Pipeline antiretrovirals: IDX899, CHX157 and bevirimat
Simon Collins, HIV i-Base
IDX899: an NNRTI in development with Idenix
An oral presentation by Rob Murphy  summarised two posters at the workshop that presented information on a new NNRTI being developed by Idenix. [2, 3]
Antiviral activity was shown in results from a seven-day Phase I/IIa dose finding study in Argentina, 40 treatment-naive patients were randomised 8:2 to once-daily monotherapy with 800mg, 400mg, 200mg or placebo. All patients switched to 28 days monotherapy or started HAART at the end of the study period.
Results were available for all but two patients in each of the 200mg and placebo arms. Viral activity was similar in each of the active drug groups which saw steady, linear viral load reductions reaching -1.8 log at day eight (using COBRAS Amplicor 1,5) from mean baseline of approximately 4.3-4.6 log copies/mL.
No clear relationship was seen between drug exposure and response and although the current formulation only has 50% bioavailability, all doses achieved drug trough concentrations that exceeded the protein adjusted EC50 by greater than 10-40 fold (with individual ranges from 5-140 fold).
Mean CD4 count increased by around +65 cells/mm3 in all groups compared to a reduction of -85 cells/mm3 in the placebo arm.
Side effects occurring in > 2 patients were mild, infrequent and distributed equally between the placebo and active drug groups, with no clear dose-related toxicity, and no grade3/4 laboratory changes. There were no cases of rash, which commonly do not show after only 7 days, but also no cases of CNS disorientation which would be expected, if induced. Toxicity has not been seen in preclinical animal studies.
A second poster presented by Jocelyn Jakubik expanded on the likely resistance and cross-resistance profile of IDX899 in a set of in vitro studies. 
IDX899 was shown to remain sensitive (<4 fold change from wild-type) to 20/23 single, double and triple site directed mutations, with greater resistance seen to Y181C (14 fold), E138K/Y181I (11 fold) and E138K/Y181I/M230L (900 fold). It also retained sensitivity to viral pools selected through exposure to efavirenz and etravirine (TMC-125), and retained greater sensitivity to each pool compared with efavirenz, etravirine of TMC-278 (rilpivirtine)
High level resistance to IDX899 was selected in vitro after >29 passages with two independent pathways initiated by mutations at either E138K or V90I/Y181C.
While showing potential benefits of IDX899 against NNRTI-resistant virus, the results also tentatively suggested that naive patients failing IDX899 might retain sensitivity to efavirenz.
This poster expanded on the first presentation of these studies at the Retrovirus Conference earlier this year. 
Idenix are still in negotiation with potential companies to take forward the clinical development programme of IDX899 in 2009, based on an improved formulation and possibly also a 100mg dose.
These promising early results should help start the Phase I/II studies. Although the in vitro data support potentially different resistance patterns to both existing NNRTIs and other pipeline drugs, this also needs to be studied in the clinical programme.
- Murphy R et al. In vitro cross-resistance profile, antiviral activity, safety and pharmacokinetics in HIV-1-infected subjects of IDX899, a novel HIV-1 NNRTI with high barrier to resistance. XVII IHDRW 2008, Sitges. Abstract 5.
- Murphy et al. Antiviral activity, safety and pharmacokinetics of IDX899, a novel HIV-1 NNRTI with high barrier to resistance, in treatment-naive HIV-1-infected subjects. XVII IHDRW 2008, Sitges. Poster 25.
- Jakubik J et al. In vitro cross-resistance profile for a next-generation NNRTI: IDX899. XVII IHDRW 2008, Sitges. Poster 26.
- Richmann D et al. Genotypic Resistance and Phenotypic Cross-resistance Profile in vitro for a Novel NNRTI: IDX899. 15th CROI, Boston, 2008. http://www.retroconference.org/2008/Abstracts/32674.htm http://www.retroconference.org/2008/PDFs/729.pdf
CHX157: a prodrug similar to tenofovir-DF
Randall Lanier, formerly with GSK and now with Chimerix, presented results on hexadecyloxpropyl tenofovir (CMX157), a new compound similar to Gilead’s tenofovir disoproxil fumerate. 
Differences claimed for this new compound include activity against K65R and that this prodrug is less efficiently cleaved to free tenofovir in plasma, which should increase tenofovir diphosphate levels in target cells, lower the apparent IC50 and reduce the rate of secretion into the kidney (and related toxicity). The primary elimination pathway is thought to be hepatic,
IC50s for CMX157 were determined against a panel of 30 NRTI single and multiple isolates including K65R, M184V, TAMS, K70E, Q151M and 69-SXX insertion and are detailed in Table 1.
Similar results were also reported against a separate panel of 14 NRTI-resistant and wild-type clinical isolates in PBMCs.
Table 1: Examples of IC50s for CMX157 and tenofovir-DF against NRTI-resistant isolates
The poster reported that tenofovir diphosphate levels were 33-fold higher in PHA/IL-2 stimulated human PBMCs after 24 hours compared to tenofovir-DF and that no toxicity has been observed in rats administered up to 100mg/kg/day for 7 seven days.
- Lanier ER et al. Hexadecyloxpropyl tenofovir (CMX157) has enhanced potency in vitro against NRTI resistant HIV relative to tenofovir and a favorable preclinical profile. XVII IHDRW 2008, Sitges. Poster 4.
Bevirimat (PA-457): baseline gag polymorphisms determine treatment response
Scott McCallister from Panacos presented an analysis of the impact of baseline Gag polymorphisms together with individual drug exposure levels on virological response to the maturation inhibitor bevirimat. 
Results of early studies were complicated by formulation difficulties and disappointing virological responses. In fact, the latest virological information from ongoing Phase IIb results has only trickled out in a series of company press statements containing ‘forward-looking statements’. [2, 3]
The analysis at this years workshop was from 44 treatment-experienced patients on currently failing treatment who added bevirimat in escalating dose groups as functional monotherapy for 14 days.
|Responders >0.5 log reduction||Non-responders <0.5 log reduction|
|Mean VL response||-1.26 log||-0.05 log|
|Baseline Gag polymorphisms||5.9||7.3|
Viral responses by presence of baseline polymorphisms in Gag at positions 369, 370 or 371 were -0.16 (n=9), -0.24 (n=22) and -0.32 logs (n=11) respectively. Patients with absence of the same mutations had viral responses of -0.69, -0.79 and -0.73 respectively. Patients with no changes at 369-371 had a reduction of -1.08 logs.
Panacos are now planning to restrict future research to patients without baseline Gag polymorphisms at the three points, citing the approximate 60% naive patients from the 567 patients in the University of British Columbia database as guide to proportion of patients likely to be eligible. They will also plan to restrict studies to patients who achieve minimum trough levels >20ug/mL.
Predefining a patient group likely to benefit from bevirimat is similar to use of a tropism assay prior to prescribing CCR5 inhibitors and should not be a block to future drug development, especially if it can reliably result in a -1.2 log drop in viral load.
- McCallister S et al. HIV-1 gag polymorphisms determine treatment response to bevirimat (PA-457). XVII IHDRW 2008, Sitges. Abstract 8.
- Panacos press statement: Results of bevirimat 350 mg cohort, data support further dose escalation in Phase 2b study (10 Dec 2007)
- Panacos press statement: Greater than 1 log10 mean reduction in 300 mg bevirimat cohort. (25 October 2007).