Effective decision-making: progressing compounds through clinical development.

In today's drug-development environment, companies are under increasing pressure to improve efficiency and maintain returns on investment. Tomorrow's environment is likely to be harsher still, and companies that survive and prosper will be those that are already responding by re-inventing their structures and culture to meet the challenges that lie ahead. In this review, we explore some of the strategies and issues that are central to this process, with particular reference to decision-making in drug development.

Disposition of [14C]velnacrine maleate in rats, dogs, and humans.

This study describes the disposition of [14C]velnacrine maleate in rats, dogs, and humans, and the isolation and identification of metabolites in dog urine. Following oral administration of [14C]velnacrine maleate, drug-related material was well absorbed in all three species, with the majority of the dose recovered in the urine. Fecal elimination of radioactivity accounted for the remainder of the dose. The majority of the radioactivity was eliminated within 24 hr. Pharmacokinetic parameters for the elimination of radioactivity from the plasma of rats and dogs were similar after oral dosing compared with intravenous dosing. In humans, the plasma and urinary levels of velnacrine maleate were substantially lower, and the elimination half-life shorter than for total radioactivity, indicating the presence of one or more metabolites with a longer half-life than the parent compound. Preliminary TLC analysis of urine, plasma, and feces showed that metabolism appeared to be similar in the three species investigated. Velnacrine maleate was extensively metabolized with only approximately 10%, 19%, and 33% of the dose appearing in the urine as unchanged drug in humans, dogs, and rats, respectively. Isolation and identification of dog urinary metabolites was conducted. The identity of the isolated metabolites was determined by GC/MS and proton NMR. One of the main metabolic routes was found to be via hydroxylation of the tetrahydroaminoacridine ring with other minor hydroxylated and dihydroxylated metabolites being detected. In addition two dihydrodiol metabolites were also identified. Phase II metabolism did not appear to be a significant route.

Studies on the biochemical effects of the aldose reductase inhibitor 2,7-difluorospirofluorene-9,5'-imidazolidine-2',4'-dione (Al 1576, HOE 843). Detection of D-glucaric and D-glucuronic acid excretion by high resolution 1H and 13C NMR spectroscopy.

The effects of two aldose reductase inhibitors on the biochemical composition of rat urine were investigated using high resolution 1H and 13C NMR spectroscopy. We report the elevated excretion of D-glucaric acid (DGA) and D-glucuronic acid (GCA) following treatment with 2,7-difluorospirofluorene-9,5'-imidazolidine-2'4'-dione (Imirestat, IM, Al 1576, HOE 843) at 50 mg/kg/day for 1 month, but not with 3-4-bromo-2-fluorobenzyl-4-oxo-3-phthalazine-1-ylacetic acid (Ponalrestat, Statil), dosed at 50 mg/kg/day for 2 weeks. Sugar aciduria was also detected following treatment with the cytochrome P450 inducer phenobarbitone (PB) at 45 mg/kg/day for 1 month, although the qualitative and quantitative pattern of excretion of sugar acids differed greatly between the IM and PB treatment groups. The levels of GCA excreted are elevated 11-fold by IM treatment from 19.0 to 210.0 mumol/24 hr, but only 2.5-fold by PB, from 9.7 to 23.9 mumol/24 hr. DGA was not detectable in control urine, although levels did increase by 30% during the study from 7.5 to 10.9 mumol/24 hr, between day 8 and day 29, with IM treatment, and by 60% from 1.7 to 4.9 mumol/24 hr following PB administration for the same time period. This predominant elevation of DGA and GCA caused by IM treatment far exceeds previous records. In contrast, PB treatment resulted in an increase in intensity of a number of partially resolved sugar resonances, but at a much lower level than resulted from IM treatment. A raised level of DGA and GCA is usually associated with hepatic P450 induction; however, we report here profound DGA and GCA uria as a result of the inhibition of the aldehyde reductase, hexonate dehydrogenase (EC 1.1.1.19, EC 1.1.1.20). This mechanism is not closely linked to P450 induction, corroborating the current view that elevated excretion of DGA is not a reliable indicator of hepatic enzyme induction. This study further demonstrates the use of high resolution NMR spectroscopy in the detection of a novel biochemical effect which may go unnoticed during routine clinical chemistry tests.

Application of 19F-n.m.r. spectroscopy to the identification of dog urinary metabolites of imirestat, a spirohydantoin aldose reductase inhibitor.

1. Urine from a dog dosed orally at 20 mg/kg with 14C-imirestat, a spirohydantoin aldose reductase inhibitor, contained 17.7 and 12.5% of the administered radioactivity at 0-48 and 48-72 h respectively. 2. Radio-h.p.l.c. of the 0-48 h urine revealed a complex mixture of metabolites and a small proportion of parent drug (1.6% of dose). Direct 19F-n.m.r. spectroscopy of this urine showed the fluoride ion, numerous metabolites which were predominantly glucuronide conjugates and, as a minor component, the parent drug. 3. After incubation with beta-glucuronidase the 0-48 h urine gave a 19F-n.m.r. spectrum showing fewer signals. This finding is consistent with aromatic ring hydroxylation followed by glucuronidation being the major metabolite pathways. 4. Deconjugated urine was analysed by proton-coupled 19F-n.m.r. and two-dimensional 19F-19F correlated spectroscopy. Results indicate that major components included three monohydroxy metabolites, a diphenol with both phenolic functions in the same ring, and a phenolic metabolite containing only one fluorine atom. 5. Semi-preparative h.p.l.c. of 0-48 h dog urine gave individual glucuronides isolated as mixtures of C-9 epimers. These fractions were hydrolysed and purified a second time by h.p.l.c. to give aglycones which were analysed by multi-nuclear n.m.r. and g.l.c.-mass spectrometry. The 3- and 4-hydroxy derivatives of imirestat were identified, as was the 2-hydroxy product obtained during or following defluorination. The other major aglycone was postulated to be the 3-fluoro-2-hydroxy metabolite. This represents a novel 'NIH-shift' type pathway for the metabolism of fluorobenzenes.

Disposition of bemitradine, a renal vasodilator and diuretic, in man.

1. 14C-Bemitradine (50 mg) was rapidly and efficiently absorbed (approximately 89%) in man following a single oral dose, as a solution in gelatine capsules. Peak 14C levels of 895 +/- 154 ng equiv./ml (mean +/- S.E.M.) were reached within 2 h, and declined with half-lives of 1.07 +/- 0.25 and 13.0 +/- 5.6h. 2. No bemitradine was detected in plasma, but peak concn. (124 +/- 29 ng/ml) of its desethyl metabolite were reached at 1.05 +/- 0.28 h, and declined with a half-life of 1.32 +/- 0.08 h. 3. Desethylbemitradine was rapidly metabolized to its ether glucuronide, a phenol and a dihydrodiol which were also present as glucuronide conjugates. The glucuronides were the major compounds in plasma from 2 h after drug administration. 4. Excretion in 5 days amounted to 88.8 +/- 2.3% and 10.4 +/- 2.1% dose in urine and faeces respectively. No bemitradine or desethylbemitradine were excreted unchanged. 8-(2-Hydroxyethyl)-7-(3,4- dihydroxycyclohexa-1,5-dienyl)-1,2,4-triazolo-1,5c-pyrimidin e-5-amine (E; 17% dose); 8-(2-hydroxyethyl)-7-(4-hydroxyphenyl)-1,2,4-triazolo-1,5c- pyrimidine-5-amine (F; 4% dose), their glucuronides (A, 19% dose and B, 6% dose respectively), desethylbemitradine glucuronide (D, 25% dose) and an unidentified metabolite (C, 12% dose) were excreted in urine. Compound F was the major faecal metabolite.

Single- and multiple-dose pharmacokinetics of nufenoxole in healthy human subjects.

1. In 12 healthy subjects, after single doses of 20, 40 and 80 mg of nufenoxole, mean peak plasma drug concentrations of 400, 815 and 1463 ng/ml were reached at 2.2, 2.5 and 2.5 h respectively. 2. Nufenoxole was absorbed with an apparent half-life of less than one hour at all three doses. Nufenoxole concentrations declined biphasically after the peak, with an initial and terminal half-life of four to five hours and about 27 h respectively. These half-lives were independent of the administered dose. 3. AUC and Cmax increased with increasing dose, but AUC did not increase proportionately to dose, due to a lower value for 80 mg than expected, possibly reflecting reduced absorption. 4. Observed nufenoxole concentrations, in another 12 healthy subjects receiving single, daily 80 mg oral doses of nufenoxole for eight days, were in excellent agreement with those predicted from single-dose pharmacokinetics.

Disposition of 3H-enisoprost, a gastric antisecretory prostaglandin, in healthy humans.

1. After oral administration of 3H-enisoprost (450 micrograms) to five healthy men, as a solution in capsules, peak 3H levels of 5624 +/- 566 pg equiv./ml (mean +/- S.E.M.) were reached within one hour. No unchanged drug was detected in plasma. 2. Enisoprost was rapidly de-esterified to SC-36067 [(+/-)11 alpha, 16 zeta-dihydroxy-16-methyl-9-oxoprost-4Z, 13E-dien-1-oic acid], a pharmacologically active analogue, which reached peak concentrations of 651 +/- 200 pg/ml within 20 min of dosing. SC-36067 was eliminated metabolically, with a half-life of 1.61 h, by a combination of beta-oxidation, omega-oxidation and 9-keto-reduction. 3. After nine days 59.0 +/- 2.98% and 17.4 +/- 1.57% of the dose was excreted in urine and faeces respectively. The majority of this excretion was complete in two days. 4. Five urinary metabolites were identified by GC-MS. These were (+/-)3-[2 beta-(4-hydroxy-4-methyl-1E-octenyl)-3 alpha-hydroxy-5-oxo-1 alpha-cyclopentanyl]propanoic acid (SC-41411; 3.6% dose), (+/-)3-[3 alpha,5-dihydroxy-2 beta-(4-hydroxy-4-methyl-1E-octenyl)-1 alpha-cyclopentanyl]propanoic acid (SC-41411 PGF analogue; 4.8% dose), (+/-)3-[2 beta-(8-carboxy-4-hydroxy- 4-methyl-1E-octenyl)-3 alpha-hydroxy-5-oxo-1 alpha-cyclopentanyl] propanoic acid (SC-41411-16-carboxylic acid; 22% dose), (+/-)3-[2 beta-(8-carboxy-4- hydroxy-4-methyl-1E-octenyl)-3 alpha,5-dihydroxy-1 alpha-cyclopentanyl] propanoic acid (SC-41411 PGF analogue-16-carboxylic acid; 8.5% dose) and its gamma lactone (2.6% dose). 5. These metabolites were also identified chromatographically in plasma, as were SC-36067, (+/-)3-[2 beta-(4-hydroxy-4-methyl-1E-octenyl)-5-oxo-1 alpha- cyclopent-3-enyl]propanoic acid and (+/-)3-[2 beta-(4-hydroxy-4-methyl-1E-octenyl)-5-oxo-cyclopent-1- propanoic acid. 6. Some 5-10% of the dose was excreted in urine as tritiated water, indicating that oxidation of the 11 alpha-hydroxy group in SC-36067 or its metabolites also occurred.

Studies on the disposition of a 5-nitroimidazole in laboratory animals.

The disposition and metabolism of a 5-nitroimidazole compound (SC 28538) was investigated in the rat, beagle dog and rhesus monkey. The absorption of [14C]-SC 28538 was rapid and essentially complete after oral dosage in male animals, and also after intravaginal dosage in the female rat. Peak plasma levels of radioactivity occurred within 2 h of dosage. In the rat and dog the radioactivity was excreted predominantly in the faeces (greater than 60%) but in the monkey more than 60% was excreted in the urine. In both the male and pregnant female rat radioactivity was concentrated in the gastro-intestinal tract, liver and harderian gland and the concentrations of radioactivity in other tissues was generally lower than in plasma. Radioactivity was cleared more rapidly from plasma than from the majority of tissues. SC 28538 was extensively metabolised to form glucuronide and amino acid conjugates. The half-life of SC 28538 was of the order of 1 h in the dog and 3.7 h in the monkey.

Identification of a plasma metabolite of 5-(2-bromo-E-ethenyl)-2'-deoxyuridine.

A prominent plasma metabolite was detected in animals and man after oral administration of 5-(2-bromo-E- ethenyl )-2'-deoxyuridine. The metabolite was isolated by solvent extraction and high-performance liquid chromatography and was identified by mass spectrometry as 5-(2-bromo-E- ethenyl )uracil.

Propantheline bromide plasma level, urinary excretion and pharmacological data in a comparison of the bioavailability of three oral formulations of Pro-Banthine.

To determine the comparative bioavailability of three oral formulations of propantheline bromide (PB) by both pharmacokinetic and pharmacodynamic parameters, six normal men received three standard Pro-Banthine 15 mg tablets, two prolonged acting (PA) Pro-Banthine 30 mg tablets or one developmental PA Pro-Banthine 45 mg capsule, in a study of balanced random crossover design. Plasma concentrations and urinary excretion of the unchanged drug were measured after each treatment using a stable isotope dilution assay. Salivary secretion rate and heart rate measurements were also made at intervals after each medication. The standard Pro-Banthine formulation was significantly more bioavailable, weight for weight, than either the developmental PA capsule (45 mg), p less than 0.05, or the two 30 mg PA tablets (60 mg), p less than 0.01, based on urinary excretion and plasma levels of PB and on salivary secretion and heart rate data. There was no evidence of significant prolonged action for the PA formulations.

The metabolism of canrenone in vitro by rat liver preparations.

1. The metabolism of [1-3H]canrenone, a primary metabolite of spironolactone and potassium canrenoate, by rat liver preparations in vitro has been investigated. 2. Canrenone was metabolized by 3-oxo-delta 4-reduction to give 3 alpha-hydroxy-5 beta-spirolactones, and also by a number of O2 and NADPH-dependent microsomal hydroxylation reactions. 3. A major metabolic route requiring the presence of a microsomal fraction, but apparently independent of oxygen and NADPH, led to the formation of a number of compounds tentatively identified as trihydroxy-spirolactones.

Co-carcinogenicity of promoting agents.

An experiment is described to investigate whether coadministration of a promoting agent would enhance the carcinogenicity of a repeatedly administered complete carcinogen. Topical application of the strong carcinogen 15,16-dihydro-11-methylcyclopenta[a]phenanthren-17-one (I) in toluene containing 1% v/v croton oil is about five times more effective than applications in toluene alone, as judged from the respective mean latent periods. A similar effect is also apparent for benzo[a]pyrene.

Metabolism of ethynodiol diacetate in the rhesus monkey before and after administration of rifampicin.

1. On administration of a single oral dose of [4-(14)C]ethynodiol diacetate (0.15 mg/kg) to rhesus monkey, plasma concn. of total 14C peaked after about 4 h. About 60% of the plasma radioactivity was present as glucuronide conjugates and no unchanged drug was detected. 2. Some 67 +/- 6% (mean +/- S.D.) of the dose of 14C was excreted in 4 days, 50 +/- 6% in urine and 18 +/- 2% in faeces. Most of the urinary excretion occurred within 24 h of dosage. 3. Glucuronide conjugates accounted for 60% of the urinary 14C, and 46% of the faecal 14C was free steroids. 4. Norethisterone and its tetrahydro metabolites were identified in the free, glucuronide and sulphate fractions of plasma and urine. Keto-4,5-dihydronorethisterones and trihydroxy metabolites were identified in the conjugated fractions of urine, and a complex mixture of polar metabolites was detected in faeces. 5. Rifampicin treatment (7.5 mg/kg/day, orally) for 8 days decreased the half-life of total 14C in plasma following a single oral dose of 4-[14C]ethynodiol diacetate (0.15 mg/kg) from 44 +/- to 24 +/- 2 h. 6. Faecal elimination of total 14C was significantly increased to 29 +/- 5% of the dose following rifampicin treatment, but urinary excretion was unchanged. 7. Rifampicin treatment increased the amount of polar metabolites and decreased the amount of norethisterone in the free and conjugated fractions of plasma and urine. The amounts of sulphate and non-hydrolysed conjugates in faeces were increased after rifampicin treatment.

Mutagenic and carcinogenic metabolites of the carcinogen 15,16-dihydro-11-methylcyclopenta[a]phenanthren-17-one.

Microsomal metabolites of the carcinogen 15,16-dihydro-11-methylcyclopenta[a]phenanthren-17-one (Structure I) were separated by high-pressure liquid chromatography, and their structures were established on the basis of their ultraviolet and mass spectra, together with considerations of their general chemical properties. This was assisted by comparisons with metabolites formed in the same way from the synthetic 15-hydroxy (Structure III), 16-hydroxy (Structure II), and 11-hydroxymethyl (Structure IV) derivatives, which themselves occur as metabolites of Structural I. Products derived from attack at the two benzo-ring double bonds occurred, but no K-region products were found. Only metabolites having a non-bay region 3,4-dihydrodiol system were mutagenic and bound to DNA after in vitro microsomal activation, and it was concluded that the 3,4-dihydro-3,4-diol (Metabolite e) was the main form and that the 3,4-diols of the monools (Structure II to IV) were minor proximate forms of this carcinogen. In a two-stage experiment, the synthetic 16-ol (Structure II) was shown to be almost as carcinogenic as was Structure I itself in mice; the 15-ol (Structure III) and 11-hydroxymethyl derivative (Structure IV) were much less active. The same order was also observed in the mutagenicity of these compounds in the Ames test.