Double Protease Combinations
Reported by Jules Levin, Executive Director of NATAP, (March 6, 1998)

Group 1:
Treatment arm A: 400 mg bid RTV + 400 mg bid SQV (n=35), every 12 hours
                      B: 600 mg bid RTV + 400 mg bid SQV (n=35), every 12 hours

Group 2:
                      C: 400 mg tid RTV + 400 mg tid SQV (n=33), every 8 hours
                      D: 600 mg bid RTV + 600 mg bid SQV (n=37), every 12 hours

The median number of antiretroviral drugs previously used was 2 for each of the 4 treatment groups.The median baseline plasma HIV RNA ranged from 4.5 log (31,600 copies/ml) to 4.7 log (50,100 copies/ml) for each of the 4 groups. The baseline median CD4 counts ranged from 266-300 for each of the 4 groups.

The two arms: 400/400 bid and 600/400 bid- were both more tolerable than the other two arms

The week 60 data includes only individuals who remained on study drugs. Study dropouts were not included. It is important to bear in mind that a number of the participants originally randomized to Groups C or D changed their dosing regimens to A or B. The actual number of such changes follow the table. 27 participants added up to 2 NRTIs (usually d4T/3TC) after week 12 (see Table 6). Although data is reported for all 4 arms, it is generally recommended that the dosing regimens to be used are those used in arms A or B. The incidence of side effects is higher in arms C and D. Also, arms C and D do not appear to be any more efficacious than arms A and B, but noncompliance may have been higher in arms C & D.

As you can see almost half of participants were randomized or switched to 400/400 bid by week 24. By week 60, 51% had switched to 400/400 bid.

All participants who were on study at week 48 and had sufficient stored plasma available had HIV RNA measured by the Roche Ultrasensitive test (lower limit of detection 50 copies/ml). 86% (66/80) who were <200 copies/ml by the standard assay also were <50 copies/ml using the Ultrasensitive test.

15 participants who were <200 copies/ml for at least 8 weeks after about 1 year on ritonavir-saquinavir alone and did not add NRTIs had CSF evaluations. Although this data is encouraging, CSF viral load was taken only once which was after being on therapy for >1 year. There was no CSF evaluation taken prior to study entry. Therefore there was no baseline measure taken to compare to the measure taken during therapy, and to see if there was an actual change and to measure the change. 14/15 (93%) had CSF <400 copies/ml. 1 person on RTV-SQV 600-600 bid had a plasma viral load <200 copies/ml but had a CSF viral load of 650 copies/ml. In the Prometheus Study, both baseline and during therapy CSF measures were taken for individuals on RTV/SQV and RTV/SQV plus nucleosides. That data is below.

The incidence of all side effects including those related to the liver were increased for those receiving the dose regimens in arms C and D. For that reason, the dose regimens in arms A and B are recommended. Abbott prefers to recommend the arm A dose regimen of 400/400 bid because of the lowest incidence of side effects compared to the other dose regimens; and, because the CD4 increases and viral load reductions appear equal to those of the other dose regimens.

The 600/400 bid regimen (Group B) has certain pharmacokinetics characteristics you may want to consider. Because you are taking 50% more ritonavir (vs 400/400 bid regimen), the peak, trough and AUC blood levels for both ritonavir and saquinavir are higher compared to if you were taking 400/400 bid. Is that an advantage? I think it offers some potential benefits. If you were taking ritonavir without saquinavir, 400 mg ritonavir would be suboptimal. However, the 60 week data shows that 400/400 bid was equivalent to 600/400 bid as measured by CD4, viral load reductions and percent below detection. But, having higher trough levels of a drug at the end of a dosing period creates a comfortability or cushion, but tolerability and consideration of long term effects are factors.

Some individuals, soon after starting ritonavir+saquinavir regimen, may develop significant elevations in LFTs to the 500-600 level. Rather than immediately discontinuing therapy over concern about those elevations, trying to work through the elevations while continuing on same regimen with very close monitoring may be successful. The elevations may peak and slowly decrease over time to a manageable level.

• 11% (16/141) of patients developed grade 3 or 4 (1500 mg/dl) elevations in triglycerides levels
• 6/16 were treated with antihyperlipidemic agents (drugs that may lower triglyceride levels)
• No cases of pancreatitis have been observed

At week 48, the mean increase from baseline in triglyceride levels for each of the 4 dose regimens were about: +175 for both Groups A and B (400/400 bid, 600/400 bid); +290 for Group D (600/600 bid); +260 for Group C (400/400 tid). The suggestion is that the regimens in A & B are less likely to cause greater increases in triglycerides.

In the week 24 data report (see the Double Protease Combinations report on the NATAP website) the investigators reported a compliance assessment of participants up to week 24. They concluded compliance is highly correlated with treatment response. At week 24, 90% of those defined as being compliant were <200 copies/ml; 97% of those defined as compliant were <1000 copies/ml. Only 66% defined as noncompliant were <200 copies/ml and only 73% were <1000 copies/ml. See the actual report for the investigators definition of compliance.

This report also was issued at week 24 and is discussed in the above mentioned report on the NATAP website. Investigators showed data that of individuals with <1001 copies/ml at week 12, 98% had <200 copies at week 24; of individuals with >1000 copies/ml at week 12, only 25% had viral load <200 copies/ml. If an individual’s viral load was <201 copies/ml at week 12, 93% were <200 copies/ml at week 24. At the Chicago Retrovirus Conference, Lisa Demeter reported that data from ACTG 320 indinavir clinical endpoint study indicates that early viral load reductions also predict treatment response in that study.

NATAP reported preliminary data from the Prometheus Study that was presented at the Hamburg AIDS Conference. See the report on Double Protease Combinations on this website for details. A 24 week update was reported in Chicago. Investigators reported CSF viral load changes between week 0 and week 12 for 8 individuals receiving RTV/SQV alone and for 9 individuals receiving RTV/SQV+d4T. Unlike the CSF data above, this has a comparison between two time points (weeks 0 and 12).

• 138 participants were randomized in this open label study to receive either RTV/SQV (400 mg bid of both) or RTV/SQV+d4T
• there were 7 discontinuations; adverse events experienced were mild/moderate diarrhea (50%); oral parasthesia (50%)- tingling and numbness around mouth; elevated liver enzymes (28%) - more common in RTV/SQV/d4T arm; elevated triglycerides (25%)
• mean baseline CD4 was 273 and 251 for protease alone and d4T arms, respectively; mean baseline HIV RNA was 4.3 log (about 20,000 copies/ml) for both arms; 54% in protease alone arm had previous NRTI experience while 4& % in d4T arm had revious NRTI experience
• at week 24, 64% taking RTV/SQV alone and 87% taking RTV/SQV/d4T were <400 copies/ml (total # of evaluable patients at week 24 is111)
• if viral load did not reach undetectable by week 18, intensification was permitted; 6 patients in the RTV/SQV arm intensified with d4T/3TC and all 6 were undetectable by week 36
• mean CD4 increase was about +145 for both arms

As you can see, of the patients with detectable VL in serum and CSF at week 0, and undetectable VL in serum at week 12 1/3 on RTV/SQV has an undetectable VL in CSF at week 12, as compared to 4/5 in the RTV/SQV/d4T arm. The results from this study are preliminary; it may be too soon to draw conclusions and too small a number of patients at this point to draw conclusions. Week 12 may be too soon to evaluate the effect of a therapy on CSF viral load. In the CSF evaluation discussed above where 14/15 had undetectable CSF viral load, the CSF viral load levels were assessed after one year of therapy. The authors said that neither ritonavir nor saquinavir was found to penetrate the CSF well in 9 patients as measured by RTV or SQV CSF drug levels. But, ritonavir is 99% protein bound in blood, and if you were looking for unbound ritonavir in blood you may not be able to find it.

Back to top

No participants had current opportunistic infections at study entry. The median baseline HIV RNA was 39,917 copies/ml (range 19,496 - 109,065). The median CD4 count was 327 cells (range 19-621).

The investigators reported that from genotypic analysis of isolates drawn from all subjects in weeks 15-22 and weeks 20-35 no occurrence of the D30N mutation was observed. At month 12, the median decrease in HIV RNA was about 2.4 log (n=9) and the median increase in CD4 was about 100 (the increase at month 11 was 172), n=9. The percent <500 copies/ml was 80% at month 12 (n=9) and 90% at month 11. As you can see, the data is based on a small number of individuals.

** patient diagnosed with H. Pylori which may have contributed to dyspepsia

The second study called SPICE was reported by M Opravil and others and is a pilot randomized study enrolling 157 individuals comparing 4 treatment arms; the number randomized to each arm in parenthesis:

• SGC SQV 1200 mg tid + 2 NRTIs (A), (26)
• NFV 750 mg tid + 2 NRTIs (B), (26)
• SQV SGC 800 mg tid + NFV 750 mg tid + 2 NRTIs (C), (51)
• SQV SGC 800 mg tid + NFV 750 mg tid without NRTIs (D), (54)

Crossovers to other arms were permitted for intolerance or virological failure. Participants had to be able to start at least 1 new NRTI. The study investigators said the data should be considered preliminary until formal analysis at week 48.

All crossovers were to NFV+SQV+2 NRTIs.

Of 16 reported serious AEs, 4 were considered possibly treatment related:

• unstable diabetes (SQV+NRTIs)
• renal colic (SQV/NFV+NRTIs)
• diarrhea and cachexia (SQV/NFV no NRTIs)
• fever (SQV/NFV no NRTIs)

By 32 weeks, 24 patients crossed over to NFV+SQV+2 NRTIs (C): 6 for toxicity and 18 for virologic failure (2 from arm A, 5 from arm B, and 11 from arm D). The authors said 11 additional patients were defined as virologic failures but had not yet crossed over at the time of this analysis. The following analysis of data was based on individuals remaining on the therapy to which they were originally randomized (crossovers and discontinations were not included). The number of evaluable patients are in parenthesis.

Back to top

21 patients were initially enrolled. They were protease naive, CD4 ( 100, HIV RNA ( 30,000 copies/ml. The median baseline viral load was 50,500 copies/ml (range 9000-316,000); the median baseline CD4 was 259 (range 12-568); 58% had no prior nucleoside experience and the remaining 42% were NRTI experienced.

Group A received NFV 500mg every 12 hrs + IDV 1000 mg every 12 hrs (q12h) for one week. Starting in the second week they received NFV 750 mg q12h + IDV 1000 mg q12h. Group B started with NFV 750 mg q12h + IDV 1000 mg q12h.

Based on the data from Agouron’s single dose studies it was expected by some that those effects seen in that study (increased IDV AUC by 51%-single dose of IDV; increased NFV AUC by 83%-single dose of NFV) would be seen when combining NFV with IDV. The investigators in this study did not observe those PK responses reflected in this study. This indicates that you cannot generally rely upon single or limited dosing studies to predict drug interactions and optimal dosing regimens. To do so is risky. It is safer to wait for additional research identifying adequate dosing regimens.

Investigators in this study found that when combining 1000 mg IDV q12h with 750 mg NFV q12h: indinavir blood levels were similar to indinavir taken at the standard dosing of 800 mg every 8 hours; that indinavir did not increase NFV steady state blood levels, resulting in low NFV trough levels. At week 8, 7/10 study participants were <500 copies/ml, and 9 patients had a mean increase in CD4 of 156 cells. Treatment was discontinued in 1 person for rash; other adverse events included: diarrhea/loose stools (6), bloating (2), and nephrolithiasis (1); nephrolithiasis can lead to kidney stones.

Because of the low NFV trough levels, a higher dose of nelfinavir was investigated. All participants raised their dose of nelfinavir from 750 to 1000 mg q12h while maintaining IDV dosing at 1000 mg q12h.

For an explanation of terms like trough, AUC, Cmax and pharmacokinetics (PK) , see the report titled Primer on PK on the web site. Without understanding these terms it is difficult to understand the considerations that go into selecting dosing, and to understand the following tables and much of the other information in NATAP’s reports.

As you can see the trough and AUC (italics) for the NFV 750 q12h dosing regimen was not increased by the addition of IDV, compared to the trough and AUC values for taking NFV monotherapy every 8 hrs (trough- 0.7 vs 1.5 mg/L, AUC- 50 vs 56 mg hr/L).

The investigators reported the mean percent change resulting from increase in NFV dose of 750 mg q12h to 1000 mg q12h in AUC, Cmax and trough for these 5 patients was 31%, 27%, and 35%, respectively.

10/21 (47%) had undetectable HIV RNA by the Roche Amplicor viral load test (<400 copies/ml). 6 of these10 were undetectable using the Roche ultrasensitive test (<50 copies/ml). Three study participants added NRTIs to their regimen after at least 12 weeks either for virological (viral load) failure or to increase antiviral effect to prevent virological failure. The median increase in CD4 from baseline (12-32 weeks) was 133 cells. There were 5 discontinuations.

Testing of this combination is in early stages. Two small initial pilot studies are ongoing, exploring different dosing regimens of these two protease inhibitors. One study is exploring 400 mg ritonavir bid + 1000 mg nelfinavir + d4T/3TC, and has just started so it will be a while before data will be available. Investigators are examining the effect of therapy on CSF and lymph tissue, as well as performing immunological evaluations.

In Chicago, JE Gallant and C Flexner of Johns Hopkins and others reported preliminary findings from a small open-label, dose escalating, non-randomized trial in 20 individuals. Participants were protease inhibitor naive while 10 were NRTI experience and 10 were NRTI naive, but all were permitted to add NRTIs after week 12 at the discretion of the investigator.

The study is an initial assessment of safety, pharmacokinetics and antiviral activity of two dosing regimens (400mg RTV bid+500mg NFV bid, 400mg RTV bid+ 750mg NFV bid). Some reasons offered by the investigators for this combination being attractive include:

• Differing primary resistance patterns - D30N for nelfinavir and V82A for ritonavir
• Convenience of twice daily dosing
• Known enhancement of pharmacokinetics of NFV by RTV. RTV slows the metabolism of other protease inhibitors, increasing blood levels and prolonging halflife, because of its potent inhibition of the CYP3A system in the liver which metabolizes protease inhibitors as well as other drugs. It is my understanding RTV increases NFV blood levels by 2 to 2.5 fold but PK data from this study are currently being analyzed
• Since both drugs are highly protein bound, one drug may displace the other from its protein binding sites, especially on alpha1-acid glycoprotein

Assessment of antiviral activity is based on use of Roche Amplicor Monitor test (400 copies/ml detection limit) and the Roche Ultrasensitive Amplicor test (20 copies/ml detection limit). The median baseline HIV RNA and CD4 for all 20 patients were 32,459 (range 4,118->750,000 copies/ml) and 325 cells (range 23-941), respectively.

The 20 participants were divided into two groups: group 1 received 500mg NFV bid with 400mg RTV bid while group 2 received 750mg NFV bid with 400mg RTV bid. Each group dose escalated for several days before reaching the full dose regimens.

Group 1 mean baseline HIV RNA was 58,045 copies/ml (range 4,912 to 279,938); mean baseline CD4 was 335 (range 29-1042). 5 persons in Group 1 were treatment naive and 5 were experienced.

Group 2 mean baseline HIV RNA was 135,300 copies/ml 7,756-939,794); mean baseline CD4 was 329 (range 186-578). 5 persons in Group 2 were treatment naive and 5 experienced.

6/10 achieved <400 copies/ml and 3/10 were <20 copies/ml before adding NRTIs or NNRTI-

• 2 persons with undetectable viral loads (HIV RNA <20 copies/ml) added NRTIs at week 16 (d4T/3TC for one, NVP/3TC for other)
• 3 persons with low but detectable virus (HIV RNA 20-400 copies/ml) added NRTIs at week 16 (n=3): AZT/3TC for 2 and d4T/3TC for 1)
• 3 persons with detectable viral load (HIV RNA >400 copies/ml) added NRTIs at week 16: d4T/ddI in 2, d4T/3TC in 1)
• 1 person with undetectable viral load (HIV RNA <20 copies/ml) at week 16 remained on NFV/RTV without adding NRTIs.
• 1 person with 2 episodes of noncompliance and alcohol abuse (4 day drug holiday at week 10 and 3 day drug holiday at week 16) has withdrawn from study

Genotypic mutations are being assessed in persons experiencing failure.

9/20 persons experienced moderate-to-severe diarrhea

Back to top

• to evaluate the efficacy of 141 in combination with other protease inhibitors for the purpose of identifying which combinations to explore further
• to examine steady state pharmacokinetics for each combination. Data from previous research from which investigators have characterized a preliminary range for increases in Cmax of 16-31% and in AUC of 22-64% for 141 when used with indinavir. They reported no appreciable change in indinavir blood levels when used with 141. commentary: These increases, if confirmed in further studies, may contribute to the potency and effectiveness of the combination of 141+indinavir; as possibly reflected by the 3.75 log reduction seen with 141+IDV in the table below.
• to assess safety and tolerability.

Participants are protease naive but could be NRTI experienced. If so they washed out prior to receiving study drugs. CD4 cells >200; HIV RNA >10,000 copies/ml.

*The usual dose of 141 is 1200mg bid. But, 800 mg tid amounts to the same daily dose. The usual SQV SGC dose is 1200 mg tid. After three weeks, participants in Group D added AZT/3TC.

After 12 weeks patients can add NRTIs if their viral load is >400 copies/nl but so far no one has exercised that option. The first 3 weeks included intensive PK. Patients tolerating study drugs could continue out to 48 weeks.

Baseline Characteristics

• Median CD4 - 393 (n=33)
• Median HIV RNA - 4.63 log, n=33 (about 42,600 copies/ml)
• 21 patients had prior NRTI experience in past 6 months
• most patients had none or minimal symptoms

This analysis was using the 20 copy limit of the ultrasensitive assay. The number of patients starting therapy in each arm is small and the number of evaluable participants at week 16 is even smaller so the results are preliminary. The number of evaluable patients is smaller at week 16 than the number of patients starting in each group not because of dropouts but it represents the numbers moving towards the 16 week endpoint. There were 2 serious adverse events resulting in two dropouts. Because the number of participants are so small it is difficult to distinguish significant differences between the arms and it is difficult to draw many conclusions.

9/34 patients discontinued study drugs due to consent withdrawal, lost to follow-up, or in two cases serious adverse events.

Table 23. Adverse Events by Regimen (no. of patients) 

back to top

Additionally, Abbott’s second generation protease, ABT-378, is currently in phase II in HIV infected individuals. ABT-378 appears to have a relatively unique resistance profile, based on in vitro research done to date. It did not appear to have a mutation at position 82, which is important for IDV and RTV resistance. ABT-378 will be used with a small dose of RTV because the addition of ritonavir greatly enhances the PK profile of ABT-378. In vitro ABT-378 appears to suppress virus with some limited degree of ritonavir resistance (possibly up to 25 fold RT resistance ). As soon as possible, Abbott hopes to start trials with ABT-378 for individuals who’ve failed protease therapy. As well, Upjohn’s new protease, PNU-140690, may offer some effectiveness against protease resistant virus.  Upjohn has just started recruiting for a small study evaluating the effectiveness of PNU-140690 in individuals who have failed protease therapy. In one arm indinavir or ritonavir failures will receive PNU140690. The second arm is for individuals who have experience with multiple protease inhbitors. Individuals will receive PNU140690 if they've failed their current protease therapy and had previous protease therapy. The objective is to make a preliminary observation if individuals who've filed protease inhibitors will be sensitive to PNU140690. If all goes well, the hope is that by this Summer a larger study will begin exploring PNU140690's effectiveness for individuals who have failed protease therapy.

At the recent Chicago Retrovirus Conference, preliminary data from Gilead Science’s new drug called PMPA showed a -1.22 log reduction over the course of a 30 day study in HIV infected. PMPA is from a new class of drugs called nucleotides. It is not expected that PMPA would be at all cross-resistant with protease inhibitors. Although the development for PMPA in the early stages it is hopeful that it will prove effective.

In the January 1998 newsletter, NATAP Reports, we reported the findings of Dr Cassy Workman from Australia. She used a 6-drug regimen for individuals who failed protease inhibitors. She recycled previously used NRTIs for 12 individuals. The 12 persons had previously used all available nucleosides and 3 protease inhibitors (indinavir, ritonavir and saquinavir). But, they were naive to nevirapine and nelfinavir. The 6-drug regimen was-- d4T, 3TC, ddI, NVP, NFV, and SQV. 3 individuals were intolerant to components to the therapy that they had previously been intolerant to. But, 9/9 individuals had undetectable viral load (<400 copies/ml) at week 12. 12 weeks is too soon to draw conclusions but follow-up information will be reported.

Hydroxyurea (HU) is a potential option for therapy. Studies have shown that in combination with ddI, reductions in viral load can be significant. Hu has been studied with ddI, ddI+d4T, and with ddI+a protease inhibior. The actual results of these studies were reported in the January NATAP Reports and separately on this web site. There is no cross-resistance bewteen HU and protease inhibitors. There are 3 concerns attached to using HU therapy: (1) HU can lower the WBC (neutropenia), (2) it may have a negative effect on platelet count, and (3) although viral load may be reduced, proportionate increases in CD4 do not occur; CD4 increases may be blunted; for a discussion of this see the HU reports on the web site.

According to Franco Lori, the researcher pioneering the research into HU application in treating HIV, neutropenia may be a concern only for individuals who already have a problems regarding their WBC. The same may be true for platelets. If your CD4 is already low, the HU component to a therapy may not raise your CD4 count. But using HU in conjunction with other drugs may have the effect of raising your CD4 count. Consult with a knowledgable physician before making treatment decisions. Within the next few months several studies will start for individuals who’ve failed protease inhibitors.