uo!le!luo:tojj!p oql u!elsns ol olqe .~offUOl ou o:te SlOAO I V~LLV zo~oI osoqJ. "suoBeamoouoo ~nap emseld gu!seo;oop dlOA!SSOJ~md ol onp s! &de:~oql VH~LV ol oOU~lS!SO.t JO OOUO.L[nOoo lenluoAo oql l~ql popnp moo oqA~ 'It'[ '£[] "1 ~ 1O !pu!nlAI ,~q poqs!lqnd se~ Oouop!AO leOtflOlOOemJeqd IeO!UlIO ~Ullep!onlo lsom oqj~ "so!.~olg.[oq~[ lg.[OAOS u.[ £pms jo snool ~ omoooq s~q VH.LV ol ,oou~ls.tso~, s!ql jo tus!u~qootu oq.L

"[ZT-6] sluotuleo~l ooueuolu!etu jo ol!ds u! osdelo:[ sluoBed jo dl!:[o[~tu oql 'V~LLV ,~q poonpu! uo~ss~tuoa moldtuoo jo sqluotu 9-17 ~olJV "pmuomnoop ,qmms!suoo uooq s~q diluouem -J~d uotsstmoJ oql UlelU!~tu ol dl!Ilqeu! s,V~&V '.IOAOmOq 'z{lI.AI.lOl~ tlOAOJd s.lql ol!dsoo "VH~LV jo ztu/fftu 001-917 jo uoiluJlst.u!tup~ [~Jo ~l!ep jo S~OOA~ 9-17 ffUI.AXOllO] UOtSSUUOJ Ololdtuo3 OAO.IqOl~ '-[dr qlI.Ax stooged jo £l!ao[mu l~oa~ V "[ET-L] oseos!p ,(aol -oe~joa ql.tax sluo!led UT Auetu 'SleUl [I~.IOAOS II.I dde~oql V~LLV ol ssouoAtsuodsoa oiqe~laetuoa pmealsuotuop seq qo!qA~ "-ldV s!souget.p qlAIV otl!oods oql ol su!elJod elep IeOIuIP olqelIeAe oql jo lsoIN "-IIAIV jo luotulgoal oql u! sp!ou!loa jo osn oam m. aoj OleUO!leJ ol_jt.:ltto.los ~uoJls e pop!AoJd sot.pnls o.a~.a u? osoq~L

LS~

"IDN tuoxl EL99I-VD luex~ l~oddn S (oaoD) ~oluo D xooueD £q pue '.sexo£ 'spuelpOOAX oq& "3u I 'SleO.qnooemJeqd sn~Jv £q pug uoge:qs!u!mpv ffm o pug pooc[ 'IIO.IS.IA.I(I slonpoxd ueqd~ O oql tuo~] ~E6000~IOd ]uezff oo!mo S q~IeOH o!Iqnd e Eq ~xed u! po~aoddn S ,

"(TCLT-96L (ELL) :xed) "V'S'CI 'O£OLL X.L 'uo]snoH 'pAIfl oqtuoolOH ~IgT '~oluoD ~ooueD uosaopuv "(I "IAI '09 xofl 'SUO.We~.I:ISOAU I Ieo!U!ID jo luotul:tedo G 'e~qoIAI I!de?,I :ol a~uapuodsa~aOD

• op.txo~Ins I~qlom!p '05/4/(/ '.dqd~ol~tuo.tqo p!nb!i omsso~d-q~!q 'D'-IdH '.p!og 3!ou!lo~-suval-He pole!oosse-omosod!i 'VE£ V-7 '.m.olmd ~u!pu!q-p!o~ o!ougo~ z~lnlioo 'dffV~IO '.ettuo~ln~I snouo~olodm omo~ '7/qV '.e!tuo~InoI o!ld3oIodmmd omo~ 'TdV '.p!oe o!ou!lo~-suv~l-ile 'V~t£V :suoz.lm.ao~qqv

"[L-£] ~IIAIV qllA~ sluo!led tuoaj pou!mqo soJmln3 A~Oazem ouoq pue souII IlO3 e!mo~Ino I uemnq [e.IOAOS jo qlA~Oa~ l!q!qu! pUe uoBm.luoaojj[.p [eU!UlaOl oonpu! ol V~LLV jo gl!l!qe poz!leIoods 0ql pole~lsuotuop s~q so!pros o~?a u~. jo ~IOIJeA V "[E'I] sIm.J~ uetunq m sJotum p.tios [e.IOAOS lSU.Ieff~ ,(B.A!IOe UAXOqS s~q V~LLV "V tI.ItUl~:~.IA JO 01!IoqelotU OA!~3~ £1Ie3!ffolo!q

ST 'u!ou!loJ, ~o '(V~IJ~V) p!oe 3!ougo:-suv~-ll V

uo!lanpmluI

"sowosoaalw '0~t~d 'oouelslsos 0nsp 'sowosodll 'ldV 'p!oe olouBoEI :sp2o~ AoX

"ldV ql!N~ m, UO!led u] suo!ss!woJ WJOl-0UO I 0upnpu! u! in~tesn AlOWOJ~,xo oq plnoo VkI.LV-q se qons VH.LV ~to UO!lelnwJo~ "^'! ue leq1~,sofifins m, lnsoa osoq.L "l~mp 041 ~to WS!loqelow u! aseoaou! lueoLqulS!s e pON~OqS SOSOp owes o41 41!M sow!l ~to aoqwnu owes oq~, VEIlV ooq po~ols!u!wpe Alle~o o JaM le41 SleW!ue wo~,t polelOS! sowoso~o!w seoaoqM "UO!lelnw~oj VEIlV--I oql po~m,s!u!wpe Alpoleodo~ o~o~ le4], slea u! po~olle lOU se~ 0rap o41=to WS!lOqm, eo le41 poleo^o~ sowoso~o!w ~O^!l pOlelOS! ~to so!pnls o11/A Ul "owp, ~o^o o0ue4o lou p!p 0rap o41 ,to SlO^Ol poolq 041 'popod po0uolo~d e ~o^o Sle~ ol po~ols!u!wpe se~ VEIIV-7 uoqM • uop, elnwaoj le~O o4~, ~o^o sofim, ue^pe leO!0OlOOew~eqd lep, um, od op!^o~d o~, AlSnOuo^ealu! po~ols!u!wpe oq ueo le4~, aowos! p!oe o!OU!lO~ s!41 =to UO!lelnw~oj e op!Ao~d ol podOlOAOp se~ (VEIIV-7) VEIlV lewosod!'l "IIOM se so!oueu0!lew ao41o '/qlep, Uolod 'pue e!wmlnOl u! AI!ABOE s,vH1 v ,to uoBe~n p pue Aouolod o41 o^oadw! ABuea!j!uO!s Aew VHIV jo uoBelnw~o j • ̂'! ue leql slsoOOns Ou!pug s!q/"luowleoJ1 poOuoloJd Jo~e uo!le~moouoo winos poonpoJ u, sllnsoJ leql uoBeldepe leO!OOlO3ewJeqd e s! 'e!wo~nol jo wJoj ouo u! 1seal le '/q!A!lOe s,VH/v jo uoBeJnp pol!w!l oql ~oj s!seq oql leql soleo]pu! oouop!^o luoooH "luo!sue~l uooq seq slooj4o sl! jo uoBemp oql lnq 'sle!J1 leO]U!lO uewnq u! pue swolsAs o~nllno onssB polelos! u] so!oueu -O!lew jo oOueJ e lSU!eBe OAgOe UOAOJd uooq seq (VEI.LV) P! oe o!ouBo~-sue~-IIV--~e~sqv

(17661 tla~tVlN OE paldaaav uol.qaa~t "1~66I davnuD£ gE Pam.aaa~t)

"V'S'f'I 'O¢OLL X& 'umsnoH '~muo3 ~aoue3 uos:topu V "O "I~ sgxo.L jo ~:[.ISIOA.IH~I oq, L 'SUO.Beff.BSOAU I [go.m~.l D jo luotulzedo(I 'sao!zae 3 ffruO pue £ffOlOlqounmm I jo uo!]oo S

ma±saaafl-zaaoq qai~tflv 9 pug NaanOoIAI vsaaa± 'IHOHOV S H~iNV~IV,.L 'VZHHIA[ ~IIdV~I

,OlDV DIONI&XH-S:NV~d,/.-TIV JO SIAISINVHD~IAI XDNV~IVX'-ID DI£VdXH ~H& S£N,'-qAIAIflDltID NOI&V'IflSdVDN~ XIAIOSOdH

00"0 + 00"L$176/9EIZ-g¢I0 potaosoa slqff~z IIV "u!m.uR leoz O u! palula d Plq ~ouo!oS ~0!A0SI~ ¢66I (~ lq'~.uXdco !~66[ '96g-LSg "dd '8 "ON 'gI "I°A tPavas~t v!u~a~lna7

(I'I~0071(I~6)9ZI~'[I~I0

uomc~aOd

588 K. MEHTA et al.

effects on leukemic cells in vivo, leading to relapse of the disease. Leukemic cells obtained from relapsed patients, however, continued to be sensitive to ATRA-induced differentiation in vitro [13]. The emergence of clinical resistance to A T R A has been attributed to decreased intestinal absorption, enhanced enzymatic catabolism, and the induction of CRABP, which leads to drug sequestration [14].

These studies provide a strong rationale for dev- eloping an i.v. formulation of A T R A . Since the limitation of A T R A , at least in APL patients, appears to be due to the drug's poor access to the malignant cells, not to resistance at the cellular level, an i.v. formulation would avoid adaptations in drug absorbance or clearance that may, in part, underlie the observed reductions in plasma levels after pro- longed administration. The lipid carrier formulation confers a pharmacological profile to the drug that can be adapted to i.v. administration. A T R A is lipid soluble and requires a solubilizer to make an accept- able solution, but its lipophilic nature and stearic configuration enables it to intercalate into the lipid compar tment of a liposome. Moreover , being extremely toxic when administered intravenously in a detergent solubilizer, it is a candidate for a liposome formulation. Particulate lipid carriers such as lipo- somes have proven useful in altering the toxicity profile of certain antibiotics and anti tumor agents, including A T R A [15-17]. We have demonstrated that liposomal A T R A (L -ATRA) is an effective i.v. formulation that is far less toxic, both in vitro and in vivo, than the free drug [18]. Fur thermore, L -A TRA is at least as efficacious in inhibiting the growth and inducing differentiation of tumor cells as the free drug [19, 20].

In this study, we evaluated the effect of chronic i.v. administration of L - A T R A on several par- ameters that are implicated in the development of acquired 'retinoid resistance' during prolonged oral administration of A T R A . The results suggest that chronic administration of L - A T R A does not affect the levels of circulating drug in rats. Liver micro- somes from these animals did not show any significant change in their ability to metabolize A T R A . In con- trast, long-term oral administration of free A T R A caused a significant decrease in circulating drug levels after 7 weeks of treatment. Liver microsomes from these animals converted A T R A into polar products much more rapidly than microsomes obtained from L-ATRA-t rea ted or untreated animals.

Materials and Methods

Liposomes and L-A TRA L-ATRA was prepared from lyophilized powder in

bottles containing 3 mg of ATRA (kindly supplied by Argus Pharmaceuticals, Inc., The Woodlands, TX) and 45 mg of a mixture of two phospholipids, dimyristoyl leci- thin and dimyristoyl phosphatidylglycerol in a 3:7 (w/w) ratio (Avanti Polar Lipids, Birmingham, AL), essentially as described elsewhere [20]. The drug was stored in the dark at -20°C. Immediately before use, L-ATRA was reconstituted by adding 3 ml of normal saline to each bottle and agitating the suspension on a vortex mixer for 2-3 min. The reconstituted preparation consisted of multilameller liposomes (average size, 3.1 ~tm).

Animals Six-week-old Lewis rats (Charles River, Wilmington,

MA) were used for these studies. Groups of eight female rats each were administered 5 mg/kg body weight of either free ATRA (mixed with mineral oil) orally or L-ATRA intravenously (in tail vein). Each rat received a total of 15 doses, twice a week (3-4 days apart) for 7 weeks. Blood samples (150 ~tl) were collected from the tail vein 5, 30, 60, and 90 rain following administration of the last dose and analyzed for A T R A levels by HPLC. Blood samples were also collected 60 min after administration of the first and sixth dose and analyzed for ATRA. Ninety minutes after the last dose, all the animals were humanely killed, and blood samples (3 ml) were collected to study the hema- tologic and blood chemistry parameters. Sections of tissues were collected on dry ice for further processing or were fixed in formalin for histopathologic analysis.

Cellular retinoic acid-binding protein (CRABP) Liver samples were collected 90 min after the last dose

and processed individually. Total CRABP (CRABP I and II) levels were quantitated by a previously described slab gel electrophoresis technique [21]. Briefly, cytoplasmic proteins were extracted and 100-200 ~tg protein were incu- bated overnight at 4°C in a 100 ~tl solution of 50 nM 3[H]- ATRA (specific activity 49.3Ci/mmol; New England Nuclear, Boston, MA) and 2 mM dithiothreitol with or without 200-fold excess of unlabeled ATRA. Reactants were fractionated over vertical slab gel polyacrylamide electrophoresis under native conditions. After electro- phoresis the gel was divided into lanes and cut into 5 mm bands; radioactivity was assessed in a liquid scintillation counter. Specific binding was determined from the radio- activity recovered with or without the 200-fold excess of unlabeled retinoid.

In vitro metabolism of A TRA Liver samples obtained from the animals at the time

of death were rinsed in ice-cold saline and homogenized individually in a three-fold volume of 0.25 M sucrose- 0.05 M Tris-HC1 (pH 7.4) using a Teflon glass homog- enizer. Microsomes were isolated by differential cen- trifugation (10,000 g, 20min; 100,000 g, 60min) as described elsewhere [22]. The microsomal pellet was sus- pended in 0.05 M Tris-HC1 (pH 7.4), portioned into ali- quots and stored at -70°C. Protein content was determined by Biorad Protein Assay using bovine serum albumin as the standard.

The assay buffer and conditions usea for determining the ability of microsomes to metabolize carboxyl-[laC]ATRA (Amersham, Arlington Heights, IL; specific activity 13.7 Ci/mmol) were essentially the same as those described by Van Wauwe et al. [22]. After 30 min, the reaction was stopped by cooling and the samples were lyophilized to

Liposomal RA circumvents hepatic clearance 589

dryness. Dried residues were extracted with methanol con- taining butylated hydroxyanisole (0.05%, v/v), and the extracts were evaporated and redissolved in small volumes of methanol (25-50 ~tl). A T R A and metabolites were then separated by thin layer chromatography by spotting 20,000-25,000 cpm on 0.25-mm silica-coated plastic sheets (EM Industries, Gibbstown, N J) and developing in a solu- tion of benzene, chloroform, and water (4 : 1 : 1). The radio- active spots were located by spraying the plates with EN3Hance (New England Nuclear) and autoradiography. The radioactive bands were scraped out, extracted with Solvable (New England Nuclear) and counted in a scin- tillation counter. The extent of A T R A metabolism was determined from the proportions of cpm in appropriate zones and expressed as a percentage of the total amount of radioactivity recovered.

HPLC analysis The extent of A T R A metabolism by isolated liver micro-

somes was also determined by HPLC analysis. The reac- tants were lyophilized and the residues were extracted twice with 2 ml of methanol containing 0.05% butylated hydroxyanisole (Sigma Chemical Co., St. Louis, MO). After centrifugation, the supernatants were aspirated and evaporated. The resulting pellets were re-extracted in a methanol: dichloromethane solution (75:25) and again evaporated in oacuo. More than 80% of the added A T R A was recovered. The final pellet was mixed with 200 ~tl of mobile phase for reverse-phase HPLC. A portion of each sample (150 ~tl) was analyzed on a 10 ~tm C18 ~tBondapack

column (3.9 x 300 mm; Waters Associates, Farmingham, MA). Samples were eluted with a solution of methanol, water, and formic acid (60:40:0.05) containing 10 mM ammonium acetate at a flow rate of 2 ml/min. After 20 min, the solvent was changed to 100% methanol in order to elute ATRA.

Reverse phase HPLC was also used for determining the blood concentrations of A T R A . All procedures were performed in a room with the lights dimmed. Whole blood samples (200 gl) were extracted twice (1 ml each) with methanol. After centrifugation, the supernatants were vacuum-dried and the dried pellets were reconstituted in 200 ~tl of methanol. The recovery of A T R A under these conditions was calculated to be 85 -+ 7%. The HPLC sys- tem included two pumps and a Zorbax-C8 reverse phase column (4 mm x 8 cm; Supelco, PA). The mobile phase consisted of a linear gradient between solvent A (THF and water (25 : 75) containing 0.04% ammonium acetate, pH 4) and solvent B (100% THF) during a 16 min run at a flow rate of 1.8 ml/min. The absorbance was monitored at 346 nm. Retention time for A T R A under these conditions was approximately 9.8 min.

Determination of P450 levels Cytochrome P450 levels in liver microsomes were deter-

mined spectrophotometrically as described elsewhere [23]. The assay system is based on the carbon monoxide (CO) difference spectra of dithionite-reduced samples, assuming a value of 91 mM/cm for the molar extinction increment between 450 and 490 m~t. The P450 activity was calculated

TABLE 1. HEMATOLOGICAL AND BLOOD CHEMISTRY PARAMETERS OF RATS FOLLOWING LONG- TERM ADMINISTRATION OF FREE A T R A AND L - A T R A *

Control A T R A (p.o.) L - A T R A (i.v.) (5 mg/kg body weight)

Hematological parameters WBCs (×103/mm 3) 5.1 +-- 1.9 4.3 -+ 1.3 6.3 -+ 1.8 RBCs ( x 106/mm 3) 7.1 -+ 0.1 7.0 -+ 0.3 6.6 -+ 0.5 Hgb (gm/dl) 13.5 +- 0.4 13.3 -- 0.6 12.6 -+ 1.0 Pits ( x l03 /mm 3) 607.0 -+ 106 577.0 -+ 252 644.0 -+ 107 Segs (x l03 /mm 3) 59.3 +-- 419 28.6 +- 8.6 46.0 -+ 14.3 Lymph (× 103/mm 3) 40.0 -4- 4.3 69.0 -+ 10.0 54.1 +- 9.0

Blood chemistry parameters Electrolytes (mEq/1) K + 3.8 -+ 0.2 4.1 --- 0.2 3.8 -+ 0.4 Na ÷ 142.6 - 0.6 142.7 --- 0.9 142.5 - 1.3 C1- 99.3 - 1.5 99.0 --- 2.5 98.7 - 3.5 Creatinine (mg%) 0.43 -+ 0.05 0.56 - 0.1 0.51 - 0.1 BUN (mg/dl) 21.7 --- 3.2 22.7 --- 3.1 22.4 --- 4.7 SGOT (IU) 171.0 -+ 33.1 144.4 -+ 55.2 131.9 --- 72.1 SGPT (IU) 62.3 --- 7.6 76.5 --- 34.0 73.9 --- 17.9 Alk. Phos. (IU) 144.7 -+- 5.0 188.6 --- 23.3 203.0 --- 20.7 Bilirubin (mg%) 0.2 -+ 00 0.16 -+ 0.05 0.13 - 0.05

Abbreviations: WBCs, white blood cells; RBCs, red blood cells; Hgb, hemoglobin; Pits, platelets; Segs, segmented neutrophils; Lymph, lymphocytes; BUN, blood urea nitrogen; SGOT, serum glutamic oxaloacetic transaminase; SGPT, serum glutamic pyruvic transaminase; Alk. Phos., alkaline phosphatase.

* Groups of rats were administered p.o. free A T R A or i.v. L - A T R A twice a week for 7 weeks (15 doses). After the last dose, blood was collected and analyzed for various parameters. The results shown are the average values from four to eight rats --- standard deviation from the mean.

590 K. MEHTA et al.

FIG. 1. Histologic sections of spleens from rats after 7 weeks of treatment (15 semi-weekly doses) with p.o. free ATRA (A) or i.v. L-ATRA (B). The tissues were fixed with 10% buffered formalin and then embedded in paraffin using standard procedures. Sections of 4 ~tm were cut and

stained with hematoxylin and eosin.

by the following formula: (change in absorbance between dithionite-reduced sample and CO sample alone)/ 91 × 1000; it was expressed as nmol/mg protein.

Statistical analysis The mean values for the groups were analyzed by using

Student's t-test for paired samples.

Results

Hematologic and blood chemistry analysis of samples drawn 90 min after administration of the last dose of free ATRA or L - A T R A , summarized in Table 1, revealed no significant changes, except that the number of circulating segmented neutrophils was significantly decreased in animals treated with either drug formulation. This decrease in circulating neu- trophils was more pronounced in rats treated with free ATRA than in those treated with L-ATRA or

untreated controls (p < 0.05). Similarly, no appreci- able change was observed in most of the blood chem- istry parameters studied, except both the free ATRA and L-ATRA-treated rats showed slight increases in alkaline phosphatase levels (Table 1).

Microscopic examination of tissue sections from the liver, lung, spleen, brain, ovary, skin, kidney, and bone marrow of the treated rats revealed no significant changes in the histopathologic charac- teristics. Interestingly, spleens from seven of the eight L-ATRA-treated rats showed the presence of numerous small vacuoles throughout the red pulp area (Fig. 1). These structures might represent entrapped liposomes that were removed during pro- cessing. They were seen throughout the sinusoids and in phagocytes. No such vacuolization was observed in animals that were treated with free ATRA or in control animals treated with saline alone.

Figure 2(A) shows the levels of ATRA in the blood 60 rain after oral administration of free ATRA or i.v. administration of L-ATRA. In general, these blood levels were higher in rats treated with L-ATRA than in those treated with free ATRA. This difference became most striking after 7 weeks of continuous drug treatment. The mean level of ATRA in the blood of rats treated with free ATRA decreased from 3.01 -+ 0.33 ~tg/ml on day 1 to 1.97 + 0.17 ~tg/ml (p < 0.01) after 7 weeks of treatment, whereas the mean blood ATRA levels of rats treated with L-ATRA did not change significantly. The mean ATRA concentration on day 1 (4.42-+ 1.2 ~g/ml) was similar to that at the end of treatment (4.41 -+ 0.2 ~tg/ml). We also studied blood clearance of ATRA following administration of the last dose of ATRA. Results shown in Fig. 2(B) demonstrated that ATRA could be detected in the blood by HPLC 30 rain after ingestion of free ATRA. The drug reached its maximum level (2.01 - 0.24 ~g/ml) after 60 rain and remained constant for at least 90 min (1.97 + 0.17 ~tg/ml). In contrast, much higher con- centrations of ATRA (7.57- + 1.2~tg/ml) were observed in the blood 5 min following i.v. adminis- tration of L-ATRA. The clearance of L-ATRA from blood occurred in two phases, the initial rapid phase (t½~ = 16 min) followed by a slower terminal phase (t½~ = 55 min). Nonetheless, blood levels of the drug were significantly higher in animals treated with L- ATRA at each time point studied (p < 0.001) than in animals administered free ATRA.

Because cytochrome P450-dependent accelerated catabolism and induction of CRABP have been implicated in the acquisition of clinical resistance to ATRA [14], we determined the CRABP and cyto- chrome P450 levels in liver tissues of rats that had been treated with either ATRA formulation. No

Liposomal RA circumvents hepatic clearance 591

A.

5

E

v

). 3

n- l-- 2

Wk. 0 Wk. 3 Wk. 7

B.

m

v

-$

. . . I

e e

2 f 0 0 u i u u

5 30 60 90

T i m e af ter a d m i n i s t r a t i o n ( m i n )

FIG. 2. Blood concentrations of ATRA in rats after 7 weeks treatment with free ATRA or L-ATRA. (A) Groups of eight rats were administered (5 mg/kg body weight) either p.o. free ATRA ([]) or i.v. L-ATRA (11) twice a week for a total of 7 weeks. Blood samples (200 ~1) were collected 60 min after the administration of the first, sixth, and fifteenth doses, and 150 ~tl aliquots of the blood were analyzed for ATRA by HPLC as described under Materials and Methods. (B) Following administration of the last dose, blood samples were collected from animals treated with free ATRA (-O-) or L-ATRA (--I~) at indicated time intervals and analyzed by HPLC for ATRA content as described under Materials and Methods. The results shown are mean plasma drug concentrations in six

rats - S.D.

appreciable differences in CRABP levels were observed between liver samples of rats that had been treated with free A T R A or L - A T R A and untreated

controls (data not shown). Similarly, there were no significant changes in total cytochrome P450 levels in liver microsomes from rats treated with free A T R A (0.63 -+ 0.13 nmol/mg; n = 7) or L - A T R A (0.59 +- 0 .1nmol /mg; n - - 7) or untreated rats (0.68 +- 0.15 nmol/mg; n = 6).

In vitro, however, the liver microsomes isolated from rats that were t reated with free A T R A exhibited much rapid catabolism of A T R A . As shown in Fig. 3, incubation of [14C]ATRA with isolated liver micro- somes in the presence of N A D P H resulted in rapid conversion of A T R A into two polar products as determined by thin layer chromatography. Incu- bation under similar conditions of liver microsomes from untreated rats or rats t reated with L - A T R A revealed a much slower rate of metabolism of A T R A into these polar metabolites (Fig. 3). When combined, these metabolites accounted for about 33-+ 0.8% of the microsomes from untreated rats and 2 8 . 8 - 2.57% of those from L-ATRA- t rea ted rats, whereas they accounted for 57 - 11.2% of the microsomes from animals t reated with free A T R A (Fig. 4(A)). Individual values for intact A T R A and its polar metabolites, generated in the presence of N A D P H by liver microsomes that were isolated from five different rats t reated with either free A T R A or L -A TRA or from three untreated rats, are shown in Fig. 4(B). Microsomes from all L-ATRA-t rea ted and control animals induced much slower catabolism of A T R A than those from rats administered free A T R A (Fig. 4(B)). Liver microsomes isolated from rats that were treated with 'empty liposomes' without A T R A showed rates of conversion of A T R A to its metab- olites similar to those of the untreated controls (data not shown).

The reaction products generated by incubating A T R A in the presence of N A D P H and liver micro- somes were further analyzed by reverse phase H P L C (Fig. 5). Results shown in Fig. 5(A) demonstrated that microsomes from rats t reated with free A T R A converted the drug into four major products (reten- tion times, 7.5-11.5 min). Two of the metabolites were eluted at the same positions as authentic 4- keto A T R A (retention time, 7.8 min) and 4-hydroxy- A T R A (retention time, 9 .5min) . Incubation of microsomes from rats injected with L - A T R A also converted A T R A into polar metabolites (Fig. 5(B)), but these metabolites were different, quantitatively and to some extent qualitatively, f rom those in the group treated with free A T R A . For example, the metabolite that eluted at 11.1 min from the free A T R A microsome reaction mixture was not seen in the L - A T R A microsome reaction mixture. Similarly, the amounts of the other three products (retention times, 7 .8-9 .6min) , as shown in Fig. 5(B), were

592 K. MErrrA et aL

FIG. 3. Autoradiogram of thin layer chromatogram of ATRA metabolites. Liver microsomes isolated from untreated rats (lane 1) or rats treated with free ATRA (lanes 2, 3) or L-ATRA (lanes 4, 5) were incubated in the presence of NADPH and the reaction products were analyzed by thin layer chromatography as detailed under

Materials and Methods.

much smaller in the reaction mixture incubated with microsomes from L-ATRA-treated rats.

Discussion

All-trans-retinoic acid is an effective therapeutic agent for inducing remission in APL patients [9-11]. In fact, ATRA may be superior to conventional chemotherapy in inducing remissions because (i) the disseminated intravascular coagulopathy often accompanying APL is rapidly controlled during ATRA treatment; (ii) as a result of the induction of differentiation of the leukemic promyelocytes by ATRA, the blood cell count increases and bone marrow suppression is rare; and (iii) adverse side effects are often mild and well tolerated by the patients. However, ATRA's inability to maintain the remission state permanently has been well docu- mented. Even when ATRA administration is con- tinued after remission has been achieved, many APL patients still experience relapse [9-11]. Clearly, some mechanism of resistance develops in relapsed patients whereby the ability of A T R A to induce cellular differentiation is diminished substantially. Several in vitro studies have attempted to explain the

evolution of this resistance mechanism, which can be induced in culture after continuous exposure to very high concentrations of the retinoid. These studies have suggested several possible mechanisms of resist- ance including mutations in retinoid nuclear recep- tors [24], changes in messenger RNA expression [11], and alterations in the quantity and binding characteristics of the retinoic acid-binding proteins [25, 26].

Interesting clinical pharmacological evidence regarding ATRA resistance was published recently by Muindi et al. [13, 14]. These authors concluded that the reason for the eventual occurrence of this drug resistance during ATRA therapy is pro- gressively decreasing plasma drug concentration levels. In most patients the onset of the decrease in plasma drug concentration levels is within 2-6 weeks after beginning treatment. Although these lower ATRA plasma levels cannot sustain the differenti- ation effects on leukemic cells in vivo, in culture the leukemic cells from these patients continue to demonstrate cytodifferentiation sensitivity to ATRA [13]. This resistance is not seen with other retinoids such as isotretinoin or etretinate [27, 28]. Published reports of similar effects in both mice and monkeys

Liposomal RA circumvents hepatic clearance 593

A . 80

60

40

2 0

0 CONTROL A T R A L - A T R A

B . 1 0 0 0 0

z D.

8 0 0 0

6 0 0 0

4 0 0 0

2 0 0 0

1 2

CONTROL

_# •

1 2

A T R A

1 2

L - A T R A

FIG. 4. The effect of long-term ATRA administration on drug metabolism by liver microsomes. (A) At the end of the 7 week treatment period, animals were killed and their liver microsomes isolated. The ability of microsomes to metabolize in vitro [14C]ATRA was then determined by incubating microsomes in the presence of NADPH and radiolabeled ATRA (50 nM). The reaction products were fractionated by thin layer chromatography and extent of drug metabolism was determined by counting the metab- olite fractions (as shown in Fig. 3). Results are expressed as percentage of ATRA metabolized to polar products (~) or percentage of ATRA remaining intact (11). The values shown are averages from five rats - S.D. (B) Radioactivity (cpm) recovered from intact ATRA (lane 1) or its polar metabolites (lane 2), as discussed in (A), were plotted

individually for five different rats.

© w

__, ' ,

* e

=.,

B

, g

FIG. 5. Analysis of ATRA metabolites by reverse-phase HPLC. Liver microsomes from rats treated with free ATRA (A) or L-ATRA (B) were incubated with ATRA and NADPH in the presence of 1% DMSO as described in Materials and Methods. Separation was achieved using a C18 ~tBondapack column using methanol-water-formic acid (60: 40 : 0.05) containing 10 mM ammonium acetate at 2 ml/min. After 20 min, unreacted A'I'RA was recovered by changing the solvent to 100% methanol. The elution positions for 4-hydroxy-ATRA and 4-keto-ATRA were

determined at a UV absorbance of 350 nm.

corroborate that this effect is unique to A T R A [29, 30].

These results led us to design L - A T R A , a lipid- based i.v. formulation of A T R A . We hypothesized that, since the limitation of A T R A in A P L appears to be due to access of the drug to the malignant cells, not resistance at the cellular level, an i.v. formulation would avoid adaptations in drug absorbance or clear- ance that may, in part, underlie the observed drop in plasma level after prolonged administration.

594 K. MEHTA et al.

Another advantage of L-ATRA is that the lipid for- mulation bypasses the clearance mechanism that evolves in the livers of patients treated with the oral formulation, so that the liposomal formulation should not be subject to the same relapse rates as have been demonstrated in clinical trials of the free formulation. In addition, the toxic effects of L- ATRA should be less severe than those associated with free ATRA because liposome encapsulation of ATRA decreases direct exposure of the drug during circulation to levels below the orally administered toxic dose. The latter allows greater total exposure of the drug on initial dose accompanied by slower clearance of the ATRA from the site of stem cell seeding.

Liposomes are being evaluated both clinically and experimentally as an alternative drug delivery system. The potential advantages of using liposomes as a delivery system include increased biological activity through specific targeting and decreased tox- icity because of altered pharmacokinetics [17]. We have demonstrated that encapsulating retinoids in liposomes can provide in vivo protection from the drug's toxic effects [18] and preserve the anti- proliferative and differentiation-inducing properties of A T R A against a variety of human leukemia cells and cell lines [18, 19]. Nastruzzi et al. [31] observed even more dramatic effects and reported that L- ATRA is 300 times more effective than free ATRA in inhibiting the growth of leukemia and melanoma cells.

A T R A is metabolized by a hydroxylation reaction of the cyclohexenyl ring to produce 4-hydroxy metab- olites which are further oxidized to the 4-oxo metab- olites. The hydroxylation of ATRA to the 4-oxo- ATRA metabolite is known to be mediated by cyto- chrome P450-dependent enzymes [32, 33]. The most favored explanation of the pharmacological mech- anism of A T R A resistance is that continuous ATRA treatment acts to induce catabolic enzymes that are responsible for conversion of the drug. Animal studies in which ATRA was administered in com- bination with cytochrome P450 enzyme inhibitors (e.g. ketoconazole or liarozole) showed a significant prolonging of the ATRA plasma half life [22, 34, 35], thereby supporting the contention of accelerated enzymatic degradation. The results reported here confirm the previous observations that chronic oral administration of ATRA in rats results in decreased drug plasma concentrations, whereas i.v. adminis- tration of L-ATRA at a similar dose and regimen did not alter the pharmacological behavior of the drug and the blood levels remained stable throughout the study period (Fig. 2). The observed differences in pharmacological behavior of the two formulations

were consistent with induction of an enzymatic process. Although no differences were observed in total P450 levels in rats treated with either formu- lation, microsomes from rats treated with free A T R A metabolized the drug much more rapidly than those from rats treated with L-ATRA (Figs 3, 4 and 5).

Another factor that might contribute to the reti- noid relapse phenomenon involves the role of high- affinity retinoic acid-binding proteins CRABP I and II. These proteins are believed to mediate the trans- fer of the retinoid from cytoplasm to the nucleus of the cell. Increased levels of CRABP may cause the pooling of retinoids in tissues, resulting in low plasma levels and accelerated clearance of the drug from the circulation. In normal body tissues the expression of CRABP is thought to increase with continuous exposure to retinoids. An increase in CRABP has been documented in human skin as a result of repeated topical application of A T R A [36]. A similar increase in skin CRABP levels was also observed by Adamson et al. in rhesus monkeys following chronic i.v. administration of ATRA [37]. These authors concluded that the increase in CRABP expression was not related to the increase in plasma drug clear- ance observed with continuous A T R A adminis- tration but rather was related to catabolic enzyme induction [37]. In the study presented here, no increase in levels of liver CRABP was observed in rats administered either free ATRA or L-ATRA on a continuous basis. We did not study CRABP expression in the skin of these rats, which might represent a major site of drug distribution, as evi- denced by skin toxic effects following free ATRA administration. In this context, it is worth mentioning that liposomal-mediated delivery of drugs may sig- nificantly reduce the distribution of the drug to the skin tissue [17].

The results of this study, coupled with the recent data obtained in clinical trials suggesting that long- term oral administration of ATRA is associated with the rapid clearance of the drug from plasma that, in turn, contributes to the relapse of the disease in APL patients, strongly supports the rationale of using L-ATRA to induce long-term remissions in APL patients. Based on the results of this study, a Phase I clinical study of the i.v. L-ATRA formulation was initiated recently at this institute in patients with hematological malignancies.

Acknowledgements--The authors wish to thank Dr L. Clifton Stephens, a veterinary pathologist, for evaluating the histopathologic characteristics of the tissues and Ms Kathryn Hale for editorial review of the manuscript.

Liposomal RA circumvents hepatic clearance 595

References

1. Lippman S. M., Kessler J. F. & Meyskens F. L. (1987) Retinoids as preventive and therapeutic anti-cancer agents. Cancer Treat. Rep. 17, 493.

2. Smith M. A., Parkinson D. R., Cheson B. D. & Friedman M. A. (1992) Retinoids in cancer therapy. J. clin. Oncol. 10, 839.

3. Breitman T. R., Selonick S. E. & Collins S. J. (1980) Induction of differentiation of the human promy- elocytic cell line (HL-60) by retinoic acid. Proc. natn. Acad. Sci. U.S.A. 77, 2936.

4. Tobler A., Dawson M. I. & Koeffler H. P. (1986) Retinoids: structure-function relationship in normal and leukemic hematopoiesis in vitro. J. clin. Invest. 78, 303.

5. Breitman T. R., Collins S. J. & Keene B. R. (1981) Terminal differentiation of human promyelocytic leu- kemia cells in primary culture in response to retinoic acid. Blood 57, 1000.

6. Chomienne C., Ballerini P., Balitrand N., Daniel M. T., Fenaux P. & Degos L. (1990) All-trans-retinoic acid in acute promyelocytic leukemias. II. In vitro studies: structure-function relationship. Blood 76, 1710.

7. Flynn P. J., Miller W. J., Weisdorf D. J., Arthur D. C., Brunning R. & Branda R. F. (1983) Retinoic acid treatment of acute promyelocytic leukemia: in vitro observations. Blood 62, 1211.

8. Nilsson B. (1984) Probable in vivo induction of dif- ferentiation by retinoic acid of promyelocytes in acute promyelocytic leukemia. Br. J. Haemat. 57, 365.

9. Huang M., Chen S., Chi J.-R., Lu J.-X., Zhoa L., Gu L.-J. & Wang Z.-Y. (1988) Use of all-trans-retinoic acid in treatment of APL. Blood 72, 567.

10. Castaigne S., Chomienne C., Daniel M. T., Ballerini P., Berger R., Fenaux P. & Degos L. (1990) All- trans-retinoic acid as a differentiation therapy for acute promyelocytic leukemia. I. Clinical results. Blood 76, 1704.

11. Warrell R. P. Jr, Frankel S. P., Miller W. H. Jr, Scheinberg D. A., Itri L. M., Hittelman W. N., Vyas R., Andreeff M., Tafuri A., Jakubowski A., Gabrilove J., Gordon M. & Dmitrovsky E. (1991) Differentiation therapy of acute promyelocytic leukemia with tretinoin (all-trans-retinoic acid). New Engl. J. Med. 324, 1385.

12. Chen Z. X., Xue Y. Q., Zhang R., Tao R.-F., Xia X-M., Li C., Wang W., Zu W.-Y., Yao X.-Z. & Ling B.-J. (1991) A clinical and experimental study on all- trans-retinoic acid-treated acute promyelocytic leu- kemia patients. Blood 78, 1413.

13. Muindi J., Frankel S. R., Miller W. H., Jakubowski A., Scheinberg D. A., Young C. W., Dmitrovsky E. & Warrell R. P. Jr (1992) Continuous treatment with all-trans-retinoic acid causes a progressive reduction in plasma drug concentration: implication for relapse and retinoid 'resistance' in patients with APL. Blood 79, 299.

14. Muindi J., Frankel S. R., Huselton C., DeGrazia F., Garland W. A., Young C. W. & Warrell R. P. Jr (1992) Clinical pharmacology of all-trans-retinoic acid in patients with acute promyelocytic leukemia. Cancer Res. 52, 2138.

15. Emmen F. & Storm G. (1987) Liposomes in treatment of infectious diseases. (1987)Pharmaceut. Weekbl. Sci. 9, 162.

16. Yatvin M. B. & Lelkes P. I. (1982) Clinical prospects of liposomes. Med. Phys. 9, 149.

17. Weinstein J. N. (1984) Liposomes as drug carriers in cancer therapy. Cancer Treat. Rep. 68, 127.

18. Mehta K. (1989) Interaction of liposome-encapsulated retinoids with normal and leukemic cells. In Phar- macology and the Skin (Reichert U. & Shroot B., Eds), p. 74. Karger AG, Basel.

19. Drach J., Lopez-Berestein G., McQueen T., Andreeff M. & Mehta K. (1993) Induction of differentiation in myeloid leukemia cell lines and acute promyelocytic leukemia cells by liposomal all-trans-retinoic acid. Can- cer Res. 53, 2100.

20. Sacks P., Oke V. & Mehta K. (1992) Antiproliferative effects of free- and liposome-encapsulated retinoic acid in a squamous carcinoma model: monolayer cells and multicellular tumor spheroids. J. Cancer Res. Clin. Oncol. 118, 490.

21. Siegenthaler G. & Saurat J.-H. (1987) A slab gel electrophoresis technique for measurement of plasma retinol-binding protein, cellular retinol-binding and retinoic acid-proteins in human skin. Eur. J. Biochem. 166, 209.

22. Van Wauwe J. P., Coene M.-C., Goossens J., Van Ninjen G., Cools W. & Lauwers W. (1988) Keto- conazole inhibits the in vitro and in vivo metabolism of all-trans-retinoic acid. J. Pharmac. exp. Ther. 245, 718.

23. Johannessen K. A. & DePierre J. W. (1978) Measure- ment of cytochrome p450 in the presence of large amounts of contaminating hemoglobin and methemo- globin. Analyt. Biochem. 86, 725.

24. Robertson K., Emami B. & Collins S. J. (1992) Reti- noic acid-resistant HL-60R cells harbor a point mutation in the retinoic acid receptor ligand-binding domain that confers dominant negative activity. Blood 180, 1885.

25. Cornic M., Delva L., Guidez F., Balitrand N., Degos L. & Chomienne C. (1992) Induction of retinoic acid- binding protein in normal and malignant human myeloid cells by retinoic acid in APL patients. Cancer Res. 52, 3329.

26. Boylan J. F. & Gudas L. (1991) Overexpression of the cellular retinoic acid binding protein-1 (CRABP-I) results in a reduction in differentiation specific gene expression in F9 teratocarcinoma cells. J. Cell Biol. 112, 965.

27. Brazell R. K. & Colburn W. A. (1982) Phar- macokinetics of the retinoids isotretinoin and etret- inate: a comparative review. J. Am. Acad. Dermat. 6, 643.

28. Brazell R. K., Vane F. M., Ehmann C. W. & Colburn W. A. (1983) Pharmacology of isotretinoin during repetitive dosing to patients. Eur. J. clin. Pharmac. 24, 695.

29. Creech K. J., Slikker W., Bailey J. R., Roberts L. G., Fisher B., Witthoft W. & Nau H. (1991) Plasma pharmacokinetics and metabolism of 13-eis and all- trans-retinoic acid in the cynomolgus monkey and the identification of 13-cis and all-trans-retinoyl-fl-glu- curonides: a comparison to one human case study with isotretinoin. Drug Metab. Dispos. 19, 317.

30. Kalin J. R., Starling M. E. & Hill D. L. (1981) Dis- position of all-trans-retinoic acid in mice following oral doses. Drug Metab. Dispos. 9, 196.

31. Nastruzzi C., Walde P., Menegatti E. & Gambari R.

596 K. MEHTA et al.

(1980) Liposome-associated retinoic acid. Increased in vitro antiproliferative effects on neoplastic cells. FEBS Lett. 259, 293.

32. Roberts A., Lamb L. & Sporn M. (1980) Metabolism of all-trans-retinoic acid in hamster liver microsomes: oxidation of 4-hydroxy to 4-keto retinoic acid. Archs Biochem. Biophys. 199, 374.

33. Leo M., Iida S. & Liebner C. (1984) Retinoic acid metabolism by a system reconstituted with cytochrome P-450. Archs Biochem. Biophys. 234, 305.

34. Van Wauwe J., Van Nyen G., Coene M-C., Stoppie P., Cools W., Goossens J., Borghgraef P. & Janssen P. A. J. (1992) Liarozole, an inhibitor of retinoic acid metabolism, exerts retinoid-mimetic effects in vivo. J. Pharmac. exp. Ther. 261, 773.

35. Van Wauwe J., Coene M-C., Goossens J., Cools W. & Monabliu J. (1990) Effects of cytochrome P-450 inhibitors on the in vivo metabolism of all-trans-reti- noic acid in rats. J. Pharrnac. exp. Ther. 252, 365.

36. Hirschel-Scholz S., Siegenthaler G. & Saurat J. H. (1989) Isotretinoin differs from other synthetic reti- noids in its modulation of cellular retinoic acid-binding protein (CRABP) Br. J. Dermat. 120, 639.

37. Adamson P. C., Boylan J. F., Balis F. M., Murphy R. F., Godwin K. A., Gudas L. J. & Pop|ack D. G. (1993) Time course of induction of metabolism of all-trans- retinoic acid and the up-regulation of cellular retinoic acid-binding protein. Cancer Res. 53, 472.