2.1.

Fluorophore Preparation

Aluminum phthalocyanine tetrasulfonate (AlPcS ₄) was purchased from Porphyrin Products (Logan, Utah). The compound was dissolved in 0.9% sterile saline and was used as a stock solution for administration to the animals and measurements in tissue-simulating phantoms.

2.2.

Animal Procedures

Sixteen New Zealand White (NZW) specific-pathogen free (SPF) rabbits weighing 2 to 4 kg were used. Rabbits were weighed and then injected with the desired dose of AlPcS ₄, which ranged from 0.5 to 4 mg/kg. One rabbit was injected with 1 mL of sterile 0.9% saline as a control. Fluorescence measurements were performed under general anesthetic 18 to 28 hours after the injection of the fluorophore. The animals were initially anesthetized with an intramuscular injection of lidocaine hydrochloride (Xylazine®, 4 mg/kg), followed by an injection 10 min later of a mixture of ketamine hydrochloride (Ketalean®, 40 mg/kg) and acepromazine maleate (Atravet®, 0.75 mg/kg). Anesthesia was maintained by gaseous anesthetic (isofluorane, 1.5%) for 12 rabbits, and by injections of the Ketalean®/Atravet® mixture for the other four. Animal ethics approval was obtained for these experiments.

2.3.

Experimental Apparatus

Estimates of fluorophore concentrations were performed using two different instruments. The first instrument used the fluorescence/reflectance (F/R) ratio to estimate fluorophore concentrations. A schematic for this instrument is shown in Fig. 1. The excitation light source was a 640-nm steady-state diode laser (OZ Optics, Ottawa, Ontario). The laser was coupled to the source fiber of a fiber-optic probe (Fiberguide Industries, Stirling, New Jersey) which had two detection fibers with source-detector separations of 0.65 mm and 1.35 mm. The detection fibers were connected to an Ocean Optics (Dunedin, Florida) SD2000 dual channel CCD spectrometer. A 665-nm cut-on filter (Oriel, Stratford, Connecticut) was used in the fluorescence channel to reduce the intensity of the excitation line. A background spectrum obtained from the skin surface with the laser turned off was subtracted from all measured spectra. The instrument was interfaced to a laptop computer where the data were processed and displayed. The excitation component of the reflectance signal was measured by integrating the spectrum from 635 to 645 nm. The fluorescence signal was the integral from 675 to 685 nm. A sample spectrum is shown as an inset to Fig. 1, with the fluorescence integration limits indicated.

Figure 1

Experimental apparatus for F/R ratio measurements. The optical fibers of the F/R probe were 200 μm in diameter. The inset spectrum is an overlay of the fluorescence and reflectance channels. The region of integration for the fluorescence channel is shaded gray.

The second instrument was based on exciting and detecting time-resolved fluorescence using a single optical fiber. A schematic of the apparatus is shown in Fig. 2. A 635-nm pulsed diode laser (BHL-150, Becker & Hickl, Germany) was used to excite the fluorophore. The laser had a repetition rate of 50 MHz and a pulse width of approximately 150 ps. Light from the laser was coupled into a mirror assembly (ThorLabs, Newton, New Jersey) that separated fluorescence from the excitation light using a thin dichroic mirror (Oriel, Stratford, Connecticut). The mirror/lens assembly coupled the excitation light into a single 200-μm optical fiber probe, which was used to excite and detect the fluorescence. The fluorescence collected by the probe was reflected from the mirror into an optical fiber coupled to a photomultiplier tube (H5783P-01, Hamamatsu, Bridgewater, New Jersey). The fluorescence was filtered with 665-nm cut-on and 680-nm band-pass filters (Oriel, Stratford, Connecticut) to reduce the signal from scattered excitation light (from the mirror and the sample) and to limit the spectral width of the detected fluorescence. The pulse from the photomultiplier tube was amplified, time delayed, and used as a “start” input to a time-correlated single photon counting card (SPC-630, Becker & Hickl, Germany) operated in reversed “start-stop” mode. The “stop” input was from an electronic synchronization pulse from the laser driver.

Figure 2

Experimental apparatus for single fiber measurements. The solid gray line is the excitation path and the dotted line the fluorescence. Dash-dots are electrical connections. The time-correlated single photon counting (TCSPC) module was run in reverse time mode, with the time measurement started by the detection of a photon, and stopped by the synchronization pulse from the laser.

2.4.

Measurements in Tissue-Simulating Phantoms

Measurements using the F/R and single fiber instruments in vivo were compared to separate calibration measurements. These were made on tissue-simulating phantoms with five combinations of optical properties containing known concentrations of AlPcS ₄. The optical properties of the phantoms at 635 nm were μ_s ^′=0.6 mm ⁻¹, μ_a=0.001 mm ⁻¹; μ_s ^′=0.6 mm ⁻¹, μ_a=0.1 mm ⁻¹; μ_s ^′=1.2 mm ⁻¹, μ_a=0.01 mm ⁻¹; μ_s ^′=2.4 mm ⁻¹, μ_a=0.001 mm ⁻¹; μ_s ^′=2.4 mm ⁻¹, μ_a=0.1 mm ⁻¹. The tissue-simulating phantoms comprised a dilute aqueous suspension of Higgins India ink (Eberhard Faber Inc., Lewisburg, Tennessee) for absorption and Intralipid-20% (Pharmacia Corporation, Peapack, New Jersey) for scattering. The concentrations of ink and Intralipid-20% required to achieve these combinations of optical properties were determined in previously published work.¹⁵ AlPcS ₄ concentrations ranging from 0.1 to 5.0 μg/mL were added to the phantoms prior to fluorescence measurements.

The line of best fit to the measured signal (either F/R or the time-integrated fluorescence measured with the single optical fiber) versus concentration for all phantoms was used to estimate tissue concentrations from in vivo measurements. Note that both techniques minimize the dependence of signal on optical properties so that a single calibration curve can be used for the wide range of μ_s ^′ and μ_a. The measured time-resolved fluorescence curves contained contributions due to the fluorophore and instrumental autofluorescence (which was measured when tissue-simulating phantoms containing no AlPcS ₄ were used). The fluorescence due to AlPcS ₄ was the difference of these two curves (each curve normalized to the collection time). The area under this difference curve was the total fluorescence signal used for determining concentration. The fluorescence lifetime was estimated using a weighted least-squares fit to a single exponential decay.¹⁶

2.5.

Measurements In Vivo

Fluorescence measurements were performed at several locations on the rabbits. The two instruments were used sequentially to prevent optical cross talk. Sites on the skin were located on the back (over the dorsal muscle) and on the leg. These sites were shaved and then depilated with a hair removal product (Nair®). After data were acquired for the skin, an incision was made and the skin reflected to expose the underlying muscle. Fluorescence was measured on the muscle in roughly the same locations as on the overlying skin. Interstitial measurements in muscle (fiber embedded ∼5 mm below the surface) were also performed using the single optical fiber. An 18-gauge needle was used to guide the fiber into the muscle, and was removed before commencing the fluorescence measurement. After sacrifice, the liver and kidney were surgically exposed. Measurements were performed on the surface (F/R and single fiber) and interstitially (the end of the fiber placed 2 to 3 mm into the organ using the same technique as in muscle). Skin and muscle samples were harvested from the marked measurement sites immediately after sacrifice, and from the liver and kidney shortly after the measurements were performed. The tissue samples were stored at −80 °C for subsequent chemical extraction.

The background for the time-resolved system was the average of the signal measured with the fiber tip in water and in air. This was found to be close to the signal found for the control (no AlPcS ₄ injected) animal. Time-resolved data were fitted to either a single or double exponential decay.¹⁶ For those data fitted with a double exponential decay, the time integrated fluorescence signal, F, was corrected for shorter lifetimes. The corrected fluorescence signal is given by the expression:

2.6.

Chemical Extractions

Measurements of concentration in tissue were performed by fluorimetry based on the extraction protocol of Lilge et al.¹¹ Three samples were cut from each piece of harvested tissue, each weighing approximately 0.1 g, and 1 mL of Solvable™ tissue solubilizer (Packard Bioscience BV, Groningen, the Netherlands) was added to each piece. Mechanical homogenization was performed by sonification (Sonics & Materials Inc., Danbury, Connecticut) at intervals of 60 to 90 sec, until no large tissue fragments remained. An additional aliquot of 1 mL of Solvable™ was added to the sonicator tip to avoid tissue mass loss. The samples were then placed in a 50 °C shaking water bath for 2 to 3 hours. All samples were transparent by this time. Five known concentrations of AlPcS ₄ were taken from the stock solution (after dilution in 0.9% saline) and processed in the manner described above. The fluorescence signals from these known concentrations were used to calibrate the spectrometer. Four 100-μL aliquots of each of these samples were diluted into 2 mL of double-distilled water, and the fluorescence from these dilutions was measured. This dilution ensured that the optical density of all samples at the excitation wavelength was less than 0.1 so that correction for optical absorption within the sample was unnecessary.¹¹

3.1.

Calibration Measurements in Tissue-Simulating Phantoms

Measurements were performed on tissue-simulating phantoms to calibrate the F/R ratio and time-resolved single fiber instruments. The calibration line for the F/R probe was the best fit straight line for the five combinations of optical properties (in Section 2.4) over the range of fluorophore concentrations. The root-mean-square percent error in the calibration line was 16.5%. The calibration of the single fiber probe, using both surface and interstitial measurements, yielded root-mean-square errors of 11%. These measurements are not shown here because they were very similar to the data presented in the papers^{12 13} that describe each technique in detail. An additional calibration, with baseline absorption coefficients ranging from approximately 0.2 to 0.5 mm⁻¹ (μ_sx ^′=1.2 mm ⁻¹), was performed for use with the liver and kidney measurements, because the absorption coefficient for these two organs was anticipated to be well outside the range of μ_a used to develop calibration curves in previous studies.^{12 13}

Estimates of the fluorescence lifetime were made for each combination of optical properties and fluorophore concentration. These data are shown in Fig. 3, plotted as a function of the reduced scattering coefficient at the excitation wavelength. The horizontal line on the graph represents the mean fluorescence lifetime (± standard error in the mean), 5.10±0.02 ns, measured in clear solutions at four concentrations using the same stock of AlPcS ₄. The fluorescence lifetime measured at the surface of the phantoms was slightly dependent on the optical properties, with values (± standard error in the mean, six concentrations, and three measurements per concentration) ranging from 5.16±0.01 ns (μ_sx ^′=0.6 mm ⁻¹, μ_a=0.1 mm ⁻¹) to 5.33±0.01 ns (μ_sx ^′=2.4 mm ⁻¹, μ_a=0.001 mm ⁻¹). Interstitial measurements yielded results that were more dependent on the optical properties. In this configuration, the measured fluorescence lifetime ranged from 5.16±0.01 ns (μ_sx ^′=0.6 mm ⁻¹, μ_a=0.1 mm ⁻¹) to 5.53±0.01 ns (μ_sx ^′=2.4 mm ⁻¹, μ_a=0.001 mm ⁻¹). The difference in the measured (apparent) fluorescence lifetime between the interstitial and surface geometries for the μ_sx ^′=2.4 mm ⁻¹, μ_a=0.001 mm ⁻¹ phantom was possibly a result of long pathlength photons scattered from behind the plane of the flatcut end of the fiber. The surface and interstitial time-resolved curves for the μ_sx ^′=0.6 mm ⁻¹, μ_a=0.1 mm ⁻¹ (surface) and μ_sx ^′=2.4 mm ⁻¹, μ_a=0.001 mm ⁻¹ (interstitial) phantoms are shown in Fig. 4. The instrumental autofluorescence has been subtracted from each of the curves, and the data normalized to the maximum value for comparison. Even in this extreme comparison the influence of the geometry (interstitial versus surface) and optical properties on the lifetime was only 0.3 ns.

Figure 3

AlPcS ₄ fluorescence lifetime measured in aqueous tissue-simulating phantoms. Squares represent interstitial measurements, circles measurements at the surface. Open and closed symbols represent μ_a=0.001 mm ⁻¹ and μ_a=0.1 mm ⁻¹ respectively. The solid line is the fluorescence lifetime measured in cuvettes of 0.9% saline.

Figure 4

Time-resolved fluorescence curves measured in aqueous tissue-simulating phantoms. The thin line represents an interstitial measurement for μ_sx ^′=2.4 mm ⁻¹ and μ_a=0.001 mm ⁻¹. The thick line represents a surface measurement for μ_sx ^′=0.6 mm ⁻¹ and μ_a=0.1 mm ⁻¹. Both phantoms contain 2 μg/mL AlPcS ₄.

3.2.

Specific Uptake of AlPcS ₄ in Various Tissues

The specific uptake was determined for each tissue type by dividing the results of the chemical extraction by the injected dose. (The injected doses for the first four rabbits were unavailable). The specific uptake in the various tissues measured in this study were consistent with other reported values for NZW rabbits^{7 12} and are shown in Fig. 5. The specific uptake of AlPcS ₄ in skin, muscle, and liver reported in Ref. 12 were from a single rabbit and so should be interpreted with caution given the inter-animal variation evident in Fig. 5. The specific uptake in skin was 0.53±0.05 as determined by the slope of the best-fit line [Fig. 5(a)]. This is comparable to the value of 0.41±0.02 measured in Ref. 7. The specific uptake from the rabbit measured in Ref. 12 was approximately 0.7±0.1, as determined by measuring at five different skin sites. Studies in rodents have produced a range of specific uptake for AlPcS ₄ ranging from 0.09 in mouse skin¹⁷ to 1.02 in rat skin,¹⁸ both determined at 24 hours post-administration. The uptake of AlPcS ₄ for muscle (both leg and dorsal muscle) and fascia are shown in Figs. 5(b) and 5(c), respectively. Values of AlPcS ₄ uptake in fascia were not available from the literature, but the measured specific uptake for these measurements was 0.39±0.05. The specific uptake in muscle was 0.062±0.006, which is consistent with a reported value of 0.052±0.008.⁷ However, the specific uptake reported in Ref. 12 was approximately 0.15±0.05. Measurements in rodents by other researchers produced a range of specific uptake for muscle: two studies determined that the specific uptake in rat muscle was 0.14 (Ref. 19) and 0.46 (Ref. 18), while the specific uptake in mouse muscle was 0.02.¹⁷ Liver [Fig. 5(d)] retained considerably more of the fluorophore, with a specific uptake of 2.2±0.1. This compared well to the single rabbit of Ref. 12, where the specific uptake in liver was approximately 2.00±0.05, but was about twice the value measured in Ref. 7, 0.89±0.08. Our measured value fell between those reported for other rodents, which ranged from approximately 1.2 in rats¹⁹ to 3.5 in mice.¹⁷ In kidney [Fig. 5(e)], we measured a specific uptake of 3.0±0.2. This value is considerably higher than the measured values in two mouse models, where a specific uptake of approximately 0.2 was measured in BALB/c mice,¹⁷ and a specific uptake of 0.7 was measured in a later study by the same group in the same model system.²⁰ It should be noted that a much lower specific uptake, 0.78±0.07, was measured at the core of the kidney [Fig. 5(f)].

Figure 5

The specific uptake of AlPcS ₄ measured in (a) skin, (b), muscle, (c) fascia, (d) liver, (e) kidney cortex, and (f) kidney medulla. The tissue sites corresponding to the different symbols are shown in each panel. Horizontal error bars were based on the uncertainty in the mass of the rabbits and the quantity of the fluorophore injected. The vertical error bars are the standard error in the mean for 12 total measurements of the three tissue samples taken from the measurement site.

3.3.

Quantitative Measurements In Vivo

The fluorescence lifetimes determined by the time-resolved instrument are shown in Fig. 6. The standard error in the mean for the individual data points (determined from at least three repeated measurements) ranged from 0.05 to 0.2 ns (error bars not shown). In skin, muscle, and fascia, the fluorescence decay was fitted using a single exponential decay. In liver and kidney, a double exponential decay was required to fit the data. The mean values (± standard error in the mean) of the fluorescence lifetimes shown in Fig. 6 are presented below.

Figure 6

The scatter plot shows the average fluorescence lifetime (three measurements) from the various tissues in 11 rabbits. The standard error in the means ranged from 0.05 to 0.2 ns, error bars not shown.

The tissue concentrations measured by the single fiber probe and the F/R ratio probe were compared with concentrations determined by chemical extraction. The concentration estimates for the skin, muscle and fascia are shown in Fig. 7. Horizontal error bars were derived from extractions from three samples cut from the tissue harvested from the measurement site. Measurements made on skin on the back and on the leg are shown in Fig. 7(a). Measurements with the single optical fiber and the F/R probe yielded similar results except at the highest concentrations. The average fluorescence lifetime measured in skin was 5.00±0.06 ns. Measurements with the two probes were in good agreement with the chemical extractions at low concentrations (<0.6 μg/g) but underestimated the tissue concentration at higher injected doses of AlPcS ₄. Measurements in muscle are shown in Fig. 7(b). The average lifetime measured in muscle (including both the surface and interstitial geometries) was 5.09±0.08 ns. The measured concentration was scattered around the line of equality, with the majority of the measurements showing very low concentrations (<0.1 μg/g). Figure 7(c) shows the measured concentration of AlPcS ₄ in fascia. Both single fiber and F/R probes tended to underestimate the concentration, but were in good agreement with one another. The average fluorescence lifetime measured from the fascia sites was 5.1±0.1 ns.

Figure 7

AlPcS ₄ concentration measured in (a) skin, (b) muscle, and (c) fascia. Closed circles represent surface measurements made with the single optical fiber, and open circles using the F/R probe. Inverted triangles represent interstitial measurements made using the single fiber, and the solid lines are the lines of equality. Vertical error bars are the standard error in the mean for at least three measurements repeated at the same location after removal and replacement of the probes. The horizontal error bars are the same as the vertical error bars in Fig. 5.

Measurements made on liver immediately post mortem using a separate high-absorption calibration are shown in Fig. 8. Figure 8(a) shows the measured concentrations compared to the chemical extractions. Both the F/R ratio probe and the single fiber tended to underestimate the concentration, but the single fiber measurements were more accurate. The average long and short lifetime components measured in the liver were 5.0±0.1 ns and 1.62±0.06 ns respectively. The short lifetime component of the fluorescence contributed approximately 10% of the integrated signal. Figure 8(b) shows a comparison of the concentration measured using the single fiber with and without a correction performed to account for the quenching associated with the short lifetime component. The ratio of the slopes between the two linear regressions (corrected for lifetime compared to not corrected) is approximately 1.27. Similar results (also using the separate calibration) were observed in kidney. Figure 9(a) shows the concentration of AlPcS ₄ measured by the single fiber and F/R ratio probes. Both probes underestimated the concentration. The average fluorescence lifetimes measured in the cortex were 4.40±0.04 ns and 1.28±0.07 ns. Interstitial kidney (medulla) measurements had slightly different fluorescence lifetimes, with averages of 4.76±0.05 ns and 1.3±0.2 ns. Figure 9(b) compares the corrected and non-corrected concentration estimates. The corrected estimates are approximately 1.4 times higher than the uncorrected concentration estimates. No differences in shape between the fluorescence spectra measured on the skin/muscle (by the F/R probe) and the liver and kidney were observed.

Figure 8

AlPcS ₄ concentration measured in liver. (a) Estimates for surface measurements with the single fiber (closed circles), the F/R probe (open circles), and interstitial measurements (inverted triangles). The effect of accounting for differences in the fluorescence lifetime is demonstrated in (b). Closed circles are the surface data for a single fiber, shown in (a); open circles show the concentration estimates that would be obtained by a steady-state measurement. Linear regressions are shown for the corrected (long dashes) and uncorrected (dotted) data. The ratio of the corrected and uncorrected slopes is 1.27. The solid line in each panel is the line of equality. The horizontal error bars are the same as the vertical error bars in Fig. 5.

Figure 9

AlPcS ₄ concentration measured in kidney. (a) Estimates for cortex surface measurements with the single fiber (closed circles), the F/R probe (open circles), and interstitial medulla measurements (inverted triangles). Only five medulla samples were large enough for chemical extraction. The effect of accounting for differences in the fluorescence lifetime for cortex measurements is shown in (b). Closed circles are the surface data for a single fiber, shown in (a); open circles show the concentration estimates that would be obtained by a steady-state measurement. Linear regressions are shown for the corrected (long dashes) and uncorrected (dotted) data. The ratio of the corrected and uncorrected slopes is 1.4. The solid line in each panel is the line of equality. The horizontal error bars are the same as the vertical error bars in Fig. 5.