Publication of the International Union Against Cancer Publication de I Union InternationaleContre le Cancer

Itit. J. Cancer: 52,491 -498 (1992) 0 1992 Wiley-Liss, Inc.

EXPERIMENTAL PHOTODYNAMIC THERAPY WITH A COPPER METAL VAPOR LASER IN COLORECTAL CANCER J-M. L A N T Z ~C. , ~MEYER'. , C. SAUSSINE~, C. LEBERQUIER', F. HEYSEL3, J. MIEHE', J. MARESCAUX?, R. SULTAN4 and M. KEDINGER' IUnit6 61 INSERM, 3 avenue Moliere, 67200 Strasbourg; ZChinirgieA , Hospices Civils, 67000 Strasbourg; 'CNRS, Groupe d'Optique AppliqirCe, 23 rue du Loess, 67037 Strasbourg; and Tentre Europten de Recherche Medico-Chirurgicale Laser, 92501 Rueil Malmaison, France. In an attempt to define the best conditions for an adjunctive treatment of residual colonic microtumors by photodynamic therapy (PDT), an experimental model has been defined. S.C. HT29 colonic-cancer-cell tumors grown in nude mice were used and, 48 hr after i.p. administrationof 30 mg/kg Photofrin (PH), laser illumination was performedwith 75 or I 5 0 Joules/cm2.The efficiencies of 2 lasers, the classically used rhodamine laser (RL) and a copper metal vapor laser (CMVL), were compared, The effects of PDT were assessed by histological and immunocytochemical (detection of a digestive enzyme, dipeptidylpeptidase IV, as a marker of cell viability) follow-up and by the growth curve of the tumors after illumination. We conclude that, although the depth of necrosis resulting from PDT was nearly 3 mm at 75 J/cm2and nearly 4-5 rnm at I 5 0 J/cm2with both lasers, complete necrosis was obtained only with the CMVL at 150 J/cm2 (in 50% of the tumors). Under the other conditions, a layer of unaffected cells persisted at the pole opposite to laser illumination, resulting in growth curves lower than but parallel to those of the controls. Analysis of drug concentrations in the tumors and various organs, 48 hr after injection, i.e., at the time of laser illumination, revealed the presence of 2 I pg/g dry weight PH in the tumors. The tumor vs. host-organ ratios were equal to or higher than I for the small bowel, colon, stomach, lung, skin and muscle. In contrast, the ratios were below I for the spleen, pancreas, kidney and liver. 8 1992 Wile)-Liss, Inc.

Colorectal cancer is one of the leading causes of mortality from cancer in Western countries (Silverberg and Lubera, 1986). Although 70% of the patients are treated with curative intent, and despite recent advances in radiotherapy and chemotherapy, 50% of them die from locoregional recurrence or distant metastasis (Moertel et a/., 1990). It has been shown that 37 to 63% of all colorectal cancer recurrences are confined to the abdomen (Willett et al., 1984; Sugarbaker ef al., 1987). These recurrences are probably linked with micrometastases which are present at the time of the presumed curative resection but can be neither detected nor treated (Buyse et al., 1988). Photodynamic therapy (PDT) has been described as a technique with the potential for selective local destruction of cancer. It is based upon the use of photosensitizers which, after injection, are selectively retained in the tumors and are able to produce a photochemical reaction after they have been excited by light, leading to cellular death in the target tissue (Ash and Brown, 1989). Photodynamic therapy has nowadays several applications in oncology but its perioperative development is only in its early stages (Herrera-Ornelas et a/., 1986; Nambisan et al., 1988; Sindelar et al., 1991). Photofrin (PH) is the most widely used photosensitizer, which is also the only one to have clinical applications. P H is usually excited with a dye laser emitting at 630 nm despite the weak absorption peak at this wavelength, because the penetration of red light in biological media is theoretically far better than with a shorter wavelength (Svaasand, 1984). However, the low power of the dye laser available at the present time precludes a perioperative use because the duration of illumination would be too long to obtain a curative light dosimetry in the whole operative field.

The necessity of developing adjuvant therapy in colorectal cancer, the proved efficiency of photodynamic therapy in oncology and the limits inherent to the low power of dye lasers available at present led us to promote the concept of adjunctive PDT using a copper metal vapor laser (CMVL). The potential advantage of this laser is that it produces a 30- to 60-cm2 beam of light with a power of 0.5 to 1 W/cm2 which could treat disseminated tumoral residues after surgical resection within a short time. However, since the laser emission at 510 and 580 nm is theoretically less penetrating than the classically used 630-nm illumination, the ability of the CMVL to cure tumors has to be proven. Bellnier et al. (1985), using a copper vapor laser, have shown that hematoporphyrin derivative plus 514,5 nm laser radiation may be an effective treatment for small and superficial malignant lesions of the urinary bladder. In the present study, we tested the efficiency of illumination with a CMVL and compared it with the classical rhodamine laser (RL) after administration of P H on colonic tumors grown in nude mice and reaching 1cm in diameter. The tumors used derive from a human colonic cancer cell line (HT29 cells, Fogh et af., 1977) injected S.C. The effects of the illumination on the tumors were analyzed using the following parameters: morphological features, immunocytochemical detection of a digestive enzyme (dipeptidylpeptidase IV) and comparative growth curves. The concept of an intra-abdominal PDT supposes an illumination in the peritoneal cavity, which raises the question of possible side-effects of P D T on intra-abdominal organs (Douglas et a/., 1981). For this purpose, the concentration of the photosensitizer in the tumors and various organs of the host at the time of laser illumination was determined, allowing an estimation of the tumor-to-organ ratio to be made. MATERIAL AND METHODS

Cells, material

Animals and tumor Jystern The tumor cells used (HT29 cell line) were derived from a human colo-rectal adenocarcinoma (Fogh et aL, 1977). The cells were routinely grown as monolayers in Dulbecco's modified Eagle medium supplemented with 10% fetal calf serum (Gibco, Glasgow, UK) at 37°C in a 95% air, 5% COz humidified atmosphere. The cells used to inoculate nude mice were obtained from actively dividing cultures, before confluence. Single-cell suspensions were obtained with 0.25% trypsin, 0.53 mM EDTA in PBS.

jTo whom correspondence and reprint requests should be sent Abbreviations: PDT, photodynamic therapy; PH, Photofrin; s.c., subcutaneous; i.p., intraperitoneal; CMVL, copper metal vapor laser; RL. rhodamine laser; J, joule; PAS, periodic-acid-Schiff PCA, perchloric acid.

Received: April 29,1992 and in revised form May 22,1992.

492

LANTZ E T A L .

Non-inbred Swiss athymic male nu/nu mice were obtained from Iffa Credo (l’Arbresle, France). The mice, 4-6 weeks old, average weight 30 g, received lo7 HT29 cells suspended in 200 p1 of culture medium, injected into the flanks. The tumors were generally palpable at 5 days and reached a size of 8-10 mm after 14-16 days, at which time the experiments started.

Photosensitizer PH was obtained from QLT Phototherapeutics (Pearl River, N.Y.) as a freeze-dried powder and was dissolved in a solution of dextrose (5%) to a concentration of 5 mg/ml. At the time of injection, P H was further diluted with a 5% dextrose solution. Laser light deliveiy system Two different lasers were used. Illumination at 630 nm was performed with an argon-pumped dye laser. The dye used was rhodamine B with a tuning range of 580-630 nm and a power of 150 mW at 630 nm. Illumination at 510 nm and 580 nm was performed with a CMVL (Aka Laser, Saint Louis, France), whose physical properties are listed in Table I. PI1 concentrations in the tumors and organs at the time of laser illumination PH (30 mg/kg) was injected i.p. in nude mice bearing HT29 tumors. Eight tumor-bearing animals were killed by cervical dislocation 48 hr after PH injection. The tumors, liver, spleen, pancreas, small and large bowel, kidneys, lungs, muscle and skin were removed from each animal and stored at -20°C until analysis. The amounts of P H in the different samples were determined as follows.

Extraction procedure The tissues were desiccated and subjected to porphyrin extraction by 2 successive sonications (30 sec) in 1 N PCA/ methanol (l:l, v/v) containing 1%SDS according to Pengetal. (1991), as adapted from Gomer et al. (1985). After each sonication, the homogenate was centrifuged at 3,000 g for 10 min and finally the supernatants were pooled. Fluorescence photometry Fluorescence was measured in the pooled supernatants with a spectrofluorophotometer (Shimatzu RF-540, Roucaire, France) using the following settings: excitation 400 nm, emission 622 nm. A standard plot of fluorescence vs. P H concentration in SDS/PCA-methanol was linear from 0.1 pg to 1 pg/ml. There was no fluorescence quenching by any of the tissues studied and the autofluorescence observed, mainly due to an overlapping with hemoglobin, was subtracted from the total fluorescence value. Determinations were done in duplicate and there was very little variation between assays. Photodynamic therapy Sixty-four nude mice were illuminated 48 hr after i.p. injection of 30 mg/kg P H and after anesthesia with tribromoethanol. Direct illumination was done by dissecting the tumors TABLE 1 CHARACTERISTICS OF THE COPPER METAL VAPOR LASER ~

Average power Wavelength Pulse repetition Pulse energy Pulse duration Peak power Warm-up time Cu reload time Beam diameter Divergence

>25 W 510.5 nm-578.2 nml 4 to 4.5 kHz 5.5 to 6.5 mJ about 20 nsec 60 kW < 2 hr > 500 hr 40 mm About 3 mrad

‘Approximately % of the total power on 510.5-nm line.

from the abdominal muscle and reversing the tumor-bearing skin fragment. Thirty-two mice were treated with the R L with a power density of 150 mW/cm2: 16 mice received a total illumination of 75 J/cm2 (group 1) and 16 received 150 J/cm2 (group 3). Thirty-two mice were treated with the CMVL with a power density of 500 mW/cm2: 16 received a total illumination of 75 J/cm2 (group 2) and 16 received 150 Jicm2 (group 4). The temperature of the tumors during illumination was measured with a thermic contact probe (Nutherm P, Chauvin Amoux, Strasbourg). Control mice received neither light nor PH, or light only, or PH only. Histological and immunocytochemical study Six animals from each group were killed 24 hr, 48 hr or 7 days after illumination for morphological analysis. The tumors were removed and 3-mm slices were cut in a plane parallel to laser illumination in the middle of the tumors for histological and immunocytochemical studies. Subsequent embedding and sections allowed visualization of the whole plane of the tumor.

Hstological analysis The tumor slices were fixed in Bouin’s fluid, embedded in paraffin and processed routinely; 5-pm sections were cut, stained with periodic-acid-Schiff (PAS) and hematoxylineosin, observed with a Zeiss microscope and photographed. The depth of necrosis and the tumor diameters were measured. tmmunofluorescence study The tumor slices were embedded in Tissue-Tek, frozen in freon cooled in liquid nitrogen and stored at -70°C until use. Frozen sections 5 pm in thickness were prepared, air-dried and kept frozen. Dipeptidyl peptidase IV (DPP IV), a brushborder enzyme, was revealed with a monoclonal anti-DPP-IV antibody (kindly provided by Dr. H.P. Hauri, Biozentrum, Basel, Switzerland) at a 1/75 dilution in PBS for 2 hr in a moist chamber. After several washings, the sections were incubated with a fluorescein-isothiocyanate-conjugated sheep antimouse antibody (1/200 in PBS; Institut Pasteur, Paris, France). The preparations were mounted in glycerol/PBS/phenylenediamine under a coverslip and observed with a Zeiss Orthoplan microscope. Tumor-growth study This study was performed on 10 mice each from groups 1, 2 and 4. After laser illumination, the animals were kept in the dark and the tumors were measured twice weekly for 3 weeks. The tumor volume was evaluated by the formula V = a / 6 x length x width x depth assuming that the tumor was spheroidal. The responses were classified as complete when the tumor was no longer palpable, and partial when there was a decrease in size as compared to that of control tumors. No response corresponded to continuous growth. For partial responses, growth curves were constituted and the absolute response rate was evaluated by calculating the ratio of the mean volumes of the tumors after PDT and the mean volume of the control groups. The relative response rate was calculated by relating the absolute response rate to the volume of the tumors at the time of treatment. RESULTS

PH concentrations in the tumors and organs At the time of laser illumination, 48 hr after P H injection, the concentrations of the photosensitizer in the tumors and in various host organs as well as the selectivity ratios between the tumor and the other tissues were measured and are listed in Table 11.

493

PHOTODYNAMIC THERAPY IN H T 29 TUMORS

TABLE I1 - PH CONCENTRATIONS AND TUMORIHOST-ORGAN RATIOS, 48 HR AFTER INTRAPERITONEAL INJECTIONS

PH concentration

HT29 tumor Muscle Liver Spleen Pancreas Kidnev Lung Small bowel Colon Stomach Peritumoral skin

20.7 2 1.5 15.7 2 1.9 116.5 2 20.1 74.8 2 8.2 35.7 2 4.1 43.3 4.7 15.5 2 1.8 15.7 2 1.7 20.0 2 2.4 13.3 2 1.8 10.8 2 1.4

*

Tumoriorgan ratio

1.3 0.2 0.3 0.6 0.5 1.3 1.3 1.0

1.6 1.9

The values are expressed as wg per g of dry tissue:mean values

of 8 animals (16 individual tumors and 8 individual host organs) t

SEM.

The mean value of PH within the tumor was 21 Fg/g dry weight. The ratio between the tumor and the adjacent skin was the highest (1.9). At 48 hr, the ratios for intra-abdominal organs were equal or superior to 1 only for the small bowel, colon and stomach as for skin, lung and muscle. Conversely, the ratios of tumor to pancreas, liver, spleen and kidney were always particularly low. Photodynamic therapy

Histoloscal arid immunocytochemical analysis of the tumors after PDT In the control groups, the tumors were surrounded by a regular fibroblastic cell sheath. Within this fibroblastic outer envelope, HT29 cells were arranged into clumps of cells delineated by stromal areas containing fibroblasts and vessels (Fig. la). The latter were revealed immunocytochemicallyby a basement membrane staining with anti-laminin or antiheparan sulfate proteoglycan antibodies (not shown). The HT29 areas were composed of cells forming small glands around the lumen, either filled with PAS-positive material or exhibiting intense red PAS staining around their borders (Fig. lb). Such glands were organized into small clusters of polarized cells with apical brush borders facing the lumen and carrying digestive hydrolases such as DPP IV, as visualized immunocytochemically with a specific monoclonal antibody (Fig. 2a). The very large tumors (grown for 3 weeks) exhibited non-specific, central, focal areas of necrotic cells. During illumination, the temperature at the surface of the tumor rose progressively from 28°C to 31 % 1°C in groups 1 and 3, to 36 2 1°C in group 2, and to 40 ? 1°C in group 4 (Table 111). Despite this partial hyperthermic effect, none of the control groups, illuminated without photosensitizer, displayed any necrosis. The combined effects of laser and photosensitizer resulted in a photochemical reaction in the 4 treatment groups and were characterized 24 and 48 hr after illumination by an area of necrosis with nuclear pyknosis and karyorrhexis, with a loss of the basic cellular shape. The body of the tumor was filled with granular debris (Fig. lc-f). The side nearest the source of illumination could be recognized by the disappearance of the fibroblastic layer surrounding the tumor (Fig. lc). The cell damage led to a loss of specific fluorescence after immunocytochemical staining of DPPIV (Fig. 2b-d).A small rim of living HT29 cells persisted beneath the fibroblasts at the pole opposite to the source of illumination in groups 1 , 2 and 3 (Fig. Id). The viability of these cells was confirmed by the persistence of specific fluorescence after immunocytochemical staining (Fig. 2).In group 4, 5 tumors out of 6 were completely necrosed after laser treatment and 1 tumor exhibited unal-

tered cells near the fibroblastic envelope at the pole opposite to the source of illumination. Seven days after treatment, the areas of necrosis were colonized by stromal cells embedded in amorphous, light-pink, PAS-positive material (Fig. le). The peripheral crown of HT29 cells had grown in groups 1 , 2 and 3 (Figs. lfand 2d). Analysis of the depth of necrosed areas in the tumors under the various conditions of illumination is summarized in Table 111. Necrosis of the tumors was never complete in groups 1, 2 and 3. The mean depth of necrosis was equal for both lasers in the case of a 75 J/cm2 illumination and never exceeded 3 mm. It increased to 5 mm after 150 J/cm2 illumination with the RL but none of the tumors treated showed complete necrosis, a small rim of viable cells being persistent even in 3-mm-thick tumors. The depth of necrosis reached 4 mm after a 150 J/cm2 illumination with the CMVL. In this case, all the tumors smaller than 5 mm showed complete necrosis. In one 5-mmthick tumor, a 1-mm crown of living cells persisted opposite to the illumination pole, leading us to assume a 4-mm penetration of the CMVL at 150 J/cm2.

Follow-up of the growth curves and response rate The results of the different protocols of PDT are summarized in Table IV and Figure 3. All the tumors treated with 75 J/cm2 illumination (groups 1 and 2) had a partial response characterized by a continuous but lower rate of growth as compared to the controls, leading to parallel growth curves. None of these tumors showed any necrosis of the adjacent skin. In group 4, 5 tumors out of 10 were definitively cured and showed marked skin necrosis around the tumor which healed spontaneously within 3 weeks. The 5 remaining tumors exhibited a partial response (Fig. 3), only one of them developing a peritumoral skin necrosis. In group 3, all mice died within 48 hr after illumination, presumably as an indirect consequence of the long period of anesthesia needed for this intensity with the RL (illumination time: 16 min per tumor). Photodynamic therapy resulted in a 47% reduction of mean tumor volume in group 1, a 55% reduction in group 2 and a 34% reduction in the 5 tumors showing a partial response in group 4, when compared with the controls (Table IV). The corresponding relative response ratios were 65% in group 1, 68% in group 2, and 76% in group 4 when referred to the mean tumor volume at laser time (respectively 89, 55 and 179 mm3). DISCUSSION

The present study was designed to explore the potential of a copper metal vapor laser compared to the rhodamine laser in photodynamic therapy. The experimental model used was a human colonic carcinoma implanted S.C. in nude mice, and PH was used as a photosensitizer. Our data show a complete disappearance of 5 out of 10 tumors and a 70% regression of the remaining tumors after a 150 Jlcm2 illumination with the CMVL, 48 hr after i.p. injection of 30 mg/kg PH. Only partial tumor destruction was obtained under the same conditions with the RL, and with lower illumination doses with both lasers. The logarithmic relationship between light dose and depth of tumor necrosis in the PDT response (Berenbaum et al., 1982; Van Gemert et al., 1985), is supported by our histologic data with both CMVL and RL. The similar depth of necrosis observed with the 2 lasers (2.8 mm and 4-5 mm after respectively 75 and 150 J/cm2 illumination) is somewhat in opposition to the observations that the penetration of light in biological media, one of the determining factors for a photodynamic reaction, is far less efficient for green than for red light (1 to 3 mm at 630 nm and 0.5 to 1.5 mm at 514 nm; Svaasand, 1984). However, the observation that an increase in the light dose by a factor of 2 results in an increase in the depth of

FZGURE 1 - Histological views of S.C.HT29 tumors illuminated with the copper metal vapor laser at 75 J without photosensitizer (controls: a, b ) or 48 hr after i.p. administration of 30 m$/kg PH (c-f). Tumors were removed 48 hr ( a d ) and 7 days ( e , f ) after laser treatment. The sections were stained with periodic-acld-Schiff and hematoxylin-eosin. f fibroblastic outer envelope; D glandular structures; lumen filled (,#) or bordered (0)with PAS-positive material; Z:pole of laser illumination; n: necrosis; Fi: fibrosis. *: persistence of living HT29 cells. **: growth of the unaffected HT29 cell rim. Bars: SO pm.

FIGURE 2 - Immunocytochemical localization of a brush-border digestive enzyme: dipeptidyl peptidase (DPP IV) revealed with an anti-DPP IV monoclonal antibody in a control tumor ( a ) and in tumors submitted to the combined effect of PH and CMVL illumination at 75 J, and removed 48 hr (b,c) or 7 days ( d ) after treatment. The specific fluorescence is localized either at the luminal surface of glandular structures 0 )or in the whole lumen of other glands (0).No specific labeling can be seen in necrosed areas (n). Other details are the same as in Figure 1. Bars: 30 pm.

496

LANTZ E T A L . TABLE U1- ANALYSIS OF TEMPERATURE DURING ILLUMINATION AND DEPTH OF TUMOR NECROSIS UNDER VARIOUS ILLUMINATION CONDITIONS (PDT 48 HR AFTER PH INJECTION) Number Mean tumor Groups of tumors diameter SEM examined (mm)

*

1

2 3 4

6 6 6 6

4.4 2 1.2 5.5 2 1.1 4.1 2 2.4 4.5 2 2.4

Dose rate (mW/cm')

Type of laser

RL1 (75J/cm2) CMVL2 (75J/crn2) RL1 (150J/crn2\ CMVL~(i5OJ/&2)

150 150

snn 500

Temperature increase durlng illumination 2 SEM ("C)

Maximal depth of

(mm)

Number of tumors complete,y necrotized

28to31" 2 1 28 to 36" _c 1

3 2.9

0 0

2 8 t n 3 1 0 - c1

5

n

28 to 40" 5 1

4

5

'RL, Rhodarnine laser.-2CMVL, copper metal vapor laser. TABLE IV - RESPONSE RATE AFTER PDT 48 HR AFTER PH INJECTION

Number of tumors followed CR' PR? NR3 ARR4 during 3 weeks

Group 1 (RL, 75J/crn?) Group 2 (CMVL. 75J/cm2)

mm

RRRS

10

0

10

0

47%

65%

10

0

10

0

55%

58%

.'"

300-

T

200-

ICR, complete response.-?PR, partial response.--'NR, no response.-"ARR, absolute response rate calculated as the mean tumor volume reduction over the whole period studied after PDT.IRRR, relative response rate, taking into account the volume of the tumors at the time of illurnination.-6Calculated for the 5 tumors which exhibited a partial response. necrosis from 2.8 mm to 4 mm and 5 mm for the CMVL and RL respectively might indicate the following values for HT29 tumors, optical penetration depth 6: 4 - 2.8

for CMVL laser: 2 = e - ie., SCMVL= 1.7 (mm)

TMVL

5 - 2.8 and for RL laser: 2 = e - i.e., SRL

=

3.2 (mrn)

8~~

These optical penetration depths are slightly higher than those observed by Svaasand (1984), but possible for tissues with limited scattering and blood content. In addition, the pulsed emission of the CMVL with high peak power of 60 KW can partly explain a better optical penetration (Hisazumi et al., 1985). The observation that the depths of necrosis were the same for the 2 lasers after 75/cm2 illumination (2.8 mm) can be expressed as follows: ccMvLe-?R/I.7

= cRLe-29'3.2

i.e., '

CCMVL - 2.2 CRL

where C is the efficiency parameter i.e., CCMVL = 2.2 x C R ~ . A higher efficiency parameter for the photodynamic reaction of the green light compared to the red light will result in a higher efficacy for the green light in the superficial layers, whereas the efficacy of the red light will, due to the greater optical penetration depth, be higher in distal layers. The observed break-even distance of 2.8 mm corresponds, as discussed above, to a ratio between the efficiency of green light and red light equal to 2.2. This result is in good agreement with expectations; the ratio of the absorption in H P D at 510 nm to that of the 630 nm is 2.5 (Doiron et al., 1983). The emission for %1 of the total power density in yellow light (578 nm) with the CMVL, which corresponds to an absorption peak of PH, can also partly enhance the photodynamic reaction but only in superficial layers because the optical penetration at 578 nm,

100-

U'

0

10

20

30

days

RGURE 3 - Growth curves of the control tumors (---X---)and of the tumors in group 1 (-0-), 2 (-0-)and 4 (-m-). Mean values mm3 + SEM.

which is very close to the absorption peak of oxyhemoglobin at 577 nm, is usually expected to be lower than for 514 nm light. The better efficiency of the CMVL could be linked-in addition to its physical properties mentioned above-to an associated hyperthermic side-effect and to a peripheral mechanism of vascular photodynamic infarctus involved in the complete destruction of tumors. In our study the probability of a non-specific hyperthermic effect or a partial synergistic hyperthermic effect can be excluded. Gomer et al. (1988) have reported an increase in the PDT response rate in experimental tumors after raising the power density from 150 to 500 mW/cm* and attribute this fact to a non-specific hyperthermic effect. In our experiment, although the temperature rose to 40°C during the 150 J/cm2 illumination with the CMVL (500 mW/cmz), control tumors failed to show any necrosis. A synergistic effect of P D T and hyperthermia has also been demonstrated by Waldow and Dougherty (1984), by increasing the temperature to 4045°C immediately after PDT, and confirmed by Mang (in Dougherty, 1987) who improved the cure rate of an experimental pulmonary tumor from 27% to 47% by adding to their PDT a 30-min exposure to 44°C. However, compared to this, the hyperthermic dose in our experiment was small (40°C for 5 min) and its role seems to have been rather insignificant. The mechanism of peritumoral vascular infarctus seems to be of importance in the determination of tumor destruction with PDT. During PDT, tumor cells can be destroyed either directly, as demonstrated with tumor cells explanted in vitro (Henderson and Fingar, 1989), or indirectly by a destruction of tumoral vessels leading to cellular anoxia. Both occurrences are probably involved in the photodynamic reaction. But the

PHOTODYNAMIC THERAPY IN HT 29 TUMORS

prime importance of an indirect vascular mechanism in obtaining definitively cured tumors is suggested by the following arguments. We only observed complete tumor necrosis and durable complete response when the skin overlying the tumor was destroyed. A similar observation was also reported by Fingar and Henderson (1987), who showed that R I F tumor cure by PDT could only by achieved under conditions which resulted in a considerable amount of normal tissue surrounding the tumor being also destroyed. The hypothesis that stromal tumor cells and vessels could be important targets in PDT has also been supported by experimental arguments based on: (1) pharmacokinetic studies with low video fluorescence microscopy of pcritumoral tissues (Nelson et al., 1985); (2) measurements of flux velocity with a laser Doppler velocimeter after PDT (Wieman et al., 1988); (3) experiments on cell clonogenicity following PDT and compared to experiments under anoxic conditions (Fingar et al., 1987); (4) morphologic studies of the subendothelial zone of the tumor capillaries (Nelson et al., 1987). The particular sensitivity to P D T of the stromal and vascular cells within the tumors could be related to a leaky vasculature, a poor lymphatic drainage, the presence of macrophages and/or an elevated concentration of lipoprotein receptors (Moan, 1986; Pantelides et al., 1989). From all these observations, it seems that tumor cure is linked to a peritumoral vascular infarctus which is only possible when the peritumor contains the photosensitizer. The weak differences in P H uptake between transplantable S.C. tumors in mice and the skin surrounding the tumor [ratios of 1.1 in a mammary carcinoma (Bellnier et al., 1989) and 2.2 in Lewis lung carcinoma (Kessel, 1986)] correlate well with our findings in HT29 colonic tumor (1.9) and could allow a peritumoral photochemical reaction. Thus, in our colonic cancer model, the study of P H concentrations showed a very partial and weak selectivity. The tumor/organ ratios were lower than 1 for several organs. The amounts of photosensitizer present in the tumor were only higher than those observed in the muscle, small and large bowel, lung and peritumoral skin. The lack of selectivity due to a high and persistent retention of PH in many organs has

497

already been described by several authors using lower doses of P H (5-10 and 20 mg/kg) (Bellnier et al., 1989; Pantelideset al., 1989; Peng et al., 1990). Liver, spleen and kidney exhibit the highest concentrations of PH in all experiments (Gomer and Dougherty, 1979; Bugelski et al., 1981). The high retention of porphyrin we observed within the pancreas was also seen by Mang and Wieman (1987) in Syrian hamster and Wistar-Lewis rat. Pantelides et al. (1989), Bellnier et al. (1989) and Peng et al. (1990) reported accumulation in almost all intra-abdominal organs studied (stomach, bladder, colon, kidney, prostate, adrenal gland). The existence of high porphyrin levels in the intra-abdominal organs up to 75 days after injection (Bellnier et al., 1989) raises the question of side-effects in intraabdominal PDT. Although the sensitivity of normal organs to P D T remains a matter of controversy (Douglass et al., 1981; Tochner et al., 1985; Bown et al., 1986; Mang et Wieman, 1987; Suzuki et al., 1987), the persistence of photosensitizer at the time of illumination points to the necessity of a second-phase study of i.p. PDT with CMVL before it can be applied in clinical experiments. In conclusion, our results support the view that the copper metal vapor laser could be as efficient as the classic rhodamine laser in PDT. Its major advantage in the concept of perioperative photodynamic therapy would be to reduce the length of the illumination period. Indeed, the phase-I treatment of disseminated i.p. neoplasms initiated by Sindelar et al. (1991) shows that, despite the use of 2 lasers emitting at 630 nm, the total duration of illumination was 4 hr to achieve a dose of 3 J/cm2. In the same manner the CMVL could do the same within 30 min.

ACKNOWLEDGEMENTS

We are grateful t o Miss C. Arnold and Mrs E. Alexandre for invaluable technical assistance, and to Mrs I. Gillot for excellent secretarial help. This work was supported by INSERM, the Conseil RCgional d’Alsace and the Ligue Nationale Francaise contre le Cancer du Haut-Rhin.

REFERENCES ASH.D. and BROWN,S.B., Photodynamic therapy. Achievements and prospects. Brit. J. Cancer. 60,151-152 (1989). BELLNIER, D.A., Ho, Y.K., PANDEY,R.K., MISSERT,J.R. and DOUGHm r y . T.J., Distribution and elimination of PHOTOFRIN I1 in mice. Phofochetn Photobiol.. 50,221-228 (1989). B ~ L L N I ED.A., R , PROUT,G.R. and LIN,C.V., Effect of 514.5-nm argon ion laser radiation on hematoporphyrin derivative-treated bladder tumor cells in vitro and in vivo. J. nut. Cancer Inst., 74, 617-625 (1985). BERENBAUM, M.C., BONETT,R. and SCOURIDES, P.A., In ii1.0 biological activity of the components of haematoporphyrin derivative. Brit. J. Cancer, 45,571-581 (1982). BOWN.S.G.. TRALAU, C.J., COLERIDGE SMITH,P.D., AKDEMIR, D. and WiEMAN, T.J., Photodynamic therapy with porphyrin and phthalocyanine sensitisation: quantitative studies in normal rat liver. Brit. J. Cancer, 54,43-52 (1986). BUGELSKI, P.J., PORTER,C.W. and DOUGHERTY, T.J., Autoradiographic distribution of hematoporphyrin derivative in normal and tumor tissue of the mouse. Cancer Rex, 41,4606-4612 (1981). BUYSE,M., ZELENIUCH-JACQUO~TE, A. and CHALMERS, T.C., Adjuvant therapy of colorectal cancer; why we still don’t know. J. Amer. med. AS., 259,3571-3611 (1988). DOIRON, C.D.R., SVAASAND, L.O. and PROFIO,A.E.. Light dosimetry on tissues: application to photoradiation therapy. In: D. Kessel and T.J. Dougherty (eds.), Porphyrin sensitization, pp. 63-76, Plenum. New York (1983). DOIIGHERTY, T.J., Photosensitizers: therapy and detection of malignant tumors. Photochem. Photobiol., 45,879-889 (1987). DOUGLASS. H.O., JR., NAVA, H.R., WEISHAUPT,K.R., BOYLE,D., SUGGERMAN, M.G., HALPERN,E. and DOUGHERTY,T.J., Intra-

abdominal applications of hematoporphyrin photoradiation therapy. Arch. exp. Med. Biol., 160,15-21 (1981). FLNGAR, V.H. and HENDERSON, B.W.. Drug and light dose dependence of photodynamic therapy. A study of tumor and normal tissue response. Photochem. Photobiol., 46,837-841 (1987). FINGAR,V.H., POTTER,W.R. and HENDERSON, B.W., Drug and light dose dependence of photodynamic therapy: a study of tumor cell clonogenicity and histologic changes. Photochem. Photobiol., 5, 643650 (1987). FOGH,J., FOGH,J.M. and ORFEO, T., One hundred and twenty seven cultured human tumor cell lines producing tumors in nude mice.J. nut. Cancer Inst., 59,221-226 (1977). GOMER,C.J. and DOUGHERTY, T.J., Determination of (3H) and (14C) hematoporphyrin derivative distribution in malignant and normal tissue. CuncerRes.,39,146-151 (1979). GOMER,C.J., FERRARIO, A,, HAYASHI,N., RUCKER,N., SZIRTH,B.C. and MURPHREE, A.L., Molecular, cellular, and tissue responses following photodynamic therapy. Lasers S q .Med.. 8,450463 (1988). GOMER,C.J., JESTER,J.V., RAZUM,N.J., SZIRTH,B.C. and MURPHREE, A.L., Photodynamic therapy of intraocular tumors: examination of hematoporphyrin derivative distribution and long-term damage in rabbit ocular tissue. Cancer Res., 45,3718-3725 (1985). HENDERSON, B.W. and FINGAR, V.H., Oxygen limitation of direct tumor cell kill during photodynamic treatment of a murine tumor model. Photochetn. Photobiol., 49, 299-304 (1989). HERRERA-ORNELAS, L., PETRILLI, N.J., MITTELMAN,A. and DOUGHERTY, T.J., Photodynamic therapy in patients with colorectal cancer. Cancer, 57,677-684 (1986). HISAZUMI. M., NAITO,K., MISAKII, T., KOSHIDA,K. and YAMAMOTO,

498

LANTZ E T A L .

H., An experimental study of photodynamic therapy using a pulsed light-sensitive video intensification microscopy. Int. J. Cancer, 45, gold vapor laser. In: G. Jori and C. Perria (eds.). Photodynamic therapy 972-979 (1990). of tumors and other discuses, pp. 251-254, Libreria Progetto, Padua SILVERBERG, E. and LUBERA, J., Cancer statistics 1986. CA: Cancer J. (1985). Chi., 36,9-25 (1986). KESSEL,D., Porphyrin lipoprotein association as a factor in porphyrin SINDELAR, W.F., DELANEY, T.F., TOCHNER, Z., THOMAS, G.F., DACHOlocalization. Cancer Lett., 33, 183-188 (1986). SWKI, L.J., SMITH, P.D., FRIAUF, W.S., COLE,J.W. and GLATSTEIN, E., MANG,T.S. and WIEMAN, T.J.. Photodynamic therapy in the treatment Technique of photodynamic therapy for disseminated intraperitoneal of pancreatic carcinoma: dihematoporphyrin ether uptake and photomalignant neoplasms. Arch. Sue.., 126,318-324 (1991). bleaching kinetics. Photochem. Photobiol., 46,853-858 (1987). SUGARBAKER, P.H., GIANOLA, F.J. and DWYER, A., A simplified plan MOAN,J., Porphyrin photosensitization and phototherapy. Phoroclzem. for follow up of patients with colon and rectal cancer supported by Photobiol., 43,681-690 (1986). prospective studies of laboratory and radiologic test results. Surgery, MOERTEL,C.G., FLEMING, T.R., MACDONALD, J.S.. HALLER,D.G., 102,79-87 (1987). LAURIE,J.A., GOODMAN, P.J., UNGERLEIDER, J.S., EMERSON, W.A., SUZUKI, S., NAKAMURA, S. and SAKAGUCHI, S., Experimental study of TORMEY, D.C., GLICK,J.H., VEEDER,M.H. and MAILLIARD, J.A., intra-abdominal photodynamic therapy. Laser Med. Sci., 2, 195-203 Levamisol and 5FU for adjuvant therapy of resected colon carcinoma. (1987). N. Engl. J. Med., 322,352-358 (1990). L.O., Optical dosimetry for direct and interstitial photoraNAMBISAN, R.N., KAKAKOUSIS,C.P., HOLYOKE, E.D. and DOUGHERTY,SVAASAND, diation therapy of malignant tumors. In: D.R. Doiron and C.J. Comer T.J., lntraoperative photodynamic therapy for retroperitoneal sarco- (eds.), Porphyrin localization and treatment of tumors, pp. 91-114, A.R. mas. Cancer, 61,1248-1252 (1988). Liss, New York (1984). NELSON,J.S., LIAW,L-H.L. and BERNS,M.W., Tumor destruction in TOCHNER, Z . , MITCHELL, J.B., HARRINGTON, F.S., SMITH,P.. Russo, photodynamic therapy. Photochem. Photobiol.. 46,829-835 (1987). D.T. and Russo, A,. Treatment of murine intraperitoneal ovarian NELSON,J.S., WRIGHT,W.H. and BERNS,M.W., Histopathological ascitic tumor with hematoporphyrin derivative and laser light. Cancer comparison of the effects of hematoporphyrin derivative on two R ~ s 45,2983-2987 ., (1985). different murine tumors using computer-enhanced di ital video fluoVANGEMERT, J.C., BERENBAUM, M.C. and GIJSBERS, G.H.M., Wave rescence microscopy. Cancer Res., 45,5781-5786 (19857. length and light-dose dependence in tumor phototherapy with haemoPANTELIDES, M.L., MOORE,J.V. and BLACKLOCK, N.J., A comparison toporphyrin derivative. Brit. J. Cancer, 5 2 , 4 3 4 9 (1985). of serum kinetics and tissue distribution of Photofrin I1 following WALDOW,S.M. and DOUGHERTY, T.J., Interaction of hyperthermia intravenous and intraperitoneal injection in the mouse. Photochem. and photoradiation therapy. Radiat. Rex, 97,380-385 (1984). Photobiol., 69,67-70 (1989). T.J., MANG.T.S., FINGAR, B.S.. HILL,T.G.. REED,M.W.R., PENG.Q., MOAN,J. and CHENG. L.-S., The etfect of glucose administra- WIEMAN, tion on the uptake of Photofrin I1 in a human tumor xenograft. Cancer COREY,T.S., NGUYEN,V.Q. and RENDER,E.R., JR., Effect of photodynamic therapy on blood flow in normal and tumor vessels. Lett., 58,29-35 (1991). Sueery, 104,512-517 (1988). PENG.Q., NESLAND,J.M.. MOAN,J., EVENSEN, J.F., KONGSHAUG, M. and RiMiNGToN, C.. Localization of fluorescent Photofrin I1 and WILLETT,C.G., TEPPER,J.E. and COHENA,M., Failure patterns aluminum phthalocyanine tetrasulfonate in transplanted human malig- following curative resection of colonic carcinoma. Ann. Surg., 200, nant tumor LOX and normal tissues of nude mice using highly 685-690 (1984).