Analyst, 1980,105, 455-461

Download as pdf or txt
Download as pdf or txt
You are on page 1of 7

View Article Online / Journal Homepage / Table of Contents for this issue

Analyst, May, 1980, Vol. 105, pp. 455-467

455

Published on 01 January 1980. Downloaded by UNIVERSITY OF BRIGHTON on 26/03/2014 10:25:47.

Quality Control of Prednisolone Sodium Phosphate


N. Stroud, N. E. Richardson, D. J. G. Davies and D. A. Norton
Centre for Drug Formulation Studies, School of Pharmacy and Pharmacology, University of Bath, Claverton
Down, Bath, B A 2 7 A Y

Prednisolone sodium phosphate is believed to undergo hydrolysis of the


phosphate ester group as its primary degradation pathway. Most published
assay methods do not determine the phosphate ester directly, and therefore
a high-performance liquid chromatographic method has been developed for
prednisolone sodium phosphate in the presence of its breakdown products,
which has been validated in the presence of excipients used in ophthalmic
solutions. Stability data are presented that are comparable to those obtained
for related steroid phosphate esters. The stability data indicate t h a t a
simpler ultraviolet spectrophotometric assay method can be used for routine
stability testing.

Keywords : Pvednisolone s o d i u m plzospliate determination ; high-performance


liquid chvomatograplay ; prednisolone s o d i u m plaosphate stability

The efficiency of corticosteroids such as prednisolone for the treatment of ocular inflammatory
conditions is now well established. Prednisolone has a low solubility in water and for
aqueous formulations the more water-soluble phosphate ester is used, which can be formulated
for both parenteral and topical administration. Several workers have reported that corticosteroids such as prednisolone undergo thermal degradation in aqueous solution, involving
the 17-dihydroxyacetone ~ide-chain.l-~Transformations and eliminations have been shown
to occur in both the presence and absence of air. I n the presence of air under alkaline
conditions, the predominant reaction appears to involve cleavage of the C17 side-chain to
yield the corresponding etianic acid. In the absence of air, two reactions predominate,
yielding the 17-keto steroid and the hydroxy acid. Degradation of the A ring has also been
shown to occur in a related steroid, hydrocortisone, formulated in a polyethylene glycol
base.5 However, the A ring is an inherently stable structure and the rate of degradation
was much slower than that for the C,, side-chain. The degradation of steroid phosphate
esters has not been studied as extensively, although Marcus6 has reported that the degradation of hydrocortisone phosphate in aqueous solution involved hydrolysis as the only significant degradative pathway and was dependent on the hydrogen-ion concentration. It would
appear, therefore, that the thermal degradation of prednisolone sodium phosphate in aqueous
solution would involve the pathways illustrated in Fig. 1 and that hydrolysis of the phosphate
group on the C17 side-chain would be predominant.
Most published assay methods do not determine the phosphate ester directly. Both
prednisolone sodium phosphate and the parent prednisolone possess the 3-keto group and
related conjugated system and have similar absorption spectra in the ultraviolet region.
Kaplan and Levine7 have developed a column chromatographic method for separating the
two compounds using ion-pair formation between the ester and trihexylammonium chloride.
However, the method is tedious and lengthy for routine analysis. The C,, side-chain of
prednisolone has been determined by complexation with tetrazolium blue followed by
spectrophotometric determination of the coloured complex.* This method, however, is
specific for the C,, side-chain and prednisolone sodium phosphate would require preliminary
hydrolysis to the parent alcohol, which is difficult to achieve quantitatively. Other methods
reported are the determination of the inorganic phosphate produced as the ester hydrolysese
and gas - liquid chromatography. Upton et aL9 have reported a high-performance liquid
chromatographic (HPLC) method for steroid phosphates using a reversed-phase column.
However, preliminary work in our laboratories indicated that prednisolone sodium phosphate
was eluted immediately after a non-retained compound (potassium dichromate) on a
Spherisorb S5 ODS reversed-phase column. It is essential that any assay method distinguishes between the parent compound and its degradation products and we have therefore
developed an HPLC assay for prednisolone sodium phosphate using an anion-exchange
column, as the phosphate ester is present in an anionic form in aqueous solution.

View Article Online

Published on 01 January 1980. Downloaded by UNIVERSITY OF BRIGHTON on 26/03/2014 10:25:47.

456

STROUD

et d.: QUALITY CONTROL


Experimental

Amzlyst, Vol. 105

Apparatus
Chromatograms were determined routinely using a Pye LC20 system, which has a fixedwavelength detector set at 254nm. Injections were made on-column with a Pye Unicam
fixed-volume 10-pl loop valve. All measurements were made a t ambient temperature in
replicate.
Spectrophotometric determinations were made using a Pye Unicam SP1800 spectrophotometer.
p H determinations were performed using either a Pye Unicam 291 pH meter or a Radiometer Type 27 pH meter fitted with a PHA 630P scale expander. Both p H meters were
used in conjunction with Pye-Ingold combined glass - silver electrodes. All p H measurements were carried out on solutions equilibrated to 25 & 0.1 "C; meters were standardised
with two appropriate standard buffers.
CH,OPO,

Na2

Prednisolonesodium
phosphate

H : =&

CH20H

I
5

H : =&

Prednisolone
Prednisolone

COOH

Fig. 1. Thermal degradation


sodium phosphate.

pathways of

prednisolone

Materi a1s
Prednisolone sodium phosphate was a gift from Smith and Nephew Ltd. and was used as
received. All buffer salts were of analytical-reagent grade and other reagents were of at
least laboratory-reagent grade. Potassium hydrogen phthalate was of an NPL certificated
grade supplied by BDH Chemicals Ltd. Solvent:; were of analytical-reagent grade. Water
was freshly distilled from an all-glass still. Poly(viny1 alcohol) (Gohsenol N300, Nippon
Goshei) was supplied by British Traders and Shippers Ltd. Trihexylammonium chloride
was prepared from trihexylamine (Eastman Kodak Co.) according to the method of Kaplan
and L e ~ i n e . ~

View Article Online

May, 1980

O F PREDNISOLONE SODIUM PHOSPHATE

457

The stationary phase was Partisil 10 SAX (Whatman), an anion-exchange material, packed
into either 250 x 4.6 mm or 100 x 4.6 mm stainless-steel columns. The mobile phase
consisted of a 1
9 V / V mixture of methanol and one-fifth strength McIlvaines citrate phosphate buffer (pH 5.2), and was de-gassed before use. The actual pH of the mobile
phase was 5.6.

Published on 01 January 1980. Downloaded by UNIVERSITY OF BRIGHTON on 26/03/2014 10:25:47.

Chromatography of Prednisolone Sodium Phosphate in Aqueous Solution


Aqueous solutions are prepared to contain 0.004-0.024~0m/ V prednisolone sodium
phosphate and 0.12% m/V potassium hydrogen phthalate as internal standard and 10 pi
are injected on to the column with a mobile phase flow-rate of 1.5 ml min-l. The chromatogram is recorded at a detector wavelength of 254 nm. The concentration of prednisolone
sodium phosphate is then determined by comparing the peak-height ratio of drug to internal
standard with that obtained with a standard solution containing 0.02% m/V of prednisolone
sodium phosphate and 0.12% m/V of potassium hydrogen phthalate.
Chromatography of Prednisolone Sodium Phosphate in the Presence of a Viscoliser
Place 2 ml of prednisolone sodium phosphate solution (concentration range 0.1-0.5y0
m / V )in a 10-ml glass centrifuge tube containing 3 ml of double-strength Sprrensens phosphate
buffer (pH 5.0), mix and add 5 ml of a 5% V / V solution of trihexylammonium chloride in
dichloromethane. Stopper the tube, shake it vigorously for 30s, then centrifuge it at
4000 rev min-l for 15 min. Remove the aqueous phase, transfer 3 ml of the organic phase
into a fresh centrifuge tube containing 4 ml of 0.1 M sodium hydroxide solution, shake for
30 s and then centrifuge a t 4000 rev min-l for 5 min. Pipette 3 ml of the aqueous phase
into a 25-ml calibrated flask containing 3 ml of 0.1 M hydrochloric acid and 3 ml of 1%
m/V potassium hydrogen phthalate solution and dilute to volume with water. The solution
in the flask is then chromatographed as described above.
Thermal Degradation of Prednisolone Sodium Phosphate
Aliquots of 10 ml of a 0.5% m/V solution of prednisolone sodium phosphate in an appropriate buffer were sealed into glass ampoules and heated in an oil-bath. The ampoules were
removed after known periods and assayed for residual drug.
Results and Discussion
Prednisolone sodium phosphate forms an anion in aqueous solution and should therefore
be retained by an anion-exchange column to an extent dependent on the p H of the mobile
phase, which would be reflected in longer retention times with increase in pH. Fig. 2 shows
the capacity factors for 0.02yo m/V of drug using benzyl alcohol as the non-retained com1

PH

Fig. 2. Influence of p H of mobile


phase on capacity factors of: A, prednisolone sodium phosphate; and B,
potassium hydrogen phthalate.

View Article Online

et al. : QUALITY CONTROL


Analyst, Vol. 105
pound, over the mobile phase pH range 2.95-6.10. It is apparent that the capacity factor

Published on 01 January 1980. Downloaded by UNIVERSITY OF BRIGHTON on 26/03/2014 10:25:47.

458

STROUD

decreases with increase in pH, and this is probably due to competition by the buffer components dominating the extent of the interactions between column and solute. Knox and
VasvarilO have shown that the capacity factor of phthalic acid on an anion-exchange column
can be selected by manipulation of the mobile phase pH. Fig. 2 also shows the capacity
factors for the more water-soluble potassium hydrogen phthalate over the mobile phase pH
range 3.50-6.10, and again it is observed that an increase in pH decreases the retention time.
However, it is apparent that over this pH range good resolution is obtained between prednisolone sodium phosphate and potassium hydrogen phthalate, and the latter was therefore
selected as the internal standard using a mobile phase pH of 5.2. The addition of 10%
V / V of methanol as organic modifier was found to reduce the analysis time and improve
peak symmetry, although the final pH of the mobile phase increased slightly to 5.6.
Initially, chromatograms were obtained using; the 250-mm column with a mobile phase
flow-rate of 1.5 ml min-l. Subsequently, a 100-mm column with a mobile phase flow-rate
of 1.2 ml min-l was shown to give improved pleak symmetry and a reduction in analysis
time. A typical chromatogram using this system is shown in Fig. 3(a).

C)

t
C

i
L

10

10

10

Time/mi n

Fig. 3. Chromatograms of prednisolone


sodium phosphate and its degradation products. (a) Prednisolone :sodium phosphate,
0.02'70 m / V ; ( b ) prednisolone sodium phosphate, 0.02% m/V
prednisolone, 0.018 6%
m / V ; and (c) prednisolone sodium phosphate,
0.02% m/V in p H 8 buffer heated a t 110 "C
for 24 h. Conditions: ambient temperature;
flow-rate 1.2 ml min-I; stationary phase 100
x 4.6 mm of Partisil 10 SAX (10 p m ) ; mobile
phase 10% V / V methanol in McIlvaines
citrate - phosphate p H 5.2 buffer and ionic
strength 0.1 M (final pH of mobile phase was
5.6) ; detector, ultraviolet a t 254 n m ; sensitivity 0.16 a.u.f.s. 1 , Prednisolone sodium
phosphate ; 2, potassium hydrogen phthalate
(internal standard) ; 3, prednisolone and 4,
degradation products.

The principal degradation product of prednisolone sodium phosphate in aqueous solution


is probably the parent steroid, prednisolone, which may break down further to give the

corresponding etianic acid, hydroxy acid or 17-keto steroid.

Prednisolone is un-ionised in

View Article Online

Published on 01 January 1980. Downloaded by UNIVERSITY OF BRIGHTON on 26/03/2014 10:25:47.

May, 1980

O F PREDNISOLONE SODIUM PHOSPHATE

459

aqueous solution and should be non-retained by an anion-exchange column. Fig. 3(b)


shows a chromatogram of 0.02y0 m/V prednisolone sodium phosphate in the presence of
0 . 0 1 8 6 ~ 0m/V of prednisolone, and it is apparent that the peak due to the latter is nonretained. A 0.018670 m/V solution of prednisolone buffered a t pH 7.4 was heated a t 100 "C
for 48 h and the chromatogram again showed only one non-retained peak, indicating that
any further degradation products would not interfere in the assay for prednisolone sodium
phosphate. Finally, a 0.5% m/V solution of prednisolone sodium phosphate buffered a t
pH 8.0 was heated a t 100 "C for 20 h and, after appropriate dilution, the chromatogram
[Fig. 3(c)3 shows that adequate resolution is obtained between drug, internal standard and
degradation products.
The linearity of the response was checked by injecting solutions of prednisolone sodium
phosphate over the concentration range 0.004-0.024~0 m/V in the presence of 0.12% m/V
of potassium hydrogen phthalate and calculating the peak-height ratios. Replicate calibration graphs constructed on two consecutive days were linear, with slopes of 63.08 [standard
deviation (s.d.) 0.181 and 63.8 (s.d. 0.22) and intercepts of -0.014 (s.d. 0.01) and -0.039
(s.d. 0.18), respectively. Comparison by a Student's t distribution showed them to be not
significantly different (tt;:: = 0.42; tintercept
ca,c.
- 1.20; ttabulated
= 2.45; n = 10, p = 0.05).
Simple aqueous formulations of prednisolone sodium phosphate containing only the drug,
buffer and a preservative such as benzalkonium chloride can be chromatographed directly
after appropriate dilution and the addition of an internal standard. The peak-height ratios
of drug to internal standard can then be compared between the sample and a standard
solution of prednisolone sodium phosphate. In the presence of formulatory excipients such
as polymeric viscolisers, pre-extraction of the drug is necessary. Preliminary extraction
experiments were monitored by determining the absorbance of prednisolone sodium phosphate
in the aqueous phase at 248 nm. An aqueous phase consisting of double-strength McIlvaines
citrate - phosphate buffer (pH 5.0) and an organic phase consisting of a 5% V / V solution of
trihexylammonium chloride in methylene chloride was found to transfer 98.0% (s.d. O.lyo,
n = 3) of the drug to the organic phase. The extent of subsequent re-extraction into 0.1 M
sodium hydroxide solution was 1 0 l . l ~ o(s.d. 0.86y0, n = 3). Formulations of 0.1% m/V
and 0.5% m/V prednisolone sodium phosphate containing 4.25% m/V of Gohsenol N300 as
viscoliser, O.Olyo m/V of benzalkonium chloride and 0.01 yo m/V of EDTA, disodium salt,
buffered at pH 8 were extracted and assayed by HPLC as described. Recoveries were
100.8% (s.d. 1.7%, n = 3) and 98.6% (s.d. 0.62y0, 12 = 3) compared to injection of the
standard aqueous prednisolone sodium phosphate solutions, which was considered t o be
satisfactory.

Applicability of the Assay to Stability Studies


Stability studies were carried out as described above and residual prednisolone sodium
phosphate was determined using the simple non-extraction HPLC assay procedure.
Degradation was generally followed to below 50%. Preliminary experiments showed that
in Smensens phosphate buffer (pH 6.1 and 8.2) a t 90 "C the data for the degradation of
prednisnlone sodium phosphate could be fitted to first-order rate plots, leading to values for
and 6.52 x 10-3h-1, respectively, which are close to the
the rate constants of 4.7 x
values of about 4 x
h-I determined by Marcus6 for the hydrolysis of
and 8 x
hydrocortisone phosphate at pH 6 and 7.5 and 91 "C. However, it was observed that
prednisolone sodium phosphate underwent an initial rapid degradation of about 5y0, and
this was overcome by the addition of O.Olyom/V of EDTA disodium salt. The influence of
temperature on the degradation of prednisolone sodium phosphate was determined over the
range 80-110 "C in Smensens phosphate buffer (pH 8) containing 0.01% m/V of EDTA,
disodium salt. When the data (Table I) were plotted according t o the Arrhenius relationship,
a value for the activation energy of 126.2 kJ mol-l was obtained, which compares with a
value a t pH 7.5 of 113 kJ mol-l for methylprednisolone phosphate reported by Flynn and
Lambll and 71 kJ mol-I determined by Marcus6 at the same pH for hydrocortisone phosphate.
The low value of 71 kJ mo1-I for hydrocortisone phosphate reported by Marcus has been
attributed to the reaction system being of a significant micellar character a t the drug concentration studied and that the micellar fraction changes with temperature.ll

View Article Online

Published on 01 January 1980. Downloaded by UNIVERSITY OF BRIGHTON on 26/03/2014 10:25:47.

460

STROUD

et al. : QUALITY

CONTROL

Analyst, Vol. 105

Applicability of the Assay to other Dosage Forms and Related Drugs


Most liquid formulations of prednisolone sodium phosphate are simple aqueous solutions,
which may contain a buffer (pH 5-8), a preservative such as benzalkonium chloride, EDTA,
disodium salt, or a viscoliser, and the drug can be determined by the procedures described.
The method should also be applicable to solid dosage forms utilising the extraction procedure
described.
Related steroid phosphate esters can also be determined directly by the HPLC procedure.
Fig. 4 shows the chromatogram of 0.02% m/V dexamethasone sodium phosphate, which is
similar to that obtained for prednisolone scldium phosphate. The retention time for
prednisolone sodium phosphate on this column it; the same as that for dexamethasone sodium
phosphate. Burgess,12however, has reported good resolution of related steroidal esters using
an elevated temperature and reversed-phase ZIPLC. Whether the method reported here
can also be used for this purpose must await further work, which is proceeding in our
laboratories.

FIRST-ORDER
RATE

CONSTANTS FOR THE DEGRADATION OF PREDNISOLONE


SODIUM PHOSPHATE AT pH 8.0 AT DIFFERENT TEMPERATURES

Temperaturel'C
80
90
100
110

IFirst-order rate
constaat/h1.13 x
4.145 x
1.271 x
3.323 x

Routine Quality Control of Aqueous Prednisolone Sodium Phosphate Formulations


During the stability studies it was observed thad if the degraded drug solution was extracted
prior to HPLC assay, the chromatographic peak due to the degradation products was negligible
down to about 60% of residual drug. It was thought possible, therefore, that the residual
drug could be determined by UV spectroscopy of the extraction solution. A solution containing 0.5:/, m/V of drug, O.Olyo m/V of benzalkonium chloride and O.Olyo m/V of EDTA,
disodium salt, in Sorensens phosphate buffer (pl3 7.4) was prepared and the degradation of
the prednisolone sodium phosphate a t 100 "C was determined using the following assay
procedures :
1. Solutions were appropriately diluted with water, internal standard was added and
the mixture was subsequently assayed by HPLC.
2. Solutions were extracted and assayed by IIPLC as described above.
3. Solutions were extracted and 2 ml of the aqueous 0.1 M sodium hydroxide phase were
added to a 100-ml calibrated flask containing 2 ml of 0.1 M hydrochloric acid, diluted
to volume with water and the UV absorbance of a 1 cm layer of this solution was
determined at 248 nm against an appropriate blank.
The percentage residual drug concentrations were calculated relative to the values of the
drug to internal standard peak-height ratios or absorbance a t 248nm determined at zero
time. Fig. 5 shows graphs of the percentage residual concentration on a logarithmic scale
against time for the three assay procedures. Both the direct HPLC and extraction - HPLC
methods gave straight lines, leading to rate constant values of 1.87 x
and 1.88 x
10-2 h-l, respectively. With the extraction - 1JV method the percentage residual concentration of drug is in agreement with that determined by the HPLC techniques down to
about 60% of residual drug. I t is apparent, t'herefore, that until 20-30% of the drug is
degraded, insufficient degradation products are transferred in the extraction procedure to
interfere significantly in the UV determination, and this method can be used for routine
quality control purposes.

View Article Online

May, 1,980

461

O F PREDNISOLONE SODIUM PHOSPHATE

Published on 01 January 1980. Downloaded by UNIVERSITY OF BRIGHTON on 26/03/2014 10:25:47.

aJ
t
0

6+
tII
C

0
.-+J

.-al

K
-

10

20

Time/min

Fig. 4. Chromatogram
of 0.02% m/V dexamethasone sodium phosphate.
Conditions as for Fig. 3;
detector, ultraviolet a t
254 nm, sensitivity 0.16
a.u.f.s.
1, Dexamethasone sodium phosphate ;
and 2, potassium hydrogen phthalate (internal
standard).

40

T i me/m in

Fig. 5. Comparison of assay techniques


to determine the degradation of 0 . 5 %
prednisolone sodium phosphate a t p H 7.4
and 100 "C. 0, Direct HPLC; 0,
extraction - HPLC; and 4, extraction ultraviolet.

Conclusions

It has been shown that an anion-exchange column can be used for the HPLC assay of
prednisolone sodium phosphate in the presence of its degradation products and that the
assay is suitable for stability studies. Using a simple extraction procedure, prednisolone
sodium phosphate can be separated from interfering formulatory excipients such as viscolisers.
The extraction - UV assay procedure described is particularly useful for the routine analysis
of prednisolone sodium phosphate in quality control laboratories.
References
1. Mason, H. L., J . Biol. Chem., 1938, 124, 475.
2. Hertzig, P. T., and Ehrenstein, M., J . Org. Chem., 1951, 16, 1050.
3. Wendler, N. L., and Graber, R. P., Chem. Ind. (London), 1956, 549.
4. Guttmann, D. E., and Meister, P. D., J . A m . Pharm. Assoc., Sci. Ed., 1958, 47, 773
5. Allen, A. E., and Das Gupta, V., J . Pharm. Sci., 1974, 63, 107.
6. Marcus, A. D., J . Am. Pharm. Assoc., Sci. E d . , 1960, 49, 383.
7. Kaplan, G. B., and Levine, J . , J . Assoc. Ofl.Anal. Chem., 1973, 57, 735.
I). Mader, W. J., and Buck, R. R., Anal. Chem., 1952, 24, 666.
9. Upton, L. M., Townley, E. R., and Sancilio, F. P., J . Pharm. Sci., 1978, 67, 913.
10. Knox, J . H., and Vasvari, G., J . Chromatogr. ScZ., 1974, 12, 449.
1 1 . Flynn, G. L., and Lamb, D. J., J . Phavm. Sci., 1970, 59, 1433.
12. Burgess, C., J . Chromatogr., 1978, 149, 233.

Received September loth, 1979


Accepted December 12th, 1979

You might also like