• The CSF Concentration/IC50 Ratio: A Predictor of CNS Antiretroviral (ARV) Efficacy (Poster 614)
  Authored by S. Letendre, R. Ellis, I. Grant, A. McCutchan, and the HNRC Group
  View the original abstract

The central nervous system (CNS) is a potential reservoir for HIV and a site where inadequate drug levels may lead to the development of resistance. In addition, HIV-associated neurologic problems are dependent, to some extent, on the CNS viral load. As a result, a goal of antiretroviral (ARV) therapy should be to obtain good penetration into all reservoirs, including and especially the CNS, and complete viral suppression in these sites.

Due to varying abilities to cross the blood-brain barrier, ARVs differ greatly in their penetration of the CNS. ARVs that reach higher levels in the CNS are likely to be more effective at controlling HIV replication in that site. This study proposes that clinicians consider using a ratio of the antiretrovirals cerebrospinal fluid (CSF) concentration to its inhibitory concentration for 50% of viruses (IC50).

The authors examined the CSF concentrations (CSF conc) of all FDA-licensed ARVs and compared that to the IC50 ranges for each of those drugs, as determined by the Virologic Phenosense Assay (CSF conc/IC50). Not unexpectedly, they found that CSF conc/IC50 ratios varied widely among the ARVs. Of the nucleoside reverse transcriptase inhibitors, zidovudine had the most favorable ratio (22.00) and zalcitabine the worst (0.08). Among the non-nucleoside reverse transcriptase inhibitors, nevirapine was best (24.00) and efavirenz the worst (5.7). Indinavir had the best ratio of the protease inhibitors (108) and ritonavir, nelfinavir and saquinavir were virtually the same (<5.0).

Based on these findings, the authors advocated the use of ARVs with high CSF conc/IC50 values and urged clinicians to consider phenotyping to determine the IC50 of HIV in the CSF. I am not sure that either of these is absolutely necessary. First, there are many factors that may determine a drug's effectiveness in the CSF besides a simple CSF conc/IC50 ratio, for example, an ARV's intrinsic potency, its overall pharmacologic barrier, and its CSF protein binding may have a substantial effect, regardless of a low CSF conc/IC50 ratio. Further, several studies, including some presented at this conference, have shown excellent suppression of HIV in the CSF, especially in naive patients treated with three drugs that achieve relatively high CSF levels. So, while the point made by this study is certainly one clinicians should consider, I am not sure that it provides them with any new or important information in making the decision regarding which ARVs are best for treatment.