Effect of irradiation on bromodeoxyuridine incorporation in human colon cancer xenografts

Int. J. Radiation

Oncology

Biol.

Phys., Vol. 34. No. 3, pp. 617-621. 1996 Copyright 0 1996 Elsevier Science Inc. Printed in the USA. All rights reserved 0360-3016196 $15.00 + .oO

0369-3016(95)02117-5

ELSEVIER

0 Biology Original Contribution EFFECT

OF IRRADIATION HUMAN

ON BROMODEOXYURIDINE INCORPORATION COLON CANCER XENOGRAFTS

IN

THEODORE S. LAWRENCE, M.D., PH.D.,* EMILY Y. CHANG,* MARY A. DAVIS, PH.D.,* PHILIP L. STETSON, M.D., PH.D.+ AND WILLLAM D. ENSMINGER, M.D., PH.D.+* Departments of *Radiation Oncology, ‘Pharmacology, and %temal Medicine, University of Michigan Medical Center, Ann Arbor, MI 48109 : Although we have characterized the incorporation of the thymidhte analog bromodeoxyuridine %3iF rd) into human colon cancer xenoRtaf& under a wide variety of conditions, iittie is known about the (B effect of radiation on subsequent incorporation. Because ciinicai protocols include, as one component, BrdUrd administration after radiation, it was important to con&m that irradiation did not prevent subsequent BrdUrd incorporation. Therefore, we studied the effect of irradiation on BrdUrd incorporation into IIT human colon cancer xenograits. Methods and Materials: Two types of experiments were performed. In the first, the effect of radiation on subsequent incorporation was measured. Tumors received doses of 0,2,8, and 12 Gy, animals were infused with BrdUrd for 4 days, and incorporation was amessed at the end of the infusion. In the second, the effect of radiation on the elimination of BrdUrd from tumors was determined. Auhuais were infused with BrdUrd, tumors were irradiated with either 0 or 12 Gy, and tumor incorporation of BrdUrd was measured 1 and 3 days later. Results: Radiation affected neither the incorporation into nor the elimination of BrdUrd from human tumor xenografts. Conclusions: These tindings support the feasibility of clinical trials htterdigitating BrdUrd infusion and radiation. Bromodeoxyuridme,

Radiation sensitizers, Human tumor xenografts, Nucieoside analogs.

‘he finding that incorporation of the thymidine analogs bromodeoxyuridine (BrdUrd) and iododeoxyuridine (JdUrd) into deoxyribonucleic acid (DNA) increases the radiation sensitivity of cultured cells and implanted tumors has led to the development of clinical protocols using these agents. Two types of protocols are underway. The first type is designed to treat tumors that are hypothesized to be rapidly proliferative compared to the surrounding normal tissue (e.g., brain and intrahepatic tumors) (2, 4, 12, 19). In these protocols, patients are infused with analog a few days prior to radiation, radiation and analog are administered concurrently, and this entire course is repeated. The second type has been designed using preclinical data (6) to address the condition in which the critical normal tissue is predicted to divide more rapidly than the tumor (e.g., the intestine compared to a retroperitoneal sarcoma). In this approach, infu-

sions ate alternated with irradiation, providing a period of drug elimination from the normal tissue, which is in excess of the elimination from the tumor (13). In both approaches described above, patients receive courses of analog infusion after irradiation. Because irradiation can cause cell cycle arrest both in G2 and at the Gl/S boundary (11,20,21), it seemed possible that analog infused after irradiation has begun might not be efficiently incorporated. If this were the case, infusions, after the first, might cause toxicity (such as bone marrow depression) rather than improved efficacy. We designed a study using athyrnic nude mice bearing HT29 human tumor xenografts to determine if irradiation affected subsequent thymidine analog incorporation. In the first part, tumors were irradiated prior to administration of BrdUrd. In the second part, we assessed the effect of radiation on the elimination of BrdUrd after an infusion

Reprints requests to: Theodore S. Lawrence, M.D., Ph.D., Department of Radiation Oncology, University of Michigan Medical Center, 1331 E Ann St., Ann Arbor, MI 48109-0582. AcknowledgementsWe would like to thank Marlene Langley and Tammara Johnson for secretarial assistance, Zhaomin Yang

for help with BrdUrd immunohistochemistry, and Mary Martel, Ph.D., for verification of the mouse dosimetry. This work was supported by NIH Grants R29-CA53440, POl-CA42761, and Cancer Center Core Grant CA46592. Accepted for publication 23 August 1995.

INTRODUCTION

617

618

I. J. Radiation Oncology 0 Biology 0 Physics

is completed. We found that radiation, even at doses that produced a substantial increase in the volume doubling time, did not affect the subsequent incorporation of BrdUrd. Likewise, we found that elimination of BrdUrd from the tumor was not affected by irradiation. These findings suggest that the use of interdigitating analog and radiation does not compromise analog incorporation and efficacy. METHODS

AND MATERIALS

Tumor implantation HT29 human colon cancer cells were cultured under standard conditions as described previously (7). For tumor implantation, 5 X lo6 viable (trypan blue-excluding) cells/ 0.2 ml phosphate-buffered saline (PBS) were injected subcutaneously into the flank of athymic nude female mice (CD-l). Experiments were conducted with animals bearing 5-10 mm tumors. Infusion of BrdUrd Athymic nude mice were infused using osmotic pumps’ filled with BrdUrd dissolved in a HCO$C03 buffer (pH 10.2). Pumps were verified to deliver a constant infusion of 1 j&h for 4 days, which administers 200 mg/kg/day of BrdUrd to a 25 g mouse. Irradiation conditions Mice with tumors were anesthetized with 2 mg sodium pentobarbital, injected IP. A mouse in a Lucite restrainer was then placed on a Lucite rack suspended beneath the primary collimator and a secondary heavy alloy collimating device on the 6oCo teletherapy unit.’ Mice were positioned such that the apex of each flank tumor was at the center of a 2.4 cm aperture in the secondary collimator and irradiated for 2, 8, or 12 Gy. Assessment of tumor growth Tumors were measured in two dimensions three times a week. Tumor volume (TV) was calculated according to the equation for a prolate spheroid: TV = % (ab’), where a and b are the longer and shorter dimensions of the tumor, respectively. Data are expressed as the ratio of TV at varying times after treatment compared to the day of irradiation (day 0). Measurements were made until day 40 (I 8 Gy) or day 60 (12 Gy) or until the tumor volume had increased by approximately a factor of 8, at which point the animals were sacrificed to avoid potential discomfort. Animals were handled according to the estab-

’ Alza Corporation, Palo Alto, CA.

Volume 34, Number 3, 1996

lished procedures of the University of Michigan Laboratory Animal Maintenance Manual. BrdUrd incorporation Incorporation of BrdUrd into DNA was measured as described previously using a gas chromatographic, mass spectrometric assay (9, 18). Briefly, tissues were lysed in a buffer containing 10 mM Tris, 10 r&I EDTA, and 0.6% sodium dodecyl sulfate (pH 8.0). DNA was isolated from the lysate by proteinase and RNase digestion, followed by ethanol precipitation. Halouracil and thymine were liberated from the DNA by hydrolysis with DNase I, phosphodiesterase, alkaline phosphatase, and thymidine phosphorylase. Chlorouracil was added as an internal standard, and the bases were extracted into ethyl acetate and derivatized with N, 0-bis-(trimethysilyl) trifluoroacetamide. The derivatized products were measured with a Hewlett-Packard 5987A GCA4S in selected ion monitoring mode. Samples were assessed in triplicate. BrdUrd incorporation is expressed as the percent of dThd replaced. Immunohistochemistry The mitotic index was assessed using immunohistochemistry as described previously (1, 10). Briefly, tumors were prepared and animals infused as for the incorporation experiments. At the time of sacrifice, the tissues were fixed in cold 70% ethanol, dehydrated, and embedded in paraffin. Deparaffinized microtome sections of these tissues were then treated with HCl and Triton X-100, followed by 10 min boiling, then by the primary antibody (B44 mouse anti-BrdUrd antibody) and the antimouse IgG secondary antibody, which is coupled to peroxidase. Sections were treated with diaminobenzidine and hydrogen peroxide and counterstained lightly with hematoxylin. The labeling index was scored as the fraction of nuclei that stained positive for BrdUrd (minimum of 300 nuclei in 10 random fields scored for each determination). Statistics Averages are presented as the mean 2 the standard error of at least six animals, except for the immunohistochemistry experiments, in which they are the mean of two tumors that differed in labeling index by < 10%. Means of the results of the different conditions were compared using the F-test. Statistical significance was defined at the level of p < 0.05 (two tailed). RESULTS We wished to determine the effect of radiation on subsequent BrdUrd incorporation after doses of radiation that produced a range of growth inhibitory effects. We found

’ Atomic Energy of Canada Limited, Kanata, Ontario, Canada.

Effect of irradiation on BrdUrd incorporation l

T. S. LAWRENCE

er al.

619

Table 1. Effect of radiationon BrdUrd incorporationandthe influence of BrdUrd incorporationon radiation-induced growth delay in HT29 humancolon cancerxenografts Volume doubling time (days) Dose Incorporation Radiation (Gy) (% thymidine replaced)* only

0' 0

10

Time

after

20 irradiation

30 (days)

40

Fig. 1. Effect of irradiation on growth of HT29 humantumor xenografts.NudemicebearingHT29 xenograftswereirradiated on day 0 with 0 (Cl), 2 (W), 8 (0), or 12 (A) Gy. Tumor size was measuredas describedin Methods and Materials and is expressedasa ratio comparedto the tumor size on the day of irradiation.

that a single fraction of 2, 8, and 12 Gy inhibited

tumor growth minimally, moderately, and substantially (Fig. 1). The next step was to determine conditions of BrdUrd infusion that were potentially radiosensitizing. We found that, when given alone, an infusion of 200 mg kg-’ day-’ for 4 days produced a minimal effect on tumor growth, but that this infusion significantly increased the volume doubling time produced by an 8 Gy dose of radiation (Fig. 2 and Table 1). We then were able to test the hypothesis that irradiation affected the incorporation of BrdUrd. To test this hypothesis, animals bearing xenografts were irradiated with 2, 8, or 12 Gy (or were left unirradiated), infused with BrdUrd (200 mg kg-’ day-‘) for 4 days, and the tumors

0

10

Time

20

after irradiation

30

40

(days)

Fig. 2. Influenceof an infusionof BrdUrd (200 mg kg-’ day-’ for 4 days) on the radiation sensitivity of HT29 humantumor xenografts.Nudemice bearingHT29 xenograftswere kept under control conditions (O), infused with BrdUrd (W), treated with a single 8 Gy fraction of radiation (0), or infusedwith BrdUrd followed by an 8 Gy fraction (0) . Tumor size was measuredas describedin Methods and Materials, and is expressedas a ratio comparedto the tumor size on the day of irradiation.

0 2 8 12

7.4 2 0.9 6.2 +- 1.5 4.5 2 1.0 5.2 2 0.7

BrdUrd/ radiation’

9t 1 10 t 1 11 t 1 not determined 202 4 34 t 4$ 55 +- 10 not determined

* Irradiation immediatelyprior to BrdUrd infusion (200 mg kg-‘day-’ for 4 days). ’ BrdUrd infusedprior to irradiation. * BrdUrd/radiationsignificantly greaterthan radiationalone.

were assessed for BrdUrd incorporation. There was no significant difference in incorporation among these four conditions @ = 0.24) (Table 1). We also used BrdUrd immunohistochemistry to assess the fraction of cells capable of incorporating BrdUrd in tumors from animals irradiated with 12 Gy and then infused with BrdUrd (200 mg kg-’ day-‘) for 4 days. We found that 92% of the cells were positive for BrdUrd, which does not differ significantly from the labeling index after infusion with BrdUrd alone under the same conditions (8). Therefore, radiation did not appear to affect incorporation of BrdUrd into xenografted tumors. Because we have initiated several clinical protocols based on the kinetics of thymidine analog elimination from tumors compared to the intestine (13, 14), we wished to determine if radiation affected the time course of BrdUrd elimination from tumors. To assess the effect of radiation on BrdUrd elimination from tumors, we infused animals with BrdUrd (200 mg kg-’ day-‘) for 4 days and either irradiated tumors with 12 Gy or gave no additional treatment. Tumors were assessed for BrdUrd incorporation either 1 or 3 days after irradiation. We found that tumors irradiated with 12 Gy eliminated BrdUrd at the same rate as unirradiated tumors (Fig. 3). We carried out analogous experiments using BrdUrd immunohistochemistry and found that 1 and 3 days after irradiation with 12 Gy, 91% and 57% of the cells were labeled, respectively. (Although it would have been interesting to assess the effect of radiation on the elimination of BrdUrd from the intestine, preliminary studies suggested that a single fraction of S- 12 Gy to the abdomen was poorly tolerated.) Irradiation with 8 Gy also had no effect on BrdUrd elimination (data not shown). DISCUSSION In this study, we have found that irradiation does not affect the incorporation of BrdUrd in human tumor xenografts. This conclusion is supported by two kinds of experiments. First, radiation administered prior to a BrdUrd

I. J. Radiation Oncology 0 Biology 0 Physics

620

q

Radiation

1 day

Time after infusion (days)

(12 Gy)

3 days

completion

Fig. 3. Effect of radiation on the eliminationof BrdUrd from tumors.Mice wereinfusedfor 4 dayswith BrdUrd (200 mgkg-’ day-‘) and then left under control conditions(opencolumn)or irradiatedwith 12 Gy (hatchedcolumn). BrdUrd incorporation into tumorswasmeasured1 or 3 days after irradiation, as describedin Methods and Materials. infusion did not affect analog incorporation.

Second, radiation delivered after a BrdUrd infusion did not alter the rate of analog elimination from the tumor. These findings suggest that it is rational to deliver repeated BrdUrd infusions during a course of radiation treatments. Although we are unaware of other studies in which BrdUrd has been administered after irradiation, the extent of incorporation into tumors after a continuous infusion, as well as the kinetics of elimination, can be compared to other studies. These findings confirm our previous investigations of BrdUrd incorporation using HT29 cells with respect to the extent of incorporation during an infusion (8). They are also in general agreement with a study of human colon xenografts using IdUrd (at 100 mg kg-’ day-‘) (16). In that study, the extent of thymidine replacement was approximately one-third of that reported here, which would be anticipated based on the poorer incorporation of IdUrd compared to BrdUrd under the same infusion conditions (8). Our results concerning the elimination of BrdUrd also confirm our previous investigations (6), as well as earlier studies using less quantitative methods (3, 17).

Controversy currently exists concerning the optimal

Volume 34, Number 3, 1996

method of administration of the tbymidine analogs as radiation sensitizers. We have advocated the use of different infusion strategies based on the relationship between the ability of the tumor and surrounding normal tissue to incorporate analogs. In the case of intrahepatic tumors, in which the tumor proliferates more rapidly than the surrounding hepatic parenchyma, we have used a continuous hepatic arterial infusion of BrdUrd during two 14-day courses (separated by a 2-week break), with radiation starting on day 8. We have found that a l-week hepatic arterial infusion of BrdUrd (25 mg-’ kg day-‘) leads to approximately 11% thymidine replacement in >80% of the tumor cells. The corresponding figures for normal hepatic parenchyma are 1% replacement in 7% of the cells (5). In the case of rem peritoneal sarcomas, cervix cancer, and pancreatic cancer, in which the surrounding intestine may proliferate more rapidly than the tumor, we have used a strategy of alternate weeks of analog infusion and radiation therapy. As described above, this approach is based on the concept of more rapid drug elimination from the normal tissue than from the tumor. In confirmation of our hypothesis, we have found in patients that during a 3day period after the infusion is discontinued, there is minimal decrease in the incorporation of analog in the tumor, but a significant decrease in the bone marrow. There is a modest decrease in incorporation in the normal rectum but, more importantly, there is a redistribution of incorporating cells from the crypt to the villi. This would be anticipated to decrease radiosensitization of the bowel, because it should provide relative sparing of the crypt cells, which are required to maintain bowel wall integrity compared to the villar cells which are fated to be shed (14). Others have argued that sensitization is limited by the fraction of cells that do not incorporate analog. According to this approach, a long continuous infusion that produces only modest thymidine replacement in a high fraction of cells (>99%) might produce the greatest sensitization (15). Regardless of which method of administration is superior, all approaches are based on the hypothesis that analog incorporation continues during a fractionated course of radiation. The findings from this study confirm this hypothesis. Current clinical trials should determine which (if any) of these methods described above can produce clinically meaningful sensitization.

REFERENCES Beisker,W.; Dolbeare,F.; Gray, J. W. An improvedimmunocytochemicalprocedurefor high sensitivity detectionof incorporatedbromodeoxyuridine.Cytometry 8:235-239; 1987. Chang,A. E.; Collins, J. M.; Speth, P. A. J.; Smith, R.; RowlandJ. B.; Waltom, L.; Begley, M. G.; Glatstein, E.; Kinsella,T. J. A phaseI study of intraarterialiododeoxyuridine in patients with colorectal liver metastases. J. Clin. Oncol. 7:662-668; 1989. Clifton, K. H.; Szybalski, W.; Heidelberger,C.; Gollin, F. F.; Ansfield, F. J.; Vermund, H. Incorporation of I”‘labeled iododeoxyuridine into the deoxyribonucleic acid

of murine and humantissuesfollowing therapeuticdoses. CancerRes.23:1715-1723; 1963. 4. Hegarty, T. J.; Thornton, A. F.; Diaz, R. F.; Chandler, W. F.; Ensminger,W. D.; Junck,L.; Page,M. A.; Gebarski, S. S.; Hood, T. W.; Stetson, P. L. In&a-arterialbromodeoxyuridine radiosensitizationof malignant gliomas. Int. J. Radiat.Oncol. Biol. Phys. 19(2):421-428; 1990. 5. Knol, J. A.; Walker, S. C.; Robertson,J. M.; Yang, 2.; DeRemer,S.; Stetson,P. L.; Ensminger,W. D.; Lawrence, T. S. Incorporationof 5-Bromo-2’-deoxyuridineinto colorectal liver metastases and liver in patientsreceiving a 7day hepatic arterial infusion. CancerRes. 55:3687-3691; 1995.

Effect

of irradiation

on BrdUrd

incorporation

6. Lawrence, T. S.; Davis, M. A.; Stetson, P. L.; Maybaum, J.; Ensminger, W. D. Kinetics of bromodeoxyuridine elimination from human colon cancer cells in vitro and in vivo. CancerRes.54(11):2964-2968; 1994. 7. Lawrence,T. S.; Davis, M. A.; McKeever, P. E.; Maybaum, J.; Stetson,P. L. Normolle,D. P.; Ensminger,W. D. Fluorodeoxyuridine-mediated modulationof iododeoxyuridineincorporationand radiosensitizationin humancolon cancer cells in vitro and in vivo. CancerRes.51(15):3900-3905; 1991. 8. Lawrence,T. S.; Davis, M. A.; Maybaum, J.; Mukhopadhyay, S. K.; Stetson,P. L.; Normolle, D. P.; McKeever, P.E.; Ensminger,W. D. The potentialsuperiorityof bromodeoxyuridine to iododeoxyuridineas a radiationsensitizer in the treatmentof colorectalcancer.CancerRes.52:36983704; 1992. 9. Maybaum, J.; Kott, M. Cl.; Johnson,N. J.; Ensminger, W. D.; Stetson,P. L. Analysis of bromodeoxyuridineincorporation into DNA: Comparisonof gaschromatographic/ massspectrometric,CsCl gradientsedimentation,and specific radioactivity methods.Anal. Biochem. 16l( 1):164171; 1987. 10. McKeever, P. E.; Letica, L. H.; Shakui, P.; Averill, D. R. A multiple-well methodfor immunohistochemical testing of manyreagentson a singlemicroscopicslide.Lab. Invest. 59: 409-413; 1988. 11. Painter, R. B.; Young, B. R. Radiosensitivity in ataxiatelangiectasia:A new explanation.Proc. Natl. Acad. Sci. USA 77:7315-7317; 1980. 12. Phillips, T. L.; Leven, V. A.; Ahn, D. K.; Gutin, P. H.; Davis, R. L.; Wilson, C. B.; Prados,M. D.; Wara, W. M.; Flam, M. S. Evaluation of bromodeoxyuridinein glioblastoma multiforme: A Northern California Cancer Center phaseII study. Int. J. Radiat. Oncol. Biol. Phys. 21:709714; 1991.

l T. S. LAWRENCE

et al.

621

13. Robertson,J. M.; Sondak,V. K.; Weiss, S. A.; Sussman, J. J.; Chang,A. E.; Lawrence,T. S. Preoperativeradiation therapy andiododeoxyuridinefor largeretroperitonealsarcomas.Int. J. Radiat.Oncol. Biol. Phys. 31:87-92; 1995. 14. Robertson,J. M.; Johnson,D.; Eisbruch,A.; Reynolds,K.; Roberts,J.; Lawrence,T. S. A phaseI/II study of bromodeoxyuridine(BrdU) andradiationtherapy (RT) for locally advancedcervix cancer.Proc. ASCO 14:201; 1995. 15. Rodriguez, R.; Miller, E.; Fowler, J. F.; Kinsella, T. J. Continuousinfusion of halogenatedpyrimidines. (Correspondence).Int. J. Radiat. Oncol. Biol. Phys. 20:1378: 1991. 16. Rodriguez,R.; Ritter, M. A.; Fowler, J. F.; Kinsella,T. J. Kinetics of cell labeling and thymidine replacementafter continuousinfusionof halogenatedpyrimidinesin vivo. Int. J. Radiat. Oncol. Biol. Phys. 29:105-l 13; 1994. 17. Rotenberg,A. D.; Bruce, W. R.; Baker,R. G. Incorporation of 5-iododeoxyuridine “‘I in spontaneousC3H mouse mammarytumours.Br. J. Radiol. 35:337-342; 1962. 18. Stetson,P. L.; Maybaum, J.; Shukla, U. A.; Ensminger, W. D. Simultaneousdeterminationof thymine and 5-bromouracilin DNA hydrolysatesusinggaschromatographymassspectrometrywith selected-ionmonitoring. J. Chromatogr.375:l-9; 1986. 19. Sullivan, F. J.; Herscher,L. L.; Cook, J. A.; Smith, J.; Steinberg,S. M.; Epstein,A. H.; Oldfield, E. H.; Goffman, T. E.; Kinsella, T. J.; Mitchell, J.; Glatstein, E. National Cancer Institute (PhaseII) study of high-grade glioma treatedwith acceleratedhyperfractionatedradiationandiododeoxyuridine: resultsin anaplasticastrocytoma.Int. J. Radiat.Oncol. Biol. Phys. 30:583-590; 1994. 20. Tolmach, L. J.; Jones,R. W.; Busse,P. M. The action of caffeine on X-irradiated HeLa cells. I. Delayed inhibition of DNA synthesis.Radiat.Res.71:653-665; 1977. 21. Weinert, T. A.; Hartwell, L. H. The RAD9 genecontrols the cell cycle responseto DNA damagein Saccharomyces cerevisiae.Science241:317-322; 1988.