Pharmacokinetic and tolerability profile of twice-daily saquinavir hard gelatin capsules and saquinavir soft gelatin capsules boosted with ritonavir in healthy volunteers


Dr Michael Kurowski, HIV-Laboratory c/o Auguste-Viktoria Krankenhaus, Rubensstr. 125, D-12157 Berlin, Germany. Tel: + 49 (30) 7903 3920; fax: + 49 (30) 7903 3922; e-mail:



To evaluate the pharmacokinetics and safety of a boosted saquinavir (SQV)/ritonavir (RTV) combination, administered as either the hard gelatin capsule (HGC) or soft gelatin capsule (SGC) formulation of SQV, in 24 healthy volunteers.


This was a single-centre, open-label, randomized, 2 × 2 crossover study. Twelve subjects were randomized to receive SQV/RTV 1000 mg/100 mg twice daily (BID) orally for 10 days, as either the HGC or SGC formulation. The pharmacokinetic profile of SQV was determined on day 10. Subjects then crossed over to the opposite SQV formulation, and the pharmacokinetic profile was determined again on day 20. The primary analysis was the assessment of bioequivalence based on logarithmically transformed values for AUC(0−24 h) and Cmax for the two formulations.


There was a statistically significant increase in the geometric means of all the pharmacokinetic variables evaluated for SQV-HGC/RTV compared with SQV-SGC/RTV. A mean AUC0−24 h-value of 15.798 µg/mL/h was reported for the HGC formulation compared with 11.655 µg/mL/h for the SGC formulation (P = 0.0043). The SQV-HGC/RTV combination was better tolerated in terms of gastrointestinal system disorders. Furthermore, no elevations in triglycerides or total cholesterol were reported with SQV/RTV during the entire study period.


In healthy volunteers, RTV boosting of SQV-HGC produces plasma exposures at least comparable to SQV-SGC, which is accompanied by an improvement in gastrointestinal system disorders.


Saquinavir (SQV) is a well-established component of current highly active antiretroviral therapy (HAART) regimens and has demonstrated potent and sustained viral load suppression, and is well tolerated [1–3]. More recently, the acceptance of pharmacokinetic enhancement, or boosting, as a means to increase protease inhibitor (PI) exposures has resulted in clinical trials investigating boosted SQV/ritonavir (RTV) 1000 mg/100 mg twice-daily (BID) and 1600 mg/100 mg once-daily regimens. Preliminary results, such as those from the 24-week interim analysis of the ongoing MaxCmin I trial, have shown that these boosted SQV/RTV regimens provide potent virological suppression and appear to be well tolerated [4–7].

Saquinavir is available in two formulations; the original hard gelatin capsule (HGC) formulation, Invirase®, and the soft-gelatin capsule (SGC) formulation, Fortovase®. Unboosted, the bioavailability of SQV with the HGC formulation is lower, around 4% when administered as a single 600 mg dose, than that observed with the SGC formulation (relative bioavailability of 3331%) [8,9]. This difference in bioavailability is mainly due to the fact that the glyceride excipient component of the SGC formulation, capmul, allows SQV to dissolve and disperse rapidly upon administration [9]. However, when SQV is boosted with the PI RTV, the pharmacokinetic parameters of SQV are increased such that twice- and once-daily dosing of SQV are possible [10], and the bioavailability of SQV from either of these formulations (SGC and HGC) becomes much more comparable [4,11,12].

Since the capmul component of the SGC formulation of SQV is associated with potential adverse events such as diarrhoea, it may be beneficial for patients in some circumstances to use the HGC formulation in a boosted regimen. To verify this improvement in bioavailability with RTV boosting, a pilot bioequivalence study was initially conducted in HIV-1-infected patients. Encouragingly, the pharmacokinetic data showed that the boosted SQV/RTV 1000 mg/100 mg BID combination produced comparable SQV exposures with either the SGC or HGC formulation [13]. However, there was some baseline imbalance in this initial pilot study. Patients taking the HGC formulation had lower plasma exposures to SQV at the early pharmacokinetic assessment time points, while at 12 h SQV plasma levels appeared to be similar for the two formulations. This difference is thought to be due to the time the last dose of medication was administered; patients were not observed taking their medication the evening before their planned hospital visit for pharmacokinetic assessments [13]. Hence, it was difficult compare the pharmacokinetic data obtained in this initial pilot study. Therefore, a second, randomized, bioequivalence study with a more stringent methodological protocol was undertaken in healthy volunteers, to confirm these earlier findings. The results of this second study to assess the pharmacokinetic and tolerability profile of a boosted SQV/RTV 1000 mg/100 mg BID combination given as either the SGC or HGC formulation, in healthy volunteers, are presented here.


Study design and treatments

This was an open-label, randomized, 2 × 2 crossover study evaluating the pharmacokinetic and tolerability profile of the two formulations of SQV (HGC and SGC [1000 mg BID]) when given in combination with RTV (100 mg BID) to healthy male volunteers (aged ≥ 18 and ≤ 70 years). The study was conducted at a single centre in Poland, between May and July 2001. Exclusion criteria included acute infection or malignancy, impaired liver function, and treatment with any concomitant medication.

Subjects were randomly assigned to either group A or group B, after initial screening. In group A, subjects received SQV-HGC/RTV (1000 mg/100 mg BID) orally for 10 days followed by SQV-SGC/RTV (1000 mg/100 mg BID) orally for a further 10 days. In group B, subjects received SQV-SGC/RTV (1000 mg/100 mg BID) orally for 10 days followed by SQV-HGC/RTV (1000 mg/100 mg BID) orally for a further 10 days.

Subject adherence to the study protocol was closely monitored by the study coordinator. All study medication was administered in the presence of the study coordinator. Mouth checks were performed following drug administration and all subjects were observed for a further 30 min.

On day 10, the pharmacokinetic profile of SQV was determined after an overnight fast. Blood samples (5 mL) for SQV analysis were taken at: 0, 0.5, 1, 2, 3, 4, 6, 8, 10, 11.5, 12, 12.5, 13, 14, 15, 16, 18, 20, 22, 23.5 and 24 h, following SQV/RTV administration. The next day subjects crossed over to the opposite SQV formulation (those previously receiving SQV-HGC switched to SQV-SGC and vice-versa), and the pharmacokinetic profile was determined again 10 days later. On the 2 days of the pharmacokinetic assessments, SQV and RTV were taken with meals at 0 and 12 h. In addition, data on adverse events were collected each day following the morning drug administration and lipid levels (total cholesterol and triglycerides) were assessed at screening, and on days 1, 10, 20 and 28.

This study was performed in accordance with the principles of Good Clinical Practice, as set out in the Declaration of Helsinki and its amendments. Ethical approval was obtained from the local ethics committee. All participants provided written informed consent.

Pharmacokinetic evaluations

Saquinavir was isolated from human plasma by solid phase extraction and determined by tandem mass spectrometry using deuterated SQV as an internal standard [14]. The lower limit of quantification for SQV was 0.4 ng/mL. SQV concentration–time profiles were constructed, from which 24-h pharmacokinetic parameters, including peak plasma concentrations of SQV (Cmax), trough plasma concentrations of SQV (Cmin) and the area under the plasma concentration–time curve (AUC) of SQV, were determined by noncompartmental methods.

Statistical analysis

The primary analysis was the assessment of equivalence based on logarithmically transformed values for AUC and Cmax for the combined data for each treatment regimen from the two treatment groups. The null hypothesis (H0) was that the antilogged 90% confidence intervals (CIs) for the mean ratio SQV-HGC:SQV-SGC were not included in the equivalence region of 0.8–1.25.

To determine whether the treatment order (HGC to SGC or SGC to HGC) had any effect on the pharmacokinetic parameters determined, mean differences between the SQV formulations for group A and group B were compared in terms of log10 transformed Cmin, Cmax and AUC values. A P-value of < 0.05 was considered statistically significant.

The incidence of adverse events was examined using logistic regression on drug type (SQV-HGC vs SQV-SGC) and plasma drug levels (AUC0−24). Again, a P-value of < 0.05 was the level of statistical significance.


Patient characteristics

A total of 24 male subjects were enrolled: 12 subjects each in groups A and B. All 24 subjects completed the study. The mean age was 24.6 years.

Saquinavir pharmacokinetics

Data from Groups A and B were combined for SQV-HGC/RTV and SQV-SGC/RTV. Cross-over test analyses to assess the effect of treatment, however, did indicate a significant treatment sequence effect for the pharmacokinetic parameters Cmin 0−12 and Cmin 0−24 (P = 0.0429 and P = 0.0484, respectively).

Mean combined SQV plasma concentrations vs. time profiles for subjects receiving SQV-HGC/RTV or SQV-SGC/RTV are presented in Fig. 1. When the data were combined, different absorption phase and elimination phase pharmacokinetics were seen for SQV-HGC/RTV and SQV-SGC/RTV.

Figure 1.

The 24-h pharmacokinetic profile of saquinavir (SQV) for subjects receiving SQV as the hard gelatin capsule (HGC) formulation or soft gelatin capsule (SGC) formulation in combination with ritonavir twice daily (geometric mean with 90% CI).

The primary analysis was the assessment of bioequivalence based on AUC and Cmax for the two formulations. The 90% CIs around the mean ratio SQV-HGC:SQV-SGC fell within the predefined equivalence regions as specified by the primary analysis (region of 0.8–1.25) and hence bioequivalence between the two formulations could not be disproved [Table 1]. The geometric mean AUC0−12, AUC12−24, AUC0−24, Cmin 12, Cmin 24, Cmax 0−12 and Cmax 12−24, values are presented in Table 1. In terms of relative bioavailability, there was a statistically significant increase in the geometric means of all the pharmacokinetic variables evaluated for SQV-HGC/RTV compared with SQV-SGC/RTV when data from groups A and B were combined [Table 1].

Table 1.  Comparison of 24-h saquinavir (SQV) pharmacokinetic parameters following SQV hard gel capsule(HGC)/ritonavir (RTV) or SQV-soft gel capsule (SGC)/RTV (1000 mg/100 mg BID)
ParameterSQV-HGC/RTVSQV-SGC/RTVBioequivalence resultsP-value
(group A + B combined)(group A + B combined)Mean (95% CI)*(t-test)
  • *Values back transformed from log10 data.

  • **

    Geometric means.

Cmin 12µg/mL (90% CI)0.217**0.153**0.0004190.0147
(0.163, 0.290)(0.127, 0.185)(0.000130, 0.000780) 
Cmin 24µg/mL (90% CI)0.232**0.166**0.0004000.0143
(0.181, 0.298)(0.139, 0.198)(0.000126, 0.000741) 
Cmax 0−12µg/mL (90% CI)1.318**1.001**0.0003500.0023
(1.093, 1.590)(0.862, 1.163)(0.000162, 0.000569) 
Cmax 12−24µg/mL (90% CI)1.228**0.983**0.0002880.0087
(1.006, 1.499)(0.850, 1.137)(0.000107, 0.000498) 
AUC 0−12µg mL−1 h (90% CI)7.979**5.745**0.0003890.0020
(6.452, 9.868)(5.021, 6.573)(0.000182, 0.000632) 
AUC 12−24µg mL−1 h (90% CI)7.792**5.894**0.0003220.0112
(6.274, 9.678)(5.183, 6.703)(0.000111, 0.000572) 
AUC0−24µg mL−1 h (90% CI)15.798**11.655**0.0003560.0043
(12.764, 19.554)(10.233, 13.275)(0.000150, 0.000598) 


All adverse events were considered to be mild or moderate, with no severe or serious adverse events reported. Overall, 22 and 24 subjects reported at least one adverse event during the SQV-HGC/RTV and SQV-SGC/RTV treatment period, respectively. Gastrointestinal system disorders occurred most frequently and all reported adverse events are presented in Table 2.

Table 2.  Summary of adverse events by body system – all subjects
Body system
adverse event
phase only
phase only
Both phases
  1. GI, gastrointestinal system; CNS, central nervous system.

GI system
 Abdominal distension0414
 Abdominal pain428
 Light abdominal
  muscle aches

The number of subjects reporting gastrointestinal system disorders was lower in the SQV-HGC/RTV treatment period compared with the SQV-SGC/RTV treatment period (eight vs 23 subjects). Diarrhoea, abdominal distension and vomiting were most frequently reported during the SQV-SGC/RTV treatment period. Furthermore, the incidence of diarrhoea was significantly higher during the SQV-SGC/RTV treatment period compared with SQV-HGC/RTV (15/24 vs 4/24 subjects, P < 0.01). However, when the incidence of gastrointestinal adverse events was examined for a correlation with SQV plasma drug levels, no relationship between the incidence of diarrhoea or abdominal symptoms was shown (P > 0.05 for both comparisons).

Median values for lipid levels (total cholesterol and triglycerides) assessed throughout the study period are presented in Fig. 2, which shows values for both SQV groups combined. No elevations in triglycerides or total cholesterol were reported with SQV/RTV.

Figure 2.

Total cholesterol and triglycerides levels for all subjects receiving saquinavir/ritonavir 1000 mg/100 mg twice daily (median ± standard deviation).


Overall, this study in healthy male volunteers shows that improved SQV exposures were obtained when SQV/RTV 1000 mg/100 mg BID was administered using the HGC formulation compared with the SGC formulation. In addition, the SQV-HGC/RTV combination appeared to be the better tolerated combination in terms of gastrointestinal system disorders.

Although unboosted the bioavailability of SQV with the HGC formulation is lower than that with the SGC formulation, several studies, including the pilot bioequivalence study in HIV-infected patients, have indicated that boosting with RTV results in increased SQV exposures [4,11–13]. Further investigations undertaken in this study show that SQV exposures with the boosted HGC are slightly, but significantly, higher than those with the boosted SGC formulation. The HGC formulation provided higher SQV exposures for all the pharmacokinetic parameters assessed. In particular, a mean AUC0−24 h-value of 15.798 µg/mL/h was reported for the HGC formulation compared with 11.655 µg/mL/h for the SGC formulation (P = 0.0043).

The improvement in SQV exposure presented here with RTV boosting of the HGC compared with the SGC formulation is similar to that reported in a pharmacokinetic study in 13 HIV-1-infected patients receiving SQV/RTV 1600 mg/100 mg once-daily [15]. A median AUC0-24 h-value of 50 µg/mL/h was reported for the boosted SQV-HGC formulation compared with 35.5 µg/mL h for the boosted SQV-SGC formulation (P = 0.056) [15]. Although these AUC0−24 values in HIV infected patients were slightly higher than those reported in our study, this is a well-recognized phenomenon, with differences in pharmacokinetic values seen between healthy volunteers and HIV-infected patients [16,17].

In addition, the improvement in SQV exposure with the boosted HGC reported here was not seen in the initial pilot study in HIV-infected patients [13]. This may have been due to the baseline imbalance observed in the initial pilot study, which was probably seen because patients were not monitored taking their study medication the evenings prior to their hospital visit. However, in our study in healthy volunteers, adherence to study medication was closely monitored throughout the entire study period.

It has been postulated that the increased SQV exposure obtained with the boosted SQV-HGC/RTV (1000 mg/100 mg BID) combination may be related to the order of dissolution of the RTV and SQV capsules. The SQV-SGC formulation may dissolve rapidly, being absorbed and metabolized before RTV has an opportunity to act on the P-glycoprotein transport and cytochrome P450 systems. In contrast, the SQV-HGC formulation may dissolve slowly, thus providing time for RTV itself to dissolve and to inhibit P-glycoprotein transport and the cytochrome P450 metabolic pathway before SQV is released, thus facilitating SQV absorption and distribution. However, if this is the case, it is unclear why Tmax for SQV-SGC is reached at a later time point than for the SGV-HGC [Fig. 1], and further investigations are required to confirm this hypothesis.

A potential drawback of the study was that a significant treatment sequence effect was observed for the pharmacokinetic parameters Cmin 0−12 and Cmin 0−24. This observation may be explained by both the lack of a washout period and the time taken to achieve steady-state inhibition of CYP3A4 by RTV.

Inhibition of metabolism by RTV is generally highest during the first few days of drug administration, leading to the highest plasma concentrations of saquinavir. However, after an initial adaptation process, some of the inhibition effect of RTV on the metabolism of saquinavir is compensated for and the effect of RTV on saquinavir plasma concentrations, although still apparent, is less pronounced. This adaptation process involves autoinduction of the 3A4 enzymes of cytochrome P450 system, and reaches a steady state 10–14 days after the start of RTV administration [18]. This steady state may, therefore, have not been achieved until the second PK evaluation (at 20 days).

Both SQV formulations were well tolerated. Overall, the incidence and severity of adverse events reported were in accordance with the side-effect profile of SQV, where gastrointestinal disorders were the most commonly observed. The SQV-HGC/RTV combination was better tolerated in terms of gastrointestinal adverse events than the SQV-SGC/RTV combination, with a lower incidence of diarrhoea and vomiting. The differences observed in the tolerability profile of the two SQV formulations were comparable with those in a study of 171 antiretroviral-naïve, HIV-1 infected patients [19]. Mitsuyasu et al. showed that dyspepsia, vomiting and abdominal pain were all more frequently observed with SQV-SGC [19].

The increased incidence of adverse events reported here with the SQV-SGC/RTV combination may be due to capmul, the glyceride excipient component of the capsule formulation of SQV [20]. This hypothesis is supported by the fact that, in this study, there was no significant correlation between SQV plasma drug levels and the incidence of diarrhoea or abdominal symptoms (P > 0.05 for both comparisons) reported. It has also been postulated that the nonabsorbed portion of the drug may be responsible for the diarrhoea. Hence, a negative correlation with plasma drug levels may exist, although this was not examined in our study.

Furthermore, an ongoing study in 69 antiretroviral-experienced patients who previously received SQV-SGC 1400 mg BID plus two nucleoside reverse transcriptase inhibitors (NRTIs) for 162 weeks, and were switched to a SQV-SGC/RTV 1600 mg/100 mg once-daily regimen, showed that at 48 weeks, this regimen was well tolerated [21]. Despite the increased SQV exposures obtained, no increase in the incidence of gastrointestinal disorders was reported [6,21]. Adverse events are usually related to the drug dose; however, this study illustrates that toxic effects can also arise from the excipient components of drug formulations.

Long-term HIV antiretroviral therapy is associated with changes in fat distribution and metabolism [22]. In one such study, the effect of different NRTIs on fasting lipid levels revealed that patients receiving stavudine had significantly elevated triglyceride levels compared with those receiving zidovudine and lamivudine in combination with indinavir, over a period of 48 weeks [23]. Furthermore, a switch study involving 23 patients receiving SQV/RTV 400 mg/400 mg BID who switched to a SQV/RTV 1000 mg/100 mg as part of a four-drug BID regimen, showed reductions in fasting triglyceride levels and cholesterol levels following the switch over 6 months [24]. Changes in triglycerides with SQV/RTV combinations containing RTV doses of 200 mg and above have been reported as early as three days after the initiation of therapy in healthy volunteers [11]. Importantly, the findings presented here show that the SQV/RTV 1000 mg/100 mg BID combination did not appear to have any effect on triglycerides and total cholesterol values, over a 20-day period.

In conclusion, in healthy volunteers, RTV boosting of SQV with the HGC formulation produced improved SQV plasma levels at least as high as that achieved with the SGC formulation. In addition, it appears that in terms of gastrointestinal system disorders, the HGC may be the better-tolerated formulation of SQV. This would suggest that RTV boosting of the HGC formulation may be considered to be as acceptable as the boosted SGC formulation for inclusion in HIV-treatment regimens. Furthermore, the HGC formulation may be a better therapeutic option for those patients who experience problems with the SGC formulation, including those who are particularly vulnerable to gastrointestinal disorders, or have difficulties with the larger size of the soft gelatin capsules or their need for refrigeration. Importantly, the findings presented here in healthy volunteers may be of clinical relevance and warrant further evaluation in HIV-infected individuals.


This study was supported by a grant from Roche Pharmaceuticals. This information has been presented in part at the 9th Conference on Retroviruses and Opportunistic Infections, Washington, Seattle, USA, 24–28 February 2002 (Abstract 432).