Treatment of advanced colorectal and gastric adenocarcinomas with 5-fluorouracil and high-dose folinic acid.

We report the results of an expanded trial of 5-fluorouracil (5-FU) combined with high-dose folinic acid for treatment of patients with advanced colorectal or advanced gastric adenocarcinoma. In each treatment course, the patients received both 5-FU (340 to 400 mg/m2/d by intravenous (IV) infusion for a period of 15 minutes) and folinic acid (200 mg/m2/d by IV bolus) for 5 consecutive days, with a 21-day interval between courses. Eighty-six patients with colorectal carcinoma were evaluated. The combined complete and partial response rates were 39% for 54 patients who did not receive prior chemotherapy and 22% for 32 patients who had previously received chemotherapy. Four patients who were previously resistant to 5-FU attained objective responses. The median time to disease progression for the 28 responders was 10 months. The median survival time of responders was 19.5 months, and the probability of their being alive at 2 years was 40%. Of 27 patients with gastric adenocarcinoma, 13 (48%) responded to therapy. Their median time to disease progression was 5.5 months. The median survival time of responders was 11 months, and their probability of being alive at 15 months was 30%. Toxicity was within acceptable limits. Toxic effects included stomatitis, diarrhea, conjunctivitis, skin rash, and mild myeloid hypoplasia. In a separate study, plasma concentrations of L-folates greater than 10(-5) mol/L were achieved after a rapid single IV injection of 200 mg/m2 of folinic acid. Comparisons of our results with those reported in previous studies on 5-FU administered as a single agent suggest that, in advanced colorectal and gastric adenocarcinoma, folinic acid administered in high doses enhances the effectiveness of 5-FU administered concomitantly. Furthermore, some colorectal tumors that were previously resistant to 5-FU become sensitive to this drug. The survival of the patients who responded to therapy was markedly improved over that observed in reported series of untreated patients with advanced colorectal and gastric adenocarcinomas.

Pharmacokinetic study of cefmenoxime (SCE 1365-CMX) in healthy adults.

Cefmenoxime pharmacokinetics were investigated in six healthy volunteers after intravenous and intramuscular administration of 0.5, 1, and 2 g. Blood and urine samples were analyzed by reversed-phase high-pressure liquid chromatography using ultraviolet detection at 275 nm. The assay is precise and linear up to 200 micrograms/ml-1, with 0.02 micrograms/ml-1 as the limit of detection. Linearity of cefmenoxime kinetics was demonstrated because the area under the plasma concentrations is proportional to studied doses. Eight hours after 1 g of cefmenoxime intramuscularly, mean plasma concentrations are, respectively, 0.6 +/- 0.1 and 0.3 +/- 0.1 microgram/ml-1. Intramuscular cefmenoxime is rapidly absorbed (Ka = 7.28 hours-1) with complete bioavailability (F = 0.99); apparent volume of distribution is 0.35 liters/kg-1 and elimination half-life 1.5 hours. The fraction of cefmenoxime excreted unchanged in the urine after intramuscular administration is 0.72, indicating a major contribution of renal clearance in total clearance. Experimental data after intramuscular administration were well fitted with a two-compartment model.

Pharmacokinetics of rilmenidine in healthy subjects.

Rilmenidine is a novel alpha 2-adrenoceptor agonist, used in the treatment of mild or moderate hypertension at the oral dose of 1 mg once or twice daily. The pharmacokinetic parameters were investigated after single or repeated administration in healthy subjects, using labeled and unlabeled compounds. Rilmenidine was rapidly and extensively absorbed, with an absolute bioavailability factor close to 1 and a maximal plasma concentration achieved within 2 hours. Rilmenidine was not subject to presystemic metabolism. Distribution was independent of the free fraction because rilmenidine was weakly bound to plasma proteins (less than 10%). The volume of distribution was approximately 5 l.kg-1 (315 liters). Elimination was rapid with a total body plasma clearance of approximately 450 ml.min-1 and an elimination half-life of approximately 8 hours. Renal excretion was the major elimination process (two-thirds of the total clearance). Metabolism was very poor, with a renal elimination of rilmenidine as the parent drug (urinary fraction of rilmenidine was about 65% and no metabolite plasma levels were detected). Linear pharmacokinetics were demonstrated for rilmenidine from 0.5 to 2 mg but, at 3 mg, a slight deviation from linearity was observed. In repeated administration, the linear disposition of rilmenidine with dose was confirmed.

Diffusion of ofloxacin into human lung tissue.

Ofloxacin 200mg twice daily was administered to 17 patients with pulmonary disorders, which necessitated surgery, during the preceding 48 hours and 200mg was administered 1 hour before the operation. During surgery, blood samples and specimens of healthy and diseased lung tissues were taken simultaneously. Ofloxacin levels were determined by HPLC. The mean values of the tissue concentration/plasma concentration ratio were 3.5 +/- 0.4 for the healthy tissue and 3.9 +/- 0.4 for the diseased tissue. These values reflected good penetration of ofloxacin into both healthy and atelectasic pulmonary parenchyma.

Effect of moderate or severe hepatic impairment on clarithromycin pharmacokinetics.

The pharmacokinetic and safety profiles of clarithromycin (C) and its 14-hydroxy-clarithromycin (HC) metabolite were determined after a multiple-dose oral clarithromycin regimen (250 mg twice daily for five doses) in six healthy subjects and seven patients with moderate or severe hepatic impairment (Pugh grades B and C). Plasma and urine C and HC concentrations were determined using high-performance liquid chromatography. Hepatic impairment resulted in increased harmonic mean C terminal disposition half-life and mean +/- SD C renal clearance (CLR) compared with normal volunteers (5.0 vs. 3.3 hr and 170 +/- 69 vs. 111 +/- 17 mL/min, respectively). Hepatic impairment also resulted in decreased metabolite peak plasma concentration and area under the plasma concentration-versus-time curve and decreased metabolite/parent concentration ratios compared with normal volunteers. These data suggest that 14-hydroxylation of C was reduced by moderate to severe hepatic impairment. No adverse events were noted in either study group and there were no study-related clinically significant changes in laboratory parameters. The decrease in C metabolic clearance appears to be partially offset by an increase in C CLR, resulting in comparable steady-state concentrations of parent drug. In those indications in which the metabolite may be a necessary element of the antimicrobial activity of C, it would seem prudent to be cautious in using C in patients with moderate to severe hepatic impairment due to reduced production of HC. Otherwise, no dosage adjustment for C appears necessary for subjects with moderate or severe hepatic impairment provided that renal function is not impaired.

Phase I study of percutaneous 4-hydroxy-tamoxifen with analyses of 4-hydroxy-tamoxifen concentrations in breast cancer and normal breast tissue.

4-OH-tamoxifen is an active metabolite of tamoxifen that is detectable in the serum and tumour tissue of patients treated by oral tamoxifen. As this metabolite penetrates through the skin, it is possible to compare percutaneous 4-OH-tamoxifen (4-OH-TAM) and oral tamoxifen treatments. We report herein a randomized study of percutaneous 4-OH-TAM versus oral tamoxifen in women with breast cancer. This pharmacology study was designed to compare the 4-OH-TAM concentration in breast cancer and normal breast tissue according to the route and dose used for administration of tamoxifen after a 3-week period prior to surgery and tissue sampling. Women were randomized into one of the five following groups: group I, oral tamoxifen given at 10 mg twice a day; group II, 4-OH-TAM delivered percutaneously at 0.5 mg day to both breast areas; group III, 4-OH-TAM applied percutaneously at 1 mg/day to both breast areas; group IV, 4-OH-TAM delivered percutaneously at 1 mg/day to a large cutaneous area excluding the breasts; and group V, 4-OH-TAM applied percutaneously at 2 mg/day to a large skin area excluding the breasts. 4-OH-TAM plasma and tissue concentrations were significantly higher in the oral tamoxifen group as compared with either the high- or the low-dose percutaneous 4-OH-TAM group. In group II, percutaneous 4-OH-TAM treatment resulted in tissue concentrations of 1,446 and 352 pg/g in tumour tissue and normal breast tissue, respectively. In group I these concentrations were as follows: tumour tissue, 12, 453 pg/g; and normal tissue, 10,214 pg/g. 4-OH-TAM concentrations in tumour tissue and normal breast tissue did not significantly differ in any group. In the oral group we observed classic effects on coagulation and lipid metabolism when pre- and post-treatment values of these biological variables were compared, whereas no difference was observed in the percutaneous group. Although percutaneous administration of 4-OH-TAM led to a low plasmatic concentration of this active metabolite, the breast tissue concentration remained lower than those observed after oral tamoxifen treatment. Therefore, at the doses described in this study, percutaneous 4-OH-TAM cannot be proposed as an alternative tamoxifen treatment.

Effect of ponsinomycin on the pharmacokinetics of dihydroergotamine administered orally.

Ponsinomycin is a new macrolide antibiotic. Its effect on DHE pharmacokinetics was investigated in this study. Twelve young healthy volunteers received a single 9 mg oral dose of DHE before and on the 8th day of treatment (800 mg twice daily) with ponsinomycin. DHE was assayed in plasma by RIA. Because of low plasma levels, only peak concentrations could be accurately compared for a ponsinomycin effect. We observed a 3-40-fold increase in maximum DHE plasma levels in the majority of cases and a much more important effect on one occasion, when DHE was administered in the presence of ponsinomycin. These data are consistent with an increase of DHE bioavailability in the presence of ponsinomycin, probably related to a reduction of its first-pass elimination. This pharmacokinetic interaction is likely to have clinical consequences and administration of ponsinomycin should be avoided in patients treated orally with DHE.

Fluorometric high-performance liquid chromatographic determination of sotalol in biological fluids.

A high-performance liquid chromatographic (HPLC) method for sotalol in biological fluids is presented. Sample preparation involves a simple acid-base double-extraction procedure, and analysis is carried out on a reversed-phase chromatographic system using a muBondapak C18 column. Procainamide was used as the internal standard. As little as 20 ng of sotalol/ml can be detected with good precision by a fluorometric detector. HPLC, which allows more rapid analysis of plasma sotalol levels than does GLC, appears suitable for pharmacokinetic studies as well as for therapeutic drug monitoring.

Mass fragmentographic determination of timolol in human plasma and urine.

A mass fragmentographic procedure was developed for measuring quantities of less than 1.0 ng of timolol/ml of plasma or urine. The lower limit of sensitivity was 0.5 ng of timolol maleate/ml of plasma. The unchanged drug was extracted into heptane--4% isopentyl alcohol from alkalinized plasma or urine, together with propranolol hydrochloride as the internal standard. The compounds were subsequently back-extracted into 0.1 N HCl and then into chloroform following adjustment of the acidic phase to an alkaline pH. The chloroform layer was evaporated to dryness, and the compounds were derivatized with N-methyl-N-trimethylsilyltrifluoroacetamide-acetonitrile to form the trimethylsilyl derivatives; these derivatives were quantitated by mass fragmentography. Recovery of timolol added to normal plasma and urine was quantitative and reproducible, and no interfering substances were observed in normal biological samples. After a 20-mg oral dose of timolol maleate, plasma levels of approximately 3.0 ng/ml were observed at 12 hr.

Serum levels and urinary excretion of mequitazine after a single oral dose.

Pharmacokinetics of mequitazine, a recently introduced peripheral H1-histamine receptor antagonist of phenothiazine type, was followed up to 72 h after the single oral dose of 5 mg of the drug to eight fasted healthy volunteers. Each subject was treated thrice with a dosing interval of 15 days or more. Thus all the results were triplicated. Serum mequitazine was measured by mass fragmentography using a gas-liquid chromatograph/mass spectrometer set in the electron impact mode. Urine phenothiazines were determined fluorometrically before and after cleaving phenothiazines from their glucuronide conjugates. Peak concentration of mequitazine in serum was 3.19 +/- 1.70 (s.d.) ng.ml-1, time to peak concentration 5.67 +/- 1.68 h, elimination half-life 45 +/- 26 h, and elimination rate constant 0.018 +/- 0.007 h-1. Only 10.9 +/- 3.3% of the dose appeared in urine in unconjugated plus the glucuronidated form during the first 72 h. About 46% of the urinary phenothiazines were glucuronide conjugates. The results suggested that after the oral administration only low mequitazine concentrations appeared in serum, most of the drug seemed to be deactivated by the extrarenal route, and the kinetic properties of the drug resembled those of several phenothiazines used for psychiatric therapy.

Kinetic profiles of tianeptine and its MC5 metabolite in plasma, blood and brain after single and chronic intraperitoneal administration in the rat.

This study was conducted to determine kinetic profiles of the antidepressant tianeptine and its main metabolite (named MC5) in plasma, blood and brain after single (10 mg.kg-1) and chronic (10 mg.kg-1.day-1 and 2 x 10 mg.kg-1.day-1 for 15 days) intraperitoneal administration in Wistar rats. In plasma, tianeptine and MC5 reached a peak very quickly (5 and 15 min post administration respectively). Following the peak, disappearance of tianeptine was faster than that of MC5 with terminal half-lives close to 2.5 and 6.5 h respectively. During chronic treatment, accumulation of tianeptine was negligible, that of MC5 was slight as indicated by accumulation ratios equal to 1.28 (once daily administration) and 1.73 (twice daily administration). Tianeptine and MC5 rapidly appeared in brain tissue but at concentrations less than those observed in plasma. Peak concentrations of tianeptine in brain were higher than those of MC5. A very low penetration of tianeptine and MC5 into erythrocytes was observed.

Pharmacokinetics and bioavailability of viqualine, a new antidepressant.

A high performance liquid chromatographic assay was developed for plasma and urine levels measurement of viqualine. The assay was used to study the disposition of 25 mg intravenous and oral single doses in five healthy subjects. Two exponential terms were required to describe the disposition of the drug after intravenous and oral administration. The bioavailability of oral viqualine averaged 80%. The mean apparent half-life was 12.1 +/- 1.9 and 11.9 +/- 1.4 h (mean +/- s.e.m., n = 5) after 25 mg I.V. and oral dose respectively. The apparent volume of distribution (Vdss) were 1578 +/- 132 1 (after I.V. administration) and 1572 +/- 201 1 (after oral administration). The body and renal clearances were respectively 1.56 +/- 0.31 1.h-1. kg-1 and 1.57 +/- 0.31 1.h-1. kg-1, after 25 mg bolus I.V.

Cross-over study of the pharmacokinetics of cefonicid administered intravenously or intramuscularly to healthy adult humans.

The pharmacokinetics of cefonicid were investigated in eight healthy adults. A one-gram dose was administered either intramuscularly or intravenously in a cross-over design study. Mean peak cefonicid plasma concentrations of 186 to 204 mcg/ml and 88 to 123 mcg/ml were achieved after intravenous and intramuscular injection, respectively, with elimination half-lives of 4.9 h and 5.3 h. Cefonicid concentrations were measured by both microbiological (M.A.) and high-performance liquid chromatography (HPLC) assays. Results were quite similar with the two techniques, except for the urinary recovery of cefonicid in the first 24 hours (83% of the dose with MA - vs 53% with HPLC method). The apparent volume of distribution (Vd area) was 0.18 1/kg; the total body clearance (CT) and the renal clearance (CR) were 24-26 ml/min and 15-19 ml/min, respectively. The kinetic data of cefonicid were not significantly different for the two routes of administration. A one-gram i.v. or i.m. cefonicid dose produced high and prolonged plasma concentrations with a longer half-life than obtained with commonly used cephalosporins.

Pharmacokinetics of prazosin administered as gastro-intestinal-therapeutic-systems to 24 healthy volunteers.

Pharmacokinetics of prazosin was investigated after single dose administrations of four different formulations, according to a randomized trial, to 24 young healthy volunteers: immediate release tablets (3 x 1 mg) used as reference (Treatment A), two new Gastro-Intestinal-Therapeutic-Systems (GITS) containing 2.5 mg (Treatment B) and 5 mg (Treatment C) of prazosin, and a traditional sustained release formulation (4 mg, treatment D). Relative bioavailability of prazosin administered as GITS was only 49.4 +/- 19.5% (B) and 45.5 +/- 18.7 (C) versus 73.8 +/- 13.9% (D) (area under the curve normalized at 3 mg dosing); but absorption was sustained and plasma concentrations were maintained at a virtually constant level for a time period close to 24 h. As a result, mean residence time (MRT) of prazosin was considerably increased after GITS administration: 21.6 +/- 1.0 h (B), 22.5 +/- 1.6 h (C) instead of 5.9 +/- 0.2 h for the reference formulation (A) and 10.8 +/- 0.8 h for the traditional sustained release formulation (D). Although extrapolation to multiple dosing situations is difficult, this study demonstrates the potential suitability of prazosin GITS for once daily administrations.

[Pharmacokinetics of sulindac in aged patients presenting with inflammatory joint disease]. / Pharmacocinétique du sulindac chez des patients âgés présentant une pathologie articulaire inflammatoire.

The kinetic of sulindac and its two metabolites (sulfide and sulfone) was investigated in twelve elderly patients, following multiple oral dose administration of 400 mg/d. Data were compared to those obtained previously in ten healthy volunteers who received the same dosage regimen. Following multiple dose administration, accumulation ratios indicate that sulindac do not accumulate either in elderly patients (R = 1.35; R = AUC0-24 J8/AUC0-24 J1) or in healthy young subjects (R = 1.38; R = U0-24 J1). No significant modification of sulindac and sulfide kinetic parameters was observed. The apparent bioavailability of the inactive metabolite, sulfone, was found to be doubled in elderly patients (p less than 0.05). We conclude that there is no need to modify the dosage regimen of Arthrocine (400 mg once a day) in elderly patients.

The kinetic profiles of amlodipine and of a sustained release form of diltiazem.

This study in normotensive subjects compared plasma concentrations of amlodipine (5 mg) and of a sustained release form of diltiazem (300 mg) after single and multiple oral dosings of the two drugs. As a consequence of the galenic form of administered formulations, plasma concentration of diltiazem versus time curves exhibited two peaks corresponding to fast and slow releases of diltiazem. Conversely, the curves of amlodipine plasma concentration depicted only one peak. There was less variability in plasma concentrations and in pharmacokinetics with amlodipine than with diltiazem after both single and multiple oral dosings of the two drugs. These results suggested that amlodipine displayed less variability in blood pressure response at steady-state. The rate of decrease in plasma levels of diltiazem between 24 and 48 hours post-dose was higher than that of amlodipine. So, even after a missed dose, there is only a small decline in plasma concentrations of amlodipine and therefore it suggests a small repercussion on the blood pressure attenuation.

[Pharmacokinetics of ponsinomycine with repeated ingestion in healthy volunteers]. / Pharmacocinétique de la ponsinomycine administrée en prises répétées chez le volontaire sain.

Ponsinomycin or miocamycin (MOM) is a new macrolide which is totally metabolized in vivo. The disposition of its 3 major metabolites (Mb12, Mb6 and Mb9a), was investigated following multiple dosing with ponsinomycin at a dose of 800 mg every 12 h, for 8 days, in healthy volunteers. Drug measurements were conducted by high performance liquid chromatography. In agreement with the low values of their apparent elimination half-lives, respectively less than 1.5 h and 3.0 h, metabolites Mb12 and Mb9a did not accumulate with time. Their pharmacokinetics was apparently stable with time. Conversely Mb6 did accumulate, by approximatively a factor 2, although its apparent elimination half-life was only close to 2 h. This value must therefore be considered with caution. A dose dependency effect was previously observed, Mb6 pharmacokinetics could be non linear with time as well. The relative importance of this metabolite is therefore greater at steady-state, following multiple administration than after single dosing with ponsinomycin.

[Pharmacokinetic properties of Bilobalide and Ginkgolides A and B in healthy subjects after intravenous and oral administration of Ginkgo biloba extract (EGb 761)]. / Propriétés pharmacocinétiques du Bilobalide et des Ginkgolides A et B chez le sujet sain après administrations intraveineuses et orales d'extrait de Ginkgo biloba (EGb 761).

The pharmacokinetics of Ginkgolide A, Ginkgolide B and Bilobalide, which are compounds extracted from the dried leaves of the Ginkgo biloba tree, were investigated in 12 young healthy volunteers (six men and six women; mean +/- SD age = 25 +/- 5 years) after single-dose administration of Ginkgo biloba extract. Subjects were given, on three occasions, Ginkgo biloba extract as a solution either orally (in fasting conditions and after a standard meal) or intravenously; corresponding to single doses of Ginkgolide A, Ginkgolide B and Bilobalide ranging from 0.90 mg to 3.36 mg. After each dosing, blood and urine samples were collected for up to 36 h and 48 h, for measurements of Ginkgolide A, Ginkgolide B and Bilobalide. Plasma and urine concentrations of these compounds were quantitatively measured by gas chromatography/mass spectrometry using negative chemical ionization, by applying a very sensitive method which allowed plasma concentrations as low as 0.2 ng/ml of each compound to be measured. When given orally, while fasting, the extents of bioavailability are high, as shown by bioavailability coefficients (FAUC) mean (+/- SD) values equal to 0.80 (+/- 0.09), 0.88 (+/- 0.21) and 0.79 (+/- 0.30) for Ginkgolide A, Ginkgolide B and Bilobalide respectively. Food intake does not change AUC quantitatively but increases Tmax. For the three compounds of interest, after oral dosing while fasting, differences can be noted for the elimination half-lives (T1/2Z), which exhibit mean values equal to 4.50, 10.57 and 3.21 h, as well as mean residence times (MRT), equal to 5.86, 11.25 and 4.89 h, for Ginkgolide A, Ginkgolide B and Bilobalide respectively.(ABSTRACT TRUNCATED AT 250 WORDS)

[Pharmacokinetic behavior of pefloxacin in man]. / Comportement pharmacocinétique de la péfloxacine chez l'homme.

The pharmacokinetic profile of pefloxacin is one of the major assets of this new antibiotic of the quinolone family. Digestive absorption is rapid and complete after oral administration. The pharmacokinetic pattern is the same with both routes. The high apparent volume of distribution (AVD = 117 +/- 6 litres) reflects good diffusion in extravascular compartments. Pefloxacin can therefore be used not only for the treatment of systemic infections, but also for that of extravascular infections, whatever the perfusion rate of the infected organ. The predominant route of excretion of pefloxacin is extrarenal, after hepatic degradation; it is responsible for prolonged elimination half-life in patients with hepatic insufficiency. Since renal excretion of the unchanged drug is less important, there is no need to modify the dosage in case of renal impairment. When pefloxacin is eliminated by biotransformation, it is excreted slowly, and plasma or tissue levels of its unchanged form remain much higher. In subjects with normal liver function, the mean elimination half-life is almost 12 h, which makes it possible to administer pefloxacin twice a day. During treatment with 400 mg doses, the steady state is reached within 48 h; mean peak plasma concentrations then are 10 micrograms X ml-1 and mean trough concentrations 4 micrograms X ml-1. The pharmacokinetic properties of pefloxacin give this antibiotic an unquestionable advantage over third-generation cephalosporins and aminoglycosides.

[Crossover pharmacokinetic study of quinupramine in 6 subjects]. / Etude pharmacocinétique de la quinupramine en cross over - chez six sujets.

The pharmacokinetic study of quinupramin was developed on healthy volunteers, after cross-over administrations of three doses (2.5 and 7.5 mg P.O. and a 2.5 mg intravenous bolus dose). A specific and sensitive Gas chromatographymass spectrometry (GC-MS) method was applied to the determination of Quinupramin in plasma and urine. The calculated kinetic data for Quinupramin indicated a good bioavailability for the oral doses (2.5 and 7.5 mg). The coefficient of bioavailability (F) was found to be (0.75), with small individual variations. After oral administration, the area under experimental points of the curves (AUC 0 leads to infinity): 43.7 +/- 10.2 (S.E.M.) and 132.5 +/- 8.3 ng.ml-1.h, respectively, were proportional to doses. Blood levels (CMAX) were about 1.1 and 3.4 ng.ml-1 at 3 hours, for oral doses of 2.5 and 7.5 mg. The pharmacokinetic of Quinupramin was not dose-dependent. The high value of the apparent volume of distribution (A.V.D.) 34.2 +/- 4.8 l.kg-1 after bolus I.V., indicated a very high distribution of Quinupramin, as for all anti-depressants in a deep compartment (CNS). The apparent terminal half life of elimination was very high: 33.5 h (+/- 2.7 S.E.M.) for 2.5 mg bolus I.V. and 28.8 h (+/- 5.01 S.E.M.), 33.1 (+/- 3.2 S.E.M.) for 2.5 mg and 7.5 mg oral doses respectively. The lack of metabolism for Quinupramin, demonstrated by many authors, explained the absence of interindividual variations of bioavailability. The renal elimination (CLr) after bolus I.V. was 2.6% of the dose, after 72 h. Many variations of plasma concentrations illustrated a predominant biliary excretion. The decline in plasma Quinupramin concentrations (Ct) was fitted to a bi-exponential equation, using least-square regression analysis, for the three formulations.