Photodynamic Therapy with Verteporfin in the Radiation-induced

Document Sample
Photodynamic Therapy with Verteporfin in the Radiation-induced Powered By Docstoc
					[CANCER RESEARCH 63, 1025–1033, March 1, 2003]

Photodynamic Therapy with Verteporfin in the Radiation-induced Fibrosarcoma-1
Tumor Causes Enhanced Radiation Sensitivity1
Brian W. Pogue,2 Julia A. O’Hara, Eugene Demidenko, Carmen M. Wilmot, Isak A. Goodwin, Bin Chen,
Harold M. Swartz, and Tayyaba Hasan
Thayer School of Engineering, Dartmouth College, Hanover, New Hampshire 03755 [B. W. P., I. A. G., B. C.]; Department of Diagnostic Radiology, Dartmouth Medical School,
Hanover, New Hampshire 03755 [J. A. O., C. M. W., H. M. S.]; Biostatistics, Norris Cotton Cancer Center, Dartmouth-Hitchcock Medical Center, Lebanon, New Hampshire
03756 [E. D.]; and Wellman Laboratories of Photomedicine, Massachusetts General Hospital, Department of Dermatology, Harvard Medical School, Boston, Massachusetts
02114 [B. W. P., T. H.]

ABSTRACT                                                                                    ultimately causing necrosis or rapid apoptosis in the parenchyma. The
                                                                                            mechanism for cellular killing is thought to be different from that of
   Photodynamic therapy (PDT) with verteporfin (lipid form of benzopor-
                                                                                            radiation treatment, where unrepaired DNA strand breaks provide the
phyrin derivative, benzoporphyrin derivative monoacid ring A) was used
                                                                                            destructive effect. Several studies have investigated the combination
to treat radiation-induced fibrosarcoma tumors before X-ray treatment.
When verteporfin was injected 3 h before light irradiation, the tumor                       effect of PDT and -radiation, and many have concluded that the
partial pressure of oxygen (pO2) rose from a pretreatment value of 2.8 1                    combined treatment provides an additive, but not synergistic, effect
to 15.2    6.9 mm Hg immediately after light application was complete                       (5, 6). In this study, we examine a non-vascular-targeting type of PDT
(P    0.048). When the optical irradiation was given 15 min after verte-                    treatment with verteporfin photosensitizer (7), which induces in-
porfin injection, the tumor pO2 decreased slightly after treatment [i.e.,                   creased oxygen tension in the tumor (8) and, when combined with
6.8 1.6 mm Hg (pretreatment) versus 4.1 0.3 mm Hg (posttreatment)],                         radiation treatment, induces a greater than additive effect in tumor
whereas control tumor pO2 did not change significantly. In vitro study of                   killing.
the cellular oxygen consumption rate before and after PDT treatment                            Previous studies have examined specific interactions between PDT
indicated that the consumption rate decreased linearly with delivered
                                                                                            and radiation therapy to determine what the effect of combined
optical dose and quantitatively matched the loss of cell viability as meas-
ured by a mitochondrial tetrazolium assay. Doppler measurements show
                                                                                            therapy would be. Several early studies have examined whether the
that red cell flux is still patent immediately after treatment, indicating that             photosensitizer molecule itself acts as a radiosensitizer, and whereas
oxygen should still be delivered to the tumor. Computational simulations                    some have shown effects with hematoporphyrin derivative and ion-
of the oxygen supply from the vessels and the consumption from mito-                        izing radiation (9), the majority of studies have concluded that there
chondrial activity confirmed that if oxygen consumption is decreased in                     was no radiation sensitivity enhancement due to the photosensitizer
the presence of unhindered blood flow, the tumor oxygenation should rise,                   (10 –12). In comparison, many more studies have examined the utility
and the hypoxic fraction of the tumor should decrease. Combination                          of combined PDT treatment with radiation treatment, but most have
treatments with PDT delivered (100 J/cm2 optical dose, with 1 mg/kg                         been carried out in monolayer cell culture (5, 13–15), largely with the
benzoporphyrin derivative monoacid ring A injected 3 h before treat-                        conclusion that synergistic effects are not observed. However, studies
ment) after radiation treatment (10 Gy from 300 keV source) were com-
                                                                                            with both aluminum phthalocyanine and aminolevulinic acid have
pared with PDT delivered simultaneously with radiation. Tumor regrowth
assay showed that the delays to reach double the tumor volume for PDT                       indicated that timing may be a key factor in determining whether
alone and radiation alone were 2.7 1.6 and 3.2 1.7 days, respectively.                      synergistic effects are observed (13, 16) because synergistic effects
When radiation was given before PDT, the delay was 5.4 1.4 days, and                        are observed at longer temporal separations between the radiation and
when PDT was given at the same time as radiation, the delay was 8.1 1.5                     PDT treatments. In addition, any tissue response characteristics that
days. This observation indicates that the combined effect in the latter case                occur would not be present with in vitro assays. In vivo studies have
was greater than additive (P 0.049).                                                        demonstrated mixed effects: Winter et al. (17) demonstrated no syn-
                                                                                            ergistic effect of PDT and radiation in tumors; and Benstead and
INTRODUCTION                                                                                Moore (18) indicated that combined treatment of meta-tetra(hydroxy-
                                                                                            phenyl) chlorin (mTHPC-PDT) with X-rays increased the normal
   PDT3 is a treatment modality for dysplastic tissues, which uses a
                                                                                            tissue damage but that the effect of time sequence of the two treat-
light-activated drug to kill targeted regions of tissue (1– 4). A photo-
                                                                                            ments upon outcome was not significant. To date, none of these
sensitizer with delivery vehicle is applied to the tissue topically or
                                                                                            studies have examined specifically whether PDT could be delivered in
systemically, and moderate intensity light is used to selectively excite
                                                                                            a manner that changes the tissue metabolism such that, when used in
these molecules to their first excited triplet state, which is then
                                                                                            conjunction with radiation therapy, it provides a superadditive or
efficiently quenched by molecular oxygen to produce singlet state
                                                                                            synergistic effect.
oxygen. The cellular response to high doses of singlet oxygen is
                                                                                               Our studies have shown that mitochondria function can be acutely
complex but generally causes phospholipid peroxidation leading to
                                                                                            impaired when verteporfin treatment is applied directly to the cells
cellular membrane damage and vessel occlusion-mediated ischemia,
                                                                                            and that this effect is observed in vivo as well, as measured by a loss
                                                                                            of NADH fluorescence (19). This effect is likely a direct result of the
    Received 7/10/02; accepted 1/6/03.
    The costs of publication of this article were defrayed in part by the payment of page   fact that BPD-MA (the photoactive molecule in verteporfin) predom-
charges. This article must therefore be hereby marked advertisement in accordance with      inantly localizes on the mitochondria membranes, and therefore PDT
18 U.S.C. Section 1734 solely to indicate this fact.                                        action would likely cause widespread mitochondria damage (20, 21).
      Supported by National Cancer Institute Grants RO1CA78734 and PO1CA84203 and
by the Electron Paramagnetic Resonance Center for the Study of Viable Systems at            If PDT can be used to halt cellular metabolism shortly after treatment,
Dartmouth Medical School (supported by the National Center for Research Resources,          then there should be a net beneficial effect by increasing the oxygen
NIH Grant P41 RR11602).
      To whom requests for reprints should be addressed, at Thayer School of Engineering,
                                                                                            available to redistribute to previously hypoxic areas (8). The hypoth-
8000 Cummings Hall, Dartmouth College, Hanover NH 03755. E-mail: pogue@                     esis of this work was that if a PDT treatment could be applied that                                                                              preserved blood flow immediately after treatment yet reduced the
      The abbreviations used are: PDT, photodynamic therapy; RIF-1, radiation-induced
fibrosarcoma; BPD-MA, benzoporphyrin derivative monoacid ring A; EPR, electron              oxygen consumption by the parenchyma cells, then the tumor tissue
paramagnetic resonance; pO2, partial pressure of oxygen; LiPc, lithium phthalocyanine.      oxygenation would increase. This reoxygenation phenomenon would
                                                       NONVASCULAR PDT ENHANCES RADIATION SENSITIVITY

be most dominantly observed where preexisting hypoxia is present                     reduction in cellular respiration. The continuation of blood flow
due to the larger diffusion distances from capillaries (22). In our                  coupled with acute reduction in cellular respiration could increase
studies, we have chosen to follow this effect in the RIF-1 tumor model               or redistribute oxygen during PDT. This effect, combined with
(23), which is known to have large regions of hypoxic cells (22, 24,                 reduced killing in regions that were hypoxic before PDT, could
25) and has also been well studied in our previous studies (22, 26 –29)              lead to reoxygenation of chronically hypoxic areas, as has been
as well as in PDT studies of several other groups (25, 30 –35).                      observed in our earlier study with both verteporfin and amino
   pO2 changes that occur in vivo during PDT are complex and have                    levulinic acid-protoporphyrin IX treatment in the RIF-1 tumor (8).
been shown to vary with many factors (36). The necessity for oxygen                     In this study, changes in tumor pO2 were quantified both during and
to be present for tissue killing by PDT has been shown both in vitro                 after treatment with non-vascular-targeting verteporfin-PDT. The cel-
and in vivo (31), and the consumption of oxygen in this process can                  lular oxygen consumption rate was determined for RIF-1 cells in vitro.
be readily observed as an acute decrease of tissue pO2 (37–39). These                Computer simulations are used to interpret the observations and
observations led to a suggestion that high optical dose rates could lead             examine what the change in hypoxic fraction is expected to be. This
to less tumor killing due to the transient depletion of oxygen (39). This            treatment is then combined with X-ray irradiation to look for benefi-
concept has been supported by the observation that the effectiveness                 cial combinatorial effects of the dual therapy. Preliminary evidence
of treatment decreases in some tumor models when high optical dose                   suggests that non-vascular-targeting PDT could be a good mechanism
rates are used for treatment (35, 40 – 42). In addition to this transient            to decrease the hypoxic fraction of tumors, thereby increasing the
decrease in tumor oxygenation, it has been observed that rapid and                   efficacy of radiation therapy.
permanent reduction in blood flow and tumor oxygenation (43– 45)
can occur due to massive vascular occlusion during or soon after PDT                 MATERIALS AND METHODS
when a large fraction of the exogenous photosensitizer is present in
the vasculature (46). Whereas this vascular-targeting approach to                         Cells and Tumor Model. RIF-1 cells were used for in vitro and in vivo
delivering PDT has been a dominant line of study, it is also interesting               studies. RIF-1 cells were originally supplied by James B. Mitchell (National
to consider the longer time regime, where the photosensitizer has                      Cancer Institute), and they were grown in vitro in RPMI 1640 supplemented
largely leaked out of the vasculature or been cleared from the blood.                 with 10% fetal bovine serum, glutamine, and antibiotics and grown with no
                                                                                      more than four passes from the original stock. In vitro studies were carried out
Iinuma et al. (46) have shown that verteporfin is initially confined to
                                                                                      with the cells in 96-well plates, given 3 days between plating and use for
the vasculature at 5 min postinjection but that at 1 h postinjection, it
                                                                                      treatment. For in vivo studies, cells were grown in large culture flasks,
appears to be predominantly perivascular in the orthotopic rat prostate               removed, resuspended in medium without fetal bovine serum, and injected at
NBT-II tumor. Fingar et al. (45) showed that the effect of vascular                   4 105 cells/mouse in 50 l volume. Intradermal injection on the upper right
stasis appears acutely in chondrosarcoma and muscle when light is                      leg was used in female C3H/HeJ (5– 6 weeks old) mice (The Jackson Labo-
delivered 15 min after injection with verteporfin-based therapy but                    ratory, Bar Harbor, ME), approximately 11–14 days before the anticipated
that it was inhibited or delayed by several hours when light was given                 treatment time.
3 h after injection. Furthermore, Major et al. (47) have shown that                       Photosensitizer. The photosensitizer, verteporfin, was obtained from QLT
blood vessels can dilate after verteporfin treatment, potentially in-                  Phototherapeutics Inc. (Vancouver, British Columbia, Canada) for this exper-
creasing the flow after verteporfin-based PDT. Our own studies have                    imental study. Verteporfin is the active ingredient in Verteporfin For Injection
                                                                                       (VFI), the lipid formulated drug that is marketed as Visudyne for the treatment
indicated that even when verteporfin-based PDT is given in a vascu-
                                                                                       of ocular disease. The photoactive molecule present in this preparation is
lar-targeting approach, the response is heterogeneous at the micro-                    BPD-MA (57). For in vivo use, the drug as supplied was reconstituted in PBS
scopic scale, indicating that individual vessels vary in their response                at 0.2 mg BPD-MA/ml and then injected into the mice through the lateral tail
to the treatment. Our studies in the RIF-1 tumor model with verte-                     vein to give a dose of 1 mg BPD-MA/kg body weight. For in vitro use,
porfin treatment have shown that blood flow is still patent immedi-                    BPD-MA was reconstituted in PBS at 1 mg BPD-MA/ml solution in PBS and
ately after treatment and decreases steadily during the hours following                diluted in medium to expose cells to 1 g/ml BPD-MA. For both in vivo and
therapy but that if an incubation time of 3 h is used, then the decrease               in vitro work, an incubation period of 3 h was used in these experiments,
in blood flow is only to 55% of the pretreatment value.4 Thus, current                 except where otherwise specified. The drug was kept in the dark and frozen
evidence would indicate that when verteporfin-based PDT with opti-                     between experiments.
                                                                                          Light Delivery. A diode laser system up to 200 mW average power was
cal irradiation is given several hours after injection, the blood vessel
                                                                                       used throughout these studies (Applied Optronics, South Plainfield, CT), with
occlusion does not predominantly occur immediately during treat-                       a wavelength of 690 nm. The beam was delivered to the animals through a 140
ment, preserving blood flow during and after therapy.                                    m fiberoptic and expanded onto the tumor or cell plate in a circular top-hat
   In contrast to blood flow, the metabolic consumption rate of the                   beam, using a fiberoptic collimator (Thor Labs, North Newton, NJ). The beam
tumor parenchyma cells has not been adequately examined in PDT.                       diameter for in vivo work was 1.1 cm, using an irradiance of 133 mW/cm2, and
Whereas this parameter is very challenging to measure directly in                      for in vitro work, it was a 2-cm-diameter beam for an irradiance of 63.7
vivo (48 –51), it could be significantly altered by PDT for photo-                     mW/cm2. In vivo light treatment was given transcutaneously, with the animals
sensitizers that primarily target the mitochondria. The rate of                        cleanly shaven before treatment.
oxygen consumption would be inhibited and the mechanism would                             In Vitro PDT Treatment. Cells were plated in black plastic 96-well plates
                                                                                      with a transparent bottom (Fisher Scientific, Springfield, NJ) at a density of
depend upon the light fluence, irradiance, and localization of the
                                                                                      5000 cells/200 l medium/well. The cells were used for PDT treatment after
photosensitizer. It has been observed that one of the earliest events                  3 days of growth in 95% oxygen with 5% CO2, at 37°C in a humidified
in PDT damage in vitro is the loss of mitochondrial membrane                           incubator. On the day of treatment, the medium was replaced with 100 l/well
integrity, which would lead to a loss of cellular respiration and,                    medium with 1 g/ml BPD-MA in the verteporfin preparation. After 3 h in the
ultimately, apoptosis (52–56). In necrotic cell death, loss of res-                   incubator, the BPD-MA was decanted off, cells were rinsed once with HBSS,
piration would be instantaneous, but it is also likely that sublethal                 and then 100 l of medium were added to each well. Cells were illuminated
damage to the mitochondria induces a transient or permanent                            in groups of four wells, with blank wells between the treated groups to ensure
                                                                                       that each group of wells received the correct dose. For each 96-well plate,
   4                                                                                   squares of four were treated with increasing total light dose with 8 groups/
     B. Chen, B. W. Pogue, I. A. Goodwin, C. M. Wilmot, J. O’Hara, and P. J. Hoopes.
Blood flow dynamics following PDT with verteporfin in the RIF-1 tumor, submitted for   plate, including the control with no light. Cells were then assayed for cell
publication.                                                                           viability or cellular oxygen consumption.
                                                       NONVASCULAR PDT ENHANCES RADIATION SENSITIVITY

    Cell Viability Assay (MTS). Cell viability was assayed 24 h after treat- linewidth. The LiPc was precalibrated to determine the ambient pO2 from the
ment, using the MTS assay kit (CellTiter 96AQ; Promega, Madison, WI), linewidth, such that the data could be used to calculate effective tissue pO2
which is used to measure the reduction of a tetrazolium compound by the during the experiment.
cellular mitochondria, producing an optically active soluble formazan. After            Laser Doppler Flow Measurement. Laser Doppler measurements of RBC
PDT treatment, the medium from the 96-well plate was removed, and the MTS flux were taken on a section of animals given PDT treatment, using an
tetrazolium reagent was added. After a 1-h incubation at 37°C, the absorbance implantable fiberoptic probes (Moor Instruments, Wilmington, DE). The meas-
of each well was measured in an automated plate reader (Thermo Max; urement of red cell flux can be used to monitor microcirculatory function and
Molecular Devices, Menlo Park, CA) at the 490 nm absorption of the MTS provides a proportionate measurement of blood flow, assuming that hematocrit
soluble formazan product. For data analysis, the absorbance values from levels are constant throughout a given measurement. Two fiber probes had
groups of wells that received the same light dose were averaged, and the data external diameters of 200 m and were implanted in the tumor tissue through
were normalized by the average value from the control (no light) wells.             holes made with a 26-gauge needle immediately before insertion. Both probes
    Cellular pO2 Measurement. Immediately after PDT, cells were trypso- were implanted within the light field, and automated data collection was done
nized and suspended in medium, and cell viability was determined by trypan          with the supplied software on a computer. Blood flow was sampled at 40 Hz,
blue exclusion assay before use in the oxygen consumption measurement. and data points were averaged together over 5-min and 30-min periods after
Cells were resuspended at 2        107 viable cells/ml in complete RPMI 1640. the experiment. All flow values during the treatment time were standardized to
Just before use, the cell suspension was diluted to 107 cells/ml with a solution the average value during the first 2 min after probe implantation. Animals were
of 10% dextran in complete medium and used for oxygen consumption assay. divided into three control groups: (a) no light, no drug; (b) drug, no light; and
    The rate of oxygen consumption was measured from the cells using EPR. (c) light, no drug. Two treatment groups were used including a 15-min as well
Cells suspensions with dextran were extracted in 0.100 ml volumes and mixed as one treatment group using a 3-h interval between verteporfin injection and
with a neutral nitroxide, 15N PDT (4-oxo-2,2,6,6-tetramethylpiperidine-d16- light irradiation. Control animals were further divided into three subgroups: (a)
   N-1-oxyl) at 0.2 mM (Cambridge Isotope Laboratories, Quebec, Canada) in no verteporfin, no light; (b) verteporfin, no light; and (c) light, no verteporfin.
a 4- l aliquot. This mixture was then drawn into glass capillary tubes, which Verteporfin was injected i.v. at 1 mg/kg BPD-MA concentration 3 h before
were then sealed with Critoseal (Sherwood Medical, St. Louis, MO). The animal treatment. In the treatment group, blood flow was monitored for 2 min,
sealed tubes were placed into quartz ESR tubes, and samples were maintained followed by 12 min of light treatment, followed by 1 h of blood flow
at 37°C by a heated flow of gas through the resonator. All spectra were measurement. Blood flow during laser treatment was meaningless because
recorded on a Bruker EMX EPR spectrometer operating at 9 GHz. Because the during this time, the irradiation light saturated the Doppler probe. All values
resulting linewidth was proportional to oxygen concentration, oxygen con- were normalized to their pretreatment value before averaging together the
sumption rates were obtained by measuring the concentration repeatedly over numbers from each animal in each treatment group. The animals in the PDT
10 –20 min, and the slope of the resulting data was determined by linear            treatment group (as opposed to controls) were not sacrificed after the 30-min
regression. Three repeated trials were completed for each sample of cells, and time point but were later reanesthetized, and their tumors were reimplanted
the slope values were averaged.                                                     with blood flow probes to measure flow at 6 h posttreatment.
    Tumor pO2 during in Vivo PDT. Animals were anesthetized with inha-                  Combined PDT and Radiation Therapy Studies. Experimental studies
lation of isoflurane at 1.5% mixed with 26% oxygen in a continuous flow,            examining the combination effect of PDT and radiation therapy were com-
delivered by a nose cone to loosely cover the head. The animal was maintained pleted in this tumor model. The injection of BPD-MA or saline (for control)
at 37°C on a heated water pad, and warm air flowed over the animal. The             was given 3 h before treatment in all cases. Light at 690 nm wavelength was
animals were used for continuous monitoring of tumor pO2 before, during, and delivered to the PDT-treated animals while the animal was positioned in the
after the treatment using EPR oximetry as described below. In the first irradiator cone. An optical irradiance of 133 mW/cm2 was used over 12 min,
treatment group, tumors were treated with light 3 h after the verteporfin for a total optical dose of 100 J/cm2. Radiation was a single dose of 10 Gy (300
injection. In the second group, the photosensitizer was injected just 15 min keV, 10 mA, half value layer (HVL)                2.33 Gy/min). The treatment groups
before optical irradiation, and these animals were followed for tumor pO2 were as follows: (a) sham-irradiated controls; (b) radiation treatment alone; (c)
before, during, and after treatment. This type of short-term treatment is thought PDT alone; (d) radiation treatment followed by PDT; and (e) PDT with
to elicit a strong vascular occlusion response and thus provides a contrast to the radiation treatment given simultaneously. In the fourth treatment group, X-ray
effect from a cellular targeting treatment, where a 3-h delay is given between radiation was given for the full 3-min treatment time after the photosensitizer
injection and irradiation. The third group was control animals that were treated was injected 3 h after photosensitive injection, and then optical irradiation was
with light alone to simulate the effects of light absorption in the tissue without   given for the 12-min treatment. In the fifth treatment group, the optical
photosensitizer.                                                                     irradiation was given 3 h after the drug injection, and during the last 3 min of
    Tumor pO2 was monitored with an oxygen-sensitive EPR probe material, the optical irradiation, X-ray irradiation was also given to the tumor at the same
synthetic LiPc, which can be implanted into the tissue and provide stable            time. This latter timing would result in the tumor being exposed to X-rays
measurements of tissue pO2 over several weeks (27, 58 – 61). Initial experi- during the maximal reoxygenation time of the tumor tissue.
ments demonstrated that pO2 could be monitored with EPR during the PDT                  Irradiation of the 10-Gy dose using approximately 300 keV was delivered in
treatment without any interaction between the probe and the photosensitizer or 3 min from a Pantak Therapax 300 Orthovoltage irradiator, using a 2-cm
light (8, 29). Small particles (approximately between 50 and 200 m) of LiPc          collimation cone. This cone was placed in intimate contact with the mouse leg,
were implanted in the animal tumors at least 24 h before PDT treatment to            surrounding the tumor. The end of the collimation cone was transparent plastic,
allow for resolution of any acute effects due to the injection. A 23-gauge allowing radiation and optical treatment to occur at the same time, as required
syringe needle was used to implant the autoclaved, dry material at a depth of in the last group, where PDT and radiation were simultaneously applied. All
1–3 mm within the tumor. Only one implantation of oxygen-sensitive material animals were treated with the cone in place, even those treated with sham
was necessary, and pO2 could be monitored as often as desired during the irradiation and PDT alone, to ensure that equal doses of light, including
experiments.                                                                         changes in the beam from reflection and refraction through the irradiator cone,
    The animals were placed in an L-band (1.2 GHz) EPR spectrometer with a were given to all groups.
microwave bridge custom-built for measurements of tissue pO2 in small                   Tumor Regrowth Assay. Mice were assigned to the different treatment
animals. The external loop resonator was positioned over the tumor to record groups in the combined radiation and PDT study in a manner that produced
the EPR signal of the LiPc within the tissue. Typical settings for the spec- matching initial tumor volumes at the time of treatment to minimize any
trometer were incident microwave power of 50 mW, magnetic field of 400 G, systematic errors associated with groups having different initial volumes at the
scan range of 0.5 G, modulation frequency of 27 kHz, modulation amplitude time of treatment. The resulting treatment effect was assayed by calculation of
of 15 mG, and scan time of 30 s. The settings were not significantly changed the tumor volume, as determined by measurement of the three major axis
between animals in this study nor during an acquisition of an individual dimensions (volume                   length    width     height/2). The time for a tumor to
animal. After accumulation of the linewidths as a function of time before, reach double its volume on the day of treatment was calculated for each animal
during, and after treatment, the data were fit with a custom-written software separately by estimating the data by a linear fit to the logarithmic growth curve
program to match the Voigt line shape of the EPR spectrum and extract the versus time and estimating when it had reached twice the tumor volume as on
                                                         NONVASCULAR PDT ENHANCES RADIATION SENSITIVITY

the day of treatment. The average values from these times to double in volume       treatment, and 1, 4, and 24 h posttreatment. The pO2 values for each
were calculated for each treatment group.                                           group of animals were averaged together, and the SDs were calcu-
   To analyze the tumor regrowth data and obtain estimates of the mean              lated. These values are shown in Fig. 1.
doubling time and its SE, individual growth curve data were combined for each          Control animals were given the laser light but no photosensitizer
group and fitted using a mixed effects model (28, 62). It was assumed that after    injection. The control animals all had pO2 values initially at 3.6 1.1
day 3, the tumor regrows exponentially, which corresponds to a linear function
                                                                                    mm Hg, indicating that the tumors have a high hypoxic fraction, as
when plotted on a semi-log graph. To address mice heterogeneity, a mixed
effects modeling technique was applied, which has been previously applied in
                                                                                    has been shown in previous studies (22, 61). Throughout the laser
O’Hara et al. (28) and described by Demidenko (62). The time for a tumor to         treatment and afterward, the control animals maintained hypoxic pO2
reach double its volume on the day of treatment was calculated for each group       values averaging 3.7 1.2 mm Hg. In the treatment group, the initial
along with the respective SE using the delta-method (63). These SEs give rise       pO2 was 2.8 1.0 mm Hg. Initial changes in pO2 had a high variance
to the group comparison using Z-test to test for significant differences between    between animals, with some increasing in pO2 immediately, some
the mean values of the difference treatment groups.                                 staying constant, and some fluctuating. By the end of the treatment
   Oxygen Distribution Simulation Studies in RIF-1 Tumors. A previously             period, all five of the treated tumors had risen significantly in their
reported simulation study was used to evaluate the effect of changes in             pO2 (8). The final value immediately after treatment was 15.2 6.9
metabolic consumption on the resulting tumor tissue pO2 (22). This study used       mm Hg, which is significantly different from the control value, as
a finite element solution to the steady-state diffusion equation to solve for the
                                                                                    calculated by a paired Student’s t test with P          0.048. At 1 h
oxygen concentration everywhere within an arbitrary volume of tissue. The
differential diffusion equation is given in steady state by
                                                                                    posttreatment, the pO2 was back to within control or pretreatment
                                                                                    values. A group was also included with the photosensitizer injected
                    D       CO2 r   kmet r, O2   SO2 r     0                 (1)    just 15 min before treatment, which allows observation of the “vas-
                                                                                    cular-targeting” type of therapy. In this case, the average tumor pO2
where CO2(r) is the oxygen concentration at position r, D is the diffusion
                                                                                    decreased slightly during treatment while staying in the hypoxic range
coefficient for oxygen in tissue (which is generally taken to be spatially
                                                                                    (i.e., 6.8 1.6 mm Hg before light treatment and 4.1 0.3 mm Hg
independent), kmet(r,O2) is the metabolic oxygen consumption rate, and SO2(r)
is the supply of oxygen by the capillaries at each point r.                         immediately after treatment).
   The capillary geometries for the simulation were derived from eight sepa-           In Vitro Cellular Oxygen Consumption and Viability. Measure-
rate H&E-stained sections of RIF-1 tumor tissue, digitized, and manually            ments of oxygen in suspensions of RIF-1 cells show that the oxygen
thresholded. The capillary oxygen supply rates were estimated based on fitting      consumption rate of untreated cells is higher than that of cells treated
to boundary information, which was given by pimonidazole staining of adja-          with BPD-MA-based PDT, as shown in Fig. 2. The average oxygen
cent sections of the tissue. This staining provides demarcation of regions of       consumption rate in the untreated cells was 1.84        0.14 nmol/min/
hypoxia, thereby allowing estimation of the neighboring capillary oxygen            million cells. When treated with light alone, there are small increases
supply rates, assuming that the oxygen diffusion coefficient is known               in this oxygen consumption rate on the order of 0.2 nmol/min/106
(D     2    10 5 cm2/s). Our earlier study indicated that a metabolic oxygen
                                                                                    cells, and when treated with the photosensitizer and light, there is a
consumption rate of kmet       10 M/s was appropriate to simulate an oxygen
                                                                                    monotonic decrease in oxygen consumption correlated to the deliv-
histogram distribution that was quantitatively similar to experimental meas-
urements with Eppendorf electrode measurements taken from a set of RIF-1            ered light dose. At the highest light dose used here (16 J/cm2), the
tumors (22).                                                                        oxygen consumption rate fell to 0.37 0.04 nmol/min/106 cells. The
   In the present study, the same eight sections of tumor tissue were used to       data for these measurements are plotted in Fig. 2, shown by open
simulate the oxygen distribution within the tumor tissue. The oxygen supply         symbols (squares for treated cells and circles for control), with the
rates of the capillaries were given by the previous fitting to the pimonidazole     absolute units of oxygen consumption plotted on the right-hand
data, and then the value of kmet was varied to appreciate the change anticipated    vertical axis. The control group received light doses without incuba-
in the tumor under normal conditions (i.e., kmet           10 M/s) and when         tion with photosensitizer.
metabolic consumption is halted (i.e., kmet 0 M/s). The calculated oxygen              The cell viability as measured by MTS assay is plotted on the same
distributions were then integrated into a histogram so that the data could be
presented in the format typically obtained with an Eppendorf electrode.
Whereas all calculations were completed in micromolar concentration units,
the final data are presented in units of pO2 (mm Hg), by multiplication by the
solubility of oxygen in water, assuming that the solubility of the tumor tissue
would be similar.


   In Vivo Tumor pO2. Measurements of tumor pO2 were taken in
control and treated tumor-bearing animals. Mean SD tumor volume
for all groups at the time of treatment was 203       23 mm3 (range,
110 –300 mm ). There was no significant difference between the
tumor volumes of the groups at the time of treatment. The EPR probe
was placed within the top 2–3 mm of tumor tissue to ensure that it was
reporting from a region that had a full effect of the PDT. This
measurement is localized and was repeated at several different times
before, during, and after the treatment in each animal. Typically, the
pO2 was recorded for 10 min before light irradiation, then recorded                     Fig. 1. Tumor pO2 measurements in RIF-1 tumors in mice treated with verteporfin-
                                                                                    based PDT and control animals with light alone but no photosensitizer (n 5). There were
every minute throughout the treatment, and recorded for 10 –15 min                  two treatment groups, each of which received 1 mg/kg BPD-MA in the verteporfin
after the treatment. The mice were then reanesthetized at 1, 4, and 24 h            preparation (n 5 each) either 15 min or 3 h before light treatment. In both cases, the light
posttreatment for pO2 measurement, requiring approximately 10 –15                   was delivered with a total dose of 144 J/cm2 at 690 nm wavelength. The error bars in the
                                                                                    figure represent the SD, and the values for the 3-h-treated group immediately after
min. The average values at each time point were calculated separately               treatment (labeled after Tx on the graph) are significantly different from control values
for each animal for pretreatment, during treatment, immediately after               (P 0.048).
                                                            NONVASCULAR PDT ENHANCES RADIATION SENSITIVITY

graph to illustrate the similarity between the dose responses, with
survival plotted on the left-hand vertical axis, normalized to 1.0 for
the untreated cell results. The closed symbols in Fig. 2 show these
data, and the line on the graph shows a best fit to the survival data.
The dose response is quantitatively identical in this study between cell
survival and oxygen consumption change. At the maximal light dose,
oxygen consumption is 21 4% of the control oxygen consumption
rate, and correspondingly, the MTS assay for the same light dose
indicated a decrease to 20 5% cell survival.
   Laser Doppler Flow Measurements. Doppler RBC flux measure-
ments were taken on six animals using the same treatment protocol
and are shown summarized in Fig. 3. The flux immediately after
treatment is equivalent to the flux in the control groups of no drug and
no light, as well as drug only. Interestingly, the light alone control
group shows a modest increase in blood cell flux immediately after
the light delivery. The treatment group then has a reduction in blood
cell flux over a longer time scale, with a reduction to 70% at 30 min
after PDT and to 55% at 6 h after PDT.
   Calculation of Changes in pO2 Based on Changes in Metabolic
Consumption Rate. Using the capillary distributions and values ob-
tained from RIF-1 tumor sections in our previous study (22), oxygen
histograms were simulated for the two conditions of kmet 10 M/s

                                                                                               Fig. 4. Calculated oxygen histograms from oxygen diffusion calculations where the
                                                                                            metabolic oxygen consumption term (kmet in Eq. 1) was varied. In a, kmet         10 M/s,
                                                                                            whereas in b, kmet was 0 M/s. In both cases, the simulations were carried out on the same
                                                                                            capillary distribution patterns with D 2 10 5 cm2/s. In a, the median pO2 is 2 mm
                                                                                            Hg, and in b, the median pO2 is 7 mm Hg.

                                                                                            and kmet 0 M/s. The values of pO2 were summed from the images
                                                                                            used, and the data are presented in a histogram format. These are
                                                                                            shown in Fig. 4, a and b, respectively. The median pO2 values for
                                                                                            these two distributions are 2 and 7 mm Hg, respectively, indicating
                                                                                            that when the metabolic oxygen consumption rate of the cells is
                                                                                            minimal, the pO2 of the tumor tissue could increase by a median of 5
                                                                                            mm Hg. These data predict that the hypoxic fraction (the fraction of
   Fig. 2. Cellular viability (closed symbols) and oxygen consumption rate (open symbols)
                                                                                            the tumor less than 10 mm Hg pO2) is initially 98%, whereas after the
as a function of the delivered light dose to RIF-1 cells incubated for 3 h with 1 g/ml      change in metabolic consumption, the hypoxic fraction is reduced to
BPD-MA. Cell viability was determined using the MTS assay. Oxygen consumption rate          65%. This simulation represents the extreme case of zero oxygen
was measured from cells in suspension using EPR oximetry. In both cases, the data points
represent the average of three successive trials, and the error bars are the SD.            consumption after treatment, which is not the case from the measure-
                                                                                            ments. Nonetheless, the increase in oxygenation is monotonically
                                                                                            correlated with the oxygen consumption rate of the cellular regions,
                                                                                            and so the two extremes simulated here present a simulated maximal
                                                                                            range for the effect observed.
                                                                                               PDT Combined with Radiation Treatment. Measurements of
                                                                                            tumor volumes after treatment with radiation, PDT, or both are shown
                                                                                            in Fig. 5. In this experiment, the five treatment groups were as
                                                                                            follows: (a) control with saline injection alone; (b) radiation therapy
                                                                                            alone with BPD-MA injection but no light; (c) PDT alone with sham
                                                                                            X-ray irradiation; (d) X-ray irradiation followed by PDT; and (e) PDT
                                                                                            and radiation treatment given together. The data in this graph show
                                                                                            that the tumor regrowth rate in all groups is slightly lower than that of
                                                                                            the control group but that these slopes are similar between all treated
                                                                                            groups. The slopes of the natural logarithm of tumor volume versus
    Fig. 3. Measured average normalized blood flow from Doppler flow probes implanted
                                                                                            days were 0.20, 0.11, 0.13, 0.13, and 0.11, respectively, for the five
in the tumor. Each bar represents the average of six animals, where the flow value was      treatment groups. PDT alone (group 3) and radiation alone (group 2)
normalized by the pretreatment flow measured over 2 min, and the error bars are the SE      induce the same approximate time for the tumor to reach twice the
between animals. Three control groups (no drug and no light, verteporfin alone, and light
alone) are shown as an average measurement over 30 min. The PDT-treated animals are         volume of the treatment size (i.e., doubling times were 8.3 and 8.9
shown both immediately after light treatment as well as 30 min and 6 h posttreatment.       days, from Table 1). In comparison, group 4 and group 5 indicate
                                                             NONVASCULAR PDT ENHANCES RADIATION SENSITIVITY

                                                                                             they essentially both measure mitochondria activity, but because the
                                                                                             MTS assay has been shown to be proportional to long-term cell
                                                                                             survival in other studies, it is likely that the loss of oxygen consump-
                                                                                             tion rate indicates acute damage to the mitochondria that can ulti-
                                                                                             mately cause cell death. This change in oxygen consumption is also
                                                                                             directly correlated to the optical dose delivered, so that the effect is
                                                                                             likely due to singlet oxygen-mediated cellular death. The fact that
                                                                                             these changes are observed immediately after the light treatment is an
                                                                                             indicator that the predominant effect causing cell death is necrosis or
                                                                                             some acute damage mechanism.
                                                                                                Tumor pO2 Increases Immediately after PDT. In vivo tumor
                                                                                             pO2 measurements indicated that the tissue oxygen increased acutely
                                                                                             in response to the treatment (Fig. 1). Our studies show that blood flow
                                                                                             is not significantly reduced immediately after this type of treatment in
                                                                                             the RIF tumor. Whereas this flow eventually degraded over time for
                                                                                             all treatment groups, this indicates that there is a time period after
    Fig. 5. Average tumor volume is plotted relative to the number of days after treatment
(days posttreatment, X axis) for the five treatment groups [control (n 6 mice), radiation
                                                                                             PDT in this treatment regime where oxygen is still being delivered to
only (n 5), PDT alone (n 6), radiation followed by PDT (n 7), and PDT together               the tissue. Because we observe that oxygen consumption is acutely
with radiation (n     7)]. The points represent the average values of each group, and the    reduced by the PDT treatment, we conclude that the tissue pO2 could
error bars are the SE. The tumor volumes on the day of treatment and for doubling the
treatment volume are denoted by horizontal dotted lines.                                     rise in response to therapy in which the blood vessels are not acutely
                                                                                             occluded. This observation also agrees with the results we have
                                                                                             reported for treatment of the RIF-1 tumor with aminolevulinic acid,
increasingly better treatment effects, with doubling times of 11.0 and                       where the rise in tumor oxygenation was also observed (8), and the
13.7 days, respectively.                                                                     decay in the oxygen level after treatment was observed to decay
                                                                                             over the time course of over 1 h, indicating that this time period
DISCUSSION                                                                                   was probably when blood was flowing. Further study of the micro-
                                                                                             regional kinetics after PDT would be useful to help characterize this
   In this study, verteporfin-based PDT has been explored as an                              phenomenon.
adjuvant to radiation therapy, using a 3-h interval between injection                           Simulations of oxygen distribution within the RIF-1 tumor tissue
and irradiation because this time point allows continuous blood flow                         were carried out, and the histogram calculations shown in Fig. 4
in tissue after treatment, as well as increased tumor pO2 (8). This work                     present representative values that indicate the relative effect that
began under the hypothesis that non-vascular targeting of PDT could                          changes in cellular oxygen consumption rate would have. In Fig. 4a,
be achieved using a drug interval of several hours between injection                         the histogram is very similar to that typically observed in the RIF-1
of the photosensitizer and the time of optical irradiation. Studies have                     tumor model, on average, showing a dominant hypoxic fraction and a
shown that the verteporfin photosensitizer leaks into the surrounding                        large fraction less than 5 mm Hg (22). By changing the oxygen
parenchyma on time scales that are longer than a few hours, whereas                          consumption rate in these simulations to zero, the oxygen distribution
intravascular concentrations have a metabolic lifetime of about 6 h                          in the tumor is simply given by the distribution of capillary pO2 values
(46). Thus, a treatment protocol that allows several hours of latency                        that supply the region. In this case, the median pO2 significantly
provides preferential targeting to the parenchyma while minimizing                           increases from 2 to 7 mm Hg. This model approach assumes that the
the effect on the vasculature. It is important to note that whereas this                     blood flow is not altered at all during therapy, in agreement with our
photosensitizer has been most used clinically for vascular targeting in                      measurements here. These simulations illustrate that changes in met-
age-related macular degeneration treatment (64), it is not limited to                        abolic consumption rate can have a measurable effect on the tumor
this type of therapy in principle. A longer drug-light interval is key to                    pO2 value, decreasing the hypoxic fraction from 98% initially to 65%
minimizing the vascular occlusion within the tumor4 and maximizing                           after loss of oxygen consumption. Alterations in blood flow would
the direct cell damage, which causes acute mitochondrial damage. Our                         change the supply rate of oxygen and further perturb the measured
blood flow studies with verteporfin treatment in the RIF-1 tumor show                        pO2, yet our studies did not demonstrate significant changes in flow.
results similar to the observations of Fingar et al. (45) in the chon-                       The limitation of this calculation is that it is for the specific tumors
drosarcoma tumor. When the 3-h drug-light interval is used, then the                         studied in our earlier paper (22) and that because this tumor is highly
flow decreases after treatment, but at a slower rate than when 15 min                        heterogeneous, it is likely that the effect will be higher in some
is used. When the 15-min drug-light interval is used, the flow de-                           regions and lower in other regions.
creases steadily, decaying to 50% of initial values at 1 h after treat-                         These observations have implications for PDT, suggesting that
ment and down to zero at 6 h after treatment.4 The potential applica-                        PDT-induced cellular damage may be used to increase oxygen in
tions for vascular-targeting versus parenchyma-targeting therapies are                       tumor tissue, including regions that were in a chronically hypoxic
quite different, and yet little study has been devoted to contrasting                        state. In our case, the pO2 of the RIF-1 tumor increased from an
these two potential dosimetry regimes (3, 21, 65– 67). This study
focuses on the effect of PDT in cellular respiration and pO2 changes
that occur in this parenchyma-targeting regime.                                                             Table 1 Tumor doubling time and regrowth delay
   Mitochondrial Function Is Inhibited after Verteporfin PDT. In                                                                Time for tumor to      Regrowth delay in
our experiments, cells that have been treated in vitro show a reduction                         Group       Treatment          double initial volume     doubling time

in oxygen consumption rate (shown in Fig. 2). These data indicate that                            1     Control                      5.6   0.4
                                                                                                  2     Radiation only               8.3   1.7               2.7   1.6
the change in oxygen consumption is directly proportional to the loss                             3     PDT only                     8.9   1.7               3.2   1.7
of cell mitochondrial function, as measured by the MTS assay. It is                               4     Radiation then PDT          11.0   1.5               5.4   1.4
not surprising that these two assays should be proportional because                               5     PDT and Radiation           13.7   1.6               8.1   1.5

                                                               NONVASCULAR PDT ENHANCES RADIATION SENSITIVITY

                              Table 2 Student’s t test comparisons                     a pretreatment, it should enhance the radiation effect. It is also
                 Difference test                                       P               possible that the blood flow shunting that could occur as some
                 Groups   2   vs.   3                                0.77              microvessels are occluded may also cause oxygen enhancement in
                 Groups   2   vs.   5                                0.12              areas of continued flow, and pathological analysis and Doppler blood
                 Groups   3   vs.   5                                0.0013a
                 Groups   4   vs.   5                                0.049a
                                                                                       flow measurements are ongoing to examine this as a potential alter-
  a                                                                                    native hypothesis. Independent of the mechanism, combining PDT in
      Significant difference (P         0.05).
                                                                                       a non-vascular-targeting regime (i.e., where there is a long latency
                                                                                       between drug injection and the time of optical treatment) may be used
average of 3.6 mm Hg (control) to 15.6 mm Hg immediately after                         effectively to enhance the efficacy of radiation therapy.
treatment. This would be sufficient to raise the average tumor oxy-                       Experiments of combined therapies can often have potentially
genation out of the region that is considered radiobiologically hypoxic                confounding causes, and it is important to not attribute the supra-
and allow it to be effectively targeted by conventional radiation                      additive effect observed here to a mechanism that cannot be directly
therapy. Whereas microscopic regions of hypoxia could still exist, the                 proven. We hypothesize that a major mechanism causing this effect is
overall average oxygenation is sufficiently high that the beneficial                   the shift in tumor oxygenation toward higher pO2 and possibly spa-
effect of the oxygen enhancement ratio should be apparent. Thus,                       tially to reduce the preexisting hypoxic fraction within the tumor. One
PDT-induced tumor reoxygenation could be used as a pretreatment to                     potentially confounding issue, which needs to be addressed, is the
radiation therapy, thereby enhancing the radiation killing effect. These               potential for hyperthermia induced by the high fluence rate of 133
effects are quite time dependent, and therefore measurements of the                    mW/cm2. Whereas some authors have indicated that this can induce
tissue pO2 could be a useful guide for timing subsequent therapy with                  hyperthermia when delivered at 633 nm, our previous measurements
ionizing radiation. Ultimately, the tumor oxygenation falls back to                    (16) have shown that the temperature rise with a higher fluence rate of
hypoxic levels, indicative of something else happening at longer                       200 mW/cm2 at 690 nm is 3– 4 degrees C maximally and thus does not
times, and this would correlate with the reduction in blood flow                       correspond to a hyperthermic rise, which requires 8 –10 degrees C
observed in our study at longer times after irradiation. Additional                    increase for a measurable effect on the tissue (51, 70). Nevertheless,
studies are needed to explore this effect and the potential for syner-                 it is possible that localized microscopic regions of the tumor experi-
gistic interaction with radiation therapy.                                             ence hyperthermic rises in temperature. Nonetheless, there is a clear
   Combined PDT and Radiation Therapy Induce a Greater than                            observation that combining PDT with radiation therapy simulta-
Additive Killing Effect. After the observation of increased tumor                      neously induces a greater than additive effect of tumor killing, and the
pO2 in response to PDT, it was concluded that there would likely be                    mechanism remains to be conclusively delineated.
a significant benefit of PDT and radiation therapy if used together in                    In summary, the cellular respiration of in vitro RIF-1 tumor cells
this model. The tumor regrowth delay data shown in Fig. 5 and                          can be significantly reduced in response to PDT treatment under the
analyzed in Tables 1 and 2 support this hypothesis. Whereas the tumor                  conditions examined here, and the dose response of this effect is
doubling times increased from 5.6 0.4 days in the control group to                     identical to the results of a cell survival (MTS) assay. These results in
8.3 1.7 and 8.9 1.7 days, respectively, for the radiation and PDT                      vitro support the conclusion that this loss in cell metabolism is directly
alone groups, the combined treatment times illustrated that the order                  proportional to the loss of cell viability. The decrease in metabolism
of the radiation versus PDT delivery matters significantly. A doubling                 likely leads to a greater diffusion of oxygen throughout the remaining
time of 11.0      1.5 days was observed for the delivery of radiation                  tissue in vivo, when the blood flow remains patent immediately after
first, followed by PDT, and an increase to 13.7            1.6 days was                treatment. Experiments confirm that an increase in tumor tissue pO2
observed when PDT was used as the pretreatment adjuvant for the                        occurs after treatment and that the tumor tissue actually can rise above
radiation. Using the tumor regrowth delay (last column of Table 1) as                  the radiobiologically hypoxic level. These observations, taken to-
the measure of damage, this indicates that group 4 treatment (when                     gether, indicate a new direction and potentially a new set of applica-
radiation is given before PDT) induces an additive effect of                           tions for PDT in modulating the oxygenation of tissue.
the radiation alone (i.e., group 2, delay     2.7 days) and PDT alone                     Combining this type of PDT treatment with radiation treatment
(group 3, delay         3.2 days) where given separately, because                      using a single 10-Gy dose of radiation demonstrated a greater than
2.7 3.2 5.9 1.9 days. This additive delay is in good agreement                         additive effect of tumor killing, as measured by tumor volume re-
with the observed effect of group 4 treatment at 5.4 1.4 days. In the                  growth delay assay for the tumor doubling time. The mechanism for
last group (group 5), when PDT and radiation were delivered together,                  this synergistic effect may be due to the enhanced oxygen available
the regrowth delay was 8.1        1.5 days, indicating an effect that is               during the treatment, effectively reducing the fraction of the tumor
more than additive (68, 69). Student’s t test comparisons were com-                    that is chronically hypoxic. Calculations of the oxygen change are in
pleted between groups 2 and 3 and between group 5 and all others to                    reasonable agreement with experimental observations. Additional ex-
determine whether there were significantly differences (P          0.05),              periments are required to help verify that this change in hypoxic
with the results shown in Table 2. The significant difference between                  fraction is the cause of the enhanced killing effect.
groups 4 and 5 indicates that the mechanism underlying the effect may
be different or that some feature of the treatment was more dominant
in the latter case.
   These results indicate that there is a significant effect of vertepor-                B. W. P. acknowledges thoughtful discussions and use of resources with
fin-based PDT when used as a pretreatment adjuvant to radiation                        Brian C. Wilson (University of Toronto) and P. Jack Hoopes (Dartmouth
therapy when given as a single 10-Gy dose to the RIF-1 tumor. The                      Medical School) in the course of this work. Verteporfin was supplied by QLT
tumor regrowth delay is increased by a factor that is more than                        Phototherapeutics Inc.
additive, indicating a potential benefit for PDT when used as an
adjuvant to radiation. This result is consistent with our central hy-                  REFERENCES
pothesis that PDT can be used to increase the oxygen available in the
                                                                                          1. Weishaupt, K., Gomer, C., and Dougherty, T. Identification of singlet oxygen as the
tumor by suppressing the oxygen consumption of the remaining                                 cytotoxic agent in photo-inactivation of a murine tumor. Cancer Res., 36: 2326 –2329,
tumor. Also, when this non-vascular-targeting form of PDT is used as                         1976.
                                                             NONVASCULAR PDT ENHANCES RADIATION SENSITIVITY

 2. Gibson, S. L., and Hilf, R. Interdependence of fluence, drug dose and oxygen on HPD      31. Henderson, B. W., and Fingar, V. H. Oxygen limitation of direct tumor cell kill during
    induced photosensitization of tumor mitochondria. Photochem. Photobiol., 42: 367–            photodynamic treatment of a murine tumor model. Photochem. Photobiol., 49:
    373, 1985.                                                                                   299 –304, 1989.
 3. Henderson, B. W., and Dougherty, T. J. How does photodynamic therapy work?               32. Fingar, V. H., Wieman, T. J., Park, Y. J., and Henderson, B. W. Implications of a
    Photochem. Photobiol., 55: 145–157, 1992.                                                    pre-existing tumor hypoxic fraction on photodynamic therapy. J. Surg. Res., 53:
 4. Dougherty, T. J., Gomer, C. J., Henderson, B. W., Jori, G., Kessel, D., Korbelik, M.,        524 –528, 1992.
    Moan, J., and Peng, Q. Photodynamic therapy. J. Natl. Cancer Inst. (Bethesda), 90:       33. Zaidi, S. I. A., Kenney, M. E., and Mukhtar, H. Photodynamic therapy of RIF-1
    889 –905, 1998.                                                                              tumors with topical application of silicon phthalocyanine: evidence for induction of
 5. Luksiene, Z., Kalvelyte, A., and Supino, R. On the combination of photodynamic               apoptosis during tumor ablation. Clin. Res., 41: A507 A507, 1993.
    therapy with ionizing radiation. J. Photochem. Photobiol. B Biol., 52: 35– 42, 1999.     34. Bremner, J. C., Bradley, J. K., Adams, G. E., Naylor, M. A., Sansom, J. M., and
 6. Allman, R., Cowburn, P., and Mason, M. Effect of photodynamic therapy in combi-              Stratford, I. J. Comparing the anti-tumor effect of several bioreductive drugs when
    nation with ionizing radiation on human squamous cell carcinoma cell lines of the            used in combination with photodynamic therapy (PDT). Int. J. Radiat. Oncol. Biol.
    head and neck. Br. J. Cancer, 83: 655– 661, 2000.                                            Phys., 29: 329 –332, 1994.
 7. Scott, L. J., and Goa, K. L. Verteporfin. Drugs Aging, 16: 139 –146; discussion,         35. van Geel, I. P. J., Oppelaar, H., Marijnissen, J. P. A., and Stewart, F. A. Influence of
    147–138, 2000.                                                                               fractionation and fluence rate in photodynamic therapy with photofrin or mTHPC.
 8. Pogue, B. W., O’Hara, J. A., Goodwin, I. A., Wilmot, C. J., Fournier, G. P., Akay,           Radiat. Res., 145: 602– 609, 1996.
    A. R., and Swartz, H. M. Tumor pO2 changes during photodynamic therapy depend            36. Veenhuizen, R. B., and Stewart, F. A. The importance of fluence rate in photody-
    upon photosensitizer type and time after injection. Comp. Biochem. Physiol. A, 132:          namic therapy: is there a parallel with ionizing-radiation dose-rate effects. Radiother.
    172–184, 2002.                                                                               Oncol., 37: 131–135, 1995.
 9. Kostron, H. The interaction of hematoporphyrin derivative, light and ionizing radi-      37. Tromberg, B. J., Orenstein, A., Kimel, S., Barker, S. J., Hyatt, J., Nelson, J. S., and
    ation in a rat glioma model. Cancer (Phila.), 57: 964 –970, 1986.                            Berns, M. W. In vivo tumor oxygen tension measurements for the evaluation of the
10. Moan, J., and Petersen, E. O. X-irradiation of human cells inculture in the presence         efficiency of photodynamic therapy. Photochem. Photobiol., 52: 375–385, 1990.
    of hematoporphyrin. Int. J. Radiat. Biol., 50: 107–109, 1981.                            38. Tromberg, B. J., Kimel, S., Orenstein, A., Barker, S. J., Hyatt, J., Nelson, J. S.,
11. Bellnier, D., and Dougherty, T. J. Hematoporphyrin derivative photosensitization and         Roberts, W. G., and Berns, M. W. Tumor oxygen tension during photodynamic
      -radiation damage interaction in Chinese hampster ovary fibroblasts. Int. J. Radiat.       therapy. J. Photochem. Photobiol. B Biol., 5: 121–126, 1990.
    Biol., 50: 659 – 664, 1986.                                                              39. Foster, T. H., Murant, R. S., Bryant, R. G., Knox, R. S., Gibson, S. L., and Hilf, R.
12. Roberts, D. J. H., Cairnduff, F., Dixon, B., and Brown, S. B. The response of a rodent       Oxygen consumption and diffusion effects in photodynamic therapy. Radiat. Res.,
    fibrosarcoma to combined treatment with photodynamic therapy and radiotherapy.               126: 296 –303, 1991.
    Int. J. Oncol., 6: 197–202, 1995.                                                        40. Gibson, S. L., Van Der Meid, K. R., Murant, R. S., Raubertas, R. F., and Hilf, R.
13. Ramakrishnan, N., Clay, M. E., Friedman, L. R., Antunez, A. R., and Oleinick, N. L.          Effects of various photoradiation regimens on the antitumor efficacy of photodynamic
    Post-treatment interactions of photodynamic and radiation-induced cytotoxic lesions.         therapy for R3230AC mammary carcinomas. Cancer Res., 50: 7236 –7241, 1990.
    Photochem. Photobiol., 52: 555–559, 1990.                                                41. Gibson, S. L., Foster, T. H., Feins, R. H., Raubertas, R. F., Fallon, M. A., and Hilf,
14. Dobler-Girdziunaite, D., Burkard, W., Haller, U., Larsson, B., and Walt, H. The              R. Effects of photodynamic therapy on xenografts of human mesothelioma and rat
    combined use of photodynamic therapy with ionizing radiation on breast carcinoma             mammary-carcinoma in nude-mice. Br. J. Cancer, 69: 473– 481, 1994.
    cells in vitro. Strahlenther. Onk., 171: 622– 629, 1995.                                 42. Hua, Z. X., Gibson, S. L., Foster, T. H., and Hilf, R. Effectiveness of -aminolevulinic
15. Kukielczak, B., Romanowska, B., and Bryk, J. Radiation and MC540 photosensi-                 acid-induced protoporphyrin as a photosensitizer for photodynamic therapy in vivo.
    tization of melanoma in the hamster’s eye. Melanoma Res., 9: 115–124, 1999.                  Cancer Res., 55: 1723–1731, 1995.
16. Berg, K., Luksiene, Z., Moan, J., and Ma, L. Combined treatment of ionizing              43. Reed, M. W., Wieman, T. J., Schuschke, D. A., Tseng, M. T., and Miller, F. N. A
    radiation and photosensitization by 5-aminolevulinic acid-induced protoporphyrin IX.         comparison of the effects of photodynamic therapy on normal and tumor blood
    Radiat. Res., 142: 340 –346, 1995.                                                           vessels in the rat microcirculation. Radiat. Res., 119: 542–552, 1989.
17. Winter, I., Overgaard, J., and Ehlero, N. The effect on photodynamic therapy alone       44. Fingar, V. H., Wieman, T. J., Wiehle, S. A., and Cerrito, P. B. The role of
    and in combination with misonidazole or x-rays for management of retinoblastoma              microvascular damage in photodynamic therapy: the effect of treatment on vessel
    like tumor. Photochem. Photobiol., 47: 419 – 423, 1988.                                      constriction, permeability, and leukocyte adhesion. Cancer Res., 52: 4914 – 4921,
18. Benstead, K., and Moore, J. V. The effect of combined modality treatment with                1992.
    ionising radiation and TPPS-mediated photodynamic therapy on murine tail skin.           45. Fingar, V. H., Kik, P. K., Haydon, P. S., Cerrito, P. B., Tseng, M., Abang, E., and
    Br. J. Cancer, 62: 48 –53, 1990.                                                             Wieman, T. J. Analysis of acute vascular damage after photodynamic therapy using
19. Pogue, B. W., Pitts, J. D., Mycek, M-A., Sloboda, R. D., Wilmot, C. M., Brandsema,           benzoporphyrin derivative (BPD). Br. J. Cancer, 79: 1702–1708, 1999.
    J. A., and O’Hara, J. A. In vivo NADH fluorescence monitoring as an assay for            46. Iinuma, S., Schomacker, K. T., Wagnieres, G., Rajadhyaksha, M., Bamberg, M.,
    cellular damage in photodynamic therapy. Photochem. Photobiol., 74: 817– 824,                Momma, T., and Hasan, T. In vivo fluence rate and fractionation effects on tumor
    2002.                                                                                        response and photobleaching: photodynamic therapy with two photosensitizers in an
20. Runnels, J. M., Chen, B., Ortel, B., Kato, D., and Hasan, T. BPD-MA-mediated                 orthotopic rat tumor model. Cancer Res., 59: 6164 – 6170, 1999.
    photosensitization in vitro and in vivo: cellular adhesion and B1 integrin expression    47. Major, A., Kimel, S., Mee, S., Milner, T. E., Smithies, D. J., Srinivas, S. M., Chen,
    in ovarian cancer cells. Br. J. Cancer, 80: 946 –953, 1999.                                  Z., and Nelson, J. S. Microvascular photodynamic effects determined in vivo using
21. Morgan, J., and Oseroff, A. R. Mitochondria-based photodynamic anti-cancer ther-             optical doppler tomography. IEEE J. Sel. Top. Quan. Elec., 5: 1168 –1175, 1999.
    apy. Adv. Drug Deliv. Rev., 49: 71– 86, 2001.                                            48. Olive, P. L., Vikse, C., and Trotter, M. J. Measurement of oxygen diffusion distance
22. Pogue, B. W., Paulsen, K. D., O’Hara, J.A., and Swartz, H. M. Estimation of oxygen           in tumor cubes using a fluorescent hypoxia probe. Int. J. Radiat. Oncol. Biol. Phys.,
    distribution in RIF-1 tumors by diffusion model-based interpretation of pimonidazole-        22: 397– 402, 1992.
    hypoxia and Eppendorf measurements. Rad. Res., 155: 15–25, 2001.                         49. Secomb, T. W., Hsu, R., Ong, E. T., Gross, J. F., and Dewhirst, M. W. Analysis of
23. Twentyman, P. R., Brown, J. M., Gray, J. W., Franko, A. J., Scoles, M. A., and               the effects of oxygen supply and demand on hypoxic fraction in tumors. Acta Oncol.,
    Kallman, R. F. A new mouse tumor model system (RIF-1) for comparison of endpoint             34: 313–316, 1995.
    studies. J. Natl. Cancer Inst. (Bethesda), 64: 595– 604, 1980.                           50. Dewhirst, M. W., Secomb, T. W., Ong, E. T., Hsu, R., and Gross, J. F. Determination
24. Fenton, B. M., and Way, B. A. Vascular morphometry of KHT and RIF-1 murine                   of local oxygen consumption rates in tumors. Cancer Res., 54: 3333–3336, 1994.
    sarcomas. Radiother. Oncol., 28: 57– 62, 1993.                                           51. Gulledge, C. J., and Dewhirst, M. W. Tumor oxygenation: a matter of supply and
25. Busch, T. M., Hahn, S. M., Evans, S. M., and Koch, C. J. Depletion of tumor                  demand. Anticancer Res., 16: 741–749, 1996.
    oxygenation during photodynamic therapy: detection by the hypoxia marker EF3             52. Kessel, D., Luo, Y., Deng, Y., and Chang, C. K. The role of subcellular localization
    [2-(2-nitroimidazol-1[H]-yl)-N-(3, 3, 3-trifluoropropyl)acetamide ]. Cancer Res., 60:        in initiation of apoptosis by photodynamic therapy. Photochem. Photobiol., 65:
    2636 –2642, 2000.                                                                            422– 426, 1997.
26. Liu, Y. H., Hawk, R. M., and Ramaprasad, S. In vivo relaxation time measurements         53. Kessel, D., and Luo, Y. Mitochondrial photodamage and PDT-induced apoptosis.
    on a murine tumor model: prolongation of T1 after photodynamic therapy. MRI                  J. Photochem. Photobiol. B Biol., 42: 89 –95, 1998.
    (Magn. Reson. Imaging), 13: 251–258, 1995.                                               54. Kessel, D., and Luo, Y. Photodynamic therapy: a mitochondrial inducer of apoptosis.
27. Goda, F., Bacic, G., O’Hara, J. A., Gallez, B., Swartz, H. M., and Dunn, J. F. The           Cell Death Differ., 6: 28 –35, 1999.
    relationship between partial pressure of oxygen and perfusion in two murine tumors       55. Oleinick, N. L., and Evans, H. H. The photobiology of photodynamic therapy: cellular
    after X-ray irradiation: a combined gadopentetate dimeglumine dynamic magnetic               targets and mechanisms. Radiat. Res., 150: S146 S156, 1998.
    resonance imaging and in vivo electron paramagnetic resonance oximetry study.            56. Xue, L., He, J., and Oleinick, N. L. Promotion of photodynamic therapy-induced
    Cancer Res., 56: 3344 –3349, 1996.                                                           apoptosis by stress kinases. Cell Death Differ., 6: 855– 864, 1999.
28. O’Hara, J. A., Goda, F., Demidenko, E., and Swartz, H. M. Effect on regrowth delay       57. Richter, A. M., Waterfield, E., Jain, A. K., Canaan, A. J., Allison, B. A., and Levy,
    in a murine tumor of scheduling split-dose irradiation based on direct pO2 measure-          J. G. Liposomal delivery of a photosensitizer, benzoporphyrin derivative monoacid
    ments by electron paramagnetic resonance oximetry. Radiat. Res., 150: 549 –556,              ring A (BPD), to tumor tissue in a mouse tumor model. Photochem. Photobiol., 57:
    1998.                                                                                        1000 –1006, 1993.
29. Pogue, B. W., O’Hara, J. A., Liu, K. J., Hasan, T., and Swartz, H. M. Photodynamic       58. Swartz, H. M., Boyer, S., Gast, P., Glockner, J. F., Hu, H., Liu, K. L., Moussavi, M.,
    treatment of the RIF-1 tumor with verteporfin with online monitoring of tissue oxygen        Norby, S. W., Vahidi, N., Walczak, T., Wu, M., and Clarkson, R. B. Measurements
    using electron paramagnetic resonance oximetry. Proc. SPIE, 3601: 108 –114, 1999.            of pertinent concentrations of oxygen in vivo. Magn. Reson. Med., 20: 333–339,
30. Fingar, V. H., Mang, T. S., and Henderson, B. W. Modification of photodynamic                1991.
    therapy-induced hypoxia by fluosol-DA (20%) and carbogen breathing in mice.              59. Liu, K. J., Gast, P., Moussavi, M., Norby, S. W., Vahidi, N., Walczak, T., Wu, M.,
    Cancer Res., 48: 3350 –3354, 1988.                                                           and Swartz, H. M. Lithium phthalocyanine: a probe for electron paramagnetic
                                                              NONVASCULAR PDT ENHANCES RADIATION SENSITIVITY

      resonance oximetry in viable biological systems. Proc. Natl. Acad. Sci. USA, 90:        66. Cincotta, L., Szeto, D., Lampros, E., Hasan, T., and Cincotta, A. H. Benzophenothia-
      5438 –5442, 1993.                                                                           zine and benzoporphyrin derivative combination phototherapy effectively eradicates
60.   O’Hara, J. A., Goda, F., Liu, K. L., Bacic, G., Hoopes, P. J., and Swartz, H. M.            large murine sarcomas. Photochem. Photobiol., 63: 229 –237, 1996.
      Oxygenation in a murine tumor following radiation: an in vivo electron paramagnetic     67. Henderson, B. W., Busch, T. M., Vaughan, L. A., Frawley, N. P., Babich, D., Sosa,
      resonance oximetry study. Radiat. Res., 144: 224 –229, 1995.                                T. A., Zollo, J. D., Dee, A. S., Cooper, M. T., Bellnier, D. A., Greco, W. R., and
61.   O’Hara, J. A., Goda, F., Liu, K. J., Bacic, G., Hoopes, P. J., and Swartz, H. M. The        Oseroff, A. R. Photofrin photodynamic therapy can significantly deplete or preserve
      pO2 in a murine tumor after irradiation: an in vivo electron paramagnetic resonance         oxygenation in human basal cell carcinomas during treatment, depending on fluence
      oximetry study. Radiat. Res., 144: 222–229, 1995.                                           rate. Cancer Res., 60: 525–529, 2000.
62.   Demidenko, E. Asymptotoic properties of mixed effects models. In: T. G. Gregorie        68. Berg, K., Steen, H. B., Winkelman, J. W., and Moan, J. Synergistic effects of
      (ed.), Modeling Longitudinal and Spatially Correlated Data. New York: Kluewer               photoactivated tetra(4-sulfonatophenyl)porphine and nocodazole on microtubule as-
      Press, 1997.                                                                                sembly, accumulation of cells in mitosis and cell survival. J. Photochem. Photobiol.
63.   Rice, J. A. Mathematical Statistics and Data Analysis. Belmont, CA: Duxbury Press,          B Biol., 13: 59 –70, 1992.
      1995.                                                                                   69. Castro, D. J., Saxton, R. E., Haghighat, S., Reisler, E., Plant, D., and Soudant, J. The
64.   Schmidt-Erfurth, U., Bauman, W., Gragoudas, E., Flotte, T. J., Michaud, N. A.,              synergistic effects of rhodamine-123 and merocyanine-540 laser dyes on human
      Birngruber, R., and Hasan, T. Photodynamic therapy of experimental choroidal                tumor cell lines: a new approach to laser phototherapy. Otolaryngol. Head Neck
      melanoma using lipoprotein-delivered benzoporphyrin. Ophthalmology, 101: 89 –99,            Surg., 108: 233–242, 1993.
      1994.                                                                                   70. Dewhirst, M. W., Sim, D. A., Gross, J. F., and Kundrat, M. A. Effects of heating rate
65.   Henderson, B. W., and Bellnier, D. A. Tissue localization of photosensitizers and the       on normal and tumor microcirculatory function. In: R. Diller (ed.), Heat and Mass
      mechanism of photodynamic tissue destruction. Ciba Found. Symp., 146: 112–125;              Transfer in the Microcirculation of Thermally Significant Vessels, pp. 75– 80. Amer-
      discussion, 125–130, 1989.                                                                  ican Society of Mechanical Engineers, 1986.


Shared By: