`
`The Official Publication of the American
`Association of C(illeges of Pharmacy
`
`TEVA - EXHIBIT 1019
`
`
`
`Editorial office: School of Pharm acy, University
`of N o rth Carolina, C hapel Hill NC 275 14.
`Telephone: (91 9) 966-1121. Address all com(cid:173)
`munications co ncerning manuscripts
`to
`the
`editorial office. All manuscripts submitted are
`subject to peer-rev iew a nd a pproval by the editor
`before publication. M anuscripts publi shed
`elsew here cannot be considered. Authors must
`prepare their manuscripts to conform to the style
`specificat ions published on pages 88-90 of the
`February 1980 issue, otherwis e such are subject
`to return. The Journal does not assess authors o r
`their institutions for set-up or page charges . All
`repo rts and papers prese nted at the AACP An(cid:173)
`nual Meeting as well as those manuscripts accep(cid:173)
`ted for publicati on in the Journal become the
`prope rty of the AAC P. All requests for reassign(cid:173)
`ment of the abo ve rights should be directed to the
`business manager at the address below.
`
`All expressions of opinion and statements o f su p(cid:173)
`posed fact appearing in the Journal are not to be
`regarded as necessa ril y expressing the policies or
`views o f the editor or of the American Associa(cid:173)
`tion of Colleges of Pharm acy.
`
`in February, May, A ugu st and
`Publis hed
`November. Subscripti on price $35.00. All new
`subscriptions begin with the February issue. In(cid:173)
`di vidual membership in the AACP is $35.00
`yearly whi ch entitles o ne to a subscription to the
`Ameri can Journal of Pharmaceutical Education
`a nd a subscripti on to the monthly AACP News
`plus participation privileges. Any full-time
`under-
`graduate o r gradu ate student attending an
`AACP-affiliated school or college of pharm acy
`may subscribe to the Journal at the special stu(cid:173)
`dent rate o f $15.00. A ddress a ll subscriptions,
`change of address and all business com munica(cid:173)
`tion s to the business office.
`
`Articles in the J ournal are indexed in : Biological
`Abstracts; Chemical Abstracts ; Currelll Contents.
`Education; Current Contell/s, Life Sciences; Ex(cid:173)
`cerptaMedica; Hospital Literature Index; Inter(cid:173)
`national Pharmaceutical Abstracts; a nd Iowa
`Drug Info rmati on Service a nd aremicrofilmed by
`Univers ity Microfilms of Ann Arbor MI.
`
`Business office: 4630 Montgomery ave., Suite
`201, Beth esda MD 20014. Telep hone: (301) 654-
`9060 . The Association cannot accept respo n(cid:173)
`sibility for foreign delivery when its reco rds in(cid:173)
`dicate shipment has been made.
`
`Postmaster: Undelivered co pies should be retur(cid:173)
`to: Ameri can J ourna l of Phanna(cid:173)
`n e d
`ceutical Education, School o f Pharmacy, Univer(cid:173)
`sity o f North Carolina, C hapel Hill NC 275 14.
`
`Co pyri ght © 198 1, American Association of
`Colleges o f Pharmacy, 4630 Montgomery Ave.,
`Suite 201 , Bethesda MD 20014. All rights reser(cid:173)
`ved . Second-class pos tage paid at Bethesda MD
`an d at add itional mailing office . Printed by
`Creative Printers, 503 W. Franklin St., Chapel
`Hill NC 275 14. Telephone: (919) 929- 11 21.
`
`ii
`
`REPORTS
`71 Minutes of the Meeting of the Board of Directors
`RODOWSKAS, Christopher A ., Jr.
`
`PRESIDENT'S SECTION
`75 Food For Thought
`SORBY, Donald L.
`
`COUNCIL OF DEANS CHAIRMAN'S SECTION
`76 Political Action Revisited
`GRANBERG, C . Boyd
`
`COUNCIL OF FACULTIES CHAIRMAN'S SECTION
`77 Obtaining Advanced Degrees in Pharmacy by Nontraditional Means
`LOWENTHAL, Werner
`
`COUNCIL OF SECTIONS CHAIRMAN'S SECTION
`78 Role-Model Congruence - A Catalyst to Learning
`LEMBERGER, Max A .
`
`EDITOR'S SECTION
`79
`Instructions to Authors
`COCOLAS, George H.
`
`ANNOUNCEMENTS
`81 General Announcements; Changes in Staff Titles; New Staff Members
`COCOLAS, George H.
`
`THE RECORD
`83 General News; Grants and Awards
`COCOLAS, George H .
`
`Jrn
`
`:!vi
`rsh.
`111
`'igh
`.n .
`i~e
`ch:
`io
`b e
`?ig.
`
`er~
`So•
`:ly'
`ive
`ri s
`::CI
`
`BOOK REVIEWS
`87 GUILLORY , K.J., LONGE, R.L., SNOW, B., KILSDONK , G.F., ABOOI
`R.R., GUMBHIR, A.K., HUNT, M.L., LAWRENCE, G .D ., IANN/,1
`RONE, M ., SPEEDIE, M .K ., RITSCHEL, W .A., ABRAMSON, H
`KELLEY , C.J., SOWELL, J.W., THOMPSON, E.B., LOEFFLER, L.fu
`BARFKNECHT, C.F., CLARKE, D .E., GRINGAUZ, A., JUN, H.~'v1
`TIMMONS , H .F., STAUBUS, A.E., CARLSTEDT, B., SIDDONS, L.~11~
`JOHNSON, H .D ., BEAMER, R.L., WEART, C.W ., BARLETTA, J.~
`FLAGSTAD, M.S., RUSSI, G., MITSCHER, L.A., LAMY, P.P., COKER, S:1~
`RECENT PUBLICATIONS
`104 New Journals and Special Publications; New Books
`COCOLAS, George H.
`
`rr
`
`·j
`
`CORRIGENDUM
`106 COCOLAS, George H .
`
`ADVERTISEMENTS
`(iii)
`Pfizer Pharmaceuticals
`E. R. Squibb & Sons, Inc.
`(iv)
`(vi)
`Parke-Davis Division
`(vii)
`Prentice-Hall Publishers
`(vii)
`McNei l Labora tories
`(viii)
`Roche Laboratories
`CIBA
`(x)
`Eli Lilly and Company
`(xi)
`(xii)
`Ortho Pharm aceutical Corporation
`The Upjohn Company
`(xiii)
`Burroughs Wellcom e Company
`(xiv)
`Abbott La boratories
`(xv)
`A merican Association of Colleges of Pharmacy
`(xvi)
`
`
`
`Calculation of Drug Solubilities by Pharmacy Students
`
`Lindley A. Cates
`College of Pharmacy, University of Houston, Houston TX 77004
`
`A method of estimating the solubilities of drugs in water is reported which is based on a principle applied
`in quantitative structure-activity relationships. This procedure involves correlation of partition coefficient
`values using the octanol/water system and aqueous solubility. After identifying the atoms or groups com(cid:173)
`prising a compound the students need to employ but a few approximate hydrophilic or lipophilic numbers
`assigned to these in calculating the log P value of the drug or chemical and then place the agent in the ap(cid:173)
`propriate soluble or insoluble category. Although this method does not always provide exact categoriza(cid:173)
`tion it does so in a great majority of cases and permits the student to recognize certain potential chemical
`and therapeutic incompatabilities.
`
`"
`
`One of the most frequent questions asked medicinal
`chemistry faculty by pharmacy students is, "How do I
`know if a drug is soluble or insoluble in water?" These stu(cid:173)
`dents were cognizant or the importance of such informa(cid:173)
`tion
`in predicting chemical and
`therapeutic
`incom(cid:173)
`patabilities. Prior to the introduction of the procedure
`described in this paper, along with a discussion of the acid(cid:173)
`base character of drugs, many students were incapable of
`determining if an insoluble material will be formed during
`a reaction and whether or not water can be used as the sol(cid:173)
`vent for a certain drug. A method, therefore, was devised
`to enable the students to estimate a drug's solubility by
`assigning a numerical value to a molecule which relates to
`this property. This procedure, which is based on a scientific
`rationale, requires the use of only a few numerical values
`and a brief calculation time.
`
`PROCEDURE
`With
`the advent of quantitative structure-act1v1ty
`relationship (QSAR) concepts has come an increased
`awareness and use of physiochemical parameters such as
`partition coefficients and steric and electronic factors for
`correlation with biologic properties. The former constant is
`deemed most critical to a drug's overall effect and most
`students are exposed to this principle during their phar(cid:173)
`macy education.
`Most work has been done with partition coefficients
`based on the octanoljwater system expressed as the log10
`or log P. Although this is a measure of the solubility
`characteristics of the whole molecule, one normally uses
`the sum of the fragments of the molecule which have been
`assigned relative hydrophilic-lipophilic values, (1r),
`to
`calculate log P. Using this procedure, a positive value for
`7r means the substituent, relative to H, favors the octanol
`Phase (i.e., lipophilic). And negative 1r value indicates its
`greater affinity for water (i.e., hydrophilic). 1 The environ(cid:173)
`ment of the substituent can influence the relative 1r value,
`but, for the most part, such changes are small and can be
`neglected for our purposes .
`The method of calculating log P values of drugs was
`
`introduced while teaching a course in the medicinal chem(cid:173)
`istry sequence to about 80 second professional year stu(cid:173)
`dents and then applied to those agents being discussed
`throughout the semester. The students learned eight 1r
`values as follows : C (aliphatic or C J2) = 0.5; phenyl = 2.0;
`C(O)O or C(O)N = -0.7; 0 or N (in amines, hydroxyls and
`ethers but not in hydrazines or N-0 compounds) = -1.0;
`and -S- = 0. These numbers were obtained by rounding off
`literature values; exceptions being the sulfide3 and the
`amido group. 4 The students then needed only to identify
`these fragments in the molecule and calculate the sum of
`the 1r values to calculate an approximate Jog P. After being
`given several examples in lecture and solving problems
`themselves at the chalkboard all students knew the 1r values
`and the only difficulty they occasionally experienced was
`identification of the appropriate fragments in a molecule
`(e .g., in cyclic drugs).
`The USP provides official definitions of water
`solubilities wherein "soluble" is defined as 3.3 to 10 per(cid:173)
`cent. For our purposes
`therefore,
`those drugs with
`solubilities above 3.3 percent are considered soluble and
`those below, insoluble. The solubilities of drugs used in this
`paper were taken from the Merck Index(!), Remington's
`Pharmaceutical Sciences(2) or the Handbook of Chemistry
`and Physics(3). The octanoljwater Jog P values were from
`Hansch and Leo( 4) and the 1r values are from three dif(cid:173)
`ferent sources( 4, 5 and 6).
`Having a definition of solubility and a means of
`calculating log P, what remains is a method of correlating
`these two parameters. Through the examination of a large
`number of log P and solubility values, an arbitrary stan-
`
`'The term "1r" more correctly refers to the system of substituting atoms or
`groups for hydrogen while the fragment system of calculating log P
`values involves the summing of appropriate structural elements. In ap(cid:173)
`pro ximating log P values this distinction normally is not critical.
`' Average of 0.71 (aromatic) a nd 0.39 (aliphatic) values.
`3Taken from aliphatic SCH 3 (0.45) and aromatic SCH 3 (0.61 ), each minus
`a methyl (0.5).
`4 Although the literature 1r values are -1.49 (aromatic) and -I. 71 (aliphatic)
`the value of -0.7 gives more correct results when using the approximate
`calculation method (q. v .. barbiturates, phenacetin , dibuca ine, nicotin(cid:173)
`amide and phenoxymethyl penicillin).
`
`American Journal of Pharmaceutical Education
`
`Vol. 45, Feb. 1981
`
`11
`
`
`
`dard was adopted whereby those drugs with positive log P
`values over 0.5 are considered water-insoluble and those
`with less than 0.5 log Pare deemed soluble. An early use of
`log P and 1r in correlating chemical structure with aqueous
`solubility
`involved
`free-energy changes of
`liquids(7).
`Although this study included only four of !56 compounds
`with log P values less than 0.5, the dividing line between
`soluble and insoluble appears to be in the same range.
`Although this method is applicable to a large number of
`drugs it is, of course, restricted to those containing only C,
`C I, N and 0. Other limitations should also be recognized,
`chief among these is the acid-base character of drugs.
`When dealing with acids or bases, log P values are normally
`determined at a pH , either very acid or alk a line, so that
`ionization is suppressed and only the neutral, most lipo(cid:173)
`philic form is present. Since most drugs are either weak
`acids or bases this possible discrepancy must be taken into
`consideration. Scherrer and Howard(8) have pointed out
`that when a n ionizable compound is equilibrated in a two(cid:173)
`phase system at a pH at which it is partially ionized, its
`concentration in the organic phase is not determined by log
`P alone . These investigators, therefore, introduced dis(cid:173)
`tribution coefficients (log D) as a correction based on the
`pKa of the compound . Log Dis also termed the apparent
`partition coefficient (P app). which is in turn related to the
`true (corrected) partition coefficient (Pcorr), Pcorris equal
`to Papp/ (1-a), where a
`is
`the degree of ionization.
`Although this correction is not readily adaptable to our es(cid:173)
`tim ation method there have been a few drugs whose log P
`values have been determined using octanol j water at pH
`values that approximate those imparted to water by these
`agents and some examples will be presented later.
`
`APPLICATIONS AND DISCUSSION
`
`This method was initially applied to those CNS drugs
`covered in the medicin al chemistry course and described by
`the textbook(9). This consisted of 29 sedative-hypnotics,
`six central relaxants, three benzodiazepines, 14 pheno(cid:173)
`thiazines, 12 anticonvulsants and 12 miscellaneous drugs.
`A few more such agents could have been included if 1r
`values for Br and F were introduced. Of this total the
`solubilities of 72 (95 percent) were correctly determined
`with three anomalies and one 'borderline' estimation. The
`success rate with this classification of drugs is not unexpec(cid:173)
`ted in view of the relationship between their log P values
`and depressant activity. It has been established that most
`organic drugs affecting the CNS require a log P of approx(cid:173)
`imately 2 to pass the blood-brain barrier and gain access to
`the brain( I 0). A partition coefficient of this magnitude
`would translate to a water-insoluble compound . Consider(cid:173)
`ing all drugs , there are relatively few that are soluble in the
`non-salt form, a situation that should make easier the
`teaching of solubilities. To simply state that drugs in their
`free form are insoluble is not, however, satisfactory . In ad (cid:173)
`dition, there are situations when it is important to be cogni(cid:173)
`zant of relative solubilities, a comparison made possible us(cid:173)
`ing the calculation method.
`Since one of the larger members of this class of CNS
`depressants are barbiturates it might be instructive to ex(cid:173)
`amine the heterocycle common to these and at least one
`specific agent. Being cyclic ureides the barbiturates contain
`two amido groups (-0.7 each), two carbons (0.5 each) and
`one carbonyl oxygen (-1.0) for a log p of -1.4. This com(cid:173)
`pares favorably with the -1.35 reported for this barbituric
`
`is
`ac id portion( I 0) . The most water-soluble drug
`diethylbarbital with a calculated log P of 0.6. This agent is
`also one of the few weakly acidic drugs whose log P has
`been determined in octanol j water at other than a low pH .
`At pH 8. 1 its log P value is 0.18 and has a log P value of
`0. 71 at pH 5. An environment closer to neutrality on the
`acid side would have been preferred for our comparison
`purposes but the value falls in the insoluble range ac(cid:173)
`cording to the established definition; the actual solubility is
`0.7 percent.
`An examin ation of the anomalies and 'borderline'
`drugs , which fall in the anticonvulsant and central relaxant
`classes, may also be of interest ethosuximide, containing
`a n amido group, six carbons and a carbonyl oxygen, has an
`estimated log P of 1.3 but is water-soluble. Trimethadione,
`a neutral drug with one each amido and carboxy groups
`and six ca rbons , calculates to 0.6 and is 5 percent soluble
`and is considered 'borderline' . The carbamates methocar(cid:173)
`bamol, 2.5 percent soluble, and chlorphenesin carbamate,
`almost insoluble, have simplified and incorrect calculated
`log P values of -0.7 and 0.3, respectively . This discrepancy
`can be accounted for on the basis that the infrequently en(cid:173)
`countered carbamyl moiety actually has a 1r value of -1.15
`instead of the -1.7 used and an aromatic methoxy 1r value is
`-0.2 as compared to our value of -0.5 . This situation exem(cid:173)
`plifies the errors that ca n be introduced when an attempt is
`ma de to simplify the calculation process.
`The method was completely successful when applied to
`the 35 local anesthetics described in the student's text. In(cid:173)
`terestingly, the ba sic drug procaine which was recorded log
`P (octanol j water) values of -0 .32 (pH 7) and 0.14 (pH 8) is
`only 0 .5 percent soluble. Our calculated value places this
`drug in the correct water-insoluble category. The proce(cid:173)
`dure was also correct in assigning 34 analgesics and
`analgesic antagonists, 30 antihistamines and 25 nonquater(cid:173)
`nized autonomic blocking agents, all water-insoluble.
`The salicylic acid derivatives, aspirin, salicylamide and
`salicyclic acid itself, were also examined. A true calculation
`of the latter requires the introduction of an additional 1r
`value, that for intramolecular hydrogen bonding (IMHB).
`Without
`this factor salicylic acid
`log P value easily
`ca lculates as 0.3 but is only 0.2 percent soluble. If the 0.65
`IMHB value is added we get a log P value of 0.95 which
`places it
`in
`the correct water-insoluble category . The
`literature value for thi s acid is 0.95 (pH 4); the pH of a ·
`saturated solution is 2.4. The need for applying the IMHB
`factor is infrequent but can be used during instruction in
`emphasizing this phenomenon which is of importance in
`biological action. It could also be pointed out that the
`isomer, p-hydroxybenzoic acid , cannot undergo IMHB
`and is eight times more soluble. Salicylamide has very close
`values in all respects to salicyclic acid while aspirin, 0.3 per(cid:173)
`cent soluble, calculates to 1. 1 log P without IMHB. As was
`the case with procaine, our procedure gives correct
`"' ca tegori zation of solubility while the experimental log P
`values for aspirin of -0.02 (pH 5) and -0.9 (pH 5.6) would
`not. It appears that there are situations when exact, or even
`simplified , calculated values are more meaningful than ex·
`peri men tally derived ones. One reason for this is the many
`experimental va lues, supposedly measured under like con·
`ditions but in different laboratories, that may vary for the
`same compound by as much as 1.5 log units . A variance of
`0.5 units is common.
`Although there are few soluble drugs in the free form,
`two CNS depressants chloral hydrate and paraldehyde fall
`
`12
`
`American Journal of Pharmaceutical Education
`
`Vol. 45, Feb. 1981
`
`
`
`Table I. Additional representative solubilities
`LogP
`LogP
`Calculated
`Observed
`5.0
`5.3
`4.3
`4.2
`2.5
`3.1
`2.5
`1.8
`2.1
`2.4
`2.1
`2.1
`1.8
`1.5
`1.6
`1.6
`1.5
`2.4
`1.5
`1.4
`1.3
`2.0
`0.5
`0.4
`0.5
`0.3
`0.5
`0.3
`-0.3
`0.0
`-0.2
`-0.4
`-0.7
`-0.6
`-1.7a
`-1.5
`-1.6
`-1.7
`
`Literature
`Predicted
`solubility
`solubility
`~or chemical
`Jb
`CtiiOriJromazine
`I
`I
`I
`Dibucaine
`I
`I(1.5%)
`phenytoin .
`sse
`I
`Amphetamme
`I
`phenoxymethyl penicillin
`I (0.08%)
`I
`I (0.08%)
`Amobarbital
`I (0. 1%)
`I
`phenacetin
`I
`I(O. l %)
`Phenobarbital
`I
`Parachlorophenol
`SS(2.7%)
`I
`I (0.57%)
`Ethyl chloride
`I
`I (0.33%)
`Benzoic acid
`ss
`BC
`Thiazole
`s
`B
`Propanol
`s (12.5%)
`B
`Acetylacetone
`s
`sU
`Ethanol
`s
`s
`Nicotinamide
`s
`s
`Lactic Acid
`s
`s
`Glycerol
`s
`s
`Citric Acid
`afrom Pomon a College Medicinal Chemistry Project dat a. bJnsoluble. CBo rderline solubility.
`
`dSoluble.
`
`estightl y soluble.
`
`in this category and may be considered exceptions to the
`SAR requirement. Chloral hydrate is highly ionized as a
`result of the inductive influence of the chlorine atoms and
`is very soluble. Neither its log P, nor that of paraldehyde,
`has been determined in o.ctanoljwater but it calculates by
`our method to 0.5 or 'borderline'. Doubtless the true value
`is considerably lower because of the halogen effect. The
`neutral paraldehyde calculates correctly giving a 0 log P
`and is 12 percent soluble. The solubilities of some ad(cid:173)
`ditional drugs and chemicals, arranged by
`increasing
`hydrophilicity and containing a variety of chemical group(cid:173)
`ings, are shown in Table I.
`After mastering the determination of drug solubilities
`using the eight constants the students will be able to
`proceed to drugs containing atoms or groups not yet con(cid:173)
`sidered. Examples of these are the nitro and nitrate groups.
`The former has a 1r value of -0.85 (aliphatic) and -0.28
`(aromatic) which can be averaged and rounded off to -0.6,
`and not the -0 .3 calculated by the previous method .
`Similarly the nitrate group, found in several vasodilators,
`has a 1r value of ca. 0.2 and not -4.0 .
`It should be emphasized that this simplified method of
`estimation has only general application and cannot,
`without becoming cumbersome, be applied with success in
`all cases . This is particularly the case when electronic fac(cid:173)
`tors play an important role. When examining the ampho(cid:173)
`teric antibacterial sulfonamides, for example, we find that
`the addition of one or two methyl groups to the pyrimidine
`
`o: sulfadiazine to give sulfamerazine and sulfamethazine
`
`Yields a progressive increase, instead of the expected
`decrease, in solubility. The effect of the methyl is to in(cid:173)
`crease the lability of the N 1 amide hydrogen and, thu s the
`molecule's hydrophilicity.
`
`In general, the log P of heterocycles, such as those
`found in sulfonamides, can be estimated by subtracting 0.5
`from phenyl ( 1r = 2.0) or naphthalene ( 1r = 3.4) for each
`carbon substituted by a heteroatom and adding the 1r value
`for the latter. Thus, the calculated 1r values for pyridine
`and isoquinoline are 0.5 (0.64 observed) and 1.9 (2.0 obser(cid:173)
`ved), respectively, and permit the solubility determination
`of such drugs as nicotinamide and dibucaine, Table I.
`This method of determining drug solubilities has been
`enthusiastically received by the pharmacy students in our
`medicinal chemistry course. It has done much to dispel
`confusion and to increase their confidence in dealing with
`drugs as chemicals capable of causing therapeutic and dis(cid:173)
`pensing problems. Provided its limitations are considered,
`it can be a useful tool in the teaching of an important and
`relevant topic.
`
`Am. J. Pharm. Educ .. 45, 11-13(1981 );
`11 / 18/80.
`
`received 9/ 3/80, accepted
`
`References
`(I) Th e M erck Index, 9th ed., Merck and Co. , Inc. Rahway NJ ( 1976).
`(2) Remington's Pharmaceutical Sciences. 15th ed., Mack Printin g Co.,
`Eas ton PA ( 1975).
`(3) Handbook of Chemistry and Physics, (edit.) Weast, R.C. , Chemical
`Rubber Pub. Co ., Cleveland OH (1977).
`(4) H a nsc h, C. and Leo, A. , Substitution Constants for Correlation
`Analysis in Chemistry and Biology. John Wiley and Sons, New York
`NY (1979).
`(5) Leo, A. , Han sch, C. and Elkin s, D ., Chem. Rev .. 71 , 552(1971).
`(6) Tute, M .S. , Adv. Drug R es .. 6, 67(1971).
`(7) Ha nsch, C., Quinlan, J .E. a nd Lawrence, G.L. , J. Org. Chem .. 33,
`347( 1968).
`(8) Scherrer, R .A . a nd Howa rd, S.M ., J. Med. Ch em .. 20, 53(1977).
`(9) Wil son, C.O., Gisvold, 0 . and Doerge, R.F. , Tex tbook of Organic
`Medicinal and Pharmaceutical Chemistry. 7th ed. J .B. Lippincott Co.,
`Phil adelphia PA ( 1977).
`(10) Daniels, T.C. and Jorgensen , E.C. , Op. cit. (9) pp. 22-23 .
`
`..
`
`American Journal of Pharmaceutical Education
`
`Vol. 45, Feb. 1981
`
`13