Decreased fertility in male rats administered the 5 alpha-reductase inhibitor, finasteride, is due to deficits in copulatory plug formation

Oral administration of 80 mg/kg/day of finasteride, a potent specific inhibitor of 5 alpha-reductase, to sexually mature male Sprague-Dawley rats for 24 to 38 weeks caused an approximate 30% to 40% decrease in fertility. There were no effects on mating indices or implants per pregnant female. From the mating trials, a selected group of treated males with poor reproductive performance was compared to a selected group of control males with good reproductive performance. Observed matings showed no qualitative effects on mating behavior or ejaculation. However, finasteride-treated males did not form or formed small and improperly positioned copulatory plugs, which are required in rats to transport sperm into the uterus. Intrauterine insemination of epididymal sperm from males that were nonfertile by natural mating resulted in similar numbers of embryos and unfertilized oocytes recovered from controls and finasteride-treated males, confirming that there was no effect of finasteride on the ability of sperm to fertilize. Decreased fertility of finasteride-treated males was due to failure to form copulatory plugs and is related to decreased weight of seminal vesicles and prostate, an expected pharmacologic effect. Testes weight was unaffected. Decreased fertility in male rats after finasteride administration is considered a species specific effect. The mechanism of the decrease in rats is not likely to be relevant to species that do not form copulatory plugs.

Treatment of non-small-cell lung cancer with vinblastine and very high-dose cisplatin. A Southwest Oncology Group study

Suggestions of a dose-response effect for cisplatin in non-small-cell lung cancer have contributed to the development of very high-dose cisplatin regimens (200 mg/m2 per cycle). We treated 53 eligible patients with metastatic or recurrent non-small-cell lung cancer with a combination of 100 mg/m2 cisplatin and 4 mg/m2 vinblastine, each given on days 1 and 8 of a 28-day cycle. We observed no complete response and 4 partial responses (8%). Median survival was 6 months. Toxicities of grade III or greater included leukopenia (11 cases), nausea/vomiting (6 cases), thrombocytopenia (2 cases), anemia (2 cases), and elevation of transaminase (1 case). Neurotoxicity has been reported to be a major problem in several other very high-dose cisplatin regimens. The low level of neurotoxicity observed in this study may be attributable to the median cumulative cisplatin dose of less than 600 mg/m2. This vinblastine/very high-dose cisplatin regimen showed minor activity against non-small-cell lung cancer. The level of activity did not surpass that of standard-dose (100 mg/m2 per cycle) cisplatin-containing regimens.

Reversible decreases of fertility in male Sprague-Dawley rats treated orally with finasteride, a 5 alpha-reductase inhibitor

Finasteride, a 5 alpha-reductase inhibitor, was investigated for its effects on fertility in male rats as part of its preclinical safety assessment. Studies were initiated when the male Sprague-Dawley rats were either young (4 to 6 weeks old) or mature (15 weeks old). Treatment duration ranged from 6 to 32 weeks. Each male was cohabited with two untreated females at various periods during and after treatment. Litter parameters were evaluated on either day 14 or 20 of gestation. Males were necropsied at the end of treatment or 7 to 11 weeks following the end of treatment. The major findings of these studies were that 1) young rats given 20 to 80 mg/kg/day of finasteride first showed mild to moderate decreases in fertility after 12 weeks of treatment, whereas mature males (given only 80 mg/kg/day) did not show a similar decrease until 24 weeks of treatment, 2) fewer copulatory plugs and atrophy of prostates and seminal vesicles were associated with finasteride treatment, 3) the decreased fertility was only partial (ie, fertility index did not decrease below 48% of control in any study) and was not due to decreases in mating, 4) formation of copulatory plugs, organ weights, and fertility returned to normal levels after at least 6 weeks of drug withdrawal, and 5) the testes showed no histologic or weight changes that would explain the effect on fertility. These results show that the decreased fertility in male rats was associated with finasteride-induced inhibition of accessory gland secretions, an expected pharmacologic effect.

External genitalia abnormalities in male rats exposed in utero to finasteride, a 5 alpha-reductase inhibitor

A series of studies was conducted to determine the developmental toxicity of the 5 alpha-reductase inhibitor finasteride (MK-0906) in rats. This compound was administered orally once daily to pregnant rats during various extended treatment periods during gestation. F1 offspring were evaluated on Day 20 of gestation as well as postnatally through mating to produce an F2 generation. MK-0906 treatment induced dosage-related incidences of hypospadias (penischisis) in male offspring with a threshold dosage level near 0.1 mg/kg/day and a 100% effect level of 100 mg/kg/day (with dosing through Day 20 of gestation). MK-0906 also caused decreased anogenital distance in male offspring. The dosage response for this effect (ranging from a 4.2% decrease at 0.003 mg/kg/day to a 38% decrease at 100 mg/kg/day) was more shallow than that for hypospadias. The decreases in anogenital distance were at least partially reversible postnatally with essentially complete recovery at dosages up to 0.1 mg/kg/day. There was also a dosage-related, temporary induction of nipples in F1 males. All of these effects were apparent following treatment on Days 6 through 17 of gestation but were more pronounced when dosing extended to Day 20 of gestation. Slight maternal toxicity consisting of minor decreases in body weight gain occurred only at dosages of 3 mg/kg/day and higher, indicating the selective nature of the developmental toxicity. The 5 alpha-reductase enzyme located in the rat fetal genital tubercle was studied in vitro and compared to that in the adult ventral prostate. The values for Km, Vmax, and IC50 for inhibition by MK-0906 were similar in the two tissues, suggesting that the enzymatic proteins in the genital tubercle and ventral prostate may be similar.

Maternotoxicity and fetotoxicity of an angiotensin-converting enzyme inhibitor, enalapril, in rabbits

When enalapril, an angiotensin-converting enzyme (ACE) inhibitor, was orally administered to inseminated rabbits at dosages of 0.1 to 30 mg/kg/day for 13 days in a range-finding study, nephrotoxicity, as measured by elevated serum urea nitrogen concentrations, occurred at 1 mg/kg/day and higher dosages and significant (p less than or equal to 0.05) increases in fetal wastage were observed at dosages as low as 3 mg/kg/day. Saline supplementation during treatment prevented this rise in urea nitrogen. Fetal wastage was significantly (p less than or equal to 0.05) increased in the absence of maternotoxicity when saline-supplemented females were treated with enalapril at 30 mg/kg/day. A developmental toxicity study of enalapril in saline-supplemented rabbits produced no evidence of teratogenicity at 3, 10, and 30 mg/kg/day. The period of sensitivity of fetuses to the toxic effects of enalapril was found to be limited to middle-to-late gestation (Gestational Days 14-27). A single oral dose of enalapril (30 mg/kg) on Day 26 of gestation resulted in 100% fetal deaths. On the basis of the work done by Broughton Pipkin et al. [1982, J. Physiol. (London) 323, 415-422] and Broughton Pipkin and Wallace (1986, Brit. J. Pharmacol. 87, 533-542), which demonstrated that the sheep fetus becomes markedly hypotensive when the dam is treated with captopril or enalapril during late pregnancy, we believe that the observed fetotoxicity of enalapril in rabbits is also due to fetal hypotension.(ABSTRACT TRUNCATED AT 250 WORDS)

On the etiology of subcapsular lenticular opacities produced in dogs receiving HMG-CoA reductase inhibitors

The administration of high dosages of various hydroxymethylglutaryl-CoA (HMG-CoA) reductase inhibitors has resulted in the development of subcapsular lenticular opacities in dogs. While dogs receiving cataractogenic doses of HMG-CoA reductase inhibitors experienced profound decreases in circulating serum cholesterol concentrations (40-60% reductions in total serum cholesterol), a causal relationship between serum cholesterol lowering and cataractogenesis was not established. A strong relationship was demonstrated, however, between the systemic exposure to inhibitor (plasma drug levels) and the cataractogenic potential of the various compounds studied. Analysis of lenses from dogs chronically dosed with various HMG-CoA reductase inhibitors revealed the presence of low drug levels in the lens (less than 500 ng equivalents g-1), but no correlation was observed between the amount of drug associated with the lens after chronic treatment and cataract development. In addition, no abnormalities in cholesterol content or sterol composition were observed in clear and/or cataract containing lenses from dogs chronically dosed with HMG-CoA reductase inhibitors. The kinetics of drug appearance in the aqueous and lens cortex was assessed after doses of various HMG-CoA reductase inhibitors, and suggested somewhat higher but not statistically significant peak concentrations of inhibitor were achieved by compounds which produced a higher incidence of cataracts. These data have suggested that high doses of HMG-CoA reductase inhibitors may increase lenticular exposure to drug via the aqueous humor by producing a substantial systemic exposure to drug substance. This may result in an increased concentration of inhibitor in the outer cortical region of the lens where cholesterol synthesis is critical, thereby resulting in the development of opacities. The production of lenticular changes by a HMG-CoA reductase inhibitor of diverse chemical structure establishes, with reasonable assurance, that these lens changes are mechanism based (i.e. a product of the biochemical mechanism of action of this class of compounds). An extrapolation of these findings to patients receiving therapeutic dosages enables a favorable risk evaluation since the doses to be employed clinically are much lower and result in a far lower systemic exposure to drug substance.

High-dose cisplatin in disseminated melanoma: a comparison of two schedules

A total of 38 patients with metastatic melanoma received monthly chemotherapy with cisplatin at a dose of 200 mg/m2, per cycle; 14 received 20 mg/m2 cisplatin i.v. on days 1-5 and 24 were given 100 mg/m2 i.v. on days 1 and 8. Objective responses were seen in 2/14 treated on days 1-5 and in 5 of 22 evaluable subjects receiving cisplatin on days 1 and 8, for an overall response rate of 22%. The median survival of all patients was 6 months, with no significant difference observed between the two schedules. Severe neurotoxicity and myelosuppression were more common in patients treated on days 1-5. Two patients treated in this manner were bedridden due to neurotoxicity and four developed grade 4 leukopenia after the first cycle of chemotherapy. Only one patient treated with the divided-dose schedule became leukopenic during the first cycle, and none of the patients were debilitated by neurotoxicity. Thrombocytopenia was statistically more severe. Nausea and vomiting, fatigue, ototoxicity, and paresthesia were seen with equal frequency. Very high doses of cisplatin can be delivered with acceptable toxicity using a divided-dose schedule. As the response rate on this schedule appeared to be comparable with that achieved on the more toxic consecutive 5-day schedule, the former deserves to be tested in diseases known to show a dose response to cisplatin. However, in melanoma, administration of 200 mg/m2 per course did not appear to be associated with a markedly improved response rate, compared with cisplatin alone at "standard" doses.

Animal safety and toxicology of simvastatin and related hydroxy-methylglutaryl-coenzyme A reductase inhibitors

Simvastatin, a hydroxy-methylglutaryl-coenzyme A reductase inhibitor intended for use as a hypocholesterolemic agent, has undergone a thorough preclinical toxicology evaluation. This review describes preclinical toxicology findings associated with simvastatin administration in animals and provides the rationale for our conclusion that these changes are not indicative of potential human toxicity. Although it was not surprising to find that a potent inhibitor of this key biochemical pathway produces toxicity at high dosages in animals, none of the observed changes poses a significant risk to humans at clinical dosages. Many of the toxicities produced by high dosage levels of simvastatin in animals are directly related to the drug's biochemical mechanism of action and are the result of a profound, sustained inhibition of the target enzyme that is not anticipated at clinical dosages. Furthermore, several of the simvastatin-induced changes are species-specific responses to this agent and are not relevant to human risk assessment. Of the treatment-related changes reported for simvastatin, the development of cataracts in dogs has received considerable attention. The available data demonstrate a wide margin of safety in terms of dosage levels required to elicit this response as well as the plasma concentrations associated with the development of these ocular lesions. The data suggest that the development of lenticular opacities at clinical doses of simvastatin is highly improbable. Overall, simvastatin is highly improbable. Overall, simvastatin was well-tolerated by animals in preclinical toxicology studies, and no findings contraindicating its use in humans were identified.

Toxicity of the HMG-coenzyme A reductase inhibitor, lovastatin, to rabbits

Lovastatin, a specific inhibitor of the rate-limiting enzyme in cholesterol biosynthesis, HMG-CoA reductase, has been shown to be highly effective in lowering serum cholesterol in animals and humans and thus represents a promising approach to the treatment and prevention of cardiovascular disease. During the preclinical safety assessment of lovastatin, oral doses that were tolerated by dogs, rats and mice were found to be lethal to rabbits in subacute studies. Postmortem findings in rabbits consisted of centrilobular hepatic necrosis, frequently accompanied by renal tubular necrosis and occasionally gallbladder necrosis. The liver lesions were associated with up to 300-fold elevations in serum aspartate and alanine aminotransferase activities, whereas the kidney lesions resulted in accumulations of serum urea nitrogen and creatinine. The organ damage was preceded by a progressive decline in food consumption and loss of body weight. All histopathological and serum biochemical changes induced by lovastatin were completely prevented by coadministration of mevalonate, the product of the inhibited HMG-CoA reductase enzyme. In addition, administration of mevalonate after the onset of lovastatin-induced hepatotoxicity effectively reversed the toxicity despite continued drug treatment. These findings indicated that the toxicity of high doses of lovastatin to rabbits is a consequence of a highly exaggerated pharmacologic action in blocking mevalonate synthesis. However, supplementation of lovastatin-treated rabbits with oral doses of the major product of mevalonate metabolism, cholesterol, paradoxically enhanced the liver and kidney damage, which suggested that the toxicity of lovastatin stemmed from depletion of a nonsterol metabolite(s) of mevalonate critical for cell viability.

Preclinical evaluation of lovastatin

Administration of lovastatin to animals at high dosage levels produces a broad spectrum of toxicity. This toxicity is expected based on the critical nature of the target enzyme (HMG CoA reductase) and the magnitude of the dosage levels used. The information reviewed in this paper demonstrates that these adverse findings in animals do not predict significant risk in humans. The reason for this derives from the fact that all the available evidence suggests that the adverse effects observed are produced by an exaggeration of the desired biochemical effect of the drug at high dosage levels. The presence of clear and high no-effect doses for these toxic effects along with the fact that most of the changes observed are clearly mechanism-based (directly attributable to inhibition of mevalonate synthesis) indicate that it is unlikely that similar changes will be observed at the therapeutic dosage levels in humans. This hypothesis is supported by the extensive human safety experience described by Tobert in the following report.

Brain and optic system pathology in hypocholesterolemic dogs treated with a competitive inhibitor of 3-hydroxy-3-methylglutaryl coenzyme A reductase

The cholesterol lowering compound lovastatin, a competitive inhibitor of 3-hydroxy-3-methylglutaryl coenzyme A reductase (EC 1.1.1.34 HMG CoA reductase), was given in nine separate experiments to normocholesterolemic dogs at rates up to 180 times the maximum therapeutic dose in man (1 mg/kg/day). Mean serum total cholesterol concentrations were reduced as much as 88% below normal. Clinical evidence of neurotoxicity occurred in up to 37% of animals given 180 mg/kg/day lovastatin for 11 or more days, especially in one laboratory where the dosing regime resulted in higher concentrations of plasma drug levels. Dogs receiving 60 mg/kg/day or less never exhibited neurologic signs. The central nervous system (CNS) of affected dogs exhibited endothelial degeneration and hemorrhagic encephalopathy. Focal extravasation of horseradish peroxidase occurred frequently (6/8) in the retrolaminar optic nerve of asymptomatic or clinically affected dogs given 180 mg/kg/day lovastatin, with endothelial degeneration and discrete optic nerve degenerative lesions interpreted as ischemic. The association between the degree of hypocholesterolemia and occurrence of clinical signs was not exact. Total brain cholesterol was similar in treated and control dogs. Hypocholesterolemic dogs had proportionally lowered serum concentrations of alpha-tocopherol, but oral supplementation of this vitamin did not prevent the neurologic syndrome. Endothelial degeneration in the CNS and optic nerve may have reflected in vitro morphologic effects of HMG CoA reductase inhibitors due to extreme inhibition of nonsterol isoprene synthesis. Retinogeniculate axonal (Wallerian-like) degeneration occurred in greater than or equal to 12% of dogs given 60 mg/kg/day or more lovastatin, with central chromatolysis of occasional retinal ganglion cells. These neuroaxonal changes may have been secondary to vascular effects, but superimposed direct neurotoxic action at the high dosage levels of lovastatin could not be excluded. There was no evidence of drug induced adverse effects in the CNS of dogs given up to 30 mg/kg/day lovastatin for 2 years.

Mechanism of action and biological profile of HMG CoA reductase inhibitors. A new therapeutic alternative

Lovastatin (MK-803, mevinolin) and simvastatin (MK-733, synvinolin), 2 highly potent 3-hydroxy-3-methylglutaryl coenzyme A (HMG CoA) reductase inhibitors, have been heralded as breakthrough therapy for the treatment of atherosclerotic disease. This paper discusses the biochemical attributes of these HMG CoA reductase inhibitors, their structures and inhibitory properties in a variety of biological systems and presents the rationale for their therapeutic use. Not only do lovastatin and simvastatin potently inhibit cholesterol biosynthesis; they also can result in the induction of hepatic low density lipoprotein (LDL) receptors, thus increasing the catabolism of LDL-cholesterol. Lovastatin and simvastatin are the first HMG CoA reductase inhibitors to receive regulatory agency approval for marketed use. Their safety profiles are reviewed and 2 aspects of this evaluation are stressed. First, the objective in the clinical use of these inhibitors is to normalise plasma cholesterol levels in hypercholesterolaemic individuals. This contrasts with the profound reductions in cholesterol obtained when normocholesterolaemic animals are treated by the high doses of these drugs required for toxicological assessment. Second, both lovastatin and simvastatin are administered as prodrugs in their lactone forms. As lactones, they readily undergo first-pass metabolism, hepatic sequestration and hydrolysis to the active form. Consequently, lovastatin and simvastatin achieve lower plasma drug levels than do other HMG CoA reductase inhibitors in clinical development. Low plasma levels have been established as an important determinant of safety in the use of HMG CoA reductase inhibitors in both animal and human studies.

Carbapenems, a new class of beta-lactam antibiotics. Discovery and development of imipenem/cilastatin

The discovery of Streptomyces cattleya and its antibiotic product, thienamycin, has ushered in a new era of beta-lactam agents, the carbapenems. Numerous carbapenems were subsequently discovered; however, none had the potency, broad-spectrum activity, and lack of cross-resistance exhibited by thienamycin. Chemical instability encountered with thienamycin was overcome by the N-formimidoyl derivative, imipenem. Imipenem is distinguished from other beta-lactams by its outstanding activity against gram-positive organisms as well as against Enterobacteriaceae, Pseudomonas aeruginosa, and Bacteroides. However, development was hindered by extensive renal metabolism of imipenem, resulting in low urinary concentrations of antibiotic. A renal dipeptidase, dehydropeptidase-I, was responsible for hydrolyzing imipenem and other carbapenems. To counter its action, a specific inhibitor, cilastatin, was developed. Coadministered with imipenem in a one-to-one ratio, cilastatin provides prolonged, reversible blockade of imipenem metabolism, dramatically improving urinary recoveries to therapeutically significant levels. Cilastatin also contributes to the safety of imipenem, since its coadministration prevents proximal tubular necrosis which has been observed in sensitive animals receiving imipenem alone in high doses. Thus, the combination imipenem and cilastatin overcame the pharmaceutical and metabolic challenges presented by thienamycin, and allowed for the evaluation in humans of the outstanding antimicrobial activity of this new class of beta-lactam antibiotics.

Mevalonate supplementation in pregnant rats suppresses the teratogenicity of mevinolinic acid, an inhibitor of 3-hydroxy-3-methylglutaryl-coenzyme a reductase

Mevinolin is a fungal metabolite, and in the hydroxyacid form, mevinolinic acid, it is an inhibitor of 3-hydroxy-3-methylglutaryl-coenzyme A (HMG-Co A) reductase, an enzyme essential in cholesterol biosynthesis. Oral administration of 800 mg/kg/day of mevinolin to rats from days 6 through 17 of gestation produced fetal malformations of the vertebrae and ribs in 29% of the litters, and there was a treatment-related increase in the incidence of gastroschisis. Mevinolinic acid at 60 and 90 mg/kg/day also produced fetal malformations of the vertebrae and ribs, and these teratogenic manifestations were markedly suppressed by coadministration of the product of HMG-Co A reductase, mevalonic acid, at a dosage level of 500 mg/kg b.i.d. A diet supplemented with 0.5% or 1.0% cholesterol had no effect on the teratogenicity of mevinolinic acid. Teratology studies in rats with a dihydroxyheptanoic acid derivative of mevinolin, a compound 1/700 as potent as mevinolinic acid as an inhibitor of HMG-Co A reductase, and dihydromevinolinic acid, an inhibitor of this enzyme comparable in activity to mevinolinic acid, indicated that the teratogenicity of these compounds was related to their relative enzyme inhibitory activity. The dihydroxyheptanoic acid derivative was not teratogenic at doses as high as 150 mg/kg b.i.d.; in contrast, when dihydromevinolinic acid was administered at 50 and 100 mg/kg/day, its potency as a teratogenic agent was comparable to that of mevinolinic acid. These studies demonstrated that inhibitors of HMG-CoA reductase produced terata in rats and that the teratogenic effects could be antagonized by coadministration of the enzyme product, mevalonic acid.(ABSTRACT TRUNCATED AT 250 WORDS)

Detection of breast carcinoma metastases in bone: relative merits of X-rays and skeletal scintigraphy

Of 1116 patients receiving primary treatment for breast carcinoma at the Royal Marsden Hospital since 1976, 651 had an abnormal bone scintigram either at primary diagnosis (378) or on subsequent follow-up (273) and 167 developed radiographically detectable bone metastases (21 at the time of primary diagnosis). Comparison of bone scintigrams and X-rays showed that scintigraphy was an inaccurate localiser of existing radiographic detectable metastases. If X-rays alone are used to detect bone metastases a limited examination with five plates will detect metastases with 92% accuracy. After primary surgery, routine X-ray screening for bone metastases is not necessary since it is possible to identify patients at risk on the basis of clinical examination, chest X-ray, and serum alkaline phosphatase and gamma-glutamyl transpeptidase levels.

The effect of sequential addition of the nitrosourea, chlorozotocin, to the FAM combination in advanced gastric cancer

Twenty-three patients with advanced measurable gastric cancer were treated with FAM-chlorozotocin in an attempt to demonstrate improvement in rate and duration of response over FAM (5-fluorouracil, Adriamycin and mitomycin C). Six (26%) partial responses were recorded with a median duration of two months. FAM-chlorozotocin was well tolerated with moderate myelosuppression as the major dose-limiting toxicity; the leukocyte nadir was 3.0 X 10(3)/mm3 (range, 0.9-6.0) and the platelet nadir was 100 X 10(3)/mm3 (range, 45-100). The authors were unable to find significant differences in prognostic factors such as performance status, sites of metastatic disease, intensity of therapy to account for the discrepancy in response rates between FAM (42%) and FAM-chlorozotocin (26%).

Tumorigenicity of 7,12-dimethylbenz[a]anthracene, its hydroxymethylated derivatives and selected dihydrodiols in the newborn mouse

The newborn mouse lung adenoma model has been shown to be a sensitive test for studying the tumorigenicity of bay region diol epoxides and their precursor dihydrodiols. When a total dose of 28 nmol of 7,12-dimethylbenz[a]anthracene (DMBA) or its derivatives was injected i.p. into the preweaning mice, it was found that the 3,4-dihydrodiols of both DMBA and 7-hydroxymethyl-12-methylbenz[a]anthracene caused 13.3 and 4.1 times more lung adenomas than DMBA, respectively. The mice treated with the 5,6- and 8,9-dihydrodiols of DMBA, 7-hydroxymethyl-12-methylbenz[a]anthracene and its 5,6- and 8,9- and 10,11-dihydrodiols, 7-methyl-12-hydroxymethylbenz[a]anthracene and 7,12-dihydroxymethylbenz[a]anthracene developed a level of lung adenomas/mouse less than 2-fold higher than that found in the DMSO-treated control group. Liver tumors also developed in some of the mice. The percentage of mice with liver tumors also indicated that the 3,4-dihydrodiols of both DMBA and 7-hydroxymethyl-12-methylbenz[a]anthracene were more tumorigenic than DMBA itself. These data indicate that the 3,4-dihydrodiols of both DMBA and its 7-hydroxymethyl derivative may be proximate carcinogenic metabolites of DMBA in the newborn mouse.

5-fluorouracil, Adriamycin, and mitomycin-C (FAM) chemotherapy for advanced adenocarcinoma of the pancreas

Thirty-nine patients with advanced metastatic adenocarcinoma of the pancreas were treated with a combination of 5-fluorouracil, Adriamycin, and mitomycin-C (FAM). Twenty-seven of these patients had measurable disease, and ten (37%) achieved a partial response. An additional three patients had evidence of disease stabilization. The median survival period of responding patients was 12 months, compared with 3.5 months for nonresponders (P < 0.01). The median survival period for all patients was 6 months. Moderate myelosuppression constituted the treatment-limiting toxicity. The FAM combination is an active and well-tolerated regimen for pancreatic cancer and may have an application in the management of patients with less advanced disease.