A durable suppression of viral replication in HIV-1 infection can be achieved by giving a combination of three antiretroviral agents [1–3] Highly active antiretroviral therapy (HAART) can result in clinical benefit in terms of a prolonged (disease-free) survival [4,5] The use of more than three agents or a higher exposure to a specific antiretroviral agent might even increase the potency of an antiretroviral regimen [6–9] However, the occurrence of side effects may hamper the use of intensified treatment regimens [10,11]
Protease inhibitor combinations are pharmacokinetically favourable in antiretroviral therapy [12,13] All protease inhibitors show mild to moderate gastrointestinal side effects in addition to a rather unique safety and tolerability profile. Recently, the association between the use of protease inhibitors and the occurrence of a new onset of diabetes mellitus, hyperlipidaemia or peripheral lipodystrophy has been described  However, there is only limited data regarding the safety profiles of PI combinations, such as ritonavir or nelfinavir plus saquinavir [15–19]
A quadruple induction regimen, consisting of stavudine, lamivudine, nelfinavir, and saquinavir, was used for at least 26 weeks in antiretroviral therapy-na]ve patients in the Amsterdam Duration of Antiretroviral Medication (ADAM) study  In this study, the feasibility of an induction-maintenance strategy in HAART was investigated. The preliminary results of the efficacy of this induction-maintenance regimen have already been described  In this paper, we focus on the toxicity of the quadruple induction regimen used in the ADAM study and its relationship with the exposure to nelfinavir and saquinavir.
Materials and methods
The enrolment of HIV-1-infected patients in this open-label randomized-controlled study started in March 1997 and was prematurely discontinued on April 6, 1998, as the interim analysis demonstrated increased viral replication in patients on maintenance therapy  Patients, aged 18 years or more, were eligible for participation if they had a CD4+ cell count of at least 200 × 106/l in their peripheral blood, 1000 or more HIV-1 RNA copies/ml in plasma, and if they were antiretroviral therapy-na]ve. Exclusion criteria were the existence of an active opportunistic infection, active hepatitis C or presence of the hepatitis B surface antigen, women who were breast-feeding or pregnant, the use of immunomodulatory drugs or investigational drugs up to 1 month prior to the start of the study medication, and certain laboratory parameters [Hb < 7 mmol/l (male) or < 6.5 mmol/l (female), neutropenia < 0.75 × 9/l, aspartate-amino transferase (ASAT)/alanine-amino transferase (ALAT) > 5 × upper limit of normal (ULN), serum creatinine > 1.5 × ULN]. Drugs such as rifampin or ketoconazole, which have a strong pharmacokinetic interaction with protease inhibitors, were not allowed. Written informed consent was obtained from all patients. The Institutional Review Boards of all participating centres approved the study protocol.
All patients started therapy with a quadruple drug regimen consisting of stavudine (40 mg twice daily, or 30 mg twice daily for those with body weight < 60 kg), lamivudine (150 mg twice daily), nelfinavir (750 mg three times a day) and saquinavir hard- gelatin-capsules (saquinavir-HGC, 600 mg three times a day). When saquinavir soft-gelatin-capsules (saquinavir-SGC) became available (1 November 1997), all patients using saquinavir-HGC switched to saquinavir-SGC (800 mg three times a day). Roche (Roche NL, Mijdrecht, the Netherlands) provided both nelfinavir and saquinavir-SGC. Patients were instructed to take their medication with food.
During the induction phase, patients were scheduled to visit the outpatient clinic for clinical assessment and routine laboratory monitoring at the start of treatment and at weeks 1, 2, 4, 8, 16, 24, 25 (for plasma HIV-1 RNA concentration assessment only) and 26. Laboratory monitoring included plasma HIV-1 RNA concentration [Nuclisens HIV-1 RNA QT assays (Organon Teknika, Boxtel, The Netherlands)], CD4+ and CD8+ cell count and plasma concentrations of nelfinavir and saquinavir. At baseline no plasma concentrations of nelfinavir and sauqinavir were assessed.
The occurrence of side effects (signs, symptoms, or laboratory abnormalities) was assessed during each study visit. The severity of each event was graded according to the World Health Organization (WHO) classification (grades 1 to 4;Appendix 1) and the probability of a relationship with the study medication used was indicated (see also Appendix 1)  As the abnormalities in plasma cholesterol were not graded in the WHO classification, a more recently defined grading system by the AIDS Clinical Trial Group (ACTG) was used for the analysis of this parameter  The probability of a relationship with the study medication used was also indicated and the use of concomitant medication to relieve side effects was recorded.
Assessment of drug exposure
The quantification of plasma concentrations of nelfinavir and saquinavir was performed using a validated and sensitive reverse-phase high-performance liquid chromatography (RP-HPLC) assay  For each sample the drug plasma concentration was divided by the expected drug concentration at the corresponding time-point, to adjust for the time interval between drug ingestion and the drawing of the sample. The expected concentrations of nelfinavir or saquinavir at different time points were obtained from full (8 h) pharmacokinetic profiles of both drugs assessed in 18 patients participating in the ADAM study  These plasma concentration ratios were used as a measure of exposure to both protease inhibitors.
Treatment failure and discontinuations
After attaining a plasma HIV-1 RNA concentration below the quantification limit of the ultrasense assay (< 50 copies/ml), a plasma HIV-1 RNA concentration above 400 copies/ml at two consecutive time points was considered as a treatment failure.
In case of a grade 4 toxicity (WHO classification) or grade 3 toxicity with no improvement after temporary discontinuation (maximum 2 weeks), or the recurrence of a grade 3 toxicity after re-challenge, permanent discontinuation of the study medication was obligatory  Further therapy was at the discretion of the investigator after treatment failure or discontinuation of the study medication.
The baseline values and the changes from baseline to week 4, 8, 16 and 26 for plasma HIV-1 RNA concentration, and CD4+- and CD8+ cell counts of all patients were assessed. Patients with and without a specific side effect were compared for baseline values of these parameters and the change from baseline to week 26 (t-test).
The proportion of patients experiencing a specific side effect was assessed. In addition, the duration of the side effect within the first 6 months was calculated. Only side effects with an incidence of > 10% and a possible relationship with the study medication used were taken into account for analysis.
The correlation coefficient between nelfinavir and saquinavir per patient was assessed. For each patient, the median plasma concentration ratios of nelfinavir and saquinavir over time were calculated. To evaluate whether the plasma concentration ratio of nelfinavir and saquinavir varied over time, a generalized linear model with mixed effects was estimated (Mixed Models procedure of the statistical package SAS 6.12 for Windows, SAS Institute, Cary, North Carolina, USA).
The median plasma concentration ratios of nelfinavir and saquinavir of patients on treatment with or without a specific side effect were compared (Wilcoxon test). In addition to this comparison between groups of patients, the odds ratio for experiencing a specific side effect at a specific time point for the corresponding drug exposure was estimated by using the generalized estimating equations (GEE) method, taking into account the `within patient' correlation between time points  Time (as a factor) and plasma concentration ratio (as a covariate) were included in the model as main effects, as well as the interaction between these two variables. The analysis was based on the logistic link function and the exchangeable working correlation matrix. Calculations were performed by using the GENMOD procedure of the statistical package SAS 6.12 for Windows. To calculate P-values from the results of the GENMOD procedure, a SAS macro was written using the Wald statistic.
Patients and discontinuations
In total, 65 patients (61 males, 94%) were enrolled. The baseline characteristics of these patients are summarized in Table 1. Two patients were lost to follow-up (one for logistic reasons, one withdrew from further medical follow-up after moving to another city). One patient discontinued his study medication due to virologic failure at week 24. Of the remaining 62 patients, seven patients discontinued the study medication due to side effects within the first 26 weeks (see `Toxicity').
CD4+ and CD8+ cell count and plasma HIV-1 RNA concentrations during therapy
The median CD4+ cell count rose from 410 × 106 at baseline to 560 × 106 cells/l at week 26. The median CD8+ cell count decreased from 1050 × 106 at baseline to 970 × 106 cells/l at week 26. The median baseline plasma HIV-1 RNA concentration was 4.51 log10 copies/ml. At weeks 4, 8, 16 and 26, the median plasma HIV-1 RNA concentration declined to 2.58, 2.42, 2.25 and 1.54 log10 copies/ml, respectively.
The frequency of the occurrence of side effects, the median of the total duration of the side effect per patient in this population and the percentage with a grade 1, 2, 3, or 4 severity are summarized in Table 2. If the use of loperamide was longer than the duration of diarrhoea, the duration of diarrhoea was extended for the period during which loperamide was used and the severity was graded as 1.
Clinical side effects
Forty-nine of the 65 patients suffered from diarrhoea, mostly consisting of loose or watery stools, two or three times daily. Loperamide was used in 30 patients for a mean duration of 128 days (SD: 61) to relieve the complaint. Diarrhoea led to discontinuation of the study medication in one patient, after 8 weeks. Other side effects leading to discontinuation of the study medication were headache (one patient, week 22), and peripheral neuropathy, probably related to the use of stavudine (one patient, week 24). In addition to gastrointestinal side effects such as abdominal pain, nausea and abdominal distension, fatigue and headache were frequently reported. No peripheral lipodystrophy was observed within the first 26 weeks of treatment.
Abnormalities of laboratory values
Mild to severe elevations of the liver enzymes occurred in 12 patients. In four of these patients the elevated liver enzymes led to a discontinuation of the study medication (at weeks 4, 8, 16, and 20, respectively). The occurrence of elevated cholesterol in plasma was frequent; 23 out of 65 patients (35%). However, the cholesterol level rose above 7.7 mmol/l in only six of these 23 patients. Hypertriglyceridaemia was less frequent and did not exceed grade 2 severity. No hyperglycaemia or glycosuria was observed within the first 26 weeks of treatment.
Toxicity and patient characteristics
Patients with or without a specific complaint did not differ with respect to their baseline characteristics. Only patients complaining about fatigue had a lower baseline CD4+ cell count and a higher plasma HIV-1 RNA concentration compared to patients without this complaint (t-test, P = 0.02 and P = 0.03, respectively). The patients complaining of abdominal pain had a greater decrease in their plasma HIV-1 RNA concentration at week 26 than patients without abdominal pain (t-test, P = 0.02; data not shown).
The correlation coefficient between the nelfinavir and saquinavir plasma concentrations per patient was calculated. The mean correlation coefficient in this group of patients was 0.55 [95% confidence interval (CI) 0.41–0.67]; calculation of the confidence intervals was based on the Fisher transformation of the mean. As a result of unknown time intervals (e.g. either an unknown time of medication intake and/or the drawing of the sample), plasma drug concentration ratios could only be calculated at two or more time points in 56 patients. The median plasma concentration ratio over time per patient for nelfinavir and saquinavir varied between 0.11 and 1.31 for nelfinavir [median 0.60; inter-quartile range (IQR) 0.44–0.86] and between 0.11 and 6.50 for saquinavir (median 0.75; IQR 0.42–1.37). For both nelfinavir and saquinavir, the median plasma drug concentration ratios were significantly lower than 1 (95% CI 0.59–0.93 and 0.51–0.68, respectively); calculation of the confidence intervals was based on the geometric mean. The plasma concentration ratios of nelfinavir and saquinavir in these patients did not change over time (Fig. 1; analysis of repeated measures P = 0.76 and P = 0.68 for nelfinavir and saquinavir, respectively).
Toxicity and drug exposure
The median plasma drug concentration ratios over time of patients with and without a specific complaint, are tabulated in Table 3. Patients with abdominal pain appeared to have higher plasma concentration ratios for nelfinavir and saquinavir than patients without this complaint (Wilcoxon test, P = 0.02 and P = 0.05, respectively). The patients with diarrhoea had lower saquinavir plasma concentration ratios than patients without diarrhoea (Fig. 2; Wilcoxon test, P = 0.04). No differences in drug exposure were found for any of the other side effects.
Using the GEE method, the occurrence of diarrhoea was found to be associated with the exposure to nelfinavir and not with the exposure to saquinavir (Fig. 3a), even if the diarrhoea was not corrected by the use of loperamide (data not shown). For the other gastrointestinal complaints, there was evidence of a time-dependent relationship with the plasma drug concentration ratios (Figs 3b–d). Both abdominal distension and nausea became associated with a low exposure to nelfinavir and saquinavir after the first weeks of therapy and for nausea, this association seemed to increase over time (P < 0.001). The association between abdominal pain and a high drug exposure remained present during the entire treatment period (P < 0.001). For all other complaints, no evidence was found for a relationship with plasma drug exposure (data not shown).
The quadruple induction therapy in the ADAM study was quite well tolerated. Seven out of 65 patients had to discontinue their study medication within the first 26 weeks of therapy. The incidence and severity of the observed side effects were comparable with those reported in other studies [15,16,27,28] The most frequently reported side effect, diarrhoea, led to a discontinuation of the antiretroviral drug combination in only one patient. Although tolerable with the use of loperamide, diarrhoea was persistent in almost all patients. Furthermore, the occurrence of diarrhoea was not consistently associated with the exposure to nelfinavir or saquinavir, in contrast to other gastrointestinal complaints.
Recently, many studies have focused on the association between protease inhibitor use and the occurrence of a new onset of diabetes mellitus, lipodystrophy or hyperlipidaemia  The median time to the development of peripheral lipodystrophy was reported to be 10 months which might explain why this side effect was not observed within our population  Hyperlipidaemia was present within the observed treatment period, although the elevations of triglycerides and cholesterol were usually mild and infrequent in comparison with those reported in studies using a combination of ritonavir and saquinavir [17–19]
Surprisingly, the plasma concentrations of nelfinavir and saquinavir obtained were relatively low in comparison with the expected levels obtained under standardized conditions. These expected plasma drug concentrations, obtained in a clinical setting and after the use of a standard meal, probably resulted in higher plasma drug levels than those obtained in daily practice. This was found even though all the patients in this study were instructed to take their medication with food. Despite the relatively low exposure, virologic efficacy of the quadruple regimen was found to be satisfactory in the first 26 weeks. Only one patient failed to attain a plasma HIV-1 RNA concentration below 400 copies/ml. However, it remains to be seen whether these plasma drug concentrations are sufficient for durable suppression of the viral replication.
In contrast to the findings of Khaliq et al. the plasma concentration ratios of the protease inhibitors in the antiretroviral combination used in the ADAM study did not decline during the first 26 weeks  However, by replacing saquinavir HGC with saquinavir SGC, the exposure to saquinavir within this population may have been maintained over time [30,31]
Figure 1 illustrates the variation of plasma concentration ratios of nelfinavir and saquinavir in patients over time. Considering this wide variation, the odds ratio for the occurrence of a specific side effect at a specific time point as a result of drug exposure was estimated. Most gastrointestinal complaints were found to have a time-dependent relationship with drug exposure. For both nausea and abdominal distension, the association with drug exposure varied within the first weeks of treatment. This change of association within the first weeks of treatment could indicate that the pharmacokinetics of the drugs and the tolerance of the patient had not yet reached an equilibrium. In addition, the use of stavudine and lamivudine might have concealed a possible relationship of nausea with the exposure to protease inhibitors. After the first weeks of therapy, however, abdominal distension and nausea were both associated with a low exposure to the protease inhibitors used. For nausea this association became even stronger over time. This could indicate a reduced uptake of the protease inhibitors in these patients, but could also illustrate that patients with nausea may well have developed a less compliant behaviour, resulting in low plasma drug concentrations. Plasma drug monitoring in these patients could be useful when virologic efficacy appears insufficient.
The patients with abdominal pain were found to have both a high drug exposure as well as a greater decrease in HIV-1 RNA within the first 26 weeks. Hoetelmans et al. found that a higher clearance rate of HIV-1 RNA was found in patients with a higher exposure to nelfinavir and saquinavir  It may therefore be concluded that the elevation of drug exposure to nelfinavir and saquinavir in order to increase virologic effectiveness might be hampered by the occurrence of abdominal pain.
In dose-finding studies, diarrhoea appeared to be the dose-limiting side effect of nelfinavir [32,33] while a higher bio-availability of saquinavir was more likely to cause diarrhoea, suggesting a relationship between the exposure to a PI and the occurrence of this side effect [31,34] The patients with diarrhoea in our study were found to have lower levels of saquinavir than patients without diarrhoea (Wilcoxon test). This was however not confirmed using the GEE method. In fact, using this more sensitive method, the occurrence of diarrhoea was not related to the exposure to saquinavir, but had an inconsistent association with the exposure to nelfinavir. It is possible that the occurrence of diarrhoea has been both a cause and an outcome of drug exposure, which might have concealed a significant association.
None of the other clinical or laboratory abnormalities could be related to the observed drug exposure. The occurrence of fatigue and headache might be explained by the (continuous) use of nucleoside reverse transcriptase inhibitors as well. In addition, fatigue is a side effect that is more likely to occur in patients with a more advanced disease stage. This was confirmed in our study since patients experiencing fatigue had a significantly lower CD4+ cell count and a higher HIV-1 RNA concentration in plasma at baseline than patients without this complaint.
In conclusion, the quadruple induction therapy was quite well tolerated. Diarrhoea was the most frequently reported side effect. During the first 26 weeks of therapy, the elevations of plasma triglycerides and cholesterol were mild and infrequent in comparison with those observed in other PI combinations. Clear and time-dependent relationships with the amount of drug exposure were only found for nausea, abdominal distension and abdominal pain. The exposure to the protease inhibitors in this population was lower than expected, suggesting that the compliance to timing, dosing and food intake required may not be achieved in daily practice. Nevertheless, the virologic efficacy of the quadruple drug regimen was quite satisfactory.
The authors thank Roche NL for supplying the study medication and financial support, and Bristol-Myers Squibb for financial support. All physicians and research nurses involved in the study are acknowledged for their care of the trial patients and commitment to the study protocol. Most of all, the patients in the study are acknowledged for their participation.
1. Gulick RM Mellors JW Havlir D et al Treatment with indinavir, zidovudine, and lamivudine in adults with human immunodeficiency virus infection and prior antiretroviral therapy.
N Engl J Med 1997 337734–739.
2. Hammer SM Squires KE Hughes MD et al A controlled trial of two nucleoside analogues plus indinavir in persons with human immunodeficiency virus infection and CD4 cell counts of 200 per cubic millimeter or less.
N Engl J Med 1997 337724–733.
3. Montaner JSG Reiss P Cooper D et al A randomized, double-blinded trial comparing combinations of nevirapine, didanosine and zidovudine for HIV infected patients - the INCAS trial.
JAMA 1998 279930–937.
4. Palella FJ Delaney KM Moorman AC et al Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection.
N Engl J Med 1998 338853–860.
5. Mouton Y Alfandari S Valette M et al Impact of protease inhibitors on AIDS defining events and hospitalization in 10 French AIDS reference centres: Federation National des Centres de lutte contre le SIDA.
AIDS 1997 11F101–F105.
6. Kempf DJ Rode RA Xu Y et al The duration of viral suppression during protease inhibitor therapy for HIV-1 infection is predicted by plasma HIV-1 RNA at the nadir.
AIDS 1998 12F9–F14.
7. Raboud JM Montaner JSG Conway B et al Suppression of plasma viral load below 20 copies/ml is required to achieve a long-term response to therapy.
AIDS 1998 121619–1624.
8. Weverling GJ Lange JMA Jurriaans S et al Alternative multidrug regimen provides improved suppression of HIV-1 replication over triple therapy.
AIDS 1998 12F117–F122.
9. Hoetelmans RMW Reijers MHE Weverling GJ et al The effect of plasma drug concentrations on HIV-1 clearance rate during quadruple drug therapy.
AIDS 1998 12F111–F115.
10. Lange JMA Richman DD The limitations of simplicity.
Antiretroviral Ther 1998 34–5.
11. Montaner JSG Hogg R Raboud J Harrigan R O'Shaughnessy M Antiretroviral treatment in 1998.
Lancet 1998 3521919–1922.
12. Merry C Barry MG Mulcahy F et al Saquinavir pharmacokinetics alone and in combination with ritonavir in HIV-infected patients.
AIDS 1997 11F29–F33.
13. Merry C Barry MG Mulcahy F et al Saquinavir pharmacokinetics alone and in combination with nelfinavir in HIV-1 infected patients.
AIDS 1997 11F117–F120.
14. Carr A Samaras K Burton S Freund J Chisholm DJ Cooper DA A syndrome of peripheral lipodystrophy, hyperlipidemia and insulin resistance in patients receiving HIV protease inhibitors.
AIDS 1998 12F51–F58.
15. Deeks SG Smith M Holodnly M Kahn JC HIV-1 Protease Inhibitors.
:A review for clinicians.
JAMA 1997 277145–153.
16. Flexner C HIV-Protease Inhibitors.
N Engl J Med 1998 3381281–1291.
17. Kaufmann GR Duncombe C Cunningham P et al Treatment response and durability of a double protease inhibitor therapy with saquinavir and ritonavir in an observational cohort of HIV-1 infected individuals.
AIDS 1998 121625–1630.
18. Rhone SA Hogg RS Yip B et al The antiretroviral effect of ritonavir and saquinavir in combination amongst HIV-infected adults: results from a community-based study.
AIDS 1998 12619–624.
19. Lorenzi P Yerly S Abderrakim K et al Toxicity, efficacy, plasma drug concentrations and protease mutations in patients with advanced HIV infection treated with ritonavir plus saquinavir.
AIDS 1997 11F95–F99.
20. Reijers MHE Weverling GJ Jurriaans S et al Maintenance therapy after quadruple induction therapy in HIV-1 infected individuals: Amsterdam Duration of Antiretroviral Medication (ADAM) study.
Lancet 1998 352185–190.
21. World_Health_Organisation WHO handbook for reporting results of cancer treatment.
Geneva: WHO, 1979.
22. Guidelines for grading of toxicity of the_ACTG
. AIDS Clinical Trial Group, Rockville, Maryland, USA, 1997.
23. Hoetelmans RMW van Essenberg M Meenhorst PL et al Determination of saquinavir in human plasma, saliva, and cerebrospinal fluid by ion-pair high-performance liquid chromatography with ultraviolet detection.
J Chrom B 1997 698235–241.
24. Hoetelmans RMW, Reijers MHE, van Heeswijk RPG, et al.Saquinavir (SQV) pharmacokinetics in combination with nelfinavir (NFV) in the ADAM study.6th European Conference on Clinical Aspects of HIV-infection.
Hamburg, October 1997 [abstract 255].
25. Zeger SL Liang KY Longitudinal data analysis using generalized linear models.
:The analysis of discrete and continuous longitudinal data.
Biometrics 1986 42121–130.
26. Centers for Disease Control 1993 Revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults.
MMWR 1992 41 (RR-17)1–19.
27. Slater l, Sension M, Feinberg J, Poblete R, Siemon P, Pilson R. Fortovase BID regimens in HIV-1 infected patients: combination with 2 nucleoside analogs or with nelfinavir plus 1 nucleoside analog.6th Conference on Retroviruses and Opportunistic Infections
. January 1999 [abstract 390].
28. Moyle G. Study of protease inhibitors combination in Europe (SPICE): Saquinavir Soft Gelatin Capsule (SQV-SGC, Fortovase TM and Nelfinavir (NFV) in HIV infected individuals.12th World AIDS Conference.,
Geneva, June 1998 [abstract 12222].
29. Khaliq Y, Gallicano K, Sahai J, et al
. Effect of nelfinavir (NFV) on short and long term plasma exposure of saquinavir in hard gel capsule (SQV-HGC) during TID and BID dosing regimens.AIDS
30. Mitsuyasu R Skolnik P Cohen S et al Activity of the soft gelatin formulation of saquinavir in combination therapy in antiretroviral-na]ve patients.
:NV15355 study Team.
AIDS 1998 12F103–F109.
31. Roche. Fortovase (saquinavir) soft gelatin capsules product monograph
. Nutley, NJ: Roche Laboratories; 1997 (package insert).
32. Moyle GL, Youle M, Higgs C, et al
. Extended follow-up of the safety and activity of Agouron's HIV protease inhibitors AG1343 (Viracept) in virological responders form the UK phase I/II dose finding study.Volume 1 of Program and abstracts of the 11th International Conference on AIDS
. Vancouver, July 1996 [abstract 25].
33. Agouron Pharmaceuticals. Viracept (nelfinavir mesylate) tablets and oral powder product monograph.
La Jolla, CA: Agouron Pharmaceuticals; 1997 (package insert).
34. Roche. Invirase (saquinavir mesylate) capsules product monograph
. Nutley, NJ: Roche Laboratories; 1997 (package insert).