JOURNAL OF
`
`Pharmaceutical
`
`Marc/9 1968
`volume 57, number 3
`
`
`Sciences
`
` Review Article—
`
`Biopharmaceutical Considerations in Subcutaneous
`and Intramuscular Drug Administration
`
`By BERTON E. BALLARD
`
`THE PURPOSES of this review are to consider
`some of the factors affecting the absorption
`rates of drugs administered subcutaneously and
`intramuscularly, and to consider some of
`the
`literature pertaining to their formulation and
`administration. This review will
`largely sup-
`plement rather than duplicate references con-
`tained in earlier reviews on these subjects (1—4).
`As a result, certain important but previously
`covered topics may be omitted entirely, or be
`discussed from different points of View. The
`parenteral route of drug administration is an im-
`portant one for the preclinical screening and
`evaluation of drugs as well as their clinical use in
`human and veterinary medicine. The rational
`design of products intended for parenteral use de-
`pends upon many factors including the physical,
`chemical, pharmaceutical, pharmacological, and
`toxicological properties of the drug and the adju-
`vants used in its formulation. Certain ana-
`
`tomical or physiological factors, such as injection
`site and body movement, may influence the ab-
`sorption rates of some parenteral products.
`ENDOTHELIAL TISSUE
`
`Although this review will not go into extensive
`detail on the anatomy and histology of the micro—
`Received from the School of Pharmacy, University of
`California, San Francisco Medical Center, San Francisco,
`CA 94122
`This paper was supported by general research support
`grant FR 05453 from the Division of Research Facilities
`and Resources, National Institutes of Health, U. S. Public
`Health Service, Bethesda, Md.
`The author wishes to dedicate this paper to Dean T. C.
`Daniels, whose leadership at the University of California has
`done so much to raise the standards of education and research
`in the pharmaceutical sciences.
`
`circulation at the many possible injection sites, a
`few key points are worth mentioning. At present
`there are many investigations of
`the several
`mechanisms involved in the passage of drugs and
`other substances through capillary and lymphatic
`vessels and the ground substance surrounding
`them. The subcutaneous region is well supplied
`by capillary and lymphatic vessels (5). Muscle
`tissue also has a rich supply of capillary vessels.
`However, it is generally agreed that there are few,
`if any, lymph vessels in muscle tissue proper (5,
`6). There is an abundant supply of lymphatic
`vessels in connective tissue sheaths and tendons.
`
`Lymph vessels usually exist where fascial planes
`enter muscles, and the fluid moves through
`spaces along fascial planes between muscle fibers.
`One important difference among lymphatics in
`several regions of the body concerns the state of
`their intercellular junctions.
`In active regions of
`the body one junction in two to five may be open,
`but in motionless regions there may be only one
`open junction in 50 to 100. Normal lymphatics
`in motionless regions and normal blood vessels al—
`low very few particles to pass through their junc-
`tions.
`In contrast, lymphatics in regions where
`there is much movement, injured lymphatics, and
`injured blood vessels all allow much more mate-
`rial to pass through their frequently opened junc-
`tions. Mild trauma near a lymphatic vessel in a
`motionless region can result in a marked increase
`in the vessel’s permeability, probably due to the
`opening of many of the normally closed junctions
`(7).
`357
`
`AstraZeneca Exhibit 2115 p. 1
`InnoPharma Licensing LLC v. AstraZeneca AB
`IPR2017-00904
`
`Fresenius-Kabi USA LLC v. AstraZeneca AB IPR2017-01910
`
`

`

`358
`
`The passage of various particles and fluid from
`one to the other side of blood and lymphatic
`endothelial tissue has been shown to occur in part
`by means of vesicles. Particles and fluid con-
`tacting one endothelial surface enter cells in small
`vesicles, and then leave the opposite surface from
`the same organelles. This process has been called
`cytopemphis (6, 7). There is evidence that these
`vesicles are not directed in their intracellular pas-
`sage, but may move and discharge their contents
`randomly on either side. Cytopemphis permits
`the net transport of material to be proportional to
`the concentration difference of the substance on
`either side of the endothelial tissue. The various
`
`paths by which particles of different sizes are
`thought to traverse the lymphatic endothelium
`have been summarized by Casley—Smith (7). A
`classification of the types and sizes of apertures
`found in normal and injured capillary and venous
`walls has been compiled by Landis (8). As to
`particles, there are theoretically only four possible
`paths that can be taken through endothelial
`tissue and these include:
`intercellular, intracel-
`lular in organelles, cytoplasmic matrix, or fenes-
`trae paths.
`It appears that all four of these paths
`are used to some extent in various endothelial tis-
`
`sues (7). For lipid-soluble molecules there is
`abundant evidence that
`they diffuse through
`capillary walls and can pass through regions of the
`wall that are relatively impermeable to lipid—in-
`soluble molecules (9).
`
`DIFFUSION
`
`From a macroscopic point of view, many drugs
`in solution injected into subcutaneous or intra-
`muscular sites behave as if their absorption were
`taking place passively by diflusion. For most
`drugs it has not been determined whether the
`microscopic absorption route is via capillaries,
`lymphatics, or both.
`Insofar as drug molecules
`penetrate endothelial tissues rapidly by passive
`diffusion, the penetration rate of the drug from
`the injection site can be described by Fick’s Law
`in one direction.
`
`dN/dt = 5A(dC/dx)
`
`(Eq. 1)
`
`where dN/dt is the penetration rate, N is the
`amount of drug penetrating the tissue, and t is
`time.
`In a well-stirred system, the rate is pro—
`portional to the mean diffusion coefficient (10) of
`the drug in the membrane, D, the area, A, of the
`absorbing membrane exposed to the solution, and
`the concentration gradient (dC/dx) of drug across
`the membrane. Equation 1 may be approxi-
`mated (11) by:
`
`5.4K
`dN/dt = T (C: — Cb)
`
`(EQ- 2)
`
`Journal of Pharmaceutical Sciences
`
`where the dx term of Eq. 1 has been replaced by
`6, the thickness of the thin subcutaneous or intrao
`muscular membrane that is assumed to be con—
`
`stant for each animal and site, and the dC term of
`Eq.
`1 has been replaced by the terms K and
`(C. — Cb). The term K is the equilibrium distri—
`bution ratio or partition coefficient of the lipid—
`soluble drug between the membrane lipids and the
`aqueous phases found at the injection site or body
`fluids. The term (C. —— Cb) is the difference be—
`tween the drug concentration at the injection
`site, C,, and the drug concentration, Cb, in the
`body fluids, e.g., blood and lymph, flowing past
`the absorption site at any time.
`The term C1, can usually be ignored, since it can
`be reasonably assumed that the drug‘s concentra-
`tion in the fluids of distribution is negligible com-
`pared to its concentration at the absorption site
`at any time. Thus, Eq. 2 can be written as:
`
`(Eq- 3)
`
`dN/dt = (94?) C.
`When the volume of the drug solution at the
`absorption site, V.,
`remains nearly constant
`throughout the experiment, the rate of penetra-
`tion will be equal to:
`
`5.4K
`dN/dt — ( WI) A.
`where A. is the amount of drug at the site at any
`time. Equation 4 can be further reduced to:
`
`(Eq- 4)
`
`dN/dt = PA.
`
`(Eq. 5)
`
`where P is the penetration coefficient which in-
`cludes all the terms in parentheses in Eq. 4, and
`has the units of time“. The penetration coeffi-
`cient or absorption rate constant has the same
`magnitude as the clearance constant, but the op-
`posite sign. Thus, it can be seen from Eq. 4 that
`the absorption rate is directly proportional to D,
`A, and K, but is inversely proportional to 6 and
`V.. The half-life of the drug at the absorption
`site, in, can be calculated from a plot of the
`logarithm of the fraction of drug remaining at the
`site at any time versus time. The half-life and
`penetration coefficient are related by:
`P = ln 2/to,5
`
`(Eq- 6)
`
`Area—The local distribution of solutions
`
`intramuscularly
`injected subcutaneously or
`is of interest, because the penetration rate of
`the drug depends in part upon the geometry
`and the resulting area of the depot exposed to
`the tissue. When the water-immiscible sili-
`
`cone, dimethylpolysiloxane,1 was injected sub-
`cutaneously into animals, most of it was limited
`to the tissue planes of the subcutaneous region.
`1 Dow Corning 360 medical fluid.
`
`AstraZeneca Exhibit 2115 p. 2
`
`

`

`Vol. 57, No. 3, March 1968
`
`Pathologic findings in the rat, mouse, and guinea
`pig following the administration of massive doses
`of the silicone were essentially similar.
`In most
`instances the silicone was contained in innumer-
`
`able thin-walled spherical or ellipsoidal sacs and
`several
`larger pools within the adipose tissue
`(12, 13). Brown et al. (14) studied the disposi-
`tion of subcutaneous injections of a radiopaque
`water-in-oil emulsion in guinea pigs at the injec-
`tion site. The horizontal and vertical aspects of
`the injection site were X-rayed over a period of
`days. A cross-section of the tissue showed on the
`film a lateral spread of the radiopaque material in
`the subcutaneous tissue that progressed over a
`period of days.
`What is the immediate distribution of sub-
`
`(15)
`stances injected intramuscularly? Shaffer
`injected radiopaque substances into the gluteus
`muscle of humans. He found that iodized oil and
`
`bismuth salicylate in oil were confined to the
`planes of fascia or connective tissue surrounding
`muscles and groups of muscles.
`Injections of
`metallic bismuth suspended in isotonic dextrose
`solution were similarly distributed in the fascial
`planes of
`the muscle. Fluoroscopic
`studies
`showed that viscous oily solutions, such as io-
`dized oil, tended to form a sphere—shaped deposit
`about the needle point and then spread out more
`slowly.
`It did not spread as extensively along
`the fascial septums as did the aqueous systems
`studied. The oily solutions continued to spread
`from 1 to 5 min. before they became fixed in posi-
`tion. The aqueous suspension of bismuth spread
`to its final location almost as soon as the injection
`procedure was completed. Some authors (16)
`have suggested that the term “intramuscular” is a
`misnomer and should be called “intermuscular”
`
`when it refers to injections in this region, because
`of the spread of solutions along the fibrous tissue
`between muscle fibers. Gruhzit (17) also noted
`that water-soluble bismuth tartrate injected in—
`tramuscularly was precipitated in the tissues and
`distributed along the connective tissue fasciculi
`between the muscle fibers. Zelman (18) has
`stated that deep, firm massage of the muscle tissue
`following an intramuscular injection favors the
`spread of the medication through a wider area of
`tissue, thus increasing the area for absorption to
`take place.
`From the foregoing examples it can be seen that
`with the usual needle-injection technique, it is
`difficult, if not impossible, to control the area of
`an injected solution, suspension, or emulsion in
`contact with the tissues.
`If other factors are
`
`(e.g., metabolic and excretion rates)I one
`equal
`would expect that higher initial serum levels of
`drug would result when the solution has a larger
`
`359
`
`area exposed to the tissue, since drug penetration
`rate is directly proportional to this area.
`Recently, an experimental technique in animals
`has been devised (19) to study the in viva sub«
`cutaneous absorption rate of about 1% (w/v)
`benzyl alcohol in normal saline where the area of
`tissue exposed to the solution is held constant.
`A medical silicone adhesive is used, which is
`capable of affixing a cylindrical, hollow glass ab—
`sorption cell to moist subcutaneous tissue. This
`adhesive prevents the spread of the solution from
`the site, and is reported to be completely inert to
`living tissue and will not cause irritation or sensi-
`tization. With this procedure it is possible to
`keep the solution in the cell constantly stirred,
`which cannot be conveniently done when needle-
`injection techniques are used. Furthermore, it is
`convenient when removing drug samples period-
`ically for analysis.
`Volume—The subcutaneous absorption cell
`just described has another advantage. The
`volume of solution in the cell, as well as the
`area exposed, can be held constant. Thus, the
`penetration coefficient, P, defined in Eqs. 4
`and 5 will be constant for fixed values of D, K,
`and 8. One would predict that if the volume of
`the solution in the cell were increased over con-
`
`trol values, a decrease in the magnitude of P
`would occur.
`If so, the result would be that the
`penetration rate of the drug would also decrease.
`Similarly, a decrease in volume over controls
`should result in an increase in both P and the
`
`penetration rate. These predictions have not
`been verified experimentally using the subcu-
`taneous absorption cell
`just described, but
`Sund and Schou (20) have shown that the intra-
`muscular clearance rates of radioactive mannitol
`
`and sucrose solutions were inversely prOportional
`to the volume of solution injected.
`When radioactive drugs or ions are used in ab-
`sorption studies, it has been observed that plots of
`the logarithm of the fraction of drug (e. g., radi0«
`active counts) remaining at the site versus time
`may not be linear (20). The absorption half—life
`at
`the beginning of
`the experiment
`is often
`shorter than that calculated at some later time.
`
`This may indicate that a bi- or polyexponential
`equation would fit the data better, A few of the
`reasons for this phenomenon are:
`(a) blood or
`lymph flow from the site may be impaired at the
`later time, (b) the solution at the site is not uni-
`formly mixed and there is a longer diffusion path
`for molecules at the center of the injected solu-
`tion, or (c) the magnitude of the penetration coeffi-
`cient may not remain constant with time. For
`example, if the area of tissue exposed to the liquid
`progressively decreases, or if the volume of solu-
`
`AstraZeneca Exhibit 2115 p. 3
`
`

`

`360
`
`tion progressively increases at the site, the value
`for the penetration coefiicient will decrease with
`time to give rise to a longer absorption half—life
`toward the end of the experiment. Either hy—
`drostatic (e.g., injection pressure), osmotic pres-
`sure effects, or both may also be factors in deter—
`mining the net solvent flow between the fluids at
`the site and in the vessels.
`
`Concentration—Sund and Schou (20) studied
`the effects of different drug concentrations on
`clearance rate. On the one hand, with a
`pharmacologically inert. neutral, water-soluble
`substance like sucrose, the clearance rate from
`rat muscle was independent of drug concen-
`tration over the range of 0.19—9.6 mg./ml.
`This is the predicted result on the basis Of
`Eqs. 4 and 5 when the injection volume is
`constant as it was in their experiment. On
`the other hand, the absorption rate Of a substance
`like atropine depends markedly upon its concen-
`tration at the injection site. The relative clear,—
`ance rate for atrOpine decreases with increasing
`concentrations Of the drug Over a threshold value
`of about 0.5 mg. /ml. Atropine and several other
`anticholinergic drugs when mixed with sucrose
`solutions decreased the clearance rate of sucrose.
`
`The self-depression of atropine absorption and the
`inhibited absorption of sucrose are probably both
`due to the local pharmacological action of atro-
`pine. Sund and Schou (21) conclude that al-
`though the exact mechanism is not known, atro-
`pine probably interferes with local blood flow at
`the injection site. Atropine also has interesting
`local efiects on the inflammatory process. Houck
`(22) administered intradermal injections of 50%
`croton oil in peanut oil to rats. Croton Oil, a
`water-immiscible vesicant, produces wounds that
`are reproducible. He found that
`the micro-
`circulatory insufficiency characteristic of injuries
`produced by the diluted croton oil was eliminated
`by the subcutaneous administration of atropine.
`He proposed that one of the primary effects of
`atropine was to accelerate wound healing by re-
`storing the biological continuity of the wound
`with the circulation and surrounding tissue.
`Molecular Size—As a first approximation,
`the difiusion coefficient of a spherical or nearly
`spherical drug molecule that is larger than the
`solvent molecules is inversely proportional to
`its molecular radius (or weight). Thus,
`in a
`diffusion-controlled process,
`large molecules
`would be expected to have slower penetration
`rates than smaller ones.
`
`However, none Of the equations so far pre—
`sented in this paper give the investigator any in—
`dication as to whether an injected molecule would
`be absorbed primarily via
`capillary vessels,
`
`Journal of Pharmaceutical Sciences
`
`lymphatic vessels, or both. Table I summarizes
`some of the literature relating molecular or for-
`mula weights of injected substances and their
`probable primary absorption routes. From the
`subcutaneous site it appears that molecules or
`ions having low molecular weights are absorbed
`primarily via the capillaries, while molecules hav-
`ing high molecular weights appear to be absorbed
`primarily via lymph vessels. Sund and Schou
`(20) followed the clearance rate of labeled carbo-
`hydrates from rat muscle. Table II lists the sub-
`stances used, their molecular weights, aqueous
`diffusion coefficients, and the fraction of drug
`cleared 5 min. after the injection. As expected, it
`can be seen from this table that with an increase
`
`in molecular weight, there is a decrease in clear-
`ance rate. The fact that the fraction cleared at
`the end of 5 min. correlates with the diflusion co-
`
`efficient, "they state, is evidence that diffusion is
`the main driving force for the absorption. These
`investigators did not determine experimentally
`which of the molecular species was cleared primar-
`
`TABLE I—RELATIONSHIP BETWEEN MOLECULAR OR
`FORMULA WEIGHT OF CHEMICAL SPECIES AND
`PROBABLE ROUTE OF ABSORPTION FOLLOWING AN
`INTRAMUSCULAR 0R SUBCUTANEOUS INJECTION
`Probable
`Primary
`Route of
`Admini5< Absorption
`tration
`Route
`i.m.
`Capillary
`s.c.
`Capillary
`s.c.
`Capillary
`s.c.
`Lymphatic
`s.c.
`Lymphatic
`
`M01. or
`Formula
`Wt.
`58
`>334
`270
`?
`>20,000
`
`Ref .
`(145)
`(38)
`(146)
`(146)
`(38)
`
`M01.
`Species
`“NaCl
`Strychnine
`(salt?)
`59FeCla
`“Fe-labeled
`plasma
`Black tiger
`snake
`venom
`2500—
`India cobra
`4000
`venom
`Russell viper ~30,000
`venom
`Diphtheria
`~70 , 000
`toxin
`Tetanus toxin
`?
`Iron poly~
`10,000—
`saccharide
`20 , 000
`complexes
`Lymphatic
`Neolymphius“ High
`~16%
`Iron—sorbitol— <5000
`lymphatic
`citrate
`~50—60%
`complexes
`capillary
`“The neolymphins are neomycinpolymetacrillate, neo-
`mycindextransulfate, and neomycincarboxymethyl starch.
`
`s.c.
`s.c.
`s.c.
`s.c.
`i.m.
`
`i.m.
`i.m.
`
`Capillary
`Lymphatic
`Lymphatic
`Lymphatic
`Lymphatic
`
`(38)
`(38)
`(38)
`(38)
`(37)
`
`(24)
`(147)
`
`TABLE Il—CARBOHYDRATE ABSORPTION FROM
`RAT MUSCLE (20)
`Fraction
`of Drug
`Cleared
`5 min.
`After i.m.
`Injection
`X 10
`~7
`NB
`
`Mol.
`Wt.
`182
`342
`
`Aqueous
`Diffusion
`Coefficient
`X 106
`8.7
`7 .5
`
`Substance
`D—Mannitol-
`1—14C
`SucroseJ‘C
`Inulin-_
`methonyH
`Inuliu-
`carboxyl-“C
`Dextran—
`carboxyle “C
`
`3000—4000
`3000—4000
`60 , 000—90 , 000
`
`2 . 1
`2 . l
`~0 . 5
`
`~12
`~2
`~0 . 7
`
`AstraZeneca Exhibit 2115 p. 4
`
`

`

`Vol. 57, No. 3, March 1968
`
`ily by the capillary or by lymphatic routes, al-
`though they pointed out
`that
`the drainage
`through lymph channels could be of relatively
`greater importance for the absorption of large
`molecules compared to smaller ones.
`Lewis (23) attempted to explain why sudden
`anaphylactic deaths occasionally occurred in hu<
`mans after subcutaneous injections of macro-
`molecules like diphtheria antitoxin when absorp—
`tion of this substance from the subcutaneous site
`was known to be slow. The test substance he
`used was horse serum. Table III shows that the
`
`serum was absorbed slowly from the subcutaneous
`region of dogs, probably because the absorption
`took place largely via the lymphatics. Massage
`of the wheal produced by the injection and the
`use of high injection pressures resulted in a more
`rapid appearance of the serum in the thoracic
`duct lymph. Lewis concluded that the serum’s
`absorption rate was too slow to account for cases
`of anaphylactic death in humans, and that the
`probable reason for this phenomenon was an ac-
`cidental intravenous injection of the serum.
`The effect that molecular size or weight can
`have on the route of drug clearance from an in—
`jection site has been demonstrated by Malek and
`co—workers (24). They prepared salts of an anti-
`biotic base like streptomycin or neomycin with
`high molecular weight anionic substances like
`polyacrylic acids,
`sulfonic or phosphorylated
`polysaccharides, and natural polycarboxyl acids.
`After an intramuscular injection of neomycin sul-
`fate into a dog, a peak blood level of about 20
`meg/ml. was obtained at 2 hr., which fell to zero
`at 12 hr. After an intramuscular injection of
`neomycindextransulfate (neolymphin II), a peak
`neomycin blood level of about 3 meg/ml. was
`obtained at 2 hr., and its concentration in the
`blood remained nearly constant at 1 meg/ml.
`from 8 to 24 hr. They termed these macro-
`molecular salts “antibiolymphins.” According
`to the authors,
`the antibiolymphins were ab-
`sorbed from the injection site primarily via the
`lymphatic system in contrast to the corresponding
`sulfate salt, which presumably was absorbed via
`TABLE
`III—~APPEARANCE OF
`SUBCUTANEOUSLY
`ADMINISTERED HORSE SERUM IN LYMPH AND BLOOD
`OF Docs UNDER VARIOUS CONDITIONS (23)
`Time to Detect
`r—Presence of Horse Serum infi
`Thoracic
`Duct Lymph
`40 min.
`...“
`15~20 min.
`
`Conditions
`No massage at
`site
`Massage at
`site
`40 min.
`<5 min.
`High-pressure
`injection
`0' Thoracic duct not cannulated.
`
`Blood
`3.5 hr.
`1.5hr.a
`2 hr.
`
`361
`
`the capillaries, although the authors did not state
`this.
`
`prMadison and Christian (25) studied
`the influence of pH on the absorption rate of
`22NaCl and 23NaCl administered subcutaneously
`in rats. The magnitude of the absorption rate
`was evaluated on the basis of the amount of
`
`radioactivity present in a sample of heart blood
`removed 45 min. after the injection. From pH
`2.5 to 10 inclusive there was little effect on the
`
`normal absorption rate of sodium ion. At pH
`values of 1.0 and 2.0, a decrease in sodium ion ab-
`sorption rate was observed, while at pH values of
`11.0 and 12.0 an increase in rate occurred.
`
`In the case of organic bases, the pH of the in-
`jected solution can often have a profound effect on
`absorption rate. Cutts and Walker
`(26)
`re-
`peated the work of White and Claflin (27). They
`confirmed that in mice intraperitoneal injections
`of
`the nitrogen mustard HN2, methyl-bism-
`chloroethyl)—amine hydrochloride, at two differ—
`ent pH values resulted in different LDw values.
`At pH 2 the LDw was about 5 mg. /Kg., while at
`pH Sit was about 2 mg. /Kg. White and Claflin
`(27) showed under their injection conditions that
`little HNz would have undergone cyclization.
`Since HNz has a pK' of 6.45 at 15° (28),
`the
`molecule exists almost entirely as the water-
`soluble protonated form at pH 2. But at pH 8 a
`significant fraction of the amine exists in the un-
`dissociated form. On the basis of pH-partition
`considerations (29),
`the drug’s absorption and
`distribution at the higher pH value should be
`facilitated resulting in the lower value found for
`the LDm.
`
`PHAGOCYTOSIS
`
`Although passive difiusion is an important
`mechanism for drug absorption,
`the process of
`phagocytosis may also be involved in drug ab-
`sorption from subcutaneous and intramuscular
`sites. Rees and co-workers (30) studied the sys—
`temic distribution of dimethylpolysiloxane,2 fol—
`lowing its subcutaneous administration, in mice.
`They stated that while the mechanism of absorp—
`tion and systemic distribution of the silicone fluid
`in mice is at present unknown, it may be distrib-
`uted to the viscera by gaining entrance to the
`general circulation (via capillaries?) or lymphatic
`channels. However, the authors suggest that the
`most
`likely mechanism is by the process of
`phagocytosis by histocytes.
`When sodium urate crystals are injected sub-
`cutaneously into animals and man, an acute in-
`flammatory response occurs accompanied by
`phagocytosis of the crystals by mononuclear and
`I Dow Corning MDX 44011.
`
`AstraZeneca Exhibit 2115 p. 5
`
`

`

`362
`
`polymorphonuclear leucocytes (31). A similar
`acute inflammatory response occurs after sub—
`cutaneous injections of microcrystalline ammo-
`nium and potassium urates, uric acid, xanthine,
`hypoxanthine, calcium carbonate, and creatinine.
`Seegmiller and co-workers (32) have shown that
`when microcrystalline sodium urate and sodium
`orotate suspended in a 50% glycerol~water vehi-
`cle were injected subcutaneously in man, 24 hr.
`later the site in most subjects showed warmth,
`swelling, erythema, induration, and tenderness.
`There was essentially no reaction to control sub—
`cutaneous injections of the vehicle after the im-
`mediate pain and wheal had subsided 6 to 8 hr.
`after the injection. Malawista and Seegmiller
`(33) have shown, when humans were pretreated
`with colchicine,
`the inflammatory response to
`subcutaneously administered monosodium urate
`monohydrate and sodium orotate microcrystals
`was substantially reduced over controls not so
`pretreated. Malawista (34) has
`shown that
`colchicine interferes with the amoeboid motility of
`individual human polymorphonuclear leucocytes.
`Other substances like actinomycin D, 6-mercap-
`topurine, and puromycin have been found to
`block lymphocyte emigration into a region of in-
`flammation (35). An interesting but as yet un—
`explained finding is that the inflammatory re-
`sponse to subcutaneous injections of amorphous
`sodium urate particles was much less than that
`found with microcrystalline preparations of the
`same salt (32).
`showed that when micro-
`(36)
`Gomez—Luz
`crystals of insoluble oxytetracycline were admin-
`istered subcutaneously into man, crystals could
`be found in the peripheral blood either intracel-
`lularly or free and in human bone marrow 6 to 14
`days following the injection. Patients suffering
`from nonspecific urethritis received the insoluble
`oxytetracycline subcutaneously, and microcrys—
`tals were found in their urethral discharges.
`Gemez—Luz explained these observations by sug-
`gesting that polymorphonuclear leucocytes and
`monocytes engulfed the microcrystals by the
`process of phagocytosis, and these crystals then
`were transported to the general circulation and
`finally to the site of infection.
`(37) showed that
`Beresford and co—workers
`most of the absorption of iron polysaccharide
`complexes injected into rabbit muscle tissue oc-
`curred during the initial 72 hr. The absorption
`during this time was mediated in part by the in-
`flammatory reaction evoked, and in part by
`lymphatic transport of the iron complexes. The
`iron complexes remaining after 72 hr. were fixed
`by tissue macrophages that seemed to impede
`further absorption.
`
`Journal of Pharmaceutical Sciences
`BIOLOGICAL FACTORS
`
`There are several biological factors, in addition
`to those mentioned, which can influence the rates
`of drug absorption.
`Some of these factors in-
`clude body movement, pre—existing disease con-
`ditions, the age of the animal, the anatomical re-
`gion into which the injection is made, and the
`condition of the tissue at the site.
`Body Movement—If a molecule is absorbed
`primarily via the lymphatic route, body move-
`ment or exercise can often have a profound
`effect on how rapidly it reaches the general
`circulation. Muscular movement has
`long
`been known to increase the flow of lymph
`fluid along the lymphatic vessels. When an
`animal is at rest, little or no lymph flows (5).
`The effect that body movement has on the
`absorption rate of a substance in solution can
`be seen from the work by Barnes and Trueta (38).
`They injected black tiger snake venom subcu-
`taneously into rabbits. Since this venom has a
`relatively large molecular weight (>20,000), it is
`transported from the injection site almost exclu-
`sively by the lymphatic drainage. Control ani-
`mals lived no longer than 150 min. following the
`injection, while animals whose injected limb was
`immobilized by a plaster cast lived more than 8
`hr.
`
`Controlled physical activity has also been
`shown to have an effect on the rate at which sub-
`
`cutaneously implanted procaine penicillin G im-
`plants were absorbed in rats. Ballard (39) found
`that the mean absorption rate per mean pellet
`area3 in animals rotated in a rodent activity cage
`at 3.83 r.p.m. was 2.57 X 10—3 Gm. hr.‘1cm._2,
`while that of nonrotated control animals was
`1.93 X 10—3 Gm. hr.“1 cm. ‘2. These differ-
`
`ences were shown to be statistically significant.
`Disease Conditions—Bauer and co—workers
`
`(40) noted that patients with cardiac failure
`and patients with myxedema had prolonged
`absorption half-lives following a jet injection of
`13’1 solutions into their thighs. When the myx-
`edemic condition was reversed by appropriate thy-
`roid medication, the absorption half—life of the
`drug returned to the normal range. De Groot
`and co-workers (41) noted delayed intramuscular
`absorption of a triiodothyronine formulation in
`one hypothyroid subject
`studied. Using the
`same jet injection technique, Bauer el al.
`(42)
`studied the absorption half-life of 1311 from the
`thighs of patients suffering from myocardial in-
`farctions. None of the patients had congestive
`heart failure.
`In one patient the half-lives were
`3 An error in the caption at the top of column five of Table I
`(39) indicated that all the rates per unit area were tenfold
`slower than they should be.
`
`AstraZeneca Exhibit 2115 p. 6
`
`

`

`Vol.57, No. 3, March 1968
`
`60, 29, 14, and 8 min, respectively, on the first,
`eighth,
`fourteenth, and thirty-first days after
`the infarction.
`In healthy controls the mean
`absorption half-life for ”11 was 9 (range:
`6—14)
`min.
`Irons (43)
`followed the blood levels of
`penicillin in patients suffering from a variety of
`bacterial infections after injections of an aqueous
`suspension of procaine penicillin G. He com-
`mented that a depression of penicillin blood levels
`tended to occur in a patient with severe cardiac
`failure because there was poor absorption of the
`drug from the injection site.
`Animal Age—~Lee (44) studied the effect of
`age on the acute toxicity of subcutaneously
`administered chlorpheniramine maleate and
`diphenhydramine hydrochloride in rats. He
`found that the LD50 of chlorpheniramine during
`the first 16 days of life varied between 175 and 230
`mg. /Kg., but after 25 days of life the LDw was
`360—365 mg./Kg. A similar trend was found for
`diphenhydramine. There was no apparent dif-
`ference in the activity of enzymes concerned with
`the metabolism of both drugs in the livers of
`15 and 40—day-old rats. Lee concluded that the
`development of greater resistance to both drugs
`with increasing age could be explained by a de-
`crease in the amount of drug absorption from the
`subcutaneous
`region.
`The reason why age
`should influence drug absorption from the subcu—
`taneous region is not known at present. How-
`ever, if one may speculate, there are at least two
`possible reasons for this phenomenon.
`Ballard and Menczel (19) studied the subcu-
`taneous absorption half-life of aqueous benzyl
`alcohol solutions in rats. The mean absorption
`half-life for four animals was 1.6 hr., but the
`range was from 0.96 to 2.58 hr. under conditions
`where the tissue area exposed to the drug solu-
`tion was constant. They suggested that one
`probable reason for the large interanimal vari-
`ation in t” could be attributed to differences in
`the membrane thickness exposed to the drug
`solution in the abdominal region for the several
`animals. It is well known that the thickness of the
`
`subcutaneous tissue among humans is not con-
`stant (45).
`In Eqs. 4 and 6 it can be seen that
`the penetration coefficient
`is inversely propor-
`tional to 6, while t0.5 is directly proportional to 6.
`It is reasonable to suppose that as young animals
`age the thickness of their subcutaneous tissue may
`increase as well, resulting in increases in t” and
`increases in LD50 values for drugs whose acute
`toxicity can be correlated with drug blood-level.
`Another hypothesis
`among several others
`deals with the question: what
`influence does
`animal age have on the chemical composition of
`subcutaneous fat tissue? The penetration rate
`
`363
`
`of a lipid-soluble drug should be directly propor—
`tional to the lipid/water partition coefficient of
`the drug.
`(See Eq. 4.)
`If the chemical com-
`position of this tissue changes with age, and the
`apparent partition coefficient of the drug differs
`in younger and older tissue, then variations in
`absorption rate would be expected. Using gas
`chromatographic analysis Bondrup et al.
`(46)
`studied the triglyceride composition of subcu-
`taneous fat obtained from premature stillborn and
`full«term stillborn children, and normal, lean, and
`obese adult patients