Clinical implications of intracellular protease inhibitor concentrations
     M Hennessy ¹ , M Barry ¹ , S Clarke ² , C Bergin ² , S Khoo ³ , D Back ³ , FM Mulcahy ^2,1 Department of Clinical Pharmacology and Therapeutics and ² GUIDE Clinic, St James’ Hospital, Dublin, and ³ Department of Pharmacology, University of Liverpool

Objective:
Highly active antiretroviral therapy (HAART) still lacks sufficient potency and durability. As viral replication takes place intra-cellularly, treatment success may depend upon intracellular (IC) PI concentrations. We examined the relationship between plasma and IC PI concentrations in peripheral blood lymphocytes (PBLs) of HIV patients.

Results:
Patients receiving indinavir (IDV; n=11) and nelfinavir (NFV; n=14) were studied. PBLs were isolated by density centrifugation, stained and counted. Plasma and IC PI concentrations were assayed by LC/MS. AUC, time to peak (T max ) and half-life (t 1 /2) were derived. Data were correlated with HIV plasma RNA levels. IC IDV AUC was lower than plasma (P<0.05; 9470±2239 vs 31788±6392) while NFV underwent a ninefold IC accumulation (P<0.05; 283251±6687 vs 31034±6911). Two patients receiving IDV had IC levels below the MEC despite acceptable plasma levels and were unsuppressed. In contrast, two patients with subtherapeutic plasma IDV levels had IC concentrations above the MEC and remained suppressed (VL<50cpm). IC IDV T max was delayed compared with plasma. IC t 1 /2 was longer, thus the mean residence time of IDV within the cell was prolonged (P<0.05) this may contribute to efficacy despite poor IC accumulation.

Discussion:
For discordant patients, IC IDV concentrations may provide better correlation with virological response than plasma levels. All patients on NFV remained suppressed. This highlights marked differences between PIs with respect to IC accumulation and pharmacokinetic behaviour and has important implications for drug penetration into sanctuary sites and therapeutic drug monitoring. The correlation between plasma and IC AUC was r 2 = 0.6 (P<0.05), in contrast to NRTIs.