Randomized, Placebo-Controlled, Single-Ascending-Dose Study of BMS-791325, a Hepatitis C Virus (HCV) NS5B Polymerase Inhibitor, in HCV Genotype 1 Infection

Abstract

BMS-791325 is a nonnucleoside inhibitor of hepatitis C virus (HCV) NS5B polymerase with low-nanomolar potency against genotypes 1a (50% effective concentration [EC₅₀], 3 nM) and 1b (EC₅₀, 7 nM) in vitro. BMS-791325 safety, pharmacokinetics, and antiviral activity were evaluated in a double-blind, placebo-controlled, single-ascending-dose study in 24 patients (interferon naive and experienced) with chronic HCV genotype 1 infection, randomized (5:1) to receive a single dose of BMS-791325 (100, 300, 600, or 900 mg) or placebo. The prevalence and phenotype of HCV variants at baseline and specific posttreatment time points were assessed. Antiviral activity was observed in all cohorts, with a mean HCV RNA decline of ≈2.5 log₁₀ copies/ml observed 24 h after a single 300-mg dose. Mean plasma half-life among cohorts was 7 to 9 h; individual 24-hour levels exceeded the protein-adjusted EC₉₀ for genotype 1 at all doses. BMS-791325 was generally well tolerated, with no serious adverse events or discontinuations. Enrichment for resistance variants was not observed at 100 to 600 mg. At 900 mg, variants (P495L/S) associated with BMS-791325 resistance in vitro were transiently observed in one patient, concurrent with an observed HCV RNA decline of 3.4 log₁₀ IU/ml, but were replaced with wild type by 48 h. Single doses of BMS-791325 were well tolerated; demonstrated rapid, substantial, and exposure-related antiviral activity; displayed dose-related increases in exposure; and showed viral kinetic and pharmacokinetic profiles supportive of once- or twice-daily dosing. These results support its further development in combination with other direct-acting antivirals for HCV genotype 1 infection. (This trial has been registered at ClinicalTrials.gov under registration no. NCT00664625.)

INTRODUCTION

It is estimated that between 130 and 170 million people, 2 to 3% of the world population, are chronically infected with hepatitis C virus (HCV) (1). The World Health Organization estimates that HCV infection results in approximately 350,000 deaths each year and is associated with long-term complications such as cirrhosis, liver failure, and hepatocellular carcinoma (2). Up until 2011, the standard of care for treatment of chronic HCV was based on the subcutaneous administration of pegylated alfa interferons (alfa-2a or alfa-2b) for periods of up to 48 weeks in combination with oral ribavirin (IFN-α/RBV) (3, 4). However, the efficacy of IFN-α/RBV varies according to both viral and host genotype, specifically with respect to polymorphisms in IL28B (5–7), and is associated with a significant burden of treatment-limiting adverse events (AEs), which result in a high rate of on-treatment dose reductions and discontinuations (8). The poor tolerability of IFN-α/RBV also results in its contraindication in approximately 17% of HCV-infected patients (9). For the more treatment-refractory genotype 1 infections that predominate in Europe, Japan, and the United States, a newer standard of care has emerged that combines one of the recently approved HCV NS3 protease inhibitors telaprevir and boceprevir with IFN-α/RBV for part of the overall treatment (10). Although significantly improving posttreatment sustained virologic response rates, the use of either protease inhibitor also significantly adds to the AE burden by elevating rates of some IFN-α/RBV-associated events, such as anemia, and by introducing new signature events, such as on-treatment rash (11).

The goal of further improving sustained virologic response while reducing treatment-related toxicities has directed a major and ongoing research effort into the development of new direct-acting HCV antivirals with better tolerability profiles that can be used in all-oral, IFN-α/RBV-free combinations to provide shorter and less toxic therapies with better posttreatment outcomes. All-oral regimens would also be expected to extend therapeutic benefit to those unable to receive or tolerate IFN-α/RBV-based treatment, as well as provide an alternative option for those who have previously failed to achieve a sustained response.

The HCV NS5B polymerase is an essential enzyme for HCV replication and has been validated as a target for anti-HCV therapy in clinical trials using both active-site and allosteric inhibitors (12–14). BMS-791325 (Fig. 1) is an orally bioavailable allosteric nonnucleoside inhibitor of NS5B that binds to the polymerase thumb 1 domain and inhibits HCV replication in genotype 1a and 1b subgenomic replicons at 50% effective concentrations (EC₅₀s) of 3 and 7 nM, respectively, as described in the accompanying paper (15, 16). BMS-791325 is metabolized by cytochrome P450 3A4 to an active N-desmethyl metabolite (BMS-794712) that is equipotent with the parent compound.

FIG 1.

Structure of BMS-791325 and its N-desmethyl metabolite BMS-794712.

Here, we report the results of a proof-of-concept, single-ascending-dose study to evaluate the safety, tolerability, pharmacokinetics, and antiviral activity of BMS-791325 in patients with chronic HCV genotype 1 infection, together with population sequencing, clonal analysis, and phenotyping of baseline and emergent (postdose) NS5B sequence variants.

MATERIALS AND METHODS

Study design.

This was a double-blind, placebo-controlled, randomized, single-ascending-dose study (study AI443-002; clinicaltrials.gov registration no. NCT00664625) for genotype 1 HCV infection, comprising four sequential dosing cohorts of 100 mg, 300 mg, 600 mg, and 900 mg. Six patients in each cohort were randomized 5:1 to receive a single dose of orally administered BMS-791325 or placebo. Dose escalation between cohorts occurred only after safety data from the previous cohort through postdose day 4 were reviewed by the sponsor and investigators. Within each open dose cohort, randomization to receive BMS-791325 or placebo was made via an interactive voice response telephone system according to a computer-generated scheme prepared by the study sponsor. Initially, the 100-mg, 300-mg, and 900-mg cohorts were recruited sequentially. Subsequently, the 600-mg dosing cohort was enrolled using a higher-strength formulation (described below) to reduce capsule burden.

Study conduct and ethics.

The study was conducted at four sites in the United States. Patients underwent screening within 28 days prior to administration of study medication on day 1. Patients who received 100, 300, or 900 mg BMS-791325 remained at their clinical centers for the first 2 days postdose and were furloughed on day 3; two further outpatient visits occurred on days 4 and 7. Due to gastrointestinal intolerance at the 900-mg dose level (presumed due to the 18 capsules of 50-mg strength required to deliver the dose), patients who subsequently received 600 mg BMS-791325 were administered 4 capsules of 150-mg strength and furloughed at day 7, with an outpatient visit at day 14.

The study was conducted according to local regulatory requirements and in accordance with the principles of the Declaration of Helsinki and with Good Clinical Practice, as defined by the International Conference on Harmonization and the ethical principles underlying the European Union Directive 2001/20/EC and the United States Code of Federal Regulations, Title 21, Part 50 (21CFR50). The protocol, amendments, and patient informed consent were approved by the Institutional Review Board of each participating site. Written informed consent was obtained from all patients prior to study procedures.

Study objectives.

The primary objective was to assess the safety and tolerability of a single oral dose of 100 to 900 mg BMS-791325 in patients with chronic HCV genotype 1 infection. Secondary objectives included assessment of the pharmacokinetics (PK) of BMS-791325 and its metabolite BMS-794712, assessment of antiviral activity, and characterization of resistance-associated NS5B amino acid substitutions following a single dose.

Patients.

Eligible patients were men and women aged 18 to 60 years with chronic HCV genotype 1 infection, a screening plasma HCV RNA level of at least 100,000 IU/ml, and a body mass index between 18 and 35 kg/m². Patients were noncirrhotic (screening FibroTest score of ≤0.59 with an aminotransferase/platelet ratio index of ≤2 or with absence of cirrhosis documented by biopsy within the previous 12 months) and could be either treatment naive or have previously received and discontinued alfa interferon, with or without RBV, at least 6 months before enrollment. Patients with previous exposure to HCV NS5A or NS5B inhibitors were excluded, as were patients coinfected with human immunodeficiency virus or hepatitis B virus or infected with other HCV genotypes. Pregnant or nursing women were also excluded, as were women of childbearing age unwilling to use contraception from 1 month predose through 8 weeks postdose. Men were excluded if unwilling to practice barrier contraception with female partners for at least 12 weeks postdose.

Dosing.

Study drug was administered with water in the fasted state as hard gelatin capsules containing 50 mg (100-mg, 300-mg, and 900-mg dose cohorts) or 150 mg (600-mg cohort) BMS-791325 or placebo. Patients continued to fast for 4 h postdose.

HCV RNA assessments.

Plasma HCV RNA levels were determined using the Roche TaqMan HCV quantitative assay (lower limit of quantitation, 25 IU/ml) at screening, predose, and 2, 4, 6, 8, 12, 16, 24, 36, 48, 72, and 144 h postdose.

Safety assessments.

Blood and urine samples for clinical laboratory evaluations and vital sign measurements were collected at specified time points throughout the study. Information about AEs was collected by spontaneous report or elicited by open-ended questioning, examination, or evaluation. Event intensity was categorized by the investigator as mild, moderate, severe, or very severe based on level of incapacitation or interference in everyday activities. A 12-lead electrocardiogram was recorded at the screening visit, on day −1, predose on day 1 and at 0.5, 1, 1.5, 2, 3, 4, 6, 8, 12, 24, and 48 h postdose, and on days 7 and 14 postdose. Safety assessments were based on AE reports and the results of vital sign measurements, electrocardiograms, physical examinations, and clinical laboratory tests.

Pharmacokinetics.

Serial blood samples for PK analyses were obtained predose and at 0.5, 1, 1.5, 2, 2.5, 3, 4, 6, 8, 10, 12, 16, 24, 48, and 72 h postdose. Concentrations of BMS-791325 and the BMS-794712 metabolite were determined from 100-μl aliquots of EDTA-treated plasma using liquid-liquid extraction and a validated liquid chromatography assay with tandem mass spectrometry at Tandem Labs (West Trenton, NJ, USA). Analyte concentrations were calculated by linear regression over concentration ranges of 0.1 to 50.0 ng/ml using BMS-791325–D₆ or BMS-794712-¹³CD₃ internal standards. The mass spectrometer was operated in multiple reaction monitoring mode under optimized conditions for the detection of positive analyte/internal standard ions formed by electrospray ionization. Assay precision (percent coefficient of variability) and accuracy (percent bias) were 4.5 to 6.9% and 0.0 to 6.5%, respectively, for BMS-791325 and 3.8 to 6.8% and 0.0 to 6.0%, respectively, for the BMS-794712 metabolite.

Parameters derived from plasma concentration-versus-time data included maximum observed concentration (C_max), time to maximum concentration, terminal half-life (t_1/2), concentration at 24 h postdose (C₂₄), and area under the concentration-time curve extrapolated to infinite time (AUC_inf). Dose proportionality was assessed for key BMS-791325 exposure parameters (C_max, AUC_inf, and C₂₄) by generating point estimates and 90% confidence intervals for the slope (β) of the log-transformed (linearized) dose-exposure relationship using a power model.

Pharmacodynamics.

Antiviral activity was assessed by the magnitude (mean decline at 24 h and mean maximum decline) of change in log₁₀ plasma HCV RNA levels from the day 1 predose baseline. Exposure-response relationships for AUC_inf, C_max, and C₂₄ were assessed graphically by plotting maximum change from baseline in log₁₀ HCV RNA versus estimated PK parameters and by estimation of Pearson's correlation coefficients.

Resistance analyses.

Serum samples were analyzed predose and at 12, 24, 48, and 144 h postdose for NS5B sequence variants potentially associated with BMS-791325 resistance, particularly codon 495 variants characteristically associated with loss of susceptibility to thumb 1 allosteric inhibitors, including BMS-791325 (16, 17). Viral RNA was extracted using the QIAamp MiniElute virus vacuum kit, and first-strand cDNA synthesis was performed using random hexamer primers and the Superscript III first-strand synthesis system (Invitrogen). The full NS5B coding region was sequenced from amplimers generated by sequential or nested PCRs using genotype-specific degenerate primers. To determine the sensitivity for minority populations, mixed wild-type and codon-495 proline-to-leucine (P495L) variant DNAs were sequenced at ratios of 100:0, 90:10, 80:20, and 50:50. The P495L variant was readily detected at 20% prevalence. Relevant amplimers were subcloned using a TOPO TA cloning kit (Invitrogen) and transfected into Escherichia coli for subsequent clonal sequence analysis of individual colonies.

Specific substitutions identified by sequence analysis were assessed phenotypically in genotype 1a or 1b replicon clones by site-directed mutagenesis using a Stratagene QuikChange mutagenesis kit according to the manufacturer's directions. Mutant clones were verified by sequence analysis and evaluated for replication efficiency and compound sensitivity using transient replicon assays as described elsewhere (16).

To evaluate the phenotype of the entire patient-derived NS5B gene, a shuttle replicon system was developed by creating unique SpeI/ClaI (genotype 1a) or SpeI/SnaBI (genotype 1b) restriction sites flanking the NS5B gene and reamplifying patient-derived NS5B amplicons using oligomer primers bearing the relevant sites. Clones generated by standard techniques were verified by sequence analysis, and multiple clones per time point were evaluated for replication efficiency and compound sensitivity using transient replication assays.

Statistical considerations.

Five patients receiving BMS-791325 in a dose cohort provided 80% probability of observing at least 1 occurrence of any AE occurring at a 28% incidence in the overall population.

RESULTS

Patients.

A total of 198 patients were screened, of whom 24 were randomized and treated between June 2008 and June 2009. The majority of screened patients who were not randomized were screening failures meeting one or more study exclusion criteria (153/174; 88%). A small number of patients withdrew consent or were lost to follow-up after screening (6/174 [3%] each). Other reasons for nonrandomization included administrative exclusions and failure to attend study day 1. Baseline demographics and disease characteristics are shown in Table 1, and a patient disposition diagram is shown in Fig. 2. Baseline characteristics were generally comparable across all treatment groups. All patients completed the study.

TABLE 1.

Baseline demographics and disease characteristics

Characteristic	BMS-791325				Placebo (n = 4)
	100 mg (n = 5)	300 mg (n = 5)	600 mg (n = 5)	900 mg (n = 5)
Age, median (range), yr	44 (33–59)	43 (24–50)	47 (40–55)	46 (35–50)	49 (41–60)
Male, n (%)	5 (100)	5 (100)	3 (60)	3 (60)	2 (50)
Race, n (%)
White	4 (80)	5 (100)	5 (100)	5 (100)	4 (100)
Black/African American	1 (20)	0	0	0	0
Wt, mean (range), kg	79 (60–91)	82 (61–117)	73 (61–94)	82 (63–91)	75 (67–78)
HCV RNA, mean (SD), log10 IU/ml	5.90 (0.48)	5.99 (0.62)	5.86 (0.46)	5.75 (0.53)	5.97 (0.58)
Genotype 1a/1b, n/n	5/0	4/1	4/1	3/2	4/0
Prior IFN treatment, n (%)	0 (0)	2 (40)	1 (20)	1 (20)	3 (75)

FIG 2.

Patient disposition.

Safety and tolerability.

BMS-791325 was generally well tolerated. There were no deaths, serious AEs, discontinuations due to AEs, laboratory or physical examination findings reported as AEs, or clinically relevant changes in vital signs or electrocardiogram parameters. All AEs were of mild intensity except for two moderate gastrointestinal events in the 900-mg cohort. All AEs resolved prior to study completion. There were 14 AEs in total, occurring in 10 patients, of which eight events (in five patients) were considered to be at least possibly related to treatment. Treatment-related events in the placebo arm were mild nausea and headache (n = 1) and mild decreased appetite (n = 1). All treatment-related events on BMS-791325 occurred in the 900-mg cohort: mild vomiting, diarrhea, and headache (n = 1); moderate nausea (n = 1); and moderate vomiting (n = 1). The high proportion of gastrointestinal events in the 900-mg cohort may have been due, at least in part, to high capsule burden (18 capsules). Events not considered to be related to treatment were ecchymosis (n = 1; 300 mg), urinary tract infection (n = 1; 900 mg), pruritus (n = 1; 900 mg), application-site erythema and headache (n = 1; 900 mg), and pain in extremity (n = 1; placebo). Overall, only nausea (n = 2), vomiting (n = 2), and headache (n = 2) were reported for more than a single patient.

Pharmacokinetics.

Two patients in the BMS-791325 900-mg cohort who experienced vomiting at approximately 2 h postdose were excluded from the PK analyses. Figure 3 shows mean concentration-time data for BMS-791325 and the BMS-794712 metabolite, and Table 2 summarizes key PK parameters. Peak BMS-791325 concentrations were attained between 2 and 4 h, and compound t_1/2 was 6.8 to 9.4 h. Exposure to BMS-791325 (C_max and AUC_inf) was dose dependent and appeared slightly more than dose proportional, with point estimates and 90% confidence intervals for β of 1.1 (0.99 to 1.22) for C_max and 1.18 (0.99 to 1.36) for AUC_inf. All patients had 24-hour plasma concentrations exceeding the protein-adjusted EC₉₀ for HCV genotype 1. The PK profile of the BMS-794712 metabolite was generally similar to that of the parent compound, but the plasma exposure was approximately 22% of the parent value. No assessment of dose proportionality was performed for the metabolite.

FIG 3.

Mean plasma concentration-time profile of BMS-791325 (A) and BMS-794712 metabolite (B) after single oral doses in HCV-infected patients.

TABLE 3.

Phenotypic analysis of baseline and posttreatment variants in in vitro replicon system

Variant	HCV genotype	Time observed	EC50 (ng/ml) (fold resistance)c		Replication efficacy vs wild type (%)
			Point mut.d	Full genea	Point mut.d	Full genea
L392F	1a (1 patient)	All time pointsb	2.2 (0.7)	1.0 (0.3)	120	14
L497 M	1a (1 patient)	12 h postdose only	1.7 (0.5)	2.1 (0.6)	13	24
A499V	1a (1 patient)	All time pointsb	3.5 (1.1)		48
V499A	1b (1 patient)	12 h postdose only	7.2 (1.6)		50
V494A		24 h postdose only	5.6 (1.7)		9
P495L	1a (1 patient)	24 h postdose only	105.3 (31.9)		29
P495S		24 h postdose only	210.5 (63.8)		16

^aFull patient-derived NS5B gene containing specified variant expressed in a chimeric genotype 1a replicon.

^bAlso present at baseline.

^cEC₅₀s for wild-type genotypes 1a and 1b were 3.3 and 4.6 ng/ml, respectively.

^dPoint mut., as a point mutation introduced to replicon NS5B.

TABLE 2.

Summary pharmacokinetic parameters for BMS-791325 and BMS-794712 metabolite

Parametera	BMS-791325				BMS-794712
	100 mg (n = 5)	300 mg (n = 5)	600 mg (n = 5)	900 mg (n = 3)	100 mg (n = 5)	300 mg (n = 5)	600 mg (n = 5)	900 mg (n = 3)
Cmax, ng/ml, geometric mean (% CV)	1,432 (23)	4,782 (22)	10,199 (20)	16,545 (31)	243.1 (17)	694.1 (47)	1,832 (40)	2,546 (41)
AUCinf, ng · h/ml, geometric mean (% CV)	17,167 (42)	69,360 (23)	112,033 (27)	290,247 (27)	3,651 (40)	13,199 (38)	25,551 (45)	65,637 (31)
tmax, h, median (range)	2.00 (2.0–6.0)	4.00 (2.0–4.0)	4.00 (1.5–4.0)	3.00 (3.0–4.0)	4.0 (2.5–10)	4.0 (4.0–8.0)	4.0 (2.5–4.0)	8.0 (6.0–10)
t1/2, h, mean (SD)	8.62 (2.25)	9.37 (1.94)	6.81 (0.91)	7.78 (0.29)	10.3 (2.8)	10.2 (2.1)	7.3 (1.0)	9.0 (0.3)
C24, ng/ml, geometric mean (% CV)	161.7 (63)	863.4 (44)	1,149 (49)	5,141 (20)	46.3 (57)	203.7 (57)	318.3 (56)	1,481 (27)
CLT/F, ml/min, geometric mean (% CV)	97.1 (59)	72.1 (26)	89.3 (23)	51.7 (31)
Metabolic ratio, geometric mean (% CV)					0.217 (22)	0.194 (29)	0.233 (23)	0.231 (7)

^aAbbreviations: AUC_inf, area under the plasma concentration-time curve extrapolated to infinite time; C₂₄, 24-hour plasma concentration; CLT/F, apparent total body clearance; C_max, maximum plasma concentration; t_max, time to maximum plasma concentration; t_1/2, terminal plasma half-life; CV, coefficient of variation.

Antiviral response.

Mean and individual changes from baseline in log₁₀ HCV RNA following single doses of BMS-791325 or placebo are shown in Fig. 4A through Fig. 4F. Antiviral activity was dose related and appeared to plateau between 300 mg and 600 mg, with the majority of the dose-dependent increase seen between 100 mg and 300 mg. Mean maximum log₁₀ HCV RNA declines (standard deviations and ranges) were −1.3 (0.54; range, −0.7 to −2.2) at 100 mg, −2.52 (0.55; −1.8 to −2.9) at 300 mg, −2.78 (0.42; −2.2 to −3.4) at 600 mg, and −2.55 (0.79; −1.3 to −3.4) at 900 mg. Mean and median times to maximum decline increased slightly with dose from 100 mg (mean, 16.8 h; median, 12 h) through 600 mg (mean and median, 24 h). Mean time to maximum decline at 900 mg was 34.4 h (median, 36 h) but with a large standard deviation (14.3 h). One patient in the 900-mg cohort showed consistently poorer virologic response than did the others (Fig. 4F), associated with low BMS-791325 levels at all sampling time points (data not shown). This patient was one of the two excluded from the PK analyses for early vomiting (2.3 h); however, the cause of the poor drug exposure is uncertain, as pre-emesis BMS-791325 levels at 1.0 and 1.5 h postdose were also low, and the other excluded patient had a strong virologic response (maximum RNA decline, −2.86 log₁₀) despite vomiting at 1.8 h.

FIG 4.

Mean (A) and individual changes from baseline in HCV RNA following a single dose of placebo (B) or BMS-791325 at 100 mg (C), 300 mg (D), 600 mg (E), and 900 mg (F).

Pharmacodynamic relationships.

Individual maximum declines in HCV RNA across all dosing cohorts were associated with AUC_inf, C_max, and C₁₂ (Fig. 5), with Pearson's correlation coefficients of 0.76, 0.76, and 0.75, respectively. A similar association (Pearson's correlation coefficient, 0.73) was seen for C₂₄ (data not shown).

FIG 5.

Exposure-response correlations between maximum change from baseline in log₁₀ HCV RNA and BMS-791325 AUC_inf (A), C_max (B), and C₁₂ (C).

Resistance analyses.

NS5B variations from the genotype 1a consensus were present at baseline and at all postdosing time points in two patients receiving BMS-791325: one L392F in the 100-mg cohort, and one A499V in the 600-mg cohort. Neither was associated with any change in drug susceptibility in the replicon model (Table 3), and although A499V reduced replicon replication efficiency by 50%, the patient had a virologic response similar to those of other patients in the same cohort. Interestingly, while the L392F variant slightly elevated the replication efficiency of the 1a replicon when introduced as a point mutation, a chimeric replicon containing the whole patient-derived NS5B gene with L392F showed a marked reduction in efficiency (14% of wild-type value).

Transient posttreatment emergence of NS5B variants not present at baseline was observed in three patients. In one (genotype 1a), an L497M substitution was detected at the 12-hour time point only, which conferred no resistance to BMS-791325 in replicon cultures either as a site-directed point mutant or in the context of the full-length patient-derived NS5B gene (Table 3). Virus from a second patient (genotype 1b) showed a V499A change at 12 h, which only conferred minimal (2-fold) resistance to BMS-791325 when introduced as a replicon point mutation. Viral load continued to decline in this patient between 12 and 24 h, when only the wild type was detected, and then returned to baseline over the next few days.

At the 24-hour time point only, population sequencing of virus from a third patient (genotype 1a) showed distinct but transient differences in amino acids 494 and 495 in two different amplifications from different primers. In one amplification, P495L was detected at approximately 100% prevalence, while in a second, P495L/S occurred in 55% of the population trace and V494A occurred in 45%. Clonal analysis of the second amplimer (n = 130 clones) was consistent with the population results (38% V494A; 62% P495L/S) and showed a very low frequency of linkage between the two variants (<2%). Clonal analysis of the baseline amplimer (n = 93) showed no change at either position, suggesting a low pretreatment prevalence.

In replicon cultures, the P495 variants observed in this patient conferred moderate levels of BMS-791325 resistance (Table 3), but V494A conferred only 2-fold resistance with a substantial drop in replication efficiency (7 to 11% of wild-type value). Replication of the rare, linked V494A-P495S double variant was too low to assess resistance accurately; similarly, a chimeric replicon containing the entire NS5B gene for this patient at 24 h had a replication efficiency that was too low to assess resistance. The patient's baseline NS5B gene yielded a replicon with normal BMS-791325 susceptibility but low replication (35% of control). Of interest, this patient also displayed the highest maximum HCV RNA decline in the 900-mg cohort (−3.39 log₁₀), which coincided with the transient emergence of resistance. Viremia in this patient returned toward baseline levels at the same rate as it did for other patients in the cohort.

DISCUSSION

This single-ascending-dose study represents a proof of concept for the use of BMS-791325 in the treatment of chronic HCV infection. Single doses of this agent between 100 mg and 900 mg resulted in strong and rapid dose- and concentration-dependent reductions in viral RNA between 1 log₁₀ and over 3 log₁₀ IU/ml among patients infected with HCV genotypes 1a and 1b. Median time to maximum antiviral response across all patients receiving drug was 24 h postdose (range, 12 to 48 h), which was also the modal time (12/20 patients). At the lowest dose of 100 mg, median time to maximum response was shorter at 12 h postdose (3/5 patients; range, 12 to 24 h). Maximum antiviral response was positively correlated with plasma drug exposure.

The PK of BMS-791325 indicate that exposure is slightly greater than dose proportional and are consistent with the expectation of once- or twice-daily dosing. Although t_1/2 appears to be below 10 h, achievable exposures at 24 h greatly exceeded effective concentrations established in vitro. Plasma concentrations at all doses exceeded the protein-adjusted EC₉₀ value in replicon cultures (52 ng/ml) within 1 h of dosing and remained above the EC₉₀ at 24 h in all patients, which suggests that robust antiviral responses can be expected for repeated administration even at doses comparable to the lowest tested dose of 100 mg in the current study. Additionally, preliminary exposure-response assessments suggest that single doses above 300 mg provide little additional antiviral benefit. The time to maximum concentration and t_1/2 of BMS-794712—the active, equipotent metabolite of BMS-791325—were consistent with those of the parent compound. The AUC of the active metabolite in plasma was approximately 20 to 30% of that of BMS-791325 and may therefore contribute significantly to the antiviral activity observed.

Single doses of BMS-791325 were generally well tolerated, with vomiting the only dose-associated event observed, and only at the highest dose of 900 mg. Since the 900-mg cohort was dosed with 18 capsules of 50 mg BMS-791325, it is likely that at least part of this gastrointestinal intolerance was excipient related. It is notable that no gastrointestinal or other treatment-related events were reported in the 600-mg cohort, in which patients received only four capsules of BMS-791325, 150 mg each. There were no indications of drug-associated changes in laboratory parameters, vital signs, or cardiac function.

The highest observed maximum reduction in HCV RNA was 3.4 log₁₀ IU/ml and was achieved by two patients, one in the 600-mg cohort (predose baseline, 6.6 log₁₀ IU/ml) and one in the 900-mg cohort (predose baseline, 6.5 log₁₀ IU/ml). Both patients achieved this reduction at 24 h postdose, but the patient in the 900-mg cohort exhibited a 24-hour plasma BMS-791325 level five times greater than that in the 600-mg patient (5,340 versus 1,090 ng/ml). It is therefore of note that the only patient who demonstrated transient enrichment for NS5B proline 495-associated resistance variants in this study was the 900-mg patient with the 3.4-log₁₀ reduction. Since selective pressure for the emergence of viral drug resistance is a function of antiviral potency, drug concentration, and residual replication (18), it is logical that resistance in this single-dose study would have been observed at the highest plasma concentration of BMS-791325 associated with the highest antiviral response, at the time that maximal response was achieved. Poor fitness of these early variants under decaying drug-selective pressure was evidenced by a return to NS5B wild type by 48 h postdose. Further in vivo data will be needed to establish whether ongoing drug selection results in compensatory mutagenesis and the restoration of viral fitness.

Substitutions at proline 495 are signature drug resistance mutations for the thumb 1 NS5B inhibitors (17, 19–21) and have been previously observed under BMS-791325 selection both in vitro in replicon cultures (15, 16) and in vivo in virologic rebound samples from patients receiving BMS-791325 in combination with IFN-α/RBV (22). Additional substitutions in NS5B, such as A421V and L392I, have been identified under sustained clinical or in vitro selective pressure of BMS-791325 (15, 16, 22), but the significance of the transient V494A variant observed here in a single patient under single-dose selection is difficult to assess. However, the emergence of resistance after a single dose of a potent allosteric viral polymerase inhibitor is well established in the human immunodeficiency virus model (23) and emphasizes the importance of treating HCV infection with combinations of active agents.

These data suggest that BMS-791325 has potent anti-HCV activity and holds significant promise in the treatment of HCV infection, and its further evaluation in combination with other antiviral agents is warranted. Phase 2 clinical studies of BMS-791325 at doses of 75 mg and 150 mg twice daily in combination with IFN-α/RBV (study AI443-012; clinicaltrials.gov registration no. NCT01193361) (24) and in combination with the NS5A replication complex inhibitor daclatasvir and the NS3 protease inhibitor asunaprevir (study AI443-014; clinicaltrials.gov registration no. NCT01455090) (25) have shown promising preliminary results for the treatment of HCV genotype 1 infection.