CLIN. CHEM.

34/9, 1897-1 899 (1988)

Simultaneous Liquid-Chromatographic Determination of Prednisone and Prednisolone in Plasma

Mary H. Cheng, William V. Huang,1 and Allen I. Llpsey”2

This high-performance liquid-chromatographic (HPLC) meth- 4.6 x 250 mm, 5-p.m particle size (Alltech Associates, Inc.,
od for simultaneous determination of prednisone and its Deerfield, IL 60015). A “Speed Vac Concentrator” with a
metabolite, prednisolone, in plasma is a modification of the refrigerated condensation trap was used to dry the samples
method of Frey et al. (Clin Chem 1979;25:1944-7). Heparin- (Savant Instruments, Inc., Hicksville, NY 11801).
ized plasma (1.0 mL) with 0.1 mL of internal standard Reagents. Methanol (American Burdick and Jackson,
Muskegon, Ml 49442), and &chloromethane (J. T. Baker
solution (1 1 -deoxy-1 7-hydroxycorticosterone, 2 mg/L) is ex-
Chemical Co., Phillipsburg, NJ 08865) were ‘HPLC” grade.
tracted with 7.0 mL of dichioromethane, then washed Se-
De-ionized water was prepared with a ‘Nanopure H” system
quentially with 0.1 mol/L HCI, 0.1 mol/L NaOH, and de-
(Barnstead Co., Boston, MA 02132).
ionized water, 2.0 mL each. The extract is evaporated and
Drug standarcLs and quality control. All steroids and
the residue reconstituted with 75 p.L of mobile phase, meth-
dnigs were purchased from Sigma Chemical Co., St. Louis,
anol/H20 (40/60 by vol). Thirty microliters of this is injected
MO 63178. Stock solutions ofprednisone (8 mgfL), predniso-
onto a reversed-phase C6 column, which is eluted at 1.4 lone (20 mgIL), and the internal standard (11-deoxy-17
mLlmin. Analytical recoveries of prednisone and predniso- hydroxycorticosterone) (100 mg/L) in methanol were pro-
lone were 94-98% and 1 02-1 06%, respectively. Day-to-day pared from crystalline material and stored at 4 #{176}C (stable for
precision (CV) was 3.8% for prednisone, 6. 1% for predniso- four months).
lone. We encountered no interference from the 21 other The working standard concentrations were prepared by
steroids and 25 drugs tested. This method is simple, accu- supplementing drug-free plasma with prednisone to give
rate, and precise. concentrations of6.25, 12.5, 25, 50, 100, 200, and 400 p.g/L,
and with prednisolone to give concentrations of25, 50, 100,
AddItIonal Keyphrases : chromatography, reversed-phase 200, 400, 800, and 1600 j.cg/L. The controls were prepared by
steroids . drug assay . glucocorticoid drugs adding prednisone (50 p.g/L, final concentration) and prod-
nisolone (300 p.g/L) to drug-free plasma, which was stored at
Prednisone is the most widely used synthetic glucocorti- -20 #{176}C in aliquots (stable for three months). The working 2
coid. Several high-performance liquid chromatographic mg/L internal standard solution was prepared by diluting
(HPLC) methods have been described for measurements of the stock solution 50-fold in methanol. Pooled plasma from
prednisone, prednisolone, and other corticosteroids in plas- residual citrate-phosphate-dextrose-adenine banked blood
ma (1-5). The use ofnormal-phase HPLC has been reported was checked by chromatography, and only plasma that was
(1-3). Among the problems associated with the use of free from prednisone, prednisolone, and the internal stan-
normal-phase chromatography are (a) a reaction between dard was used to prepare the standards and the controls (7).
the silica and the compound to be analyzed (e.g., hydrogen- Mobile phase. We prepared the mobile phase, methanoll
bond formation), resulting in poor chromatographic separa- water (40/60 by vol), in 4000-mL batches, ifitering and
tion; (b) a chemical reaction of solvent with the metallic degassing with a Filter-Degasser (Model F9-256; Lazar
pump (e.g., hydrogen chloride formation), resulting in an Research Labs, Inc., Los Angeles, CA 90046) before use.
unstable baseline; and (c) the higher vapor pressure of Sample preparation. Add 1.0 mL of heparinized plasma
nonpolar solvents vs polar solvents (4). Kabra et al. (5) found and 100 pLofinternalstandard(2mgIL)toa 16 x 100mm
that prednisone and prednisolone were co-eluted by the screw-cap glass tube and mix for 5 s. Add 7 mL of dichloro-
reversed-phase column (p-Bondapak C18) method. Caldar- methane and shake for 5 mm with a mechanical shaker.
ella et al. (6) reported an analysis of cortisol by a p.- Centrifuge for 5 mm, then transfer the organic (bottom)
Bondapak reversed-phase HPLC method; however, cortisol phase into another 16 x 100 mm screw-cap tube. Add 2 mL
and prednisolone were eluted at the same time. of 0.1 mollL HC1, shake, centrifuge for 5 mm, and discard
We describe here a reversed-phase HPLC method involv- the aqueous (upper) layer. Extract the lower layer again
ing a C6 column and a mobile phase of methanollwater with 2 mL of 0.2 mol/L NaOH and finally with 2 mL of do-
(40/60 by vol). The sample-preparation procedures are sim- ionized water, discarding the aqueous layer each time.
ple; the chromatographic separation of prednisone, prednis- Transfer the organic (bottom) layer into a disposable 13 x
olone, cortisol, and other steroids is good; and the method is 100 mm glass tube, place in the vacuum concentrator, and
relatively free from interference by drugs. remove all solvent by evaporation.
Reconstitute the dried extract with 75 pL of mobile phase.
Materials and Methods Mix for 20 s, then centrifuge at 10 000 x g for 2 mm, and
Apparatus. The HPLC unit included a Model U6K injec- transfer the supernatant fluid to another vial.
tor, a Model 6000A HPLC pump, and a Model 440 single- Chromatography. Set the flow rate of the mobile phase at
wavelength (254 nm) detector (all from Waters Associates, 1.4 mL/min, the detector at 254 nm, the attenuation at 0.01
A full-scale, and the chart speed at 0.5 cm/mm. Inject 30 p.L
Inc., Milford, MA 07157). We used a Spherisorb C6 column,
of the reconstituted mixture onto the column. Quanti1’ the
steroids in the sample by comparison with a standard curve
1 Clinical Laboratory, Childrens Hospital Los Angeles, 4650 based on the ratios of the peak-height of prednisone and
Sunset Blvd., Los Angeles, CA 90027 (address for correspondence). prednisolone relative to that of the internal standard.
2Department of Pathology, University of Southern California,
School of Medicine, Los Angeles, CA. Analytical recovery. Three drug-free plasma samples were
Received January 25, 1988; accepted May 16, 1988. supplemented with known amounts of prednisone and prod-

CLINICAL CHEMISTRY, Vol. 34, No. 9, 1988 1897

 0
C
0
Table 1. RetentIon TImes for 24 SteroIds
0
0 Steroid Retention time, mm
C
.5
0 .5 Aldosterone 10.0
S 4-Androstene-3,1 7-dione 35.2
.5
0
0 C 5a-Cholestan-3-one 12.6
S
0 C
0 Corticosterone 22.7
0.
0
0
S 0 Hydrocortisone 13.1
0
0
C Cortisone 12.3
0
U
U
0
11 -Dehydrocorticosterone 15.7
.5 5 Dehydroepiandrosterone 14.5
0.
0 Deoxycorticosterone 59.4
0 11 -Deoxy-1 7-hydroxycortic 29.4
0 . Dexamethasone 21.2

IU
E 17-Ethynyl estradiol 50.7
0
47.3
C) 23.8
17-Hydroxyprogesterone 56.6
11 -Ketoetiocholanolone 15.6
5 10 16 20 26 30 Methylprednisone 21.2
1 9-Nortestosterone 42.8
Time (mm) Prednisone 11.8
Prednisolone 14.0
Ag. 1. Chromatogram of plasma supplemented wIth prednisone, 5p.Pregnano-3a,1 7a,20a-triol 43.0
prednisolone, and internal standard (11 -deoxy-1 7-hydroxycorticoster- 5p-Pregnane-3a,1 7a,20a-triol-1 1-one 42.9
one) Progesterone 133.9
Testosterone 52.5

nisolone to obtain three different concentrations of added

standards: prednisone/prednisolone of 20/100 pg/L, 40/200 and a SD of 17 /.LgfL (n =38). The detection limit for
p.g/L, and 80/400 pg/L. Each sample was analyzed in prednisone and prednisolone (signallnoise ratio 3) is 5
=

duplicate and the recovery calculated. Mean recoveries at pg/L when 1.0 mL of plasma is extracted. This is well below
different concentrations ranged from 94% to 98% for predni- the therapeutic range for prednisone (10-50 p.g/L) and
sone and 102% to 106% for prednisolone. prednisolone (30-300 p.g/L).
Standards for 24 steroids were studied by chromatograph-
Results ing each steroid individually. No interference was found
Figure 1 shows a typical chromatogram. Retention times from these steroids. Retention times are shown in Table 1.
for prednisone, prednisolone, and internal standard are Standards for 25 drugs commonly used in combination
11.8, 14.0, and 29.4 mm, respectively. with prednisone and prednisolone were investigated by
Prednisone, prednisolone, and internal standard were chromatographing each compound individually. There was
identified by their retention time. The standard curves for no interference from these drugs (Table 2).
these compounds are shown in Figure 2. The concentrations Phenytoin has a retention time similar to that of predni-
of prednisone and prednisolone are linearly related to the sone. However, phenytoin was not detected after extraction
peak-height ratio to 400 p.gfL and 1600 pg/L, respectively.
Within-day and day-to-day precision during a three- Table 2. Retention Times for Drugs
month period were assessed by analyzing plasma pools. The
Drug Retention time, mm
mean value for prednisone concentration (mean ± SD) and
Acetaminophen 2.6
coefficient of variation (CV) within-day were 48 ± 1.7 pg/L
Allopurinol 2.2
and 3.6%, respectively. The within-day CV was 1.5% for a Amitnptyline 6.4
mean prednisolone concentration of 297 (SD 4) p.gfL. The Caffeine 3.6
day-to-day CV for prednisone was 3.8%, with a mean of 48 Calcitriol 6.0
pgtL and a SD of 1.8 pgfL (n 38). The mean value for
=
Cephalothin 1.3
prednisolone was 287 pg/L, with a day-to-day CV of 6.1% Chlordiazepoxide 47.0
Chiorothiazide 2.0
Diazepam 40.8
Ephednne 6.4
Furosemide 8.0
Ibuprofen 6.0
Imipramine 10.6
Indomethacin 5.6
Metolazone 7.8
Mechiorethamine 1.6
‘ Naproxen 2.2
Phenacetin 6.4
Phenobarbital 5.4
Phenytoin 11.6
ISO 100 400 500
300
1300
400
1000
P,044.InO .41L
Pl#{149}d4ooo. JIIIL
Probenecid 1.7
Conc.ntrst#{234}on 0-O p..d.00..,.
Propranolol 1.6
6-#{243} P100.100.10 Sulfasalazine 1.4
Theophylline 2.6
Fig. 2. Standard curves for prednisone (is) and prednisolone (0),with Vincristine 6.0
11-deoxy-1 7-hydroxycorticosterone as internal standard

1898 CLINICAL CHEMISTRY, Vol. 34, No. 9, 1988

with dichioromethane and washing with acid, alkali, and
prednisolone, and hydrocortisone were well separated. Oth-
water. er steroids and drugs commonly given with prednisone do
not interfere in this procedure.
Discussion
References
The separation procedure described is a simple, precise,
1. Frey FJ, Frey BM, Benet 12. Liquid-chromatographic measure-
and specific method for simultaneously measuring predni- ment of endogenous and exogenous glucocorticoida in plasma. Clin
sone and prednisolone in plasma by isocratic HPLC. In most Chem 1979;25:1944-7.
prednisone and corticosteroid procedures previously pub- 2. Rose JQ, Juako WJ. Corticosteroid analysis in biological fluids
lished, normal-phase chromatography is used, which in- by high-performance liquid chromatography. J Chromatogr
volves a polar packing material such as silica and a non- 1979;162:273-80.
polar solvent system. Reportedly, the components of the 3. Prasad VK, Ho B, Haneke C. Simultaneous determination of
sample may form hydrogen bonds with the silica molecules, prednisone acetate, prednisolone, prednisone, cortisone and hydro-
resulting in poor steroid separation (6). In addition, high cortisone in swine plasma using solid-phase and liquid-liquid
extraction techniques. J Chromatogr 1986;378:305-16.
vapor pressure of dichloromethane in the mobile phase for
4. Stewart JT, Honigberg IL, Turner EM, Davenport DA. Improved
normal-phase columns results in an unstable recorder base-
sample extraction before liquid chromatography of prednisone and
line and a yellowish color of the solvent mixture (4). In prednisolone in human serum. Clin Chem 1982;28:2326-7.
reversed-phase chromotagraphy, non-polar packing materi- 5. Kabra PM, Tim LL, Marten U. Improved liquid-chromato-
als and a polar solvent are used. Because the silica is bound graphic method for determination of serum cortisol. Cliii Chem
to a long-chain hydrocarbon, hydrogen bonds do not form in 197925:1293-6.
these reversed-phase columns, and corticosteroids are eluted 6. Caldarella AM, Reardon GE, Canalis E. Analysis for cortisol in
as sharp, symmetrical peaks (6). In earlier reversed-phase serum by liquid chromatography. Clin Chem 1982;28:538-43.
chromatography studies, hydrocortisone and prednisolone 7. Mahnovski V, Cheng MH, Lipsey Al, Keyomarsi K. Drugs in
were not well resolved (6). As with our method, prednisone, blood donors [Tech Briefi. Cliii Chem 1987;33:189.

CLIN. CHEM. 34/9, 1899-1903 (1988)

Laboratory Evaluation of the Du Pont “Dimension Clinical Chemistry System”

D. Knight, R. Singer,1 J. M. White, and C. G. Fraser

We evaluated the analytical performance of the Du Pont Materials and Methods

Dimension#{174}Clinical Chemistry System to assess its suitabil-
Matenals
ity as a replacement for the Du Pont aca ill#{174}, currently
routinely used for analysis of pediatric, high-risk, and emer- Instruments. The Du Pont Dimension was loaned for the
gency specimens. Twenty-six analyses were studied. These purpose of this study by Du Pont (UK) Ltd., Stevenage,
generally fulfilled pre-set analytical goals for precision based Herts., U.K.
on biological variation or “state of the art.” Significant econo- Operation of the Dimension is controlled by a menu-
mies were achieved and we found the Dimension to be driven computer with 2.0 megabyte memory and two floppy
reliable, easy to operate, and simple to maintain. It also discs. Each of the available tests can be requested individ-
ually or in seven user-programmable proffles of up to 20
fulfilled safety requirements for analysis of specimens of high
tests. Reagents are supplied in plastic cartridges (Flex),
nsk to laboratory staff.
each containing sufficient reagent for 15 to 240 tests. The
reagents are contained in four compartments in liquid or
The Du Pont Dimension Clinical Chemistry System is a
tablet form (hydrated automatically when required for
recently introduced, single-module, floor-standing analyzer
analysis), and are stored at 8 ± 2 #{176}C. Reagents and speci-
currently capable of performing 37 different assays. To our
mens are dispensed by Hamilton piston syringes. Disposable
knowledge, no evaluation of this instrument has yet been
reaction cuvettes are manufactured by the instrument from
reported.
twin rolls of plastic film (Surlyn#{174})by means of a heat torch
Our evaluation comprised studies of precision, compara-
and sealer. Specimens and reagents are added to the cuvette
tive bias, carryover, and practicability (1) for 26 tests. Table
and mixed by the action of ultrasonic probes. Reactions are
1 lists the tests evaluated and, for each, the methodological
performed at 37 ± 0.1 #{176}C, after which the absorbance of each
principle, specimen volume required, assay range, and pre-
specimen is measured at one, two, or three wavelengths.
cision as quoted by the manufacturer, i.e., only in terms of
The tops of the cuvettes are heat-sealed, and the web of used
coefficient of variation (CV).
cuvettes is directed into a waste container. Analysis time is
Our main aim was to assess the suitability of the Dimen-
between 2 and 12 miii per test. Up to 380 individual tests
sion for pediatric, emergency, and high-risk specimen analy-
per hour can be done, 200 photometric and 180 with the ion-
ses, as currently done in our laboratory with a Du Pont aca.
selective electrodes. Emergency tests can be performed
during routine operation. Test modes include endpoint,
endpoint with reagent blank, endpoint with specimen
Department of Biochemical Medicine, Ninewells Hospital and
Medical School, Dundee DD1 95Y, Scotland. blank, and kinetic. Bichromatic analyses or Allen correc-
‘Address correspondence to this author. tions (2) are used in the calculations, and calibration curves
Received April 4, 1988; accepted May 16, 1988. can be generated for nonlinear data. The ion-selective

CLINICAL CHEMISTRY, Vol. 34, No. 9, 1988 1899