses using refinements of the CPE scoring system and excluding dementia diagnosed within the first year of treatment did not alter the results.


The authors and editorialists struggle to explain these counterintuitive findings, which imply that high-CPE regimens actually predispose to HIV-associated neurologic disorders. Are high-CPE regimens more difficult to tolerate, with more nonadherent (and thus effectively undertreated) patients? Do these regimens cause CNS inflammation and clinical illness on that basis? Are they more likely to be prescribed by physicians who are more apt to pick up subtle neurologic findings? Until further data corroborate these results, clinicians are encouraged to concentrate on choosing well-tolerated, effective regimens, regardless of their CPE scores.



  2. Berger JR and Clifford DB.The relationship of CPE to HIV dementia: Slain by an ugly fact? Neurology 2014 Jun 6; [e-pub ahead of print]. (

– See more at: