Carcinogenicity Studies of Astemizole in Mice and Rats

(CANCER RESEARCH 55. 5589-5594. December I. 19951 Carcinogenicity Studies of Astemizole in Mice and Rats

Johanna Benze,' Ludo Gypen, John Vandenberghe, Ann Lampo, Roland De Coster, Charles Bowden, and Herman Van Cauteren

Janssen Research Foundation, Division offanssen Pharmaceutica N. V., Turnhoutseweg 30. B-2340 Beerse, Belgium

ABSTRACT The results from both an 18-month mouse study and a 2-year rat study establish clearly that astemizole is not carcinogenic. We have pub The histamine H1 antagonist astemizole (Hismanal) was tested for lished our findings in response to claims made recently by Brandes et carcinogenicity in Swiss mice and Wistar rats. Astemizole was adminis a!. (7) that astemizole, amongs other H1 antagonists, promotes the tered with the food to mice for 18 and to rats for 24 consecutive months. The doses givenâ€”approximately5,20, and 80 mg/kg body weightday growth of experimentally induced tumors in mice. were equivalent to 25, 100, and 400 times, respectively, the recommended human dose of 10 mg/day. Survival of both mice and rats was comparable MATERIALS AND METHODS between groups. Peto's age-adjusted, dose-related trend analysis for the tumor-bearing rats did not reveal a statistically significant difference for Test Animals and Animal Care. A total of 400 Swiss mice from our own males or females. There was no evidence that astemizole led to an in permanent noninbred laboratory colony and 400 specific pathogen-free rats of creased incidence of spontaneously or unusually occurring neoplastic the Wistar substrain, purchased from Charles River (Sulzfeld, Germany), were lesions in either mice or rats. used in the studies. All animals were approximately 5 weeks old at the start of Special attention was given to the effect ofastemizole on the progression the studies. of spontaneously occurring mammary gland adenomas and fibroadeno Mice were housed individually in macrolon cages with a wire mesh roof mm. Peto's analysis applied to the number of female rats bearing these (24 X 14 X 13 cm), and rats were housed individually in stainless steel mesh benign mammary gland tumors disclosed no statistically significant dose cages (20 X 20 X 20 cm) in air-conditioned rooms (temperature, 18â€”22Â°C; related trend. There was no positive trend for the onset ofthis tumor type, alternating 12-h light and dark cycles; relative humidity, 50â€”70%). and the median size of the tumor over time per rat was also not statisti Both mice and rats were given free access to fresh tap water dispensed from cally significantly different in a comparison of the control group with each drinking bottles. Mice received Huybrechts' powdered mice food, which was of the dosed groups. pelleted and administered in self-raising hoppers. Rats received Huybrechts' The findings from these carcinogenicity studies suggest that astemizole powdered rat food supplemented with a vitamin premix. is not tumorigenic and that it does not promote tumor growth. Test Article and Study Design. The test article, astemizole, was synthe sized by Janssen Pharmaceutica. Batches A2201 and C0401 (both pharmaceu tical grade) were used in the mouse and rat carcinogenicity studies, respec INTRODUCTION rively. Astemizole was administered to the animals in food pellets (mice) or powdered food (rats) available ad libitum. Astemizole was mixed into the Astemizole, 1-(4-fluorophenylmethyl)-N-[l-(2-(4-methoxyphenyl)- animals' food at concentrations of 0 (control group), 2.5, 10, or 40 mg/l00 g ethyl l)-4-pipiridinyl]-1H-benzimidazole-2-amine, is a potent and food for mice and 0 (control group), 5, 20, or 80 mg/100 g food for rats. The @ long-acting histamine receptor antagonist devoid of central, sed concentrations of astemizole were equivalent to approximately 5, 20, and 80 ative, and anticholinergic effects (1). At a once-daily administration of mg/kg body weightday.3 The low dose level of S mg/kg body weightday was 10 mg, astemizole is indicated for the treatment of perennial and chosen as a multiple (approximately 25-fold) of the recommended human dose seasonal allergic rhinitis, allergic conjunctivitis, chronic urticaria, and of 10 mg/day (i.e., 0.20 mg/kgday for a 50-kg man or 0.14 mg/kgday for a other allergic conditions (2). Astemizole and other second-generation 70-kg man), and the high dose of 80 mg/kg body weightday was chosen to I receptor antagonists have a more favorable benefit:risk ratio than produce adverse effects on the basis of findings in previously conducted first-generation histamine receptor blockers with regard to the toxicity studies in mice and rats. In both species, each study group comprised 50 males and 50 females. absence of CNS2 effects. They do not cross the blood-brain barrier Dosing started on March 4, 1980, and continued for 18 consecutive months readily and do not potentiate the CNS effects of alcohol and other in the mouse study. For rats, dosing started on March 21, 1984, and continued CNS-active chemicals (3). Adverse experiences associated with for 2 years (105 weeks). astemizole, such as dry mouth, gastrointestinal disturbances, rash, and CliniCal Observations and Evaluations. All test animals were observed at increased appetite, occur rarely, and the incidences are comparable to least once daily for signs of waning health, abnormal behavior, unusual those during placebo treatment (4). appearance, occurrences of untoward clinical effects, and manifestations of During the development of astemizole, which was marketed sub toxic and pharmacological responses. All rats underwent a physical examina sequently as Hismanal, numerous preclinical studies were carried out lion once weekly, and the location and size of any palpable cutaneous or s.c. to assess the drug's safety for human use. In laboratory animals, tissue mass were recorded. A scoring system was used to determine the size of astemizole was shown to be safe on single and repeated administra these tissue masses (score 1, 1 cm or smaller; score 2, larger than 1 cm but smaller than 3 cm; score 3, 3 cm or larger). Body weight and food consumption tion. There was no evidence of adverse effects on fertility, and of rats were determined every 4 weeks. Blood samples were taken at necropsy astemizole was devoid of primary embryotoxic, teratogenic, carcino from the carotid artery for hematological analysis from all rats surviving the genic, and mutagenic effects under widely varying test conditions study period. Hematological analysis included hematocrit, hemoglobin, RBC (5, 6). count, and WBC count. The present article provides a detailed account of the findings At the scheduled termination dates, the surviving mice or rats were exam obtained from two carcinogenicity studies conducted with astemizole. med. anesthetized with ether, and killed by exsanguination. A full autopsy was

Received 7/5/95; accepted 10/3/95. 3 The concentration of astemizole/lOO g food was kept constant throughout the study The costs of publication of this article were defrayed in part by the payment of page period. In the mouse study, neither food consumption nor actual drug intake were charges. This article must therefore be hereby marked advertisement in accordance with determined, but it is known that an adult mouse of this strain consumes about 6â€”10g I8 U.S.C. Section 1734 solely to indicate this fact. food/day. In the rat study, the actual drug intake ranges for males receiving astemizole at I To whom requests for reprints should be addressed, at Department of Toxicology, 5, 20, and 80 mg/l0O g food were 2.36â€”4.59, 9.47â€”19.2,and 37.3â€”77.5mg/kg body Janssen Pharmaceutica, Turnhoutseweg 30, B-2340 Beerse, Belgium. weight'day, respectively. The actual drug intake ranges for female rats were calculated as 2 The abbreviation used is: CNS, central nervous system. 3.48â€”5.32,14.5â€”20.7,and 61.9â€”113mg/kg body weight'day. 5589

performed, and all macroscopic changes were noted. On animals that died Table 1 Neoplastic lesions from male micefed monthsÂ°Treatmentasiemizole for 18 during the study or required killing prematurely, a necropsy was performed as groupCâ€•L soon as possible. During the necropsies, tissue samples were collected for subsequent histological evaluation. Care was taken to collect and identify MHApproximate potentially neoplastic tissue. All tissue samples for histopathology were fixed 2080No. dose (mg/kg)05 in buffered 10% formalin, processed, and examined histologically by means of 5050Survivedof animals5050 4348Survived12 mo4444 standard hematoxylin and eosin-stained sections. Slides of five animals/study 2635No. 18 mo2827 group were reviewed at a time. The order was: (a) control group; (b) low 2730animalsAdrenalof tumor-bearing343 1 dosage group; (c) medium-dosage group; and (d) high-dosage group. Statistical Analysis. In the mouse study, differences concerning mortality, 01Adrenalcortex adenoma01 10pheochromocytomaHematopoieticmedulla01 clinical observations, gross pathology, hyperplastic changes, and neoplastic changes between the control group and each dosed group were analyzed with 79tumorsLiverHepatocellularsystemI I9 @ the two-sided test (8). The same test was used in the rat study for the determination of differences between the control group and each dosed group 1614Hepatocellularadenoma1512 with regard to clinical observations, gross pathology, and nonneoplastic 02Hemangio(endothelio)ma58carcinomaI0 changes. Differences in body weight, food consumption, and hematology were 62Lung, analyzed with the two-sided Mann-Whitney U test (8). For the rat study, Peto's 97Pancreas,primary lung tumor91 1 statistic for a positive dose-related trend (9) was used to analyze mortality, total 02exocrineSoftadenoma,I0

number of tumor-bearing animals, and number of animals bearing a specific tissueFibrosarcoma01 tumor type. As suggested by Peto et a!. (9), the tumor lethality was taken into 01Histiocytoma10 account. An age-adjusted analysis was done for rats bearing fatal tumors and 01Testis, for rats bearing incidental tumors by means of the death rate method and 00ThyroidLeydig cell tumor10 gland adenoma00 prevalence method, respectively. The results were combined in a nonbiased 02 fashion afterward. a Types of neoplasms that occurred only once in the study are not listed in the table. b C, control; L, low dose; M, medium dose; H, high dose. The occurrence of mammary gland adenomas or fibroadenomas in female rats was given special attention. The time of onset of this particular tumor type was analyzed by means of Peto's trend statistic. The size of the tumor was plasia), the lungs (alveolar epithelium hyperplasia), and the pancreas analyzed by nonparametric statistics. For each rat, the median of the scores for (large or hyperplastic islets). Hyperplastic lesions were seen in fe the size of the tumor over time was computed by linear interpolation between males in the following locations: adrenals (myelopoiesis), bronchial the two size scores enclosing the 50th percentile. A vertical line through this lymph nodes (myelopoiesis), mammary gland (hyperplastic nodules), interpolated median split the histogram of the size scores into two equal parts. pancreas (myelopoiesis), Peyer's patch (hyperplasia), red splenic pulp These interpolated medians were compared among the four study groups by (hyperplasia), skin (acanthosis), and thyroid gland (follicular hyper means of a two-sided Kruskal-Wallis one-way ANOVA (8). The same medians were compared between the control group and each dosed group with the plasia). There was no statistically significant difference in the mci Mann-Whitney U test (one sided). dence of the above-mentioned hyperplastic lesions in the dosed One-sided P values are indicated by P1. and two-sided P values are groups and the control groups. indicated by P2. In female mice, the most frequently seen hyperplastic change was a cystic hyperplasia in the uterus. In the 80-mg/kg dosage group, RESULTS however, the incidence of this hyperplastic lesion was statistically significantly lower (P2 < 0.001 ; 6 dosed animals versus 22 control 18-Month Mouse Study animals). Mortality. The number of male or female animals that died per Neoplastic Histological Changes. The numbers of tumor-bearing month or were euthanized in moribund condition varied, but this male and female mice in the dosed groups were not statistically parameter was not statistically significantly different in the dosed different from those in the control group. The various tumor types groups and the control groups. occurring in all dosed groups and in the control groups are listed in Pairwise comparison of the number of deaths per dosed group Tables 1 (male mice) and 2 (female mice). The f test analysis versus the control group did not reveal any statistically significant revealed a statistically significant lesser incidence of hepatocellular differences, with the exception of female mice dosed at S mg/kg adenoma in female mice dosed at 80 mg/kg (P2 = 0.031) than in the (Tables 1 and 2). controls. No drug- or dose-related increase in the incidence of benign Clinical Observations and Gross Pathology. All of the clinical or malignant tumors occurred in any of the dosage groups. signs and macroscopic changes observed were normal for aging mice. 2-Year Rat Study They occurred in the control groups and all dosed groups with a similar incidence and severity. Among the most common findings Mortality. Peto's analysis did not reveal a positive dose-related were pneumonia, tissue masses in the lungs, swollen lymph nodes, trend for mortality in rats (P1 = 0.129 for males; P1 = 0.274 for obesity, alopecia, skin irritations, and swelling of the spleen in both females). The numbers of surviving animals per group are given in sexes, and stimulation of the mammary gland and ovarian cysts in Tables 3 (male rats) and 4 (female rats). females. In females dosed at 80 mg/kg, a decrease in the incidence of Clinical Observations. The clinical signs noted during the study liver tissue masses (P, < 0.05; 2 dosed animals versus 10 controls) period are encountered normally in aging rats. They occurred in both and dilation of the uterus (P2 < 0.01; 5 dosed animals versus 21 undosed and dosed animals, and the incidence of the observations was controls) was disclosed. mostly comparable between the dosed and the undosed groups. In Hyperplastic Changes. In male and female mice, organs or tissues both males and females, frequent clinical observations included poor affected by hyperplasia included the exocrine pancreas (focal hyper general condition, the occurrence of s.c. and cutaneous masses, alo plasia), the pituitary gland (focal hyperplasia), the liver (focal hyper pecia, and waste of food. At 80 mg/kg, an increase in the incidence of plasia and myelopoiesis), the spleen (myelopoiesis), and the stomach cachexia (P2 < 0.05; 6 dosed animals versus 0 controls) was noted in (glandular hyperplasia). Other hyperplastic changes were seen only in male rats, and an increase in waste of food (P2 < 0.001 ; 41 dosed male mice and affected organs such as the adrenals (cortical hyper animals versus 2 1 controls) was observed in female rats. 5590

Table 2 Neoplastic lesions fromfemale micefed monthsÂ°Treatmentastemizole for 18 Nonneoplastic Histological Changes. Most of the observed non groupCâ€•L neoplastic changes are seen normally in aging rats. The affected organs included the pituitary gland, adrenals, thyroids, spleen, lymph MHApproximate nodes, mammary gland, liver, pancreas, lungs, kidneys, testes, uterus, 2080No. dose (mg/kg)05 5050Survivedof animals5050 and ovaries. The incidence of some nonneoplastic changes, however, 4538Survived12 mo4642 was statistically significantly different in the dosed animals and the 3225No. 18 mo3423' 2331Adrenalof tumor-bearing animals3028 controls (Table 5). None of the listed changes were associated with an 10Hematopoieticcortex adenoma00 increased neoplastic response of the organ or tissue involved. The 118LiverHepatocellularsystem tumors1212 1 increase in incidence of small, foamy macrophage aggregates in the 52'Hemangioma42adenoma104 lungs (in males and females) and follicular pigmentation in the spleen 21Lung, and sinusoidal pigmentation in the liver (in females) is indicative of a 73Mammaryprimary lung tumor45 glandAdenocarcinoma33 response of the mononuclear phagocyte system to a toxic dose level of 22(adenoacanthoma)Carcinoma01 astemizole. The higher incidence of fibrosis in the mammary glands in females dosed at 80 mg/kg is probably associated with the reduction 00OvaryGranulosa-theca of lipoid tissue in the integument, related to the lower body weight of 00TeratomaI0 cell tumorII animals in this dosage group. 00HemangiomaII 00PancreasAdenoma,

02Adenoma,exocrine02 01Pituitary endocrine01 Table 3 Neoplastic lesions from maleratsyearsâ€•Treatment fed astemizolefor 2 21adenomaSoftgland (pars distalis),36 groupPCâ€•L 11UterusCarcinoma01tissue fibrosarcomaI0 fortrend'Approximate MH(I-sided) 00HemangiomaI0 2080No. dose (mg/kg)05 00Leiomyoma00 5050Survivedof animals5050 02SchwannomaI0 5050Survived12 mo5049 00 2926No. 24 mo3423 4346Abdominalof tumor-bearing animals4341 a Types of neoplasms that occurred only once in the study are not listed in the table. 010.4815Adrenalmesothelia, sarcoma10 S C. control; L, low dose: M, medium dose; H, high dose. glandAdenoma00 ( P2 < 0.05. 0I0.2342Ganglioneuroma00 010.2342Pheochromocytoma45 Body Weight. Body weight evolution in males and females was 460.2743Hematopoietic comparable between the control group and rats dosed at S or 20 120KidneyAdenocarcinoma12system, tumor02 130.1 mg/kg, but body weight was slightly lower in rats dosed at 80 mg/kg. 000.9038Adenoma00 Hematology. A slight statistically significant decrease (P2 < 0.05) 010.2321Lipoma00 010.2321Large in hematocrit and hemoglobin was seen in males dosed at 80 mg/kg. 100.3064fibroleiomyo(sarco)maLiver,intestine,00 No other meaningful biological differences between groups were found. 410.8064Lymphneoplastic nodule41 420.4180hemangio(endothelio)maMammarynodes,30 Gross Pathology. Most of the observed gross pathology changes are seen normally in aging rats, and their incidence was comparable glandAdenoma, between the dosed and the control groups. In both male and female 200.7775Adenocarcinoma00fibroadenoma20 110.1126Carcinoma00 rats, changes of the kidneys and the liver, pancreatic nodules, and 010.2586PancreasAdenoma, swelling of the pituitary gland with hemorrhagic foci occurred fre quently. A relatively high incidence of heart dilation, hemorrhagic and 7100.1905Adenoma,endocrine69 860.8886Pituitaryexocrine134 swollen lymph nodes, edematous salivary glands, cutaneous and s.c. glandAdenoma1510 masses, testicular changes, increases in thoracic fluid, thymic involu 13130.5386Adenocarcinoma20 010.7517Prostate tion, swollen thyroids, and dilation of the major blood vessels was 020.0617Seminalcarcinoma00 seen in male rats. In female rats, frequently observed macroscopic 000.6646SkinPapillomaI0vesicle carcinoma01 changes in all dosage groups included obesity, generalized cachexia 340@34dSquamous and anemia, swollen adrenal glands, tissue masses in the cervix, 110.1969Smallcell carcinoma01 mammary gland changes, ovarian cysts, swelling of the spleen, thymic 100.6984Softintestine adenocarcinomaI0 involution, thyroid nodules, and swelling of the uterus. tissueFibroma35 140.5211Sarcoma00 The occurrence of some macroscopic abnormalities was statisti 100.3252FibrosarcomaI3 cally significantly higher in the 80-mg/kg dosage group. For male rats, 0I0.6867Hemangioendothelial 010.2361Testis, sarcoma01 these changes included cachexia (P2 < 0.05; 15 dosed versus S 30.5623ThyroidLeydig cell tumor1 31 7 131 controls), swollen kidneys (P2 < 0.05; 18 dosed versus 8 controls), glandAdenoma87 white and/or hemorrhagic foci on the lungs (P2 < 0.01; 40 dosed 650.8217Adenoma, 320.6598Adenocarcinoma51â€œlightcellâ€•solid33 versus 24 controls), small testes (P2 <0.05; 24 dosed versus 12 630.5049 controls), and white nodules on the thyroid gland (P2 < 0.05; 18 dosed a Types of neoplasms that occurred only once in the study are not listed in the table. versus 8 controls), and for female rats these changes included the b C. control; L, low dose; M, medium dose; H, high dose. presence of white foci on the lungs (P2 < 0.001 ; 31 dosed versus 7 C Age-adjusted analysis, taking into account the context of observation. P values are either asymptotic (asymptotic P value of Peto's trend statistic, no correction for continu controls) and hemorrhagic lymph nodes (P2 < 0.05; 6 dosed versus 0 ity) or exact (exact P value of the age-adjusted Cochran-Armitage trend test). The exact controls). However, histological examination revealed that none of P value was computed when the asymptotic P value was at least marginally statistically significant (P < 0.10) and when the total number of tumor-bearing animals in all groups these macroscopic changes was linked to a higher incidence of neo was eight or less. plastic changes. dp1

Table 4 Neoplastic lesionsfrom ftmalerats2yearsâ€•C5Treatmentfed astemizolefor cate that the background incidence of skin papilloma in males ranges group from 0 to 5/50 animals. In this study, the incidence of 1/50 in the (l-sidedf control group, 0/50 in the 5-mg/kg dosage group, 3/50 in the 20- fortrendApproximate L MHP mg/kg dosage group, and 4/50 in the 80-mg/kg dosage group is, 2080No. dose (mg/kg)05 therefore, not higher than the background incidence of this benign 5050Survivedof animals5050 4950Survived12 mo5050 tumor type. 2721No. 24 mo2527 On histological examination after death or sacrifice, most of the 4243Abdominalof tumor-bearing animals4546 100.3291Adrenalmesothelia, sarcoma00 palpable tumors in females were identified as mammary gland ade glandAdenoma01 nomas or fibroadenomas (Tables 4 and 7). No statistically significant 000.7405Ganglioneuroma00 dose-related trend was detected for the number of mammary gland 100.4107PheochromocytomaII 300.6329CervixGranular adenoma- and fibroadenoma-bearing rats per study group (P1 = 0.7476). In the analysis of the more tumor-specific variables, 000.6661Sarcoma53cell tumor01 neither the time of onset (week of first recording of the existence of 1I0.9734Hematopoietic 010.9350KidneyAdenoma01system. tumor32 a palpable tissue mass smaller than 1 cm in size) for this particular tumor type (P1 = 0.740) nor the median tumor size over time per rat 000.6581Lipoma00 010.2100Large (Table 8) was statistically significantly different between groups. 110.1151fibroleiomyo(sarco)maLiver.intestine.00

590.1065Lymphneoplastic nodule64 DISCUSSION 000.9431hemangio(endothelio)maMammarynodes.I2 Astemizole, when administered ad libitum in the food to mice at glandAdenoma, daily doses of 2.5, 10, and 40 mg/l00 g food for 18 months and to rats 16190.7476Fibroma21fibroadenoma2224 100.9077Adenocarcinoma49 at daily doses of 5, 20, and 80 mg/100 g food for 2 years, produced 940.4357Carcino(sarco)maI0 110.3979OvaryGranulosa-theca Table 5yearsâ€•Treatment Nonneoplastic changes from rats fed astemizolefor 2 I00.8002Sertoli cell tumorII 010.2100PancreasAdenoma,cell tumor00 HApproximate group Câ€• L M 80MalesLiver, dose (mg/kg) 0 5 20 310.6967Adenoma,endocrine24 020.2930PituitaryexocrineII cholangiofibrosis 17 28' 15 14 34300.3158Smallgland adenoma2935 42â€•Spleen,Lungs, foamy cells 29 31 27 010.5294Softintestine. fibroleiomyosarcomaI0 6TestisAccumulationparenchymal pigmentation 11 9 3' tissueFibroma00 I20.0688Sarcoma00 22â€•Mineralizationof content 7 6 11 100.3135Fibrosarcoma00 23â€•Vasculopathy 8 15 11 200.2514Thyroid 24dFemalesLiverCholangiofibrosis 9 13 12 glandAdenoma21 100.9110Adenoma, 240.1256Adenocarcinoma01â€œlight-cellâ€•solidI3 9Ectasia 18 13 8' 000.6903Uterus, 3Focal of sinusoidal spaces 0 6' 3 200.9738HemangioendotheliomaIIadenoma44 5CSinusoidalcellular changes I5 11 12 I00.7954Adenocarcinoma2I 14'Lungs, pigmentation 4 5 5 I00.9170Carcinoma 38eMammaryfoamy cells 16 10 14 010.2339Sarcoma00scirrhous00 glandFibrosis 100.3021 43CSecretion 3 1 24 29 a Types of neoplasms that occurred only once in the study are not listed in the table. @Spleen, 35 47d b C, control; L, low dose; M, medium dose; H, high dose. 24ea follicular pigmentation 5 2 2 C Age-adjusted analysis, taking into account the context of observation. P values are The number of examined organs per study group was either 49 or 50. either asymptotic (asymptotic P value of Peto's trend statistic, no correction for continu bdose.Cp2 C, control; L, low dose; M, medium dose; H, high ity) or exact (exact P value of the age-adjusted Cochran-Armitage trend test). The exact < 005@ P value was computed when the asymptotic P value was at least marginally statistically dp2 < 0.01. significant (P < 0. 10) and when the total number of tumor-bearing animals in all groups e,,@<0.001(x2test,two-sided).Table was eight or less.

ratsAge 6 incidence of ski,, papilloma iii male Neoplastic Changes. Almost all tumor-bearing rats had benign adjusted test for positive dose-related trendâ€• P (1-sided) Lunar tumors (85â€”100%),and about one-third of the tumor-bearing rats had dAnimalsmonth Câ€• L M H Trend variance z C malignant tumors (2 1â€”36%). papilloma1â€”20bearing incidental skin There was no statistically significant, positive dose-related trend 0.00021â€”270/6 0/6 0/8 0/3 0.000 0.000 with regard to the number of tumor-bearing rats per study group 1.044Terminal0/10 0121 2/13 1/21 1.923 3.391 1.550All 1/34 0/23 1/29 3/26 3.902 6.338 either for males (P1 = 0.208) or for females (P1 = 0.424). Further 0.0434e 5.825 9.729 1.867 0.0309 more, no positive dose-related trend concerning the number of rats 4/50Animals1/500/503/50 with fatal or probably fatal tumors was established. The same applies papillomaTotalbearing fatal or incidental skin 0.0434C 5.825 9.729 1.867 0.0309 to rats with incidental or probably incidental tumors. 1/50 0/50 3/50 4/50 The various tumor types found in all dosed groups and in the a From Peto et al. (9). Dose levels 0, 1, 2, and 3. control groups are listed in Tables 3 and 4. A separate analysis was b Number observed/number at risk. C, control; L, low dose; M, medium dose; H, high performed for males and females, and no statistically significant dose. C Asymptotic P value of Peto's trend statistic (no correction for continuity). dose-related trend was detected in either sex, except for the occur d Exact P value of the age-adjusted Cochran-Armitage trend test (StatXact, version 2. rence of skin papilloma in males (Table 6). This finding, however, is 1991; Cytel Software Corp.). This P value was computed when the asymptotic P value was at least marginally statistically significant (P < 0. 10) and when the total number of not considered relevant, because historical control data obtained from tumor-bearing animals in all groups was eight or less. eight different carcinogenicity studies using the same rat strain mdi eNumberofanimalswithskinpapilloma/numberofanimalsexaminedhistologically. 5592

Table 7 IncidenceratsAge of mammary gland adenoma/fibroadenoma in female adjusted test for positive dose-related trendâ€• Lunar (I-sided)'variancezAnimalsmonth C@' L M H TrendP bearing fatal mammary gland adenoma/fibroadenoma 1.5031.2551.341I14 0/50 0/50 0/49 1/50 .2451.360186 0/50 0/50 0/49 1/48 1.5181 â€”1.4731.239â€”1.324201/48 0/48 0/47 0/45 â€”1.4671.238â€”1.319241/47 0/48 0/45 0/44 1.4961.2131.35826 0/33 0/35 0/36 1/33 .0486.8230.40127 0/29 3/34 2/32 1/29 1 0.5691.1670.526All 0/26 0/27 1/28 0/21 3.193

4/5014.1800.8480.1983Animalsâ€˜I 2/50 3/50 3/50 bearing incidental mammary gland adenoma/fibroadenoma 0.0000.0000.00021â€”221â€”20 0/3 0/4 0/5 0/7 â€”1.7604.742â€”0.808233/Il 2/6 1/6 0/2 â€”2.7333.129â€”1.54524â€”27I/I 2/5 0/3 1/6 â€”0.82810.443â€”0.256Terminal3/8 3/5 1/6 4/10 â€”3.12029.407â€”0.575All13/25 14/27 11/27 10/21 â€”8.441 15/4647.722â€”1.2220.8891Animals(I 20/48 21/47 13/47 bearing fatal or incidental mammary gland adenoma/fibroadenoma Total â€”5.248 d 22/50 24/50 16/50 19/5061.902â€”0.6670.7476

â€˜IFrom Peto et til. (9). Dose levels 0, 1 , 2, and 3. h Number observed/number at risk. C, control; L, low dose; M, medium dose; H, high dose.

( Asymptotic P value of Peto's trend statistic (no correction for continuity).

@1Number of animals with mammary gland adenoma/fibroadenoma/number @â‚¬animals examined histologically.

no evidence of a carcinogenic response. The doses given/lOO g food Carruthers Czyzewski (18) reported that the Canadian Health Protec were equivalent to about 5, 20, and 80 mg/kg body weighUday, which tion Branch had evaluated the methodology and had not been able to is approximately 25, 100, and 400 times higher, respectively, than the reproduce the findings published by Brandes et a!. (7). Another group recommended human dose of 10 mg/day. of researchers (19) also did not confirm the claim of Brandes et a!. that Astemizole did not affect survival of either mice or rats adversely. astemizole has tumor growth-promoting properties. They adminis Neither in mice nor in rats did dosing with astemizole result in a dose tered astemizole not only i.p. (3 mg/m2) but also p.o. (dose range, or drug-related increase in the number of tumor-bearing animals, and 3â€”60mg/m2) to C57BL/6 female mice inoculated with Bl6FlO benign and malignant tumor rates were comparable in the dosed and melanoma cells. Astemizole neither enhanced tumor growth in vivo control groups. The age-adjusted Peto analysis for rats revealed that nor stimulated B16F1O melanoma cell proliferation in vitro. astemizole caused neither a dose-related increase in the fatal or As part of the normal screening research done at Janssen, a set of probably fatal tumor rate nor a dose-related increase in the incidental tests was performed with the aim of investigating the effect of astem or probably incidental tumor rate. izole on tumor-related mortality in mice inoculated with murine Furthermore, we found no evidence that astemizole increases the fibrosarcoma (MO4), mammary carcinoma (TA3-Ha) or leukemia incidence of any type of malignant neoplasm in either mice or rats. A (Ll210). Astemizole was administered i.p. at 24 h after tumor cell dose-related trend for an increase in benign skin papilloma, in which, inoculation and every fourth day until day 13 (for M04 and TA3-Ha) however, the incidence was not higher than the background incidence or until day 9 (for Ll2lO). For all three tumor types, tumor-related of this tumor type, was seen in male rats only. In female mice, there mortality of mice receiving astemizole (dose range, 2.5â€”20 mg/kg, was even a statistically significant decrease in the incidence of hep equivalent to 12.5â€”100times the recommended human dose) was atocellular adenoma at the high dose. comparable with that in the control groups. The absence of any type of malignant neoplasm with increased In the carcinogenicity studies, we assessed the possibility of a incidence indicates that astemizole neither initiates nor promotes tumorigenic effect of astemizole in initially healthy, non-tumor-bear carcinogenesis in mice and rats. However, both histamine and anti ing albino mice and rats, which, in some respects, resembles the histamines might play a role in modulating tumor growth and devel situation in humans, who also do not have their tumors injected but opment. It has been shown in various experimental tumor models and undergo a natural disease process. in in vitro tumor cell cultures that histamine promotes tumor cell proliferation via the H2 receptor, and that H, receptor antagonists Table 8 Comparison of the median sizeâ€•ofa mammary gland adenomalfibroadenoma inhibit tumor cell growth and tumor development (10â€”15).There is over time perfemale rat also evidence that histamine stimulates tumor cell growth via the H, Studygroupsâ€•Test'StatisticsPC, receptor (I 1, 16, 17). Depending on the experimental model, H1 L, M,HKW2001d0.572eC, antagonists exert either an inhibitory effect (1 1, 16, 17) or a stimula LMWU+0.031o486@C, tory effect (11, 14, 7) on tumor growth or tumor cell proliferation. In MMWUâ€”1.158/O.8ll@C, HMWU-0.511O.697@a a previous publication, Brandes et a!. (7) reported that some H1 The median is computed by linearly interpolating the tumor size score distribution antihistamines, among them astemizole, promoted the growth of two perrat. transplantable murine tumorsâ€”B 16F10 melanoma and C-3 fibrosar b C, control; L, low dose; M, medium dose; H, high dose. comaâ€”in two pigmented strains of mice. Astemizole administered C KW, Kruskal-Wallis one-way analysis of variance; MWU, Mann-Whitney U test. d Chi-square value with three degrees of freedom. i.p. at doses of 1.5â€”24mg/m2(6 mg/m2 being equivalent to the human eTwo-sidedPvalue. dose of 10 mg/day) for 18â€”21days increased tumor wet weight, albeit @Standardnormal value (z value), which is positive if the median tumor size is larger in the dosed group than in the control group. with a bell-shaped dose-response curve. g One-sided P value. It is less than 0.5 when the median tumor size is larger in the This method for predicting tumor growth is not accepted widely. dosed group than in the control group. 5593

Both our carcinogenicity studies were carried out according to the 2. Janssens, M. M-L. Astemizole, a nonsedating antihistamine with fast and sustained guidelines and recommendations in force at the time (e.g., Ref. 20). activity. Clin. Rev. Allergy, 1/: 35â€”63,1993. 3. Simons, F. E. R. H -receptor antagonists, comparative tolerability and safety. Drug Although it is now recognized that existing protocols for carcinoge Saf., 10: 350â€”380,1994. nicity testing are insufficient (21, 22), regulatory agencies are hesitant 4. Ray, J. A. The development of a new antihistamine: astemizole. N. E. R. Allergy to adopt new methods as part of recommended programs because of Proc.,6(1):71â€”77,1985. 5. Awouters, F. H. L., Niemegeers, C. J. E., and Janssen, P. A. J. Pharmacology of uncertainties as to their validity and predictive value. the specific histamine H,-antagonist astemizole. Arzneim. Forsch.. 33: 381â€”388, For human risk assessment, the rodent carcinogenicity test is still 1983. accepted as the most appropriate one for evaluating carcinogenic 6. Vanparys,P.,Fabry,L., Leonard,A., and Marsboom,R.Mutagenicitytestswith astemizole in vitro and in vivo. Arch. Toxicol., 50: 167â€”173,1982. effects of drugs (23). This lifetime rodent test allows evaluation of 7. Brandes, L. J., Warrington, R. C., Arron, R. J., Bogdanovic, R. P., Fang, W., both the carcinogenic-initiating and tumor-promoting effects of drugs Queen, G. M., Stein, D. A., Tong, J., Zaborniak, C. L. F., and La Bella, F. S. on spontaneously occurring neoplasms. Enhanced cancer growth in mice administered daily human-equivalent doses of In an attempt to detect a possible link between astemizole and some H,-antihistamines: predictive in vitro correlates. J. NatI. Cancer Inst., 86: 770â€”775,1994. acceleration of tumor growth, we assessed the development of mam 8. Siegel, S. Nonparametric Statistics for the Behavioral Sciences, International Student mary gland neoplasms (adenoma and fibroadenoma), because these Ed. New York: McGraw-Hill Book Co., 1956. tumor masses are palpated easily during the live phase of the study. In 9. Peto,R.,Pike,M.C.,Day,N.E.,Gray,R.G.,Lee,P.N.,Parish,S.,Peto,J.,Richards, S., andWahrendorf,J.Guidelinesforsimple,sensitivesignificancetestsfor carci carcinogenicity studies, mammary neoplasms are seen usually after nogenic effects in long-term animal experiments. IARC (Int. Agency. Rca. Cancer) long periods in rodents that are already predisposed to the develop Monogr. Eval. Carcinog. Risks. Hum. Suppl., 2: 31 1â€”426,1980. ment of mammary neoplasms occurring spontaneously with advanc 10. Watson, S. A., Wilkinson, L. J., Robertson, J. F. R., and Hardcastle, J. D. Effect of histamine on the growth of human gastrointestinal tumours: reversal by cimetidine. ing age (24). Under the influence of a given compound with growth Gut,34: 1091â€”1096,1993. promoting properties, tumors can be expected to grow faster. This was 11. Van der Ven, L. T. M., Prinsen, I. M., Jansen, G. H., Roholl, P. J. M., Defferrari, R., decidedly not the case with astemizole. We found no indication of an Slater, R. and Den Otter, W. Growth of cultured human glioma tumour cells can be regulated with histamine and histamine antagonists. Br. J. Cancer. 68: 475â€”483, earlier onset of development of palpable mammary gland adenomas 1993. and fibroadenomas in females dosed with astemizole, and the median 12. Bartholeyns, J., and Fozard, J. R. Role of histamine in tumor development. Trends. tumor size per rat over time was not statistically significantly different Pharmacol. Sci., 6: 123â€”125,1985. 13. Tutton,P.J. M.,andBarkla,D.H. StimulationofcellproliferationbyhistamineH2 in dosed and undosed females. This suggests that astemizole, under receptors in dimethylhydrazine-induced adenocarcinomata. Cell Biol. Lot. Rep., 2: the given experimental conditions, does not enhance or potentiate the 199â€”202,1978. growth of spontaneously occurring tumors in rodents. 14. Cricco, G. P., Davio, C. A., Bergoc, R. M., and Rivera, E. S. Inhibition of tumor In addition, it was shown in a series of tests that astemizole has no growth induced by histamine: in vivo and in vitro studies. Agents Actions, 38: C175â€”C177, 1993. mutagenic potential (6). There was no evidence of primary DNA damage 15. Burtin, C., Noirot, C., Scheinmann, P., Galoppin, L., Sabolovic, D., and Bernard, Ph. (sister chromatic exchange test on human lymphocytes), point and/or Clinical improvement in advanced cancer disease after treatment combining hista gene mutations (Ames test in five strains of Salmonella typhimuriwn and mine and H2-antihistaminics (ranitidine or cimetidine). Eur. J. Cancer & Clin. Oncol., 24: 161â€”167,1988. the sex-linked recessive lethal test in Drosophila melanogaster), or chro 16. Modlin,I. M.,Kumar,R. R., Soroka,C. J., Ahlman,H.,Nilsson,0., Goldenring, mosomal aberrations (metaphase analysis test on human lymphocytes in J. R. Histamineas an intermediategrowthfactorin genesisof gastricECLomas vitro, micronucleus test in rats, and the dominant lethal test in male and associated with hypergastrinemia in mastomys. Dig. Dis. Sci., 39: 1446â€”1453, 1994. female mice) due to astemizole. 17. Tilly, B. C., Tertoolen, L. G. I., Remorie, R., Ladoux, A., Verlaan, I., de Laat, S. W., Only epidemiological evidence, however, can establish truly andMoolenaar,W.H. Histamineasa growthfactorandchemoattractantforhuman whether there is a carcinogenic hazard for man. To classify an agent carcinoma and melanoma cells: action through Ca2@-mobilizing H, receptors. J. Cell as probably not carcinogenic to humans, there must be evidence Biol., 110: 1211â€”1215,1990. 18. Carruthers Czyzewski, P. Antihistamines and the risk of cancer. Can. Pharm. J., 127: suggesting lack of carcinogenicity in experimental animals together 244, 246, 255, 1994 with evidence suggesting lack of carcinogenicity in humans (25). 19. Kurokawa,M.,Futamura,Y.,Sato,S., Obata,M.,andShibutani,Y.Confirmatory With regard to these criteria, it is noteworthy that after 12 years of study on tumor cell growth-enhancing effect of astemizoleâ€”failure to reproduce this effect. Jpn. Pharmacol. Ther., 23: 311â€”319, 1995 marketing and more than 10 million patient years ofexposure, neither the 20. Carcinogenicity studies. in: Organization for Economic Cooperation and Develop induction of malignant tumors nor the promotion of already existing ment Guidelines for Testing of Chemicals, Section 4. Vol. 451, pp. 1â€”7,1981. tumors in humans has been reported with reference to astemizole. 21. Henschler, D. Carcinogenicity testingâ€”existing protocols are insufficient. Trends Pharmacol. Sci., 6 (Suppl.): 526â€”528, 1985. 22. McAuslane, J. A. N., and Lumley, C. E. Carcinogenicity studies for pharmaceuticals: ACKNOWLEDGMENTS classification and misclassification. Drug Inf. J., 27: 509â€”514, 1993. 23. Schou, J. S. Predictive value of carcinogenicity studies for drug safety: with special We thank Yolande Devocht for her assistance in the preparation of the emphasis on survivability and study duration. In: P. F. D'Arcy and D. W. G. Harron D. W. G. (eds.), Proceedings of The First International Conference on Harmonisation, manuscript. pp. 208â€”212.Brussels:The Queen's University of Belfast, 1991. 24. Mammary Gland. in: P. Greaves (ed), Histopathology of preclinical toxicity studies: REFERENCES interpretation and relevance in drug safety evaluation, pp. 48â€”76.Amsterdam: Elsevier, 1990. I. Richards, D. M., Brogden, R. N., Heel, R. C., Speight, T. M., and Avery, G. S. 25. IARC monographs programme on the evaluation of carcinogenic risks to humans, Astemizole, a review of its pharmacodynamic properties and therapeutic efficacy. preamble. IARC (tnt. Agency. Res. Cancer) Monogr. Eval. Carcinog. Risks. Hum., Drugs, 28: 38â€”61, 1984. 50: 11â€”31,1990.

5594

Johanna Benze, Ludo Gypen, John Vandenberghe, et al.

Cancer Res 1995;55:5589-5594.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/55/23/5589

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/55/23/5589. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.