Effect of administration route and estrogen manipulation on endometrial uptake of Photofrin porfimer sodium

,

Photofrin, a complex mixture of dihematoporphyrin esters and ethers.Photofrin is the most widely used photosensitizer, although it is not always the most efficacious.Phthalocyanins and chlorins, among others, have shown considerable therapeutic promise with few side effects (i.e., reduced cutaneous phototoxicity) in experimental animal tumors."We selected Photofrin for the current work because it has been extensively characterized and it is being evaluated in multicenter human trials.In addition, the substantial effect Photofrin has on tumor microvasculature''' 7 suggests that it should be highly effective in the well-vascularized endometrium.
Currently, non-photodynamic therapy laser-induced tissue ablation is routinely performed on humans for a variety of conditions" 10 Drawbacks to these methods are that they require relatively high laser powers and have minimal tissue selectivity.In contrast, photodynamic therapy is a low-power, highly selective therapy.Several gynecologic applications have been described, II although to date it has been less extensively used on endometrial tissue.Promising results have been reported by Manyak et aI.' 2 who investigated photodynamic therapy of an endometriosis model, and Schneider et al., I" who observed tissue binding enhancement when intravenously administered photosensitizer is supplied with estradiol to ovariectomized rats.To better understand the determinants of tissue selectivity, we have systemically investigated the influence of estradiol and drug administration route on the distribution of Photofrin in rat uterine layers .Phase I of our studies evaluates the relative merits of intravenous, intraperitoneal, and intrauterine administration.In phase 2, we focus on the effect of estrogen on uptake and retention of the photosensitizer within uterine layer s.Our results indicate that excellent drug localization can be achieved 'at extremely low Photofrin doses, thus suggesting that photodynamic treatment of selected endometrial conditions, such as menorrhagia and dysfunctional uterine bleeding, may be performed with minimal cutaneous phototoxicity.

Material and methods
Phase 1: Effect of route of administration on dis.tribution.Eighty mature female Sprague-Dawley rats, weighing 263 to 330 gm, were placed in a control setting of 12 hours darkness followed by 12 hours light for I week.In this phase of the study all 80 rats were" suppressed with 0.05 rug/day of subcutaneous leupro-Iide acetate for 10 days (TAP, North Chicago) followed by intramuscular administration of 500/lJog estradiol valerate (Squibb, Princeton, N.J.) (Table I).Twenty-four hours later, 26 rats received 7 mg/kg Photofrin porfiiner sodium (QLT, Vancouver) in normal saline solution intravenously and 12 rats received the same dose intraperitoneally.Twenty-five rats received 0.7 mg/kg intrauterine Photofrin.Intrauterine injection was per-Formed through an open laparotomy, and Photofrin was placed in the right uterine ' horn: The skin was closed with staples, and all rats were housed individually in a dark room.Animals were killed 3, 6, 24, or 48 hours after Photofrin delivery.Seventeen rats that did not receive Photofrin served as controls.
Phase 2: Effect of estrogen on uptake and distribution ofPhotofrin.Forty mature female Sprague-Dawley rats were divided into four groups (Table I).One group was used as a control and was neither hormonally stimulated nor suppressed (group A).The second group (group B) was hormonally suppressed for 10 days with 0.05 mg/day of subcutaneous leuprolide acetate, followed by 500 IJog of intramuscular estradiol valerate and 24 hours later by Photofrin.The third group received only leuprolide acetate for 10 da ys before receiving Photofrin (group C).In the final group (group D) animals did not receive leupro1ide and were injected with estradiol valerate onl y, 24 hours before Photofrin administration.All rats received intrauterine Photofrin at a dose of 0.7 mg/kg, and all were killed by carbon dioxide inhalation 48 hours after Photofrin delivery.Within each group two rats were not given Photofrin and were used as baseline controls for each group.Intracardiac puncture and withdrawal of 3 to 5 ml of blood for estradiol levels were performed, followed immediately by surgical removal of the right uterine horn.
Vaginal smears (phase 1).Monitoring of estrogen suppression was achieved by means of vaginal smears.Rats were not injected with estradiol until after the smears were consistent with a hypoestrogenic state.Suppression was first noted after 5 days but was continued for 10 days to induce prolonged suppression.The 10day interval was also used in phase 2 of the experiment.
Serum estradiol levels (phase 2).Blood samples obtained by an intracardiac puncture were immediately placed in serum separation tubes and then frozen at -20°C.After all samples were obtained, serum levels of estradiol were determined by direct radioimmunoassay (Pantex, Santa Monica, Calif.).The minimum detectable estradiol concentration was 10 pg/ml.The intraassay and interassay variability was 6% ± I% and 10% ± 3%, respectively.Tissue extraction.Extraction of Photofrin from uterine tissue was conducted according to a modified porphyrin fecal extraction technique.14After the right uterine horn was dissected, the specimen for frozen section was removed and the remaining portion was placed in a closed opaque container and frozen at -70°C.Samples were lyo philized for 7 days , manually crushed, and weighed.The sa m p le s we re the n placed in a glass vortex tube, to whi ch I m l of 12 N hyd rochlo ric acid was added, and mixed for 5 min utes .To ensure maximal dissolution, the tu be was left standing for 45 to 50 min utes, with in ter mitt en t mix ing.Ethyl et her, 3 m l, was added and the sample was re mixed .After this, 3 ml of d oubl edi stilled wa ter was added , mixed, and cen trifuged for 10 minutes at 10,000 revolutions/mi n.The resultant mixture consisted of a top organic layer separated from an ac id -wa te r laye r by a thin lip id-tissue layer.The lower Photofrin-containing laye r was removed and analyzed for Photofri n content by means of either a bsorption or fluo res cence te ch n ique s.Absorbance me asu rements were reco r d e d a t the absorp tion maxim a (typica lly 405 to 407 nm) .T he emission wavelen g th p eak was measured at 610 n m after ex citation at 400 nm .Ph otofrin conten t was derived fr om calibration data and ex pr esse d as mi crogr a m s per gram dry weigh t of tissue.Drug recovery levels were d et ermined from spi ked tissu e sa m p le s to range from 8 3% to 90 %.
Frozen secti o ns.Tissue sa m p les used for fluoresce nce analysis were removed from the midportion of the o r igina l specimen and imm ediately placed in mo lds co nta in in g embedding med ium for frozen specimens (O CT , Miles, Elkhart, Ind .).The blocks were rapidl y fro zen o n dry ice and stored a t -70 0 C.All specimens were handled in the da rk .Tissu es were se ctioned in low diffu se lig ht (C ryosta t microtome, AO Reichert, Buffa lo, N.Y.), with slices 6 u rn th ick ta ke n from three loca tions, approximately 3 mm apart.
Frozen sections were analyzed for both histologic typ e and fluorescence , The cross section of the rat u terus was divi ded in to different layers for co m parative a na lysis.T he fir st laye r was th e surface gland ular endo metrial ce lls, and th e second laye r was th e underlying endometrial stromal cells .The third layer-was the myometrium, and the fina l layer was the serosa.
An e p ifluoresce nce microscope (Karl Zeiss, model RA., Oberkochen, Germany) was used for a ll tissue fluorescence stud ies.Fluorescence ex cit ation was provided by a 100 \V mercury arc la m p , and emission images were re corded by means of a low-ligh t-level video camera (Karl Zeiss, model 1V2M inten sified newvicon) .Video images were stored on one-half-inch VHS tape and a nalyzed for drug d istribution and intensity.To eva luate relative fluorescence intensity in " rea l time," images were scored on a scale of 0 to 4 where 0 equaled no fluorescence.These levels were assigned by systemically attenuati ng the excitation intensity with a series of neutral-density filters .Au tofluoresce nce values ob tai ned from drug-free control animals were subtra cted fro m ex pe rim en tal values to obtain a final fluore scence intensity score.T h is permitted a semiquantitative co m p ar iso n of both tissue di strib u tio n and amount ofPhotofr in.
Statistics.The sig nificance of differences was tested by means of the u npaired t te st and two-way analysis of variance .

Results
As shown in Fig.I , ati er Photofrin delivery the rel ative fluo rescence of the e nd o m e tr ial g lands gradua lly increased ove r time, regardle ss of route of administration .Endometrial fluores cence afte r in tr au ter ine administr ation was sign ificantly higher (p < 0 .05 ) than with the other two ro utes of inj ection, a nd this was maintained over time (Fig. 2) .As seen in Figs .3 and 4, relative fluorescence of th e endometrial stroma and myometrium a lso increased over time.There was no ap p are n t difference between stromal fluoresce nce, regardless of rou te of Photofrin administrat ion, over all time intervals (two-way analysis of variance, F 0.995, not significant).Myometrial fluorescence was significantly  higher after intraperitoneal injection of Photofrin (P < 0.05), at 3 hours and 6 hours, and after intravenous administration (p < 0.05) at 24 hours and 48 hours, as compared with the intrauterine route (Fig. 4).
As shown in Table II, total porphyrins extracted 3 and 48 hours after intrauterine drug administration (10.45 ± 16.1, 30.9 ± 20.1) was not significantly different (p > 0.2) than that of either the intravenous or intraperitoneal route (7.4 ± 9.61, 17.9 ± 18.5; 18.57 ± 7.93, 9.3 ± 8.11).The Photofrin dose used was tenfold higher in both the intraperitoneal and intravenous routes of delivery (7 vs 0.7 mg/kg intrauterine), such that intrauterine delivery resulted in the highest concentration of Photofrin absorbed per milligram of Photofrin.

Table IV. Extraction of total uterine porphyrins 48 hours after intrauterine Photofrin (0.7 mg/kg) administration
There was also a significant difference (p = 0.01) between the control group and the leuprolide-only (suppressed) group.There was no significant difference (p = 0.97) in the endometrial depth of the two estrogen-stimulated groups.
As seen in Fig. 5, the fluorescent intensity of the columnar epithelium after estrogen stimulation (group B 3.61 ± 0.55, group D 3.88 ± 0.23) is significantly higher (p > 0.05) than in the absence of estrogen (group A 2.62 ± .89, group C 2.7 ± 1.26).Prolonged leuprolide suppression did not inhibit uptake in the presence of estrogen.After estradiol stimulation there was no significant effect on uptake of Photofrin in the endometrial stroma layer (p > 0.05).In spite of this lack of statistical significance, the estrogen-only group had a more diffuse and homogeneous pattern of stromal uptake as compared with the sparse pattern of fluorescence in the other three groups.Comparing the two estrogen-stimulated groups, group D had a significantly higher (p = 0.04) uptake of Photofrin.All four 31 :±: 12.9 44.4 :±: 12.9 198.6:±: 118 groups consistently showed minimal fluorescence in the myometrium (Fig. 5).Estrogen had no effect on increasing or preventing uptake by the myometrium when there was no prolonged suppression (group A 0, group D 37 ± 0.88; P > 0.05).Estrogen stimulation after prolonged leuprolide suppression showed a significantly lower level of myometrial uptake, as compared with that of leuprolide suppression alone (group B 0.2 ± 0.25 vs group C 1.33 ± 1.29; P > 0.05), while increasing uptake within the endometrium (p = 0.04).
Extraction of total porphyrins from the rat uteri is seen in Table IV.Higher extracted levels of porphyrin corresponded to a relatively higher fluorescence.The only exception is in the group receiving both leuprolide acetate and estrogen stimulation.

Comment
Our purpose was to evaluate the effects of route of admininstration and estrogen manipulation on the uptake and distribution of a photosensitizer (Photofrin) in an estrogen-dependent tissue, to develop diagnostic and treatment modalities for endometrial pathologic conditions.Several delivery routes were compared to determine whether site-specific delivery would minimize the systemic side effects of Photofrin while maximizing endometrial uptake.Systemic application of Photofrin inherently involves a higher level of possible adverse reactions, primarily skin photosensitivity.In addition, because the drug distribution predominantly limited to the endometrium (glands and stromal tissue) is desired, an effective delivery system was needed.
After determination of the most effective route, estrogen manipulation was evaluated to determine if the hormone could be used to selectively increase Photofrin uptake or retention within uterine layers.Of concern here is whether states of high proliferative activity are more likely to retain the photosensitizer.Potential recipients of this therapeutic approach would be women with menorrhagia, dysfunctional uterine bleeding, or other endometrial disorders.Because of the high tissue destruction specificity characterized by photodynamic therapy, an endometrial application might replace other surgical approaches.
In phase I of the experiment, all animals received the gonadotropin-releasing hormone analog leuprolide acetate, followed by estradiol.This was done to suppress endogenous estrogen secretion and to synchronize the hormonal status of the experimental subjects, followed by endometrial proliferation and neovascularization at the time of photosensitization.Systemically administered (intravenous) Photofrin proved to be the least effective on a per-weight Photofrin-absorbed basis (Table II).Furthermore, intravenously injected Photofrin showed diffuse fluorescence within the uterus, regardless of time killed (Figs. 1 to 4).In spite of the fact that fluorescence increased over time, intravenous injection of the photosensitizer did not appear to promote endometrial selectivity.Myometrial fluorescence uptake (Fig. 4, intravenous injection) was not significantly different (P > 0.05) from endometrial uptake, and this lack of significance persisted over time.Higher myometrial fluorescence was observed when all other routes were compared with intrauterine delivery (P > 0.05).
Intraperitoneally administered Photofrin resulted in a definite pattern of uptake and redistribution within the uterus, as well as a higher overall Photofrin uptake than with intravenous Photofrin (Figs.I, 3, and 4; Table II).This trend suggested that initially there was a high concentration of the drug in the serosa (data not shown); however, as time elapsed, the fluorescence shifted toward the endometrium.Again, myometrial uptake and retention persisted up to 48 hours, although it was not significantly higher (P > 0.05) than intrauterine delivery.It is not clear whether this redistribution was caused by diffusion or absorption into the vascular system and subsequent redistribution.Intrauterine delivery of the photosensitizer appeared to allow for more selective retention within the surface endometrial cells over all time intervals (Fig. I, P < 0.05) and minimized myometrial uptake (Fig. 4, P < 0.05).On the basis of fluorescent intensity, the drug remained within the surface endometrial glands with limited diffusion into the deeper stromal layers.Uptake by the endometrial stroma was not significantly different at 48 hours as compared with intravenous administration.However, the relative distribution favors Chapman at al. 691 uptake within the endometrium with limited uptake by the myometrium.It appears the elevated mitotic activity and increased protein production within the surface endometrial cells (glandular) and deeper stromal cells increased the concentration and retention of the drug by cells in these two layers.Finally, in spite of a tenfold reduction in dose, intrauterine injection yielded a significant increase in extracted Photofrin, lending support to the hypothesis that site-specific delivery of the photosensitizer can achieve selective retention of the drug at a much reduced dose (Table II).The distribution of fluorescence after intravenous injection is in partial agreement with previous work by Schneider et al.;" who followed intravenously administered iodine 125-labeled dihematoporphyrin ether in both estrogen-primed and non-estrogen-primed ovariectomized rats.Pharmacokinetic differences may be caused, in part, by our use of Photofrin rather than 1125-labeled dihematoporphyrin ether.
In phase 2 of the experiment, on the basis of the above findings, all rats received intrauterine photosensitizer.Fluorescent activity within the surface endometrial glands is most prominent in the estradiol-stimulated rats (Fig. 5).There is some fluorescence in the deeper stromal cells, with all groups showing some pockets of bright fluorescence (P < 0.05 for group D only).However, except for a more homogeneous distribution and slightly more intense fluorescence in the estrogen-only group, there is no significant difference.Photofrin, however, tended to be excluded from the myometrial layer especially after estrogen stimulation, and this appears to be due to the presence of an active, thicker endometrial layer.In the Ieuprolide-only group (group C) the endometrial depth is significantly reduced (Table Ill, p < 0.05), which may account for the higher myometrial uptake in this group.There are possible explanations for the prolonged retention of Photofrin within the epithelium.As a relatively hydrophobic compound, once it is inside the epithelial cellular lining, it binds to the metabolically active components within the cytosol (recently stimulated production by estradiol).It is unclear whether estradiol affects the initial uptake.Schneider et al.I" noted the distribution of photosensitizer to be similar regardless of estrogen status but that the intensity of the fluorescence was greater in the estrogen-stimulated group.This could account for the higher Photofrin levels after estrogen stimulation.It appears the estradiol effect is indirect and secondary to the end-organ effect.It may also be a diffusion defect, because Photofrin appears not to traverse the tight gap junctions between the epithelial lining.The uptake within the endometrium itself would be greater if this were the case.Most likely, with time the Photofrin will diffuse out of the surface endometrial cells and enter the deeper stromal layer.If the endo-metrium has been recently stimulated by estradiol, stromal cells and endometrial glands will exhibit prolonged binding of the Photofrin.The distribution pattern of fluorescence within the endometrial stroma may reflect products produced from the breakdown of Photofrin within the surface endometrial cells that diffuse into the stroma, while still retaining their fluorescent photodynamic properties.
Although estradiol stimulation yielded equivocal increases in fluorescence in the endometrial layer (columnar epithelium and stroma), it did increase the overall amount of porphyrins retained within the uterus, indicative of greater photosensitizer uptake.This model system can be used to further the study of a sitedelivered system for photodynamic therapy.First, estradiol appears to promote the uptake and retention within the surface endometrial glands.When the photosensitizer enters the deeper endometrium, it appears to be retained longer.Second, this model illustrated that site delivery was successful.This study, in addition to encouraging further preclinical studies with intrauterine site delivery of light, may lead to the clinical application of photodynamic therapy for endometrial ablation.

Fig. 2 . 8 .
Fig. 2. Fluorescence micrographs of rat uteri after intrauterine administration of Photofrin.A, Collapsed surface endometrial layer 3 hours after injection.Band C, Fluorescence at 6 and 24 hours after injection.Uterine cavity distended because of estrogen stimulated secretions.D, Background fluorescence of deeper layers at 48 hours after injection.

Table I .
Randomization characteristics for Sprague-Dawley rats *Two rats served as internal controls and were not given Photofrin.