Effect of long-term metronidazole (MTZ) therapy on experimental colon cancer in rats

JOURNAL

OF SURGICAL

RESEARCH x,547-552

(1984)

Effect of Long-Term Metronidazole (MTZ) Therapy on Experimental Colon Cancer in Rats A. MAJEED A-KAREEM, M.B., B.S., DAVID M. FLEISZER, M.D., GEOFFREY K. RICHARDS, M.B., MARY K. SENTERMAN, M.D., AND RAE A. BROWN, M.D.’ Montreal

General Hospital, McGill

University, Montreal,

Quebec H3G IA4 Canada

Presentedat the Annual Meeting of the Association for Academic Surgery, Syracuse,New York, November 2-5, 1983 Current reports suggest a beneficial effect of long-term metronidazole (MTZ) therapy in Crohn’s disease.Since Crohn’s diseaseis associatedwith a higher risk of bowel cancer and long-term MTZ has been shown to have a tumorigenic potential in rodents and a cocarcinogenic effect in experimental colon cancer, more studies are required to explore this area. Eighty-one rats were divided into four groups. Group A served as a control, groups B and C were given MTZ in their food (50 mg/kg/day). In groups C and D, a 3-cm colonic segment was isolated and brought out as a blind loop fistula. All animals received 20 weekly sc dosesof 1,2dimethylhydrazine (DMH) and were killed 25 weeks after the first injection. The mean number of colon tumors per animal (+SEM) in MTZ groups B (1.65 rt 0.29) and C (2.57 +- 0.38) were higher than A (1.44 + 0.3) and D (1.18 f 0.21) but the increase was only significant for group C over groups A and D (P < 0.05) and group B (P = 0.06). The mean number of tumors per animal in the isolated loop of group C (0.95 f 0.28) was similar to group D (0.68 + 0.16) P = 0.4 1, but the mean number of tumors in the functioning colon of group C (1.62 f 0.25) was higher than group D (0.5 + 0.12) P < 0.001. These findings suggestthat (1) long-term MTZ increased the number of colon tumors per rat in the DMH model but a statistical significance (P < 0.05) was only noted in the MTZ and surgery group. (2) Surgery alone did not increase the number of tumors. (3) This effect of MTZ is dependent on the presenceof the fecal stream, since there was no significant difference between the number of tumors in the empty loops of MTZ and non-MTZ groups. INTRODUCTION

Metronidazole (MTZ), a nitroimidazole derivative antibiotic, was originally developed as a trichomonacide and its clinical use in the treatment of human trichomoniasis began in the early 1960s following the first report by Cosar and Julou [5]. Subsequent to the observation by Shinn [31] that the drug was effective in the treatment of Vincent’s acute ulcerative gingivitis, MTZ became the subject of a number of in vitro and in vivo studies demonstrating its effective role in the treatment of amoebiasis 1231,giardiasis [ 131,and in the prevention and treatment of anaerobic infections [6, 11, 34, 361. Since MTZ contains a nitro group (NO,) I Author to whom requests for reprints should be addressed.

in its molecular structure, and these groups have been associated with drugs producing leukopenia, Rustia and Shubik [28] investigated the long-term effect of high doses of MTZ in mice and concluded that it has a carcinogenic potential since it increased the incidence of spontaneous lung tumors and malignant lymphomas in mice. Since then other studies [4, 27, 291 have shown similar findings in mice and rats. MTZ and its metabolites were also found to have a mutagenic activity in the bacterial system [20, 331 but no genotoxic effect on human lymphocytes [ 18,221. With these results, human safety became a controversial issue [2 1, 261which was addressed by Beard ec al. in 1979 [l] when they retrospectively studied 77 1 women treated between 1960 and 1969 with at least one course of MTZ for vaginal trichomoniasis and with a mean observation period of about

541

0022-4804/84 $1.50 0 1984 by Academic Press, Inc. AI1 rights OF repmduction in any form reserved.

Copyngbf

548

JOURNAL

OF SURGICAL

RESEARCH:

11 years. They concluded that there was no appreciable increase in cases of cancer after such an exposure to the drug. Another similar study of nearly 2500 women for a shorter observation period also revealed a similar finding [7]. All the above-mentioned laboratory, clinical, and epidemiologic studies were reviewed and analyzed by Roe [26 ] who concluded that there is no reason why MTZ should not be used freely and without fear of carcinogenic risk, although more human long-term followup studies are needed. Association of colon cancer with MTZ in experimental animals was first described by Sloan et al. in our laboratory [32] when he demonstrated an increased incidence of 1,2dimethylhydrazine (DMH)-induced colon cancer in MTZ-fed rats. He postulated that these findings were the result of either a systemic effect of absorbed MTZ and/or its metabolites, or a local effect on the intracolonic contents which in turn enhance DMH-induced colon cancer. In the present study we have attempted to test these hypotheses. METHODS

One hundred and two male Sprague-Dawley rats weighing 70-90 g were housed individually in hanging wire mesh cagesand administered tap water ad Iibitum. All animals were fed the same basic diet which consisted of Charles River RMH-2000 (Agway, Inc., Syracuse, N. Y.) rat chow. The equicaloric consumption was assured by using the pair feeding techniques previously described by us [32] in which the amount of food was weighed out daily for each animal. Sixty-one animals underwent partial colonic bypass in which a 3-cm segment of the ascending colon beginning 2 cm distal to the caecum (Fig. 1) was transected with its blood supply preserved. Colonic continuity was restored by end-toend anastomosis using a single layer interrupted 6-O silk (Fig. 2). The distal end of the isolated segment was brought out through the abdominal wall as a mucous fistula after closure of its proximal end. Pentobarbital (40 mg/kg ip) was used to achieve anaesthesia.

VOL. 36, NO. 6, JUNE 1984

FIG. 1. Location of the isolated loop of the rat’s colon.

Forty-eight animals survived the procedure and were divided into two groups (C and D) containing 24 rats each. The remaining 41 nonoperated animals were also divided into two groups (A and B). Beginning at Week 2 of the study, MTZ 50 mg/kg body wt per rat per day was added in a powder form to the diet of nonoperated group B and operated group C animals (Fig. 2). All animals were weighed once a week and beginning at Week 5 of the experiment all animals received subcutaneous injections of 1,2-dimethylhydrazine (DMH) (Aldrich Chemical Corp.) once a week for 20 weeks at a dose of 20 mg/kg body wt in 0.9% saline solution at pH 7.2 (Fig. 2). At 29 weeks all animals were killed and complete necropsies were performed. Student’s t test was used for statistical analysis. Our institution’s guide for the care and use of laboratory animals was followed throughout the study. RESULTS

Of the 89 animals, 8 died of various causes during the early course of the study, and none of these animals were included in the results. The final numbers of rats per group are A= 18, B=20, C=21, and D=22. The growth curves of the four groups indicated

A-KAREEM ET AL.: METRONIDAZOLE

549

THERAPY

appearanceranging from ulcerated endophytic to polypoid exophytic type and with varying degrees of differentiation. Metastatic adenocarcinoma to peritoneal cavity, liver, and lungs was also seen, but was not confined to one particular group. The mean number of colon tumors per animal (+-standard error mean) of the four groups are A = 1.44 (-tO.3), B = 1.65 (+-0.29) C = 2.57 (+_0.38),and D = 1.18 (kO.21) (Fig. 4). We found that group C (surgery and MTZ) had significantly more tumors than groups A (control) and D (surgery) (P < 0.05) and more FIG. 2. Schematic representation of the four experi- than group B (MTZ) but to a P value of 0.06. mental groups’ colons. A and B: no surgery, D and C: The other three groups, A, B, and D, showed colon&segmented isolation, B and C: given MT2 (Flagyl). no statistically significant difference among All groups received DMH. them, although group B (MTZ) was slightly higher than groups A and D (Fig. 4). that the weight gain of groups B and D was Looking at groups C (surgery and MTZ) less than the control group A, but not to a and D (surgery) and comparing the loop tustatistical significance (P > 0.05). Group C, mors separately (Fig. 5), we found that the however, had significantly less weight gain mean number of loop tumors per animal was than the control group (P < 0.00 1) beginning similar in the two groups (C = 0.95 +- 0.28 at Week 16 of the study (Fig. 3). vs D = 0.68 +- 0.16, P = 0.4). The mean Bowel adenocarcinoma was the predomi- number of tumors per animal in the fimcnant malignancy in all the groups with a gross tioning colon was significantly higher in group C (1.62 r 0.25) than group D (0.5 * 0.12) (P < 0.001) (Fig. 5). Analyzing the results as a percentage of tumor bearing animals per group, we found no

3 28 2b 24 22 2 I

rurgw

7,) SURGERY L-J OfI ’ 35

DMH al mgmg sc ! L I 6

12 16 WEEKS

20

24

, 28

FIG. 3. Growth curves of the four groups (mean group weight). * = Beginning of a significant difference, C < A (P < 0.001).

IL D

COLON TUMORS PER ANIMAL t"WJ"*StMl

FIG. 4. C vs A and D, P < 0.05; C vs B, P = 0.06.

550

JOURNAL OF SURGICAL RESEARCH: VOL. 36, NO. 6, JUNE 1984 COLON & LOOP P
2.t 2.f

C : Surgery

& Metronidazole

D: Surgery

2.4 2.: 2

COLON P
T

1.f 1.t 1.1 1.: 1

.. :: . IL 5

0.f OS 0.4 0.2

-

c.

1

LOOP

P 0.4

C D ~ C D D COLON TUMORS PER ANIMAL L+krn*51VI,

FIG. 5. Colon: the remaining functioning part of the colon after a segment isolation. Loop: the isolated blind loop segment.

significant differences(&i-square test) between the groups (Fig. 6). DISCUSSION

technique 1321and a lower dose of MTZ (50 mg/kg body wt per day) to achieve a better control of the consumed amounts of diet and MTZ, thus avoiding some of the aforementioned limitations. However, there was a significant decreasein the weight gain of the high tumor incidence group C in our study starting 11 weeks after the first injection of DMH. This could be as a result of the increased number of tumors in this group, since DMH-induced colon cancer has been shown to develop as early as 8 weeks after induction [36]. MTZ has never been reported to have a colon carcinogenic effect in rodents, although high doseswere given throughout the life span of those animals. On the other hand, we have demonstrated in a previous study [32], a potentiating effect of MTZ on DMH-induced colon cancer in rats. This apparent promoting or “cocarcinogenic” effectcould be manifested along two possible pathways, in&alumina1 or systemic (Fig. 7). Preliminary results in our laboratory indicated no change in colonic mucosal turnover after a short exposure to oral MTZ in rats. MTZ is known to affect the

loo-

80.-

Previous metronidazole carcinogenicity studies [4, 27-291 in rodents have various limitations as indicated by Roe [25,26]. Some of these limitations are (1) the nutritional status of different experimental groups and the amount of MTZ consumed were not well controlled, since diet was given ad libitum, with possible variations in the caloric and MTZ intake. (2) High total doses of MTZ were administered throughout the life span of the animals. These doses,on a milligram/killigram body weight basis, are equivalent to between 350 and 3000 times that given to patients for trichomoniasis. (3) Failure to corCOLON TUMORS rect spontaneous tumors incidence data for FIG. 6. Percentageof tumor bearing animals per group. the observed longer survival of MT&treated A = 77.7%(control), B = 85%(metronidazole), C = 85.5% animals in some studies. In the present study (metronidazole and surgery), D = 72.7% (surgery). No we used the previously described pair feeding significant difference between the groups (chi-square test).

A-KAREEM ET AL.: METRONIDAZOLE

THERAPY

551

the first limb of our hypotheses, i.e., intraluminal effect (Fig. 7). Finally, the concern about human safety is justified, especially when we consider the recent reports [35] which indicated a beneficial effect of long-term MTZ therapy in Crohn’s FIG. 7. Hypotheses of the possible mechanismsof metronidazole promoting effecton experimental colon cancer. diseasepatients who are already at a higher risk of developing bowel cancer than the general population [8, 161. This question about composition of bile [ 191, and to change the human risk is still unanswered, and more gut bacterial profile in humans [2] and ex- studies are certainly needed in this area. perimental animals [32], and in view of the possible role of bile acids [ 31and bacteria [ 101 REFERENCES in colon carcinogenisis, MTZ might have poBeard,C. M., Noller, K. L., O’FaJJon,W. M., Kurland, tentiated the induced colon tumors through L. T., and Docker@,M. B. Lack ofevidence for cancer its effect on one or both of these factors. due to use of metronidazole. IV. Engl. J. Med. 301: The systemic effect of absorbed metabolites 519, 1979. of MT2 gave another possible explanation, Brass,C., Richard, G. K., Ruedy, J., Prentis, J., and since they were found to have a mutagenic Hinchey, E. J. The effect of metronidazole on the incidence of postoperative wound infection in elective activity in the bacterial system [33]. Also, colon surgery. Amer. J. Surg. 135: 91, 1978. acetamide which is a liver carcinogen in rats Chomchai, C., Bhedracheri, N., and Nigro, N. D. The [ 121, has been detected in urine of patients effect of bile on the induction of experimental intes[ 15] and rats [ i 41 treated with MTZ. Finally, tinal tumor in rats. Dis. Colon Rectum 17: 3 IO, 1974. MTZ has a documented immunodepressive Cohen, S. M., Erturk, E., Von Esch, A. M., Crovetti, A. J., and Bryan, G. T. Carcinogenicity of 5-nitroeffect in rats [ 171and mice 191,and thus posfurans, 5-nitroimidazoles, Cnitrobenzenes and related sibly enhancement of DMH-induced colon compounds. J. Nat. Cancer Inst. 51: 403, 1973. cancer, as suggestedpreviously by us [ 321.But 5. Cosar,C., and Julou, L. Activite de I’(hydroxy-2’etbyl~ others [30] have found that additional treat1 methyl-2 nitro-5 imidazole (8.823 R-P.) vis-a-vis ment with immunodepressants to DMHdes infections experimentalesa trichomonas vaginalis. Ann. Inst. Pasteur 96: 238, 1959. treated rats did not alter the incidence or the 6. Eykyn, S. J. The therapeutic use of metronidazole in induction period of tumors. anaerobic infection: Six years’ experience in a London In the present study we have shown that hospital. Surgery 93: 209, 1983. long-term MTZ therapy in rats subjected to 7. Friedman, G. D. Cancer after metronidazole. N. Engl. surgery (partial colonic bypass) significantly J Med. 302: 519, 1980. 8. Greenstein, A. J., and Sachar, D. B. Cancer in inincreases the multiplicity of DMH-induced flammatory bowel disease.Sure. Dig. Dis. 1: 8, 1983. colon cancer (Fig. 4). This increase appears 9. Grove, D. I., Mahmoud, A. A. F., and Warren, to be due to a promoting or a “cocarcinogenic” K. S. Suppressionof cell-mediated immunity by meteffect since the percentage of tumor bearing ronidazole. Int. Arch. Allergy. Appl. Zmmunol. 54: animals per group was not significantly altered 422, 1977. from one group to the other (Fig. 6). On the 10. Hill, M. 3. Some leads to the etiology of cancer of the large bowel. In L. Nyhus (Ed), Surgery Annual. other hand, no statistically significant effect New York Appelton-Century-Crofts, 1978. P. 135. of MTZ alone or surgery alone could be dem- 11. Hinchey, E. J., Richards, G. K., and Prentis, J. Metonstrated in the present study. ronidazole asa prophylactic agent in wound infection It is also evident that this possible cocarafter colon surgery. Surgery 93: 197, 1983. cinogenic effect of MTZ is dependent on the 12. Jackson, B., and Dessau, F. 1. Liver tumors in rats fed acetamide. Lab. Invest. IO: 909, 1961. presence of the feca1 stream in the colonic 13. Kbampatta, R. B. Metronidazole in giardiasis. Ann. lumen, since there was no difference in tumor Trop. Med. Parasitol. 65: 487, 1971. incidences of the isolated colonic loops be- 14. Koch, R. L., Chrystal, E. J. T., Beaulieu, Jr., B. B., tween C and D (Fig. 5). This finding supports and Goldman, P. Acetamide-A metabolite of met-

552

JOURNAL OF SURGICAL RESEARCH: VOL. 36, NO. 6, JUNE 1984

ronidazole formed by the intestinal flora. B&hem. Pharmacol. 28: 3611, 1979.

15. Koch, R. L., Beaulieu, Jr., B. B., Chrystal, E. J. T., and Goldman, P. A metronidazole metabolite in human urine and its risk. Science 211: 398, 1981. 16. Korelitz, B. I. Carcinoma of the intestinal tract in Crohn’s disease:Results of a survey conducted by the national foundation for ileitis and colitis. Amer. J. Gastroenterol. 78: 44, 1983.

27.

28. 29.

17. Kostakis, A., and Calne, R. Y. The immunosuppressive action of metronidazole. IRCS Med. Sci: Cardiovasc.; Immunol. Allergy; Transplant. 5: 142, 1977.

Pharmacol.;

Surg.

18. Lambert, B., Lindbald, A., and Ringborg, U. Absence of genotoxic effect of metronidazole and two of its urinary metabolites on human lymphocytes in vitro. Mutat. Res. 67: 281, 1979. 19. Low-Bear, T. S., and Nutter, S. Colonic bacterial activity, biliary cholesterol saturation, and pathogenesis of gallstones. Lancet 2: 1063, 1978. 20. McCann, J., Choi, E., Yamasaki, E., and Ames, B. A. Detection of carcinogens as mutagens in the Salmonella/microsome test: Assay of 300 chemicals. Proc. Nat. Acad. Sci. USA 12: 5135, 1975. 21. Med. Lett. Drugs. Ther.: Metronidazole hydrochloride (FlagyI iv) 23: 13, 1981. 22. Mitelman, F., Strombeck, B., and Ursing, B. No cytogenic effectofmetronidazole. Lancet 1: 1249,198O. 23. Powell, S. J., Macleod, I., Wilmot, A. J., and ElsdonDew, R. Metronidazole in amoebic dysentery and amoebic liver abscess.Lancet 2: 1329, 1966. 24. Pozharisski, K. M. Morphology and morphogenesis of experimental epithelial tumors of the intestine. J. Nat. Cancer Inst. 54: 1115, 1975. 25. Roe, F. J. C. Metronidazole: Review of usesand toxicity. J. Antimicrob. Chemother. 3: 205, 1977. 26. Roe, F. J. C. Toxicologic evaluation of metronidazole

30.

31. 32.

33.

34. 35.

36.

with particular reference to carcinogenic, mutagenic, and teratogenic potential. Surgery 93: 158, 1983. Rust, J. H. An assessmentof metronidazole tumorgenicity: Studies in the mouse and rat. Excerpta Medica, I.C.S. 438: 138, 1977. Rustia, M., and Shubik, P. Induction of lung tumors and malignant lymphomas in mice by metronidazole. J Nat. Cancer Inst. 48: 721, 1972. Rustia, M., and Shubik, P. Experimental induction of hepatomas, mammary tumors, and other tumors with metronidazole in noninbred Sas:MRC(WI)BR rats. J. Nat. Cancer Inst. 63: 863, 1979. Schmahl, D., Danisman, A., Habs, M., and Diehl, B. Experimental investigations on the influence upon the chemical carcinogenesis.2. Krebsforsch. 86: 89, 1976. Shinn, D. D. Metronidazole in acute ulcerative gingivitis. Lancer 1: 1191, 1962. Sloan, D. A., Fleiszer,D. M., Richards, G. K., Murray, D., and Brown, R. A. Increased incidence of experimental colon cancer associatedwith long-term metronidazole therapy. Amer. J. Surg. 145: 66, 1983. Speck, W. T., Stein, A. B., and Rose&ram, H. S. Mutagenicity of metronidazole: Presence of several active metabolites in human urine. J. Nat. Cancer Inst. 56: 283, 1976. Tally, F. P., Sutter, V. L., and Finegold, S. M. Metronidazole versus anaerobes.In vitro data and initial clinical observations. Calif: Med. 117: 22, 1972. Ursing, B., Alm, T., Barany, F., Bergelin, I., GanrotNorlin, K., Hoevels, J., Huitfeldt, B., Jamerot, G., Krause, U., Krook, A., Lindstrom, B., Nordle, O., and Rosen, A. A comparative study of metronidazole and sulfasalazine for active Crohn’s disease:The cooperative Crohn’s disease study in Sweden. Gastroenterology 83: 550, 1982. Willis, A. T., and Fiddian, R. V. Metronidazole in the prevention of anaerobic infections. Surgery 93: 174, 1983.