[Efficacy of prostaglandin derivatives and mucoprotective drugs in treatment and prevention for NSAIDs-induced ulcer].

Deficiency of prostaglandins (PGs) by non-steroidal anti-inflammatory drugs (NSAIDs) causes a loss of gastroduodenal mucosal integrity, leading to development of ulceration. PG derivatives such as misoprostol and enprostil have been proven effective in prevention and treatment of NSAIDs-induced gastroduodenal ulcers. Although side effects such as diarrhea limit the use of PG derivatives, the efficacy of these drugs in NSAIDs-induced injuries is approximately equal to that of proton pump inhibitors. Mucoprotective drugs such as rebamipide also have been reported to be effective for prevention of NSAIDs-induced ulcers. Since misoprostol and some mucoprotective drugs can prevent NSAIDs-induced small intestinal injuries, these drugs, especially misoprostol, should be used as first-line therapy for NSAIDs-induced gastrointestinal injuries with attention paid to the side effects.

Properties of Rikkunshi-to (TJ-43)-induced relaxation of rat gastric fundus smooth muscles.

The relaxant effects of Rikkunshi-to (TJ-43), a gastroprotective herbal medicine, on rat gastric fundus were investigated. Experiments were carried out using standard tension and intracellular microelectrode recording techniques. During contraction induced by enprostil (0.5 microM), a prostaglandin E(2) analog, TJ-43, produced relaxation dose dependently (0.1-5.0 mg/ml) in the rat fundic circular smooth muscle (CSM) strips. The relaxant effects of TJ-43 were not affected by tetrodotoxin or 1 H[1, 2, 4] oxadiazolo [4, 3-a] quinoxalin-1-one (10 microM), an inhibitor of soluble guanylate cyclase. TJ-43 inhibited enprostil-induced membrane depolarization. Apamin (1 microM), a blocker of small-conductance Ca(2+)-activated K(+) (SK) channel, inhibited T-43-induced membrane repolarization. TJ-43-induced relaxation was biphasic, comprising of an initial fast followed by a second slow relaxation. The fast relaxation was abolished by apamin. Application of high K(+) (29.4 mM [K(+)](o)) also abolished the fast relaxation induced by TJ-43. In diabetic Goto-Kakizaki (GK) rat fundic CSM strips, the relaxant responses of TJ-43 during enprostil-induced contraction were increased compared with control rat strips. These results indicate that TJ-43 elicited fast muscle relaxation through membrane hyperpolarization induced by the activation of SK channels; the time-dependent slow relaxation reflects an additional direct of TJ-43 on CSM in the rat gastric fundus. Because TJ-43-evoked relaxation of fundic CSM strips was more potent in diabetic GK rat than in control rat, further analysis of this herb could lead to better treatments of diabetic gastroparesis.

Combination of enprostil and cimetidine is more effective than cimetidine alone in treating gastric ulcer: prospective multicenter randomized controlled trial.

BACKGROUND/AIMS: Little is known about the clinical efficacy of co-therapy of enprostil, a prostaglandin E2 analogue, with a histamine H2-receptor antagonist. We aimed to assess the additive benefit of enprostil in combination with cimetidine for treating gastric ulcer in a prospective multicenter randomized controlled trial. METHODOLOGY: In 43 hospitals 171 intention-to-treat (ITT) patients, diagnosed as having gastric ulcer by endoscopy, were randomly allocated to receive either enprostil 25microg b.i.d. and cimetidine 400mg b.i.d. (Group E=85), or cimetidine 400mg b.i.d. alone (Group C=86) for 8 weeks. Healing was examined by endoscopy at 4 and 8 weeks. RESULTS: Per protocol (PP) analysis comprised 166 patients (E=82, C=84). Despite no significant advantage at 4 weeks (E=55.3%, C=42.2%), the combination yielded higher healing rates at 8 weeks by ITT (E=89.4%, C=68.6%; p<0.001) and PP analysis (E=92.7%, C=70.2%; p<0.001). Symptom relief rates [E, C] at 2, 4, and 8 weeks were [80.2%, 68.3%] (not significant), [97.4%, 88.3%] (p<0.05), and [95.6%, 87.0%] (p<0.05), respectively. Significant advantage was observed in the patients aged 40 or older, with solitary ulcer (>5mm in diameter), and without smoking or drinking habits. No adverse effects were critical. CONCLUSIONS: Enprostil safely and significantly augmented gastric ulcer healing and symptom relief by cimetidine.

Pharmacological characterization and identification of EP3 prostanoid receptor binding sites in hamster uterus homogenates.

The pharmacological properties of [(3)H]-prostaglandin E(2) ([(3)H]-PGE(2)) binding to washed homogenates of hamster uterus were determined. Scatchard analysis of competition data yielded dissociation constants (K(d)s) of 30.9 +/- 5.6 nM (n = 3) and apparent receptor density (B(max)) of 25.25 +/- 1.89 pmol g(-1) wet weight tissue (74 +/- 8% specific binding). Competition studies yielded the following affinity parameters (K(i)) for various prostanoids: GR63799X = 13 4 nM; PGE(2) = 17 +/- 3 nM; sulprostone = 64 +/- 5 nM; enprostil = 67 +/- 3 nM; misoprostol = 124 +/- 15 nM; cloprostenol = 187 +/- 33 nM; carba-prostacyclin = 260 +/- 167 nM; iloprost = 555 +/- 162 nM; PGF(2 alpha) = 767 +/- 73 nM; PGD(2) > 3560 nM; fluprostenol = 11 790 +/- 2776 nM; RS93520 = 21 558 +/- 14 228 nM. These data closely matched the pharmacological profile of previously described EP(3) receptors such as in bovine corpus luteum (BCLM) and the cloned mammalian EP(3) receptors. The high correlation between the current hamster uterus pharmacology data vs the EP(3) receptor binding in BCLM (r = 0.94; P < 0.0001), vs cloned human EP(3) receptor (r = 0.94, P < 0.0001), vs the cloned mouse EP(3) receptor binding (r = 0.78; P < 0.002), vs cloned rat EP(3) receptor (r = 0.9, P < 0.0004), and vs EP(3) receptor-mediated functional responses (r = 0.72, P < 0.02) substantiated the conclusion that the hamster uterus contains EP(3) receptor binding sites.

Enprostil, a prostaglandin-E(2) analogue, inhibits interleukin-8 production of human colonic epithelial cell lines.

We previously reported that intracolonic administration of enprostil, a prostaglandin-E(2) (PGE(2)) analogue, had therapeutic effects on acute colitis induced in rodents by dextran sulfate sodium (DSS). In addition, production of growth-regulated gene product/cytokine-induced neutrophil chemoattractant-1 [GRO/CINC-1; an interleukin(IL)-8 like cytokine] was suppressed in the inflamed tissues. In the present study we used a human colon cancer cell line (HT-29) to investigate enprostil effects on the IL-8 production of intestinal epithelial cells stimulated by various stimulants. In a MTT assay, concentrations of enprostil >10(-5)M had cytotoxitic effects on HT-29 cells. Furthermore, 10(-6) M enprostil suppressed IL-8 production in HT-29 cells, SW620 and CaCo2 stimulated with interleukin-1 beta (IL-1 beta) or lipopolysaccharide (LPS), but did not suppress this response when cells were stimulated with tumour necrosis factor (TNF)-alpha. These results suggest that enprostil affects a point in the pathway between the IL-1 receptor or LPS receptor and nuclear factor-kappa B(NF-kappa B), without affecting the pathway between the TNF receptor and NF-kappa B, with the latter factor being required for the IL-8 gene transcription. The therapeutic effect of exogenous enprostil on DSS colitis may involve the inhibition of IL-8 production in colonic epithelial cells stimulated by IL-1 beta or LPS.

Effect of intrarectal prostaglandin E2 analogue (enprostil) on trinitrobenzenesulphonic acid-induced colitis in rats.

Prostaglandins exert a protective effect on colonic mucosa in experimentally induced colitis. This study investigated the effect of enprostil, a prostaglandin E2 (PGE2) analogue, on trinitrobenzenesulphonic acid (TNBS)-induced colitis in rats. Each rat received a rectal enema containing TNBS (30 mg), followed 24 h later by intrarectal once-daily enprostil (200 microg). Enprostil-treated and control rats were killed on day 3 (enprostil group, n = 5; control, n = 6) or day 10 (enprostil group, n = 5; control, n = 5) after TNBS treatment. The area of damaged mucosa of the colon was measured relative to the total colonic area. We also determined the macroscopic score of mucosal damage, and measured PGE2, 6-keto-prostaglandin F1alpha (6-keto-PGF1alpha) and thromboxane B2 (TXB2) concentration in portal vein blood samples. Enprostil significantly reduced both the area of damaged mucosa (including the ulcer area) and the macroscopic score after 3 days' treatment compared with control. Similarly, enprostil significantly reduced plasma concentration of PGE2, 6-keto-PGF1alpha and TXB2 during the acute phase at day 3 of treatment compared with control, but not at day 10. These results suggest that PGE2 enema may have therapeutic potential for treating patients with proctitis or left-sided colitis.

Effect of enprostil on omeprazole-induced hypergastrinemia and inhibition of gastric acid secretion in peptic ulcer patients.

This study was performed to examine the effects of additional enprostil administration on hypergastrinemia and gastric acid suppression induced by omeprazole. Serum gastrin concentrations were measured in 10 peptic ulcer patients (six Helicobacter pylori-positive and four Helicobacter pylori-negative patients) before treatment, after two weeks of omeprazole (20 mg/day), and after two weeks of omeprazole and enprostil (50 micrograms/day). The additional acid inhibitory effect of enprostil was evaluated by 24-hr intragastric pH measurements in five healthy Helicobacter pylori-negative volunteers. After omeprazole treatment, the serum gastrin level of Helicobacter pylori-positive patients (3.5-fold of control) was markedly higher than that of Helicobacter pylori-negative patients (1.7-fold of control). Additional treatment with enprostil suppressed serum gastrin levels to 0.4-fold and 0.7-fold of omeprazole treatment levels in Helicobacter pylori-positive and Helicobacter pylori-negative patients, respectively. In healthy volunteers, median pH recorded during the nonmeal daytime interval increased significantly with additional enprostil. Thus, enprostil reduces undesirable omeprazole-induced hypergastrinemia, especially in Helicobacter pylori-positive patients, and effectively suppresses acid secretion.

Prostaglandins induce proliferation of rat hepatocytes through a prostaglandin E2 receptor EP3 subtype.

We characterized the proliferative action of prostaglandins (PGs) in relation to their membrane receptors on rat hepatocytes in primary culture. PGs in the order 16,16-dimethyl PGE2 > PGE2 > PGF2alpha >> PGD2 augmented epidermal growth factor (EGF)/insulin-induced DNA synthesis, assessed by [(3)H]thymidine incorporation, in a concentration-dependent manner, whereas PGs alone did not stimulate basal DNA synthesis without EGF and insulin. The cells exhibited [(3)H]PGE2 binding sites that were displaced by unlabeled PGs in the order PGE1 = PGE2 > PGF2alpha > PGD2. PGE2 inhibited glucagon-stimulated adenosine 3',5'-cyclic monophosphate (cAMP) accumulation concentration dependently. The mean effective concentration for DNA synthesis, median inhibitory concentration for cAMP accumulation, and dissociation constant for [(3)H]PGE2 binding at 25 degrees C were almost identical (approximately 70 nM). Treatment of the cells with pertussis toxin (100 ng/ml), which ADP-ribosylated most of the 41-kDa substrate, abolished the proliferative effects of PGs. We detected the expression of mRNA of the EP3 subtype PGE2 receptor using reverse transcription-polymerase chain reaction. Moreover, an EP3 agonist, enprostil, but not the EP1 agonist 17-phenyl-trinor-PGE2 or the EP2/EP4 agonist 11-deoxy-PGE1, stimulated EGF/insulin-induced DNA synthesis. These results indicate that PGs act as comitogenic growth factors through the EP3 subtype PGE2 receptor coupled with G(i) protein in cultured rat hepatocytes.

Distribution of prostaglandin E receptors in the rat gastrointestinal tract.

AIMS: In order to study the role of prostaglandin in the regulation of the gastrointestinal functions, gene expression of prostaglandin receptors along the rat gastrointestinal tracts were investigated. METHODS: Rats were used for the study. The combination of counterflow elutriation separation of mucosal cells and Northern blot analysis was used to detect the gene expression of prostaglandin receptors in gastrointestinal tracts. RESULTS: In small intestine and colon, prostaglandin E2 EP1 and EP3 receptor mRNAs were mainly localized in the deeper intestinal wall containing muscle layers. EP4 receptor gene expression, on the other hand, was detected in the intestinal mucosal layer. In the stomach, EP1 mRNA was detected in gastric muscle layers, whereas EP3 and EP4 receptor gene expression was mainly present in the gastric mucosal layer containing epithelial cells. In gastric epithelial cells, parietal cells were found to have both EP3 and EP4 receptors. At lower concentrations, prostaglandin E2 inhibited gastric acid secretion by parietal cells probably through EP4 receptors. At higher concentrations, however, it stimulated it. On the other hand, mucous cells possessed only EP4 receptor mRNA. CONCLUSIONS: Thus, it is suggested that prostaglandin E2 modulates gastrointestinal functions through at least three different prostaglandin receptors (EP1, EP3, and EP4), each of which has a distinct contribution in the gastrointestinal tract.

Effects of enprostil on gastric endocrine secretion during chronic administration of lansoprazole.

We investigated changes in the secretory kinetics of gastric endocrine cells related to the administration of lansoprazole, and the effects of enprostil on these altered kinetics. Male Wistar-derived 8-week-old rats were allotted to a control group, a lansoprazole administration group, an enprostil administration group, and a lansoprazole + enprostil administration group. Lansoprazole (30 mg/kg once a day for 4 weeks) and enprostil (10 micrograms/kg twice a day for 4 weeks) were administered into the gastric lumen with a gastric tube. At this time, blood was collected and immunohistological staining of gastric endocrine cells was conducted to investigate the secretory kinetics. Lansoprazole administration induced hypergastrinemia, increase of gastrin cells, and increase of enterochromaffin-like cells. Enprostil administration induced increase of somatostatin cells. The group administered lansoprazole + enprostil exhibited significant decreases in serum gastrin level, total gastrin cell count, and total enterochromaffin-like cell count, compared with the group administered lansoprazole alone. These findings suggest that enprostil may ameliorate the alteration in gastric endocrine secretion produced by the chronic administration of lansoprazole.

Prostaglandin E2 derivative, enprostil, reduces hepatic triacylglycerol content in Japanese quail.

This study investigated the effect of prostaglandin E2 derivative, enprostil, on the hepatic lipid metabolism of Japanese quail. In experiment 1, the birds received 0.1, 1, 10 or 100 micrograms enprostil/kg body weight intraperitoneally, and they were killed after 10, 30, 60 and 120 min. After 60 min, hepatic triacylglycerol content was reduced most effectively by 10 micrograms enprostil/kg body weight. In experiment 2, effect of enprostil (10 micrograms/kg body weight) on liver enzyme activities associated with lipid metabolism and plasma lipid concentrations was examined. There were no significant differences in malic enzyme and carnitine palmitoyltransferase activities. Peroxisomal beta-oxidation in birds which received enprostil was significantly lower than that in birds receiving the vehicle. Enprostil caused a significant decrease in plasma triacylglycerol and non-esterified fatty acid concentration. We conclude that the administration of enprostil reduces hepatic triacylglycerol content through the decreased influx of non-esterified fatty acid.

[A basic study on gastric emptying test using barium grains and effect of drugs on gastric emptying in rats].

We reported an experimental study on a new non-invasive method for evaluation of gastric emptying by abdominal X-ray after administration of radiopaque barium grains. Adult male Wistar rats weighing around 200 g were used. After they were fasted for 24 hours, 1 ml of gruel mixed with 10 barium grains (1 mm diameter) was introduced into the rat stomach with a catheter. The rats were sacrificed at 30, 60, 90, 120 and 150 minutes after the introduction of the gruel. X-rays were taken at each time point and grains in the stomach were counted in the X-ray photographs. All barium grains were emptied from the stomach in 150 minutes. After incision of the abdomen, the residual gastric contents were weighed. A positive correlation was found between the grains in the stomach and the weight of the contents. We studied the effects of cisapride, scopolamine buthylbromide and enprostil on the gastric emptying time by this method. Cisapride accelerated gastric emptying, whereas scopolamine buthylbromide delayed it. A prostaglandin E2 analog, enprostil delayed the gastric emptying. This method was found to be a simple procedure which is outstanding for quantitative determination and useful in evaluating gastric emptying functions.

Weakness of mucosal barrier in portal hypertensive gastropathy of alcoholic cirrhosis. Effects of propranolol and enprostil.

BACKGROUND/AIMS: It has been suggested that the vulnerability of gastric mucosa is increased in patients with cirrhosis as a result of a PGE2 deficiency. Therefore, we evaluated whether PGE2 mucosal generation, and gastric potential difference - a reflection of the gastric mucosal barrier - were correlated to endoscopic features and whether these alterations could be alleviated. METHODS: The potential difference was measured before (basal) and after a stimulation test by aspirin. The serum levels of gastrin and glucagon were also determined. Finally, the effects of a 1-week administration of propranolol or enprostil were tested on potential difference. The endoscopic grade of portal hypertensive gastropathy was assessed according to McCormack et al. The results are presented respectively for controls, patients with mild gastropathy, and patients with severe gastropathy. Comparisons were made using variance or covariance analysis after adjustment with age. RESULTS: Basal potential difference was significantly different between the three groups: -30.6, -28.8, -24.9 mV, p <0.05, respectively. The effects of aspirin administration on potential difference parameters were significantly different between the three groups (irritability index: 35 +/- 25, 92 +/- 98, 114 +/- 74 mV2.min, p <0.05, respectively) when non-responders to aspirin were excluded. PGE2 mucosal generation was significantly increased in both the antrum (9.8, 19.5, 19.7 ng/mg proteins, p<0.05, respectively) and in the corpus (8.1, 14.0, 20.2 ng/mg proteins, p<0.05, respectively). PGE2 generation was not related to potential difference. Glucagon serum levels were related to the grade of gastropathy. A 1-week administration of 160 mg/d long-acting propranolol, 35 micro g/d enprostil or placebo did not significantly modify basal potential difference. CONCLUSIONS: Portal hypertensive gastropathy is characterized by a decreased potential difference proportional to the endoscopic severity. The gastric mucosa of patients with cirrhosis seems to be more susceptible to aspirin than that of healthy subjects. It appears that the role of PGE2 is controversial in portal hypertensive gastropathy. Propranolol and enprostil do not improve this decreased potential difference.

A comparison of two prostaglandin analogues (enprostil vs misoprostol) in the treatment of acute duodenal ulcer disease.

We conducted a clinical trial to compare the efficacy and safety profile of two prostaglandin analogues, enprostil (35 micrograms twice daily) and misoprostol (200 micrograms four times daily) in the treatment of acute duodenal ulcers in 214 patients. The two agents healed approximately 80% and in excess of 90% of duodenal ulcers after 4 and 6 weeks' therapy, respectively. There was a significantly lower ulcer healing rate in both treatment groups in smokers compared with non-smokers (P < 0.05). However, daytime and nighttime ulcer pain relief was achieved in fewer than 50% of patients by either agent. Diarrhea, which occurred in more than 40% of patients, was the predominant side effect, and occurred mainly during the first 2 weeks of therapy with either agent. Nevertheless, this side effect was mild and self-limiting in the majority of patients. Both agents were found to be safe and well tolerated by the majority of patients. We conclude that these prostaglandin analogues are safe and effective duodenal ulcer healing agents. Furthermore, there was very little difference between enprostil and misoprostol. The limiting factors, however, for their routine use as ulcer healing agents are their low efficacy with regard to ulcer pain relief and the high incidence of diarrhea.

Copresence of prostaglandin EP2 and EP3 receptors on gastric enterochromaffin-like cell carcinoid in African rodents.

BACKGROUND & AIMS: Prostaglandins (PGs) have important roles in the regulation of gastric acid secretion. The aim of this study was to examine the possible presence of PG receptors on the gastric enterochromaffin-like (ECL) carcinoid of Mastomys natalensis, which might be a useful model of normal ECL cells. METHODS: A [3H]PGE2 binding experiment was performed by using the ECL tumor membrane, and intracellular signal transduction was studied in the cells. In addition, Northern blot analysis using EP2 and EP3 receptor complementary DNAs was conducted. RESULTS: [3H]PGE2 specifically bound to the tumor cell membrane, and the binding was displaced by various PGs with a potency order of PGE1 = PGE2 > enprostil > PGF2 alpha. Although PGE1 and PGE2 stimulated 5'-cyclic adenosine monophosphate (cAMP) production, neither PGF2 alpha nor enprostil had any effect. On the other hand, all of PGE1, PGE2, PGF2 alpha, and enprostil attenuated the forskolin-induced cAMP production. Moreover, enprostil inhibited histamine release induced by forskolin. However, on pertussis toxin treatment, PGE2 paradoxically enhanced the forskolin-induced increase of cAMP production. Finally, the presence of EP2 and EP3 receptor messenger RNAs was confirmed by RNA blot analysis. CONCLUSIONS: The ECL carcinoid tumor cells of Mastomys seem to possess two subtypes of PGE receptor: EP2 linked to cAMP production and EP3 coupled with inhibitory guanosine 5'-triphosphate-binding proteins mediating the inhibition of cAMP production.

Receptor subtypes involved in dual effects induced by prostaglandin E2 in circular smooth muscle from dog colon.

Smooth muscle strips and isolated muscle cells from the circular layer of dog colon, were used to study the effect of prostaglandin E2 (PGE2) and their analogs acting at EP receptors: Iloprost (IP/EP1), butaprost (EP2) and enprostil (EP3) and SC19220 (antagonist EP1) to characterize the EP-receptors involved in the control of muscle function. In strips treated with tetrodotoxin, only enprostil provoked a concentration-dependent contraction. The concentration of enprostil inducing a half maximal contraction (EC50) was 400 nM and the maximal effect was obtained at 1 microM. PGE2, butaprost and iloprost induced a dose-dependent relaxation, with an EC50 of 200, 80 and 200 nM, respectively. The maximal relaxation was obtained at 1 microM for all these agents. When the EP1 antagonist, SC19220 (10 microM), was added 20 min before PGE2 or their analogs, their respective concentration-response curves were not affected. In isolated cells, PGE2 and enprostil induced a cell contraction in a concentration-dependent manner, whereas iloprost and butaprost had no effect by themselves. The maximal contraction was 241.1 +/- 1.7% at 10 nM PGE2 and 22.5 +/- 1.6% at 10 nM enprostil. EC50 of PGE2 and enprostil was 40 pM. SC 19220, at concentrations ranging from 1 pM to 0.1 microM, failed to inhibit the contraction induced by either PGE2 or enprostil. When cells were preincubated for 1 min with butaprost or iloprost at concentrations ranging from 1 pM to 1 microM, the contraction induced by CCK8 (10nM) was inhibited in a concentration-dependent matter.(ABSTRACT TRUNCATED AT 250 WORDS)

Cholinergic stimulatory effect of intragastric administration of a prostaglandin E2 analogue on pancreatic exocrine secretion in conscious rats.

The effect of a long-acting, potent synthetic analogue of prostaglandin E2, enprostil, on pancreatic exocrine secretion was examined in conscious rats. Rats were prepared with cannulae draining bile and pancreatic juice separately. Pancreatic exocrine secretion was increased by intragastric administration of enprostil but inhibited by its intravenous administration. The pancreatic response to intragastric administration of enprostil was not inhibited by the administration of cholecystokinin antagonist or secretin antibody, or by bilateral vagotomy, but was completely abolished by atropine. Therefore, intragastric administration of enprostil seemed to stimulate pancreatic exocrine secretion via a peripheral gastro-(entero)-pancreatic reflex.

Effect of a synthetic prostaglandin E2 analogue, RS-86505-007, on plasma lipids and lipoproteins in patients with moderate hypercholesterolaemia: efficacy and tolerance of treatment and response in different apolipoprotein polymorphism groups.

A double-blind, placebo-controlled ascending dose trial was carried out to evaluate the hypocholesterolaemic efficacy and tolerance of RS-86505-007, a prostaglandin E2 analogue, in moderately hypercholesterolaemic patients. Twenty-four patients received an oral dose of RS-86505-007 3 micrograms t.i.d. and a separate group of 26 patients 6 micrograms t.i.d. for 6 weeks. Plasma total and low-density lipoprotein (LDL) cholesterol concentrations decreased after 2 weeks of treatment, and the reductions were dose dependent. After 6 weeks of treatment (6 micrograms t.i.d.), the reductions from baseline in total and LDL cholesterol concentration were 14.6% and 18.5%, respectively. No changes in the plasma concentration of triglycerides or high-density lipoprotein (HDL) cholesterol were observed. RS-86505-007 tended to reduce total and LDL cholesterol concentrations less in patients with the epsilon 4 allele of the apolipoprotein E than in those with epsilon 3 allele. In contrast, the XbaI or EcoRI polymorphisms of the apolipoprotein B gene seemed to have no effect on the hypocholesterolaemic efficacy of the drug. The drug had no effect on the lipoprotein (a) concentration. Sixty-three percent of patients receiving 3 micrograms t.i.d. and 81% receiving 6 micrograms t.i.d. had adverse events, two-thirds of which related to the gastrointestinal tract. One patient in the 3-micrograms group and three patients in the 6-micrograms group terminated the study prematurely due to adverse effects.(ABSTRACT TRUNCATED AT 250 WORDS)