Determination of the in vivo pharmacokinetics of palladium-bacteriopheophorbide (WST09) in EMT6 tumour-bearing Balb/c mice using graphite furnace atomic absorption spectroscopy

Palladium-bacteriopheophorbide (WST09), a novel bacteriochlorophyll derivative, is currently being investigated for use as a photodynamic therapy (PDT) drug due to its strong absorption in the near-infrared region and its ability to efficiently generate singlet oxygen when irradiated.

In this study, we determined the pharmacokinetics and tissue distribution of WST09 in female EMT6 tumour-bearing Balb/c mice in order to determine if selective accumulation of this drug occurs in tumour tissue.

A total of 41 mice were administered WST09 by bolus injection into the tail vein at a dose level of 5.0 ± 0.8 mg kg−1.

Three to six mice were sacrificed at each of 0.08, 0.25, 0.5, 1.0, 3.0, 6.0, 9.0, 12, 24, 48, 72, and 96 h post injection, and an additional three control mice were sacrificed without having been administered WST09.

Terminal blood samples as well as liver, skin, muscle, kidney and tumour samples were obtained from each mouse and analyzed for palladium content (from WST09) using graphite furnace atomic absorption spectroscopy (GFAAS).

The representative concentration of WST09 in the plasma and tissues was then calculated.

Biphasic kinetics were observed in the plasma, kidney, and liver with clearance from each of these tissues being relatively rapid.

Skin, muscle and tumour did not show any significant accumulation at all time points investigated.

No selective drug accumulation was seen in the tumour and normal tissues, relative to plasma.

Thus the results of this study indicate that vascular targeting resulting from WST09 in the circulation, as opposed to selective WST09 accumulation in tumour tissues, may be responsible for PDT effects in tumours that have been observed in other WST09 studies.


Bacteriochlorophyll-based derivatives are currently being studied as photochemotherapeutic agents, mainly due to their high molar extinction coefficients in the 760 to 780 nm region of the electromagnetic spectrum.1–3

The bacteriochlorophyll-based compounds have further benefits of low toxicity and fast clearance from tissues.

This results in little to no risk of dark toxicity or lengthy periods of post treatment photosensitivity, a factor which can be a drawback with Photofrin® based photodynamic therapy (PDT).

Palladium bacteriopheophorbide (Fig. 1), code named WST09, is a novel bacteriochlorophyll derivative that is currently in phase II clinical trials for the treatment of recurrent prostate cancer.

To date, investigations concerning the efficacy of WST09-based PDT in in vivo tumour therapy have involved irradiation during or immediately following intravenous injection of the WST09 chromophore,4,5 an approach that targets WST09 in the circulation rather than that which might accumulate in tumour tissues at times well after administration such as occurs with other approved PDT drugs.6–8

Results on WST09 have shown high cure rates of mice with prostate small cell carcinoma xenografts,9 and PDT treatment data collected so far provide preliminary indications that vasculature targeted PDT is effective with this drug.10–12

However the pharmacokinetics of WST09 accumulation and clearance have yet to be reported in an animal tumour model.

There is mounting evidence that direct cell kill only accounts for about 1 to 2 logs of direct tumour-cell killing,13 and that the required 7 to 8 logs of cell kill required for a tumour cure can mostly be related to a microvasculature target as the initial site of injury.

As a result, knowledge of the time dependent tissue and plasma distribution of WST09 in a tumour-bearing model can elucidate further the mechanism of PDT action of this compound.

In addition, knowledge of the pharmacokinetic behaviour of this compound provides valuable information pertaining to its potential to elicit skin photosensitivity or systemic toxicity.

In the current study we have determined the in vivo pharmacokinetics and tissue distribution of WST09 in an EMT6 tumour-bearing Balb/c mouse model after a single intravenous injection.

The analysis was carried out using a graphite furnace atomic absorption spectroscopy (GFAAS) method in order to selectively detect the Pd in WST09.

This is a complementary method to those reported previously for monitoring biological samples containing Pd-organometallics14–16.

Materials and methods

Materials and injection formulation

Palladium-bacteriopheophorbide (WST09) was obtained both in powder form and as a 5 mg mL−1 injection formulation in an alcohol/Cremophor EL®/NaOH based proprietary vehicle from Steba Biotech (Toussus-le-Noble, France).

Stock solution was diluted in the solvent in order to obtain a WST09 concentration of 0.5 mg mL−1 suitable for injection into the mice.

The stock and diluted solutions were protected from light at all times; the dilutions were performed in dim light conditions and all solution containers were wrapped in aluminum foil.

The diluted concentrations were verified by measuring the absorbance of WST09 at 758 nm.

All solvents were commercially available spectroscopic or HPLC grade.

Solvable® digestion solution was obtained from Canberra-Packard.

Cremophor EL® was obtained from Sigma.

Animal model

A total of 44 female Balb/c mice were obtained from Charles River, Canada.

At the time of injection the mice were between 6 and 16 weeks old and had an average body mass of 19.8 ± 1.6 g.

Each mouse carried an EMT6 mammary adenocarcinoma tumour (NCI Frederick Cancer Research and Development Center CDT Tumour Repository).

This animal/tumour model has been shown in our laboratory to selectively accumulate other Pd-containing drugs suitable for photoactivation.14

The tumours were maintained by serial transplantation of homogenized tumour tissue subdermally into the right flank 3 to 15 days prior to WST09 administration.

Under these conditions, the tumours studied were of roughly equal size, with an average weight at the time of sacrifice of 100 to 150 mg.

Animals were maintained on an ad libitum diet of Agribrands Purina Rodent Chow 5001 and tap water.

Animal care was performed in accordance with the guidelines set forth by both the Queens University Animal Care Committee and the Canadian Council for Animal Care.

All procedures as well as the experimental protocol were peer-reviewed and approved by the Queens University Animal Care Committee prior to the commencement of this study.

Absorption spectroscopy

The UV-Visible spectra of WST09 dissolved in Solvable®, or prepared in injection formulation as described below, were recorded on a UV-Vis spectrophotometer (Shimadzu UV-160, Kyoto, Japan).

Tissue injection and extraction procedure

The photosensitizer drug was introduced in the mice via bolus injections of the WST09 formulation into the tail vein at a dose level of 5.0 ±1.0 mg kg−1.

The mice were kept in the dark, with food and water ad libitum, until the time of sacrifice.

Prior to sacrifice the back of each mouse was depilated using Nair® Lotion Hair Remover (Carter Horner Corp., Mississauga, Ont., Canada) in order to provide a hairless skin sample; animals were rinsed thoroughly in warm water in order to ensure that all the Nair® lotion was removed.

Three to four mice (six in the case of the 1 and 48 h time points) were sacrificed by euthanizing with natural gas (propane) at each of 0.08, 0.25, 0.5, 1.0, 3.0, 6.0, 9.0, 12, 24, 48, 72, and 96 h post injection.

An additional three mice were sacrificed without having been administered WST09 in order to provide WST09-free baseline values.

Terminal blood samples were obtained via cardiac puncture using a syringe coated with 0.1 mL of 3.2% (0.105 M) buffered sodium citrate solution (obtained from a sterile Vacutainer® tube, Becton-Dickinson Vacutainer Systems).

The volume of the blood obtained in each case was noted and the samples were centrifuged for 10 minutes at 2500 rpm at 4 °C.

100 μL of plasma was then added to 1 mL of Solvable®.

Samples of liver, kidney, depilated skin, leg adductor muscle, and tumour were obtained by dissection and rinsed in sterile saline.

Accurately weighed samples of each tissue were added to Solvable® in a volume ratio of 2 mL Solvable® per 100 mg of tissue and these, as well as the plasma samples, were placed in a 55 °C oven and allowed to digest for 24 h.

After digestion, the samples were stored in the dark at room temperature until analysis.

The entire analysis procedure was carried out under subdued light in order to prevent photodegradation of the WST09.

Calibration standards

Calibration standards were made by dissolving WST09 in Solvable® to obtain a standard solution whose concentration was verified by UV-Vis spectroscopy.

2 mL calibration standards were then made by adding 100 mg of chicken breast (stored frozen and thawed just prior to use) to appropriate volumes of standard solution and Solvable®.

The standards were placed in a 55 °C lab oven for 24 h in order to allow for complete tissue digestion.

Calibration standards were used to construct calibration plots during GFAAS analysis.

Since tissue causes a light scattering effect, the inclusion of tissue in the calibration standards was deemed necessary.

The use of chicken breast to mimic the scattering effect of mouse tissues has been previously validated15.

Sample stability

In order to check stability, the standards were analysed for palladium content on days 1, 2, 3, and 7 post preparation.

Standards were stored in the dark, at room temperature for the duration of the study.

Absorbance peak area of palladium in solutions of 50 mg chicken tissue per mL of Solvable® was found to vary linearly with Pd concentration in the region of 5 to 95 ng mL−1 and the solutions as well as the instrument readings were found to be stable for up to 7 days.

A calibration curve was obtained each day sample analysis was performed.

GFAAS analysis

Calibration standards and tissue samples were analyzed using a Unicam 939 Graphite Furnace Atomic Absorption Spectrometer with a Unicam 247.6 nm (palladium atomic absorption line) hollow cathode lamp and a coated, ridged, graphite cuvette.

A deuterium source was used for background correction and absorbance peak area was measured.

The coated graphite cuvette was changed every 110 firings because of cuvette aging effects noted during previous method validation.15

Twenty microlitres of unknown sample or calibration standard were injected directly into the graphite cuvette without further preparation.

Three replicate measurements of the absorbance peak area were routinely obtained from each sample, from which the mean absorbance value and standard deviation were obtained.

Calibration curves were obtained as described above in order to permit determination of palladium content (ng mL−1 Solvable®) of the unknowns based on the equation of the best linear fit line of the calibration plot.

The concentration of palladium (and corresponding concentration of WST09 based on the mass fraction of palladium in WST09) per gram tissue or mL of plasma were then calculated from the concentration of palladium in each sample and known amount of tissue added to the digestion solution (molar ratio of palladium atom to the WST09 macrocycle is 1 ∶ 1).

The limit of detection of WST09 by this method has been determined elsewhere to be 330 ng g−1 in skin, 360 ng mL−1 in plasma, and 610 ng g−1 in liver, kidney, tumour, and muscle respectively15.

Determination of pharmacokinetic parameters

The mean concentration of WST09 in each tissue was plotted logarithmically versus time.

Assuming a two-compartment open model, the bi-exponential curves that best fit the data were determined using the method of residuals.17

In each case, the distribution and elimination rate constants (kα and kβ) were obtained graphically from each drug concentration versus time plot.

The highest drug concentration observed following administration, and the time at which it is observed (Cmax and tmax), were determined directly from the concentrations measured at each time point.

The area under the drug concentration versus time curve, both from time zero to infinity (AUC0-∞), and from time zero to the time at which the last quantifiable value was measured (AUC0-tlast), were calculated through application of the trapezoidal rule.17

Infinity was taken to be the clearance time, or the time at which 99.9% of the initial concentration of the drug was cleared from the tissues (as determined by extrapolation of the terminal elimination phase).

The rate of total clearance of the drug (Cl) was calculated as the dose (μg WST09 injected) divided by the area under the drug concentration versus time plot (AUC0-∞).

The apparent volume initial distribution (Vd) was calculated as the dose (μg WST09 injected) divided by the concentration of WST09 in the plasma when the concentration versus time plot was extrapolated back to zero. .

Results and discussion

The pharmacokinetic distribution of WST09 in the plasma and tissues at 0.08, 0.25, 0.5, 1.0, 3.0, 6.0, 9.0, 12.0, 24.0, 48.0, 72.0 and 96.0 h post intravenous injection was determined using GFAAS.

The mean and standard deviation of the concentration of WST09 present in plasma and each tissue at each time point tested are given below in Table 1.

WST09 accumulation was not observed in skin, muscle, or tumour at any time.

This lack of accumulation indicates that direct tumour cell kill would not be possible using this drug.

However, local irradiation of the tumour area shortly following iv administration of the drug will potentially target the tumour vasculature.

This treatment approach has been demonstrated with positive results using bacteriochlorophyll-serine18,19 as well as uncoupled bacteriochlorophyll.3

The lack of accumulation in the peritumoural skin and muscle may contribute to the selectivity of the resulting tumour destruction if such a technique is employed, especially if the tumour is the most highly vascularized of the three tissues.

The lack of accumulation in the skin also makes unwanted phototoxicity unlikely.

The level of WST09 in the plasma (Fig. 2) was measured to be maximal (19 μg mL−1) at the earliest time point tested, i.e. 5 minutes post injection.

Considering the small size of the mouse and the fact that circulation time is rapid, the level of WST09 in the plasma is actually very likely maximized within seconds of the delivery of the bolus dose.

Furthermore, clearance from the plasma is rapid, with the concentration of WST09 decreasing to zero within three hours of injection.

Therefore, if the vasculature is the intended target, the plasma data support an irradiation time immediately post iv injection of the WST09 in order to take advantage of maximal intravascular concentrations of the drug.

Initially high levels of WST09 (Cmax equal to 43 μg g−1 at 5 minutes post injection) are present in the liver that steadily decrease up to 24 h post injection (Fig. 3).

This affinity for the reticuloendothelial organs is typical of lipophilic drugs and represents the binding of the drug to high density lipoprotein (HDL) in the serum and its clearance via the bile-gut pathway.20–24

WST09 is also found to be present in the kidney (Cmax equal to 8.1 μg g−1 at 5 minutes post injection) with the levels decreasing to zero at 9 h post injection (Fig. 4).

The concentration of WST09 in the kidney is substantially less than that of the liver, consistent with clearance via the bile gut pathway.

The relatively rapid clearance of the drug from the liver and kidney reduces the risk of unwanted systemic toxicity.

The kinetic results indicate biphasic kinetics for plasma, kidney, and liver.

The existence of a distribution phase in the plasma data was postulated based on only two data points (0.08 and 0.25 h) and therefore the delineation of this phase may be subject to considerable error.

The application of a two-compartment open model was assumed to be correct based on the biphasic kinetics exhibited in the liver and kidney.

However, due to the dynamic equilibrium that exists between the amount of drug present in the plasma and tissues, biphasic kinetics are also expected in the plasma.

Therefore, discounting the existence of the initial fast distribution phase in the plasma would likely have introduced a greater margin of error than that incurred through the determination of the elimination phase using such a small number of data points.

Analysis of the pharmacokinetic and distribution characteristics in each tissue was performed using a two compartment open model.

The results of this analysis are shown in Table 2.

In each of plasma, kidney, and liver, tmax was observed at the earliest time point taken, i.e., 5 minutes post injection.

Relatively rapid clearance was observed in plasma, kidney and liver.

The total rate of clearance and the apparent volume of distribution were calculated to be 11 mL h−1 and 0.9 mL respectively.

This small volume of initial distribution is consistent with the drug initially distributing only within the plasma and blood-rich tissues.

This is what was expected based on the lack of drug found in the non-blood-rich tissues, and supports the view that the initial linear portion of the drug concentrations versus time curve for plasma was extrapolated correctly despite the small number of data points present.

The goal of this study was the determination of the pharmacokinetic profile of WST09 in an EMT6 tumour-bearing Balb/c mouse model for the purpose of determining its tumour selectivity.

Based on the results obtained by GFAAS, WST09 appears to be efficiently cleared from the plasma and tissues of this model via the liver and kidney.

The drug does not appear to accumulate in the skin, muscle, or tumour indicating that vascular targeting is most likely necessary for the effective use of this drug in PDT.

Furthermore, due to the rapid clearance of WST09 from the plasma, irradiation should be carried out immediately after intravenous injection in order to take advantage of maximal intravascular drug concentration.


GFAAS, graphite furnace atomic absorption spectroscopy; WSTO9/TOOKAD, palladium-bacteriopheophorbide; PDT, photodynamic therapy; AUC0-tlast, area under the drug concentration vs. time plot from 0 h to the last quantifiable value; AUC0-∞, area under the drug concentration vs. time plot from 0 h to infinity; Cl, rate of total clearance of the drug; Cmax, highest drug concentration measured following administration; kα, drug distribution rate constant; kβ, drug elimination rate constant; t1/2α, drug distribution half-life; t1/2β, drug elimination half-life; tmax, time at which the highest drug concentration is measured following administration; Vd, apparent volume of initial drug distribution; IR, infrared; UV-Vis, ultraviolet-visible; HPLC, high pressure liquid chromatography; rpm, revolutions per minute; HDL, high density lipoproteins.