New drugs

The advent of highly active antiretroviral therapy (HAART) has revolutionized the outlook for individuals living with HIV, with both physicians and patients enjoying the associated reductions in morbidity and mortality [1,2]. Unfortunately, despite an increase in the choice of antiviral agents, HAART may still fail. There may be several reasons for this, including an increase in primary HIV resistance and a relative lack of potency of some regimens. In addition, high levels of adherence are necessary and the development of toxicities may lead to difficulties in compliance with regimens. Unfortunately, when individuals develop resistance to one antiretroviral agent within a class there is often, if not always, development of cross-resistance to other agents to which they are naïve. Therefore, although the HIV treating physician has a multitude of agents to choose from, there is still a need for additional antiretroviral agents. New strategies have been developed to try and address the issues of compliance and toxicity. These include the development of combination pills, and the use of simplified regimens once viral suppression has been achieved. In addition, the use of two-agent, nucleoside-sparing regimens and even single-agent therapy is under investigation. New drugs are being developed in existing classes (Fig. 1) to reduce toxicity and to overcome resistance, and also in new classes to improve the antiviral options available to patients. CD4 cell and feplication cycle of HIV demonstrating sites of activity of new drugs in development. The presently available nucleoside combinations are difficult to surpass in terms of both compliance and toxicity. Two combination pills dosed as a single tablet once a day have recently been licensed: Truvada (tenofovir/emtricitabine) and Kivexa (abacavir/lamivudine). These combinations have relatively low toxicities, although abacavir is associated with a hypersensitivity reaction in approximately 8% of patients and tenofovir may be associated with bone and renal toxicity in the long term. The risk of development of lipodystrophy with both these agents appears to be less than with thymidine analogues. New NRTIs are therefore principally being developed to overcome resistance. Recently, two new NRTIs have been withdrawn, stavudine extended release [3,4] secondary to manufacturing problems and amdoxivir [diaminopurine dioxolone (DAPD)] secondary to the development of lens opacities [5]. However, there are many other agents in this class under investigation. SPD754 [(–)dOTC] is a novel NRTI active against NRTI-resistant strains. It is a cytidine analogue with no mitochondrial toxicity in vitro [6]. In a 10-day double-blind placebo-controlled monotherapy study in 63 naïve individuals, no drug-related adverse effects were identified and no new nucleoside mutations occurred during follow-up [7]. Activity was significantly greater than placebo at 7 and 10 days at a dose of 400 mg twice a day (bid), while the 1200 mg once a day (qd) and 800 mg bid doses performed significantly better than 400 mg bid, being associated with decreases in viral load of 1.65 and 1.59 log10 HIV-1 RNA copies/mL at 10 days, respectively, compared with −1.4 log10 copies/mL in the 800 mg arm. SPD754 is active against HIV strains resistant to zidovudine and lamivudine. Research into SPD754 is currently suspended while details of toxicity in cynomolagous monkeys are investigated. Elvucitabine, or beta LFD4C, has anti-HIV activity against strains of HIV with NRTI mutations. The mean IC50 for wild-type virus is 0.15 μm, compared with 2.24 μm for M184 V and 3.05 μm for M41L, T215Y and M184 V in combination; all of these concentrations can be achieved with clinical doses. Importantly, beta LFD4C also shows activity against the multiple NRTI-resistant mutations Q151 M and the 69SSS insertion [8]. In a study of individuals harbouring virus with the M184 V mutation, elvucitabine at doses of 50 and 100 mg qd achieved a viral load fall at 28 days of between 0.65 and 0.75 log10 copies/mL, compared with a small increase in viral load in individuals who continued with lamivudine [9]. Elvucitabine may be dosed once daily and appears to have limited mitochondrial toxicity. It also has activity against the hepatitis B virus. Reverset (D-D4FC) is a cytidine analogue which has been shown to be active in both naïve and HIV-experienced individuals and has in vitro activity against a variety of NRTI-resistant strains [10]. In naïve individuals receiving D-D4FC monotherapy, the maximum decrease in viral load obtained after 10 days of therapy was nearly 1.8 log10 copies/mL, with 88% of individuals achieving a viral load of less than 400 copies/mL [11]. No resistance mutations developed. In experienced individuals who received D-D4FC at a dose of 200 mg qd, the viral load fall was 0.8 log10 copies/mL during a similar dosing period, with four of eight individuals achieving a viral load of less than 400 copies/mL [12]. D-D4FC is dosed once daily and no toxicity has yet been demonstrated. Pharmacokinetic studies suggest that D-D4FC will be active (i.e. concentrations are achievable) against both M184 V and K65R mutations [11]. Racivir is an oxothialone analogue closely related to FTC and has been compared to lamivudine in clinical trials, as part of a triple-agent HAART regimen with stavudine and efavirenz. At 14 days, similar viral load reductions were achieved in the two study arms using doses of racivir between 200 and 600 mg [13]. Interestingly, after the dosing period, the return of the viral load towards baseline was slower in all racivir arms compared with the lamivudine arm. Further studies with racivir are presently recruiting in lamivudine-resistant individuals. Unfortunately, the development of resistance to one of the presently available NNRTIs results in cross-class resistance. Therefore, the major development strategy for new nonnucleosides, which are often the anchor drug for therapy-naïve individuals, is the development of agents with novel resistance. Capravirine has been shown in vitro to have activity against strains of virus resistant to presently available NNRTIs (both with single and multiple mutations) [14]. In one study, capravirine activity was similar in individuals with wild-type virus and those harbouring K103N or other mutations associated with NNRTI resistance [15], but superior virological success could not be demonstrated for added capravirine above that of a standard three-drug salvage regimen for NNRTI failures. Of the patients who remained on capravirine after study termination, viral loads remained undetectable in nearly half after at least 39 months [16], and of those who developed virological failure 70% remained sensitive to the agent [17]. TMC125 is a flexible diarylpyrimidine (DAPY) compound with potent in vitro activity against both wild-type and NNRTI-resistant virus [18]. In individuals who were naïve to antiretroviral therapy, 7 days of TMC125 monotherapy produced a viral load fall of almost 2 log10 copies/mL [19]. In a similar study in individuals who received TMC125 but who had received previous NNRTI therapy, the viral load fall, although less than in the naïve study, was almost 1 log10 copies/mL [20]. The development of resistance to TMC125 appears, at least in vitro, to be slower than that to available NNRTIs [21]. Several benzophenone NNRTIs are presently in development. In a HeLa-Magi assay, antiviral potency of three of these compounds was demonstrated against wild-type and virus resistant to currently used agents in the NNRTI class [22]. As with other available classes, several new PIs are being developed, principally to overcome resistance to presently available drugs. Tipranavir is a nonpeptidic PI. In a study in individuals with documented PI resistance and at least three key PI mutations at positions 33, 82, 84 and 90, a viral load fall of 1.2 log10 copies/mL occurred after 2 weeks of tipranavir monotherapy or when tipranavir was added to optimized background therapy [23]. However, this was not sustained over subsequent weeks of follow-up, as there were frequently no other active drugs in the regimen. Tipranavir may also substantially lower the levels of other PIs including amprenavir, lopinavir and saquinavir by between 40 and 70% as a result of potent induction of the cytochrome P450 system [23]. Two studies comparing ritonavir-boosted tipranavir with standard-of-care boosted PI regimens, the RESIST studies, have presented data to 24 weeks. In the RESIST I study [24], patients failing PI-containing HAART underwent resistance baseline testing with pre-selection of a regimen by the investigator. Following randomization to either investigator-chosen boosted PI or ritonavir/tipranavir, individuals commenced therapy. By an intent-to-treat (noncompleter=failure) analysis, the proportion of individuals at 24 weeks experiencing at least a reduction in viral load of 1 log10 copies/mL was 41.5% in the tipranavir arm and 22.3% in the comparator arm, with median reductions in viral load of 0.88 and 0.28 log10 copies/mL, respectively. The proportion achieving an undetectable viral load, to below 400 or below 50 copies/mL, was increased by the concomitant use of enfurvitide; this demonstrates the importance of a second effective antiviral agent in the regimen. Similar results were reported in the RESIST II study, which recruited from Europe and Latin America as opposed to the North American cohort studied in RESIST I [25]. The activity of tipranavir is in proportion to previous PI resistance. Although, initially, it was felt the important resistance were the universal protease associated mutations (UPAMs) at positions 33, 82, 84 and 90 a more complicated resistance profile has recently been reported with the number of mutations at positions 13, 20, 33, 35, 36, 45, 46, 47, 54, 59, 66, 69, 71, 74, 82 (not A), 84 and 91 being associated with activity of tipranavir [26]. TMC114 is a ritonavir-boosted PI which is being developed for PI-experienced patients. It is sulpha-based and therefore may be associated with rash. In an open-randomized study, 50 individuals failing therapy, with a CD4 count of >50 cells/μL, a viral load of >2000 copies/mL and prior use of 2–4 PI inhibitors, received TMC114 at dosages of 100, 300, 600 and 900 mg. All dose arms showed greater activity than the control arm [27]. Adverse events were few, with diarrhoea, rash and headaches being the most commonly reported. A total of 497 patients who were three-class experienced with a minimum of one major PI mutation were randomized to receive TMC114 at four separate doses: 400 mg qd, 400 mg bid, 800 mg qd, 600 mg bid or investigator-chosen PI with optimized background. The most impressive results were obtained with a dose of 600 mg bid with a mean viral load fall of 1.85 log10 copies/mL at 24 weeks and 47% with <50 copies/mL, compared with 0.27 log10 copies/mL and 9% in the comparator arm [28]. Several investigational drugs are being developed in new classes. These include entry inhibitors, attachment inhibitors, co-receptor antagonists, integrase inhibitors and drugs against other targets such as zinc finger inhibitors and trans-acting transcription (TAT) antagonists. HIV attachment and fusion: targets for inhibition. Integration is a multi-step reaction. The process of HIV attachment infusion entails CD4–glycoprotein 120 (gp120) binding followed by co-receptor binding, utilizing either the C-C chemokine receptor-5 (CCR5) or CX chemokine receptor-4 (CXCR4) co-receptor followed by viral/cell fusion. BMS4880043 is a novel attachment inhibitor which has completed a phase Ib dose ranging study in 30 treatment-naïve individuals. At a dose of 800 mg qd, the maximum viral load fall was 0.7 log10 copies/mL, while a dose of 1800 mg achieved a maximum viral fall of 1.2 log10 copies/mL. However, activity varied from individual to individual, with some displaying only minimal activity and others experiencing a greater than 1.5 log10 copies/mL decrease. Side effects were minimal and no serious adverse events were observed [29]. Other attachment inhibitors in development are TNX355, which is an anti-CD4 monoclonal antibody made from a humanized immunoglobulin G4 (IgG4) anti-CD4 molecule and which acts by inhibiting viral fusion prior to virus–CD4 binding. In a phase I dose escalation study in 30 HAART-experienced individuals, a maximum viral load change from baseline was observed with a dose of 10 mg/kg with a decrease in viral load of almost 1.5 log10 copies/mL [30]. A multidose study in HAART-experienced patients showed that intermittent TNX355 for 9 weeks resulted in peak viral load declines of approximately 1 log10 copies/mL [31]. PRO542 is a soluble attachment inhibitor formed from human immunoglobulin fused with the HIV binding domains of the CD4 molecule. It neutralizes a broad range of HIV variants in vitro [32]. In a phase II study in HAART-experienced individuals, a single dose of PRO542 produced significant reductions in viral load, most markedly in the advanced-disease group with a mean drop of 0.5 log10 copies/mL at 4–6 weeks [33]. Several co-receptor antagonists against both CXCR4 and CCR5 are presently in development. The most advanced of these is the Pfizer compound UK427857, which is an oral CCR5 inhibitor with an IC90 of approximately 2 nmol [34]. In a phase IIa placebo-controlled study of 25 mg qd vs. 100 mg bid vs. placebo for 10 days, the greatest viral load decline was in the 100 mg bid arm, with a decrease of 1.42 log10 copies/mL and >90% receptor saturation throughout the dosing period [35]. In an individual with a mixed R5 and X4 population at baseline there was an apparent relative increase in the X4 virus, although this reverted to the pre-dose R5:X4 ratio after dosing was completed. Unfortunately, the number of individuals harbouring a CCR5 virus declines with CD4 count, although some individuals utilize this co-receptor despite advanced disease. Other CCR5 antagonists being developed include the Schering D, which has shown activity of approximately 1.5 log10 copies/mL during a 15-day single-drug dosing period [36]. Development of the CXCR4 inhibitors is relatively slow. The initial drug in this class, AMD3100, did show activity against pure CXCR4 virus [37], but has been replaced by other analogues, AMD070 [38] and AMD887 [39], both of which exhibit activity against multiple-drug-resistant virus in vitro. The development of resistance to receptor antagonists is unclear. Resistance may theoretically occur as a result of (1) changes in the receptor allowing the virus, but not the drug, to bind; (2) switching to the alternative receptor; or perhaps (3) use of another co-receptor. Integration is a multiple-step reaction (Fig. 3) and integrase inhibitors are being developed by several companies at the present time. Integrase inhibitors have shown activity in rhesus monkeys with one, L870812, resulting in a sustained viral load fall of between 1 and 4 log10 copies/mL and minimal decline in CD4 count when used as monotherapy [39]. These agents are entering phase 2 studies at the present time. PA457 specifically blocks a late step in the processing of the HIV Gag protein, namely conversion of the capsid precursor to mature capsid protein. In a phase I/IIa clinical trial of PA457, when administered at a single dose to HIV-infected individuals not on other therapy, a reduction in viral load of approximately 0.7 log10 copies/mL was seen in individuals receiving high doses. PA457 was well tolerated and plasma concentrations of the compound reached levels significantly greater than those predicted to provide a therapeutic benefit. Several new drugs are under development for the treatment of HIV infection. As treatment in naïve individuals improves, new treatments should have at least equivalent, or possibly better, antiviral activity, but should also have less toxicity and be easy for the patient to comply with. The development of agents in new and existing classes gives those individuals who have previously received therapy hope in the continuation of their fight against HIV-associated disease.