No. 3
`
`Chem. Pharm. Bull.]
`20.: 3 ) 443-451 (1972)
`
`(
`
`-·-·· -- ---- -·--- - -- - -- -
`
`443
`
`UDC 615. 31. 03. 076. 9
`
`Absorption of Drugs from the Skeletal Muscle of the Rats. (3). 11 Effect of Water(cid:173)
`soluble Adjuvants and Vehicles on the Intramuscular Absorption21
`
`K11cmRo KAKEMI (the late), Hnosm SEZAKI, KATSUHIKO OKUMURA,
`Hrnosm KOBAYASHI, and SHUNJI FuRUSAWA
`
`Faculty of Pharmaccutical Sciences, ]{yoto University 3>
`
`(Received June 15, Hl71)
`
`The effect of various kinds of acljuvants or vehicles was studied with a view to examine
`the mechanism of intramuscular drug absorption. From the relationship between in
`vivo absorption studies and various in vitro diffusion experiments, it was clarified which
`step was the rate-limiting one in the intramuscular absorption.
`1) The absorption mechanism of a drug with water-soluble adjuvants did not differ
`from that in aqueous solution without any adjuvant.
`2) There was a good correlationship between the parenteral absorption rate and the
`reciprocal of viscosity of an injectable solution, provided that the molecular weight of
`an acljuvant was comparatively small such as propylene glycol, glycerin, and PEG 400.
`Effect of these solvents on the drug absorption was general in nature and not specific to
`<lrugs.
`In the case of an adjuvant having higher molecular weight such as PEG 4000
`dextran, and 1nethylcellurose, rate of drug absorption was greater than that expected
`from the viscosity, suggesting that macromolecules could hardly diffuse through the pores
`of the capillary wall.
`3) vrom in vitro diffusion stu(ly using Visking membrane, glass filter, and slice of
`muscle, it ,ms concluded that the contribution of the diffusion process through the pores
`of capillary \\·all was dominant compared with the one through the muscle fiber space.
`
`In the field of aqueous injectable solution, various kinds of adjuvants or vehicles are
`\\·idely used. Reviev.:s4 •51 are reported on their toxicity or chemical and physical properties,
`and on their effect on the hemolysis of erythrocytes, 61 but few studies have been performed
`of their effect on the absorption mechanism.
`In the previous papers, 1 • 7l it was clarified that the intramuscular absorption of aqueous
`unionized drug solution from the injection site was chiefly proceeded by the apparent first
`order process and the diffusion through the pores of the capillary vessels was predominant
`compared with the penetration through the capillary endotherial cells. Therefore physico(cid:173)
`chemical properties of injectable solution may serve as a major role in the absorption process
`provided physiological condition was controlled normally.
`The purpose of this work was to examine the effect of the viscosity and the osmotic pressure
`on the ab$orption mechanism in the presence of adjuvants. From the relation between in vivo
`absorption study and in vitro diffusion rate analysis, it was clarified that the injected solution
`was absorbed from the injected site through muscle fiber space and then pores of capillary
`walls, and the latter step might act as the rate-limiting step in the absorptive process.
`
`1) Preceding paper, Part lI: Eiichiro Eakcmi, Hitoshi Sezaki, Katsuhiko Okumura, and Chiyoko Takada,
`Chem. Phanu . Bull. (Tokyo), 19, 2058 (Hl71).
`2) Part of this "·ork was presented at 89th Annual }\Jecting of the Pharmaceutical Society of Japan, Nagoya,
`April 1969.
`3) Location: Yoshidashinioadachi-cho, Sakyo-ku, J(yoto.
`. .\.J. Spiegel and M.M. Noseworthy, J. Pharm. Sci., 52, 917 (1963).
`4)
`5) E.S. Lin, J. Anschel, and C.J. Swartz, Bull. Parenteral Drug Assoc., 25, 40 (1971).
`6) D.E. Cad·wallader, ]. Pharm. Sci., 56, 351 (1967).
`i) IC I-:akemi, H. Sezaki, IC Okumura, and S. Ashida, Chem. Pharm. Bull. (Tokyo), 17, 1332 (1969).
`
`MYLAN INST. EXHIBIT 1086 PAGE 1
`
`MYLAN INST. EXHIBIT 1086 PAGE 1
`
`

`

`444
`
`Vol. 20 (1972)
`
`Experimental
`
`Materials--All drugs, isonicotinamidc and methylisonicotinate, and adjuvants, propylene glycol,
`glycerin, polyethylene glycol 400 (PEG 400), polyethylene glycol 4000 (PEG 4000), dextran (mol. wt. 70000)
`and methyl cellulose (4000 cps) were obtained from commercially available sources. All materials, but for
`PEG 4000 and methyl cellulose, were of analytical grade, and used without further purification. PEG 4000
`and methyl cellulose were of extra pure reagent.
`Procedure of Absorption Experiments--The absorption experiments were almost identical with those
`described in the previous paper from this laboratory. 1>
`Preparation of Solutions for Injection and Diffusion--Isotonic buffer systems and the concentration of
`drug solution were the same as previously described. 1>
`Determination of Possible Complexation of Dextran and PEG 4000 with Isonicotinamide--Ultraviolet
`spectrometry and equilibrium dialysis method were used.
`Determination of Viscosities--Viscosities were determined at 36° with B-type viscometer (Tokyo
`l{eiki Seisakusho).
`In the case of propylene glycol and dextran solution, it was confirmed that the solution
`was Newtonian fluid by the use of Universal Rheometer UR-1 of Shimazu Manufacturing Co., Ltd.
`Determination of Diffusion Coefficients--(A) A Glass Filter and Visking Cellulose Membrane as
`Diffusion Barrier: The apparatus used for this study is shown in Fig. 1. The apparatus consists of a jack(cid:173)
`eted glass beaker containing 100 ml of solvent without drug maintained at 37° by circulating water through
`the jacket. The beaker was closed on the top by a rubber stopper with one hole to keep the temperature
`constant and to prevent excessive evaporation. For the diffusion barrier, a glass filter (G-3) and a Visking
`cellulose membrane (Visking Co., Ltd., 24/32, 3 cm diameter) were used, the latter was firmly fixed by rubber
`band around the diffusion cell. All solutions were kept at 37° before use.
`In the diffusion cell,8> the test
`solution with 50 mM isonicotinamide was filled. The cell was then set to contact with the surface of the
`solution. Stirring of the solution was achieved by the use of a magnetic stirring bar throughout the whole
`experiment. Sample solutions of 0.5 ml were withdrawn every one hour in the case of a glass filter and
`every 30 minutes in the case of the Visking membrane through the hole, and after dilution with 5 ml of
`water, optical densities were measured at 270 m1t. The logarithmic plots of the residual amounts of iso(cid:173)
`nicotinamide in the diffusion cell versus time were straight line in c\·ery case, and the diffusion rate wa~
`obtained from the slope.
`
`C
`
`D
`
`-
`
`Fig. 1. Schematic Diagram of the
`Apparatus used to Study the Dif(cid:173)
`fusion Coefficient
`A: diffusion cell
`B :
`test solution
`C: diffusion barrier {galss filter or Visking
`cellulose membrane)
`D: solvent
`· E: magnetic stirring bar
`F : glass tu be for sampling
`
`Fig. 2. Schematic Diagram of the
`Apparatus used to study the Per(cid:173)
`meability Constant
`A,A': cell
`B: ground-glass end
`C: Visking cellulose mc1nbrane
`D: magnetic stirring bar
`
`8) Volumes of diffusion cells of Visking membrane and of glass filter were about 4.5 cm3 and 6.8 cm3
`respectively.
`
`~
`
`MYLAN INST. EXHIBIT 1086 PAGE 2
`
`MYLAN INST. EXHIBIT 1086 PAGE 2
`
`

`

`No. 3
`
`445
`
`(B) Slice of ~foscle as Diffusion Barrier: The experiments were almost identical with those described
`in the previous paper from this Jaboratory. 1>
`Determination of Apparent Membrane Permeability-
`- The whole apparatus•> for the permeability
`study, shown in Fig. 2, consisted of two glass cells, whose volume was about 7 to 8 cm3 • The same Visk(cid:173)
`ing membranes as used in fig . 1 were prepared and attached tightly between the ground-glass end of the
`glass cell by the rubber bands. The area of the membrane available to p ermeate was about 5 to 7 cm 2 •
`. .\11 solutions were kept at 37 ° before use.
`In the cell A, the test solution with 50 mM isonicotinamide was
`filled and, in the cell A', the same solution without drug was :filled. The entire cell was maintained at 37°
`by immersing in a constant-temperature water bath. Stirring of the solution was achieved by using magne(cid:173)
`tic stirring bars and stirring rate was adju sted so as not to affect the permeability . After 2 or 3 hour ex(cid:173)
`periment, the drug concentration of each cell " ·as measured, and apparent m embrane permeability was
`calculated. 10>
`Analytical Methotls---Thc spectrophotometric determination was applied to all the drugs investigated.
`Isonicotinamide: 1 n absorption experiments, same spectrophotometric methods were used as
`i)
`desc ribed previously. 1> 1:or the determination of the samples other than the absorption experiments, 5 ml
`of distilled water was adcled to 0.5 ml sample solution and its optical density was me<J.sured at 270 m1t.
`ii) Methylisonicotinatc :
`In absorption experiments, same spectrophotometric methods were used
`as previously described. 7>
`J n absorption experiments, removed muscle was homogenized and centri(cid:173)
`iii) Propylene Glycol:
`fu gated in the manner as d escribed prcYiously. 7> Supernatant, depr.oteinized by trichloroacetic acid, was
`u sed as 0.5 ml sample solution and 0.5 ml of distilled water, 0.5 ml of 0.lN hydrochloric acid, and 5 ml of
`1.6 % w/v sodium metabisulfite were added. After centrifugation, 1 ml of supernatant was separated.
`Then 2.5 ml of chromotropic acid solution, prepared by solubilizing 200 mg of chromotropic acid in 2 ml
`of distilled water and sufficient sulfuric acid to make 50 ml under cooling in ice-bath, was added, and the
`aliquot was boiled for 30 minutes. After cooling, 10 ml of sulfuric acid was added and the optical density
`,Ya s determined at :i65 m11.
`
`Result and Discussion
`
`Contrary to such pharrn.aceutical dosage form as suspensions, 11> oily solutions,12> and
`emulsions, 13> very little has been understood about the effect of vehicles on the absorption of
`aqueous injection solutions.
`(I) Effect of Adjuvants on the Time Course of Drug Clearance
`For the purpose of examining the possibility of change in the absorption mechanism
`caused by these adjuvants, the time course of drug clearance in the rat rn.uscle was investi(cid:173)
`gated. According to our previous paper,1> it was proved that the time course of the drug
`absorption from aqueous solution followed in most cases apparent first-order kinetics. Figure
`3 shows the logarithrn.ic plots of the amounts of isonicotinamide remaining in the muscle versus
`time after injection of 10 µl of 40% propylene glycol solution and 10% dextran solution. As
`is evident from the figure, straight lines were obtained in both cases, which shows that the
`absorption was proceeded by the apparent first-order kinetics. Accodingly, it was suggested
`that the absorption mechanism of the drugs with water-soluble adjuvants did not essentially
`differ from that in aqueous solution without any adjuvant.
`(2) Effect of Injection Volume in the Presence of Adjuvants
`As reported in the previous papers,1 •7> it was suggested that the parenteral absorption
`from aqueous solution was not affected by the variation of the injection volume. So the effect
`of injection volume was also examined in the presence of adjuvants. Absorption of isonico(cid:173)
`tinamide within 3 min when 40% propylene glycol or 10% dextran was added to injectable
`solution is shown in Fig. 4. No remarkable difference on absorption was observed between
`
`9) This apparatus was originally designed by Prof. ~Iasayuki Nakagaki of I-Cyoto University and Mr. l\fasa-
`katu Yonese of Nagoya City University.
`10) M. );'akagaki a nd i\l. I(oga, Yakugaku Zasshi, 82, 1134 (1962).
`11) F.H. Buckwalter and H.L. Dickson,]. Phann. Sci., 47, 661 (1958).
`12) J .C. Bauernfeind and H.L. Newmark, Bull. Parenteral Drug Assoc., 24, 169 (l!l70).
`13) J .J. \.Vindhcuser, i\I.L. Best, a nd J .H. Perrin, Bull. Parenteral Dnig Assoc., 24, 286 (l 9i0).
`
`MYLAN INST. EXHIBIT 1086 PAGE 3
`
`MYLAN INST. EXHIBIT 1086 PAGE 3
`
`

`

`446
`
`100
`
`"Cl
`~ 50
`·;;;
`s
`
`V
`h
`~ 20
`
`40% propylene glycol
`solution
`
`1
`
`3
`
`5
`Time (min)
`
`10% dextran solution
`
`1
`
`3
`
`5
`
`Fig. 3. Clearance Curves for Isonicotinamide
`Each point represents the mean value of at least five
`experiments. Vertical bars indicate S.D.
`
`Vol. 20 (1972)
`
`100
`
`-I-I-I-
`
`-11-I-
`
`40% propylene
`glycol
`
`solution
`
`10°~ dextran solution
`
`5
`
`10
`
`10
`5
`20
`Injection volume (µl)
`
`:rig. 4. Effect of Injection \"olumc on Parcnkral
`Absorption
`
`these two vehicles in the range of injection volume investigated. This suggests that there is
`little possibility of physiological alteration caused by adjuvant.
`As the further study for confirming whether the decrease of absorption accompany with
`the increase of the amount of propylene glycol, influence of pre-treatment by adjuvant on
`parenteral absorption was examined.
`In this experiment, 10 µl of 40% propylene glycol
`without drug was injected initially and 5 min after the first injection, when propylene glycol
`was expected to be mostly absorbed, the absorption of 50 mM isonicotinamide within 3 min
`was measured by injection of aqueous drug solution into the same injection site. No signi(cid:173)
`It is conceivable, therefore, that
`ficant difference caused by pre-treatment was observed.
`there is little possibility of change in the absorption mechanism caused by the local effect of
`the adjuvants.
`(3) Effect of Osmotic Pressure
`Hydrophilic pharmaceutical solvents are usually incorporated in very high concentrations
`and, in this experiment too, fairly high concentrated solutions were used. Therefore, the
`osmotic pressure of the injected solutions with the solvents of small molecular weight is
`naturally hypertonic. Thus the effect of osmotic pressure on the absorption was examined,
`particularly on the hypertonic range.
`In this
`Effect of osmotic pressure on the absorption from muscle is shovvn in Fig. 5.
`case N,N-dimethylacetamide, a non-aqueous solvent of which the contribution of viscosity
`is almost negligible, was added to 50 mM isonicotinamide solution to make the osmotic pressure
`of the final injection solution in the range of 50 moslVI to 3 osM which exceeds the physiological
`osmotic pressure. As is evident from the figure, no significant difference of the absorption
`was observed either in isotonic or in hypertonic range, which rule out the possible effect of
`osmotic pressure of the solvents on the drug absorption.
`(4) Effect of Adjuvant on Drug Absorption
`(A) Contribution of Viscosity--In Fig. G is shmvn the result of the examination of the
`effect of propylene glycol on absorption in which the left vertical axis is for the absorption
`rate constant, the right vertical one for the reciprocal viscosity. The horizontal one for the
`concentration of propylene glycol in pH 7.0 phosphate buffer. The solid line indicates the
`absorption rate constant calculated from the percentage-absorbed within 3 min and the
`dotted line denotes the reciprocal of viscosity. Good correlationship was obtained between
`these two parameters. Absorption rate of propylene glycol itself is indicated by the mark
`((cid:143)).
`As shown in this figure, absorption rate constant of isonicotinamide and that of pro(cid:173)
`pylene glycol are very close. The observed results rationalize the view that both of the
`components are absorbed by the same route.
`In other words this suggests that in the case
`of unionized drug such as isonicotinamide, drug is not separated from the solvent but is
`transported together ,vith the solvent into blood thus the viscosity of the solvent affects the
`
`MYLAN INST. EXHIBIT 1086 PAGE 4
`
`MYLAN INST. EXHIBIT 1086 PAGE 4
`
`

`

`No. 3
`- - - - · - - - - - - - - --
`
`100:-
`:
`
`I l :~-1-I--~
`~ 5o;t
`I
`I
`
`1 - - - I -
`
`300
`
`1500
`Osmotic pressure (mosl\1)
`
`3000
`
`Fig. il. Effect of Osmotic Pressure on Par(cid:173)
`enteral Absorption
`vehicle: N,N-<limethylacdamick - wakr
`drug:
`isonicotinamidc
`
`-
`0.6 -
`c
`"' "' 0.5:>;
`§
`\
`~ 0.4 \ \ \
`J 0.1
`
`- 1.5
`
`0.5
`
`',a',,,,,:~.
`
`o.si
`
`7·r\
`1 o.
`
`•
`
`~ 0.3 -
`c 0.2 -
`
`'-
`a,
`',,,, •~
`
`----~------.\.
`o-~zc-'co- -40
`60
`80
`100
`Concentration of propylene glycol ( v/ v) %
`
`.
`
`Fig. 7. Effect of Propylene Glycol on
`the Absoption Rate Constant of Methyl
`Isonicotinate
`-------= absorption rate constann of mrthyl iso(cid:173)
`nicotinate
`- 0 -: reciprocal oi viscosity
`
`447
`
`1.0 ~
`'
`
`~
`
`0.5 .-.
`-------
`
`100
`80
`60
`40
`20
`0
`Concentration of propylene glycol (v/v) %
`
`Fig. 6. Effect of Propylene Glycol on
`the Absorption Rate Constant of Iso(cid:173)
`nicotinamide and Propylene Glycol
`
`-------: absorption rate constant of isonicotinamide
`D
`: absorption rate constant of propylene glycol
`---0---: reciprocal of viscosity
`
`1.0
`
`7
`
`~
`
`0.5 .-.
`-------
`
`100
`Concentration of glycerin (v/ v)%
`
`Fig. 8. Effect of Glycerin on the
`Absorption Rate Constant of Iso(cid:173)
`nicotinamide
`-------= absorption rate constant of iso(cid:173)
`nicotinamide
`---O---: reciprocal of viscosity
`
`absorption remarkably. This relationship was also reported by Coles, et al. in the absorption
`experiment of various vaccine formulation administered subcutaneously. 141
`Further examination was made about methylisonicotinate, a drug having high lipid
`solubility. Similar tendency with the case of isonicotinamide, shown in Fig. 7, suggests that
`the nature of the effect of these solvents on drug absorption is not of specific to drugs. The
`absorption rate of methylisonicotinate is larger than that of isonicotinamide by some de(cid:173)
`finite value, which is independent of propylene glycol concentration. This difference can be
`In general, the diffusion process
`attributed to the greater lipid-solubility of the former.
`through pores of the capillary wall and the partition process through the lipid component
`
`1-!) C.L.J. Coles, K.R. Heath, M.L. Hilton, K.A. Lees, P .W. Muggleton, and C.A. ·walton, ]. Pharm. Pharma(cid:173)
`col., 17 (Suppl .), 87s (l!l65).
`
`MYLAN INST. EXHIBIT 1086 PAGE 5
`
`MYLAN INST. EXHIBIT 1086 PAGE 5
`
`

`

`448
`
`Vol. 20 (1972)
`
`of the vascular endothelial cells have been reported for the mechanism of capillary permeability.
`In the latter process which is considered to be more dependent on lipid solubility, effect of
`the water-soluble adjuvants is comparatively small.
`In the limited sense, parenteral drug
`absorption with such adjuvant seem.s to be proceeded dominantly by the former process.
`Similar examination was made for other vehicles such as glycerin, and PEG 400. The results
`are shown in Fig. 8 and. Fig. 9, respectively. As is shown in the figures, good corre1ationship
`between parenteral absroption rate and the reciprocal of viscosity was obtained.
`On the basis of these experiments, it may be proposed that the prediction of the parenteral
`absorption rate of a drug from the injectable solution could be possible by the viscosity of the
`solvents, provided that the solvents are of comparatively small molecular weight and exert
`little local effect. On the other hand, in the case of adjuvants of macro molecules such as
`PEG 4000, dextran, and methyl cellulose, size of molecules outweighs the viscosity. As
`
`I
`- ,
`- 0.5?-
`
`c 0,4
`rs
`:§ 0.3
`
`<;;
`
`2
`e 0.2 •
`.:: c,
`e- 0.1
`.B
`<
`
`' ·
`,0.5 -
`I
`I
`····~>-- . .. .. ' .
`80 mo
`0
`40
`60
`20
`Concentration of PEG 4000
`
`Q
`
`Fig. 10. Effoct of P EG 4000 on the
`Absorption Rate Constant of lso(cid:173)
`nicotinamide
`-e-w-< Jhsorption rate con~tJnt of i:;o .
`nicotinamidc
`- (;,- : r1~ciprc~ l of vitcosi ty
`
`,,-.
`j
`7c: .0.5o
`
`:~
`
`..., .0.4
`t:::
`"'
`ti
`§ 0.3
`'-'
`e 0.2
`"'
`
`\
`
`',
`I
`I
`\ •
`
`I
`
`.. 1.0 -
`
`---~
`· 0.5 ,....
`
`100
`Concentration of PEG 400 (v/v)%
`
`F ig. 9. Effect of PEG 400 on the
`Absorption R ate Constant of Iso(cid:173)
`nicotinamidc
`¥-•··--: absorption rate constant of iso¥
`.njcotin.an1ide
`,. .... i:.)--:"""; r r.::ciprocal of vlsCO$i.ty
`
`.. 1.0
`
`.,.
`0.5 :::;
`
`20
`15
`5
`Concentration of dextran (w/v)%
`
`Fig. 11. Effect of Dextran on the
`Absorption Rate Constant of Iso(cid:173)
`nicotinamide
`--e-: absorption rate constant of iso(cid:173)
`n!co!inamide
`.... ,.C ...... : reciprocal of viscosity
`
`Fig. 12. Effect of Methyl Cellulose on
`the Absorption Rate Constant of Iso(cid:173)
`nicotinamicle
`-•-: absorption rate constant of is.onkotinaxnide
`--~()---: f.c-tlprocal of viscosity
`
`MYLAN INST. EXHIBIT 1086 PAGE 6
`
`MYLAN INST. EXHIBIT 1086 PAGE 6
`
`

`

`No. 3
`
`449
`
`shown in Fig. 10, ll, and 12, absorption rate is greater than that expected from the
`It is proper to consider that the vvater-soluble solution injected into muscle is
`viscosity.
`transported into blood from the injection site through tvvo steps ; as the first step, it diffuses
`through the intercellular space of muscle fiber or connective tissue, and as the second step, it
`diffuses through the pores of the capillary wall. According to Pappenheimer,15l the pore
`radius of capillary wall is 30-40 A, while the intercellular space is much larger than that.
`
`41.0 "' -:;;
`
`'"'?
`
`'1.0 2
`
`k
`c::
`c 0.4
`0
`·u;
`£!
`:£
`c:: 0.3 ..
`:.;
`0
`.....
`'-'
`· 0.5 0
`.3
`"' 0.2 ..
`~
`k
`\.;')
`"' k
`c::
`e 0.1
`•-,,..,
`0
`/':,. ,..._"
`"'
`•!;
`·c,
`---·---1 -a:;
`cu
`"'
`.:. <
`;:::::
`100
`60
`80
`40
`0
`20
`Concentration of propylene glyeol (v/v) %
`
`0.5~
`\
`\
`\
`
`\
`\
`
`:\
`
`!'.
`
`\
`
`:✓;
`
`0
`
`Fig. 13. Relationship between the Absorp(cid:173)
`tion Rate Constant and the Diff usion of
`Isonicotinamide with Propylene Glycol
`added as an A.djuvant
`--e-: absorption rate colliltant o! isonicolinamide
`O ; di(!nslon of isoo.icctinamidc thtough the
`Visking memhran0
`: diffusloo of lwoicotinumido through the
`glnss filter
`
`f::,
`
`Fig. 14. Relationship be tween the Absorp(cid:173)
`tion Rate Constant a11d the Diffusion of
`Isonicotinamide with Dextran added as
`an Atljuvant
`- -~ absorption rate constant of isonicotinmnid e
`- ··-: r ociprocal of viscosity
`O ; diffusion of isonicotinamicle through the Vi•
`sldng rne1nbr-.ane
`: diffusion of isonicotinamide through the glass
`Jilt~,
`
`6
`
`(B) Contribution of Diffusion Rate--ln order to gain further insightinto the nature of
`intramuscular absorption of drugs from aqueous vehicles, various diffusion experiments have
`been undertaken.
`In these experiments, cellulose Visking 1nembrane with similaT pore size
`with capillary pores (average pore size of about 24 A) and the glass filter (No. 3) (average pore
`size of about 40-20 µ) were used for diffusion b arrier. Fig. 13 is the result of diffusion of
`isonicotinamide with propylene glycol added as an adjuvant of low molecular weight. The
`left vertical axis is for absorption rate constant and the right vertical one for the value of
`the relative ratio of the diffusion rate without adjuvant to that with the adjuvant of respective
`concentration. As is evident from the figure, all these three plots have slwwngood correlation(cid:173)
`ship.
`As for dextran which belong to the adjuvants of macro molecular group, results of the
`In this adjuvant, good correlation was obtained with the
`experiments are shown in '.Fjg, 14.
`absorption curve when the Visking membrane was used. However in the case of glass filter
`the plots appeared almost on the curve of the reciprocal of viscosity. These results again
`indicate that, in the case of the solvent of small molecular weight such as propylene glycol,
`the solvent molecule itself permeates into the blood through pores of capillary wall, the drug
`and the solvent are transported together, so that the rate of transportation is rather limited
`by the solvent viscosity.
`\iv11en PEG 4000, dextran or methyl cellulose, which does not
`easily penneate the capillary wall by themselves, was added as an adjuvant, the transportation
`through the capillary wall is now becomes a rate-limiting step. The drug and the adjnvant
`
`15) J .TC Pappenheimcr, Physi<,l. Rw., 33, 387 (19flH).
`
`MYLAN INST. EXHIBIT 1086 PAGE 7
`
`MYLAN INST. EXHIBIT 1086 PAGE 7
`
`

`

`450
`
`Vol. 20 (1972)
`
`are presumably separated and only the former diffuses through pores of capillary wall as the
`simple aqueous solution. This may be considered to be the reason why the absorption rate
`It is conceivable also
`i'.n vivo is greater than the one anticipated from the solvent viscosity.
`that the pharmacological effect of dextran itself, £,e. enhancement of capillary permeability
`caused by histamine release,1m should be taken into consideration. Schou, however, demon(cid:173)
`strated that the self-depression of drug absorption was very pronounced when the injected
`drug is a chemical histamine liberator,m and in our experiments such effect is considered
`to be of little account.
`
`0.5\
`
`~ 0.4
`"' ti
`3 0.3
`
`CJ
`~ 0.2 -
`
`.
`
`(l
`b.\
`\
`\. , .. :,
`
`0
`20
`40
`60
`80
`Concentration of propylene glycol { v/v) %
`
`Fig. 15. Relationship between the
`Absorption Rate Constant ·and the
`Apparent Membrane P ermeability
`through
`the Visking Membrane,
`Permeability through the Slice of
`Muscle
`: absorption rate constant of isonicotinamjde
`/\ : pcnneahility of isonico!innmide tbroug'h
`the Visking membrane
`O : permeability of isonicotin;:nnide through
`the slice of mu.scle
`
`-
`
`CJ
`
`~ 0.2 ·
`
`C:
`0
`-~ 0.1
`0
`,:£
`<:
`
`C
`
`C
`
`20
`15
`10
`5
`0
`Concentrat.ion of dextran ( w /v) %
`
`Fig. 16. Relationship between the Absorp(cid:173)
`tion Rate Constant and the Apparent
`Membrane Permeability
`through
`the
`Visking Membrane, Permeability through
`the Slice of Musch:
`! ahsorptinn rate constant of- isonicoUoamid,~
`.. : r,~dprocal of vbcosi ty
`D. ; permeability of isonicotlnatnide through the
`Viski ng 1n<:mbranc
`O : permeahi1Hy of isonicoiinan1id11 through ill(>
`sliee t>{ musk<'
`
`...... H
`
`-
`
`(C) Contribution of Apparent .Membrane Permeahility---For the purpose of further
`confirm.ation, apparent membrane permeability of isonicotinami<le through the same cellulose
`Visking m.embrane as that used in the previous diffusion experiment was examined. Results
`of propylene glycol and dextran are shown in Fig. 15 and Fig. lG, respectively. The right
`vertical axis denoted the value of the relative ratio of the apparent membrane permeability
`without adjuvant to that when the adjuvant of respective concentration is added.
`In the
`case of propylene glycol, a good correlation was observed between the absorption rate -in vivo
`and the 1u.em.brane permeability. However, in the case of dextran, relative ratio of the
`apparent membrane penneability is greater than the ratio of diffusion rate. Particularly,
`in the range of low concentration of dextran, such as 5% w/v, the ratio of apparent membrane
`permeability is almost unity.
`It m.ay be concluded that vvith the solvent of small rnolecu1er weight such as propylene
`glycol, the pore of the cellulose membrane is filled ·with propylene glycol solution so that the
`membrane permeability of isonicotinamide is rather limited by the solvent viscosity. On
`the other hand, dextran cannot easily penneate the pores of the cellulose membrane and
`isonicotinaiuide molecules have to diffuse through the pores filled with water. At the rm1ge
`
`16) RH. l'oyser and G.B. West, Brit. ]. P/w.rmacol., 25, 602 (19G5).
`17) J. Schou, Nature, 182, 324 (Hl58).
`
`MYLAN INST. EXHIBIT 1086 PAGE 8
`
`MYLAN INST. EXHIBIT 1086 PAGE 8
`
`

`

`Nu. 3
`
`451
`
`(I)
`
`7>------
`
`B
`
`(II)
`
`of high concentration of dextran, howe,·er, the pore masking effect of dextran becomes pre(cid:173)
`dominant thus reduces the apparent membrane permeability.
`In the present report tvvo apparatus with
`Yisking membrane as the barrier have been used
`to investigate the mechanism of intramuscular
`absorption of drug with water-soluble adjuvant.
`The concentration gradients in each apparatus
`and in the case of practical intramuscular in(cid:173)
`jection are shown in Fig. 17. Considering the
`correlationship between in vivo absorption data
`and in vitro diffusion data, it may be concluded
`that, in our experim.ental condition, the diffusion
`apparatus shown in Fig. 1 is the most promi(cid:173)
`sing one for the investigation of the mecha(cid:173)
`It
`nism of intramuscular absorption of drug.
`is also suggested that in the case of macro(cid:173)
`molecules, the absorption process through the
`intracellular space or its masking effect could not
`be neglected in determining the rate of drug
`absorption.
`(D) Contribution of Permeability through Iso(cid:173)
`lated Muscular Slices--Finally to confirm
`the foregoing considerations, we have examined
`the permeability of the drug through isolated
`muscular slices. The result is shown in Fig. 15 and Fig. 16. As can be seen cleary from Fig.
`15 the ratio of diffusion rate to the control declines with the increase of the concentration of
`propylene glycol. Both isonicotinamide and prolylene glycol permeated the slice of muscle
`almost at the same rate, which agrees with the data obtained in the in vivo absorption ex(cid:173)
`periment. On the other hand, as shown in Fig. 16, diffusion rate decreases with the increase
`of the concentration of dextran, which can be explained by the increase of its viscosity. This
`data differs significantly from the absorption data in vivo. Consequently results of muscle
`slice diffusion experiment on dextran could be interpreted that the contribution of the dif(cid:173)
`fusion process through the pores of capillary wall to the drug absorption is do~inant compared
`,,·ith the process through the muslce fiber space. This again supports our conclusion.
`
`depo~;jf--_b_lo_od_
`
`Fig. 17 . Model of Drug Concentration
`Gradient
`A: apparatus shown in Fig. 1
`B : apparatus shown in Fig. 2
`C : practical intramuscular injection
`(I) Visking membrane
`(II) capillary wall
`
`MYLAN INST. EXHIBIT 1086 PAGE 9
`
`MYLAN INST. EXHIBIT 1086 PAGE 9
`
`