HIV attachment inhibitors: three more join the ranks at CROI

Three companies presented promising data on drugs which block the entry of HIV into cells at last week's Eleventh Conference on Retroviruses and Opportunistic Infections in San Francisco, indicating that even after the termination by Roche of research into its fusion inhibitor T-1249, the development of new drug classes for HIV treatment remains promising.

TNX-355

The drug that is closest to large-scale studies is a monoclonal antibody designed to coat CD4 receptors on cells. HIV uses the CD4 receptor and another co-receptor, such as CCR5 or CXCR4, to gain entry to cells, so a variety of approaches are being investigated to block attachment and fusion, with the eventual aim that they will be used in combination.

TNX-355 was unveiled at last year's Retroviruses conference. The study presented this year tested various intravenous doses in treatment-naive and treatment-experienced individuals (off treatment for at least twelve weeks or on stable background therapy with viral load above 5,000 copies/ml).

TNX-355 need only be dosed once a week in order to maintain sufficient antiviral activity. This study tested weekly dosing or fortnightly/biweekly dosing in 22 patients in the following pattern:

• 10mg/kg every seven days for 10 weeks
• 10mg/kg on the first day followed by 6mg/kg every 14 days, starting at day 7 for six doses.
• 25mg/kg every 14 days for five doses over 8 weeks.

The mean viral load in this study was in the region of 60,000-80,000 copies/ml in each arm and the mean CD4 cell count ranged from 313 cells/mm³ in the weekly arm to 390 cells/mm³ in the higher dose fortnightly arm.

After completing the treatment course, 23% of participants experienced a viral load reduction of greater than -1.4 log₁₀ copies/ml, whilst 64% experienced viral load reductions of greater than -1 log₁₀ copies/ml. In the majority of participants, virologic control appeared to wane as the study continued, suggesting that effective monotherapy with TNX-355 quickly results in drug resistance. Virus isolates from 16 patients showed reduced susceptibility to TNX-355 by week 9.

TNX-355 will now be studied in treatment-experienced individuals receiving background therapy that has been optimised by resistance testing, in order to determine whether its virologic effect can be sustained. These results suggest that functional monotherapy with this drug will be pointless.

SCH-D

Schering Plough presented the first information on its chemokine antagonist SCH-D. Development of a previous agent, SCH-C, was abandoned partly due to toxicity in animal studies but also because SCH-D proved more potent in vitro. SCH-D also has a longer half-life, better absorption and higher bioavailability in rat and monkey studies.

Schering Plough conducted a dose-escalated 14-day study of SCH-D, given as monotherapy to a total of 48 patients. In the first group, twelve individuals received SCH-D 10mg twice daily and three received placebo. In the second group, the same numbers of participants were randomised to 25mg or placebo, and in the third group to 50mg or placebo.

All participants had been off antiretroviral therapy for at least eight weeks prior to starting SCH-D treatment, which was administered on an inpatient basis. Mean viral load varied across the groups: 81,053 copies/ml in the placebo group, 36,299 copies/ml in the 10mg group, 43,684 copies/ml in the 25mg group and 105,582 copies/ml in the 50mg group. The variation in CD4 cell count was less, ranging from 369 cells/mm³ to 486/mm³.

The maximum viral load reduction occurred in the 50mg group: by day 15, viral load had fallen by an average of –1.5 log₁₀ copies/ml in this group, compared to just under –1 log₁₀ copies/ml in the 10mg group and no change from baseline in the placebo group. However, the proportions who achieved a viral load reduction of greater than -1.5 log₁₀ copies/ml did not differ substantially between the 25mg and 50mg dosing groups (46% and 45% respectively).

Viral load was still more than –0.5 log below baseline in the 50mg group ten days after treatment ceased; rebound to baseline occurred sooner at lower doses. This long-term effect is likely to be explained by the prolonged saturation of CCR5 receptors on T-cells.

Some researchers are concerned that over time, use of CCR5 antagonists will favour the emergence of more lethal viruses that use the CXCR4 receptor. Viruses that favour the CXCR4 receptor are known to infect and kill CD4 cells much more rapidly than CCR5 viruses. In this study, one patient with viral load reduction of greater than –1.5 log₁₀ copies had evidence of a transient switch to CXCR4 virus after treatment.

SCH-D was well tolerated; only one adverse event (a fever) was reported during the study that might have a link to the drug.

Bristol Myers Squibb’s HIV attachment inhibitor

Bristol Myers Squibb is pursuing the development of small molecules that will prevent attachment of HIV’s gp120 molecule to the CD4 receptor. The attachment inhibitors are designed to bind to the gp120 molecule so that HIV cannot latch on to the CD4 receptor, thus stopping viral infection of CD4 lymphocytes.

Unlike Schering-Plough’s chemokine receptor antagonist, BMS’s attachment inhibitor will act against viral variants bearing either the CCR5 or the CXCR4 receptor

Dosing data from a 14-day study in healthy volunteers showed that levels five to ten times above the minimum target concentration were achieved when BMS-488043 was given at doses of 1200mg and 1800mg twice daily with a high-fat meal.

At CROI, Bristol Myers Squibb presented results of study AI430-003, in which HIV-positive individuals were randomised in a 4:1 fashion to receive BMS-488043 or placebo at one of two doses (800mg bid or 1800mg bid) for seven days as monotherapy. The rationale for choosing a dose (800mg) that has already been clearly demonstrated to lead to suboptimal trough levels was not explained.

Participants had a mean viral load of 4.61 log₁₀ copies/ml (around 40,000 copies) and a mean CD4 cell count of 399 cells/mm³. Fourteen of 30 were antiretroviral-experienced.

The peak reduction in viral load was observed at day 9, averaging -1.23 log10 copies/ml in the 1800mg group and –1.01 log₁₀ copies/ml in the 800mg group. Viral load rebounded slowly in the majority of patients. 42% of the 1800mg recipients experienced viral load reductions of at least -1.5 log10 copies/ml, compared to 25% of the 800mg recipients.

Adverse events and laboratory abnormalities were mild, and Bristol Myers Squibb will now move forward with refining the dose in a larger phase I/II study.