Radiosensitization of a mouse tumor model by sustained intra-tumoral release of Etanidazole and Tirapazamine using a biodegradable polymer implant device

Radiotherapy and Oncology 53 (1999) 77±84 www.elsevier.com/locate/radonline

Radiosensitization of a mouse tumor model by sustained intra-tumoral release of Etanidazole and Tirapazamine using a biodegradable polymer implant device Donald T.T. Yapp a, David K. Lloyd b, Julian Zhu c, Shirley Lehnert a,* a

Department of Radiation Oncology, McGill University, MontreÂal, Canada b Dupont-Merck, Wilmington, DE, USA c Department of Chemistry, Universite de MontreÂal, MontreÂal, Canada

Received 6 July 1998; received in revised form 8 July 1998; accepted 27 August 1999

Abstract Background and purpose: Drug toxicities are often a limiting factor in long term treatment regimes used in conjunction with radiotherapy. If the drug could be localized to the tumor site and released slowly, then optimal, intra-tumoral drug concentrations could be achieved without the cumulative toxicity associated with repeated systemic drug dosage. In this paper we describe the use of a biodegradable polymer implant for sustained intra-tumoral release of high concentrations of drugs targeting hypoxic cells. Materials and methods: The RIF-1 tumor was implanted subcutaneously or rntramuscularly in C3H mice and irradiated with 60Co gamma rays. The drug delivery device was the co-polymer CPP-SA;20:80 into which the drug was homogeneously incorporated. The hypoxic radiosensitizer Etanidazole or the bioreductive drug Tirapazamine were delivered intra-tumorally by means of implanted polymer rods containing the drugs. Tumor growth delay (TGD) was used as the end point in these experiments. Results: Both Etanidazole and Tirapazamine potentiated the effects of acute and fractionated radiation in the intra-muscular tumors but neither drug was effective in sub-cutaneous tumors. Since both drugs target hypoxic cells we hypothesized that the lack of effect in the subcutaneous tumor was attributable to the smaller size of the hypoxic fraction in this tumor model. This was con®rmed using the hypoxia marker EF5. Conclusions: These results indicate that the biodegradable polymer implant is an effective vehicle for the intra-tumoral delivery of Etanidazole and Tirapazamine and that, in conjunction with radiation, this approach could improve treatment outcome in tumors which contain a sub-population of hypoxic, radioresistant cells. q 1999 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Biodegradable polymer; Intra-tumoral drug delivery; Etanidazole; Tirapazamine; Radiation

1. Introduction The toxicities of drugs used in chemotherapy or in conjunction with radiotherapy are frequently the dose-limiting factor. In many cases, this precludes the use of optimal drug doses required for tumor control or the use of multiple drug doses during fractionated radiation treatment over a prolonged period of time. If, however the drug could be localized at the disease site and released slowly over time, then systemic drug toxicities could be decreased while simultaneously increasing drug concentrations at the tumor site. There is thus a role for a sustained release intra-tumoral drug delivery system in mixed modality therapy. * Corresponding author. Radiation Oncology, McGill University, MontreÂal General Hospital, 1650 Cedar Ave., MontreÂal, Quebec, Canada HG3 1A4.

A number of polymer or gel based systems have been used to deliver chemotherapeutic drugs intra-tumorally in experimental tumor models. Hydrogels, which release drug as a function of water uptake by the gel have been used to deliver cisplatin directly to the tumor in a mouse model [13,14] and injectable gels containing ¯uorouracil, cisplatin or doxorubicin have been used to treat human tumor xenografts in mice [29]. A number of studies have been done of delivery of cisplatin by OPLA-Pt, an open cell polylactic acid polymer. Implantation of this device adjacent to a murine mammary tumor produced a greater than additive response, in terms of growth delay, when combined with radiation [15]. The same device, used for the treatment of spontaneous tumors in dogs, alone or in combination with radiation, showed improved response over conventional treatment [31,33]. A biodegradable copolymer implant of bis(p-carboxyphenoxy)propane-sebacic

0167-8140/99/$ - see front matter q 1999 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0167-814 0(99)00123-1

78

D.T.T. Yapp et al. / Radiotherapy and Oncology 53 (1999) 77±84

acid (CPP:SA;20:80), initially developed by Langer and coworkers [22] which has good drug release characteristics and benign degradation products [24] has been tested in a number of experimental studies. Phase I±III clinical trials using this polymer system demonstrated that 3.8% w/w BCNU incorporated into CPP:SA is safe and effective for treatment of patients with recurrent malignant glioma [3±5]. We have reported the use of the same polymer, CPP:SA;20:80, for intra-tumoral delivery cisplatin in a mouse tumor model (RIF-1). The concentration of drug in the tumor was enhanced and tumor growth signi®cantly delayed when drug delivery was by cisplatin/polymer implants compared to that seen when mice were treated with cisplatin administered by injection or by osmotic pump [36]. Administration of cisplatin by the polymer implant was found to potentiate the effects of acute and fractionated radiation [34±36]. The positive results observed with the cisplatin-polymer implants led us to initiate studies with compounds which are tumor-speci®c by virtue of the fact that they target hypoxic cells and, given an effective delivery system, can potentiate the effect of radiation. Etanidazole, a nitroimidazole hypoxic cell radiosensitizer, has been extensively studied experimentally and in clinical trials in conjunction with fractionated radiation. The results were disappointing partly due to systemic toxicities which develop when the drug was administered over a long treatment schedule [12,17]. In a recent review, Saunders [26] cited results of a meta-analysis of 83 clinical trials by Overgaard [25] and concluded that the overall bene®t of the nitro-imidazoles is small and does not justify their application in radiotherapy although there is evidence to support their use in certain speci®c situations. Interestingly, intra-tumoral injections of nitroimidazoles have been found to be effective as radiosensitizers and promising results have been obtained using this approach in the treatment of advanced disease of the oral cavity [26], bladder tumors [1,2] and carcinoma of the cervix [16]. In view of these ®ndings, and of the unsatisfactory results which have been obtained with systemic administration of nitroimidazoles it would appear that delivery of Etanidazole by biodegradable polymer could be the perfect scenario in which to test the ef®cacy of these radiosensitizers since the drug can be maintained in the tumor at a high level for a prolonged period of time with minimal systemic side effects. Tirapazamine is a bioreductive drug (a compound which is reduced in the cell to an active cytotoxic agent) which shows preferential toxicity to hypoxic cells [37] and is currently in clinical trials in combination with radiotherapy and with cisplatin [28]. The hypoxic cytotoxicity of Tirapazamine results from activation by reductive enzymes to produce a radical species causing DNA damage and chromosome aberrations [9]. Because of tumor speci®c hypoxia the combination of Tirapazamine and radiation treatment would be predicted to be bene®cial since Tirapazamine would kill radioresistant hypoxic cells complementing the

radiation killing of the aerated tumor cells [10] and preclinical studies have shown that Tirapazamine potentiates radiation-induced cell killing in transplanted mouse tumors [8]. The same rationale which supports the use of a sustained release system for the delivery of Etanidazole is applicable to Tirapazamine. Moreover, since the drug which is selectively toxic for hypoxic cells, is continuously present at high concentration, any tumor cells going through a period of hypoxia, however transient, would be targeted. In this paper we report on the effect of continuous delivery of Etanidazole and Tirapazamine used in conjunction with radiation on the RIF-1 mouse tumor implanted at subcutaneous or intra-muscular sites. Our results indicate that the biodegradable polymer is an effective delivery system for these two drugs and that this approach could be used to boost the effectiveness of treatment in cases where hypoxic cells limit the response to radiotherapy and systemic toxicity restricts the use of radiosensitizing drugs which target hypoxic cells. 2. Materials and methods 2.1. Radiation The tumors were irradiated with 60Co gamma rays using a Theratron 780 unit at a dose rate of approximately 1.0 Gy/ min. The mice were anesthetized and only the tumor was irradiated, the rest of the body being shielded with 5 cm thick lead blocks. A growth delay modi®cation factor (GDMF) was estimated following the method published by Kirichenko et al. [20] for calculation of a dose modifying factor. Brie¯y, two variables, AGD and NDG were de®ned as follows: ² AGD is the number of days for tumors, treated with drug or radiation, to reach the endpoint minus the number of days for control, untreated tumor to reach the same endpoint. ² NGD is the number of days for tumors treated with both radiation and drug to reach the endpoint minus the number days for tumor treated with drug alone to reach the endpoint. The GDMF is expressed as the ratio of NGD (of tumors receiving both radiation and drug) to AGD (tumors receiving radiation only). Ninety-®ve percent con®dence intervals were calculated for the GDMFs using Fieller's theorem, assuming that the AGDs and NGDs were normally distributed. 2.2. Polymer implant The CPP-SA;20:80 copolymer was synthesized according to published procedures [24]. Etanidazole or Tirapazamine (30 and 15% w/w, respectively) were ground together with the polymer to form a homogenous mixture, heated to

D.T.T. Yapp et al. / Radiotherapy and Oncology 53 (1999) 77±84

808C and extruded through an Eppendorf Combitip. The resulting drug-polymer rods were cooled at room temperature and stored in a dessicator until required. 2.3. Tissue culture RIF-1 cells were obtained from Dr R. Hill (Ontario Cancer Institute) were passaged using standard tissue culture techniques in alpha modi®cation of MEM medium supplemented with 10% fetal bovine serum and 1% antibiotic-mycotic (all supplied by Gibco-BRL). A 50 ml cell suspension (4 £ 106 cells/ml) was injected subcutaneously into the backs or intra-muscularly in the upper leg of female C3H mice (weight, 20±25 g). Tumors appeared within 10 days and reached a volume of 94±139 mm 3 within 3 weeks. Tumor volumes were calculated from measurements at 3 orthogonal angles using the formula …a £ b £ c†p=6. 2.4. Treatments Treatments were begun when the tumors reached a volume of approximately 100 mm 3. Tumor-bearing mice were separated into groups of 5±6 for different treatments. For implant of polymer rods the skin was punctured with a 20 gauge hypodermic needle, and the polymer rod inserted into the tumor. Generally, the rod (8 £ 0:5 mm) was divided into three pieces and inserted in three different positions. Fractionated doses (5 £ 6 Gy) were delivered at 24 h intervals for 5 days beginning on the same day as the insertion of the polymer rod. Acute doses (16.7 Gy) were delivered 48 h after implant. Control mice were sham irradiated. Tumor measurements were made three times per week, and the mice were sacri®ced when the endpoint (4 £ initial tumor volume) was reached. The signi®cance of the difference between TGDs for different experimental groups was determined using Fisher's two tailed t test.

79

2.5. Detection of hypoxic cells EF5, a ¯uorinated derivative of Etanidazole, generously supplied by Dr Cameron Koch, was used for evaluation of hypxic cells in vivo. Tumors were excised from anesthetized mice 3 h after injection of EF5 and rapidly cooled in ice-cold PBS. A single cell suspension was prepared by mincing and enzymatic disaggregation. Cells were ®xed and stained with the ELK3-51 antibody conjugated with the ¯uorescent dye Cy-3 as described by Koch et al. [21] and analyzed by ¯ow cytometry using a Becton Dickinson FacStar Instrument with excitation at 514 nm. To determine the response of a population of hypoxic cells RIF-1 monolayers were covered with 100 mM EF5 in 1.0 ml medium and incubated for 3 h in sealed aluminum chambers ¯ushed with 95% N2±5%CO2, conditions which had been previously shown to produce radiobiological hypoxia [23]. At the end of the incubation the monolayer was washed, suspended by trypsinization and processed for ¯ow cytometry as described above. 3. Results Tumor growth delay was used as an endpoint to assess the impact of the treatment on the tumors. The results for subcutaneous and intramuscular tumors are shown in Table 1 and Fig. lA,B. Growth delay modifying factors were also calculated to assess the extent the potentiation of the radiation treatment by the drug and these values are shown in Table 2. 3.1. Effect of radiation alone For untreated tumors there was a small but signi®cant difference in growth rate between subcutaneous and intramuscular tumors, the subcutaneous model taking a shorter

Table 1 Tumor growth delay (TGD) of sub-cutaneous and intramuscular RIF-1 tumors irradiated with or without drug delivered by polymer implant a Radiation dose (Gy)

Drug

Sub-cutaneous 1 2 3 4 5 6 7 8 9

0 16.7 5£6 0 16.7 5£6 0 16.7 5£6 a

± ± ± Eta. Eta. Eta. Tira. Tira Tira

P-value b

Tumor model

5.7 ^ 0.7 16.0 ^ 2.1 23.0 ^ l.5 7.2 ^ 1.9 16.6 ^ 1.6 25.8 ^ 3.7 7.0 ^ 0.2 15.8 ^ 1.9 26.8 ^ 3.9

c

Intra-muscular 7.2 ^ 0.8 11.0 ^ l.2 18.8 ^ 2.1 7.0 ^ 1.0 19.0 e ^ 2.5 23.4 f ^ 2.7 8.0 ^ 1.3 18.3 e ^ 1.0 24.4 f ^ 0.9

0.006 d 0.0005 d 0.0026 d n.s 0.076 (n.s) n.s 0.092 (n.s) 0.017 g n.s

Abbreviations: Eta, Etanidazole; Tira, Tirapazamine; N.s., difference not signi®cant. P-value for difference between TGDs of subcutaneous and intramuscular tumors receiving the same treatment. c TGD ^ standard deviation (days). d TGD subcutaneous . TGD intramuscular. e Signi®cantly different from control (line 2). f Signi®cantly different from control (line 3). g TGD subcutaneous , TGD intramuscular. b

80

D.T.T. Yapp et al. / Radiotherapy and Oncology 53 (1999) 77±84

Fig. 1. Tumor growth delay in RIF-1 tumors treated with polymer implants containing Etanidazole or Tirapazamine combined with acute or fractionated Radiation. (A) Subcutaneous tumors. (B) Intramuscular tumors.

time to reach the endpoint of four times the initial volume. Following acute or fractionated radiation however the TGD for the intramuscular tumor was less than that for the subcutaneous tumor indicating that the intramuscular model was more radioresistant. The difference between the TGDs for Table 2 Growth delay modi®cation factors for tumors implanted with polymer incorporating Etanidazole or Tirapazamine Tumor site

Radiation dose (Gy)

Drug

GDMF a

Subcutaneous

16.7 5£6 16.7 5£6 16.7 5£6 16.7 5£6

Etanidazole Etanidazole Tirapazamine Tirapazamine Etanidazole Etanidazole Tirapazamine Tirapazamine

0.91 (0.69±1.19) 1.08 (0.86±1.31) 0.85 (0.62±1.13) 1.14 (0.95±1.34) 3.15 (2.89±4.21) 1.43 (1.24±1.66) 2.96 (1.98±4.18) 1.51 (1.13±1.78)

Intramuscular

a

Numbers in parentheses indicate 95% con®dence intervals of GDMFs.

the two tumor models for both acute and fractionated radiation was highly signi®cant (Table 1). 3.2. Effect of polymer/drug implants Implantation of blank (no drug) polymer did not change the TGD of untreated tumors (results not shown). Polymer/ drug implant caused a small increase in TGD for both drugs in subcutaneous tumors and for Tirapazamine in intramuscular tumors but in neither case was this statistically significant. 3.3. Etanidazole implant plus radiation For subcutaneous tumors polymer/Etanidazole implant had little effect on the response of the tumor to radiation. In the same tumor model, when fractionated radiation was combined with polymer implant the TGD was longer than that for radiation alone but the difference was not statistically signi®cant. In contrast, for intramuscular tumors

D.T.T. Yapp et al. / Radiotherapy and Oncology 53 (1999) 77±84

81

implantation of Etanidazole/polymer potentiated the effect of radiation treatment to a signi®cant extent. The effect was greatest for an acute dose of 16.7 Gy where the TGD increased from 11 days for radiation alone to 19 days for radiation plus polymer implant. Treatment of intramuscular tumors with fractionated (5 £ 6 Gy) radiation increased the TGD from 18.8 days for radiation alone to 23.4 days when polymer/Etanidazole was implanted at the start of fractionated treatment. Thus, for Etanidazole-implanted tumors there is no difference between the radiation response of subcutaneous and intra-muscular tumors. This is in contrast to tumors given radiation only where the TGDs for the subcutaneous tumors are signi®cantly shorter (Table 1). 3.4. Tirapazamine implant plus radiation For subcutaneous tumors combined treatment with Tirapazamine/polymer and acute radiation produced no change of TGD from the response to radiation alone. In the same tumor model, for fractionated treatment, there was a small increase in TGD when radiation was combined with polymer implant which was not quite statistically signi®cant. The response of intramuscular tumors was affected by polymer/Tirapazamine implant to a much greater extent. The greatest increase in TGD (from 11 to 18 days) was seen when acute radiation was combined with polymer implant. When fractionated radiation was combined with polymer implant the TGD was increased from 18.8 to 23 days. There is a signi®cant difference between the response of Tirapazamine-implanted subcutaneous and intramuscular tumors to fractionated radiation but, in this case, the TGDs for the intramuscular tumors are longer than those for the subcutaneous tumors, the reverse of what is seen for radiation alone. For both drugs the radiation effect for intramuscular tumors was potentiated most when a single acute dose of 16.7 Gy was used with GDMFs of 3.15 and 2.96 for Etanidazole and Tirapazamine, respectively (Table 2). When fractionated radiation was used, the estimated GDMFs were about half those observed with the acute dose; 1.43 and 1.51 for Etanidazole and Tirapazamine, respectively. Thus a signi®cant growth delay modifying effect is seen for intramuscular tumors implanted with polymer to deliver Etanidazole or Tirapazamine. These ®ndings are summarized in Fig. lA,B. 3.5. Measurement of tumor hypoxic fraction In Fig. 2A,B representative pro®les for EF-5 binding in sub-cutaneous and intramuscular RIF-1 tumors of the same volume (l00 mm 3) are shown. Based on measurements made in vitro a gate was set such that the mean ¯uorescence intensity (indicative of the level of FF5 binding) of the gated subpopulation corresponded to that of the in vitro hypoxic population. This value was approximately 25 times the mean ¯uorescent intensity associated with non-

Fig. 2. Flow cytometric analysis of cells from RIF-1 tumors. The unshaded pro®le on the left (No. FF5) is a control for non-speci®c binding of the ELK3-51 antibody. The unshaded pro®le on the right (Hypoxic) is for cells incubated in vitro under 95% N2±5% CO2. The shaded pro®le is tumor cells dissociated 3 h after administration of FF5. The horizontal line indicates the dimensions of the gate de®ning the hypoxic cell population. The vertical arrows indicate the positions of the mean ¯uorescence intensity for each population. (A) Subcutaneous tumor. (B) Intramuscular tumor.

speci®c antibody binding. The proportion of tumor cells which was gated was taken to represent the hypoxic fraction of the tumor. The hypoxic fraction for subcutaneous tumors obtained by this method was 4:3 ^ 1:4% (mean ^ SE) while that for the intramuscular tumors was 11:7 ^ 3:2%, the difference between the two tumor models was just signi®cant (P ˆ 0:05). There were four tumors in each experimental group. The distribution of FF5 binding in the populations suggested that for both tumor models there were cells at an intermediate level of hypoxia. The relative importance of these was determined by calculating a relative ¯uorescent intensity for the total population which is the mean ¯uorescence intensity for the tumor cell population relative to the ¯uorescence intensity attributable to nonspeci®c binding [28]. For subcutaneous tumors the mean ¯uorescent intensity was 2:0 ^ 0:5 while that for intramuscular tumors was 3:4 ^ 07 the difference between the two was signi®cant (P ˆ 0:02). In summary, while hypoxic and partially hypoxic cells are detectable in both tumor models the proportion of cells in both categories is greater for intramuscular tumors.

82

D.T.T. Yapp et al. / Radiotherapy and Oncology 53 (1999) 77±84

4. Discussion Our results indicate that radiosensitization of intramuscular RIF-1 tumors occurred when Etanidazole or Tirapazamine was delivered using the biodegradable polymer. For subcutaneous tumors there was a potentiation of the effect of fractionated radiation by implanted polymer/Tirapazamine which was not quite statistically signi®cant, otherwise for this tumor model, radiation treatment combined with either drug was no more effective than radiation alone. In contrast, intramuscular tumors, which are initially more radioresistant than are subcutaneous tumors were radiosensitized by both Etanidazole and Tirapazamine. One objective of drug delivery by intra-tumoral implant is to ensure that a high level of drug is maintained in the tumor over a prolonged period. In a study in which a cisplatinloaded polymer was implanted in the RIF-1 tumor using the same conditions as those in the present experiments we found that the amount of drug released into tumor tissue was about 4% of that incorporated in the polymer when the tumor was sampled at 1, 2 and 4 days after implant, rose to approximately 10-fold that amount at 8 days after implant and had returned to the 4% level by 12 days after implant [36]. If we use these ®ndings as a basis for the interpretation of results obtained with other drugs, the level of Etanidazole which would be present in the tumor over a period of 12 days after implant would range from 0.05±0.5 mg/g tumor tissue. Administered systemically, 0.43 mg/g Etanidazole signi®cantly reduces the TCD50 dose of acute radiation for the MDAH-MCa-4 tumor in mice [32]. The daily dose of systemic Etanidazole to cause neuropathy is 0.6 mg/g [7]. Drug delivery by polymer implant thus maintains, over a 12 day period, a tumor drug concentration which, is high enough to enhance radiation response but which, if given systemically, would be in the range to cause toxicity. The acute LD50 for Tirapazamine is 0.09 mg/g [28]. Again on the basis of the results obtained with cisplatin we project that the level of Tirapazamine maintained in the tumor for 12 days after polymer implant will be in the range 0.024±0.240 mg/g. This level of Tirapazamine is selectively toxic to hypoxic cells and, if continuously present in the tumor would be expected to result in the complete elimination of hypoxic cells. 0.011 mg/g Tirapazamine given systemically produces a large enhancement of fractionated radiation in the mouse SCCVII carcinoma [28] while 0.027 mg/g daily for 8 days is the LD50 for continuous dosing with Tirapazamine [30]. Thus the intra-tumoral drug concentration resulting from Tirapazamine/polymer implant is high enough to signi®cantly in¯uence tumor response to therapy while the same intra-tumoral concentration, if it were induced by systemic administration, would cause severe toxicity. A recent study by Cardinale et al. [11] investigated the feasibility of administering tirapazamine by a slow-releasing polymer disc implanted interstitially into a U251

(human glioblastoma) xenograft. Combined intraperitoneal drug administration (0.0l4 mg=kg £ 6) and tumor drug implant (2 mg) signi®cantly delayed tumor growth when combined with 12 Gy radiation in 2 Gy fractions compared to radiation alone. This result was not achieved when either drug treatment alone was combined with radiation. In spite of the level of availability of the drugs in tumor tissue, the effect of radiation was not potentiated for subcutaneous tumors by drug/tumor implant except for the small effect seen for polymer/Tirapazamine and fractionated radiation. There are several possible explanations of this. In the case of Etanidazole the effect of the radiosensitizer would not be apparent if reoxygenation between fractions was ef®cient. While this might explain the results for fractionated radiation, a similar type of response was seen for acute radiation suggesting that reoxygenation was not a factor. In the case of Tirapazamine the toxicity towards hypoxic cells is dependent on the availability of enzymes which will reduce the pro-drug to a toxic product Since the same tumor, implanted in a different site, does show a response it seems unlikely that it is lack of reductive enzymes which is causing the subcutaneous RIF-1 to be impervious to Tirapazamine. In addition, RIF-1 tumor cells have been shown elsewhere to be sensitive to Tirapazamine under hypoxic conditions [28]. Both drugs affect the hypoxic fraction of tumors acting as a radiosensitizer or cytotoxin. The fact that the two drugs behave in similar fashion, potentiating the effect of radiation in intramuscular tumors while being ineffective in subcutaneous tumors suggests that the difference in response is related to the size of the hypoxic fraction in the two models. As noted above the longer TGD for irradiated intramuscular tumors indicates the greater radioresistance of this model. The implication of the ®nding that the radiation response of the sub-cutaneous RIF-1 tumor is not potentiated by either a hypoxic cytotoxin nor a hypoxic radiosensitizer is that the hypoxic fraction of this tumor is too small to make a difference to the overall response. This would be consistent with some of the earlier measurements of the hypoxic fraction of this tumor growing in sub-cutaneous sites [6]. It is not possible to draw any ®rm conclusion as to the oxygenation status of the RIF-1 tumor on the basis of published ®ndings. The earliest measurements of the hypoxic fraction of the RIF-1 tumor, using the paired survival curve method, found that 1.5% of cells of the subcutaneous tumor showed maximum radioresistance while, for the intramuscular tumor, the majority of cells were at an intermediate level of oxygenation [6] and the intramuscular tumors were sensitized to radiation by misonidazole. Also using the paired survival curve method, Shibamoto et al. found a hypoxic fraction of 4.7% for the sub-cutaneous RIF-1 model [27]. A recent study by Kavanagh and coworkers used three methods to determine the hypoxic fraction of intramuscular RIF-1 tumors. The result obtained by the paired survival curve method was 54% (45±65%) and while measurements made with the eppendorf pO2 histo-

D.T.T. Yapp et al. / Radiotherapy and Oncology 53 (1999) 77±84

graph did not correlate with these results they were supported by results of an assay involving binding of 3H misonidazole [19]. An earlier study from the same laboratory used the comet assay to determine the hypoxic fraction of the intra-muscular RIF-1 and reported a value of 15% [18]. The results described here indicate that, as predicted on the basis of the drug response, the level of oxygenation differs between the two tumor models with the relative mean ¯uorescence, indicative of EF5 binding for intramuscular tumors being greater than that for subcutaneous tumors (2.0 compared with 3.4). The fraction of cells de®ned as acutely hypoxic is also greater in the intramuscular RIF-1 tumors than in the subcutaneous tumor model (11% compared with 4.4%). These results support the hypothesis that subcutaneous RIF-1 tumors are not sensitized to radiation by Etanidazole or Tirapazamine because of the low level of hypoxia in these tumors. There is however, a detectable level of hypoxic cells in subcutaneous tumors so either this was not suf®cient to in¯uence the response to radiation combined with drug/polymer treatment or other factors were involved. The results reported here con®rm our prediction that the biodegradable polymer implant is an effective vehicle for the intra-tumoral delivery of Etanidazole and Tirapazamine. In conjunction with radiation this approach improves treatment outcome in tumors which contain hypoxic radioresistant cells. Future studies will focus on de®ning the relationship between the tumor microenvironment and the kinetics of drug delivery by implanted polymers. Acknowledgements This work was supported by the National Cancer Institute of Canada. We wish to thank Sano® Research for generously providing Tirapazamine for this study and the Drug Synthesis and Chemistry Branch NCI. for supplying Etanidazole. We are grateful to Dr Cameron Koch for supplying the reagents and advice for measuring tumor hypoxia using the EF5 binding method. References [1] Awwad HK, Abd el Moneim H, Abd el Baki H, Omar S, Farag H. The topical use of misonidazole. 13th International Cancer Congress, Part D, New York: Alan R. Liss Inc, 1983. pp. 303±306. [2] Balmukanov SB, Beisebaev AA, Aitkoolovo ZI, et al. Intratumoural and parametrial infusion of metronidazole in the radiotherapy of uterine cervix cancer: preliminary report. Int. J. Rad. Oncol. Biol. Phys. 1989;16:1061±1063. [3] Brem H, Mahaley S, Vick NA, et al. Interstitial chemotherapy with drug polymer implants for the treatment of recurrent gliomas. J. Neurosurg. 1991;74:441±446. [4] Brem H, Piantadosi S, Burger PC. The safety of BCNU-loaded chemotherapy with BCNU-loaded polymer followed by radiation therapy in the treatment of newly diagnosed malignant gliomas. J. Neuro-Oncol. 1995;26:111±123.

83

[5] Brem H, Piantadosi S, Burger PC. Placebo-controlled trial of safety and ef®cacy of intra-operative controlled delivery by biodegradable polymers of chemotherapy for recurrent gliomas. Lancet 1995;345:1008±1012. [6] Brown JM, Twentyman PR, Zamvil SS. Response of the RIF-1 tumor in vitro and in C3H/Km mice to X-radiation (cell survival, regrowth delay and tumor control), chemotherapeutic agents and activated macrophages. J. Natl. Cancer Inst 1980;64:605±611. [7] Brown JM. Clinical perspectives for the use of new hypoxic cell sensitizers. Int. J. Radiat Oncol. Biol. Phys. 1982;8:1491±1497. [8] Brown JM, Lemmon U. Potentiation by the hypoxic cytotoxin SR 4233 of cell killing produced by fractionated irradiation of mouse tumors. Cancer Res. 1990;50:7745±7749. [9] Brown JM. SR 4233 (Tirapazamine): a new anti-cancer drug exploiting hypoxia in solid tumors. Br. J. Cancer 1993;67:1163±1170. [10] Brown JM, Sum BG. Hypoxia-Speci®c Cytotoxins in Cancer Therapy. Semin. Radiat Oncol. 1996;6:22±36. [11] Cardinale RM, Dillehay LE, Williams JA, Tabassi K, Brem H, Lee DJ. Effect of interstitial and/or systemic delivery of tirapazamine on the radiosensitivity of human glioblastoma multiforme in nude mice. Rad. Oncol. Invest. 1998;6:63±70. [12] Coleman CN, Noll L, Riese N, et al. Final Report of the Phase I trial of continuous infusion etanidazole (SR-2508): a Radiation Therapy Oncology Group Study. Int J. Radiat Oncol. Biol. Phys. 1992;22:577±580. [13] Deurbo MIM, Bohlken D, Kop W, Lerk CF, Hennink W, Bartelink H, Begg AC. Intratumoral administration of cisplatin in slow release devices. Cancer Chemother. Pharmacol. 1990;27:135±140. [14] Deurbo MJM, Kop W, van Tellingen O, Bartelink H, Begg AC. Intratumoral administration of cisplatin in slow release devices II Pharmacokinetics and intratumoural distribution. Cancer Chemother. Pharmacol. 1991;27:347±353. [15] Douple EB, Xi FX, Yang LX, Brekke I. Potentiation of radiotherapy by cisplatin released via biodgradable implants (OPLA-Pt) in a murine solid tumor. Radiation Research Society 40th Annual Meeting, Salt Lake City, March, Abstract 1992;P1:15±17. [16] Garcia-Angulo AH. Radiosensitizer metronidazole plus standard radiotherapy for advanced cervical carcinoma. In: Sugahara T, editor. Hyperthermic oncology Vol.1, London: Philadelphia: Taylor and Francis, 1988. pp. 633±636. [17] Halberg FE, Cosmatis D, Gunderson LL, et al. A phase I study to evaiuate intra-operative radiation therapy and the hypoxic cell sensitizer etanidazole in locally advanced malignancies. Int. I. Rad. Oncol. Biol. Phys. 1994;28:201±206. [18] Hu Q, Kavanagh M-C, Newcombe D, Hill RP. Detection of hypoxic fraction in murine tumors by comet assay: Comparison with other techniques. Radiat. Res. 1995;144:266±275. [19] Kavanagh M-C, Sun A, Hu Q, Hill RP. Comparing techniques of measuring tumor hypoxia in different murine tumors: Eppendorf pO2 histograph, [3H] Misonidazole binding and paired survival assay. Radiat. Res 1996;145:491±500. [20] Kirichenko AV, Rich TA, Newman RA, Travis EL. Potentiation of murine Mca-4 carcinoma response by 9-amino-20(s)-camptothecin. Cancer Res. 1997;57:1929±1933. [21] Koch CJ, Evans SM, Lord EM. Oxygen dependence of cellular uptake of EF5 [2-(2-nitro-1H-imidazol-1-yl)-N-(2,2,3,3,3-penta¯uorylpropyl)acetamide]: analysis of drug adducts by ¯uorescent antibodies versus bound radioactivity. Br. J. Cancer 1995;72:865±870. [22] Langer R. New methods of drug delivery. Science 1990;249:1527± 1533. [23] Lehnert S, Greene G, Batist G. Radiation response of a drug-resistant variants of a human breast cancer cell line: effect of glutathione depletion. Radiat. Res. 1990;124:208±215. [24] Leong KW, Brott BC, Langer R. Bioerodible polyanhydrides as drugcarrier matrices I: characterization, degradation, and release characteristics. J. Biomed. Mater. Res. 1985;19:941±955. [25] Overgaard J. Modi®cation of hypoxia from Gottwald Schwarz to

84

[26]

[27] [28] [29]

[30] [31]

D.T.T. Yapp et al. / Radiotherapy and Oncology 53 (1999) 77±84 nicotinamide: have we learned the lesson? Danish Cancer Society. In: Kogelnik HD, editor. Progress in radiation oncology, Bologna, Italy: Monduzzi Editore, 1995. pp. 269±275. Saunders M, Dische S. Clinical results of hypoxic cell radiosensitization from hyperbaric oxygen to accelerated radiotherapy, carbogen and nicotinamide. Br. J. Cancer 1996;74(Suppl. XXVII):S271± S278. Shibamoto Y, Yukawa K, Tsutsui K, Takahashi M, Abe M. Variations in the hypoxic fraction among mouse tumors of different types sizes and sites. Jpn. J. Cancer Res. 1986;77:908±915. Sum BG, Menke DR, Done MJ, Brown JM. Tirapazamine-induced cytotoxicity and DNA damage in transplanted tumors: relationship to tumor hypoxia. Cancer Res. 1997;57:2922±2928. Smith JP, Stock F, Orenberg EK, Yu NY, Kanekal S, Brown DM. Intratumoral chemotherapy with sustained release drug delivery system inhibits growth of human pancreatic xenografts. AntiCancer Drug. 1995;6:717±726. Spiegel JF, Spear MA, Brown JM. Toxicology of daily administration to mice of the radiation potentiator SR 4233 (W1N59075). Radiother. Oncol. 1993;26:79±81. Straw RC, Withrow SI, Douple EB. Effects of cisdiamminedichlor-

[32] [33] [34]

[35] [36]

[37]

oplatinum (II) released from d,l-polylactic acid implanted adjacent to cortical allografts in dogs. J. Orthopaedic Res. 1993;12:871±877. Stone HB, Sinesi M. Testing of new hypoxic cell sensitizers in vivo. Radiat. Res. 1982;91:186±198. Withrow SJ, Straw RC, Brekke JH. Slow release adjuvant of cisplatin in slow release devices. Fur. J. Exp. Muskoskelet Res. 1995;4:105± 110. Yapp DTT, Lloyd DK, Zhu J, Lehnert SM. Tumor treatment by sustained intra-tumoral release of cisplatin: effects of drug alone and combined with radiation. Int. J. Rad. Oncol. Biol. Phys. 1997;39:497±504. Yapp DTT, Lloyd DK, Zhu I, Lehnert SM. The potentiation of the effect of radiation treatment by intra-tumoral delivery of cisplatin. Int. J. Radiat Oncol. Biol. Phys. 1998;42:413±420. Yapp DTT, Lloyd DK, Zhu J, Lehnert SM. Cisplatin delivery by biodegradable polymer implant, systemic injection and osmotic pump: differences in selected tissue distribution in C3H mice bearing subcutaneous RIF-1 Tumors. Anti-Cancer Drugs 1998;9:786±791. Zeman EM, Brown DM, Lemmon LM, Hirst VK, Lee WW. SR 4233: a new bioreductive agent with high selective toxicity for hypoxic mammalian cells. Int J. Rad. Oncol. Biol. Phys. 1986;12:1239±1242.