1. Introduction

Protease inhibitors are a group of powerful suppressors of viral replication used in the management of patients infected with HIV. International guidelines recommend that protease inhibitors be included in multidrug treatment regimens with other antiretroviral agents including nucleoside reverse transcriptase inhibitors (NRTIs) and/or non-nucleoside reverse transcriptase inhibitors (NNRTIs) both early in HIV infection and in patients with progressive HIV disease.[14]

Nelfinavir is one of several currently available HIV protease inhibitors. The pharmacology and therapeutic use of nelfinavir in the management of patients with HIV infection have been reviewed previously in Drugs.[5,6] The current review provides an update on the clinical profile of nelfinavir and its role in HIV infection. An overview of the pharmacodynamic and pharmacokinetic properties of nelfinavir is also provided.

2. Pharmacodynamic Properties

Nelfinavir is a selective, competitive, nonpeptidic HIV-1 protease inhibitor designed using protein structure-based technology.[7] The HIV protease is a homodimeric aspartic protease involved in the post-translational processing of viral gag (p55) and gag-pol (p160) polyprotein products into functional core proteins and viral enzymes.[8] This process, which occurs during or immediately after the budding of immature virions from infected cells, is critical for the production of infectious virus particles. As proteolytic maturation is one of the latest stages of the retroviral life cycle, inhibitors of HIV protease block the formation of mature virus particles in both acutely and chronically infected cells.

2.1 Antiviral Activity

2.1.1 In Vitro

Nelfinavir has an in vitro inhibition constant (K i) of 1.7 nmol/L against HIV-1 protease, but no significant activity against human aspartic proteases including pepsin, renin and gastricin. Very weak affinity for cathepsin E (74 μmol/L) and cathepsin D (435 nmol/L) has been observed.[9]

The in vitro antiviral activity of nelfinavir against a variety of HIV-1 and HIV-2 strains was demonstrated in a number of cultured human T cell lines. CEM-SS and MT-2 cells were protected against HIV-1-induced cell death by nelfinavir at 50% effective concentrations (EC50) ranging from 31 to 43 nmol/L. Nelfinavir inhibited replication of HIV-1 and -2 strains in acutely infected primary macrophages and peripheral blood mononuclear cells (PBMCs), as well as chronically infected CEM-SS cells. The mean nelfinavir concentration producing 95% inhibition of viral replication (IC95) in a variety of in vitro HIV infection models was 59 nmol/L (range 7 to 130 nmol/L).[9] Nelfinavir was also effective at inhibiting replication of the zidovudine-resistant HIV-1 strain G910-6 and the NNRTI-resistant strain A17.[9]

An accumulation of gag and gag-pol polyprotein intermediates was observed in chronically infected T cell lines H9/HIV-1IIIB, H9/HIV-1MN and CEM/HIVMN after exposure to nelfinavir and other protease inhibitors. This apparent inhibition of HIV protease activity was concentration dependent; however, concentrations up to 500 times the EC50 values of each protease inhibitor did not completely inhibit gag-pol precursor processing.[10]

In vitro antiviral synergy against HIV-1 was observed when nelfinavir was combined with zidovudine, lamivudine, zalcitabine or zidovudine plus lamivudine in CEM-SS cells.[11] Additive activity occurred with nelfinavir in combination with either stavudine or didanosine.[11] Nelfinavir and ritonavir exhibited statistically significant antiviral synergism in the presence of human serum in vitro (p = 0.02).[12] A slightly antagonistic interaction was noted between nelfinavir and indinavir against HIV-1RF in CEM-SS cells.[11] Minimal cytotoxicity was observed with nelfinavir alone or in combination with other protease inhibitors in vitro.

2.1.2 In Vivo

Combination therapy with multiple antiretroviral drugs is now considered necessary to adequately contain viral replication and delay the development of viral drug resistance.[14,13]

Nelfinavir, in combination with NRTIs and/or NNRTIs and other protease inhibitors, has been shown to significantly reduce plasma viral load and to maintain suppression during long term therapy in patients treated early in the course of HIV infection (see section 4).[14] In univariate and multivariate analyses, a positive correlation between nelfinavir exposure in plasma and HIV-1 RNA clearance from plasma was reported in a group of 29 patients receiving both nelfinavir and saquinavir in combination with stavudine and lamivudine.[15] More specifically, the independent variable of drug ‘exposure’ was defined as the median plasma drug concentration ratio when values from patients enrolled in the study were compared with those from a reference group of 18 patients in whom pharmacokinetic profiles of both drugs were assessed. A similar correlation was observed with saquinavir exposure in univariate but not multivariate analysis. However, the study authors suggested various interpretations of these results and cautioned that they do not necessarily indicate a greater effect of nelfinavir than saquinavir on the initial elimination rate of HIV-1 RNA in plasma when used in this quadruple drug regimen.[15]

2.1.3 Suppression of Viral Replication in Compartmental HIV Reservoirs

An important consideration in HIV therapy is the ability of antiretroviral agents to penetrate peripheral target tissues such as lymph nodes and CNS, as well as the genital mucosa, where HIV replicates and is transmitted. AIDS dementia complex is thought to be the result of direct brain infection by HIV, as neurophysiological functioning has been correlated with HIV load in CSF.[16] The effects of combination regimens containing nelfinavir on HIV replication in CSF and genital mucosa have been assessed in small numbers of patients.

CSF

HIV RNA in CSF from 3 HIV-infected patients was reduced by a mean of 0.95 log10 copies/ml from baseline 5 to 7 days after initiating combination therapy with nelfinavir, stavudine and lamivudine.[17] Plasma HIV RNA was reduced by a mean of 1.1 log10 copies/ml in the same period, although 2 of 3 patients showed substantially smaller reductions in CSF viral load than in plasma viral load. Despite the rapid lowering of HIV RNA levels by this multidrug regimen, the difference in HIV RNA clearance between CSF and plasma suggests a distinct CNS compartment of HIV replication.

Although nelfinavir could not be detected in CSF by high performance liquid chromatography (HPLC) methodology,[18] an ultrasensitive detection system (LC/MS assay) was able to measure the combined concentration of nelfinavir and its active metabolite M8 in CSF at 9.4 and 11.8 nmol/L in 2 patients via lumbar punctures.[19] These concentrations are well below the IC95 for nelfinavir in vitro (section 2.1.1), and it is unclear whether nelfinavir contributed significantly to the reduction in CSF HIV RNA of 0.38 and 1.18 log10 copies/ml from baseline observed in these patients on the fourth day of combined therapy (nelfinavir 750mg 3 times daily with stavudine and lamivudine).

Genital Mucosa

HIV-infected women were tested for the presence of HIV RNA in cervicovaginal lavages before and during therapy with nelfinavir, saquinavir, stavudine and lamivudine in either twice- or 3-times-daily dosage schedules in a nonblind randomised trial.[20] At baseline, 16% (9 of 56 women) had detectable HIV RNA in their genital mucosa (mean = 1240 copies/ml), which was suppressed below 400 copies/ml in 8 of 9 women during 48 weeks of therapy.

Nelfinavir and other protease inhibitors (as part of combination regimens) contributed to a decrease in the shedding of HIV RNA in seminal fluid as well as in plasma in treated patients (n = 85) relative to untreated controls (p < 0.0001).[21] Nonetheless, another study demonstrated that seminal cells could harbour proviral HIV DNA even when plasma RNA concentrations have been suppressed below the limit of detection (LOD; 50 copies/ml).[22]

2.2 Effects on Immune Function

HIV infection results in severe immunodeficiency characterised by a depletion of CD4+ cells (helper T lymphocytes). After plasma viral load, a patient’s CD4+ cell count is currently the next best predictor of progression to AIDS and death.[23] The CD4 molecule acts as a high affinity receptor for HIV, initiating fusion to the cell membrane and viral entry. CD4+ cell depletion by HIV appears to result from the induction of programmed cell death (apoptosis) in response to Fas receptor ligation,[24,25] although other contributory mechanisms may ultimately be elucidated. Combination therapy with nelfinavir, lamivudine and zidovudine was associated with a decrease in Fas receptor expression in T lymphocytes, possibly resulting in a reduction in T cell apoptosis.[26] A 23% decrease in apoptosis was observed in cultured PBMCs from patients treated with nelfinavir and reverse transcriptase inhibitors (RTIs) for 48 weeks, coinciding with a decrease in plasma HIV RNA.[27] Therapy with nelfinavir plus saquinavir in combination with 2 NRTIs significantly reduced apoptosis in both CD4 and CD8 T cells (from 17.4 to 2.6%, and from 13.7 to 1.5%, respectively; p ≤ 0.03).[25] In this study, changes in apoptosis were not associated with altered Fas expression.

In clinical trials, treatment of HIV-infected individuals with nelfinavir in combination with NRTIs led to increased CD4+ cell counts in protease inhibitor-naive patients [e.g. mean increase from baseline was >190 cells/μl at 1 year in a pivotal trial (Study 511[28,29]); see section 4]. This effect was sustainable with combination therapy; however, nelfinavir monotherapy is associated with generally transient increases in CD4+ cell counts (reviewed by Powderly and Tebas[14]) because of the rapid development of resistance (section 2.3). Kaufmann and colleagues[30] prospectively evaluated immune reconstitution in 28 patients treated with nelfinavir (n = 15) or indinavir (n = 13) in combination with 2 NRTIs during primary HIV infection and in 30 antiretroviral therapy-naive patients starting similar (undefined) highly active antiretroviral therapy during chronic HIV infection. Results of the study showed that immune reconstitution was more rapid and complete in patients with primary HIV infection than in those at more advanced stages of the disease. Median CD4+ count increased from 470 cells/μL at baseline to 758 cells/μl at 1 year and reached the normal range in 93% of the patients with primary HIV infection. In contrast, median CD4+ count was 204 cells/μl at baseline, increased to 310 cells/μl at 1 year, but reached the normal range in only 37% of patients with more advanced disease (p < 0.0005 for 93 vs 37%).

2.3 Resistance and Cross-Resistance

The degree of resistance to protease inhibitors is generally a function of the cumulative number of resistance mutations in the HIV protease gene, although the location of specific point mutations can also have a marked effect.[31,32] Primary resistance mutations affect protease inhibitor and substrate binding by changing the number and/or the strength of subsite interactions.[33] These mutations affect drug sensitivity directly and tend to arise prior to secondary mutations that may act to restore or increase the replication efficiency of the protease inhibitor-resistant viral variants. All of the currently available protease inhibitors rapidly select for viral drug resistance when used as monotherapy.[13] The time needed for a particular resistant species to dominate a viral population depends on the replicative fitness of the mutant virus and on the magnitude of the selective pressure.

An HIV-1 variant with 7-fold reduced sensitivity to nelfinavir was isolated in vitro after 22 passages of virus with increasing concentrations of the drug.[9] The variant contained a previously undescribed aspartic acid (D) to asparagine (N) substitution at residue 30 (D30N) of the HIV protease amino acid sequence. This mutation was subsequently observed as the predominant variant in clinical isolates from HIV-1-infected patients experiencing virological failure while receiving nelfinavir (25 of 55 patients; median treatment duration 13 weeks; range 2 to 52 weeks).[34] The clinical isolates exhibited 5- to 93-fold decreased susceptibility to nelfinavir in vitro. Martinez-Picado et al.[35] showed that a nelfinavir-selected protease D30N substitution substantially impaired the replicative fitness of HIV-1.

The D30N mutation was most often associated with substitutions of M36I, A71T and N88D in clinical isolates from patients who had received nelfinavir.[34] These compensatory mutations appear to increase viral fitness without directly affecting drug sensitivity. D30N may be the only characteristic mutation conferring resistance to nelfinavir; it does not arise from treatment with other protease inhibitors, and alone does not cause resistance to other protease inhibitors in vitro. In terms of phenotypic resistance, although treatment with nelfinavir was associated with high level resistance to nelfinavir, minimal reduced susceptibility to other protease inhibitors (indinavir, ritonavir, saquinavir or amprenavir) was observed (fig. 1). In contrast, treatment with indinavir or another protease inhibitor resulted in reduced susceptibility to multiple protease inhibitors. Ritonavir-resistant variants are also partially cross-resistant to both nelfinavir and indinavir.[31] Amino acid substitutions selected as primary mutations by other protease inhibitors in vivo (e.g. L90M, M46I) have been occasionally observed in isolates from patients in whom nelfinavir therapy has failed. These changes act as secondary mutations in the emergence of nelfinavir resistance.[37] Although M46I is a primary mutation selected by indinavir, at least 3 mutations in the same viral genome are required for high level indinavir resistance.

Fig. 1
figure 1

Phenotypic cross-resistance of HIV isolates from patients failing to respond to protease inhibitor-based therapies.[36] Isolates from patients treated with nelfinavir (n = 44), indinavir (n = 16) or another protease inhibitor [saquinavir (n = 5), ritonavir (n = 3) or amprenavir (n = 1)] were tested for in vitro phenotypic drug susceptibility, measured as fold change in the 50% inhibitory concentration (IC50) from control (wild type) virus.

2.4 Other Effects

All available protease inhibitors interfere with adipocyte differentiation to some degree.[38] This leads to a number of metabolic disturbances which have been observed with these drugs in clinical trials (see section 5.2).

In an in vitro model of epithelial barrier function, nelfinavir decreased transepithelial resistance, but did not stimulate ion secretion in HT-29/B6 cell polarised epithelial monolayers. This could explain the induction of the leak-flux type of diarrhoea experienced by patients receiving nelfinavir (section 5.1). Other protease inhibitors, including ritonavir and saquinavir but not indinavir, produced similar in vitro effects.[39]

Nelfinavir, along with ritonavir, saquinavir and indinavir, has been shown to interact with the multidrug resistance protein MDR1 (also known as P-glycoprotein or PgP) in human cell lines.[4042] MDR1 is an efflux pump protein expressed at the interface of major organs, including brain, and in CD4+ T lymphocytes and monocytes.[43] MDR1 protein expressed in the gastrointestinal tract and capillary endothelial cells of the brain may constitute a pharmacological barrier limiting the oral bioavailability and CNS absorption of HIV protease inhibitors that are substrates of the transporter protein. Brains of mice deficient for the gene encoding MDR1 protein incorporated higher concentrations of nelfinavir, indinavir and saquinavir than did brains of wild type mice, confirming the role of MDR1-mediated translocation of the drugs into the CNS.[40] Preliminary data in patients with HIV infection suggest that it may be possible to inhibit MDR1 and thus improve CSF penetration of protease inhibitors.[44]

Binding of the protease inhibitors to MDR1 may also inhibit the ability of MDR1 to transport other drugs, potentially increasing their systemic and organ concentrations. However, analysis of protease inhibitor interaction with MDR1 and the related drug transporter gene MRP1 (multidrug resistance-associated protein 1) showed that the concentration of protease inhibitors required to inhibit these transporters was approximately 1000-fold higher than those required to inhibit HIV replication, and is not therefore of significant clinical concern.[45]

3. Pharmacokinetic Properties

The pharmacokinetics of nelfinavir have been investigated in healthy volunteers[46,47] and in adult patients with HIV infection.[4749] No substantial differences in pharmacokinetic parameters were observed between the 2 groups. As noted in section 2.1.2, a positive correlation between nelfinavir exposure in plasma and HIV-1 RNA clearance has been reported. Some data are also available on the pharmacokinetics of nelfinavir in children with HIV infection.

3.1 Absorption

In healthy male volunteers, maximum plasma concentrations (Cmax) of nelfinavir were reached approximately 3.4 hours (tmax) after administration of single oral doses (100, 200, 400 or 800mg) with food. The mean plasma nelfinavir concentration exceeded the 95% antiviral effective dose (ED95 = 40 μg/L) for up to 8 hours after a single 100mg dose, and up to 24 hours after an 800mg dose. The area under the concentration-time curve (AUC) after administration in the fasted state was 27 to 50% of the AUC when the drug was administered with food.[46] The oral bioavailability of nelfinavir has been reported as 70 to 80% when administered with food.[50,51] This is higher than that of the other available protease inhibitors with the possible exception of amprenavir (35 to 90%).[51] The relatively slow oral absorption of nelfinavir and lack of large fluctuations between plasma trough (Cmin) and Cmax concentrations (see later in this section), along with its relatively long elimination half-life (t½; see section 3.4) compared with, for example indinavir, led to the clinical evaluation (and subsequent approval in the US) of twice-daily administration of nelfinavir in adult patients (section 6).

Nelfinavir Cmax values of 3 and 4 mg/L, respectively, were reported after 28 days’ administration of nelfinavir 750mg 3 times daily (total daily dose 2250mg) [n = 11] or 1250mg twice daily (total daily dose 2500mg) [n = 10] in patients with HIV infection.[47] Morning and afternoon Cmin concentrations with the 3-times-daily regimen were 1.4 and 1.0 mg/L, respectively, and morning and evening Cmin values with the twice-daily regimen were 2.2 and 0.7 mg/L. The AUC0-8h of nelfinavir was 23.4 μmol/L ⋅ h in 10 patients who received nelfinavir 750mg 3 times daily.[52]

3.2 Distribution

Preclinical studies using radiolabelled nelfinavir in rats showed extensive distribution of the drug to tissues, with concentrations in lymph nodes and spleen that were significantly higher than the in vitro antiviral ED95. Nelfinavir was also found at high concentrations (10 to 64 μg/g wet weight) in liver, pancreas, kidneys, submaxillary glands and heart, and at lower concentrations (0.3 to 1 μg/g) in testes, thymus and brain.[53]

Nelfinavir is highly bound to plasma proteins in humans (>98%) and the apparent volume of distribution is 2 to 7 L/kg.[47] Plasma protein binding is probably primarily to α1-acid glycoprotein.[54] With increasing interest in therapeutic drug monitoring for protease inhibitors, it is noteworthy that antiviral activity of nelfinavir and other protease inhibitors with extensive plasma protein binding was reduced against some clinical isolates of HIV when alpha-1 acid glycoprotein concentration was increased 4-fold in vitro. This magnitude of increase has been reported in human plasma during chronic infection.[55]

3.3 Metabolism

Nelfinavir, like all of the currently available protease inhibitors, is metabolised in the liver by the cytochrome P450 (CYP) enzyme system. Nelfinavir is metabolically converted to its major hydroxy-t-butylamide metabolite M8, which has anti-HIV activity comparable to nelfinavir in vitro.[52] A second metabolite, M1, has been detected in plasma but has not been shown to be active against HIV infection. At least 4 different CYP enzymes are involved in nelfinavir metabolism (CYP3A4, CYP2C9, CYP2C19 and CYP2D6), although CYP2C19 has been implicated as the primary enzyme involved in generating M8.[56] In samples from 10 patients treated with nelfinavir 750mg 3 times daily, the AUC0–8h of nelfinavir, M1 and M8 on day 1 were 23.4, 1.0 and 6.6 μmol/L ⋅ h, respectively. Similar AUC values were observed on day 28, indicating that nelfinavir and its metabolites do not accumulate at this therapeutic dose level.[52] However, a peak in AUC for both nelfinavir and M8 was observed on day 3 of a 9-day multiple-dose study (n = 6); peak values were 3 times higher than the AUCs recorded on day 1.[57] Exposure on day 9 dropped to 1.5 times the day-1 AUC value, a pattern consistent with autoinduction of nelfinavir metabolism. Similar results were obtained in another study in 12 patients, where the peak nelfinavir AUC occurred on day 2, followed by a decline to steady state around day 7.[58]

3.4 Elimination

The t½ of nelfinavir ranged from 3.5 to 5 hours in multiple-dose studies.[47,49] Most (87%) of a single oral 750mg dose of 14C-labelled nelfinavir was recovered in faeces over 7 days; 22% was unchanged nelfinavir. The remaining 78% consisted of oxidative metabolites.[47] Urinary excretion accounted for only 1 to 2% of the dose, most of which was unchanged nelfinavir.

Bodyweight, age, gender, race or dose level did not significantly affect clearance of nelfinavir in a population pharmacokinetic analysis of 509 plasma samples from 174 HIV-infected patients.[59]

3.5 Drug Interactions

Pharmacokinetic interactions between nelfinavir and other drugs are outlined in table I. All currently available HIV protease inhibitors have been shown to inhibit CYP3A4 isoenzymes,[6870] which are responsible for the metabolism of many drugs. Thus, concomitant administration of protease inhibitors with other CYP3A4 substrates can lead to increased plasma concentrations of the latter, potentially increasing the activity and/or toxic effects of these drugs. Nelfinavir does not appear to significantly inhibit other CYP isoforms.[68,69] Among the protease inhibitors, ritonavir has the greatest whereas nelfinavir and saquinavir have the weakest inhibitory effect on CYP3A4 in vitro.[68,69] Whether this reduced potential for CYP-mediated drug interactions with nelfinavir translates into fewer clinically significant drug interactions in patients is unclear. Importantly, table I includes data only for pharmacokinetic drug interactions that have been evaluated in formal analyses; therefore, some potential interactions of clinical significance are not included in the table (see later in this section and section 6).

Table I
figure Tab1

Pharmacokinetic drug interactions with nelfinavir (table does not include all potential interactions with nelfinavir)

Metabolic interactions between different protease inhibitors are potentially beneficial because they may increase plasma drug concentrations, reduce the required frequency of administration and delay the emergence of viral drug resistance.[71] Nelfinavir inhibits the elimination of saquinavir and indinavir but not ritonavir; conversely, nelfinavir metabolism is impaired by ritonavir and indinavir, and to a lesser extent by saquinavir and amprenavir (table I). The favourable interactions between nelfinavir and either indinavir or saquinavir provide good exposure for both drugs when administered twice daily.[72,73]

Nelfinavir has a lower capacity for drug interactions with currently available RTIs than with protease inhibitors (table I). However, nelfinavir AUC and Cmax were substantially increased in the presence of delavirdine (113 and 93%, respectively), whereas delavirdine concentrations were reduced (42 and 34%). Increasing the delavirdine dose from 400 or 600mg 3 times daily to 1200mg twice daily allowed for adequate exposure and effective viral suppression when combined with nelfinavir 1250mg twice daily.[74] Nelfinavir also reduces the Cmax and AUC of zidovudine by >30% (table I), although dosage adjustments are not recommended.

Competition for CYP3A4 by nelfinavir could inhibit the metabolism of other CYP3A4 substrates and potentially result in serious and/or life-threatening events. In particular, the following drugs are contraindicated with nelfinavir: astemizole, cisapride, triazolam, midazolam, ergot derivatives, amiodarone and quinidine.[47] The contraindication also applies to terfenadine, a drug withdrawn from the US and other markets.

Specific recommendations or warnings also apply to the concomitant use of nelfinavir and various other drugs.[47] Nelfinavir produced a substantial reduction in the clearance of rifabutin (table I); therefore the dose of rifabutin should be half that of the normal dose when administered with nelfinavir. Caution is advised if sildenafil is prescribed in patients receiving protease inhibitors, including nelfinavir, as concomitant administration is expected to markedly increase plasma sildenafil concentrations, potentially leading to hypotension, visual changes, priapism and other sildenafil-associated adverse events. Caution is also advised if nelfinavir is administered concomitantly with other CYP3A4 substrates such as dihydropyridine calcium antagonists, immunosuppressants and HMG-CoA reductase inhibitors. In particular, concomitant administration of nelfinavir with lovastatin or simvastatin is not recommended because of the increased risk of myopathy including rhabdomyolysis.[47]

The combination of nelfinavir and rifampicin (rifampin) resulted in a substantial reduction in nelfinavir plasma AUC; thus coadministration of these agents is not recommended.[47] Anticonvulsants such as carbamazepine, phenobarbital and phenytoin may also decrease nelfinavir plasma concentrations below clinically effective levels; alternative anticonvulsant agents such as valproic acid or gabapentin may be considered if nelfinavir is used.

Plasma concentrations of the oral contraceptives ethinyl estradiol and norethindrone were reduced in the presence of nelfinavir (table I), potentially below effective contraceptive concentrations.[47] Alternative or additional contraception is therefore recommended during nelfinavir treatment. In 2 patients with HIV infection receiving methadone for the treatment of opiate addiction, the addition of nelfinavir to their antiretroviral regimen was associated with about a 50% reduction in steady-state plasma methadone concentration.[75] Although data are limited, this interaction has potential clinical significance because reduced methadone plasma concentrations could lead to withdrawal symptoms. Mechanisms of these interactions have not yet been elucidated.

3.6 Pharmacokinetics in Special Populations

The pharmacokinetics of nelfinavir have not been extensively studied in patients with liver disease. In a small group of HIV-infected patients with moderate to severe liver disease, nelfinavir pharmacokinetics were characterised by a prolonged t½ (5 to 20 vs 3.5 to 5 hours expected) and low plasma concentrations of the active M8 metabolite (2 to 5% of nelfinavir concentrations vs ≈33% expected).[76] Target pharmacokinetic parameters of nelfinavir (AUC, Cmax and Cmin values from patients without liver disease) were achieved with nelfinavir regimens of 250mg 3 times daily and 500mg twice daily in 2 patients with moderate to severe liver disease, and 2 patients with mild disease were stabilised on nelfinavir 750mg 3 times daily.[76]

Pharmacokinetic studies with nelfinavir in patients with renal dysfunction appear to be lacking, with the exception of a case report showing that nelfinavir pharmacokinetic parameters were not influenced by severe renal insufficiency in an HIV-infected patient.[77] The need for dosage adjustment in patients with renal dysfunction seems unlikely because only 1 to 2% of an administered dose is eliminated in the urine (section 3.4).

3.7 Pharmacokinetics in Children

Clearance of nelfinavir was greater in children than in adults when adjusted for bodyweight. Mean AUC and Cmax values after a single 10 mg/kg dose in 8 children aged 2 to 13 years were 20 to 40% of those in adults after a similar weight-adjusted dose.[78]

In 8 children aged 0.5 to 10 years (weight 3.8 to 23kg), nelfinavir 30 mg/kg 3 times daily produced a steady-state AUC0–8h of 36.2 mg/L ⋅ h; Cmax was 7.0 mg/L, Cmin 3.4 mg/L, tmax 2.25 hours, and t½ 3.9 hours.[79] In an analysis of 50 children with HIV infection receiving nelfinavir (in combination with efavirenz and 1 or more NRTI), AUC0–8 values after 2 and 6 weeks of therapy were 21.6 and 19.9 mg/L ⋅ h, respectively (corresponding mean dosages were 23.6 and 24.6 mg/kg 3 times daily).[80] Twice daily nelfinavir 55 mg/kg produced a mean AUC0–12h of 51.2 mg/L ⋅ h in children aged 2 to 17 years.[81]

4. Clinical Efficacy

Trials of nelfinavir therapy have not been of sufficient length to evaluate effects on clinical end-points of HIV infection (including development of AIDS and death), and efficacy has therefore been evaluated using surrogate end-points, primarily HIV viral load and CD4+ cell count. These markers have been shown to correlate well with progression to AIDS.[82,83] Although no ‘hard’ clinical data are available for nelfinavir, the use of protease inhibitor-containing antiretroviral regimens has produced marked reductions in morbidity and mortality in patients with advanced HIV infection.[84,85]

Although many studies of viral load in HIV-infected patients have used a polymerase chain reaction (PCR) assay with a sensitivity of 400 HIV RNA copies/ml of plasma as the LOD of the virus, suppression of plasma viral load to less than 50 copies/ml has been shown to be a more significant predictor of durable virological response.[86,87] These observations support the use of ultrasensitive RNA assays to predict long term outcomes of anti-HIV therapy.

Preliminary efficacy studies of nelfinavir in HIV infection have been reviewed in a previous issue of Drugs.[5] Since then, a number of additional studies have been conducted, although many have been presented only as conference abstracts and/or posters. Some of the studies have reported on intention-to-treat analyses, the standard for US Food and Drug Administration (FDA) regulatory submission, and these have been noted where applicable in the following subsections.

Two pivotal phase II/III randomised, double-blind trials evaluated nelfinavir, administered 3 times daily, as first-line therapy with NRTIs in antiretroviral treatment-naive patients, or as an additional antiretroviral agent in protease inhibitor-naive patients already receiving NRTIs (section 4.1).

Nelfinavir has also been combined with NNRTIs, often in patients with extensive previous experience with NRTIs but also as a first-line therapy option (section 4.2). Other studies have evaluated the efficacy of nelfinavir treatment in patients switching from another protease inhibitor-based regimen for reasons of drug intolerance (section 4.5.1) or virological failure (section 4.5.2). To determine the potential for development of protease inhibitor cross-resistance to nelfinavir, salvage therapy with other protease inhibitors has been assessed in patients who experienced virological failure while receiving nelfinavir (section 4.5.3). These studies are discussed in the following sections.

The efficacy of twice-daily nelfinavir, usually in combination with other protease inhibitors (section 4.3) has also been investigated in several randomised clinical trials. Until recently, the manufacturer’s dosage recommendation for nelfinavir was 750mg 3 times daily; however, on the basis of clinical data from these randomised trials, the US FDA recently approved an alternative regimen of 1250mg twice daily in adult patients. This may improve patient adherence to treatment, which is known to be influenced by the convenience of administration.[88]

4.1 In Combination with NRTIs

Two multicentre, randomised, double-blind trials (Study 511[28,29] and AVANTI 3,[89,90] the latter being an intention-to-treat analysis), assessed the efficacy of nelfinavir combined with lamivudine and zidovudine as first-line therapy in antiretroviral therapy-naive patients (table II). Both studies have been published as posters/abstracts only.

Table II
figure Tab2

Efficacy of nelfinavir (NFV) in combination with NRTIs in randomised multicentre trials. Studies were OT analyses unless otherwise indicateda

In the AVANTI 3 study, 53 patients received triple therapy with nelfinavir 750mg 3 times daily, lamivudine 150mg twice daily and zidovudine 200mg 3 times daily, and 52 patients received lamivudine and zidovudine only (same dosage regimens).[89,90] After 28 weeks the study was unblinded and 25 patients in the latter group chose to have nelfinavir added to their regimen. Study 511 included 297 patients randomised to 3 treatment regimens, all of which included lamivudine and zidovudine (dosages as in AVANTI 3); additional therapy was nelfinavir 500mg 3 times daily in 97 patients, nelfinavir 750mg 3 times daily in 99 patients and placebo in 101 patients (nelfinavir was added at week 24 in this last group).[28,29,91] Results of study 511 are those for patients still on treatment at the time of assessment.

Both studies demonstrated greater reductions from baseline in mean/median HIV viral load in patients who received nelfinavir 750mg 3 times daily, lamivudine and zidovudine triple therapy than in patients treated with lamivudine and zidovudine dual therapy (reduction of ≥2.3 vs ≤1.5 log10 copies/ml at 24 weeks). The response was sustained during the 52 weeks of treatment with nelfinavir-containing triple therapy at this dosage (plasma HIV RNA levels reduced by ≥2.3 log10 copies/ml vs baseline). In AVANTI 3, statistical analysis was provided only for a comparison of area under the curve minus baseline (AUCMB; 0 to 28 weeks) for viral load, and results favoured nelfinavir-treated patients (−1.85 vs −0.98 HIV RNA log10copies/ml; p < 0.001).[90] In Study 511, several parameters favouring triple therapy over dual therapy were statistically significant after 6 months of therapy (table II and fig. 2), although statistical analyses have not been provided for 1-year data.

Fig. 2
figure 2

Increase from baseline in mean CD4+ cell count in patients treated with a nelfinavir-containing triple therapy regimen (study 511).[29] 297 antiretroviral treatment-naive patients were randomised to receive zidovudine, lamivudine and placebo (control) or zidovudine, lamivudine and nelfinavir (either 500 or 750mg 3 times daily). At 6 months all patients in the control group were started on nelfinavir 500 or 750mg 3 times daily. Zidovudine and lamivudine administered at standard dosages. Statistical analysis not available for 12-month data. * p = 0.02 vs control at 6 months.

In Study 511, the mean CD4+ cell count increased by >190 cells/μl from baseline in patients treated with nelfinavir 500 or 750mg 3 times daily for 52 weeks (fig. 2).[29] 61% of those treated with the higher nelfinavir dosage had plasma HIV RNA levels below the LOD (50 copies/ml) at 52 weeks, compared with 37% of those receiving triple therapy with the lower dosage of nelfinavir. Only 5% of patients who received lamivudine and zidovudine had HIV RNA levels below 50 copies/ml at 24 weeks, but after nelfinavir was added to the regimen in at least some patients, this increased to 23% at 52 weeks.

Twice-daily administration of nelfinavir 1250mg was compared with the standard 750mg 3-times-daily dosage, both in combination with stavudine and lamivudine, in 554 patients with ≥6 months prior NRTI experience. In this randomised, nonblind comparison (Study 542), both treatment regimens provided equally effective suppression of HIV RNA over a 96-week period (table II, fig. 3).[92]

Fig. 3
figure 3

Nonblind comparison of the long term antiviral efficacy of twice- and 3-times-daily administration of nelfinavir in triple therapy with nucleoside reverse transcriptase inhibitors.[92] 554 antiretroviral treatment-naive, HIV-infected patients were randomised to receive nelfinavir twice daily (bid, 1250mg) or 3 times daily (tid, 750mg) with stavudine and lamivudine and plasma HIV RNA levels were recorded for up to 96 weeks. Analysis was by intention to treat, with noncompleters recorded as failures. LOD = limit of detection.

Study 542 was an intention-to-treat analysis; however, as indicated in table II, some results are also presented for patients on treatment at the time of assessment. Plasma HIV RNA levels decreased below the LOD (400 copies/ml) in approximately 80% of all patients in both groups by 16 weeks of therapy, and was maintained in 80% of responders through week 96. Switching nelfinavir administration from 3 times daily (750mg) to twice daily (1250mg) in clinically stable antiretroviral therapy-experienced patients was also found to be effective in maintaining immunological and virological responses for up to 52 weeks in a randomised nonblind trial (other antiretroviral agents not specified).[94]

A retrospective analysis comparing viral load and CD4+ cell count changes in patients receiving nelfinavir in twice- versus 3-times-daily schedules with NRTIs (n = 178) reported similar efficacy in all groups, but noted that rebound of plasma viraemia to detectable levels (>400 RNA copies/ml) was less likely in patients receiving nelfinavir 1250mg twice daily than in those receiving 750mg 3 times daily (p ≤ 0.03).[95]

Dual therapy with nelfinavir and stavudine was effective in NRTI-experienced patients (Study 506), conferring greater reductions in HIV viral load and increases in CD4+ cell counts than those observed with stavudine alone after 24 weeks of therapy (statistical significance not provided). Nevertheless, results also suggested that such dual combination therapy was not optimal (table II).[14,28]

Primary HIV infection is defined as the period from the initial infection to complete seroconversion and is often symptomatic (acute HIV syndrome).[3] Hecht and colleagues[96] reported on 10 patients with primary HIV infection (presence of HIV-1 RNA with a negative or indeterminant HIV-1 antibody test, or documented seroconversion within 6 months of enrolment) who received triple therapy with nelfinavir, zidovudine and lamivudine. Median baseline plasma HIV RNA level was approximately 4.9 log10 copies/ml, and levels were <500 copies/ml in all 10 patients after 12 weeks of therapy. Eight patients continued therapy beyond 12 weeks, and viral load remained below 500 copies/ml after a median follow-up period of >26 weeks. Broadly similar results were reported in a group of 28 patients with primary HIV infection who received 2 NRTIs plus either nelfinavir (n = 15) or indinavir (n = 13), although details on antiviral efficacy are sketchy because the study focused on immune reconstitution and was reported only as an abstract (see section 2.2).[30]

4.2 In Combination with NNRTIs

Nelfinavir has shown efficacy when combined with NNRTIs in both antiretroviral therapy-naive and -experienced patients (table III). Unless indicated otherwise, results of studies reported in table III reflect those for patients on treatment at the time of assessment.

Table III
figure Tab3

Efficacy of nelfinavir (NFV) in combination with non-nucleoside reverse transcriptase inhibitorsa

In a randomised, phase II trial in patients with extensive NRTI but no protease inhibitor experience (Study ACTG 364; n = 189), quadruple therapy with nelfinavir, efavirenz (dosages not stated) and 2 new NRTIs provided a significantly greater and more durable viral load suppression than did triple nelfinavir- or efavirenz-based regimens [67% of patients had HIV RNA below the LOD (<50 copies/ml) at weeks 40 to 48 vs 22 and 44%, respectively; p ≤ 0.008].[98] A nonblind, multicentre study (DMP 266-024) reported that in antiretroviral- or protease inhibitor-naive patients, nelfinavir 750mg 3 times daily in combination with efavirenz 600mg once daily decreased mean plasma HIV RNA levels from baseline by 1.66 or 1.55 log10 copies/ml at 16 weeks (p < 0.05).[99] At 24 weeks, a larger proportion of antiretroviral-naive patients had undetectable plasma HIV RNA (<400 or 50 copies/ml; 71 or 64%, respectively) compared with patients who were only naive to protease inhibitors (54 or 42%), although statistical analysis was not provided (table III).

A combined quadruple regimen of nelfinavir 750mg 3 times daily with delavirdine 400 or 600mg 3 times daily, stavudine 40mg twice daily and didanosine 200mg twice daily reduced plasma HIV RNA by 2.0 to 3.0 log10 copies/ml from baseline in 22 patients in a 48-week randomised nonblind study.[102] 84% of all patients in the study achieved an undetectable plasma HIV RNA level (<400 copies/ml), which was maintained through the entire treatment period. Similar reductions in HIV viral load were observed in another randomised study in which the same 4 drugs were given twice daily for 24 weeks (n = 173) [table III].[103]

Triple therapy with nelfinavir, nevirapine and stavudine suppressed plasma HIV RNA to undetectable levels (<50 copies/ml) in 15 of 17 patients (88%) by week 13 of treatment in a small nonblind study.[104] These patients were naive to protease inhibitors and NNRTIs, and had <6 months prior stavudine experience. Other studies have demonstrated modest short term success using the nelfinavir/nevirapine combination as a salvage regimen in patients experiencing virological failure while receiving another protease inhibitor (see section 4.5.2).

4.3 In Combination with Other Protease Inhibitors

Dual protease inhibitor therapy, with or without other antiretroviral agents, is being investigated as a means of increasing the potency and extending the durability of virological and immunological response, and decreasing the probability of developing viral drug resistance to individual protease inhibitors.[105] As outlined in table IV, nelfinavir has been most often combined with saquinavir.[106109] Ritonavir and indinavir have also been used in combination with nelfinavir.[60,110112] The favourable pharmacokinetic interactions between protease inhibitors (section 3.5) allow for the drugs to be administered in more convenient twice-daily regimens, potentially increasing patient adherence to therapy. Administration of nelfinavir 1250mg twice daily has been shown to be as effective as 750mg 3 times daily. Studies included in table IV are on-treatment analyses unless otherwise indicated.

Table IV
figure Tab4

Efficacy of nelfinavir (NFV) in combination with other protease inhibitors. Studies are OT analyses unless otherwise indicated.

A randomised nonblind study in protease inhibitor-naive (≈50% antiretroviral therapy-naive) patients with a high baseline viral load (≥10 000 copies/ml) evaluated the efficacy of 4 antiretroviral drug regimens in 157 patients with HIV infection (SPICE study).[106] The study used intention-to-treat analysis, although some data are provided for on-treatment analysis (table IV). In a group of 51 patients who received a 4-drug combination of nelfinavir, saquinavir soft-gel capsule (SGC) [both 3 times daily] and 2 NRTIs, mean plasma HIV RNA level had decreased by 1.96 log10 copies/ml from baseline and mean CD4+ count had increased by 216 cells/μl after 72 weeks of therapy. These results were similar to those for patients receiving either nelfinavir plus 2 NRTIs, saquinavir SGC plus 2 NRTIs, or nelfinavir and saquinavir SGC without NRTIs (table IV). However, among patients who met the protocol-defined conditions of virological relapse, time to relapse was significantly longer in patients receiving the 4-drug regimen than in those in other treatment arms (p = 0.005).

The efficacy of dual protease inhibitor therapy with nelfinavir and saquinavir (plus 2 NRTIs) administered 3 times daily was also shown in a nonblind study in which plasma HIV RNA levels were reduced by a median 2.25 log10 copies/ml at 44 weeks, with 9 of 10 patients exhibiting suppression below 500 copies/ml.[109]

A small randomised study was conducted in HIV-infected women (n = 64) who were treated with either twice- or 3-times-daily nelfinavir (1250 or 750mg, respectively) in combination with saquinavir HGC, stavudine and lamivudine (Women First study).[107] In these women, the twice-daily regimen was at least as effective as the 3-times-daily regimen, suppressing HIV viral load below 50 copies/ml in 80 versus 60% of patients on treatment through 48 weeks, and increasing the mean CD4+ cell count by 250 vs 180 cells/μl from baseline (table IV).

The efficacy of twice-daily nelfinavir treatment was also demonstrated in a large randomised nonblind study in which nelfinavir 1250mg was combined with saquinavir SCG 1200mg twice daily and an NRTI (TIDBID study) in antiretroviral therapy-experienced and -naive patients.[74,108] This twice-daily protease inhibitor regimen was as effective as either 3-times-daily saquinavir SGC 1200mg plus 1 NRTI or twice daily saquinavir SGC 1600mg plus 2 NRTIs. All regimens achieved HIV viral load below the LOD (400 and 50 RNA copies/ml) in ≥73 and 60% (respectively) of all patients receiving study drugs at 48 weeks (table IV).

The efficacy of nelfinavir 750mg 3 times daily or 1000mg twice daily in combination with indinavir, delaviridine and NRTIs was evaluated in a retrospective analysis (n = 25).[110] Most patients had previously received extensive antiretroviral therapy and were taking at least 2 NRTIs in addition to the 2 protease inhibitors and delaviridine during the 24-week study. Overall, a good antiviral effect was observed (table IV), despite the slightly antagonistic effect between nelfinavir and indinavir demonstrated in in vitro studies (section 2.1.1). Likewise, results of another small study combining nelfinavir and indinavir (with or without NRTIs) produced a good antiviral response; 10 of 13 patients (77%) had plasma HIV RNA levels <400 copies/ml after 44[60] and 64[112] weeks of therapy.

A small prospective nonblind study evaluated the efficacy of nelfinavir 500 or 750mg twice daily in combination with ritonavir 400mg twice daily in 20 patients.[111] Efficacy results were generally favourable (table IV), suggesting a promising regimen that may need to be combined with additional agents (e.g. NRTIs) for optimal and durable efficacy; however, moderate-to-severe diarrhoea was reported in about half of the patients (9 of 20).

4.4 Comparisons with Other Protease Inhibitors

Nelfinavir has been compared with other protease inhibitors in randomised nonblind studies in combination with abacavir in antiretroviral therapy-naive patients (n = 73)[113] and amprenavir in protease inhibitor-naive patients (n = 21).[114] In these studies, both of which were probably too small to detect statistically significant differences in efficacy between treatment groups, the various protease inhibitors demonstrated approximately equivalent efficacy in terms of reduction of HIV viral load or percentage of patients achieving an undetectable viral load after 24 or 48 weeks of therapy. Larger retrospective analyses reported some variability in efficacy between different protease inhibitors, particularly in antiretroviral therapy-experienced but protease inhibitor-naive patients.[115,116] Nelfinavir was more effective than indinavir, saquinavir, ritonavir or ritonavir/saquinavir with respect to the percentage of patients achieving undetectable viral loads after 4 months and the incidence of relapse (detectable viral load) at 6 months in 1 study; however, statistical significance was not reported (fig. 4).[116]

Fig. 4
figure 4

Percentage of HIV-infected patients with undetectable viral load at 4 months (n = 847; limit of detection not specified) and subsequent relapse at 6 months (n = 186) [retrospective analysis].[116] Patients received either nelfinavir (NFV), indinavir (IDV), ritonavir (RTV), saquinavir hard-gel capsules (SQV HGC) or RTV/SQV HGC as part of their initial protease inhibitor treatment regimen. Statistical significance was not reported.

A single-centre clinical cohort of 901 patients initiating protease inhibitor-containing therapy were followed for a mean of 15 months to assess factors determining viral response (achievement of plasma HIV RNA <500 copies/ml), time to response and response durability.[117] In this patient group, nelfinavir was associated with a 1.65 times greater probability of response than indinavir (p < 0.0001), although patients taking either drug as a first protease inhibitor were more likely to respond than those taking ritonavir or saquinavir (p ≤ 0.0009). The probability of viral rebound (viral load >500 copies/ml after an initial response) was 53% overall, and was highest among saquinavir recipients (p = 0.004 vs indinavir). Although therapy-naive patients were more likely to achieve an undetectable viral load than those with prior antiretroviral experience, this was not a factor in determining the durability of viral response. It should be noted that these patients were not randomised to treatment groups and that their initial prognoses were not necessarily directly comparable; therefore the results should be interpreted with caution.

Another single-centre observational study involving 337 patients treated with nelfinavir, indinavir or ritonavir in combination with RTIs for at least 16 weeks found no significant difference in viral response between the protease inhibitors.[118] Success of the therapy was dependent more on initiation of a new NRTI with the protease inhibitor than on previous therapy experience or which new protease inhibitor was used.

4.5 Changing Protease Inhibitor Therapy

4.5.1 While Viral Load is Suppressed

Toxicities of some protease inhibitor regimens may lead to poor adherence to drug therapy, and warrant conversion to a different therapy combination.[119] Several small, nonrandomised studies have assessed the antiviral efficacy of nelfinavir after switching from another protease inhibitor-based regimen. Typically, patients with undetectable HIV RNA (<400 copies/ml) who switched to nelfinavir from another protease inhibitor regimen experienced continued suppression of viral load for at least 12 weeks,[120122] and in some studies 24[123,124] or 36 weeks.[125] A small study (n = 14) showed that the probability of maintaining suppression of HIV RNA levels was lower in patients switched to nelfinavir compared with those remaining on indinavir for 9 months after achieving undetectable viral load.[126]

4.5.2 Nelfinavir as Salvage Therapy for Virological Failure

Use of a single protease inhibitor as part of antiretroviral therapy in HIV-infected individuals may limit the potential benefit of subsequent therapy with a different protease inhibitor because of the development of cross-resistance within this class of drugs.[127] Patterns of HIV protease mutations associated with resistance to nelfinavir are different from those seen with other protease inhibitors (section 2.3). Nonetheless, activity of the drug appears to be impaired in patients refractory to saquinavir.[128] Cross-resistance has also been observed in clinical isolates with multiple mutations associated with prolonged ritonavir therapy.[31] Cross-resistance to nelfinavir after therapy with indinavir or ritonavir was correlated with mutations at I54V and V82A/F,[129] whereas L90M was associated with nelfinavir cross-resistance after saquinavir therapy.[128]

Only a few randomised studies have investigated the efficacy of nelfinavir as part of salvage therapy in patients experiencing virological failure on other protease inhibitors (table V). The largest of these, study ACTG 359[130](n = 277), reported no significant difference between dual protease inhibitor regimens containing nelfinavir 750mg 3 times daily or ritonavir 400mg twice daily (both with saquinavir plus delavirdine and/or adefovir dipivoxil) in reducing plasma HIV RNA levels at 16 weeks after indinavir failure. In ACTG 359, the combination yielding the highest proportion of patients with undetectable viral load (47%; <500 copies/ml) was that of nelfinavir with saquinavir 800mg 3 times daily and delavirdine 600mg twice daily. A similar regimen containing ritonavir instead of nelfinavir resulted in 33% of patients having undetectable viral load at 16 weeks. Viral suppression was not as great in groups receiving adefovir dipivoxil instead of delavirdine in this trial. In a separate randomised study of salvage therapy after failure of indinavir-based treatment, multiple-drug regimens containing nelfinavir, efavirenz, adefovir dipivoxil and either NRTIs or abacavir were significantly more effective than combined regimens of the same drugs but without nelfinavir (ACTG 372B) [table V].[131]

Table V
figure Tab5figure Tab5

Efficacy of nelfinavir (NFV) as salvage therapy after virological failure on another protease inhibitor-based regimena

Noteworthy among the nonrandomised trials is a prospective nonblind study of 62 patients who failed treatment after at least 3 months of antiretroviral therapy [all patients had been exposed to NRTI (median duration 35.6 months) and protease inhibitor (median duration 12.2 months) therapy].[141] In this study, median reduction in viral load was 0.38 log10copies/ml after switching to nelfinavir-based therapy for a median period of 5.3 weeks. At baseline, RTI- and protease inhibitor-resistance mutations were detected in 90 and 89% of patients respectively, and the total number of resistance mutations was the only independent predictor of virological response after this short follow-up period.

Another prospective nonblind study evaluated the antiviral effect of nelfinavir combined with 2 NRTIs in 16 patients with plasma HIV RNA levels ≥5000 copies/ml after at least 6 months of saquinavir therapy (table V).[128] This salvage regimen resulted in minimal (0.59 log10 copies/ml) and transient (8 weeks) suppression of plasma HIV RNA levels. Failure of salvage therapy was associated with the baseline presence of the protease mutation L90M in the absence of D30N, indicating cross-resistance to nelfinavir due to a saquinavir-selected mutation. Consistent with the emergence of cross-resistance to nelfinavir by saquinavir selection, salvage therapy with nelfinavir and saquinavir was ineffective after failure of ritonavir/saquinavir combination therapy.[133]

Successful salvage therapy after protease inhibitor failure may depend on incorporation of a drug from a class to which the patient is naive (usually an NNRTI). Several small, nonrandomised studies have evaluated the combination of nelfinavir and nevirapine after failure of a regimen containing another protease inhibitor (usually indinavir or ritonavir) and NRTIs (table V).[134,136138] These studies have generally shown successful short term reductions in plasma HIV RNA (>1 log10 copies/ml at 8 to 12 weeks) in the majority of patients.[134,138] Studies extending to 18[137] or 36 weeks,[136] however, show that the proportion of patients maintaining this degree of viral suppression is probably closer to 20%. The combination of nelfinavir and efavirenz (with either stavudine[139] or didanosine plus hydroxyurea[140]) provided effective viral suppression for up to 6 months in NNRTI- and nelfinavir-naive patients with extensive experience on protease inhibitor/NRTI regimens. A dual protease inhibitor regimen of nelfinavir and saquinavir with efavirenz and NRTIs suppressed HIV viral load to undetectable levels in 8 of 10 patients for at least 3 months (and up to 12 months) in another small nonrandomised study.[142]

4.5.3 Salvage after Nelfinavir Failure

Nelfinavir appears to select for a different set of HIV protease variants from the other protease inhibitors (section 2.3), suggesting that patients treated with nelfinavir as initial therapy may remain susceptible to other protease inhibitors even after developing resistance to nelfinavir.

A cohort analysis assessed the virological response to a salvage regimen of ritonavir and saquinavir in 24 patients who had experienced virological failure during treatment with nelfinavir in 2 randomised double-blind trials (studies 511 and 506).[143] The 24 patients were from 2 of the centres involved in the randomised studies. Patients were switched from their nelfinavir-containing regimens to a combination of ritonavir 400mg, saquinavir 400mg, stavudine 40mg and lamivudine 150mg (all twice daily). All patients had been naive to protease inhibitors at the start of the nelfinavir trials. At the time of virological failure, HIV isolates from most patients (13 of 18 available samples) had the D30N mutation associated with nelfinavir resistance. Other common baseline mutations at crossover to the new protease inhibitors were N88D (8 of 18) and M36I (8 of 18). Changes at residues critical to the development of resistance to other protease inhibitors (G48V, V82A/F/T, and I84V) were not found. HIV viral load in all patients decreased below 500 copies/ml on at least 1 occasion during the first 6 months on the ritonavir/saquinavir salvage regimen. Viral load suppression was maintained through the entire 6 months in 17 patients (71%), with 10 patients (59%) achieving suppression below 50 copies/ml. The patient’s HIV RNA level (above or below 30 000 copies/ml) at the time of the switch was almost a statistically significant predictor of long term response (p = 0.05).

In a similar switch from a failing nelfinavir-based regimen to one containing ritonavir and saquinavir, intention-to-treat analysis showed that 55% of patients (n = 79) maintained suppression below 500 copies/ml for 24 weeks, and 77% achieved a reduction of >1 log10 copies/ml reduction in plasma HIV RNA levels.[144] These results were consistent with those from another small study in 6 patients.[145]

A nonrandomised, nonblind study (n = 128) evaluated rescue therapy subsequent to failure of 24 weeks’ therapy with protease inhibitor-based antiretroviral regimens (failure defined as HIV RNA >10 000 copies/ml on 2 separate occasions).[146] The investigators determined that initial therapy with nelfinavir gave a better chance of successful salvage therapy (i.e. that producing viral load below the LOD) than initial therapy with indinavir or ritonavir. The data for saquinavir as initial therapy were equivocal.

A retrospective study in 33 patients with viral load breakthrough (>400 copies/ml) while taking nelfinavir (after ≥1 prior protease inhibitor) showed that viral load could be reduced by 1.1 to 2.1 log10 copies/ml on treatment with efavirenz, another protease inhibitor, or both (protease inhibitor determined by genotype analysis; time frame not reported).[147]

As indicated in section 2.3 (and fig. 1), HIV isolates from patients who had failed nelfinavuir therapy were less likely to exhibit phenotypic cross-resistance to other protease inhibitors than those from patients failing other protease inhibitor therapy, according to a retrospective analysis.[36]

4.6 Patient Adherence

Patient adherence to therapy is critical in order to maintain adequate suppression of viral replication and to avoid the emergence of drug-resistant virus. A randomised, nonblind study in 112 antiretroviral treatment-experienced HIV-infected patients who received stavudine, lamivudine, and either indinavir or nelfinavir determined that adherence to therapy was significantly better among nelfinavir than indinavir recipients after 6 months of treatment (70 vs 48%; p = 0.0311).[148] Adverse events provoked discontinuation of treatment in 34% of patients in the indinavir group and in 12% in the nelfinavir group (p = 0.0075) [fig. 5]. Importantly, other larger analyses of nelfinavir showed much lower rates of discontinuation because of adverse events (4% of 696 patients during 24 weeks of therapy in 3 clinical trials[28] and 1.5% of 3312 patients receiving nelfinavir through the French compassionate use programme in 1997[149] (see section 5).

Fig. 5
figure 5

Adherence to therapy in a randomised nonblind prospective study of 112 antiretroviral therapy-experienced patients with HIV infection.[148] Patients received triple therapy with either nelfinavir (NFV) or indinavir (IDV) plus stavudine (STV) and lamivudine (LAM). Adequate adherence was not clearly defined. * p < 0.05, ** p = 0.0075 vs IDV group.

Analysis of adherence using a medication-vial electronic monitoring system in a small patient cohort (n = 21) found better adherence to nelfinavir than to indinavir (both 3 times daily) at week 5 of treatment (93 vs 78% of doses taken; p = 0.04).[150] This was believed to reflect the advantage of nelfinavir being administered with food.

The twice-daily regimen of nelfinavir 1250mg, which was recently approved by the US FDA, may improve adherence compared with the 3-times-daily regimen of 750mg because of less frequent dosage administration. In some patients the potential advantage of less frequent dosage administration may be tempered by the fact that the pill burden is increased to 5 tablets at each dosage time, although the new film-coated formulation has made it much easier for patients to swallow nelfinavir tablets. Better adherence to a twice-daily than a 3-times-daily regimen of nelfinavir was demonstrated using a semiquantitative measure (doses missed in the previous 3 days) in a small retrospective analysis of 17 patients, although statistical analysis was not provided in this brief report.[151]

4.7 Studies in Children

The efficacy of nelfinavir in paediatric patients (<16 years of age) has not been evaluated in randomised trials. Observational studies, however, have indicated that nelfinavir used as part of a treatment regimen with saquinavir[152,153] or efavirenz[80,154,155] and NRTIs is effective in reducing viral load, at least in the short term, and particularly in protease inhibitor-naive patients. Protease inhibitor-naive patients have also responded well to triple therapy with nelfinavir and NRTIs,[156158] although patient numbers were small in these studies (n = 14 to 36). In most (but not all) paediatric cohorts, patients with prior antiretroviral treatment tended to experience a rebound of plasma viraemia after initial suppression on nelfinavir therapy. In one prospective, nonblind study, patients with prior saquinavir experience responded better to nelfinavir than did patients experienced with indinavir or ritonavir.[159] In general, response to nelfinavir as a salvage therapy in paediatric patients has been minimal.[160,161]

A marked and sustained antiviral effect was observed in what appears to be the largest cohort of HIV-infected children (n = 57; age 3.6 to 16.8 years) receiving nelfinavir and evaluated in a prospective study.[80] The treatment regimen consisted of nelfinavir 20 to 30 mg/kg 3 times daily, efavirenz and 1 or more NRTI. All children had been previously treated with only NRTIs. Median baseline plasma HIV RNA level was 4.0 log10 copies/ml and CD4+ count was 699 cells/μl. After a follow-up period of 48 weeks, intention-to-treat analysis showed that 76% of children had plasma HIV RNA levels <400 copies/ml and 63% had levels <50 copies/ml. A statistically significant but moderate increase in CD4+ cell count was also noted, despite most children having values within the normal range for age at baseline.

In order to alleviate some of the challenges to adherence in paediatric patients, one small study analysed the efficacy of twice-daily administration of nelfinavir 45 to 55 mg/kg to 10 patients aged 5 to 15 years.[162] In this patient population, the twice-daily regimen was found to be more effective than the standard 20 to 30 mg/kg 3-times-daily regimen in terms of virological and immunological response, and was associated with fewer adverse events.

A study of nelfinavir therapy in HIV-infected infants (n = 19; mean age 11.5 weeks) found significant virological response in 47% (9/19) of patients, which was sustained in 42% for a mean duration of 43 weeks. CD4+ cell counts remained stable or increased in all evaluable patients.[163]

5. Tolerability

In general, nelfinavir has been well tolerated in clinical studies and discontinuation of therapy because of adverse events has occurred in only a small proportion of patients. Adverse events of moderate or severe intensity reported in at least 2% of patients enrolled in 2 randomised double-blind studies with nelfinavir are presented in table VI. In 3 phase II/III trials, 4% of a total of 696 patients discontinued treatment as a result of adverse events during 24 weeks of therapy.[28] Of 3312 patients receiving nelfinavir through the French compassionate use programme in 1997, approximately 7% presented with an adverse event (0.7% with a serious adverse event) suspected to be related to nelfinavir use.[149] Only 1.5% of patients discontinued treatment because of an adverse event, the most frequent being diarrhoea and rash.

Table VI
figure Tab6

Most frequently reported adverse events of moderate or severe intensity and laboratory abnormalitiesa in 2 randomised, double-blind nelfinavir (NFV) trials[47]

A retrospective comparison of tolerability profiles of 4 different protease inhibitors (n = 69) indicated that nelfinavir was better tolerated than indinavir, amprenavir or saquinavir, causing no drug discontinuation due to major adverse events during 48 weeks of treatment.[164] A prospective, nonrandomised multicentre study in 1209 patients (mean follow-up = 10.7 months) showed nelfinavir to be associated with the lowest incidence of serious adverse events (grade 3 or 4; 0.45%) when compared with ritonavir (2.25%), ritonavir/saquinavir HGC (1.75%) or indinavir (1.24%).[165]

5.1 Gastrointestinal Disturbance

Diarrhoea is the most common adverse event associated with nelfinavir, but is generally of mild to moderate intensity and can usually be controlled with drugs that slow gastrointestinal motility (e.g. loperamide).[47] Thus, diarrhoea has resulted in minimal discontinuation of therapy in clinical trials.

Diarrhoea of moderate or severe intensity occurred in 20% of antiretroviral therapy-naive patients receiving triple therapy with nelfinavir 750mg 3 times daily, lamivudine and zidovudine in a 24-week randomised double-blind trial, compared with 3% of patients treated with lamivudine and zidovudine only (Study 511).[47] In a nonblind randomised study (study 542), moderate or severe diarrhoea was reported by 18% of patients receiving nelfinavir 1250mg twice daily and by 14% of those receiving nelfinavir 750mg 3 times daily combined with stavudine and lamivudine for 48 weeks (table VI).[166] In a retrospective analysis of patients receiving nelfinavir-containing regimens at a single US clinic (n = 181), 25% of patients reported diarrhoea, most of which was mild (grade 1) and did not require treatment.[167] In these studies, this adverse effect was manageable with antidiarrhoeals, was not dose-limiting, and led to study discontinuation in fewer than 2% of patients. A comparison of adverse events derived from the European summary of product characteristics gave nelfinavir the best gastrointestinal tolerability profile among the available protease inhibitors, reporting only diarrhoea and a low incidence of vomiting (4.5%), but no other gastrointestinal event.[119]

Interestingly, the incidence of ≥ grade 2 diarrhoea was relatively low (3%) in the Women First study among the 31 women randomised to receive nelfinavir 750mg 3 times daily and saquinavir HGC 600mg 3 times daily as part of a 4-drug regimen.[107] While these preliminary results might suggest a gender difference for this adverse event [i.e. most other clinical trials included only a small percentage (e.g. ≈15 to 20%) of female patients and showed a higher incidence of diarrhoea], the incidence was 15% among the 33 women randomised to receive nelfinavir 1250mg twice daily and saquinavir HGC 1000mg twice daily as part of a 4-drug regimen in the Women First study.

5.2 Metabolic Disturbances

As a class, protease inhibitors have been associated with a syndrome of hyperlipidaemia, insulin resistance and lipodystrophy. In a single-centre cross-sectional study, 64% of protease inhibitor recipients (compared with 3% of protease inhibitor-naive patients; p = 0.0001) exhibited fat redistribution associated with elevated serum triglyceride, insulin and C-peptide levels and increased insulin resistance.[168] 113 protease inhibitor recipients were reassessed after a further 8 months of therapy (mean total duration of protease inhibitor therapy 21 months), and lipodystrophy was reported in 83% of patients [compared with 1 of 28 protease inhibitor-naive patients (4%)].[169] Although patients were not randomised to protease inhibitor therapy in this study, the incidence and severity of lipodystrophy associated with nelfinavir (with or without saquinavir) was similar to that associated with other protease inhibitors. However, as there is no consensus case-definition for the syndrome, the reported incidence among patients receiving protease inhibitors varies widely, and a direct causal link between protease inhibitors and lipodystrophy has not been established.[170] Abnormal fat disposition has also been observed occasionally in HIV-infected patients treated with non-protease inhibitor-containing regimens.[169,171,172]

Hyperglycaemia[173,174] and hypercholesterolaemia[175177] have been associated with treatment with protease inhibitors, including nelfinavir. Several small studies have suggested that the risk of these disturbances may be highest with ritonavir.[175,178,179] Some improvement in hyperglycaemia has been reported in patients switching from indinavir to nelfinavir.[180]

New-onset diabetes mellitus occurs infrequently in HIV-infected patients receiving protease inhibitors, including nelfinavir, and is often associated with a family history of diabetes mellitus and elevated serum cholesterol and triglyceride levels.[181184]

An analysis of adverse events reported in 5 nelfinavir clinical trials (n = 1229 total) determined that the risk of events possibly related to increased serum lipid levels (e.g. myocardial infarction, angina, stroke, pancreatitis, peripheral vascular disease or fat redistribution) was similar between nelfinavir and non-nelfinavir treatment regimens.[185] The incidence of myocardial infarction was not increased with any available protease inhibitor during 1 year of therapy in clinical trials (amprenavir was not evaluated),[186] although this is a relatively short follow-up period to monitor for such events.

5.3 Effects in Children

Nelfinavir is the preferred initial protease inhibitor for children because of the availability of a palatable oral powder formulation and fewer adverse effects.[187] Aside from diarrhoea, adverse events associated with nelfinavir use in children included asthenia, abdominal pain, headache, nausea, flatulence and rash. Laboratory abnormalities (e.g. decreased haemoglobin, increased AST/ALT or increased triglyceride levels) occurred infrequently.[187] Although rash was the most frequent adverse event reported in a recent study of 57 children receiving nelfinavir, efavirenz and 1 or more NRTIs, the incidence was similar to that with efavirenz therapy in HIV-infected adults.[80]

5.4 Other Effects

The most frequently reported marked laboratory abnormalities (defined as a shift from Grade 0 at baseline to at least Grade 3 or from Grade 1 to Grade 4) and moderate or severe treatment-emergent adverse events occurring in patients receiving nelfinavir in 2 double-blind clinical trials are listed in table VI.

Thromboembolic disorders have been noted in a small proportion of patients receiving protease inhibitors, including nelfinavir.[188,189] Although systemic hypertension has been associated with indinavir therapy, no correlation has been observed with nelfinavir.[190]

Cases of nelfinavir-induced urticaria have been reported infrequently. Some patients were successfully desensitised to nelfinavir via dose escalation over a period of several days to 3 weeks.[191,192]

There has been one reported case of severe acute pancreatitis that was positive upon rechallenge with nelfinavir (i.e. all drugs had been discontinued, and nelfinavir was the first reintroduced drug).[193]

Severe hepatotoxicity has been reported in patients taking protease inhibitors; however, the increased incidence appears to be primarily associated with ritonavir (4.1-fold higher risk compared to other protease inhibitors, including nelfinavir).[194]

One case of worsening peripheral neuropathy was attributed to the addition of nelfinavir to NRTI combination therapy in a patient with extensive previous exposure to antiretroviral therapy.[195]

Although a causal relationship has not been established, there have been reports of increased spontaneous bleeding in patients with haemophilia receiving protease inhibitors, therefore caution is advised when using nelfinavir in patients with this condition.[47]

6. Dosage and Administration

Nelfinavir should be used in combination with other antiretroviral agents in patients with HIV infection. Because its oral absorption is greatly enhanced in the presence of food, nelfinavir should be taken with a meal or light snack. In patients aged >13 years, the recommended oral dosage is 750mg (3 x 250mg tablets) 3 times daily, and an alternative regimen of 1250mg (5 x 250mg tablets) twice daily was recently approved by the US FDA.[47] The recommended dosage in paediatric patients aged 2 to 13 years is 20 to 30 mg/kg 3 times daily. The new film-coated formulation[47] has made it easier for patients to swallow nelfinavir tablets. However, patients unable to swallow tablets may place whole tablets or crushed tablets in a small amount of water to dissolve before ingestion or they may mix crushed tablets in a small amount of food. Children unable to swallow tablets may take nelfinavir oral powder mixed with a small amount of water, milk, soymilk, milk- or soy-based formula or liquid dietary supplement. Acidic food or juice is not recommended with nelfinavir powder, as the combination may result in a bitter taste.[47]

Nelfinavir administration has not been associated with fetal abnormalities in studies in rats and rabbits.[47] However, as only limited data are available on use during human pregnancy,[196] nelfinavir should be used in pregnant women only in the absence of other therapeutic options. Nelfinavir is excreted in breast milk in lactating rats, although it is not known if it is excreted in human milk.

Nelfinavir is an inhibitor of CYP3A4 metabolism, and thus can interfere with the clearance of certain coadministered drugs (section 3.5). Concomitant administration of nelfinavir with astemizole, cisapride, triazolam, midazolam, ergot derivatives, amiodarone or quinidine is contraindicated because of the potential for serious and/or life-threatening toxicities.[47] The contraindication also applies to terfenadine, although this drug has been removed from most markets including the US. Rifabutin should be taken in half the normal dose when coadministered with nelfinavir. Caution is also advised if nelfinavir is used concomitantly with other CYP3A4 substrates such as sildenafil, dihydropyridine calcium antagonists, immunosuppressants and HMG-CoA reductase inhibitors (section 3.5). In particular, concomitant administration of nelfinavir with lovastatin or simvastatin is not recommended because of the increased risk of myopathy including rhabdomyolysis.[47]

Rifampicin and other CYP3A4 inducers including the anticonvulsants carbamazepine, phenobarbital and phenytoin may decrease nelfinavir plasma concentrations below clinically effective levels, and coadministration is therefore not recommended.[47]

In patients using oral contraceptives, alternative or additional contraceptive methods should be used during nelfinavir treatment, as the plasma concentrations of ethinyl estradiol and norethindrone are reduced by nelfinavir and thus their efficacy cannot be guaranteed.[47]

The pharmacokinetics of nelfinavir have not been studied in patients with impairment of renal function, other than a case report (see section 3.6). As renal clearance of nelfinavir is negligible, clinically relevant decreases in nelfinavir clearance are not expected in renally impaired patients. Nelfinavir should be used with caution in those with impaired hepatic function because the drug is metabolised principally by hepatic microsomal enzymes, and limited data in patients with moderate to severe liver disease indicate significant changes to the pharmacokinetics of nelfinavir (section 3.6).

7. Place of Nelfinavir in the Management of HIV Infection

The goal of antiretroviral therapy is to achieve substantial prolonged suppression of viral replication, ultimately limiting the development of symptomatic HIV disease and AIDS. The point at which antiretroviral therapy should be initiated during the course of HIV infection has not been clearly defined and many factors must be taken into consideration. Once a decision is made to start therapy, current international guidelines recommend the use of 2 nucleoside analogues and an HIV protease inhibitor as a first-line option for patients with HIV infection.[14]

HIV protease inhibitors are among the most powerful available drugs for suppressing viral replication. When used in combination with NRTIs and/or NNRTIs, plasma viraemia can be suppressed below the LOD of ultrasensitive RNA assays (<50 copies/ml) in patients with low baseline viral loads. Maintenance of this level of viral suppression is critical to avoid the development of viral drug resistance and to ensure the long term efficacy of protease inhibitor therapy.[87] For this reason, suboptimal dosages to reduce drug toxicity are not recommended, nor are concurrent therapies that reduce protease inhibitor plasma concentrations. Patient adherence is also critical. Therefore there is a need for regimens requiring fewer tablets or capsules and longer intervals between doses, fewer adverse effects and fewer drug interactions.

Nelfinavir is a protease inhibitor which exhibits several properties compatible with long term administration and convenient use in therapy, including metabolic stability allowing a relatively long elimination half-life, good oral bioavailability, and minimal toxicity. Its unique resistance profile suggests that the potential for cross-resistance to other protease inhibitors in the event of nelfinavir failure may be lower than that of other drugs of this class (section 2.3). Preclinical studies in animals indicate penetration of nelfinavir in lymph nodes, spleen, brain and testes (section 3.2). Partial reductions in viral replication in CSF and genital mucosa have been demonstrated in patients using nelfinavir-containing combination therapies (section 2.1.3).

The effects of nelfinavir on the clinical end-points of HIV infection (disease progression and death) have not yet been studied. However, in patients with less advanced disease, a substantial proportion of those remaining on therapy achieved and sustained levels of plasma HIV RNA associated with long term nonprogression of the disease, while simultaneously sustaining increased CD4+ cell numbers. A low baseline viral load at initiation of nelfinavir treatment was predictive of successful viral load suppression without relapse.[86] The improvement in these surrogate markers indicates a substantial reduction in the risk of future disease progression. Although few randomised, double-blind trials have been carried out, nelfinavir appears to be an effective, well tolerated and convenient option as part of triple and quadruple therapy regimens that result in durable suppression of viral replication over treatment periods up to 96 weeks (section 4).

Nelfinavir therapy was most successful in antiretroviral therapy- or protease inhibitor-naive patients (sections 4.1 to 4.3). In most cases viral load suppression could also be maintained in patients who changed their protease inhibitor to nelfinavir for reasons of intolerance to their previous regimen while plasma viral RNA was undetectable (section 4.5.1).

Development of resistance to antiretroviral drugs is of significant concern in HIV therapy. HIV protease inhibitors have been shown to select for HIV variants that are cross-resistant to other drugs of this class, thereby limiting subsequent therapeutic options. Nelfinavir showed little utility in patients experiencing virological failure on another protease inhibitor (section 4.5.2), indicating some degree of cross-resistance. Patient outcomes were better, however, when nelfinavir was used as initial therapy; failure of nelfinavir-based regimens did not preclude success with subsequent protease inhibitors (section 4.5.3). This is thought to be due to the fact that nelfinavir selects for a unique HIV protease variant that does not confer cross-resistance to other protease inhibitors in vitro. Nonetheless, prolonged nelfinavir therapy under suboptimal viral suppression leads to selection of secondary mutations with broad cross-reactivity to several protease inhibitors. This fact emphasises the importance of aggressive treatment, relying heavily on strict patient adherence.

The activity of the various HIV protease inhibitors in triple therapy regimens with NRTIs are similar (section 4.4);[197] therefore the choice of which drug to use will depend on tolerability profiles, pharmacokinetic and metabolic interactions with concomitant medications, convenience of administration, resistance profiles and the potential for salvage if the initial regimen fails. Compartmental penetration to reservoirs of HIV replication such as the lymph nodes, CNS and genital tract should also be considered.

The low incidence of serious adverse events associated with nelfinavir, as well as the ability to administer the drug twice daily (amprenavir and ritonavir are also approved for twice-daily administration) and with food, contribute to relatively good patient adherence (section 4.6). Diarrhoea was the only adverse event of severity grade 2 or greater reported by ≥10% of patients in clinical trials (section 5.1), and was not dose-limiting. Metabolic disturbances associated with protease inhibitor use (hypercholesterolaemia, hyperglycaemia and lipodystrophy) have been reported infrequently with nelfinavir. Although no clear advantage of nelfinavir over other protease inhibitors has been established in this regard, several small studies have suggested that the incidence of hyperglycaemia and hypercholesterolaemia may be higher with ritonavir than other protease inhibitors (section 5.2). Nelfinavir was effective and well tolerated in children, with adherence benefitting from the availability of a palatable oral powder formulation (section 5.3). In adults, twice-daily nelfinavir regimens were shown to be at least as effective as 3-times-daily regimens in suppressing HIV viral loads and increasing CD4+ cell counts in several clinical trials (sections 4.1 to 4.3).

The tolerability profile of nelfinavir is preferable to that of ritonavir, which is commonly associated with gastrointestinal disturbances, asthenia, circumoral paraesthesia and headache, or indinavir, which induces nephrolithiasis.[105] It is recommended that indinavir is taken every 8 hours on an empty stomach, and patients must maintain adequate hydration to reduce the formation of kidney stones.[198] On the basis of in vitro data, ritonavir has the greatest and nelfinavir and saquinavir have the weakest inhibitory effect on CYP3A4 activity (section 3.5), but it is unclear whether this translates into fewer clinically significant CYP3A4-mediated drug interactions with nelfinavir than with other agents such as ritonavir or indinavir in patients with HIV infection.

Because cross-resistance to multiple protease inhibitors does not appear to emerge as frequently during treatment with nelfinavir as with other protease inhibitors (sections 2.3 and 4.5.3), salvage therapies after nelfinavir failure may have more potential for success than those attempted following another failed protease inhibitor regimen.

In conclusion, nelfinavir, in combination with RTIs and/or other protease inhibitors, is effective in limiting HIV replication and increasing CD4+ cell counts in HIV-infected adults and children. The convenience of its dosage administration (twice-daily administration with food is now approved in the US), the low incidence of serious adverse events, and the potential for salvage therapies indicate that nelfinavir (as part of combined antiretroviral therapy) should be considered as a first-line option in protease inhibitor-naive patients and in those unable to tolerate other protease inhibitors.