NATAP REPORTS
January 1998 Volume 1, Number 3-4, Section 7

CSF Penetration by Protease Inhibitors

BID Dosing. Merck, Agouron and Roche are researching their protease inhibitors in a twice-a-day regimen. Merck reported preliminary 24 week data was encouraging, but they caution it is premature to experiment using indinavir twice a day. They are conducting additional research with indinavir 1200 mg every 12 hours in combination with nucleosides.

Preliminary 16 week data was reported at Hamburg for nelfinavir 1250 mg twice-a-day in combination with d4T and 3TC. Again, the data is encouraging but it may be premature to experiment; a larger study is ongoing in Europe to evaluate the bid regimen. Roche is just starting its trial to see if Fortovase, the new formulation of saquinavir, can be dosed twice a day. See NATAP’s web site for the actual data from these studies.

Effect of Indinavir on CSF. At the 8th European Congress of Clinical Microbiology and Infectious Diseases in May ‘97, Courtney Fletcher (University of Minnesota, PhD), reported findings from evaluating indinavir pharmacokinetics (PK) in HIV infected children receiving indinavir with ddI+d4T. Indinavir penetrated the CSF in 4 children measured with concentrations of 151-977 ng/mL.

Dr. Anne Collier of the University of Washington reported at IDSA in early September ‘97 her preliminary findings from two groups of study participants who underwent lumbar punctures. Investi-gators found that indinavir penetrated the CSF in study participants. Investigators measured actual indinavir drug levels at various time points. An important point is that individuals in group B were not taking indinavir at baseline but then added indinavir. So the effect of indinavir on the CSF could be evaluated by comparing the CSF HIV RNA before and after starting indinavir therapy. In group A, all individuals had been on indinavir therapy before the CSF HIV RNA was measured, so there was no pretreatment measure of CSF HIV RNA for comparison.

Of the 10 individuals in group A, 7 were taking indinavir+2 NRTIs, 1 was taking indinavir + 1 NRTI, 2 were taking indinavir alone. 9/10 individuals had <200 copies/ml CSF HIV RNA. Only 4/10 had plasma HIV RNA <200 copies/ml but the median plasma HIV RNA was 2.4 log (about 250 copies/ml, ranging from <200 to 5.4 log- 250,000 copies). The median CD4 count was 328.

In group B, 9 individuals had CSF and plasma viral load measures at baseline prior to taking indinavir therapy and at week 8 after all 9 started indinavir therapy. 7/9 were taking 2 nucleosides at baseline, 1 was taking 1 RTI and 1 person was not taking any therapy. Then, 5 individuals simply added indinavir to their current regimen, 2 changed 1 RTI, and 1 person started 3 new drugs - 2 new RTIs and indinavir. See Table 17

Table 17. Group B changes in CSF HIV RNA and Plasma HIV RNA

Most individuals in group B were not taking optimal therapy as we know it today since 5 simply added indinavir to current NRTI therapy, etc.

The results from this study as well as those from the ritonavir/saquinavir study (below) are preliminary and need further study for confirmation. Collier said this was an initial observational pilot study but results suggest that indinavir penetrates into the CSF of adults. There are several confounding factors making it difficult to determine the absolute effect of indinavir on CSF HIV RNA.

Some researchers postulate that the absolute drug concentration in the CSF is more important than the CSF-plasma ratio of the drug. In this study only one drug level measure was taken for each participant, but they were taken at different time points during the dosing cycle. Essentially, the absolute drug levels of indinavir were found to be above the IC95 level identified in vitro to be 100 nM of indinavir at all these various time points. The IC95 drug level is the amount of drug necessary to inhibit 95% of virus replication. In one case the blood level was less than 100 nM, but investigators believed because there is no protein binding in the CSF, the drug level in that case would be adequate.

Median indinavir CSF ratios ranged from 2.2 to 76%, with the highest ratios seen near the end of the dosing period. The CSF to plasma ratio of HIV RNA was found to be higher later in the dosing interval of 8 hours. That is because the plasma indinavir levels decline towards 8 hours but the CSF indinavir level remains the same, so the percentage changes.

It is uncertain if CSF penetration by a drug means that it penetrates the brain and effects HIV brain-related disease. Studies to explore this question are planned. Individuals with signs of early HIV related-brain disease will be treated with a potent protease inhibitor regimen to examine the effect of therapy. Dr. Justin McArthur, a leading researcher in this field at Johns Hopkins, has stated that HIV brain disease is reversible if detected and treated early enough (prior to neuron death), but at some point (after neuronal death starts) it becomes irreversible. So, you need to evaluate individuals with early disease to detect therapy effect.

Another way in which we should be able to determine the effect of HAART on brain disease, and this is also ongoing, is simply by following individuals receiving HAART over long periods of time and to see if less HIV brain-related disease develops.

CSF Study:Ritonavir+Saquinavir+D4T vs Ritonavir+Saquinavir.

We have been receiving results of ongoing research on the effect of HAART on opportunistic infections and viral load in lymph tissue. Now, research is beginning to turn to the effect of HAART on the CSF and the brain. I think determining this effect may be the most difficult. But some of the initial preliminary research findings are reported and reviewed in this section. In our July issue we reviewed nevirapine and nucleoside CSF penetration. Ongoing and future research will address how HAART affects viral load in the CSF and HIV related brain disease.

At Hamburg, data was reported from the Prometheus Study on the first 18 weeks of treatment for 104 individuals randomized (open-label) to ritonavir+saquinavir or the triple regimen of d4T combined with ritonavir+saquinavir. Aside from evaluating serum HIV RNA and the peripheral CD4 responses to therapy, there was a substudy to look at the effect of study treatments on CSF HIV RNA, and to measure levels of the drugs in serum and CSF. Participants were protease inhibitor and d4T naive.

All study medications were taken every 12 hours. If serum viral load was not undetectable by week 18, participants were permitted to add NRTIs. See Tables 18, 19, 20, & 21.

Table 18. Ritonavir+Saquinavir: Baseline characteristics

*Individuals who had NRTI treatment experience were pretreated for a median of 28 months.
** These are relatively low baseline viral loads

Table 19. HIV RNA and CD4 changes at Week 18 from Baseline

* undetectable is <400 copies/ml. At week 18 the graph line for the number of participants with undetectable HIV RNA was heading upward.

Table 20. Ritonavir+Saquinavir - Week 0 vs Week 12 changes in serum and CSF viral load

Table 21. Ritonavir and Saquinavir absolute serum and CSF drug levels
(concentration in ng/ml)

LLQ is below the level of quantification (25 ng/ml)

Two participants who intensified their study regimen of ritonavir+saquinavir with 3TC/d4T because of a detectable viral load at week 18 had an undetectable viral load weeks after intensification.

Safety. Discontinuations of study medications were permitted for up to 2 months. 7 participants permanently discontinued study medications because of clinical (4) or laboratory (3) toxicities: elevated liver enzymes (2), 1 with grade 4; gastrointestinal side effects (2), nausea/diarrhea; peripheral neuropathy (1); skin rash (1); thrombocytopenia grade 4 (1). Main laboratory abnormalities: elevation of liver enzymes - 10% of patients had a grade 3 or 4 elevation of SGOT any time during the study; elevation of triglycerides - 25% of patients had a grade 3 or 4 toxicity.

CSF Sub Study. 9 study participants volunteered for lumbar punctures, which were performed at weeks 0 and 12. HIV RNA was measured at both time points and drug levels in CSF were measured at week 12. One person (Pt#2, RTV/SQV) had a partial serum response (detectable) and a complete CSF response (undetectable). Patient #8 (d4t/RTV/SQV) reached undetectable serum HIV RNA but not in CSF HIV RNA. 64-69% of the overall study population had a mean duration of prior therapy of 28 months but were protease and d4T naive.

The authors stated that the reason for the low drug levels in CSF (and the low ratios) are because of the high protein binding of ritonavir and saquinavir in serum. Nevertheless, RTV/SQV (± d4T) appears "highly effective in the central nervous system." Abbott officials say, the IC50 of ritonavir in the absence of protein is 42 ng/ml. Since there is very little protein in the CSF, one would expect that low levels of ritonavir or saquinavir would have antiviral effect. The IC50 is the amount of drug (the inhibitory concentration) needed to inhibit 50% of virus replication.

One concern is that although a PI may penetrate the CSF, will the amount of drug that penetrates be adequate to achieve the suppression necessary to prevent replication and resistance? Some researchers have raised the concern that if inadequate amount of drug penetrates and there isn’t enough overall viral suppression from the combination regimen resistance mutations could develop.

Back to top