- Email: [email protected]
Modification of alkylating agent induced cell kill by 2-nitroimidazoles in unclamped and clamped sc 9L tumors
Inr. / Radralrnn Oncologic Bid Ph:, Vol. 22, pp. 727-730 Pnnted ,n the USA All rights reserved.
??Session E: Bioreductive
Copyright
0360.3016192 $5.00 + .OO C 1992 Pergamon Press plc
Therapies
MODIFICATION OF ALKYLATING AGENT INDUCED CELL KILL BY 2-NITROIMIDAZOLES IN UNCLAMPED AND CLAMPED SC 9L TUMORS C. ANNE
WALLEN, Experimental
PH.D.,
SARAH
K. MOORE,
B.S.
AND KENNETH
T. WHEELER,
PH.D.
Radiation Oncology, Department of Radiology, Bowman Gray School of Medicine of Wake Forest University, Winston-Salem, NC 27 157
Enhanced cell kill has been observed when experimental tumors were treated with alkylating agents in combination with 2-nitroimidazoles (2-NI). In this study, modification of the cell kill induced by cyclophosphamide (CY) and an analog, ifosfamide (IFO), by two radiation sensitizers, misonidazole (MISO) and etanidazole (SR-2508), was measured. Three important parameters were determined: (a) the necessity for hypoxic reduction of the 2-NI to achieve an increase in tumor cell kill, (b) the optimal timing for administration of the alkylating agents and the 2NI, and (c) the degree of enhancement of the CY- and IFO-induced cell kill. The subcutaneous (SC) 9L tumor model in male Fisher 344 rats was used in these experiments, and the endpoint measured was clonogenic cell survival 1820 hr after treatment. Under hypoxic conditions, MIS0 potentiated both CY- and IFO-induced cell kill with a sensitizer enhancement ratio of approximately 1.3 and 1.5, respectively, at the 10m3survival level. This enhancement was seen when CY was administered simultaneously or 1.5 hr prior to MIS0 administration. A similar enhancement of CY-induced cell kill was measured under hypoxic conditions when SR-2508 was used. Enhanced IFO-induced cell kill was measured under hypoxic conditions only when the IF0 was given 1 hr before MIS0 administration. No enhancement of the IFO-induced cell kill was observed when SR-2508 was used instead of MISO. Increased normal tissue damage (i.e., hemorrhagic cystitis) was observed when the MIS0 was administered along with CY or IFO. Four conclusions can be drawn from these data. Metabolism of the 2-NI by hypoxic cells is necessary for potentiation of CY- or IFO-induced cell kill. Only MIS0 can potentiate the cell kill induced by IFO. The timing of administration of the alkylating agents and the 2-NI is a critical determinant of the extent of the cell kill obtained. Cell kill induced by IF0 appears to be enhanced by MIS0 to a greater extent than the cell kill induced by CY. Ifosfamide, Chemosensitization,
2-nitroimidazoles,
SC 9L tumor, Cyclophosphamide.
ways. For example, the position of the two functional groups relative to one another are different in the two compounds, the pharmacokinetics of activation of IF0 is much slower than that of CY, and IF0 has a higher therapeutic index than CY (1). However, there are no available studies to indicate if the 2-nitroimidazoles (2NI) can modify the cell kill induced by IFO. In this study, the subcutaneous (SC) 9L tumor model was used to determine if IFO-induced cell kill can be increased by administration of either MIS0 or etanidazole (SR-2508) to a greater extent than the cell kill induced by CY. Three important parameters were determined: (a) the necessity for hypoxic reduction of the 2-NI to achieve an increase in tumor cell kill, (b) the optimal timing for
INTRODUCTION
Cyclophosphamide (CY) has consistently been shown to be a drug whose cell kill can be enhanced by administration of misonidazole (MISO) (10, 11). Clinical trials to take advantage of this chemosensitization have been suggested for tumors that are likely to contain a significant hypoxic cell population (10, 1 I), such as small cell lung carcinoma (2,4). In our institution, the chemotherapeutic agent of choice for treating small cell lung carcinoma is ifosfamide (IFO). IF0 and CY are alkylating oxazaphosphorines that must be metabolized by the liver. Although chemically the two drugs differ only in the position of one chloroethyl group, they are quite different in other
Presented at the Seventh International Conference on Chemical Modifiers of Cancer Treatment, in Clearwater, FL, 2-5 February 1991. Reprint requests to: C. Anne Wallen, Department of Radiology, Bowman Gray School of Medicine, Medical Center Boulevard, Winston-Salem, NC 27 157. ,4cknowlrdgements-The authors thank N. R. Lomax of the
drugs used in these experiments, and Dr. T. M. Morgan from the Biostatistics Unit of the Cancer Center of Wake Forest University for his help with the statistical analysis. We also thank Dr. R. Tang for help with some of the experiments, J. Anderson for technical support, and P. Cregger for preparation of this manuscript. This research was supported by grant RD-302 from the American Cancer Society. Accepted for publication 26 July 1991.
Drug Synthesis and Chemistry Branch, Division of Cancer Treatment at the National Cancer Institute for supplying the 727
728
I. J.
Radiation
Oncology
0 Biology
0 Physics
Volume
22, Number
4. 1992
administration of the alkylating agents and the 2-NI, and (c) the degree of enhancement of the CY- and IFO-induced cell kill. METHODS
AND MATERIALS
Tumor model The procedures for implanting 9L tumor cells in male Fisher 344 rats have been described previously (12, 13). The average tumor weight in these experiments was x500 mg and no tumor weighed > 1 g. These SC 9L tumors contain essentially no radiobiologically hypoxic cells, but a transient hypoxia can be produced by clamping the blood supply to the tumors for up to 2 hr with no irreversible effects on the tumor vasculature and no reduction in the colony forming efficiency (13, 15). This model allows one to measure the interaction of the alkylating agents and the 2-NI in situ under relatively uniform oxic and hypoxic conditions ( 12- 14).
0
Drug administration
Clonogenic assay the endpoint
40
60
CYCLOPHOSPHAMIDE
The drugs used in this study were supplied by the Drug Synthesis and Chemistry Branch of the Division of Cancer Treatment at the National Cancer Institute. The alkylating agents were stored at -70°C and the 2-NI at 4°C. All of the drugs were dissolved in saline before i.p. injection. Sterile saline was administered with single agent treatments so that the total volume injected per animal was held constant per kilogram of body weight. The MIS0 dose was always 2.5 mmole/kg, whereas the SR-2508 dose was always 3.75 mmole/kg. The MIS0 and SR-2508 pharmacokinetics in this model have been published (14, 15). All rats were anesthetized with -60 mg/kg of pentobarbital prior to the start of the treatment protocol. Body temperature was maintained at 35-37°C by placing the rats under heat lamps ( 14, 15). The 2-NI was given either 2.5 hr before, simultaneously, or l-l.5 hr after the alkylating agent. The 9L SC tumors were made hypoxic by clamping the blood vessel(s) supplying the tumor with a Edslab double softjaw handless clamp, as described previously ( 13, 14). The clamps when used were always applied 30 min after administration of the 2-NI and remained in place for 2 hr. Full dose response curves were obtained for both CY and IF0 for each protocol.
In this study,
20
was clonogenic
survival
of
9L tumor cells 1S-20 hr after treatment. The colony formation assay was performed as described previously ( 12. 15). No cell kill was measured in unclamped or clamped SC 9L tumors treated with either 2-NI. The exponential portion of each survival curve was fitted by a least-squares linear regression analysis of the natural log of the surviving fraction as a function of the administered dose. Each sensitizer enhancement ratio (SER) was calculated from the
80
100
(mg/kg)
Fig. I. Cyclophosphamide-induced cell kill in SC 9L tumors. Unclamped tumors treated with CY alone, unclamped tumors treated with CY and a 2-NI. or clamped tumors treated with CY (0). Clamped tumors treated with CY and MIS0 given simultaneously (A), CY and MIS0 given I hr apart (0) or CY and SR-2508 given l- 1.5 hr apart (V). The lines were fitted by a least squares linear regression analysis. The points represent the geometric mean ? I SEM from 3-16 tumors assayed individually from at least two independent experiments. The error bars are smaller than the data points when not shown.
appropriate for equality
regression equations by an F-test.
whose slopes were tested
RESULTS Under hypoxic conditions, MIS0 potentiated CYinduced cell kill with a sensitizer enhancement ratio (SER) of approximately 1.3 (p < 10-3) at the 1O-3 survival level (Fig. 1). Although the SER of 1.3 was lower than the SER of 1.5-2.0 previously reported for murine tumors with hypoxic fractions < 1 .O, it is likely that the lower SER is caused by differences in the way rats and 9L cells handle or respond to these agents (6-9). In our study, the SER was identical (p > 0.2) when CY was administered either simultaneously or 1 hr prior to administration of the MIS0 (Fig. 1). An enhancement of CY-induced cell kill (SER = 1.3, p < 10m3) was also measured under hypoxic conditions with SR-2508 (Fig. 1). No potentiation of CYinduced cell kill was observed when the MIS0 was given prior to the clamping period and CY was administered immediately after release of the clamp (data not shown). Under hypoxic conditions, MIS0 also potentiated IFOinduced cell kill. However, the SER at the lop3 survival level was 4 Cp < 10P3) when compared to IF0 alone, and
Potentiation
of ifosfamide
IO"
5
16'
5 $
16'
0 Z ;
10”
5 u-l
16"
1 IV
n5 0
50 IFOSFAMIDE
100
150
(mg/kg)
Fig. 2. Ifosfamide-induced cell kill in SC9L tumors. Unclamped tumors treated with IF0 alone, unclamped tumors treated with IF0 and a 2-NI, or tumors clamped for 2 hr and then treated with IF0 (0). Tumors clamped 1.5 hr after IF0 alone (A). Clamped tumors treated with IF0 and MIS0 given 1 hr apart (V), or IF0 and SR-2508 (0) given 1 hr apart. The exponential portion of the curves were fitted by a least squares linear regression analysis. The points represent the geometric mean of 4- 16 tumors assayed individually from at least three independent experiments. The SEMs are lo-20% and are not plotted for clarity.
I.5 (p < 0.05) when compared to tumors clamped 1.5 hr after IF0 administration (Fig. 2). Enhanced IFO-induced cell kill 0, < 0.05) was measured in clamped tumors only when the IF0 was given 1 hr before MIS0 administration (Fig. 2). No enhancement of the IFO-induced cell kill 0, > 0.5) was observed when SR-2508 was used instead of MIS0 in any of the protocols (Fig. 2). The increased tumor cell kill in these experiments was accompanied by an increase in toxicity (i.e., hemorrhagic cystitis and death). The threshold dose for observing the hemorrhagic cystitis was lower in those animals with clamped or unclamped SC 9L tumors treated with MIS0 in combination with either of the alkylating agents. This is somewhat surprising because in unclamped tumors the combination of MIS0 and IF0 or CY killed no more tumor cells than the alkylating agent alone.
DISCUSSION Both IFO- and CY-induced cell kill can be enhanced by MIS0 in the 9L tumor model (Figs. I and 2). As in most studies to date, the presence of hypoxic cells was required to obtain this enhanced cell kill. IF0 was potentiated only by MISO, whereas CY was potentiated equally
0 C. A. WALLEN e/ al.
729
by MIS0 and SR-2508. This equal enhancement of CYinduced cell kill by these 2-NI has been observed previously in the RIF-I tumor model (7, 8) but not in all tumors (6, 9). These observations suggest that the chemosensitizing activity of SR-2508 may be limited to a few tumors (10, 1 I). IFO-induced cell kill was altered by clamping the tumor (Fig. 2). When IF0 and MIS0 were combined, a large portion of the increased cell kill (SER = 4) was a reflection of this phenomenon. The exact nature of this enhancement is unknown at the present time, but may reflect the pharmacokinetics ofthe toxic products of IF0 and/or the interaction of these products with hypoxic cells. For example, the liver metabohtes of IF0 may be very toxic to hypoxic 9L cells when these metabolites reach the tumor. Alternatively, further reduction of the liver metabolites by hypoxic 9L cells may be required to generate the cytotoxic moiety. To obtain chemosensitization when IF0 and MIS0 were combined required that IF0 be given at least I hr before MISO. The chemosensitization of CY required only that CY be given before or at the same time as MIS0 or SR-2508. This schedule dependence is reversed from that determined previously for the nitrosoureas ( IO, I I). This difference probably stems from the site of “activation” of the drugs: intercellularly for the nitrosoureas or in the liver for IF0 and CY. From our data (Figs. 1 and 2) we suggest that IF0 or CY be given prior to MIS0 to enhance their cell kill. This recommended order of administration is reversed from that used in the recent clinical trials of CY and MIS0 (3, 5). This may in part explain the failure of these clinical trials. When the magnitude of the potentiation of CY- and IFO-induced cell kill are compared, IF0 was potentiated slightly more than CY. These data imply that the use of IF0 in a clinical chemosensitization trial in small cell lung cancer is appropriate. The increased toxicity of IF0 and CY combined with MIS0 is unlikely to be troublesome in the clinic. Mesna has been shown to be extremely effective in reducing the hemorrhagic cystitis induced by IF0 and CY in humans (I) and is likely to do the same for the combination of MIS0 with these alkylating agents. In summary, four conclusions can be drawn from the data in this study. As expected, metabolism of the 2-NI by hypoxic cells is necessary for potentiation of the alkylating agent-induced cell kill. Although both MIS0 and SR-2508 potentiated the SC 9L cell kill induced by CY, only MIS0 potentiated the cell kill induced by IFO. The timing of administration of the alkylating agents and the 2-NI is a critical determinant of the extent of the cell kill obtained. The cell kill induced by IF0 appears to be modified by MIS0 to a slightly greater extent than the cell kill induced by CY. However, as a cautionary note, normal tissue toxicity may also be enhanced, thereby making the therapeutic gain smaller than the SER reported on the basis of cell kill.
730
I. J. Radiation
Oncology
0 Biology 0 Physics
Volume 22. Number
4, I992
REFERENCES 1 Brade, W. P.; Nagel, G. A.; Seeber, S., eds. Ifosfamide in tumor therapy. Munich, Germany: Karger, 1987. 2. Chapman, J. D.; Urtasun, R. D.; Franko, A. J.; Raleigh. J. A.; Meeker, B. E.; McKinnon, S. A. The measurement of oxygenation status of individual tumors. In: Palliwal, B. R., Fowler, J. F., Hebert, D. E., Kinsella, T. J., Orton, C. G., eds. Prediction of response in radiation therapy (Part I): the physical and biological basis. NY: Am. Inst. of Physicists in Medicine; 1989:49-60. 3. Davila, E.; Klein, L.; Vogel, C. L.; Johnson, R.; Ostroy, F.; Browning, S.; Gorowski, E.. Furrier, R. L.; Presant, C. A. Phase I trial of misonidazole (NSCX261037) plus cyclophosphamide in solid tumors. J. Clin. Oncol. 3: 12 1-127; 1985. 4. Gatenby, R. A.; Kessler, H. B.; Rosenblum, J. S.; Coia, L. R.; Modldofsky. P. T.; Hartz, W. H.; Broden, G. J. Oxygen distribution in squamous cell carcinoma metastases and its relationship to outcome of radiation therapy. Int. J. Radiat. Oncol. Biol. Phys. 14:831-838; 1988. 5. Glover, D.; Trump, D.; Kvols, L.; Elson. P.: Vogl, S. Phase II trial of misonidazole (MISO) and cyclophosphamide (CYC) in metastatic renal cell carcinoma. Int. J. Radiat. Oncol. Biol. Phys. 12: I405- 1408; 1986. 6. Hinchliffe, M.; McNally. N. J. A comparison of the ability of some radiosensitizers undergoing clinical trials to act as chemosensitizers. Int. J. Radiat. Oncol. Biol. Phys. 10: 16351640: 1984. 7. Hirst, D. G.; Hazlehurst. J. L.; Brown, J. M. Sensitization of normal and malignant tissues to cyclophosphamide by nitroimidazoles with different partition coefficients. Br. J. Cancer 49:33-42; 1984.
8. Law, M. P.: Hirst. D. G.: Brown, J. M. Enhancing effect of misonidazole on the response of the RIF- 1 tumor to cyclophosphamide. Br. J. Cancer 44:208-2 18: I98 I. 9. McNally, N. J.; Hinchliffe. M.; de Ronde, J. Enhancement of the action of alkylating agents by single high, or chronic low doses ofmisonidazole. Br. J. Cancer 48:27 I-278; 1983. IO. Mulcahy. R. T. Update: the potential for chemosensitization of alkylating agents by nitroimidazoles. Oncology 12: I727: 1988.
I I. Mulcahy,
R. T.; Trump, D. L. Clinical chemosensitization by misonidazole and related compounds, a critical evaluation. J. Clin. Oncol. 6:569-573; 1988.
12. Wallen. C. A.; Michaelson,
S. M.; Wheeler, K. T. Evidence of an unconventional radiosensitivity of 9L subcutaneous tumors. Radiat. Res. 84:529-54 I ; 1980.
13. Wheeler, K. T.; Wallen, C. A.; Wolf, K. L.; Siemann. D. W. Hypoxic cells and in situ chemopotentiation of the nitrosoureas by misonidazole. Br. J. Cancer 49:787-793; 1984. 14. Wong. K.-H.; Wallen, C. A.; Wheeler, K. T. Biodistribution of misonidazole and 1.3 bis(2-chloroethyl)1-nitrosourea (BCNU) in rats bearing unclamped and clamped 9L subcutaneous tumors. Int. J. Radiat. Oncol. Biol. Phys. 17: I35143: 1989. 15. Wong, K.-H.: Wallen, C. A.: Wheeler, K. T. Chemosensitization of the nitrosourea by 2-nitroimidazoles in the subcutaneous 9L tumor model: pharmacokinetic and structureactivity considerations. Int. J. Radiat. Oncol. Biol. Phys. 18: 1043-1050: 1990.
??Session E: Bioreductive
Copyright
0360.3016192 $5.00 + .OO C 1992 Pergamon Press plc
Therapies
MODIFICATION OF ALKYLATING AGENT INDUCED CELL KILL BY 2-NITROIMIDAZOLES IN UNCLAMPED AND CLAMPED SC 9L TUMORS C. ANNE
WALLEN, Experimental
PH.D.,
SARAH
K. MOORE,
B.S.
AND KENNETH
T. WHEELER,
PH.D.
Radiation Oncology, Department of Radiology, Bowman Gray School of Medicine of Wake Forest University, Winston-Salem, NC 27 157
Enhanced cell kill has been observed when experimental tumors were treated with alkylating agents in combination with 2-nitroimidazoles (2-NI). In this study, modification of the cell kill induced by cyclophosphamide (CY) and an analog, ifosfamide (IFO), by two radiation sensitizers, misonidazole (MISO) and etanidazole (SR-2508), was measured. Three important parameters were determined: (a) the necessity for hypoxic reduction of the 2-NI to achieve an increase in tumor cell kill, (b) the optimal timing for administration of the alkylating agents and the 2NI, and (c) the degree of enhancement of the CY- and IFO-induced cell kill. The subcutaneous (SC) 9L tumor model in male Fisher 344 rats was used in these experiments, and the endpoint measured was clonogenic cell survival 1820 hr after treatment. Under hypoxic conditions, MIS0 potentiated both CY- and IFO-induced cell kill with a sensitizer enhancement ratio of approximately 1.3 and 1.5, respectively, at the 10m3survival level. This enhancement was seen when CY was administered simultaneously or 1.5 hr prior to MIS0 administration. A similar enhancement of CY-induced cell kill was measured under hypoxic conditions when SR-2508 was used. Enhanced IFO-induced cell kill was measured under hypoxic conditions only when the IF0 was given 1 hr before MIS0 administration. No enhancement of the IFO-induced cell kill was observed when SR-2508 was used instead of MISO. Increased normal tissue damage (i.e., hemorrhagic cystitis) was observed when the MIS0 was administered along with CY or IFO. Four conclusions can be drawn from these data. Metabolism of the 2-NI by hypoxic cells is necessary for potentiation of CY- or IFO-induced cell kill. Only MIS0 can potentiate the cell kill induced by IFO. The timing of administration of the alkylating agents and the 2-NI is a critical determinant of the extent of the cell kill obtained. Cell kill induced by IF0 appears to be enhanced by MIS0 to a greater extent than the cell kill induced by CY. Ifosfamide, Chemosensitization,
2-nitroimidazoles,
SC 9L tumor, Cyclophosphamide.
ways. For example, the position of the two functional groups relative to one another are different in the two compounds, the pharmacokinetics of activation of IF0 is much slower than that of CY, and IF0 has a higher therapeutic index than CY (1). However, there are no available studies to indicate if the 2-nitroimidazoles (2NI) can modify the cell kill induced by IFO. In this study, the subcutaneous (SC) 9L tumor model was used to determine if IFO-induced cell kill can be increased by administration of either MIS0 or etanidazole (SR-2508) to a greater extent than the cell kill induced by CY. Three important parameters were determined: (a) the necessity for hypoxic reduction of the 2-NI to achieve an increase in tumor cell kill, (b) the optimal timing for
INTRODUCTION
Cyclophosphamide (CY) has consistently been shown to be a drug whose cell kill can be enhanced by administration of misonidazole (MISO) (10, 11). Clinical trials to take advantage of this chemosensitization have been suggested for tumors that are likely to contain a significant hypoxic cell population (10, 1 I), such as small cell lung carcinoma (2,4). In our institution, the chemotherapeutic agent of choice for treating small cell lung carcinoma is ifosfamide (IFO). IF0 and CY are alkylating oxazaphosphorines that must be metabolized by the liver. Although chemically the two drugs differ only in the position of one chloroethyl group, they are quite different in other
Presented at the Seventh International Conference on Chemical Modifiers of Cancer Treatment, in Clearwater, FL, 2-5 February 1991. Reprint requests to: C. Anne Wallen, Department of Radiology, Bowman Gray School of Medicine, Medical Center Boulevard, Winston-Salem, NC 27 157. ,4cknowlrdgements-The authors thank N. R. Lomax of the
drugs used in these experiments, and Dr. T. M. Morgan from the Biostatistics Unit of the Cancer Center of Wake Forest University for his help with the statistical analysis. We also thank Dr. R. Tang for help with some of the experiments, J. Anderson for technical support, and P. Cregger for preparation of this manuscript. This research was supported by grant RD-302 from the American Cancer Society. Accepted for publication 26 July 1991.
Drug Synthesis and Chemistry Branch, Division of Cancer Treatment at the National Cancer Institute for supplying the 727
728
I. J.
Radiation
Oncology
0 Biology
0 Physics
Volume
22, Number
4. 1992
administration of the alkylating agents and the 2-NI, and (c) the degree of enhancement of the CY- and IFO-induced cell kill. METHODS
AND MATERIALS
Tumor model The procedures for implanting 9L tumor cells in male Fisher 344 rats have been described previously (12, 13). The average tumor weight in these experiments was x500 mg and no tumor weighed > 1 g. These SC 9L tumors contain essentially no radiobiologically hypoxic cells, but a transient hypoxia can be produced by clamping the blood supply to the tumors for up to 2 hr with no irreversible effects on the tumor vasculature and no reduction in the colony forming efficiency (13, 15). This model allows one to measure the interaction of the alkylating agents and the 2-NI in situ under relatively uniform oxic and hypoxic conditions ( 12- 14).
0
Drug administration
Clonogenic assay the endpoint
40
60
CYCLOPHOSPHAMIDE
The drugs used in this study were supplied by the Drug Synthesis and Chemistry Branch of the Division of Cancer Treatment at the National Cancer Institute. The alkylating agents were stored at -70°C and the 2-NI at 4°C. All of the drugs were dissolved in saline before i.p. injection. Sterile saline was administered with single agent treatments so that the total volume injected per animal was held constant per kilogram of body weight. The MIS0 dose was always 2.5 mmole/kg, whereas the SR-2508 dose was always 3.75 mmole/kg. The MIS0 and SR-2508 pharmacokinetics in this model have been published (14, 15). All rats were anesthetized with -60 mg/kg of pentobarbital prior to the start of the treatment protocol. Body temperature was maintained at 35-37°C by placing the rats under heat lamps ( 14, 15). The 2-NI was given either 2.5 hr before, simultaneously, or l-l.5 hr after the alkylating agent. The 9L SC tumors were made hypoxic by clamping the blood vessel(s) supplying the tumor with a Edslab double softjaw handless clamp, as described previously ( 13, 14). The clamps when used were always applied 30 min after administration of the 2-NI and remained in place for 2 hr. Full dose response curves were obtained for both CY and IF0 for each protocol.
In this study,
20
was clonogenic
survival
of
9L tumor cells 1S-20 hr after treatment. The colony formation assay was performed as described previously ( 12. 15). No cell kill was measured in unclamped or clamped SC 9L tumors treated with either 2-NI. The exponential portion of each survival curve was fitted by a least-squares linear regression analysis of the natural log of the surviving fraction as a function of the administered dose. Each sensitizer enhancement ratio (SER) was calculated from the
80
100
(mg/kg)
Fig. I. Cyclophosphamide-induced cell kill in SC 9L tumors. Unclamped tumors treated with CY alone, unclamped tumors treated with CY and a 2-NI. or clamped tumors treated with CY (0). Clamped tumors treated with CY and MIS0 given simultaneously (A), CY and MIS0 given I hr apart (0) or CY and SR-2508 given l- 1.5 hr apart (V). The lines were fitted by a least squares linear regression analysis. The points represent the geometric mean ? I SEM from 3-16 tumors assayed individually from at least two independent experiments. The error bars are smaller than the data points when not shown.
appropriate for equality
regression equations by an F-test.
whose slopes were tested
RESULTS Under hypoxic conditions, MIS0 potentiated CYinduced cell kill with a sensitizer enhancement ratio (SER) of approximately 1.3 (p < 10-3) at the 1O-3 survival level (Fig. 1). Although the SER of 1.3 was lower than the SER of 1.5-2.0 previously reported for murine tumors with hypoxic fractions < 1 .O, it is likely that the lower SER is caused by differences in the way rats and 9L cells handle or respond to these agents (6-9). In our study, the SER was identical (p > 0.2) when CY was administered either simultaneously or 1 hr prior to administration of the MIS0 (Fig. 1). An enhancement of CY-induced cell kill (SER = 1.3, p < 10m3) was also measured under hypoxic conditions with SR-2508 (Fig. 1). No potentiation of CYinduced cell kill was observed when the MIS0 was given prior to the clamping period and CY was administered immediately after release of the clamp (data not shown). Under hypoxic conditions, MIS0 also potentiated IFOinduced cell kill. However, the SER at the lop3 survival level was 4 Cp < 10P3) when compared to IF0 alone, and
Potentiation
of ifosfamide
IO"
5
16'
5 $
16'
0 Z ;
10”
5 u-l
16"
1 IV
n5 0
50 IFOSFAMIDE
100
150
(mg/kg)
Fig. 2. Ifosfamide-induced cell kill in SC9L tumors. Unclamped tumors treated with IF0 alone, unclamped tumors treated with IF0 and a 2-NI, or tumors clamped for 2 hr and then treated with IF0 (0). Tumors clamped 1.5 hr after IF0 alone (A). Clamped tumors treated with IF0 and MIS0 given 1 hr apart (V), or IF0 and SR-2508 (0) given 1 hr apart. The exponential portion of the curves were fitted by a least squares linear regression analysis. The points represent the geometric mean of 4- 16 tumors assayed individually from at least three independent experiments. The SEMs are lo-20% and are not plotted for clarity.
I.5 (p < 0.05) when compared to tumors clamped 1.5 hr after IF0 administration (Fig. 2). Enhanced IFO-induced cell kill 0, < 0.05) was measured in clamped tumors only when the IF0 was given 1 hr before MIS0 administration (Fig. 2). No enhancement of the IFO-induced cell kill 0, > 0.5) was observed when SR-2508 was used instead of MIS0 in any of the protocols (Fig. 2). The increased tumor cell kill in these experiments was accompanied by an increase in toxicity (i.e., hemorrhagic cystitis and death). The threshold dose for observing the hemorrhagic cystitis was lower in those animals with clamped or unclamped SC 9L tumors treated with MIS0 in combination with either of the alkylating agents. This is somewhat surprising because in unclamped tumors the combination of MIS0 and IF0 or CY killed no more tumor cells than the alkylating agent alone.
DISCUSSION Both IFO- and CY-induced cell kill can be enhanced by MIS0 in the 9L tumor model (Figs. I and 2). As in most studies to date, the presence of hypoxic cells was required to obtain this enhanced cell kill. IF0 was potentiated only by MISO, whereas CY was potentiated equally
0 C. A. WALLEN e/ al.
729
by MIS0 and SR-2508. This equal enhancement of CYinduced cell kill by these 2-NI has been observed previously in the RIF-I tumor model (7, 8) but not in all tumors (6, 9). These observations suggest that the chemosensitizing activity of SR-2508 may be limited to a few tumors (10, 1 I). IFO-induced cell kill was altered by clamping the tumor (Fig. 2). When IF0 and MIS0 were combined, a large portion of the increased cell kill (SER = 4) was a reflection of this phenomenon. The exact nature of this enhancement is unknown at the present time, but may reflect the pharmacokinetics ofthe toxic products of IF0 and/or the interaction of these products with hypoxic cells. For example, the liver metabohtes of IF0 may be very toxic to hypoxic 9L cells when these metabolites reach the tumor. Alternatively, further reduction of the liver metabolites by hypoxic 9L cells may be required to generate the cytotoxic moiety. To obtain chemosensitization when IF0 and MIS0 were combined required that IF0 be given at least I hr before MISO. The chemosensitization of CY required only that CY be given before or at the same time as MIS0 or SR-2508. This schedule dependence is reversed from that determined previously for the nitrosoureas ( IO, I I). This difference probably stems from the site of “activation” of the drugs: intercellularly for the nitrosoureas or in the liver for IF0 and CY. From our data (Figs. 1 and 2) we suggest that IF0 or CY be given prior to MIS0 to enhance their cell kill. This recommended order of administration is reversed from that used in the recent clinical trials of CY and MIS0 (3, 5). This may in part explain the failure of these clinical trials. When the magnitude of the potentiation of CY- and IFO-induced cell kill are compared, IF0 was potentiated slightly more than CY. These data imply that the use of IF0 in a clinical chemosensitization trial in small cell lung cancer is appropriate. The increased toxicity of IF0 and CY combined with MIS0 is unlikely to be troublesome in the clinic. Mesna has been shown to be extremely effective in reducing the hemorrhagic cystitis induced by IF0 and CY in humans (I) and is likely to do the same for the combination of MIS0 with these alkylating agents. In summary, four conclusions can be drawn from the data in this study. As expected, metabolism of the 2-NI by hypoxic cells is necessary for potentiation of the alkylating agent-induced cell kill. Although both MIS0 and SR-2508 potentiated the SC 9L cell kill induced by CY, only MIS0 potentiated the cell kill induced by IFO. The timing of administration of the alkylating agents and the 2-NI is a critical determinant of the extent of the cell kill obtained. The cell kill induced by IF0 appears to be modified by MIS0 to a slightly greater extent than the cell kill induced by CY. However, as a cautionary note, normal tissue toxicity may also be enhanced, thereby making the therapeutic gain smaller than the SER reported on the basis of cell kill.
730
I. J. Radiation
Oncology
0 Biology 0 Physics
Volume 22. Number
4, I992
REFERENCES 1 Brade, W. P.; Nagel, G. A.; Seeber, S., eds. Ifosfamide in tumor therapy. Munich, Germany: Karger, 1987. 2. Chapman, J. D.; Urtasun, R. D.; Franko, A. J.; Raleigh. J. A.; Meeker, B. E.; McKinnon, S. A. The measurement of oxygenation status of individual tumors. In: Palliwal, B. R., Fowler, J. F., Hebert, D. E., Kinsella, T. J., Orton, C. G., eds. Prediction of response in radiation therapy (Part I): the physical and biological basis. NY: Am. Inst. of Physicists in Medicine; 1989:49-60. 3. Davila, E.; Klein, L.; Vogel, C. L.; Johnson, R.; Ostroy, F.; Browning, S.; Gorowski, E.. Furrier, R. L.; Presant, C. A. Phase I trial of misonidazole (NSCX261037) plus cyclophosphamide in solid tumors. J. Clin. Oncol. 3: 12 1-127; 1985. 4. Gatenby, R. A.; Kessler, H. B.; Rosenblum, J. S.; Coia, L. R.; Modldofsky. P. T.; Hartz, W. H.; Broden, G. J. Oxygen distribution in squamous cell carcinoma metastases and its relationship to outcome of radiation therapy. Int. J. Radiat. Oncol. Biol. Phys. 14:831-838; 1988. 5. Glover, D.; Trump, D.; Kvols, L.; Elson. P.: Vogl, S. Phase II trial of misonidazole (MISO) and cyclophosphamide (CYC) in metastatic renal cell carcinoma. Int. J. Radiat. Oncol. Biol. Phys. 12: I405- 1408; 1986. 6. Hinchliffe, M.; McNally. N. J. A comparison of the ability of some radiosensitizers undergoing clinical trials to act as chemosensitizers. Int. J. Radiat. Oncol. Biol. Phys. 10: 16351640: 1984. 7. Hirst, D. G.; Hazlehurst. J. L.; Brown, J. M. Sensitization of normal and malignant tissues to cyclophosphamide by nitroimidazoles with different partition coefficients. Br. J. Cancer 49:33-42; 1984.
8. Law, M. P.: Hirst. D. G.: Brown, J. M. Enhancing effect of misonidazole on the response of the RIF- 1 tumor to cyclophosphamide. Br. J. Cancer 44:208-2 18: I98 I. 9. McNally, N. J.; Hinchliffe. M.; de Ronde, J. Enhancement of the action of alkylating agents by single high, or chronic low doses ofmisonidazole. Br. J. Cancer 48:27 I-278; 1983. IO. Mulcahy. R. T. Update: the potential for chemosensitization of alkylating agents by nitroimidazoles. Oncology 12: I727: 1988.
I I. Mulcahy,
R. T.; Trump, D. L. Clinical chemosensitization by misonidazole and related compounds, a critical evaluation. J. Clin. Oncol. 6:569-573; 1988.
12. Wallen. C. A.; Michaelson,
S. M.; Wheeler, K. T. Evidence of an unconventional radiosensitivity of 9L subcutaneous tumors. Radiat. Res. 84:529-54 I ; 1980.
13. Wheeler, K. T.; Wallen, C. A.; Wolf, K. L.; Siemann. D. W. Hypoxic cells and in situ chemopotentiation of the nitrosoureas by misonidazole. Br. J. Cancer 49:787-793; 1984. 14. Wong. K.-H.; Wallen, C. A.; Wheeler, K. T. Biodistribution of misonidazole and 1.3 bis(2-chloroethyl)1-nitrosourea (BCNU) in rats bearing unclamped and clamped 9L subcutaneous tumors. Int. J. Radiat. Oncol. Biol. Phys. 17: I35143: 1989. 15. Wong, K.-H.: Wallen, C. A.: Wheeler, K. T. Chemosensitization of the nitrosourea by 2-nitroimidazoles in the subcutaneous 9L tumor model: pharmacokinetic and structureactivity considerations. Int. J. Radiat. Oncol. Biol. Phys. 18: 1043-1050: 1990.