Registration Dossier

Toxicological information


Currently viewing:

Administrative data

other: subcutaneous and drinking water
Type of information:
experimental study
Adequacy of study:
supporting study
3 (not reliable)
Rationale for reliability incl. deficiencies:
other: Insufficient information on materials and methods.

Data source

Reference Type:
Effect of warfarin on cell kinetics, epithelial morphology and tumour incidence in induced colorectal cancer in the rat.
Goeting G, Trotter GA, Cooke T, Kirkham N, Taylor I
Bibliographic source:
Gut, 26, 807-815

Materials and methods

Test guideline
no guideline available
Principles of method if other than guideline:
The effect of low dose warfarin and high dose warfarin on epithelial cell kinetics as determined by stathmokinetic techniques and preneoplastic morphological changes was studied during azoxymethane induced carcinogenesis in the rat.
GLP compliance:
not specified

Test material

Details on test material:
Study only mentions the test substance name as Warfarin, but mentions dissolving in water and the word therapeutic. Both of these terms point toward the test substance as a sodium or potassium analog, each having different CAS numbers.

Test animals


Administration / exposure

Route of administration:
other: subcutaneous and drinking water
Details on exposure:
Rats received 12 weekly subcutaneous injections of azoxymethane at a dose of 10 mg/kg per week. At week eight they were divided into three groups: group 1 - azoxymethane alone (control), n=50; group 2 - Azoxymethane + low dose warfarin (LDW), n=40; group 3 - azoxyrnethane + high dose warfarin (HDW), n=40. Group 1 continued to receive azoxymethane alone. Groups 2 and 3 additionally received warfarin in their drinking water (LDW and HDW respectively.)
Analytical verification of doses or concentrations:
not specified
Duration of treatment / exposure:
5 weeks.
Frequency of treatment:
One injection per week and continuous exposure in drinking water.
Doses / concentrationsopen allclose all
Doses / Concentrations:
0.6-1.2 mg/l

Doses / Concentrations:
1.2-1.8 mg/l

No. of animals per sex per dose:
Control animals:
Details on study design:
Therapeutic anticoagulation occurs when clotting times are between two and three times the mean normal clotting time. In the LDW group the warfarin dose was 0.6-1.2 mg/l and clotting times were in the range between normal and therapeutic values. In the HDW group, therapeutic anticoagulation was maintained at a dose of between 1-2-1.8 mg/l. Clotting times were monitored using a thrombotest kit on citrated venous blood taken from the tail vein.


Sacrifice and pathology:
Ten animals were killed 10, 15, 20, and 25 weeks after initial azoxymethane injection from each of the three groups, and in addition 10 animals were killed from the control group at five weeks. Ten normal animals, which had received neither azoxymethane nor warfarin were killed at time zero. Animals were injected intraperitoneally with the vinca alkaloid, vincristine (1 mg/kg ) which, at this dose, causes metaphase arrest. The number of arrested metaphases increases with time after injection. To prevent variation in kinetic results due to diurnal changes each procedure started at 9 am. In each group, the first animal was killed 30 minutes after the injection with vincristine, and thereafter one animal was killed every 15 minutes up to 22 hours. The colon was removed from each animal, opened, and washed in saline to remove fecal contents. Samples of approximately 1 cm square were excised from the caecum, ascending colon, transverse colon, descending colon, and rectum. They were fixed in Carnoy's solution for four to six hours, and then transferred to 70% alcohol for storage. Subsequently, specimens were rehydrated on an alcohol gradient, and hydrolyzed tissue was stained with periodic acid Schiff's reagent as part of the Feulgen reaction. Crypts were microdissected in 15% acetic acid, and a wet squash preparation of each sample made for microscopic examination. Each group of animals killed, at each time interval, generated 50 samples (10 animals, five areas per animal). The number of arrested metaphases was counted in 10 crypts for each sample, and the mean metaphase count plotted against the time that the sample was taken (10 points, 30 minutes to 24 hours). The crypt cell production rate (CCPR) - that is, the mean number of cells produced per crypt per hour, was calculated for each site, in each group, from the slope of the graph. In those animals killed at 25 weeks, tumor incidence, distribution, and volume were also noted in each colon before excision of samples for cell kinetic studies.
Other examinations:
After sampling for CCPR, samples of descending colon and rectum were taken from 28 animals for scanning electron microscopy. These included two normal, 10 control (two each at 5, 10, 15, 20, and 25 weeks), eight LDW and eight HDW (two each at 10, 15, 20, and 25 weeks). Samples measuring approximately 1 cm square were excised and pinned flat in 10% formol saline. They were then post-fixed in osmium tetroxide for two hours. Subsequently they were washed in distilled water, then dehydrated on an acetone gradient and subjected to critical point drying with liquid carbon dioxide. Samples were mounted on aluminium stubs, sputter coated with gold and palladium, then scanned using the JEOL 35 scanning electron microscope. Micrographs were taken at x40 magnification and the following parameters were determined: (1) The number of microadenomas per low power field. (2) The area of each microadenoma (using the Apple II graphics tablet). (3) The number of crypts per microadenoma. For the purposes of this study, no distinction was made between distal colon and rectum, as results from an earlier study have shown that morphological epithelial changes during carcinogenesis are similar in both areas.
In the calculation of CCPR, linear regression analysis was carried out to determine the correlation coefficient. The standard deviations of the slopes were also calculated but, as is normal practice in this analysis, they were not quoted. Cell kinetic data between the different groups were compared for statistical significance using Student’s t test modified for this model. All other statistical analyses were done using the Mann Whitney U test for non-parametric variables.

Results and discussion

Results of examinations

Clinical signs:
not examined
not examined
Body weight and weight changes:
not examined
Food consumption and compound intake (if feeding study):
not examined
Food efficiency:
not examined
Water consumption and compound intake (if drinking water study):
not examined
Ophthalmological findings:
not examined
Haematological findings:
not examined
Clinical biochemistry findings:
effects observed, treatment-related
Urinalysis findings:
not examined
Behaviour (functional findings):
not examined
Organ weight findings including organ / body weight ratios:
not examined
Gross pathological findings:
no effects observed
Histopathological findings: non-neoplastic:
not examined
Histopathological findings: neoplastic:
effects observed, treatment-related
Details on results:
Clotting times determined by the thrombotest method were as follows: in normal animals the median clotting time was 24.7 seconds (range 23.6-29.6), in animals receiving HDW the clotting times were greater than 49.4 seconds and in animals receiving LDW they were between 24.7 and 49.4.

In the LDW and HDW groups, CCPR increased with time after initial azoxymethane administration. The CCPR values obtained for both LDW and HDW groups were, however, not significantly different from controls at any time interval. The values for CCPR in LDW and HDW groups at any site, either caecum, ascending colon, transverse colon, descending colon, or rectum, at any time interval studied, either 10, 15, 20, or 25 weeks, were not significantly different from the corresponding values in the control group

In those animals killed at 25 weeks, the number of malignant tumors was significantly reduced in both LDW and HDW groups when compared with controls.

In the distal colon and rectum, at low magnification, the mucosa was relatively flat, and crypt orifices had a round, regular appearance. In contrast theultra structural appearance changed with time after initiation of carcinogenesis. There was loss of crypt regularity, with marked slit like distortion or crypt orifices.

In the LDW and HDW groups, microadenomas were also observed, and similarly the number increased with time. The numbers of microadenomas in both LDW and HDW groups were significantly lower than in control animals, at all time intervals

The mean area of microadenomas was not significantly different between controls and LDW and HDW groups at any time interval. There was, however, a gradual increase in area with time in each of the groups. The mean number of crypts per microadenoma in controls, LDW and HDW groups were also not significantly different at any time interval.

Applicant's summary and conclusion

Warfarin does not exert an overall effect on colonic epithelial cell kinetics, therefore it acts on specific cells. The action of warfarin is independent of the degree of anticoagulation because the tumor incidence is equally reduced in both high and low dose warfarin groups. When microadenomas arise during warfarin administration, the area of the lesions is unaffected when compared with that of controls. Therefore if pre-malignant cells are refractory to the effect of warfarin, their subsequent growth is unaffected. As microadenoma formation is reduced in animals treated with warfarin, it acts prior to this stage of carcinogenesis.