Am J Hosp Pharm.
1980; 37:95-8
Solubility and Stability of Phenytoin Sodium
when Mixed with Intravenous Solutions
Ronald R. Carmichael, Charles D. Mahoney,
and Louis P. Jeffrey
The stability of intravenous admixtures of phenytoin sodium
in dextrose 6% in water (D5W) and in 0.9% sodium chloride in­
jection (NS) was studied.
Phenytoin sodium solutions were prepared by mixing 100-
mg/2 ml solutions of drug with 25, 50,100, or 150 ml of D5W or
NS; pH was measured and solutions were inspected for crystals
for up to one hour after admixture. Samples of phenytoin ad-.
mixtures (100 mg/50 ml of D5W or NS) infused at a rate of 1
ml/minute through i.v administration sets (with or without a
1-pm fdter) were collected for up to one hour and analyzed by
using an enzyme-mediated immunoassay technique.
The pH of admixtures did not vary greatly over the 60-min-
ute period, but pH decreased with increasing dilution. Solubili­
ty decreased with decreasing pH; crystallization of phenytoin
solution occurred rapidly in D5W admixtures (pH 9.44-10.15)
but was not noted in NS admixtures (pH 9.82-10.81). Pheny­
toin concentration of D5W samples was significantly lower (p <
0.01) than that of NS samples after infusion through an i.v. ad-
minstration set; filtration had no significant effect on phenytoin
availability.
The results suggest thati^hen direct i.v, injection of pheny­
toin sodium is not practical, the drug thay be infused over a pe­
riod of ho more than one hour if used immediately after dilution
with no more than 50 ml of NS.
Index terms: Additives; Anticonvulsants; Dextrose; Hydrogen
ion concentration; Phenytoin sodium;Sodium chloride;Solubil­
ity; Stability
Phenytoin sodium (Na 5,5-diphenylhydantoin, molecu­
lar weight.274.25) was proven to have anticonvulsant prop­
erties as early as 1938 in experiments by Merritt and Put-
man. ^ Since that time the drug has been used extensively in
the treatment of epilepsy^^ and cardiac arrhythmias.®"'' For
.rapid control of seizures, such as status epiilepticus, clinicians
quickly realized the need for a more readily usable form of
phenytoin than an oral dosage form. Murphy and Schwab®
developed a diluent (40% propylene glycol, 10% alcohol, and
50% water for injection adjusted with sodium hydroxide to
pH 12) that enabled clinicians to administer the drug par-
enterally. Intramuscular injections have proven to be un­
desirable because of pain at the injection site and inadequate
serum levels.®''®
The intravenous route is not without its problems. First,
ill any solution witb a pH of 11.7 or below, phenytoin sodium
(equivalent to 91.98% phenytoin acid'') readily hydrolyzes.
Ronald R. Carmichael is Pharmacist, Medical College of Virginia, Rich­
mond. Charles D. Mahoney, M.S., is Assistant Director of Pharmacy Ser­
vices and Director of Eiiucation and Training, and Louis P. Jeffrey, M.S.,
is Director of Pharmacy'Services," Rhode Island Hospital, Providence
02902. , '
Address reprinLrequests to Mr. Jeffrey,
The assistance of Horace F. Martin, M.D., Ph.D., John Pezzullo, Ph.D.,
Paul A. Ullucci, and Dennis W. Welch is acknowledged. /
Presented at the 13th Annual ASHP Midyear Clinical Meeting, San An­
tonio, TX, December 5,1978.
Copyright© 1980, American Society of Hospital Pharmacists, Inc. All rights
reserved.
0002-9289/80/0101-0095$0.1.00
thereby generating insoluble pbenytoin.i^ Phenytoin (mo­
lecular weight 252.26) is a weak organic acid which is poorly
soluble in water but becomes increasingly soluble in aqueous
solutions as the pH increases. Dill et aP^ determined that
phenytoin, in aqueous solution at 26 °C, is soluble to the
extent of 14 jig/ml over the pH range of 1-7,165 jug/ml at pH
9.1 (in borate buffer), and 1.52 mg/ml at pH 10.0 (in sodium
hydroxide). In addition, the dissociation of phenytoin so­
dium, resulting in the precipitation of free phenytoin, can
be produced by numerous weak acids, including carbon
dioxide (CQ2) absorbed upon exposure to air.®4i.i2 Because
of these specific physical and chemical properties, the
Downloaded from https://academic.oup.com/ajhp/article-abstract/37/1/95/5199353 by guest on 07 November 2019
manufacturer recommends that the drug be given only by
direct i.v. injection.
The rate of administration is an important factor, because
too rapid administration can have adverse effects. Three
fatalities, resulting from rapid i.v. administration, have been
reported.i''-^^ Research in this area has indicated that the
propylene glycol solvent, combined with the initial additive
effects of phenytoin, is the cause of the problem and the
reason for the rate restriction. Numerous authors,'^''"^^ in
an effort to overcome this administration problem, have
deviated from the manufacturer's recommendation and have
proposed administering phentoin sodium by i.v. infusion in
various commercial i.v. solutions. A number of other au-
thors^'2®-^® have opposed administering phenytoin sodium
by any means other than that recommended by the manu­
facturer.
Objectives
The objective of this research was to determine by quali­
tative and quantitative techniques the effects on solubility
and stability of phenytoin sodium injection (Dilantin, Parke,
Davis and Company) upon mixing with dextrose 5% in water
(D5W) and 0.9% sodium chloride injection (NS).
Methods
Qualitative Analysis—pH Determinations. All experi­
ments were conducted in a laboratory setting at room tem­
perature. All glassware was cleaned with Contrad 70 Glass
Cleaning Solution and rinsed three times with distilled water
before use.
A 25-ml sample of D5W was withdrawn from a 1000-ml
bag using a 50-ml syringe and was injected into a 50-ml
beaker. The pH of thissolution was tested. Then, (previously
withdrawn) lOO-mg/2 ml phenytoin sodium solution was
injected into the P5W and the pH of the resulting admixture
was tested at 0, 5, 10, 15, 20, 25, 30, and 60 minutes after
a:dmixture. The same procedure was used to test the pH and
solubility of 100 mg of phenytoin sodium in 25-ml, 50-ml,
100-ml, and 150-ml volumes of D5W and of NS.
The pH and crystallization characteristics were studied
over a period of only one hour because, in almost all clinical
situations, 100 mg of phenytoin will be infused over a rela­
tively short period of time. In many instances, it is admin-
Vol 37 Jan' 1980 American Journal of Hospital Pharmacy 95
 Phenytoln sodium

istered within two minutes.

Quantitative Analysis—Drug Assays. All experiments
were conducted under the conditions described above.
Phenytoin sodium, in a concentration of 100 mg/2, ml, was
injected into a 150-ml capacity burette solution adminis­
tration fet with a membrane filter to which 50 ml (measured
by using calibration on burette) of D5W hai^en added;
Immediately after mixing by inverting thi burette three
. times, the first 1.0-ml sample was withdraX from the bur­ 4 • • • •
ette, using a 3-ml syringe with an 18-gauge, 1.5-inch heedle,
and was injected into a 10-ml volumetric flask. Then, with
an infusion flow rate of approximately 60 drops/minute (1
ml/minute), seven more LO-ml samples were collected. This
IS an enzyme-mediated immuhoassay technique that is based
on the principle that the drug in solution to be analyzed will-
standard test solution. The
EMIT assay has been found to be as accurate, precise, and
specific as other assay methpds. As an experimental quality
control check, tests were duplicated on a random basis and
all the variations^were found to be in the range of ±1
Mg/ml. • '
Results and Discussion
Qualitative. The pH values determined after injecting 100

Downloaded from https://academic.oup.com/ajhp/article-abstract/37/1/95/5199353 by guest on 07 November 2019

time the sample was collected at the end of the i.v. tubing mg of phenytoin sodium into various volumes of D5W and
(point of entry into the patient in a clinical situation) at 5, NS are shown in Table 1. From this table significant trends
10,15, 20, 25, 30, and 60' minutes after admixture.' can be distinguished. First, the'pH of the D5W solution is
noticeably lower than that of the NS solution used in the
As each sample was collected, it was diluted to 10 ml with
• qualitative experiments. A second point is the effect of
. 0.1 N N,aOH; this returned to solution any phenytoin that
may have precipitated. The dilutions were made using a dilution on the pH; as the volume of i.v. solution mcrcascd,
. 10-ml burette. The volume of NaOH used to dilute the se­ the pH of the resulting admixture decreased. For example,
ction to 10 ml was recorded. By subtracting this volume 100 mg of phenytoin sodium in 25 ml of NS at 0 minutes
from 10 ml, one could accurately determine the amount of produced an admixture with a pH of 10.81; in 100 ml of NS
i.v. solution taken as a sample. Once diluted, the eight the admixture had a pH of 10.08. Also, these same ad­
samples were analyzed using the EMIT assay technique mixtures had mean pH values of 10.75 and 10.02, respec­
(Sayya Corp., Palo Alto, CA). Immediately before per- tively. It should be noted that the pH of the admixtures did
not vary greatly over the 60-minute period.
, forming the assay, 1.0 ml of each sample was further diluted -
to 10 ml to get the sample in the range of 0 to 30 gg/ml The A visible precipitate of phenytoin crystals appeared in the
second diluent used was EMIT AED Buffer-Concentrate admixture solutions prepared with D5W as early as five
Tris-HCl with surfactant, diluted to 150 ml before use with minutes and no later than 25 minutes after admixture The
distilled water to make a 0.055 M Tris-HCl pH 7.9 buffer crystallization of phenjdoin in the 25 ml
solution. of D5W, as opposed to the larger volumes of diluent used,
To ensure reproducible data, each test was repeated three can be explained by analyzing the data collected in this re­
additional times to obtain four sets of samples for analysis. search plus the results of Schroeder and DeLuca^^ and
In addition, this procedure was duplicated with the variables others. .33 This research supports the conclusion that not
being (1) the presence or absence of the membrane filter and only IS pH a'^determinant for maintaining solubility (or
(2) the diluent being 50 ml of either D5W or NS. The con­ causing crystallization) but the volume of diluent used is also
centration-time profiles studied were designed to approxi­ an important consideration. In the case of D5W, the pH
mate the typical clinical situation that would be seen in any overrides the dilution effect at the 150-mr dilution (Table
hospital. The concentration range of phenytoin sodium 1). With NS, the combination of a pH above 10.0 ahd suffi­
• studied was 67-400 mg/dl. Other authors, notably Cloyd et cient dilution results in a stable solution devoid of visible
al, have published findings on the effects of mixing large precipitate for at least one hour. One must be aware, though
doses of phenytoin with small volumes of i.v. fluids. that microcrystals of phenytoin may be present but unho-
ticeable in the solution;
The EMIT assay is a homogenous immunoassay technique
used for microanalysis of specific compounds. Basically, it . Quantitative. The results of the quantitative analysis are
presented in Table 2. One would expect the Concentrations
Table 1. pH of i^arious Concentrations of Phenvtoin

10.79 10.76
10.63 10.75
10.36 10.35
10.25 10.33
10.05 10.03
9.92 10.02
9;89 • 9.88
9.82 9.87
° PHTIO 0-1-g/2 ml ampul; pH 12.0,
'pH5.51.
" Crystals visibly noted.

96 American Journal of Hospital Pharmacy Vol 37 Jan 1980

 Phenyloin sodium

of. samples to be low initially. This is because the phenytoin
in solution is being mixed with the solution in the tubing.
Based on a flow rate of approximately 1 ml/min, no pheny­
toin is delivered to the patient until 5 to 10 minutes after the
start of the infusion or drip. To determine if any significant
differences resulted from using a burette solution adminis­
tration set with or without a membrane filter, the Wilcoxon's
sum of ranksftest®® was used.-The data from Table 2 were
examined and it was determined that p > 1; therefore, there
is no significant difference between the two. This cotnparison
revealed that the membrane filter had no effect on the
quantity of phenytoin delivered.
phenytoin sodium in 100 ml of NS should be avoided. A
suitable volume appears to he 90 ml.
3. Time is another important elem^t;any dilution of phen­
ytoin sodium in NS should be prepared immediately before
use and infused within one hour. From a practical stand­
point, there seldom is a need to infuse the drug over any
longer period.
4. A statistical analysis of the data using the Wilcoxon's sum
of ranks test revealed that the presence or absence of the
1.0-pm membrane filter in the burette solution adminis­
tration had no significant effect on the ultimate availability
of phenytoin. sodium. Although delivered concentrations
are not appreciably different when comparing filtered
versus nonfiltered solutions, it is appropriate to use an inline
filter because of the possibility of microcrystal formation.

Downloaded from https://academic.oup.com/ajhp/article-abstract/37/1/95/5199353 by guest on 07 November 2019

^To determine if the i.v. solutions used as the diluent (D5W
versus NS) affected the quantity of phenytoin sodium de­
livered, the Wilcoxon's sum of ranks test again was used. The
results proved that p = 0.01, which is a significant difference.
The probability of getting that significant a difference from
random sanipling fluctuations alone is 1 in 100; therefore,
it may he concluded that the values for D5W are significantly
less than those for NS for some reason other than chance.
The decreased availability of phenytoin sodium in D5W
resulted from rapid crystallization. Crystals were seen in the
burette solution administration set within one hour after,
admixture. In NS, crystallization was not noted in the bur­
ette solution administration set during the 60-minute period.
Phenytoin sodium in NS consistently produced the expeAed
concentration upon assay, as evidenced in Table 2.
The values were exceptionally good for phenytoin sodium
administration in NS without the membrane filter burette
solution administration set.
Conclusions
The following conclusions were reached as a result of this'
research: , Am Heart J 77: 315,1969.
1. The pH of the diluent solution affec^he solubility of
phenytoin sodium. For this reason, D5W proved to be an
uriacceptable solution while NS proved to he acceptable,
providing certain volume restrictions were adhered to.
2. As the volume of i.v. solution used as a diluent increases,
the pH of the resulting admixture decreases. If the pH is
lowered through dilution to the extent that the admixture
has a pH of 10.0 or below, the solubility of phenytoin sodium
will be decreased sufficiently to cause its precipitation from
the solution relationship, any further dilution of 100 mg of
In summary, this research demonstrates that 25-50 ml
of NS is an acceptable volume and solution for dilution of
ICQ mg of phenytoin sodium when the method described
here is followed for infusing the drug. The admixture should
he prepared immediately before use and infused within one
hour. However, this procedure should he used only when the
accepted practice of direct i.v. injection is neither desirable
nor practical.
References
1. Meritt HH and Putman TJ: Na diphenyl hydantoinate in the
treatment of convulsive disorders, JAMA III: 1068-1073,
1938.
2. Easton JD: Diphenylhydantoin and epilepsy management, Ann
Intern Med 77: 421-423,1972.
3. Kutt H and McDowell F: Management of epilepsy with diphe­
nylhydantoin sodium, JAMA 203:167-170 (Mar 11) 1968.
4. Wallis W, Kutt H and McDowell F: Intravenous diphenylhy­
dantoin in the treatment of acute repetitive seizures. Neurology
18: 513-535,1968.
5. Bigger JT, Jr: Arrhythmias and antiarrhythmic drugs, Adv
Intern Med 18: 251-281,1972.
6. Helfant RH et al: The clinical use of diphenylhydantoin (Di­
lantin) in the treatment and prevention of cardiac arrhjrthmias,'
7. Mercer EN and Osborne JA: Current status of diphenylhy­
dantoin in heart,disease, Aran Intern Med 67: 1084-1107,
1967.
8. Murphy JT and Schwab RS: Diphenylhydantoin sodium used
parenterally in the control of convulsions—a five-yeM report,
JAMA 160: 385-388,1956.
9. Serrano EE, Roye DB, Hammer RH et al; Plasma diphenyl­
hydantoin values after oral and intramuscular Administration
of diphenylhydantoin. Neurology 23: 311-317,1973.
10. Wilensky AJ and Lowden JS: Inadequate serum levels after
intramuscular administration of diphenylhydantoin. Neurology
23: 318-324,1973.

Table 2. Stability of Phenytoin Sodium Solutions^ In Burette Solution Administration Set

Phenytoin Sodium Concentration (pg/mi)'*
Time Dextrose 5% in Water 0.9% Sodium Chloride inlection
(mln) 1-pm Filter No Filter 1-pm Filter No Finer
0 18.0 ±1.5 17.6 ± 0.8 18.1 ± 3.0 18.4 ± 0.3
5 ND= ND ND ND
10 7.5 ± 0.8 19.1 ± 4.3 5.8 ±1.4 4.0 ± 1.5
15 10.2 ±2.6 11.6 ±3.8 13.5 ± 1.9 10.9 ± 3.6
20 14.5 ±1.6 13.9 ± 1.9 16.2 ± 1.3 14.8 ± 3.7
25 17.0 ± 2.3 14.1 ± 0.8 16.9 ± 1.3 18.0i1.8
30 15.4 ±1.6 16.8 ± 2,3 18.3 ±2.7 17.0 ± 1.7
60 17.2 ±2.1 . 16.6 ±0.7 18.4 ± 2.7 , 19.0 ± 0.8

® Prepared with Dilantin, Parke, Davis and Company; 0.1 g/2 ml arnpul; pH 12.0.
'' The data represent the mean value ± 1 standard deviation Of the assays performed.
<= Not detected.

Vol 37 Jan 1980 American Journal of Hospital Pharmacy 97

 Phenjrtoln sodlum/Cefamandole nafale

11. Woodbury DM, Penry JK, Schmidt RP et al: Antiepileptic

drugs, Raven Press, New York, 1972, p 103. 7: 418 (Sep) 1973. ,
12. Wmdholz M (ed): The Merck index, 9th ed, Merck and Com­ 7n°rw/rr^ 11^ ^ use (letter). Drug
pany, Inc. Rahway, NJ, 1976, p 952 /nfeHC/m Pharm 7: 418 (Sep) 1973. >.eri, i^rug
A' 5,5-diphenyl- ?4T9?S^ep?J97? P^arm
270 279 1956^"™ ^ Ther 118:
Sp°?9?3" Pharm
intravenous
7A£
15. Gellerman GL and Martinez C: Fatal ventricular fibrillation

• SSVSnS """""
16. Louis S, Kutt H and McDowell F: The cardiocirculatory changes

J'wsTsTiS''*"''''"''''""""""" ® s»!: wSrsSS'"""'

Downloaded from https://academic.oup.com/ajhp/article-abstract/37/1/95/5199353 by guest on 07 November 2019

"• N'SITZ 2^ m?'N»51~ ?etSrA77
11 n i
^Ihanol intoxication
109 (Jul 8) 1976.

19. Bauman JL, Siepler JK and Fitzloff j; Phenytoin crystallization

fluids. Drug Intell Clin Pharm 11:646-649 (Nov)
11 295: 109-110 (Jul 8) 1976.
20. Bauman JL and Siepler JK: Phenytoin crystallization intra- fluSr(fetTeD^n?"J f •" intravenous
venous fluids (letter). Am J Hasp Pharm 35: 20-21 (Jan) 34 ?hr i uri. 7nleH Clin Pharm 12: 120 (Feb) 1978.
34. Schroeder HG and DeLuca PP: A study on the "in vitro" ore
cipitation of poorly soluble drugs from nonaqueous vehicles in
SoffleSnhZt Sawchuck RJ: Concentration-time
profile of phenytoin after admixture with small volumes of in- 15 v"™n" I 1-14 1974
35. VanDerLinde LP and Campbell RK: Guidelines for the intra-
22 Ralkh." NH 1978.
Sil^r W P Neurologic disorders in renal 3nr(Jan7i97?*'°" intell Clin Pharm 11:
91 Q ^^3-148 (Jan 15) 1976.
23. Sachtler G: Dilantin for iv use (letter). Drug Intell Clin Pharm 6??^^ T' simply explained, Dover Pub­
lications Inc., New York, 1979, p 166-178.
\

c Am J Hosp Pharm. 1980; 37:98-101

Stability of Frozen Soiutions of Cefamandoie
Nafate
Michael Bornslein, Paul R. Klink, Brett T. Farrell,
Carl L. WInely, and James 0. Boylan
The chemical, microbiological and visual stability of frozen
solutions of cefamandoie nafate was studied. " and cefamandoie sodium are shown in Figure 1
Solutions of cefamandoie nafate were prepared by diluting 1
g of drug with ,3 ml of Water for Injection, USP, or 0.9% Sodium
Chloride Injection, USP, or 5% Dextrose Injection, USP (i.m
d. utions ; or with 50 or 100 ml of the latter two diluents (i.w
dilutions) Stability of samples stored in glass and polyvinyl
chloride plastic containers for up to 52 weeks at -10 and -20 C
was measured by microbiologic, polarograpbic, iodometric ne­
phelometric and chromatographic assay and pH was measured.
/ITj-V ° P"fo™ed "Sing the i.m. dilutions
I.M. dilutions of cefamandoie nafate were stable for 52 weeks
when stored at -20 C; at -10 C, however, some samples did not
freeze completely and were turbid when thawed. I.V. dilutions
were stable for 26 weeks when stored at -20 C. I V dilutions
with D5W stored at -10 C developed a transient haze A gmd
uj decrease in pH, which was a function of storagd time, was
riter th T? 9'" '"""ths of freezing did not
alter the LDso in mice.
Solutions of cefamandoie nafate are stable for at least 26
weeks when stored at -20 C in glass or PVC containers.
Index terms: Cefamandoie nafate; Containers; Contamination;
extrose Diluents; Glass; Hydrogen ion concentration; Injec-
tme'Waterchloride; Sodium chloride; Stability; Tempera-
Cefamandoie nafate for injection is a new semisynthetic,
broad-spectrum cephalosporin antibiotic fof intramuscular
and intravenous administration.1.2 Chemically, it is the so­
dium salt of 7-D-mandelamido-3-[[(l-methyl-lH-tetrazol-
5-yl)-thio]methyl]-3-cephem-4-carhoxylic acid formate
(ester). The empirical formula for cefamandoie nafate is
CigHjvNeOeSaNa. The structures for cefamandoie nafate
The convenience of extended storage by freezing recon­
stituted antibiotic solutions has made this technique popular
in many hospital pharmacies. Frozen solution stability data
previously reported for cephalosporins,3-6 have not included
cefamandoie nafate solutions. The objective of this study was
Michael Bornstcin, Ph.D., and Paul R. Klink, Ph.D., are Senior Phar
maceutica Chemists, Parenteral Products Development; Brett T. Farrell
^ Assoemte Analytical Chemist, Analytical Development; and Carl L
H Lflf' R 'I TT' Microbiological Assay Development
lily Research Laboratories, Indianapolis, IN 46206. James C. Boylan
rC^cl"? Development, G. D. Searle'
Address reprint requests to Dr. Bornstein.
The assistance of W. D. Broddle, Ph.D., S. J. Conine, E. I Pursell M I
COM © 1980, American Society of Hospital Pharmacists..Inc. All rights

98 , American Journal of Hospital Pharmacy Vol 37

Jan 1980
0002-9289/80/0101-0098$01.00