Journal of Clinical and Hospital Pharmacy ( 1986) 1 1,2 15-223.

STABILITY OF HYDRALAZINE HYDROCHLORIDE IN

AQUEOUS VEHICLES

V. Das Gupta*, K. R. Stewartt and C. Betheat

*University of Housron, 1441 Moursund Sr., Houston, T X 77030, U.S.A.and tBen Taub General Hospiral,
Houston, T X 77030, U.S.A.

SUMMARY

The stability of hydralazine hydrochloride in aqueous vehicles which

contain either dextrose, fructose, lactose, maltose, mannitol, sorbitol or
sucrose has been studied using a stability-indicating high-performance
liquid chromatographic method. Dextrose, fructose, lactose and maltose
had adverse effects on the stability of hydralazine. In mannitol (better than
sorbitol) and sorbitol, hydralazine was stable for about 21 days (loss in
potency of less than 100;) and sucrose had an adverse effect only after its
hydrolysis to fructose and dextrose. T h e optimum pH range of stability in
dextrose was approximately between 3.2 and 4.4. The first-order rate of
decomposition increased with an increase in the concentration of dextrose
but not with an increase in the concentration of hydralazine. In the absence
of other excipients the phosphate and citrate buffers did not adversely
affect the stability of hydralazine hydrochloride.

INTRODUCTION
Hydralazine hydrochloride is extensively used against hypertension. It is available in
the form of tablets and injection. It is often prescribed for small children who cannot
swallow tablets. Most of the hospitals in this area prepare oral liquid dosage forms by
mixing the tablet powder with commercially available strawberry-simple syrup. The
usual concentration ofhydralazine hydrochloride in oral liquid is 1 mg/ml and an arbi-
trary expiry date of 10-21 days is assigned. T h e purpose of these investigations was to
study the stability of hydralazine hydrochloride in a number of oral aqueous vehicles
containing either dextrose, fructose, lactose, maltose, mannitol, sorbitol or sucrose.

MATERIALS AND METHODS

Chemicals and reagents. All the chemicals and reagents were either USP-NF or ACS
grade and used as received. Hydralazine hydrochloride (lot KL4109KL, Aldrich
Chemical Co., Milwaukee, WI) was used without further purification.
Correspondence:Dr V. D. Gupta, College of Pharmacy, University of Houston, 1441 Moursund St.,
Houston, T X 77030, U.S.A.

215
216 V . Das Gupta, K . R . Stewart and C . Bethea
Apparatus. A high-pressure liquid chromatograph (ALC 202 equipped with U6K
universal injector, Waters Associates, Milford, MA) equipped with a multiple wave-
length detector (Schoeffel SF770,Kratos, Inc., Ransey, NJ), a recorder (Omniscribe
1513-12, Houston Instruments, Austin, TX), and a digital integrator (Autolab
minigrator, Spectra Physics, Santa Clara, CA) were used as well as a nonpolar column
(30 cm long x 3.9 mm i.d., Bondapak/C,,, Waters Associates, Milford, MA).

Chromatographic conditions. T h e mobile phase was 2 yo (v/v) methanol, 0.1 Yo (v/v)

glacial acetic acid in water containing 0.015 M KH,PO,, and the flow rate was
3*0ml/rnin. T h e detector was set at 256nm (0.04, AUFS), the temperature was
ambient and the chart speed was 30 cm/h.

Preparation of solutionsfor stability studies. T h e list of solutions prepared are presented

in Table 1. All solutions were prepared using a simple solution method. The solutions
were assayed (see procedure below), pH values were determined (Model 4500 digital
pHmeter, Beckman Instruments, Fullerton, CA), transferred to 60-ml amber bottles
(Brockway Glass Co., Brockway, PA) and stored at room temperature (24"Cf 1°C).
T h e data were recorded again after appropriate intervals.

Preparation of stock solutions for analysis. A 2.00,d aqueous solution of phenylpro-

panolamine hydrochloride (the internal standard) and a O.lYo aqueous solution of
hydralazine hydrochloride were prepared fresh daily.

Preparation of standard solutions. T h e standard solutions were prepared as needed

by diluting the stock solution(s) with water. T h e usual concentrations of hydralazine
hydrochloride and phenylpropanolamine hydrochloride were 30 pg/ml and 1.6 mg/ml,
respectively.

Preparation of assay solutions. All assay solutions were prepared by diluting an

appropriate quantity of the assay solution with water to contain 30 pg/ml of hydrala-
zine hydrochloride. Before diluting, the stock solution of the internal standard was
added if needed. No internal standard was added in most of the solutions in order to
detect the presence of new peaks from the degradation products.

Assay procedure. A 2 0 4 aliquot of the assay solution was injected into the
chromatograph using the described conditions. For comparison, an identical volume
of the standard solution was injected after the components of the assay solution eluted.
T h e results were calculated using:

Ph,
- x 100 =Per cent of the label claim,
Ph,
where Ph, is either the peak heights ratio of hydralazine: phenylpropanolamine
or the peak height of hydralazine, of the assay solution and Ph, that of the standard
solution. Preliminary investigations indicated that the peak heights ratios (or peak
heights) versus concentrations were linear between 15-45 pg/ml of hydralazine
hydrochloride.
 Stability of hydralazine hydrochloride 217
Table 1. List of aqueous solutions prepared for stability studies

Concentration of
Solution hydralazine HCI Other ingredients Buffering agent Initial pH
number (mgiml) if any (M) if any ( f0.05)

1 1 Dextrose* (0.28) 4.0

2 1 Dextrose (0.28) - 4.6
3 1 Dextrose (0.28) 0 . 1 KHzPO,
~ 4.4
4 1 Dextrose (0.28) 0.1M KH,PO, 4.4
5 1 Dextrose (0.28) 0.2M KHJ’O, 4.4
6 2 Dextrose (0.28) 0.lM KHJ’O, 4.4
7 1 Dextrose (0.28) 0.1M HCI 0.9
8 1 Dextrose (0.28) 0.lM Phosphate 2.2
9 1 Dextrose (0.28) 0 . 1 Phosphate
~ 3.2
10 1 Dextrose (0.28) 0 . 1 Phosphate
~ 5.7
11 1 Dextrose (0.28) 0 . 1 Phosphate
~ 6.7
12 1 Dextrose (0.28) 0 . 1 Phosphate
~ 7.5
13 1 Fructose (0.28) - 4.5
14 1 Fructose (0.28) 0 . l M KHJ’O, 4.4
15 1 Lactose (0.28) - 4.1
16 1 Lactose (0.28) 0 , l M KH,PO, 4.4
17 1 Maltose (0.28) - 4.3
18 1 Maltose (0.28) 0 . 1 KH,PO,
~ 4.4
19 1 Mannitol(O.28) - 4.9
20 1 Manr.itol(O.28) 0 . 1 KHJ’O,
~ 44
21 1 Sorbitol(O.28) - 4.5
22 1 Sorbitol(O.28) 0.lM KHJ’O, 4.4
23 1 Sorbitol(O.56) - 4.6
24 1 Sucrose (0.28) - 4.7
25 1 Sucrose (0.28) 0 . 1 KH,PO,
~ 4.4
26 1 Sucrose 8500 w/v - 5.4
27 1 Sucrose 8So, w/v+ - 5.7
0.1O 0 sodium benzoate
28 I Sucrose 42 500 w/v+ 0.025~
Citric acid 2.8
0.05°0 sodium benzoate
29 It Sucrose8Soo w/v+ 5.7
0.1 O 0 sodium benzoate
30 1: Sucrose85O0 w/v+ 5.7
0.1 O 0 sodium benzoate
31 1 Strawberry syrup$ 2.6
32 1 Simple syrup‘ - 4.0
33 1 - 0 . 0 5Citric
~ acid 2.6
34 1 - 0.lM Phosphate 26
35 Dry mixture: 100.0 mg of hydralazine hydrochloride mixed with 4.9 g of dextrose and left overnight

* From dextrose 5u/b injection, Travenol laboratories, Deerfield, IL.

t From 50 mg tablets powder of hydralazine hydrochloride.
: From 20 mg/ml injection of hydralazine hydrochloride. T h e injection also contained parabens as the
preservative.
$ Commercial lot from C & D Flavor Co., Sugarland, TX. Also contained 0.1O 0 sodium benzoate, citric
acid (quantity not disclosed on the label), colour and strawberry flavour.
1 Commercial lot from C & D Flavor Co., Sugarland, TX. Also contained O . l o 0 sodium benzoate and
citric acid (quantity not disclosed on the label).
218 V . Das Gupra, K . R . Stewart and C . Bethea
R E S U L T S AND D I S C U S S I O N
Assay method
T h e assay method developed appears to be stability-indicating (Figs 1 and 2). In
Fig. 1, Chromatogram A is from a standard solution, B from strawberry syrup without
hydralazine, but diluted as assay solutions, and C from a 1-day-old solution of
hydralazine hydrochloride (1 mg/ml) in strawberry syrup. Obviously, almost all of
the hydralazine hydrochloride (peak 1) had decomposed in 1 day. This was further
confirmed using a phenyl column and a mobile phase containing 0.7% v/v of methanol
instead of 2% v/v. All other chromatographic conditions were the same. The method
was reproducible when tried on synthetic mixtures and had a per cent relative standard
deviation based on six readings of 1.2-1.9. Furthermore, the results were reproducible
on a day to day basis.
If required, phenylpropanolamine hydrochloride (1.6 mg/ml) can be used as an
internal standard with all the solutions except when fructose is present. Phenylpro-
panolamine elutes before hydralazine. With fructose, hydrochlorothiazide (60 pg/ml)
can be used as the internal standard (with phenyl column) which elutes after hydrala-
zine. The increase in the per cent relative standard deviation, based on six readings,
was too small (0.1) without the use of an internal standard.
Further investigations indicated that the peak immediately after peak 1 in Fig. 1C
was from the decomposition produc: formed by the interaction of hydralazine with
dextrose and the peak immediately before peak 1 was due to an interaction between
hydralazine and fructose.

Timr (rnin)
Fig. 1. Sample chromatograms. Peak 1 is from hydralazinc. Chromatogram A is from a standard solution; B
from a commercial strawberry syrup (diluted 0.75 rnl to 25 ml with water); and C from 1-day-old solution 31
(Table 1). The peak immediatelybefore peak 1 is from the decomposition produn of reaction of hydralazine
with fructose and the peak immediately after peak 1 is from the decomposition product of reaction between
dextrose and hydralazine. For chromatographic conditions, see text.
 Stability of hydralazine hydrochloride 219
Table 2. Assay results and h a 1 pH values of the solutions

Per cent of the label claim found* after (days)

Solution Final dayt pH value
number 1 2 3 7 14 21 ( 0.05)

1 59.4 41.9 % % % t 3.3

2
3
4
58.3

29.6
40.8 %
- %
See Fig. 3 (K value 0.54 day- ')
+ 10.0 %
%

%
%

%
3.9
4.4
4.2
5 55.0 + 18.2 % % % 4.3
6 59.6 + 20.8 + % + 4.2
7-12 See Fig. 4 No changes in pH values
13 79.6 66.2 % 2 % % 3.3
14 62.3 43.0 % % % % 3.9
15 46.4 27.0 % % % % 3.5
16 34.2 18.8 % 3 + t 4.3
17 67.0 47.1 % % % t 3.9
18 59.3 35.7 . 22.3 % % + 4.3
(Kvalue -0.52 day-')
19 100.6 + 99.5 4.4
+ 99.3 99.6 99.3
20 100.0 98.8 % 98.9 % 3 4.4
21 100.0 99.3 % 97.2 % 95.7 4.4
22 99.2 97.1 % 96.8 % 1 4.4
23 99.8 % 99.5 97.2 95.0 92.4 3.4
(Kvalue -G 0034day-')
24 100.2 99.8 % 99.1 % 83.4 3.4
25 100.0 1 100.2 99.1 % % 4.4
26 94.6 93.5 % 90.0 86.59 % 44
27 96.8 95.2 % 90.0 88.05 % 5.5
28 93.1 ++ 64.2 17.6 3 2.6
29-30 Results were similar to solution 27
31 2 4 + % % + + 2.6
32 6.6** ++ % % % % 3.9
33 100.1 + 3 100.0 99.7 99.5 2.6
34 101.0 ++ 1 100.3 100.1 99.8 2Wt
35 99.1 No reaction-not followed funher
(Drymixture)

*Based on 100°, at zero day. The per cent relative standard deviations based on six readings were
between 1.3-1.9.
t That is on the last day of analysis as reported in the table.
% Not determined on this day.
9 These are 12-day results.
7 Results after 2 h were 31.400.
**Resultsafter 2 h were 54.900.
In pH 4.4 phosphate buffer, results were similar and no change in pH value was noticed.

A sample prepared in fresh simple syrup (sucrose 85% w/v) lost only 5.4% of
hydralazine in 1 day (Fig. 2C and solution 26 in Table 2). However, a solution in 0.2 M
sucrose ( 9 5 % w/v) did not lose any potency in 2 days (solution 21,Table 2). A sample
in a commercial simple syrup lost more than 93% of potency in one day and more than
220 V . Das Gupta, K . R . Stewart and C . Bethea

Time (mid
Fig. 2. Sample chromatograms from 1-day solutions. Peak 1 is from hydralazine hydrochloride. Chromato-
gram A is from solution 33 (Table 1); B from solution 1 and C from solution 26 (Table I). For chromato-
graphic conditions, see text.

45O, in 2 h (solution 32, Table 2). The commercial syrup also contained 0.10/, w/v of
sodium benzoate and citric acid (quantity not disclosed on the label).
The solution prepared in 0.28 M dextrose (solution 2, Table 2) lost more than 59”/,
of the potency in 2 days and in 0 . 2 8 fructose
~ (solution 13, Table 2) more than 330,, in
2 days. It is obvious that there was no interaction of hydralazine hydrochloride with
sucrose, as such, but only with the products of hydrolysis of sucrose, i.e. dextrose and
fructose. This reaction is probably similar to the formation of a Schiff base.
Since the pH of strawberry syrup (commercial lot) was about 2.6, almost all the
sucrose must be in a hydrolysed form, thus explaining the rapid hydralazine decompo-
sition in it (solution 31, Table 2). Further investigations indicated that the loss in
potency in 2 h was 58.6O,,. Results in commercial simple syrup, with pH close to 4,
showed only 45.1O; loss in potency in 2 h (solution 32 in Table 2). The strawberry
syrups also contained 0.1So sodium benzoate and citric acid (quantity not disclosed on
the label). These two compounds did not interfere with the assay procedure and on
their own did not affect the stability of hydralazine (solutions 27 and 33 in Table 2).
However, when citric acid was added to fresh simple syrup (solution 28, Table 2), the
decomposition was very fast. For example, in 7 days, the potency decreased by 82.40,,
versus only 109, in solution 27 which did not contain citric acid. Citric acid apparently
increased the hydrolysis of sucrose to dextrose and fructose. The decomposition in
dextrose solutions was first-order (Fig. 3). T h e increase in the concentration of dex-
trose increased the rate of decomposition (solution 4 versus 2 or 3 which had similar
results, Table 2) and there was no change in the rate of decomposition when the con-
centration of hydralazine was increased (solution 6 versus 2 or 3, Table 2). The
phosphate buffer did not change the rate of decomposition significantly (solution 5
versus 2 or 3, Table 2) in solutions containing dextrose (0.28 M).
 Stability of hydralazine hydrochloride 22 1

p" 2.0 -
L

1.0-

, 1 , ,

PH
Fig. 4. pH versus per cent decomposed (after 2 days) curve of solutions 3 and 7-12 (Table 1) in 0.28 M
dextrose.

T h e phosphate buffer without dextrose did not affect the stability of hydralazine. After
21 days, a solution in phosphate buffer (0.1 M) was still 99.800 potent (solution 34
in Table 2). T h e pH value, as expected, did affect the rate of decomposition (Fig. 4 ) .
T h e optimum range for the stability of hydralazine in dextrose (C.28 M) was determined
to be about 3.2-4.4.This is in agreement with the USP-NF recommendation (1) of the
222 V . Das Gupta, K . R. Stewart and C . Bethea

Fig. 5. Sample chromatograms from 3-day-old solutions. Peak 1 is from hydralazine hydrochloride and
peak 2 from the product of decomposition (reaction between maltose and hydralazine). ChromatogramA is
from solution 18 and B from solution 21 (in mannitol, Table 1). For chromatographicconditions,see text.

pH value of the injection to be between 3.4 and 4.4. In a dry mixture of dextrose with
hydralazine (number 35, Table 2), there was no reaction between the ingredients.
Fructose had an adverse affect on the stability of hydralazine hydrochloride (solution
13, Table 2). In the presence of phosphate buffer, the rate of decomposition in fructose
solution increased (solution 14, Table 2). For example, in 0.28 M fructose solution,
potency in 2 days decreased to 66.2% versus 43% when phosphate buffer was present.
This is in contrast to the effect of phosphate buffer in the presence of dextrose which
was insignificant. The decomposition product with fructose eluted immediately before
hydralazine versus immediately afterwards, when the reaction was with dextrose
(Fig. 1C). The first-order reaction was not followed in the presence of fructose with or
without phosphate buffer indicating some complex reaction.
In the presence of lactose, the phosphate buffer hastened the rate of decomposition
(solution 16 versus 15, Table 2) as with fructose. Once again, the first order of reaction
was not followed, even in the presence of phosphate buffer, indicating a complex
reaction. In the case of lactose, the product of decomposition eluted at the same time as
that of thereaction between dextrose and hydralazine. The effect of phosphate buffer was
probably due to an increase in the rate of hydrolysis of lactose to galactose and dextrose.
In the presence of maltose, the phosphate again increased the rate of decomposition
slightly (solutions 17 and 18, Table 2). In this case, the first-order of reaction was
followed and in the presence of phosphate buffer the decomposition constant value
was slightly less (0.52 day-') than in the presence of 0.28 M dextrose (0.54 day- '). The
product of decomposition eluted approximately 7 min after hydralazine (Fig. 5).
Apparently, this reaction must have been with maltose and not with the products of
hydrolysis (2 moles of dextrose); otherwise, the product of decomposition would have
eluted immediately after hydralazine as was the case with dextrose.
Neither mannitol nor sorbitol affected the stability of hydralazine (Fig. 5 ) for up to
7 days (solutions 19-23, Table 2) and solutions with shelf-lives of up to 21 days could
 Stability of hydralazine hydrochloride 223
be formulated, which is the usual expiry duration for extemporaneously prepared
dosage forms. With mannitol and sorbitol, phosphate buffer had very slight effects on
the stability.
In conclusion, it can be stated that hydralazine hydrochloride oral liquid dosage
forms should not be prepared using fructose, dextrose, lactose, maltose and sucrose.
Sucrose can adversely affect the stability following its own hydrolysis. For extempor-
aneous purposes, only mannitol or sorbitol can be used as sweetening agents. Mannitol
proved better than sorbitol.

REFERENCE
1. Anonymous (1980) Supplement 3 to The United Stares Pharmacopeia 20th edn-The National Formulary
15th edn, p. 144. U.S.Pharmacopeial Convention, Rockville, MD.