Recent ARVs and the blood/brain barrier: CSF drug concentrations of darunavir/r and raltegravir

The issue of drug penetration into the cerebrospinal fluid (CSF) is an increasing focus for all antiretrovirals, but has particular importance given the interest in nucleoside-sparing regimens.

These include studies looking at boosted-PI monotherapy as a maintenance strategy after initial viral suppression with triple drug therapy and dual NNRTI + PI/r or raltegravir + PI/r combinations.

Scott Lettendres group presented encouraging results from two studies at ICAAC: on darunavir/r and on raltegravir, supported by Tibotec and Merck respectively.

The darunavir study measured 29 CSF and plasma pairs from 16 HIV-positive patients between August 2006 and August 2008, and compared levels to the median IC50 for wild-type virus (2.75 ng/mL). [1]

Participants were a median 48 years, 62% were Caucasian and 19% had HCV co-infection. Median CD4 cell count was 197 cells/mm3. Viral load was detectable in 38% of blood and 10% of CSF samples. Median duration of darunavir was 7.5 months (IQR 3.6 – 14.6).

Darunavir was detected in all CSF specimens with a median concentration of 56.9 ng/mL (IQR 39.6 – 81.4). The median total plasma concentration (AUC) was 4,094 ng/mL (IQR 2,993 – 6,410) with a median CSF-to-plasma ratio of 1.4%. The median unbound plasma concentration was 542 ng/mL (IQR 376 – 971) with a median CSF-to-plasma unbound ratio of 9.4% (IQR 6.8 – 14.2%).

DRV concentrations in CSF exceeded the IC50 of wild-type HIV in all specimens by a median of 20-fold.

The raltegravir study had a similar design, with 22 matched plasma/CSF samples from 18 HIV-positive patients. Demographics were also similar: median age 46 years, 89% Caucasian, 12% HCV coinfection with a median CD4 of 276 cells/mm3. [2] The median raltegravir inhibitory concentration (IC50) for wild-type HIV is 3.4 ng/mL.

Raltegravir was present in all CSF specimens with a median concentration of 14.5 ng/mL. The median plasma concentration was 260.9 ng/mL (IQR 2.0 – 640.4) with a median CSF-to-plasma ratio of 5.8% (IQR 2.1%-17.8%). CSF concentrations correlated with plasma concentrations (r = 0.49, p = 0.02) but not with post-dose sampling time (p > 0.50). Raltegravir concentrations in CSF exceeded the IC50 of wild-type HIV in all specimens by a median of 4.3-fold (IQR 2.7-7.7). HIV RNA levels were undetectable in 20 of 21 (95%) CSF specimens and in 13 of 21 (62%) plasma specimens.

Each study concluded that the study drug penetrated CSF in concentrations that are in the therapeutic range to suppress wild-type HIV and would be expected to contribute antiviral activity in the CSF as part of combination therapy.

Drug levels of both drugs in CSF correlated better with total levels in plasma suggesting that therapeutic drug monitoring could indicate effectiveness in the nervous system. Unbound plasma concentrations of darunavir were less closely correlated.