Received 14 February 2002Revised 28 March 2002

Copyright # 2002 John Wiley & Sons, Ltd. Accepted 2 April 2002

BIOPHARMACEUTICS & DRUG DISPOSITIONBiopharm. Drug Dispos. 23: 227–231 (2002)

Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/bdd.317

Effect of Dose and Input Rate on the Brain Penetration ofBMS-204352 following Intravenous Administration to Rats

Rajesh Krishna*, Holly Palme, Jianing Zeng and Nuggehally SrinivasClinical Discovery, Bristol-Myers Squibb Pharmaceutical Research Institute, Princeton, NJ 08543-4000, USA

ABSTRACT: BMS-204352 is a novel maxi-K channel opener that is being developed for thetreatment for stroke. The current study was designed to evaluate the plasma and brainpharmacokinetics of BMS-204352 in rats, in particular, assessing the effect of dose and input rateon brain penetration of BMS-204352. Rats (3 animals/group/time point) received a singleintravenous dose of BMS-204352 as 5mg/kg bolus, 5mg/kg 30min infusion, 5mg/kg 60mininfusion, and 10mg/kg bolus dose, into the jugular vein. Terminal blood (for plasma) and brainsamples were collected for up to 9 h post-dose and samples were analyzed for the concentrations ofintact BMS-204352 using a validated liquid chromatographic tandem mass spectrometric method(LC/MS/MS). As dose increased from 5 to 10mg/kg, both BMS-204352 Cmax and AUC valuesincreased in plasma and brain, somewhat greater in proportion to the increment in dose. Whereasthe peak concentrations of BMS-204352 were affected by infusion time, overall AUCs werecomparable across the bolus and infusion groups. Terminal disposition (T-half ranged from 1.6 to2.7 h) of BMS-204352 was unaltered as a function of input rate. BMS-204352 crossed the blood–brainbarrier with brain-to-plasma (B/P) ratios of approximately 7–11. Brain-to-plasma ratios appeared tobe independent of dose and infusions produced somewhat higher brain penetration (B/P of ca. 11)as compared to bolus (B/P of ca. 7–8) dose. The decline of BMS-204352 in the brain paralleled thatof plasma independent of the input rate and dose. Copyright # 2002 John Wiley & Sons, Ltd.

Key words: BMS-204352; rats; pharmacokinetics; brain

Introduction

BMS-204352 (MaxiPostTM) is being developedfor the treatment of stroke [1–3]. Chemically,([3S]-[+]-[5-chloro-2-methoxyphenyl]-1,3-dihydro-3-fluoro-6-[trifluoromethyl]-2H-indol-2-one)BMS-204352 is a novel fluorooxindole maxi-Kchannel opener that has been shown to beefficacious in acute forms of stroke in experi-mental animal stroke models [4]. BMS-204352effectively shifted the half-maximal activationvoltage of maxi-K channels in isolated outside-

out membrane patches excised from HEK 293cells, with an estimated EC50 under these condi-tions of 125 ng/g [1].

Several factors favor the permeation of acompound across the blood–brain barrier,namely, high lipophilicity, low degree of ioniza-tion, low molecular size, minimal plasma proteinbinding, non-substrate for drug efflux transpor-ters, and lack of blood–brain barrier integrity.BMS-204352 exhibits several of these desirableproperties; however, its protein binding in hu-man plasma is >99% [5]. Given that the target siteof action for BMS-204352 is the ischemic neuronsof the brain, it is imperative for a neuroprotectivedrug to penetrate the brain at high concentrationsfollowing onset of stroke. Two key issues, then,are: (a) the extent of brain penetration, and

*Correspondence to: Dr. Rajesh Krishna, Clinical discovery,Bristol-Myers Squibb Pharmaceutical Research Institute, Route206 and Province Line Road, Princeton, NJ 08540, USA. E-mail:rajesh.krishna@bms.com

(b) whether brain penetration is dependent ondose and input rate in the context of dosingflexibility, i.e. with regard to intravenous bolusand infusion strategies.

The rat was chosen as an animal model toperform these investigations for several reasons;it is a toxicology species for this compound, thedisposition of unlabeled and radiolabeled BMS-204352 has been investigated in this species, priorinformation is available on the dose proportion-ality of the compound in rats, and feasibility ofobtaining paired plasma and brain concentra-tions [6, 7].

The objectives of the current investigationwere: (a) to assess the plasma pharmacokineticsof BMS-204352, and (b) to assess the effect of doseand input rate on the brain penetration of BMS-204352.

Experimental

Animals

Adult male Sprague Dawley rats (weighing ca.180–220 g). The rats were acclimatized prior tostart of study with appropriate temperature andhumidity control. Each rat was tagged with aunique identification number and housed instainless-steel cages. Animals were provided freeaccess to water and Purina rat chow (RalstonPurina) and weighed prior to drug administra-tion. Prior to experimentation, rats were surgi-cally prepared for placement of jugular venouscatheter for drug administration. Terminal bloodand brain samples were obtained and processedas described below.

Chemicals and formulations

BMS-204352 was obtained from Bristol-MyersSquibb Pharmaceutical Research Institute (Prin-ceton, NJ). A stock solution of BMS-204352was prepared by dissolving the compound in asolvent mixture composed of polyethyleneglycol-300, polysorbate 80, ethanol, and 5%dextrose in water. The dosing solution was sterilefiltered using a 0.22 mm filter prior to drugadministration.

Study design

This was an open, parallel group experimentaldesign. Sprague–Dawley rats (3/group/timepoint) received a single intravenous dose ofBMS-204352 as 5mg/kg bolus, 10mg/kg bolus,5mg/kg 30min constant rate infusion (0.33ml/kg/min), and 5mg/kg 60min constant rateinfusion (0.167ml/kg/min) via the jugular veincatheter (dose volume= 10ml/kg). A 1ml vo-lume of saline was used to flush the catheter. Fordrug administration, an appropriate volume ofthe dosing solution was infused over the speci-fied duration using a calibrated infusion pumpthat employed a 20-gauge Luer stub adapterattached to a syringe.

Sample collection and preparation

Terminal blood (cardiac puncture) and brainsamples from 3 rats/group/time point wereobtained at: 3, 5, 10, and 30min, 1, 2, 4, and 8post-dose (5 and 10mg/kg bolus doses); 15 (mid-infusion), 30 (end of infusion), 35, and 40min, 1,1.5, 2.5, 4.5, and 8.5 h from start of infusion(5mg/kg, 30min infusion); and, 30 (mid-infu-sion), 60 (end of infusion), 65, and 70min, 1.5, 2,3, 5, and 9h from start of infusion (5mg/kg,60min infusion). Blood samples were collected inpotassium EDTA containing vacutainers andprocessed for plasma as follows. Blood sampleswere stored on ice until centrifugation in arefrigerated centrifuge to separate plasma. With-in 30min of blood collection, plasma washarvested by centrifuging at 1000 g for 10min at48C, and then transferred to screw cappedpolypropylene tubes and stored at or below�208C until analysis. Each brain sample wascarefully rinsed with saline, gently blotted dry,and weighed individually. The brain sampleswere frozen at or below �208C until homoge-nized in 4ml acetonitrile. Following homogeni-zation under ice and in acetonitrile, volumeswere measured and recorded. Brain homoge-nates were subjected to centrifugation at 1000 gfor 10min, and 1ml of supernatant was trans-ferred to labeled screw capped polyethylenetubes and stored at or below –208C until analysis.

Copyright # 2002 John Wiley & Sons, Ltd. Biopharm. Drug Dispos. 23: 227–231 (2002)

R. KRISHNA ET AL.228

Analysis of BMS-204352

Concentrations of BMS-204352 in plasma andbrain acetonitrile homogenates were determinedusing a validated LC/MS/MS method that wasaccurate, precise, specific, sensitive and reprodu-cible. Analyses were carried out using a PerkinElmer System fitted with a Series 200 LCautosampler (Perkin Elmer Corporation, Nor-walk, CT), Shimadzu LC-10AD pump (ShimadzuScientific Instruments, Inc., Columbia, MD) and aYMC Basic S5, 2� 50mm� 5 mm column (YMC,Inc., Wilmington, NC), operating with a Quat-troLC tandem mass spectrometer (Micromass,Inc., Beverly, MA), monitoring m/z 358.1 for BMS-204352 and m/z 362 for the internal standard.Selected reaction monitoring included monitor-ing of m/z 358.1 to m/z 338 transitions for BMS-204352 and m/z 362 to m/z 342 transitions for theinternal standard. The MS was operated in anegative ion electrospray mode using nitrogen asa nebulizing gas at a flow rate of 80 l/h and as thedesolvation gas at a flow rate of 1000 l/h. Argonwas used as the collision gas. Sample cone,extractor, and capillary voltages were set at 37, 2,and 3 kV, respectively. Briefly, to a 0.05ml volumeof plasma, internal standard, [13CD3]BMS-194549,and 5mM ammonium acetate buffer were added.The mixture was gently vortex mixed andextracted with toluene. In the case of brainsamples, only the internal standard was addedto 0.1ml of rat brain homogenate. The organicphase was separated and evaporated to drynessunder a gentle stream of nitrogen at 408C. Theresidue was reconstituted in the mobile phase

(mixture of methanol, water, and 5mM ammo-nium acetate). Standard curves were fitted usinga 1/X2 weighted quadratic regression model andwas over the concentration range 0.5–250 ng/ml,with a lower limit of quantitation of 0.5 ng/ml.Intra- and inter-assay precisions and assayaccuracy were 510% CV.

Pharmacokinetic analyses

Plasma data were subjected to non-compartmen-tal pharmacokinetic analysis [8,9]. The terminallog–linear phase of the plasma concentration-time curve was identified by least-squares linearregression of data points which yielded a mini-mum mean square error. The area under theplasma concentration-time curve from time zeroto time infinity (AUC) was determined using acombination of trapezoidal and log-trapezoidalmethods plus the extrapolated area. The extra-polated area was determined by dividing thepredicted concentration at the time of last non-zero plasma concentration by the slope of theterminal log–linear phase. The elimination half-life (T-half) was calculated by dividing ln2 (or0.693) by the absolute value of slope. Meanresidence time in the body (MRT) was estimatedas equal to [AUMC/AUC]-T/2, where AUMC isthe area under the first moment curve and T, theinfusion time. Total body clearance (CLT) wasestimated as dose/AUC and steady-state volumeof distribution was estimated as a product ofMRT and CLT. The peak plasma concentration(Cmax) was obtained from observed data.

Table 1. Mean pharmacokinetic parameters for BMS-204352 in male rats after intravenous administration ofBMS-204352

Dose Input rate Matrix Cmax

(ng/ml)AUC(ngh/ml)

AUCBrain/AUCPlasma

MRT(h)

T-half(h)

CLT(ml/h)

VSS(ml)

5mg/kg Bolus Brain 9100 13263 7.4 2.4 1.8 Na naPlasma 2016 1787 2.1 1.8 2798 5968

30min Infusion Brain 7254 13749 11.8 2.7 2.3 Na naPlasma 610 1163 2.5 2.1 4299 10703

60min Infusion Brain 5309 14315 11.4 3.0 2.8 Na NaPlasma 534 1261 2.9 2.6 3966 11310

10mg/kg Bolus Brain 20624 37460 8.3 2.4 1.8 Na naPlasma 4509 4526 1.8 1.6 2209 4041

Na: Not applicable

Copyright # 2002 John Wiley & Sons, Ltd. Biopharm. Drug Dispos. 23: 227–231 (2002)

RAT PLASMA AND BRAIN PHARMACOKINETICS OF BMS-204352 229

Results and Discussion

Mean (SD) plasma BMS-204352 concentration-time profiles are illustrated in Figures 1 (5 and10mg/kg bolus doses) and 2 (input rate).Table 1 summarizes the mean pharmacokinetic

parameters for BMS-204352 in plasma andbrain of rats receiving various doses and inputrates.

Plasma concentration-time curves indicate thatBMS-204352 is rapidly distributed upon dosing(Figures 1 and 2). The decline of Maxipost in the

0 2 4 6 8 10

T ime (h)

10

100

1000

10000

BM

S-2

0435

2 C

on

cen

trat

ion

5-BPl.

5-BBr.

5-I-30Pl.

5-I-30Br.

5-I-60Pl.

5-I-60Br.

Figure 2. Mean brain (ng/g) and plasma (ng/ml) concentration-time curves of BMS-204352 in male rats receiving 5mg/kgintravenous bolus, 30min infusion and 60min infusion doses of BMS-204352 (Note: 5-B–5mg/kg bolus; 5-I-30–5mg/kg, 30mininfusion; 5-I-60–min infusion)

0 2 4 6 8 10

Time (h)

10

100

1000

1000030000

BM

S-2

0435

2 C

on

cen

trat

ion

5-BPl.

5-BBr.

10-BPl.

10-BBr.

Figure 1. Mean brain (ng/g) and plasma (ng/ml) concentration-time curves of BMS-204352 in male rats receiving single 5 and10mg/kg intravenous bolus doses of BMS-204352 (Note: 5-B and 10-B refer to 5mg/kg, bolus and 10mg/kg, bolus, respectively)

Copyright # 2002 John Wiley & Sons, Ltd. Biopharm. Drug Dispos. 23: 227–231 (2002)

R. KRISHNA ET AL.230

brain paralleled that of plasma independent ofthe input rate and dose.

As dose increased from 5 to 10mg/kg, bothBMS-204352 Cmax and AUC values increased inplasma and brain in a dose-related manner.Although the increase in exposure was some-what greater than the proportion to the incre-ment in dose, it must be noted that the rats werenot serially bled to allow generation of intra-individual measurements; rather, the data gener-ated were composite in nature. Whereas the peakconcentrations of BMS-204352 were affected byinfusion time, overall AUCs were comparableacross the bolus and infusion groups. Terminaldisposition (T-half ranged from 1.6–2.7 h) ofBMS-204352 was unaltered as a function of inputrate.

Non-compartmental pharmacokinetic analysis(based on AUC of brain and plasma) yieldedbrain-to-plasma (B/P) ratios that appeared to beindependent of dose (ca. 7–8 for bolus and 11 forinfusion), with infusions producing somewhathigher brain penetration as compared to bolusdose.

BMS-204352 was found to cross the blood–brain barrier with significant amount ofdrug in the brain (B/P 7–11), the target site ofaction for this compound. Dose related increasesin BMS-204352 exposure were observed as doseincreased from 5 to 10mg/kg. Terminal disposi-tion of BMS-204352 was unaltered as a function ofinput rate.

Acknowledgements

The authors acknowledge the assistance of thestaff at Primedica Ltd. for animal experimentation.

References

1. Gribkoff VK, Starrett JE Jr, Dworetzky SI, et al. Targetingacute ischemic stroke with a calcium-sensitive opener ofmaxi-K potassium channels. Nat Med 2001; 7: 471–7.

2. Cheney JA, Weisser JD, Bareyre FM, et al. The maxi-Kchannel opener BMS-204352 attenuates regional cerebraledema and neurologic motor impairment after experimen-tal brain injury. J Cereb Blood Flow Metab. 2001; 21: 396–403.

3. Starrett JE, Dworetzky SI, Gribkoff VK. Modulators of largeconductance calcium-activated potassium (BK) channels aspotential therapeutic targets. Current Pharm Design 1996; 2:413–428.

4. Starrett JE, Hewawasam P, Oritz AA, et al. Synthesis,pharmacokinetic analysis and MCAO stroke activity of themaxi-K opener BMS-204352. Proceedings of the KeystoneSymposium: Potassium Channels. Keystone, CO, 2000; 57.

5. Krishna R, Yao M, Kaczor D, Vachharajani N, Srinivas NR.In vitro protein binding studies with BMS-204352: lack ofprotein binding displacement interaction in human serum.Biopharm Drug Dispos 2001; 22: 41–44.

6. Krishna R, Yao M, Vachharajani N, et al. Disposition ofradiolabeled BMS-204352 in rats and dogs. Biopharm DrugDispos 2002; 23: 41–46.

7. Krishna R, Shah V, Srinivas NR. Pharmacokinetics and doseproportionality of BMS-204352 after intraarterial adminis-tration to rats. Biopharm Drug Dispos 2002; 23: 243–247.

8. Riegelman S, Collier P. An application of statistical momenttheory to the evaluation of in vivo dissolution time andabsorption time. J Pharmacokinet Biopharm. 1980; 8: 509–534.

9. Gibaldi M, Perrier D. Pharmacokinetics, (2nd edn), MarcelDekker: New York: 1982.

Copyright # 2002 John Wiley & Sons, Ltd. Biopharm. Drug Dispos. 23: 227–231 (2002)

RAT PLASMA AND BRAIN PHARMACOKINETICS OF BMS-204352 231