In Vivo Influenza Virus-Inhibitory Effects of the Cyclopentane Neuraminidase Inhibitor RWJ-270201

ABSTRACT The cyclopentane influenza virus neuraminidase inhibitor RWJ-270201 was evaluated against influenza A/NWS/33 (H1N1), A/Shangdong/09/93 (H3N2), A/Victoria/3/75 (H3N2), and B/Hong Kong/05/72 virus infections in mice. Treatment was by oral gavage twice daily for 5 days beginning 4 h pre-virus exposure. The influenza virus inhibitor oseltamivir was run in parallel, and ribavirin was included in studies with the A/Shangdong and B/Hong Kong viruses. RWJ-270201 was inhibitory to all infections using doses as low as 1 mg/kg/day. Oseltamivir was generally up to 10-fold less effective than RWJ-270201. Ribavirin was also inhibitory but was less tolerated by the mice at the 75-mg/kg/day dose used. Disease-inhibitory effects included prevention of death, lessening of decline of arterial oxygen saturation, inhibition of lung consolidation, and reduction in lung virus titers. RWJ-270201 and oseltamivir, at doses of 10 and 1 mg/kg/day each, were compared with regard to their effects on daily lung parameters in influenza A/Shangdong/09/93 virus-infected mice. Maximum virus titer inhibition was seen on day 1, with RWJ-270201 exhibiting the greater inhibitory effect, a titer reduction of >104 cell culture 50% infective doses (CCID50)/g. By day 8, the lung virus titers in mice treated with RWJ-270201 had declined to 101.2 CCID50/g, whereas titers from oseltamivir-treated animals were >103CCID50/g. Mean lung consolidation was also higher in the oseltamivir-treated animals on day 8. Both neuraminidase inhibitors were well tolerated by the mice. RWJ-270201 was nontoxic at doses as high as 1,000 mg/kg/day. These data indicate potential for the oral use of RWJ-270201 in the treatment of influenza virus infections in humans.

mice were used per drug dosage, with 16 to 20 infected animals treated with PSS only. Three to five mice were used in toxicity control groups and as normal controls. These animals were weighed immediately prior to initial treatment and again 18 h after final treatment and observed for 21 days for signs of toxicity (lethargy, prostration, hunching, hyperactivity, and ruffled fur), and deaths were recorded. The SaO 2 levels in the normal controls were also determined in parallel to 1 those in the infected mice. Parameters for evaluation of antiviral activity included prevention of death, prolongation of mean day to death, and reduced decline in SaO 2 determined on days 3 through 11 of the infection. In addition, groups of three to five mice were killed at various times after infection, and their lungs were removed, weighed, assigned a consolidation score ranging from 0 (normal) to 4 (maximal plum coloration), and assayed for virus titer.

Procedure for comparison of the effects of RWJ-270201 and oseltamivir therapy on influenza A/Shangdong/09/93 (H3N2), A/Victoria/3/75 (H3N2), A/NWS/33 (H1N1), and B/Hong Kong/05/72 virus infections in mice.
Mice infected with each virus were treated p.o. with RWJ-270201 at doses of 1, 10, and, in some cases, 100 mg/kg/day; with oseltamivir at 1 and 10 mg/kg/day (10 mg/kg/day only with Victoria virus); with ribavirin at 75 mg/kg/day (with A/Shangdong and B/Hong Kong viruses); or with PSS twice daily for 5 days beginning 4 h pre-virus exposure. Ten mice in each drug-treated group and 20 PSS-treated controls were observed for death daily for 21 days, and SaO 2 levels were determined. In the A/Shangdong and B/Hong Kong virus experiments, 5 additional mice in each high-dose drug-treated group and 10 saline-treated controls were killed on day 3, and 5 additional mice at all doses (10 saline-treated controls) were killed on day 6. In the A/Victoria virus experiment, five mice in each treatment group were killed on days 3, 6, and 9 for determination of lung consolidation and virus titer. The A/NWS virus experiment compared 10-and 1-mg/kg/day doses of both RWJ-270201 and oseltamivir, with lung parameters evaluated on days 3 and 6 using three drug-treated mice and five PSS-treated animals at each time of sacrifice. The A/Shangdong virus experiment was performed twice. Oseltamivir could not be tested at higher doses because of a shortage of the drug.
Procedure for comparison of the effects of RWJ-270201 and oseltamivir therapies on daily lung parameters in influenza virus-infected mice. Groups of 34 influenza A/Shangdong virus-infected mice were treated p.o. with RWJ-270201 or oseltamivir at doses of 10 or 1 mg/kg/day. A group of 80 infected mice were treated with PSS as controls. Treatment was twice daily for 5 days beginning 4 h pre-virus exposure. Three animals in each drug treatment group and 10 mice from the placebo-treated controls were randomly selected on days 1 through 8 and sacrificed, and their lungs were assigned a consolidation score, weighed, and assayed for virus titer. To provide background data, two uninfected, untreated mice were killed on days 4 and 8 in parallel with those described above, and their lungs processed as described above.    c Animals dying prior to day 21.

Comparison
SaO 2 decline was seen [day 11] are shown in Table 1), inhibition of lung consolidation, and inhibition of lung virus titers. Oseltamivir was significantly inhibitory to this infection as seen by prevention of death at a dose of 10 mg/kg/day only. At 1 mg/kg/day, only a delay in the mean day to death was seen, with moderate lessening of SaO 2 decline and inhibition of lung consolidation. In experiment 1, a significant inhibition of lung virus titer was seen at the 1-mg/kg/day dose of oseltamivir, but this inhibition was not seen in experiment 2. Ribavirin was significantly inhibitory to the infection in both experiments. Toxicity control mice used in this study all survived; those treated with RWJ-270201 and oseltamivir gained weight at a rate comparable to that for normal controls. Those receiving ribavirin lost a mean of 0.7 g during therapy (data not shown). In a separate experiment, RWJ-270201 was evaluated for induction of toxicity in uninfected mice at a dose of 1000 mg/kg/ day using the treatment schedule of p.o. twice daily for 5 days. The animals gained 0.7 g during the period of treatment, and no adverse effects were seen (data not shown).
The results using these compounds in mice infected with influenza A/Victoria virus are shown in Table 2, with daily mean SaO 2 values shown in Figure 3. RWJ-270201 significantly prevented deaths of the mice at 100 and 10 mg/kg/day. This antiviral effect was also reflected by significant inhibition of lung consolidation and lung virus titers, with the greatest virus titer inhibition seen at the day 9 sampling time. Oseltamivir, run in parallel at a dose of 10 mg/kg/day, was less effective against this virus infection, with only significant inhibition of lung score seen. The SaO 2 data (Fig. 3) also indicate that RWJ-270201 inhibited SaO 2 decline at the 100-and 10-mg/kg/ day doses, but at 1 mg/kg/day, a drop in SaO 2 level began to occur by day 9. Oseltamivir at the 10-mg/kg/day dose was similarly inhibitory to SaO 2 decline until day 9. Both compounds were again well tolerated by the toxicity control animals (data not shown).

Comparison of RWJ-270201 and oseltamivir therapies for influenza A (H1N1) virus infections.
The results of the study with influenza A (H1N1) virus are shown in Table 3. Each compound at the 10-mg/kg/day dose prevented deaths from occurring in the infected mice, lessened SaO 2 decline (Fig. 4), and inhibited lung score development and lung weight increases. Lung virus titers were reduced by at least 0.6 log 10 unit by both doses of each compound on day 3; by day 6, the high doses of each compound still significantly inhibited the lung virus titers. A lesser lung virus-inhibitory effect was seen using the 1-mg/kg/day doses of each compound. In this experiment, oseltamivir appeared to be slightly more inhibitory to lung virus titers than RWJ-270201. At the 1-mg/kg/day dose, all of the mice receiving RWJ-270201 survived the infection, but 30% of those treated with this dose of oseltamivir died. The lung parameters also reflected this difference in efficacy. Table 4, RWJ-270201 therapy significantly prevented influenza B/Hong This antiviral effect was also seen by inhibition of SaO 2 decline (Fig. 5), lung consolidation, and lung virus titers on day 6. Oseltamivir treatment at the 10-and 1-mg/kg/day dosages did not prevent death of the mice but did significantly prolong the mean day to death and lessen lung consolidation. Both doses also reduced SaO 2 decline, but to a lesser extent than RWJ-270201 at the same doses. The 10-mg/kg/day dose of oseltamivir also significantly inhibited day 6 lung virus titers. Ribavirin was highly active against this virus infection by all parameters studied. Toxicity control animals treated with all compounds survived and gained weight (data not shown). Effect of treatment on daily lung parameters. The daily lung scores, weights, and virus titers on infected animals treated with each compound are summarized in Fig. 6. In this experiment, treatment with both compounds significantly inhibited development of lung consolidation, although by day 6, 2 days after termination of therapy, the lung weights in the oseltamivir-treated groups were 18 to 23% higher than those in the mice treated with RWJ-270201. The day 7 and 8 lung data were also higher in the oseltamivir-treated mice. By day 7, the placebo-treated animals had all died in this experiment. The lung virus titers were particularly inhibited on day 1 of the infection, with the 10-mg/kg/day dose of RWJ-270201 inhibiting the mean virus titers by nearly 5 log 10 units at this time. Oseltamivir treatment at the same dose lessened this titer from 10 6.9 cell culture 50% infective doses (CCID 50 )/g in the PSS-treated animals to 10 5.75 CCID 50 /g; the lower dose of RWJ-270201 yielded a lung virus titer of 10 5.25 CCID 50 /g, whereas the same dose of oseltamivir resulted in a mean lung titer of 10 6.5 CCID 50 /g at this same time. The later assay times indicated markedly less differences in lung virus titers between the treated and the control groups. It is noteworthy that by day 8, the mean lung virus titers in mice treated with both doses of RWJ-270201 were reduced to 10 1.2 CCID 50 /g whereas the virus titers in the mice receiving oseltamivir were 10 4.0 and 10 3.2 CCID 50 /g in the groups receiving dosages of 10 and 1 mg/kg/ day, respectively. Statistical analysis of the differences between the equivalent doses of each drug showed a significant difference (P Ͻ 0.05) in all lung parameters on day 8.

DISCUSSION
These data indicate that the orally administered influenza virus neuraminidase inhibitor RWJ-270201 is highly effective against experimentally induced influenza A (H1N1), A (H3N2), and B virus infections in mice. This in vivo activity appeared to often be more efficacious than that of oseltamivir when the effects of the same dosages of each compound were compared, although such differences were not seen in every experiment. The data showing a decline in SaO 2 level in the infected mice treated with oseltamivir late in the infection, which also correlated with increased lung weights at this same time (i.e. approximately 2 days after termination of treatment), suggests that the latter compound may clear from the host sooner than RWJ-270201. Li et al. (9) have reported oseltamivir to persist as the parent compound, GS4071 (oseltamivir carboxylate), for approximately 24 h in rat, dog, and ferret plasma. Pharmacokinetics studies on orally administered RWJ-270201 indicate a similar profile in mice (G. Drusano, Albany Medical College,   personal communication). No prodrug is involved with this compound. However, as shown by Eisenberg et al. (3), the levels in plasma may be misleading, since the concentration of GS4071 in the bronchoalveolar fluid in rats treated p.o. with oseltamivir declined at a lower rate than that in the plasma; such data are not yet available for RWJ-270201. RWJ-270201 was well tolerated in a dose of up to 1,000 mg/kg/day by the toxicity control mice. Acute-toxicity studies with rats have shown no untoward effects using doses as high as 3,000 mg/kg/day, and multiple-dose treatments for 5 days using doses of 1,000 mg/kg/day in rats were similarly well tolerated (T. Coogan, The R. W. Johnson Pharmaceutical Institute, Rar-itan, N.J. personal communication). The 1,000-mg/kg/day dose in rats is equivalent, on a milligram-per-square meter basis, to a dose of 2,000 mg/kg/day in the mouse. This lack of toxicity indicates a high in vivo margin of safety for the use of RWJ-270201 against influenza virus infections in mice.
The treatment schedule used in all of these experiments was twice daily for 5 days beginning 4 h pre-virus exposure. A study has also been done in which p.o. treatment twice daily for 5 days using RWJ-270201 was delayed for up to 60 h after influenza A/NWS virus exposure, with all infected mice still surviving (R. W. Sidwell and D. F. Smee, unpublished data). These data were similar to those we reported using oseltamivir   (20) and indicate the potential for RWJ-270201 to be used in a therapeutic, and not just prophylactic, manner. Additionally, the antiviral effects of other dosing schedules, including once, twice, and three times daily, are currently under investigation to determine the pharmacokinetic parameter that drives efficacy.
The results using oseltamivir in this study correlate quite well with those reported previously by us (20), where the A/Shangdong virus infections were strongly inhibited with 100and 10-mg/kg/day doses given on the same treatment schedule; in those studies, a 1-mg/kg/day dose was marginally inhibitory.
Previously, we found oseltamivir at 10 mg/kg/day to be highly protective to mice infected with influenza A/Victoria virus; in that experiment, however, the viral challenge dose was such that 5 of 16 placebo-treated mice survived (69% lethality), whereas in the present experiment the viral challenge was increased, killing 14 of 16 placebo controls (87.5%). In our earlier studies using influenza B/Hong Kong virus infection, efficacy was seen using 10-and 3.2-mg/kg/day doses of oseltamivir but no effect was seen at 1 mg/kg/day; in that study, however, the influenza B virus infection was lethal to 16 of 18 mice (88.8%) among the placebo-treated controls, with a mean  day to death of 10.5 days. In the present experiment using this virus, the viral challenge was greater, killing all placebo-treated mice, and the mean day to death of those control animals was 7.1 days. It has been well established that the severity of the viral challenge dose will influence the antiviral activities of test compounds (16). It would be informative to test and compare the efficacies of these drugs against even higher viral challenge doses. Studies to investigate this effect are under way. Efficacy using RWJ-270201 was demonstrated using all disease evaluation parameters. Inhibition of SaO 2 decline was particularly useful, since this parameter could definitively indicate, on a daily basis, the disease condition of the mouse. As seen in the experiment where lungs were taken daily for determination of consolidation score, lung weight, and virus titer, the increasing consolidation, especially seen using the more objective lung weight, correlated well with SaO 2 decline. Virus titers were highest (mean ϭ 10 6.5 CCID 50 /g) in the placebotreated mice 1 day after virus exposure; since the viral challenge dose was 10 5.1 CCID 50 /g, this was an approximate 2-log 10 -unit increase in a 24-h period. Treatment with either RWJ-270201 or oseltamivir inhibited this initial virus replication by 0.4 to 4.6 log 10 units at this time, depending on the compound and dose used. In our experience, a reduction of viral challenge dose of 0.5 log 10 unit is sufficient to prevent the majority of challenged mice from dying, so it would be expected that the infected mice treated with these compounds would survive the infection if such an initial lung virus inhibi-tion was seen. It is noted that the mice receiving the 1-mg/kg/ day dose of oseltamivir, where a maximum titer reduction of 0.4 log 10 unit was seen, did not survive the infection (Table 1). Maximal lung virus inhibition was also seen relatively late (day 8 or 9) in the infection; this time point was not routinely selected for lung assay because often the placebo-treated animals had all died by this time.
As reported by Smee et al. (21), the A/Shangdong and A/Victoria viruses were highly sensitive to RWJ-270201 in vitro, with 50% effective concentrations ranging from 0.03 to 0.17 M, depending on the assay system. The A/NWS and B/Hong Kong viruses appeared to be less sensitive in vitro, with 50% effective concentrations of 2.0 to 21 M. All the virus infections responded to RWJ-270201 therapy to approximately the same extent in the present in vivo experiments, however. These differences between in vitro and in vivo viral sensitivities to antiviral compounds have been observed by us previously using oseltamivir (20).
As indicated earlier, the influenza virus serotypes A (H1N1), A (H3N2), and B represent those viral serotypes currently responsible for the influenza outbreaks in the world today. These data indicate that the orally administered cyclopentane influenza virus neuraminidase inhibitor RWJ-270201 is an effective inhibitor of in vivo infections caused by representatives of each serotype; these data, coupled with the broad therapeutic window of the compound, indicate it to be of potential for human use.