Anal. Chem. 2003, 75, 1458-1462
Continuous Monitoring of Adriamycin in Vivo Using Fiber Optic-Based Fluorescence Chemical Sensor Lu Wen-xu*,† and Chen Jian‡
Department of Pharmacy, XinJiang Tumor Hospital, Urumqi 830011, People’s Republic of China, and Pharmacy College, XinJiang Medical University, Urumqi 830000, People’s Republic of China
A novel method using a fiber optic-based fluorescence chemical sensor (FOCS) was developed for the preparation of on-line continuous monitoring of a drug in animals. First, an accurate optical design was used to enhance the intensity of light from a 100-µm optic fiber so the fluorescence signal can be detected. Second, A new solgel method was used to fix the fluorescence substance 4-(N,N-dioctyl)amino-7-nitrobenz-2-oxa-1,3-diazole (D70) on the tip of the fiber. The variation in the quenching means variation in the concentration of Adriamycin (ADM) in rabbit’s blood. ADM is determined by FOCS based on fluorescence multiple quenching. In a simple animal model, the carotid artery was catheterized with a cannula, housing a 100-µm optic fiber. The average recovery of all the tested compounds within the set concentration range was 99.4-106.2%; the within-day and between-day repeatability values were acceptable between 6.6-11.4% and 5.9-11.7%. The method permitted detection limits as low as 0.057 µg‚mL-1 at a signal-to-noise ratio of 3. The fiber optic chemical sensor is potentially useful for monitoring blood concentrations of drugs and metabilities in the biomedical field.
Figure 1. Structure and UV-visible spectrum of ADM (2 µg‚mL-1).
Figure 2. Structure and fluorescence spectrum of D-70 in Et-OH.
Adriamycin (ADM; Figure 1) is an anthracycline antibiotic with antineoplastic activity. Adverse effects include myelosuppression, alopecia, gastrointestinal disturbances, and cumulative, doserelated cardiotoxicity, which limits the amount that can be given to any patient.1 So special attention must be given to the patients receiving ADM, and blood drug concentration monitoring should be advised. A variety of methods are employed for the determination of ADM such as fluorometry, chromatography, enzyme immunoassay, and HPLC (fluorometric detector). The HPLC method is more sensitive but uses expensive apparatus, has long procedures, and a number of chemicals are involved.2 So, simple and convenient new methods for the determination of ADM are welcome.2-4
In this report, a simple, rapid, and sensitive fiber optic chemical sensor (FOCS) was developed for the determination of ADM in rabbit blood. The sensor is so small that the carotid artery can be catheterized with it. 4-(N,N-Dioctyl)amino-7-nitrobenz-2-oxa1,3-diazole (D-70, MW ) 404; Figure 2) sensing membrane immobilized on the tip of an optic fiber can distinguish the analyte ADM signals. The chemical action information was a tranduced optic signal. The assay was validated rigorously and applied successfully to the analysis of plasma samples from pharmacokinetic studies of ADM in rabbit. Compared with traditional methods, the new method would not destroy the analyte and there is no need to separate sample. This method makes it possible to monitor the object on-line and in real time and obtain more information.
* Corresponding author. E-mail:
[email protected]. † XinJiang Tumor Hospital. ‡ XinJiang Medical University. (1) ] Reynolds, J. E. F. Martindale: The Extra Pharmacopoeia, 31th ed.; London Royal Pharmaceutical Society: London, 1996; pp 565-566. (2) Chen, J.; Li, W.; Yan, C,; et al. Sci. Sin. C. 1997, 40 (4), 414-415. (3) Gu, Y. Q.; Qing, Z. Y.; Tao, B. Q.; et al. Sens. Actuators, B 2000, 66 (10), 197-199. (4) Li, W.; Chen, J. Anal. Chim. Acta 1996, 331 (8), 103-109.
EXPERIMENTAL SECTION Apparatus and Reagents. The following equipment and chemicals were used: fluorospectrophotometer (RF-540, Shimadu), bifurcated optic fiber (common leg bundle diameter 100 µm, Ossen Co.), electric microbalance (Sartorius), high-speed centrifuge (4000 rpm), ultraviolet-visible photometer (HP8453E), Adriamycin (ZheJiang HaiZhen Medicine), D-70 (Molecular
1458 Analytical Chemistry, Vol. 75, No. 6, March 15, 2003
10.1021/ac0260894 CCC: $25.00
© 2003 American Chemical Society Published on Web 02/15/2003
Figure 3. Schematic diagram of the fiber optic chemical sensor system. Figure 5. Scheme of the system used for the preparation of standard curves.
Figure 4. Scheme of the probe and the polyhole structure of D-70 sensing membrane.
Probes Inc.), and tetraethoxysilane (TEOS, TianJing Chemical Reagent Factory No. 1). All reagents are analytical grade. Male and female white rabbits weighing 2.0-2.5 kg, obtained from the animal department of Xinjiang Medical University, were used in this study. Fiber Optic Sensor System. The fiber optic sensor system (Figure 3) included an accurate optics apparatus, sensor probe, detecting apparatus, and data collecting and processing program. The light emitted from the light source, a xenon lamp (Janpen Vshio Inc.), is directed into one branch of the bifurcated optical fiber beam, transmitted to the end of the probe that is inserted into blood vessels, and induces the D-70 sensing membrane to emit signal fluorescence. The structure of the sensor probe is as follows: The D-70 sensing membrane (Figure 4) was immobilized on the naked tip of the optic fiber; the fluorescence reagent molecules were exposed on the surface and resulted in a polyporous structure. Selective reagent deposition at the fiber optic probe tips formed a polyporous structure, and the pore sizes were on the order of 1-2 µm. A useful general method of selective reagent deposition at fiber optic probe tips was established and resulted in the polyporous structure. The signal fluorescence was quenched by ADM in blood. The resulting signal was collected by the same probe, propagated along another fiber optic beam, penetrated through a filter lens (Shang Hai Spectrum Lens Co.), wavelength of 530 nm, and then arrived at the photomultiplier (PMT, 9292B), as shown in Figure 3. The quenching signal was then converted by A/D and displayed on the screen in real time. An application program was written in Visal Basic (VB) 5.0 to control fluorometry. Real-time analyses of the admittance data based on the simultaneous nonlinear fitting of both blank and sample data to the equivalent circuit were also achieved using the same VB program. Immobilization of the Probe. Preparation of the Colloidal Solution. To 10 mL of TEOS was added 40 mL of absolute alcohol, and the resultant solution A was stirred under magnetic stirring. To 2.4 mL of water and 0.4 mL of HCLwas added 17.0 mL of absolute alcohol as solution B. A and B solutes were mixed under magnetic stirring at room temperature and kept for 24 h as the colloidal solution. Preparation of D-70. About 4.35 mg of D-70 was accurately weighed, transferred to a 10-mL volumetric flask, diluted with
absolute alcohol to volume, and mixed. A portion of this solution, equivalent to about 1.0 mL, was pipetted, diluted with diluent to 3.0 mL; this solution was transferred to 5.0 mL of prepared colloidal solution, and the resulting solution was kept for use. Immobilization. The ends of a multimode optical fiber were soaked in the D-70 colloidal solution for 24 h at room temperature and then removed from the solution and kept for 8 h at room temperature; this process was repeated. The sol-gel D-70 film thickness is ∼0.01 mm. The following reaction shows the sol-gel chemical composition.
Preparation Device for the Standard Curve. The recirculating perfusion vehicle is shown in Figure 5. The flow-through cell (inner diameter 2 mm, volume 0.2 mL), which connects the bifurcated fiber bundle (BFB; one end connected with light source and the other end connected with photomultipliter), is joined to the pump (0.1 mL/min). The body’s blood circulation was imitated by the sample flowing through the cell drived by the pump. Animal Model. A simple animal model was designed for the experiment and the carotid artery was catheterized with a cannula, housing a 100-µm optic fiber. As the carotid artery is thicker than most blood vessels and has better elasticity, it was chosen as the detection point; see Figure 6:
Figure 6. Fiber optic sensor probe as inserted into a rabbit’s carotid.
The probe is inserted in the direction facing the heart at the juncture of the neck artery and left bone artery; blood circlation remains good. The blood circlation blockage was not a problem even after a long monitoring time. The animal recovered rapidly after the small wound was sewed when the experiment was finished.5 Analytical Chemistry, Vol. 75, No. 6, March 15, 2003
1459
Design Sensor. The study on designing the sensor probe refers to the following points. (1) Selection of the Carrier Matrix. The fluorescence probe was immobilized on the naked surface of the optic fiber. As a good carrier matrix, silica glass involves low-temperature hydrolysis of suitable monomeric precursors and is highly suitable for microencapsulation of a variety of molecules that have good optical stability. It can deposit enough of the indicated reagent by absorption. The initial hydrolysis and polycondensation reactions in a localized region lead to formation of colloidal particles. As the interconnection between these partides increases, the viscosity of the sol starts to increase and leads to the formation of a solid gel. The nature of the final polymeric gel can be regulated to a certain extent by controlling the rates of the individual steps. (2) Selection of a Sensing Reagent. D-70 was selected as the sensing substance based on the quenching attributes. ADM can quench the intensity of the fluorescence emission from the reagent D-70. Different concentrations of ADM can result in different quenching levels. (3) Method of Immobilization. The fluorescence probe was encapsulated by the sol-gel method and then was immobilized on the tip of the optic fiber at room temperature several times. The resulting pore sizes on the order of 1-2 um. Determination of ADM in Serum by a Sensor Method. Rabbits were used as the test subjects. Twelve hours after eating, the rabbit was placed on its back and the neck fur was cleaned. Procaine hydrochloride was injected subcutaneously along the femoral artery separating the neck artery and fixed at the side to the back of the heart; the probe was inserted into the side facing the heart. At the same time, 2 mL of sodium citrate was administered iv in order to avoid blood coagulation. The instrument was turned on and data were collected from the vein of the ear after the administration of 10 mL of ADM. Data were collected automatically and continuously by computer. Fluorescence Method To Determine the Concentration of ADM in Serum. ADM in serum was determined by fluorometry6 as the comparison method at the same time that the blood concentration of ADM was being determined by the sensor method. Sample collections was carried out before and after the administration of medicine at 5, 15, and 30 min and 1, 2, 4, 8, 12, and 24 h. The blood samples (2 mL) obtained from the ear vein were kept for 30 min at room temperature and then centrifuged for 20 min. The serum was isolated, and then the pure solution was diluted 2-fold with a salt solution. The prepared sample was determined by fluorometry at an excitation wavelength of 481 nm and an emission wavelength of 550 nm. The calibration curve of ADM in serum was in good linearity over the concentration range of 0.01-10 µg‚mL-1, and the coefficient of correlation was 0.9996. The limit of detection was 0.01 µg‚mL-1, the average recovery was 106.4%, and the RSD was 3.9%. RESULTS AND DISCUSSION Calibration Curves and Determination Limit. To 1.0 mL of drug-free rabbit whole blood, containing 0.1 mL of the anticoagulant sodium citrate, was then added the ADM standard (5) Editorial group of experimental animal anatomy in Nankai University. Experimental animal anatomy; People’s Education Press: Beijing, 1980; pp 30-32. (6) Shao, Z. G. Chin. Hosp. Pharm. 1991, 11 (5), 200-201.
1460 Analytical Chemistry, Vol. 75, No. 6, March 15, 2003
Table 1. Precision and Recovery of the Assay of Adriamycin in Rabbit Blood (n ) 15) concentration/µg‚mL-1
RSD/%
added
found
within-run
between-run
recovery/%
1.0 10.0 25.0
1.06 10.16 24.86
7.4 6.6 11.4
5.9 7.2 11.7
106.2 101.6 99.4
Figure 7. On-line monitoring I-t curve of ADM in blood at different times after iv administration of 10 mg per rabbit. The legend on the right is the ID number of the rabbit.
solutions at known concentrations, to yield final concentrations of 0.5, 0.99, 1.96, 4.76, 9.09, and 16.67 µg‚mL-1. The blood samples flowed through the recirculating perfusion vehicle, which mimicked the body’s circulatory system. Real-time analyses of the admittance data of the fluorescence intensity (F) including blank data (F0) were achieved by computer when different concentrations of ADM passed by. The regression equation was log(F0/F) ) 0.126 + 0.011C, r ) 0.9976, and N ) 5 (where C is blood concentration of the ADM). The limit of detection was 0.057 µg‚mL-1 at a signal--to-noise ratio of 3. Precision and Recovery. The intraday precision and accuracy of the analytical method were established by making five replicate measurements of whole blood samples containing added ADM at three concentrations (1.0, 10.0, and 25.0 µg‚mL-1; see Table 1). The sample was determined using the recirculating perfusion vehicle, and the data were stored in the computer. Absolute recoveries were calculated by comparing the blank data obtained with those obtained by direct injection of ADM. Reversity and Repeatability. Sample solution at concentrations of 1.0 and 20.0 µg‚mL-1 ADM in whole blood were prepared and respectively flowed through the recirculating perfusion vehicle. The repeatability curve was shown as the average relative standard deviations of 2.4 and 3.6% (n ) 5). The system was turned on, the probe was inserted into the carotid artery of drug-free rabbit, and the change in fluorescence intensity was monitored for as long as 12 h. The results showed that there was no distinct variation, but light fluctuations appeared in the response signal, and the optic fiber system expressed good recovery and repeatability. Concentration-Time Curve. The blood concentration-time curve of five rabbits after iv administration of 10 mg of ADM is shown as Figure 7, which is fit for a two-compartment model, and the peak time of arrival was 5 min. The intensity of fluorescence was increasing while the drug was deposited from the body blood. Data Processing and Analysis. The experimental data were calculated in a simulated way by means of the computing program
Table 2. Concentration of ADM in Rabbit Blood after iv Administration of a Single 10-mg Dose by Two Methods (n ) 5, χ ( s) concentration of ADM/µg‚mL-1 time/min
FOCS method
fluorescent method
1 5 10 20 30 60 120 240 480 720 1440
17.3 ( 1.1 27.3 ( 2.4 14.2 ( 1.0 9.88 ( 0.6 8.75 ( 0.4 7.15 ( 0.2 5.86 ( 0.3 5.09 ( 0.1 4.51 ( 0.3 4.03 ( 0.3 3.08 ( 0.2
15.7 ( 1.3 21.9 ( 1.9 12.8 ( 1.4 9.32 ( 0.6 8.18 ( 0.8 6.18 ( 0.2 5.32 ( 0.2 4.77 ( 0.2 4.36 ( 0.2 3.97 ( 0.3 2.95 ( 0.1
Figure 8. Comparison of the UV-visible spectrum of ADM (A) and D-70 (B).
Table 3. Pharmacokinetics Parameters of Adriamycin Extracted by Two Methods (n ) 5, χ ( s) parameter A/µg‚mL-1
B/µg‚mL-1 Vc/L‚kg-1 T1/2,R/min T1/2,β/h K21/1‚min-1 K10/1‚min-1 K12/1‚min-1 AUC/µg‚mL-1‚h CL/L‚kg-1‚min
FOCS method
fluorescent method
14.2 ( 1.8 6.02 ( 0.1 0.24 ( 0.02 12.6 ( 1.5 23.0 ( 2.3 0.0158 ( 0.002 0.0017 ( 0.0003 0.039 ( 0.005 204 ( 21 0.00041 ( 0.00005
12.8 ( 1.6 5.38 ( 0.2 0.27 ( 0.03 13.6 ( 2.1 27.5 ( 3.6 0.0148 ( 0.001 0.0015 ( 0.0002 0.035 ( 0.008 216 ( 22 0.00004 ( 0.00038
of pharmacokinetics 3p87. The mean pharmacokinetic parameters of ADM in rabbit blood extracted by two methods are listed in Table 2. The concentration of ADM in rabbit blood after iv administration of a single 10-mg dose by two methods is shown in Table 3. It has been observed from the data given in Tables 2 and 3 that the blood concentration of ADM in rabbit determined by the sensor method is slightly higher than that by the fluorescence method. This can be explained by the fact that the sample determined by the fluorescence method was lost in the dissociation process and some may be deposited on blood cells or protein. However, the sensor probe can collect nearly all the fluorescence near the top of the probe and avoid the loss of sample. Response Mechanism of D-70 to Quencher ADM. Figure 8 shows the UV-visible spectrum of ADM (A) and the emission spectrum of D-70 (B). The absorbing spectrum of ADM is similar to the emission spectrum of D-70, The quenching mechanism was inner-filter effects. As a quencher, ADM reduced the fluorescence intensity of probe D-70. The possible added factors about the response mechanism of ADM to D-70 needed to be investigated further. Figure 9 shows the fluorescent emission spectrum of D-70 with different ADM concentrations in rabbit blood. Improving the Immobilization Method of D-70. Recent research has demonstrated that silicate glasses obtained by the sol-gel method can provide a host matrix and that biomolecules immobilized by this method retain their functional characteristics to a large extent. These biofunctional glasses make it possible to retain the specificity and reactivity of biological molecules in the solid state and provide morphological and structural control that
Figure 9. Fluorescence emission spectrum of D-70 with different ADM concentrations in blood of a rabbit. Concentration (µg‚mL-1) (a) 0.0; (b) 0.5; (c) 0.99; (d) 1.96; (e) 4.76; (f) 9.09; (g) 16.67.
Figure 10. Electron micrograph of the sensor probe (magnification 2000×; signed scale, 10 µm).
is not available when the biological molecules are simply dissolved in aqueous media. Sol-gel encapsulation by the biomolecular method was widely applied for biosensors.7-9 The method is simple and affords the biosensors alternative environments that stabilize them and preserve their reactivity. Because of blood factors, the sol-gel method was used to encapsulate the D-70 molecule and was improved in this paper. As usual, the formed (7) Grant, S. A.; Glass, R. S. SPIE Proc. 1997, 2976 (12), 64-70. (8) Damian, A.; Maruszewski, K.; Podbielska, H.; et al. SPIE Proc. 1997, 2976 (19), 137-141. (9) Chan, M. A.; Lawless, J. L.; Lam, S. K.; et al. Anal. Chim. Acta 2000, 408 (23), 33-37.
Analytical Chemistry, Vol. 75, No. 6, March 15, 2003
1461
pore size is ∼10 nm after aging at higher temperature.10 The improved method was employed at room temperature, and the pore sizes formed were 1-2 µm (see Figure 10. The formed structure not only makes them suitable for sensors with drug molecules transferring freely but also prevents blood cells from entering the probe. So the obstructions of monitoring were overcomed at such a extent. CONCLUSION In this work, the feasibility of a sensor as a method to monitor a blood drug on line has been reported for the first time. The sensor based on fluorescence quenching was fabricated and showed good analytical performance. The proposed method has been successfully applied to the determination of ADM in a serum medium. The results are in good agreement with those obtained (10) Bakul, C. D.; Dunn, B.; Selverstone, J. V.; Zink, J. E. Anal. Chem. 1994, 67 (22), 1120A-1127A.
1462
Analytical Chemistry, Vol. 75, No. 6, March 15, 2003
by the last traditional method. This approach to sensor design shows some advantages including a direct communication between the coating and the surface of the transducer which is created in a very simple way. Moreover, it offers the possibility of miniaturization of the sensor which is one of the major goals of chemical sensor technology. The fiber optic chemical sensor is potentially useful for monitoring blood drugs and metabilities in the biomedical field. ACKNOWLEDGMENT This work was supported by the Natural Sciences Foundation of China (Grant 29775022). The electron micrograph of the sensor probe was offered by EM laboratory of XinJiang University operated by Li Xingxia. AC0260894
