Journal of Chromatographic Science, Vol.

33, July 1995

Determination of Propofol in Blood by HPLC.

Comparison of the Extraction and Precipitation Methods
Andrzej L. D a w i d o w i c z *
Department of C h e m i c a l Physics and P h y s i c o c h e m i c a l Separation Methods, Faculty of Chemistry, M a r i a C u r i e S k l o d o w s k a University,
20-031 Lublin, M . C S k l o d o w s k a Square 3 , Poland

A n n a Fijalkowska
Department of Anaesthesia a n d Intensive Therapy, University School of M e d i c i n e , 2 0 - 0 9 0 L u b l i n , J a c z e w s k i e g o Street 8 , Poland

Abstract body where they can accumulate, for example, in the adipose
tissue (5). The accumulated drugs can be released from such
The use of intravenous agents to maintain anesthesia has become tissues in the postoperative period, which puts not only the pa­
increasingly popular since the introduction of propofol. Its tient's life at risk but also all those people and things that de­
popularity results from the fact that in 95% of patients given 2.5 pend on the psychophysical functioning of the patient. This
mg propofol per kilogram of body weight, the induction of the drug problem is especially important for short-acting anaesthetic
is rapid and smooth and is followed by rapid recovery with a low drugs frequently used in one-day surgery (6). Such anesthetic
incidence of postoperative side effects. In order to study the agents are often applied for short medical interventions after
physiological effects of an intravenous anesthetic agent, it may be
which the patient leaves the hospital immediately.
necessary to determine its concentration in blood. This paper deals
Propofol (Diprivan) (2,6-diisopropylphenol) is one of the in­
with the high-performance liquid chromatographic analysis of
propofol in blood. The concentration values are calculated from
travenous anesthetic drugs used for anesthetic induction and its
the application of extraction and precipitation procedures. maintenance (7,8). It is also employed as a sedative (9). The pa­
Observed differences resulting from the two analytical methods are tient loses consciousness 30-50 s after receiving the drug and
discussed and interpreted with reference to investigations of blood, remains asleep for about 4-6 min (6-8). For this reason, physi­
plasma, and water solutions of propofol and the precipitation of cians readily use propofol for one-day surgery.
solid elements in blood. Despite its rapid action in the body, propofol is not metabo­
lized completely in a few minutes. It is present in the blood, the
adipose tissue, the liver, and the brain for a few hours, its con­
centration decreasing with time (8-10). Of course, its pres­
Introduction ence influences the patient's psychophysical efficiency, which is
confirmed by the changes in the electroencephalographic
Significant progress in the field of biochemical and biomed­ records of the brain waves (6).
ical analysis has opened numerous possibilities for measuring Analyzing and monitoring the concentration of propofol in
the concentration of various drugs in blood (1). Presently, drug blood are important for determining when a patient can safely
monitoring helps to decrease the risk of drug-induced compli­ leave the hospital. In addition, in the case of the continuous in­
cations, which is especially important when strong therapeutic travenous infusion of propofol, the estimation of drug concen­
agents are used. tration in blood helps to find the minimal dose sufficient for an
Individual sensitivity to medicines varies because of differ­ individual patient to maintain sleep.
ences in metabolism and renal elimination, which are deter­ Two methods are described in the literature for the analysis
mined by genetic factors and the physical condition of an or­ of propofol in blood (11,12). According to the first procedure
ganism or an individual. Consequently, the same drug dose (11), propofol was extracted from blood using cyclohexane.
can lead to different effects in different patients. Based on phar­ After solvent evaporation, the remaining sample, which was
macokinetic developments and modern analysis techniques, it diluted in the mobile phase, was injected into the column. In
may be possible to optimize individual doses by measuring the the second method (12), the supernatant obtained from blood
therapeutic agent concentration in blood (2). after protein precipitation and centrifugation was directly in­
Currently, anesthesiologists employ many groups of anes­ jected into the chromatographic column. In preliminary in­
thetic agents (3). Regardless of the administration technique, all vestigations, different concentrations were found in the same
anesthetic drugs penetrate the cells of the central nervous blood sample depending on which of the two methods was ap­
system (4). They are also transported to other tissues of the plied. Both calibration curves for propofol in blood, plasma, and
* Author to whom correspondence should be addressed. water samples and concentration values of propofol in blood

R e p r o d u c t i o n ( p h o t o c o p y i n g ) of editorial c o n t e n t of this j o u r n a l is prohibited w i t h o u t p u b l i s h e r ' s p e r m i s s i o n . 377


samples differing in hematocrit values (i.e., in the amount of
solid elements in blood) were considered in order to explain this
difference.
Experimental
Equipment
A Gilson solvent delivery system (Middleton, WI) composed
of two high-pressure pumps (Models 305 and 306), a Model 805
manometric module, and a Model 811 C dynamic mixer was
used. The ultraviolet-visible variable wavelength detector was
a Model 308 from MIM (Budapest, Hungary).
Chromatographic separations were carried out using a 250-
x 4-mm i.d. column packed with a laboratory-made reversed-
phase octadecyl silane sorbent. The samples were injected into
the column by a Model 7125 injection valve from Rheodyne
(Cotati, CA).
A mixture composed of 67% acetonitrile (gradient grade for
chromatography, LiChrosolv series, Merck; Germany) and 33%
double-distilled water (pH 4.0) were used as the mobile phase.
The desired pH was achieved by the addition of acetic acid (ap­
proximately 0.3-0.5 mL depending on the pH of the water).
Propofol was obtained from ICI-Pharma (Goteborg, Sweden).
Methods
Calibration standards
The samples for the calibration curves were prepared from
a solution containing 10 mg propofol in 100 mL acetonitrile.
The solution was appropriately diluted using water (to prepare
water standards), blood (to obtain blood standards), or plasma
(to prepare plasma standards) to obtain the following concen­
Figure 1. Concentration of propofol in the blood of a patient after an intravenous dose of 2.5 mg per
kilogram body weight. The solid and dashed lines correspond to the extraction and precipitation
methods, respectively. Mean values plus or minus the standard deviation (n = 7) are shown.
378
Journal of Chromatographic Science, Vol. 33, July 1995
trations: 0.5,1.0,2.0,4.0,7.0, and 10.0 μg/mL. By using these
three sets of solutions and following either the extraction (11)
or precipitation (12) procedure, five calibration curves were
plotted: propofol blood solutions, extraction method; propofol
water solutions, extraction method; propofol blood solutions,
precipitation method; propofol plasma solutions, precipita­
tion method; and propofol water solutions, precipitation
method. To construct these curves, the ratios of the peak
heights from propofol and the internal standard were used.
Following the methodologies described in the literature
(11,12), thymol was employed as an internal standard in the ex­
traction procedure, and n-dibutylphthalate was used in the
precipitation procedure. Thymol was added to the blood,
plasma, or water samples as a single methanol solution,
whereas n-dibutylphthalate was added to the samples as a
component of the precipitating solution.
Reagents
All chemicals, except for those mentioned previously, were
obtained from the Polish Factory of Chemical Reagents-POCh
(Gliwice, Poland) and were of analytical grade. The blood used
for the preparation of standards and the blood samples used for
analysis were collected in tubes containing sodium citrate and
stored at 4°C for no longer than 5 days. A stock solution of
thymol in methanol (1 mg/mL) was diluted with methanol to
the appropriate concentration. Tetraethylammonium hydroxide
(TEAH) (25% in ethanol) was diluted with ethanol (3:37). The
precipitating solution, which contained an internal standard,
was prepared by diluting 2 mL n-dibutylphthalate solution (1
mg in 1 mL acetonitrile) in 100 mL acetonitrile-65% per­
chloric acid (67:33, v/v).
Extraction procedure
The internal standard solution (thymol
dissolved in methanol) (20 pL), phosphate
buffer (1 mL of 0.1M NaH PO ), and cyclo-
2 4
hexane (5 mL) were added to 1 mL blood. A
vessel containing all the components was
vigorously shaken for 15 min at 200 rpm.
After centrifugation (1200 ×g for 5 min) to
separate the phases, an aliquot of the cyclo-
hexane layer (4.5 mL) was transferred to a
clean tube to which TEAH solution (50 μL)
was added. The solvent was evaporated to
dryness using a stream of nitrogen. The
residue was redissolved in a mobile phase
and injected into the chromatographic
column. Water standards were subjected to
the same procedure.
Extraction of propofol from the solid phase
separated from blood
Blood samples (40 mL) containing
propofol (400 μg each) were centrifuged
(1200 ×g for 10 min) in order to separate the
two phases. The whole amount of the solid
phase from one sample (12-22 mL, de-
Journal of Chromatographic Science, Vol. 33, July 1995
pending on the blood hematocrit) was subjected to the extrac­
tion and concentration procedures previously described and fi­
nally to chromatographic separation.
Precipitation procedure
A known amount of blood was centrifuged in order to sepa­
rate plasma from the solid elements of blood. The precipi­
tating solution containing n-dibutylphthalate as the internal
standard (0.5 mL) was added to the plasma (0.5 mL) and mixed
for 2 min with a vortex mixer. The precipitated substances
were separated by centrifugation (1200 ×g for 5 min), and the
supernatant was injected into the chromatographic column.
The same procedure was used for plasma and water standard
solutions, except without preliminary centrifugation.
Sorbent
The sorbent for chromatographic separation was obtained by
the chemical modification of a laboratory-made controlled
porosity glass (CPG) (mean pore diameter, 128 A; specific surface
2 3
area, 318 m /g; pore volume, 1.17 cm /g; particle size, 7 μm)
with ODS according to a procedure described elsewhere (13,14).
The coverage density of the support surface by octadecyl radicals
2
was 4.01 μmol/m (21.94% carbon content).
The specific surface area measurement of the support (nec­
essary for coverage density calculation) was carried out by the
Brunauer-Emmet-Teller (BET) method using a Model 1800 ni­
trogen sorptomat from Carlo Erba (Milan, Italy).
The mean pore diameter and pore volume of the CPG were
calculated from the porosimetric data. Porosimetric measure-
ments were CARRIED out with a Model 4000 Mercury Porosimeter(Carlo Erba).
porosimeter (Carlo Erba).VVV1VVU11
Figure 2. Calibration curves obtained using the extraction method. The height ratios of propofol and
the internal standard peaks are plotted against the propofol concentrations. The solid line with rings tion curves constructed in the precipitation
(slope, 0.5649) corresponds to blood solutions. The standard deviation values (n = 7) for concentra­
tions of 0.5,1.0, 2.0, 4.0, 7.0, and 10.0μg/mLwere 0.007, 0.046,0.061, 0.130, 0.271, and 0.353,
respectively. The dashed line with triangles (slope, 0.5811) corresponds to water solutions. The stan­
dard deviation values (n = 7) for the same concentrations used for the blood solution were 0.006,
0.042,0.057,0.123,0.259, and 0.341.
The amount of bonded ODS was calculated on the basis of an
equation from reference 15, and the data was determined by a
Hewlett-Packard Model 185 CHN analyzer.
Controlled porosity glass was prepared from Vycor glass ac­
cording to a procedure described elsewhere (16,17).
Results and Discussion
The aim of this paper was to compare the results of propofol
analysis carried out by means of the extraction and precipita­
tion methods (11,12). The curves obtained by plotting propofol
concentration in blood as a function of time (from the injection
moment of a single propofol intravenous bolus dose, which is
2.5 mg of propofol per 1 kg of the patient's body weight) are
shown in Figure 1. The values for these curves were quanti­
tated using the regression parameters obtained from appro­
priate calibration graphs. Significant differences between the
corresponding magnitudes are seen. The propofol concentra­
tions obtained from the precipitation analysis (dashed line)
are about 30% lower than those obtained from the extraction
procedure.
On the basis of this comparison, at least one procedure
seems to include either a wrong preparative assumption or
incorrectly chosen internal standards. Systematic or manual
errors made during sample preparation are also possibilities.
The calibration curves relating to the extraction and the
precipitation procedures are plotted in Figures 2 and 3, re­
spectively. Solid lines correspond to the blood standards. The
calibration curves for the water series are drawn in dashed
lines. The dashed-dotted line (Figure 3)
refers to the plasma solutions.
By analyzing the two sets of calibration
curves independently, one for the extrac­
tion method (Figure 2) and one for the pre­
cipitation method (Figure 3), small differ­
ences are seen. The slopes of the water
calibration curves (dashed lines) are a little
higher in comparison with the calibration
line constructed for blood and plasma stan­
dards. In the precipitation method, the dif­
ference is more explicit. These small differ­
ences can result from the different behavior
of the internal standards during the sample
preparation procedures. This behavior is
probably connected with the use of two dif­
ferent systems: one that contains blood pro­
teins (blood or plasma standards) and one
that is protein-free (water standards). This
conclusion is additionally confirmed by a
very slight difference between the calibra­
procedure when blood and plasma standard
solutions were used (compare the solid and
dashed-dotted lines in Figure 3).
The chemical structure of thymol (5-
methyl-2-isopropyl-l-phenol) is very sim-
379
 Journal of Chromatographic Science, Vol. 33, July 1995

ilar to that of propofol (2,6-diisopropylphenol). Both the parti­ greater difference between the structures of propofol and n-
tion coefficients of the substances in the extraction process dibutylphthalate leads to a somewhat greater difference be­
and the runs of the calibration curves are very similar. The tween the calibration curves (compare standards for blood,
plasma, and water series) obtained in the pro­
cedure involving the precipitation process.
The small differences within Figures 2 and
3 suggest the influence of protein presence
on the behavior of propofol and the internal
standards, but their presence does not ex­
plain the significant differences (exceeding
30%) seen in Figure 1.
In Figures 4 and 5, the chromatograms of
the extracts from water (4A) or blood (4B)
containing propofol and the chromatograms
of the supernatants obtained from the precip­
itation procedure of plasma (5A), blood (5B),
or water (5C) solutions of propofol are shown.
All these chromatograms correspond to the
samples containing the same amount of
propofol: 4 μg in 1 mL blood, plasma, or water.
As shown in Figure 4, in the extraction
Figure 3. Calibration curves obtained using the precipitation procedure. The height ratios of propofol procedure, the heights of the propofol peaks
and the internal standard peaks are plotted against the propofol concentrations. n-Dibutylphthalate was are almost equal regardless of the type of sol­
used as the internal standard. The solid line with rings (slope, 0.4711) corresponds to the blood solu­
vent used (blood or water). This proves that
tions. The standard deviation values (n = 7) for concentrations of 0.5,1.0,2.0,4.0, 7.0, and 10.0 μg/mL
almost the same amount of propofol is ex­
were 0.009,0.058,0.072,0.148,0.299, and 0.401, respectively. The dashed-dotted line with squares
tracted from blood as from water standards.
(slope, 0.4789) corresponds to plasma solutions. The standard deviation values (n = 7) for analogous
standards were 0.008,0.055,0.067,0.141,0.283, and 0.374, respectively. The dashed line with tri­
The case of the precipitation procedure is
angles (slope, 0.5026) corresponds to the water standards. The standard deviation values (n = 7) for different (see Figure 5). As in Figure 4, the
the same concentrations were 0.007, 0.044,0.064,0.132,0.272, and 0.356. similarity of the height ratio of the propofol
peak to the internal standard peak is notice­
able and confirms the conclusions made on
the basis of the plots presented in Figures 2 and 3 about the sim­
ilar behavior of propofol and the internal standards. The chro­
matograms in Figure 5 indicate that the propofol concentration
in the supernatant (after a partial protein precipitation from
blood or plasma solution) is lower than that in the mixture ob­
tained by adding the precipitating solution to the water solution
(compare the heights of propofol peaks marked P). The differ­
ence is even more explicit between chromatograms 5A (plasma
solution) and 5C (water solution). Because of a small diminution
of plasma sample volume (the result of removing denatured
proteins from plasma after precipitation and centrifugation),
an inverse situation can be expected. Co-precipitation of a por­
tion of propofol together with denatured proteins and a probable
propofol decrease due to its adsorption on precipitated proteins
seem to be reasonable explanations of the observed phenomena
(Figure 5) and explain the differences shown in Figure 1.
When chromatograms 5A (plasma solution) and 5B (blood
solution) are compared, the difference in the heights of the
propofol peaks is also observed. This case is of a more complex
nature. Assuming co-precipitation of propofol with denatured
proteins, it should also be taken into account that, in agreement
with the procedure, the solid elements of blood were separated
during the preliminary centrifugation process, decreasing the
sample volume significantly (about 40%). Consequently, despite
Figure 4. Chromatograms of extracts from the water (A) and blood (B)
the propofol concentration diminution (as a result of its co-
solutions containing propofol (4 μg/mL). Peaks P and T are propofol and
precipitation and/or adsorption), the propofol peak observed in
thymol, respectively.
Figure 5B is higher than that in 5A.

380
Journal of Chromatographic Science, Vol. 33, July 1995
Figure 5. Chromatograms of supernatants from the precipitation procedure: A, plasma; B, blood; C,
water. The peaks P and n-DBPhT are propofol and n-dibutylphthalate, respectively.
Table I. Concentrations of Propofol Extracted from Solid
Elements in Blood (Calculated per 1 mL of Blood)*
Hematocrit (%)
35 38
Propofol
cone. [μg/mL] 0.47 0.55
*lnitial blood s a m p l e ( 4 0 m L ) c o n t a i n e d 1 0 μg propofol in 1 m L b l o o d .
The precipitation procedure required the removal of solid el­
ements in blood during the first stage of sample preparation.
The next step involved protein precipitation from the obtained
plasma. According to Plummer (11), propofol is significantly
associated with the formed elements of blood, and therefore,
whole blood samples are preferred for pharmacokinetic
analysis. This raises the question of if a portion of the propofol
is lost during the preliminaryμcentrifugationwhen the solid el­
ements of blood are removed. A comparison of the chro­
matograms in Figures 5B and 5C suggests such a possibility.
Table I lists the concentrations of propofol extracted from
solid elements of blood. These values are based on 1-mL samples
of blood and were calculated from the analysis of extracts from
solid elements separated during the centrifugation process. Each
sample (40 mL) contained 400 μg propofol (10 vg/mL). In order
to differentiate the amount of solid elements in a simple way,
blood samples of various hematocrit values were used for this in­
vestigation. Approximately 4-6% of the total amount of propofol
contained in blood was extracted from the solid phase after cen­
trifugation. A part of this propofol was probably dissolved in
plasma that remained among the solid elements in the solid
phase. Another part was probably bonded with the basic centers
of the solid elements. The data in Table I suggest an increasing
tendency of the propofol concentration re­
maining in the solid phase with an increase
in hematocrit values. Of course, this con­
clusion is not unequivocal because various
quantitative relations of solid components
in blood (dependent on individuals) should
be taken into account.
Conclusion
Concentration values of propofol in blood
calculated using the precipitation method
were significantly lower than those obtained
by the extraction method. The calibration
curves constructed using water, plasma, and
blood standard solutions did not show a sig­
nificant difference in behavior (during the
extraction or precipitation process) of in­
ternal standards and propofol. During the
calibration procedure, the influence of the
presence of protein was very small. The
chromatograms prove that, in the precipi­
tation method, a portion of propofol was
lost from the sample, probably as a result of co-precipitation with
denatured proteins and/or adsorption on them. Additional mea­
surements showed that a small amount of propofol remained in
the solid phase separated during the centrifugation process.
43 46 49
When using the precipitation method, the propofol standard
solution should be prepared from the blood of individual patients
who are to receive propofol. This conclusion results from the fact
0.52 0.56 0.59 that propofol can bond with some solid elements in blood; the
concentration of these elements changes from patient to patient.
The extraction method seemed to be more precise. It involved ex­
traction of propofol from whole blood without the protein pre­
cipitation process. Consequently, the co-precipitation of propofol
and its possible adsorption on denatured proteins was not a
problem.
Reference
1. G . W . Peng and W . L . Chiou. Analysis of drugs and other sub­
stances in biological samples for pharmacokinetic studies. J.
Chromatogr. Biomed. Appl. 531: 3 (1990).
2. Farmakologia kliniczna (Clinical Pharmacology). A . Chodera
and Z.S. Herman, Eds. P Z W L , Warszawa, Poland, 1986, p 8 1 .
3. Textbook of Anaesthesia, 2nd ed. A.R. Aitkenhead and G . Smith,
Eds. Churchill Livingstone, N e w York, NY, 1990, pp 1 5 3 , 1 7 5 .
4. M. Naguib and A. Sari-Konzel. Thiopentone-propofol hypnotic
synergism in patients. Br. J. Anaesth. 6 7 : 4 (1991).
5. W . D . Wylie and H.C. Churchill-Davidson. A Practice of Anaes­
thesia, 5th ed, A Lloyd-Luke Publication. H.C. Churchill-Davidson,
Ed. Year Book M e d i c a l Publisher, C h i c a g o , IL, 1 9 8 4 ,
p626.
6. A . Fijalkowska, A . Nestorowicz, B. Darwaj, A . Klos, R. Sekrecki,
and J. Falk. Effect of short-term propofol anaesthesia on brain
function. Anestezjologia, Intensywna Terapia XXV: 33 (1993).
7. A . Shafer, V.A. Doze, S.L. Shafer, and P.F. White. Pharmacoki­
netics and pharmacodynamics of propofol infusions during gen­
eral anesthesia. Anesthesiology 6 9 : 3 4 8 (1988).
381

8. P. Dailland, I.D. Cockshott, J.D. Lirzin, P. Jacquinot, J.C. Jorrot, J.
Devery, J . L . Harmey, and C . Conseiller. Intravenous propofol
during cesarean section: Placental transfer, concentrations in
breast milk, and neonatal effects. A preliminary study. Anesthe­
siology 71: 827 (1989).
9. J . Albanese, C . Martin, B. Lacarelle, P. Saux, A . Durand, and
F. Gouin. Pharmacokinetics of long-term propofol infusion used for
sedation in I C U patients. Anesthesiology 7 3 : 2 1 4 (1990).
10. B. Marsh, M. White, N. Morton, and G . N . C . Kenny. Pharmacoki­
netic model driven infusion of propofol in children. Br. J. Anaesth.
67:41 (1991).
11. G.F. Plummer. Improved method for the determination of propofol
in blood by high performance liquid chromatography with fluo­
rescence detection. J. Chromatogr. 421:171 (1987).
12. I. Pavan, E. Buglione, M. Massiccio, C . Gregoretti, L. Burbi, and
M. Berardino. Monitoring propofol serum levels by rapid and
sensitive reversed-phase high-performance liquid chromatography
during prolonged sedation in I C U patients.j. Chromatogr. Sci. 30:
164-66(1992).
382
Journal of Chromatographic Science, Vol. 33, July 1995
13. B. Buszewski, A . Jurasek, J . Garaj, L. Nondek, and D. Berek. The
effect of the reaction medium on the coverage density of C 1 8
chemically bonded phases. J. Liq. Chromatogr. 10:2325 (1987).
14. H. Engelhardt and H. Muller. Optimal conditions for the reversed-
phase chromatography of proteins. Chromatographia 19: 77
(1984).
15. G . E . Berendsen and L. de G a l a n . Preparation and chromato­
graphic properties of some chemical bonded phases for reversed-
phase liquid chromatography. J. Liq. Chromatogr. 1:561 (1978).
16. A . L . D a w i d o w i c z , J. Matusewicz, and J. Wysocka-Lisek. Deter­
mination of residual boron in thermally treated controlled-porosity
glasses, by colorimetry, spectrography and isotachophoresis.
Talanta 36:581 (1989).
17. A.L. Dawidowicz, A. Waksmundzki, and A. Derylo. Column pack­
ings for liquid permeation chromatography obtained by crushing
coarse fractions of porous glasses. I. Evaluation of applicability of
glass packings. Chem. Anal. (Warsaw) 24:811 (1979).
Manuscript accepted October 1 3 , 1 9 9 4 .