regulatory toxicology and pharmacology€¦ · a pharmaceutical is tumourigenic in animals and...

Regulatory Toxicology and Pharmacology xxx (2011) xxx–xxx

Contents lists available at ScienceDirect

Regulatory Toxicology and Pharmacology

journal homepage: www.elsevier .com/locate /yr tph

Evaluation of carcinogenicity studies of medicinal products for human useauthorised via the European centralised procedure (1995–2009)

Anita Friedrich a,⇑, Klaus Olejniczak b

a Granzer Regulatory Consulting and Services, Zielstattstrasse 44, 81379 Munich, Germanyb Federal Institute for Drugs and Medical Devices, Kurt-Georg-Kiesinger-Allee 3, 53175 Bonn, Germany

a r t i c l e i n f o a b s t r a c t

Article history:Received 25 October 2010Available online xxxx

Keywords:Medicinal productsCarcinogenicityRodents

0273-2300/$ - see front matter � 2011 Elsevier Inc. Adoi:10.1016/j.yrtph.2011.04.001

⇑ Corresponding author. Fax: +49 89 68 98 15.E-mail address: [email protected] (A. Friedrich

1 Member States of the European Union plus Norwa

Please cite this article in press as: Friedrich, A., OEuropean centralised procedure (1995–2009). R

Carcinogenicity data of medicinal products for human use that have been authorised via the Europeancentralised procedure (CP) between 1995 and 2009 were evaluated. Carcinogenicity data, either fromlong-term rodent carcinogenicity studies, transgenic mouse studies or repeat-dose toxicity studies wereavailable for 144 active substances contained in 159 medicinal products. Out of these compounds, 94(65%) were positive in at least one long-term carcinogenicity study or in repeat-dose toxicity studies.Fifty compounds (35%) showed no evidence of a carcinogenic potential. Out of the 94 compounds withpositive findings in either carcinogenicity or repeat-dose toxicity studies, 33 were positive in both miceand rats, 40 were positive in rats only, and 21 were positive exclusively in mice. Long-term carcinogenic-ity studies in two rodent species were available for 116 compounds. Data from one long-term carcinoge-nicity study in rats and a transgenic mouse model were available for eight compounds. For 13compounds, carcinogenicity data were generated in only one rodent species. One compound was exclu-sively tested in a transgenic mouse model. Six compounds were tumourigenic in repeat-dose toxicitystudies in rats.

The majority of tumour findings observed in rodent carcinogenicity studies were considered not to berelevant for humans, either due to a rodent-specific mechanism of carcinogenicity, a high safety marginbetween exposures at the NOAEL (No Observed Adverse Effect Level) in rodents and recommended ther-apeutic doses in humans, or based on historical control data, a small effect size and lack of dose–responserelationship and tumours typically observed in rodent strains used, or were considered not to be relevantfor humans based on literature and clinical data or likely differences in metabolism/local concentrationsbetween rodents and humans.

Due to the high number of rodent tumour findings with unlikely relevance for humans, the value of thecurrently used testing strategy for carcinogenicity appears questionable. A revision of the carcinogenicitytesting paradigm is warranted.

� 2011 Elsevier Inc. All rights reserved.

1. Introduction

1.1. Marketing authorisation of pharmaceuticals for human use inEurope

In the European Economic Area (EEA),1 a marketing authorisa-tion (MA) can either be issued by the competent authority of a Mem-ber State (or EEA country) for its own territory (nationalauthorisation) or for the entire Community (Community authorisa-tion). Regulation (EC) No. 726/2004 of the European Parliamentand of the Council lays down a centralised Community procedurefor the authorisation of medicinal products, for which there is a sin-gle application to the European Medicines Agency (EMA), a single

ll rights reserved.

).y, Iceland and Liechtenstein.

lejniczak, K. Evaluation of carcegul. Toxicol. Pharmacol. (2011

evaluation by the Committee for Medicinal Products for HumanUse (CHMP) within the EMA and a single marketing authorisationgranted by the European Commission (EC). Community authorisa-tions can be granted for medicinal products that fall under the man-datory scope of the centralised procedure which includes medicinalproducts derived from biotechnology, new active substances forwhich the therapeutic indication is the treatment of acquired im-mune deficiency syndrome, cancer, neurodegenerative disorder, dia-betes, auto-immune diseases or viral diseases, and applications formedicinal products designated as orphan medicinal products. Othernew active substances may be accepted for consideration under thecentralised procedure when the applicant shows that a new activesubstance or the medicinal product constitutes a significant thera-peutic, scientific or technical innovation, or the granting of a Com-munity authorisation for the medicinal product is in the interestsof patients at Community level (optional scope of the centralisedprocedure). Generic applications of medicinal products authorised

inogenicity studies of medicinal products for human use authorised via the), doi:10.1016/j.yrtph.2011.04.001

http://dx.doi.org/10.1016/j.yrtph.2011.04.001

mailto:[email protected]


http://www.sciencedirect.com/science/journal/02732300

http://www.elsevier.com/locate/yrtph


2 A. Friedrich, K. Olejniczak / Regulatory Toxicology and Pharmacology xxx (2011) xxx–xxx

via the centralised procedure may also be authorised via the centra-lised procedure.

1.2. Carcinogenicity testing of pharmaceuticals for human use

1.2.1. Objective of carcinogenicity testingThe objective of carcinogenicity studies is to determine whether

a pharmaceutical is tumourigenic in animals and whether thistumourigenic potential poses a relevant risk to humans (ICH S1Aguideline; CPMP note for guidance on carcinogenic potential).

1.2.2. Need for carcinogenicity studiesCarcinogenicity studies are generally required for pharmaceuti-

cals which are expected to be used continuously for at least sixmonths or intermittently for the treatment of chronic or recurrentconditions (ICH S1A guideline).

Carcinogenicity studies are also necessary if there is a concernabout the carcinogenic potential of a pharmaceutical. Relevant fac-tors include: (1) previous demonstration of carcinogenic potentialin the product class that is considered relevant to humans, (2)structure–activity relationship suggesting a carcinogenic risk, (3)positive genotoxicity findings, (4) evidence of preneoplastic lesionsin repeat-dose toxicity studies, and (5) long-term tissue retentionof the pharmaceutical or its metabolite(s) resulting in local tissuesreactions or pathophysiological responses (ICH S1A guideline).

Pharmaceutical administered infrequently or for short duration,e.g. anaesthetics and radiolabelled imaging agents do not need car-cinogenicity studies unless there is a cause for concern (ICH S1Aguideline).

Carcinogenicity studies are also not required for therapeuticsintended to treat patients with advanced cancer who have a shortlife-expectancy, for pharmaceuticals administered by the dermalor topical route, unless there is significant systemic exposure,and for biotechnology-derived pharmaceuticals, such as endoge-nous peptides and proteins especially when they are given asreplacement therapy (ICH S1A guideline; ICH S9 guideline; ICHS6 guideline).

1.2.3. Test systems for carcinogenicity1.2.3.1. Traditional approach. Long-term carcinogenicity studies inrodents have been required since the 1970s for marketing author-isation of pharmaceuticals in Europe, the USA and Japan. Thestudies have traditionally been conducted in mice and rats usinglife-time treatment.

Until now, the traditional approach of conducting long-termcarcinogenicity studies in mice and rats has remained the mostfrequently chosen testing strategy. However, discussions havebeen ongoing for many years whether a single study alone wouldbe adequate for assessing the carcinogenic potential ofpharmaceuticals.

1.2.3.2. Alternative approach. The analysis of several databases didnot support the concept of conducting long-term carcinogenicityin two rodent species. Positive carcinogenicity findings were oftennot relevant to humans. These analyses led to the introduction byICH of a more flexible weight of evidence approach for carcinoge-nicity testing, that is the use of scientific judgement in the evalua-tion of the data derived from one long-term carcinogenicity studyalong with other appropriate investigations (Smith, 1996; ICH S1Bguideline).

According to the ICH S1B guideline, one long-term carcinoge-nicity study should be supplemented by another study thatsupplements the long-term carcinogenicity study and providesinformation that is not readily available from the long-term assay.Appropriate experimental models include short or medium term

Please cite this article in press as: Friedrich, A., Olejniczak, K. Evaluation of carcEuropean centralised procedure (1995–2009). Regul. Toxicol. Pharmacol. (2011

in vivo rodent assays such as the p53+/� deficient mouse modelor the TgHras2 mouse model.

The ICH S1B guideline recommends that a single long-termstudy (usually in the rat) should be conducted and should be com-plemented by a short- or medium-term in vivo rodent study. Theshort or medium-term test should provide additional informationthat is not readily available from the long-term assay. Usuallythe mouse is the preferred species to be used in a short or med-ium-term assay, especially when the rat is used in the long-termstudy.

Several transgenic mouse assays have proven useful for replace-ment of the second long-term study and are generally accepted byregulatory authorities in all ICH regions. Among these are thep53+/� deficient model, the TgHras2 model and the Tg.AC model.The regulatory authority concerned should always be consultedbefore a decision on the choice of a particular model is made.

1.2.3.3. Toxicokinetic studies. Toxicokinetic assessments are anessential component of carcinogenicity studies. They allow to re-late systemic exposure levels to the toxic and/or carcinogenic find-ings observed in carcinogenicity studies and to contribute to theassessment of the relevance of these findings to clinical safety(ICH S3A guideline).

1.2.3.4. Mechanistic studies. Mechanistic studies are useful for theinterpretation of tumour findings in a carcinogenicity study andcan provide a perspective of their relevance to human risk assess-ment (ICH S1B guideline). As such, mechanistic studies are oftenprovided for compounds with positive carcinogenicity findings.For example, the measurement of hormone levels, such asthyroxine or prolactin, have been used to confirm species- orrodent-specific hormonal imbalances and related tumour develop-ment induced by test compounds (ICH S1B guideline). Additionalgenotoxicity tests, such as the UDS test or the Comet assay, oradditional studies in transgenic or neonatal mice have been usedfor compounds with equivocal findings in the standard batteryfor genotoxicity testing (ICH S2B guideline) in order to exclude agenotoxic mechanism of carcinogenicity.

1.2.3.5. Photocarcinogenicity studies. The SKH1 (hr/hr) albino hair-less mouse model is currently the most widely used model to as-sess photocarcinogenicity in animals (CPMP note for guidance onphotosafety testing). The model is designed to induce squamouscell carcinoma and their precursors by chronic UV radiation in allanimals and to assess the effect of simultaneously applied test sub-stances on the time to tumour development. The model is beingused to assess the co-carcinogenic potential of dermally appliedproducts that intended for long-term or chronically intermittenttreatment of skin disorders. However, the predictivity of the albinohairless mouse model for the human situation is at present unclear(CPMP note for guidance on photosafety testing).

2. Methods

This paper presents an evaluation of carcinogenicity data ofmedicinal products for human use that have been authorised viathe European centralised procedure (CP) between 1995 and 2009.Data were retrieved from European Public Assessment Reports(EPARs) and Summary of Product Characteristics (SPCs) as the pub-licly available sources of scientific and labelling information on thewebsite of the European Medicines Agency (EMA). Additionalinformation on the approved products that is eventually availablefrom the scientific literature was beyond the scope of thisevaluation.



Table 1Reference medicinal products which were tested for carcinogenicity.

Referencemedicinalproduct

ATC code Internationalnonproprietary name(INN)

Pharmacotherapeutic group Results of genotoxicity studies Results of carcinogenicity studies

Mouse Rat

ATC code ADrugs acting on the gastrointestinal system and on the metabolismATC code A02 Drugs for acid related disordersPantozol

controlA02BC02 Pantoprazole Proton pump inhibitors Not genotoxic Positive Positive

ATC code A03 Drugs for functional gastrointestinal disordersResolor A03AE04 Prucalopride Drugs acting on serotonin

receptorsNot genotoxic (mutagenic in Ames testin TA100 strain)

Positive Positive

ATC code A04 Anti-emetics and anti-nauseantsAloxi A04AA05 Palonosetron Antiemetics and antinauseants,

serotonin (5HT3) antagonistsNot genotoxic Negative Positive

Emend A04AD12 Aprepitant Antiemetics and antinauseants Not genotoxic Negative PositiveATC code A08 Anti-obesity preparations, excluding diet productsXenical A08AB01 Orlistat Anti-obesity agents Not genotoxic Negative NegativeATC code A10 Drug used in diabetesLantus A10AE04 Insulin glargine Insulins and analogues for

injection, long-actingNot genotoxic Positive Positive

Avandamet A10BD03 Rosiglitazone/metformin Combinations of oral bloodglucose lowering medicinalproducts

Rosiglitazone not genotoxic Metforminnot gneotoxic

Rosiglitazone negativeMetformin negative

Rosiglitazone positive (benign lipoma)Metformin positive (benign uterus polyps)

Avaglim A10BD04 Rosiglitazone/glimepiride Rosiglitazone not genotoxicGlimepiride not genotoxic

Rosiglitazone negative (positive inAPCmin model)Glimepiride positive

Rosiglitazone positiveGlimipiride positive

Competact A10BD05 Pioglitazone/metformin Pioglitazone not genotoxicMetformin not genotoxic

Pioglitazone negativeMetformin negative

Pioglitazone positiveMetformin positive

Tandemact A10BD06 Pioglitazone/glimepiride Pioglitazone not genotoxicGlimepiride not genotoxic

Pioglitazone negativeGlimepiride positive

Pioglitazone positiveGlimepiride positive

Efficib A10BD07 Sitagliptin/metformin Sitagliptin not genotoxicMetformin not genotoxic

Sitagliptin negativeMetformin negative

Sitagliptin positiveMetformin positive

Eucreas A10BD08 Vildagliptin/metformin Vildagliptin not genotoxicMetformin not genotoxic

Vildagliptin positiveMetformin negative

Vildagliptin negativeMetformin positive

Avandia A10BG02 Rosiglitazone Oral blood glucose loweringdrugs; thiazolidinediones

Not genotoxic Negative Positive

Glustin A10BG03 Pioglitazone Not genotoxic Negative PositiveXelevia A10BH01 Sitagliptin Dipeptidyl peptidase 4 (DPP-4)

inhibitorsNot genotoxic Negative Positive

Galvus A10BH02 Vildagliptin Not genotoxic Positive NegativeOnglyza A10BH03 Saxagliptin Not genotoxic Negative NegativeNovoNorm A10BX02 Repaglinide Other blood glucose lowering

drugs excluding insulinsNot genotoxic Negative Positive

Starlix A10BX03 Nateglinide Not genotoxic Negative NegativeByetta A10BX04 Exenatide Not genotoxic Negative PositiveVictoza A10BX07 Liraglutide Not genotoxic Positive PositiveATC code A16 Other alimentary tract and metabolism productsZavesca A16AX06 Miglustat Other alimentary tract and

metabolism productsNot genotoxic Positive Positive

Kuvan A16AX07 Sapropterin Genotoxic (in vitro) Negative NegativeATC code B Blood and blood forming organsATC code B01 Anti-thrombotic agentsPlavix B01AC04 Clopidogrel Platelet aggregation inhibitors

excluding heparinNot genotoxic Negative Negative

Ventavis B01AC11 Iloprost Not genotoxic Negative NegativeEfient B01AC22 Prasugrel Not genotoxic Positive Negative

(continued on next page)

A.Friedrich,K

.Olejniczak

/Regulatory

Toxicologyand

Pharmacology

xxx(2011)

xxx–xxx

3

Pleasecite

thisarticle

inpress

as:Friedrich,A

.,Olejniczak,K

.Evaluationofcarcinogenicity

studiesofm

edicinalproductsfor

human

useauthorised

viathe

Europeancentralised

procedure(1995–2009).R

egul.Toxicol.Pharmacol.(2011),doi:10.1016/j.yrtph.2011.04.001


Table 1 (continued)




Mouse Rat

ATC code C Cardiovascular systemATC code C01 Cardiac therapyCorlentor C01EB17 Ivabradine Other cardiac preparations Not genotoxic Negative NegativeRanexa C01EB18 Ranolazine Not genotoxic (clastogenic in chromo-

somal aberration test)Negative Positive

ATC code C02 Anti-hypertensivesPolaris C02KX02 Ambrisentan Other antihypertensives Not genotoxic Negative NegativeTracleer C02KX01 Bosentan Not genotoxic Positive PositiveThelin C02KX03 Sitaxentan sodium Not genotoxic (clastogenic in vitro at

cytotoxic concentrations)p53+/� model negative Positive

ATC code C03 DiureticsSamsca C03XA01 Tolvaptan Vasopressin antagonists Not genotoxic Negative NegativeATC code C09 Agents acting on the renin-angiotensin systemAprovel C09CA04 Irbesartan Angiotensin II antagonists Not genotoxic Negative NegativeMicardis C09CA07 Telmisartan Not genotoxic Negative NegativeCoAprovel C09DA04 Irbesartan/

hydrochlorothiazideAngiotensin II antagonists,combinations

Hydrochlorothiazide equivocal Irbesartan negativehydrochlorothiazide positive

Irbesartan negativehydrochlorothiazide negative

Micardis Plus C09DA07 Telmisartan/hydrochlorothiazide

Angiotensin II antagonists,diuretics

Telmisartan not genotoxichydrochlorothiazide equivocal

Telmisartan negativehydrochlorothiazide positive

Telmisartan negativehydrochlorothiazide negative

Exforge C09DB01 Amlodipine besylate/valsartan

Calcium channel blockers,angiotensin II antagonists

Amlodipine not genotoxic Valsartan notgenotoxic

Amlodipine negative Valsartannegative

Amlodipine negativevalsartan negative

Exforge HCT C09DX01 Amlodipine besylate/valsartan/hydrochlorothiazide

Angiotensin II antagonists, plain(valsartan), combinations withdihydropyridine derivatives(amlodipine) and thiazidediuretics (hydrochlorothiazide)

Amlodipine not genotoxic Valsartan notgenotoxic Hydrochlorothiazideequivocal

Amlodipine negative Valsartannegative Hydrochlorothiazidepositive

Amlodipine negative valsartan negativehydrochlorothiazide negative

Riprazo C09XA02 Aliskiren Renin inhibitors Not genotoxic Alternative model negative NegativeRasilez HCT Aliskiren hemifumarate/

hydrochlorothiazideRenin inhibitors, combinationswith diuretics

Aliskrien not genotoxichydrochlorothiazide equivocal

Aliskrien alternative modelnegative Hydrochlorothiazidepositive

Aliskrien negativehydrochlorothiazide negative

ATC code C10 Lipid modifying agentsCholestagel C10AC04 Colesevelam Bile acid sequestrants Not genotoxic Negative (low survival rate) PositiveTredaptive C10AD52 Nicotinic acid/laropiprant Nicotinic acid and derivatives Nicotinic acid not genotoxic (literature)

laropiprant not genotoxicNicotinic acid negative (literature)laropiprant positive

Nicotinic acid negative (literature)laropiprant negative

ATC code D DermatologicalsATC code D06 Antibiotics and chemotherapeutics for dermatological useAldara D06BB10 Imiquimod Chemotherapeutics for topical

use, antiviralNot genotoxic Topical negative Not performed

ATC code D11 Other dermatological preparationsProtopic D11AX14 Tacrolimus Other dermatologicals Not genotoxic Topical positive PhotoCA positive

oral negativeTopical not performedoral negative

Vaniqa D11AX16 Eflornithine Not genotoxic Topical negativephotoCA negativeoral negative

Topical not performedoral negative

ATC code G Genito-urinary system and sex hormonesATC code G02 Other gynecologicalsTractocile G02CX01 Atosiban Other gynaecologicals Not genotoxic Not performed PositiveATC code G03 Sex hormones and modulators of the genital systemEVRA G03AA13 Norelgestromin/ethinyl

estradiolNorelgestromin and estrogen Norelgestromin not genotoxic

ethinyl estradiol not genotoxicNot performed Combination of norgestimate and ethinyl

estradiol positivecombination negative in a 10-year monkey

4A

.Friedrich,K.O

lejniczak/R

egulatoryToxicology

andPharm

acologyxxx

(2011)xxx–

xxx

Pleasecite

thisarticle

inpress

as:Friedrich,A

.,Olejniczak,K

.Evaluation

ofcarcinogenicitystudies

ofmedicinalproducts

forhum

anuse

authorisedvia

theEuropean

centralisedprocedure

(1995–2009).Regul.Toxicol.Pharm

acol.(2011),doi:10.1016/j.yrtph.2011.04.001


Table 1 (continued)




Mouse Rat

studyEvista G03XC01 Raloxifene hydrochloride Selective oestrogen receptor

modulator (SERM)Not genotoxic Negative Negative

Fablyn G03XC01 Lasofoxifene Not genotoxic Positive PositiveConbriza G03XC02 Bazedoxifene Not genotoxic TgHras2 model positive PositiveATC code G04 UrologicalsEmselex G04BD10 Darifenacin Urinary antispasmodics Not genotoxic Negative Positive (considered not treatment-related)Toviaz G04BD11 Fesoterodine Not genotoxic Negative NegativeViagra

RevatioG04BE03 Sildenafil Drugs used in erectile

dysfunctionNot genotoxic Negative Negative

Cialis G04BE08 Tadalafil Not genotoxic Negative (numerical increase) Negative (numerical increase)Levitra G04BE09 Vardenafil Not genotoxic Negative NegativeATC code H Systemic hormonal preparations, excluding sex hormones and insulinsATC code H01 Pituitary and hypothalamic hormones and analoguesIncrelex H01AC03 Mecasermin Somatropin and somatropin

agonistsNot genotoxic Not performed Positive

ATC code H05 Calcium homeostasisForsteo H05AA02 Teriparatide Parathyroid hormones and

analoguesNot genotoxic Not performed Positive

Preotact H05AA03 Parathyroid hormone(rDNA)

Not genotoxic Not performed Positive

Mimpara H05BX01 Cinacalcet Anti-parathyroid agents Not genotoxic Negative NegativeATC code J Anti-infectives for systemic useATC code J02 Anti-mycotics for systemic useVfend J02AC03 Voriconazole Triazole derivatives Not genotoxic Positive PositiveNoxafil J02AC04 Posaconazole Not genotoxic Negative PositiveATC code J05 Antivirals for systemic useRebetol J05AB04 Ribavirin Nucleosides and nucleotides

excluding reverse transcriptaseinhibitors

Genotoxic Negative Positive (upper range of historical controls)

Invirase J05AE01 Saquinavir Protease inhibitors Not genotoxic Negative NegativeCrixivan J05AE02 Indinavir Not genotoxic Negative PositiveNorvir J05AE03 Ritonavir Not genotoxic Positive NegativeViracept J05AE04 Nelfinavir Not genotoxic Negative PositiveAgenerase J05AE05 Amprenavir Not genotoxic Positive PositiveKaletra J05AE06 Lopinavir/Ritonavir Not genotoxic Combination positive Combination negativeTelzir J05AE07 Fosamprenavir Not genotoxic Positive PositiveReyataz J05AE08 Atazanavir sulphate Not genotoxic (clastogenic in chromo-

somal aberration test)Positive Negative

Aptivus J05AE09 Tipranavir Not genotoxic Positive PositivePrezista J05AE10 Darunavir Not genotoxic Positive PositiveZerit J05AF04 Stravudine Nucleoside and nucleotide

reverse transcriptase inhibitorsGenotoxic Positive Positive

Epivir J05AF05 Lamivudine Genotoxic Negative NegativeZiagen J05AF06 Abacavir sulphate Genotoxic Positive PositiveViread J05AF07 Tenofovir disoproxil Genotoxic Positive Positive (benign lipoma, within historical

control range)Hepsera J05AF08 Adefovir dipivoxil Genotoxic Negative NegativeEmtriva J05AF09 Emtricitabine Not genotoxic Negative NegativeBaraclude J05AF10 Entecavir Genotoxic (clastogenic in chromo-somal

aberration test)Positive Positive

Sebivo J05AF11 Telbivudine Not genotoxic TgHras2 model negative PositiveViramune J05AG01 Nevirapine Not genotoxic Positive Positive


A.Friedrich,K

.Olejniczak

/Regulatory

Toxicologyand

Pharmacology

xxx(2011)

xxx–xxx

5

Pleasecite

thisarticle

inpress

as:Friedrich,A

.,Olejniczak,K


studiesofm

edicinalproductsfor

human

useauthorised

viathe

Europeancentralised




Table 1 (continued)




Mouse Rat

Stocrin J05AG03 Efavirenz Non-nucleoside reversetranscriptase inhibitors

Not genotoxic Positive Negative

Intelence J05AG04 Etravirine Not genotoxic Positive NegativeTamiflu J05AH02 Oseltamivir Neuraminidase inhibitors Not genotoxic Negative Dermal Tg.AC model

negativePositive (trend)

Combivir J05AR01 Lamivudine/zidovudine Antivirals for treatment of HIVinfections, combinations

Lamivudine genotoxic Zidovudinegenotoxic

Lamivudine negativezidovudine positive

Lamivudine negative Zidovudine positive

Kivexa J05AR02 Abacavir sulphate/lamivudine

Abacavir genotoxiclamuvudine genotoxic

Abacavir positivelamivudine negative

Abacavir positive Lamivudine negative

Truvada J05AR03 Emtricitabine/tenofovirdisoproxil

Emtricitabine not genotoxicTenofovir genotoxic

Emtricitabine negativetenofovir positive

Emtricitabine negative Tenofovir negative

Trizivir J05AR04 Abacavir sulphate/lamivudine/zidovudine

Lamivudine genotoxiczidovudine genotoxicabacavir genotoxic

Lamivudine negativezidovudine positiveabacavir positive

Lamivudine negativezidovudine positive abacavir positive

Atripla J05AR06 Efavirenz/emtricitabine/tenofovir disoproxil

Efavirenz not genotoxicemtricitabine not genotoxictenofovir genotoxic

Efavirenz positiveemtricitabine negativetenofovir positive

Efavirenz negativeemtricitabine negativetenofovir negative

Isentress J05AX08 Raltegravir Other antivirals Not genotoxic Negative NegativeCelsentri J05AX09 Maraviroc Not genotoxic TgHras2 model negative PositiveATC code L Anti-neoplastic and immunomodulating agentsATC code L01 Anti-neoplastic agentsXeloda L01BC06 Capecitabine Antimetabolites Genotoxic Negative Not performedIressa L01XE02 Gefitinib Protein kinase inhibitors Not genotoxic Positive PositiveAfinitor L01XE10 Everolimus Not genotoxic Positive NegativeOnsenal L01XX33 Celecoxib Other antineoplastic agents Not genotoxic Positive (within historical control

values)Positive (within historical control values)

ATC code L02 Endocrine therapyFareston L02BA02 Toremifene Anti-oestrogens Not genotoxic Positive NegativeFaslodex L02BA03 Fulvestrant Not genotoxic Positive (literature) PositiveFirmagon L02BX02 Degarelix Other hormone antagonists and

related agentsNot genotoxic Positive Positive

ATC code L04 ImmunosuppressantsCellCept L04AA06 Mycophenolate mofetil Selective immunosuppressive

agentsGenotoxic Negative Negative

Arava L04AA13 Leflunomide Leflunomide not genotoxicminor metabolite genotoxic in vitro

Positive Negative

Rapamune L04AA10 Sirolimus Calcineurin inhibitors Not genotoxic Positive PositiveOrencia L04AA24 Abatacept Not genotoxic Positive Not performedAdvagraf L04AD02 Tacrolimus Not genotoxic Oral negative

topical not performedOral negativetopical positive

Modigraf L04AD02 Tacrolimus Other immunosuppressiveagents

Not genotoxic Oral negativetopical not performed

Oral negativetopical positive

ThalidomideCelgene

L04AX02 Thalidomide Not genotoxic Negative Negative

ATC code M Musculo-skeletal systemATC code M05 Drugs for treatment of bone diseasesZometa M05BA08 Zoledronic acid Bisphosphonates Not genotoxic Positive NegativeBondenza M05BA06 Ibandronic acid Not genotoxic Negative NegativeFosavance M05BB03 Alendronate sodium/

ColecalciferolBisphosphonates, combinations Alendronate not genotoxic Colecalciferol

not genotoxicAlendronate negativecolecalciferol not performed

Alendronate negativecolecalciferol not performed

Osigraft M05BC02 Eptotermin alfa Bone morphogenetic protein Not genotoxic (in vitro) Not performed Positive

6A

.Friedrich,K.O

lejniczak/R

egulatoryToxicology

andPharm

acologyxxx

(2011)xxx–

xxx

Pleasecite

thisarticle

inpress

as:Friedrich,A

.,Olejniczak,K

.Evaluation


ofmedicinalproducts

forhum

anuse

authorisedvia

theEuropean



acol.(2011),doi:10.1016/j.yrtph.2011.04.001


Table 1 (continued)




Mouse Rat

Osseor M05BX03 Strontium ranelate Other drugs affecting bonestructure and mineralisation

Not genotoxic Negative Positive

ATC code N Nervous systemATC code N01 AnestheticsQutenza N01BX04 Capsaicin Other local anesthetics Not genotoxic (weakly positive in mouse

lymphoma assay)Dermal Tg.AC model negative Not performed

ATC code N03 Anti-epilepticsInovelon N03AF03 Rufinamide Carboxamide derivatives Not genotoxic Positive NegativeExalief N03AF04 Eslicarbazepine acetate Not genotoxic (clastogenic in some

in vitro assays)Positive Not performed

Keppra N03AX14 Levetiracetam Other antiepileptics Not genotoxic Negative NegativeZonegran N03AX15 Zonisamide Not genotoxic Negative NegativeLyrica N03AX16 Pregabalin Not genotoxic Positive NegativeDiacomit N03AX17 Stiripentol Not genotoxic (clastogenic in vitro at

cytotoxic concentrations)Positive Negative

Vimpat N03AX18 Lacosamide Not genotoxic (clastogenic in mouselymphoma assay)

Negative Negative

ATC code N04 Anti-Parkinson drugsStalevo N04BA03 Levodopa/carbidopa/

entacaponeDopa and dopa derivatives Entacapone genotoxic (clatogenic

in vitro)levodopa/Carbidopa not genotoxic(literature)

Entacapone negativelevodopa/carbidopa not performed

Entacapone positivelevodopa/carbidopa negative (literature)

Sifrol N04BC05 Pramipexole Dopamine agonists Not genotoxic Negative PositiveNeupro N04BC09 Rotogotine Not genotoxic (clastogenic in mouse

lymphoma assay)Negative Positive

Azilect N04BD02 Rasagiline Monoamine oxidase B inhibitors Genotoxic (clastogenic in vitro) Positive NegativeTasmar N04BX01 Tolcapone Anti-Parkinson agents Not genotoxic Negative PositiveComtess N04BX02 Entacapone Genotoxic (clastogenic in vitro) Negative PositiveATC code N05 PsycholepticsZyprexa N05AH03 Olanzapine Diazepines, oxazepines and

thiazepinesNot genotoxic Positive Positive

Zypadhera N05AH03 Olanazapine Not genotoxic Depot form not performed(technical reasons)

Depot form negative

Abilify N05AX12 Aripiprazole Other antipsychotics Not genotoxic Positive PositiveInvega N05AX13 Paliperidone Not genotoxic Positive PositiveZerene N05CF03 Zaleplon Benzodiazepine related drugs Not genotoxic Negative NegativeCircadin N05CH01 Melatonin Melatonin receptor agonists Not genotoxic Tg NK model negative PositiveATC code N06 PsychoanalepticsExelon N06DA03 Rivastigmine Anticholinesterases Not genotoxic Negative NegativeAriclaim N06AX21 Duloxetine Other antidepressants Not genotoxic Negative Negative (multinucleated liver cells)Thymanax N06AX22 Agomelatine Not genotoxic (clastogenic in chromo-

somal aberration test)Positive Positive

Ebixa N06DX01 Memantine Other anti-dementia drugs Not genotoxic Negative NegativeATC code N07 Other nervous system drugsChampix N07BA03 Varenicline tartrate Active substances used in

nicotine dependenceNot genotoxic Negative Positive

Suboxone N07BC51 Buprenorphinehydrochloride/naloxonehydrochloride

Drugs used in opioiddependence

Combination not genotoxic Not performed Combination positive

Rilutek N07XX02 Riluzole Other nervous system drugs Major active metabolite genotoxic(clastogenic in vitro)

Negative Negative


A.Friedrich,K

.Olejniczak

/Regulatory

Toxicologyand

Pharmacology

xxx(2011)

xxx–xxx

7

Pleasecite

thisarticle

inpress

as:Friedrich,A

.,Olejniczak,K


studiesofm

edicinalproductsfor

human

useauthorised

viathe

Europeancentralised




Table 1 (continued)




Mouse Rat

Xyrem N07XX04 Sodium oxybate Not genotoxic Sodium oxybate not performed c-butyrolactone equivocal (NTPstudy)

Sodium oxybate positive (upper rangehistorical control values) c-butyrolactonenegative

ATC code R Respiratory systemATC code R01 Nasal preparationsAvamys R01AD12 Fluticasone furoate Corticosteroids Not genotoxic Intransal negative Intransal negativeATC code R06 Antihistamines for systemic useAerinaze R06AX27;

R01BA52Desloratadine andPseudoephedrine (assulphate)

Antihistamines-H1 antagonist;nasal decongestants for systemicuse group

Not genotoxic Negative Not performed

ATC code S Sensory organsATC code S01 OphthalmologicalsAzopt S01EC04 Brinzolamide Antiglaucoma preparations and

miotics, carbonic anhydraseinhibitors

Not genotoxic (clastogenic in mouselymphoma assay)

Oral positive Oral negative

DuoTrav S01ED51 Travoprost/timolol Antiglaucoma preparations andmiotics-beta-blocking agents-timolol, combinations

Travoprost not genotoxic Timololgenotoxic (low potential)

Subcutaneous travoprost negativeoral timolol positive

Subcutaneous travoprost negativeoral timolol positive

Ganfort S01ED51 Bimatoprost/timolol Bimatoprost not genotoxic Timololgenotoxic (low potential)

Oral bimatoprost negativeoral timolol positive

Oral bimatoprost negativeoral timolol positive

Azarga S01ED51 Brinzolamide/timolol Brinzolamide not genotoxictimolol genotoxic (low potential)

Oral brinzolamide positiveoral timolol positive

Oral brinzolamide negativeoral timolol positive

Lumigan S01EE03 Bimatoprost Antiglaucoma preparations andmiotics – prostaglandinanalogues

Not genotoxic Negative Negative

Travatan S01EE04 Travoprost Not genotoxic Negative NegativeEmadine S01GX06 Emedastine Decongestants and antiallergics;

other antiallergicsNot genotoxic Negative (numerical increase) Negative

Opatanol S01GX09 Olopatadine Not genotoxic Negative NegativeATC code V VariousATC code V03 All other therapeutic productsRenvela V03AE02 Sevelamer (carbonate) Treatment of

hyperphosphataemiaNot genotoxic (clastogenic inchromosomal aberration test)

Negative (within historical controlrange)

Positive

Exjade V03AC03 Deferasirox Iron chelating agents Not genotoxic p53+/� model negative NegativeNot yet classifiedMultaq – Dronedarone – Not genotoxic Positive PositiveOnbrez

Breezhaler– Indacaterol Long-acting, ß2-adrenergic

agonistNot genotoxic TgHras2 model negative Positive

ATC: Anatomical Therapeutic Chemical classification system.NTP: National Toxicology Program.PhotoCA: Photocarcinogenicity.rDNA: Recombinant DNA.

8A

.Friedrich,K.O

lejniczak/R

egulatoryToxicology

andPharm

acologyxxx

(2011)xxx–

xxx

Pleasecite

thisarticle

inpress

as:Friedrich,A

.,Olejniczak,K

.Evaluation


ofmedicinalproducts

forhum

anuse

authorisedvia

theEuropean



acol.(2011),doi:10.1016/j.yrtph.2011.04.001


A. Friedrich, K. Olejniczak / Regulatory Toxicology and Pharmacology xxx (2011) xxx–xxx 9

It is important to note that this analysis is based on informationavailable for centrally approved pharmaceuticals in the EEA anddoes not consider carcinogenicity studies for medicinal productsthat were approved via national procedures. Compounds whichwere tested for carcinogenicity by pharmaceutical companies butwhich have not been pursued or approved for marketing were alsonot considered in this analysis.

All medicinal products authorised via the CP between 1995 and2009 were analysed with respect to the number of products andactive substances with genotoxicity and carcinogenicity data, thetests systems used for their assessment, and the potential rele-vance of the data for the human situation and their impact onthe labelling.

Table 2Categorisation of active substances (INN) with carcinogenicity data according to typeand number of studies performed.

Active substances with carcinogenicity data Number %

Two long-term carcinogenicity studies 116 81One long-term carcinogenicity study in rats and one

transgenic mouse study8 5.5

One long-term carcinogenicity study in mice or rats 13 8.5One transgenic mouse model 1 1No carcinogenicity studies performed 6 4

Total 144 100

3. Results

3.1. Overview of carcinogenicity data

3.1.1. Medicinal products authorised via the centralised procedureAccording to the alphabetical listing of medicinal products that

have an EPAR on the EMA website, the total number of centrallyauthorised medicinal products until end of 2009 was 521. Out ofthe 521 authorised medicinal products, 146 (28%) products ac-count for duplicate, informed consent, generic or biosimilar prod-ucts, for which no new carcinogenicity data have been generated.Carcinogenicity data were available for 280 medicinal productsaccounting for 54% of all authorised products.

3.1.2. Fixed combination medicinal productsFixed combinations are used in oral blood glucose lowering

medicinal products, in blood pressure lowering products and top-ical antiglaucoma preparations (see Table 1 for reference prod-ucts). All active substances contained in these fixed combinationmedicinal products have been individually tested for genotoxicityand carcinogenicity.

In the treatment of human immunodeficiency virus (HIV-1) in-fected patients, the combined use at least three active substances iscurrently considered essential based on the inherent high mutationrate in HIV (guideline on carcinogenicity evaluation of medicinalproducts for the treatment of HIV infection). Therefore, fixed com-binations have been developed aiming at improving adherence byreducing pill burden (Guideline on the clinical development ofmedicinal products for the treatment of HIV infection) (see Table 1for reference products). With the exception of lopinavir/ritonavir,which were tested in combination, genotoxicity and carcinogenic-ity studies were performed using the individual active substances.This is in accordance with the guideline on carcinogenicity evalu-ation of medicinal products for the treatment of HIV infection,which does not require carcinogenicity studies of drug combina-tion, if the individual components have been adequately tested.

Active substances that have been tested for carcinogenicity incombination include norgestimate/ethinylestradiol and buprenor-phin hydrochloride/naloxon hydrochloride (see Table 1 for refer-ence products).

3.1.3. Reference medicinal productsReference medicinal products included medicinal products con-

taining new chemical entities (NCEs), combinations of NCEs, NCEsin combination with known active substances or products contain-ing known active substances developed for new indications.

375 centrally authorised reference medicinal products had a va-lid MA a by the end of 2009. Carcinogenicity data were available for144 individual compounds or fixed combinations that were con-tained in 159 reference medicinal products (Table 1). The highernumber of reference medicinal products compared with the num-


ber of active substances is either due to the combination of activesubstances in different products, e.g. hydrochlorothiazide in com-bination with irbesartan (CoApprovel) and telmisartan (MicardisPlus), or due to the fact that some compounds have been approvedin different indication and/or are available in different formulation,e.g. the use of sildenafil in Viagra and Revatio for the treatment oferectile dysfunction and arterial pulmonary hypertension, respec-tively, and the use of tacrolimus in Protopic and Advagraf for top-ical treatment of atopic dermatitis and systemic treatment oftransplant rejection following organ transplantation, respectively.

Thus, from all reference medicinal products authorised via theCP between 1995 and 2009, carcinogenicity data have been gener-ated for approximately 42% (159/375). Products that have not beentested for carcinogenicity include biotechnology-derived pharma-ceuticals such as endogenous peptides and proteins, monoclonalantibodies and vaccines produced by recombinant DNA technol-ogy. This is consistent with current requirements (ICH S1A guide-line; ICH S6 guideline). Furthermore, carcinogenicity testing isnot required for products intended for short-term clinical useand for anti-neoplastic agents for the treatment of patients withshort life expectancy.

Carcinogenicity studies in at least one rodent species are avail-able for 138 compounds that were either tested individually (in themajority of cases) or in combination. Six additional compoundsexhibited neoplastic lesions in repeat-dose toxicity studies and itwas thus considered that carcinogenicity studies would not needto be performed for these compounds (Table 2).

Conventional long-term carcinogenicity studies in mice and ratsare available for the majority of compounds, while transgenicmouse models were infrequently used in carcinogenicity testing.A number of compounds were tested for carcinogenicity in onlyone rodent species (Table 2).

Photocarcinogenicity data have been provided for two com-pounds intended for dermal use (ATC code D11). The studies wereperformed in addition to conventional long-term rodent carcinoge-nicity studies using the systemic and dermal route of administra-tion (see Table 1).

Out of the 144 compounds tested for carcinogenicity, 50 (35%)yielded negative results and 94 (65%) were positive in at least onecarcinogenicity study or in repeat-dose toxicity studies (Table 3).

Two long-term carcinogenicity studies were available for 116compounds, 44 of which were negative and 32 of which were po-sitive in both mice and rats. Twenty-two compounds were positivein rats only and 18 were positive exclusively in mice (Table 3).

For 13 compounds, carcinogenicity data were available in onlyone rodent species. Within this group, three compounds were neg-ative in mice and one compound in rats, while two and seven com-pounds were positive in mice and rats, respectively (Table 3). Onefixed combination product for contraceptive treatment (ATC codeG3) yielded positive results in rats, but was shown to be negativein a monkey study of 10-year duration (see Table 1).

Transgenic mouse models were infrequently used as a replace-ment of a long-term carcinogenicity study. Data from one long-term carcinogenicity study in rats and a transgenic mouse model



Table 3Categorisation of active substances (INN) with carcinogenicity data according to studyresults.

Active substances with carcinogenicity data Number %

All compounds 144 100- Negative in mice and/or rats- Positive in mice and/or rats

50 3594 65

Two long-term carcinogenicity studies 116 80.5- Negative in mice and rats- Positive in mice and rats- Negative in mice and positive in rats- Positive in mice and negative in rats

44 30.532 2222 1518 12.5

One long-term carcinogenicity study in rats and onetransgenic mouse study

8

- Negative in mice and rats- Positive in mice and rats- Positive in rats and negative in mice- Negative in rats and positive in mice

2150

One long-term carcinogenicity study in mice or rats 13- Negative in mice- Negative in rats- Positive in mice- Positive in rats

3127

One transgenic mouse study 1- Negative- Positive

10


were available for eight compounds. The TgrasH2 mouse modelwas used in four studies followed by the p53+/� model whichwas used twice.

Two compounds showed negative results in the long-term car-cinogenicity study in rats and the transgenic mouse study, whilefive were positive in the long-term rat study and negative in thetransgenic mouse study. Only one compound was positive in boththe long-term rat study and the transgenic mouse study (Table 3).

For one compound, the only carcinogenicity data available werefrom a transgenic mouse study, which showed negative results(Table 3).

Of the two compounds tested for photocarcinogenicity in the al-bino hairless mouse model, one exhibited a negative and one a po-sitive result.

With regard to the rodent species that produced positive carcin-ogenicity findings, 33 out of 94 compounds were positive in bothmice and rats (35%), 40 were positive in rats (43%) and 21 (22%)were positive in mice.

3.2. Overview of genotoxicity data

Out of the 144 compounds with carcinogenicity data, 114 wereclearly negative in the standard battery of genotoxicity tests (ICHS2B guideline) or in a more comprehensive test battery. Twelvecompounds had positive findings in one or more in vitro test(s);however, when considering the results of all genotoxicity tests,the weight of evidence suggested that these compounds had nogenotoxic potential. For 18 compounds, the overall results of geno-toxicity tests indicated that the compounds may have intrinsicgenotoxic properties (Table 4). A more detailed discussion of the

Table 4Genotoxicity findings for active substances (INN) with carcinogenicity data.

Active substances with carcinogenicity data

All compounds- Negative in a battery of genotoxicity tests- Positive in one or more genotoxicity test(s)- Compounds for which the overall results of genotoxicity tests suggested that the- Compounds for which the overall results of genotoxicity tests suggested that the


compounds with positive or equivocal genotoxicity findings canbe found in section 3.3.2.1.

3.3. Individual active substances tested for carcinogenicity

The individual 138 compounds tested for carcinogenicity areshown in Table 1.

In Table 1, active substances (international nonproprietaryname, INN) and their reference medicinal products are groupedaccording to their ATC code and pharmacotherapeutic group. Boththe results of genotoxicity and carcinogenicity studies arepresented.

In term of genotoxicity, the overall assessment of genotoxicityas stated in the EPAR and/or SPC is presented for each individualcompound. However, positive findings in individual genotoxicitystudies were also indicated.

With regard to tumour findings in carcinogenicity studies inmice and rats, both benign and malignant neoplasms have beenconsidered in the evaluation. Carcinogenicity studies were as-sessed as positive if statistically significant increases in tumourincidences were observed. Studies with statistically significant tu-mour increases that fell within the range of historical control datawere classified as positive.

The six compounds exhibiting neoplastic lesions in repeat-dosetoxicity studies are shown in Table 5.

3.3.1. Active substances with negative carcinogenicity findingsOut of the 144 compounds with carcinogenicity data, 50 yielded

negative results (Table 3). Forty-four compounds were negative inlong-term carcinogenicity studies in mice and rats, while two com-pounds were negative in a long-term study in rats and a transgenicmouse model (Table 3). For four compounds, negative long-termcarcinogenicity data were available in only one rodent species (Ta-ble 3). One compound was exclusively tested in a transgenic mousemodel and was shown to be negative (Table 3). The individualcompounds with negative carcinogenicity findings are listed inTable 6.

3.3.2. Active substances with positive carcinogenicity findingsOut of the 144 compounds tested for carcinogenicity, 94 (65%)

were positive in at least one carcinogenicity study or in repeat-dose toxicity studies (Table 3).

3.3.2.1. Compounds with positive or equivocal genotoxicity find-ings. The majority of compounds evaluated for carcinogenicitywere found to be devoid of genotoxic properties (Table 4). For 18compounds, the EPARs concluded that they may have intrinsicgenotoxic properties according to the results of the genotoxicitytesting battery (Table 7). However, a carcinogenic risk for humanswas excluded for those compounds with negative results inlong-term rodent carcinogenicity studies and for compoundswith rodent-specific tumour findings in carcinogenicity studies(Table 7). The majority of the compounds with positive or equivo-cal genotoxicity tests and positive long-term rodent carcinogenic-ity studies are anti-retroviral agents for the treatment of HIV-1infections and belong to the two classes of anti-retroviral agents

Number %

144 100

y have no genotoxic potentialy have intrinsic genotoxic properties

114 7930 2112 10.518 12.5



Table 5Reference medicinal products for which neoplastic changes were observed in repeat-dose toxicity studies.

Referencemedicinal product

ATC code Internationalnonproprietaryname (INN)

Pharmacotherapeutic group Results ofgenotoxicitystudies

Results of repeat-dose toxicity studies

ATC code A10 Drug used in diabetesNovorapid A10AB05 Insulin aspart Insulins and analogues for injection,

fast-actingNot genotoxic Positive in 12-month toxicity study in rats

Protaphane A10AC01 Insulin human(rDNA)

Insulins and analogues for injection,intermediate-acting, insulin (human)

Not genotoxic Positive in 12-month toxicity study in rats

ATC code J02 Anti-mycotics for systemic useMycamine J02AX05 Micafungin Other antimycotics for systemic use Not genotoxic Foci of altered hepatocytes in 6-month toxicity

study in rats - developing into tumours duringrecovery

ATC code J05 Antivirals for systemic useVistide J05AB12 Cidofovir Antivirals for systemic use Genotoxic Positive in repeat-dose toxicity studies in rats

ATC code L01 Anti-neoplastic agentsXagrid L01XX35 Anagrelide Other antineoplastic agents Not genotoxic Positive in 12-month toxicity study in rats

ATC code V03 All other therapeutic productsSavene V03AF02 Dexrazoxane Detoxifying agents for antineoplastic

treatmentGenotoxic Positive in 12-month toxicity study in mice

and rats (NCI study)

NCI: National Cancer Institute.


that are known to be clastogenic, i.e. the nucleoside/nucleotide re-verse transcriptase inhibitors and the non-nucleoside reversetranscriptase inhibitors. The carcinogenicity findings observed forthese compounds and their potential relevance to humans accord-ing to the EPAR and SPC are presented in Table 8.

3.3.2.2. Compounds with rodent-specific or species-specific tumourfindings. A high number of tumour findings in rodents wereconsidered to be of no relevance for humans since a species- orrodent-specific mechanism has been identified (Greaves, 2007).These findings are unlikely to pose any carcinogenic risk tohumans. Active substances (INN) for which a species- or rodent-specific mechanism was identified in carcinogenicity studies arepresented in Table 9.

Liver tumours in mice and rats and thyroid gland follicular celltumours in rats as a consequence of hepatic microsomal enzymeinduction were found most frequently, followed by tumours ofthe endocrine system including endocrine glands (pituitary gland,adrenal gland), male and female reproductive organs (testes, ovary,uterus) and mammary gland for which rodents are particularlysensitive (Greaves, 2007).

Most of the mechanisms identified were either relevant for bothmice and rats, e.g. liver tumours due to hepatic enzyme inductionor tumours of endocrine and reproductive system due to distur-bances of the hypothalamic–pituitary–gonadal axis, or they wereunique to the rat, e.g. thyroid gland follicular cell tumours. Onlyfew tumours were considered to be mouse-specific (Greaves,2007) (Table 9).

For 38/94 (40%) of the compounds listed in Table 9, tumourfindings were exclusively attributable to a rodent-specific mecha-nism. For these compounds, regulatory actions are not requiredand wordings indicating that the compounds pose no carcinogenicrisk to humans or that the relevance of the tumour findings to hu-mans is minor or limited can be found in the SPC Section 5.3 (fordetails see Table 10).

3.3.2.3. Compounds with tumour findings due to exaggerated phar-macodynamic effects or secondary to toxic damage. Tumour develop-ment related to long-term administration of supraphysiological ortoxic doses of compounds is frequently observed in rodent carcin-ogenicity studies. Active substances (INN) which fall into this cat-egory are shown in Table 11. Examples include liver and kidneytumours secondary to hepatic or renal toxicity and lymphomas fol-


lowing administration of immunosuppressive compounds(Greaves, 2007). The mechanisms of carcinogenicity found in ro-dents are principally also relevant to humans, however, it is unli-kely that such tumours will be induced at therapeutic doses.

For many compounds of this group, a sufficiently high safetymargin in terms of systemic exposure has been demonstrated intoxicokinetic evaluations (see Section 3.3.2.6). No safety marginhas been established for micafungin, tacrolimus, leflunomide, sirol-imus, abatacept and amprenavir. The potential relevance of the tu-mour findings in animals has been included in SPC sections 4.4, 4.8and/or 5.3 (for details see Table 10).

3.3.2.4. Compounds with photo co-carcinogenic potential. Topicalapplication of tacrolimus (Protopic) shortened the time to skin tu-mour development induced by UV radiation in albino hairlessmice. The underlying mechanism was considered to be systemicimmunosuppression based on high exposure levels observed inthe animals.

It was included in the SPC Section 5.3 that a risk for humanscannot be completed ruled out as the potential for local immuno-suppression with the long-term use of tacrolimus is unknown. Awarning to avoid exposure of the skin to sunlight during use ofthe ointment was included in Section 4.4 of the SPC (see Table 10).

3.3.2.5. Compounds with tumour findings for which the relevance forhumans could not be established. This heterogeneous group com-prises active substances (INN) with carcinogenicity findings thatwere either not considered treatment-related or compounds withtumour findings of unknown relevance to humans. An assessmentof tumour findings for all compounds with positive carcinogenicitystudies and their potential relevance to humans and the corre-sponding SPC wording is presented in Table 10.

3.3.2.6. Compounds with a high safety margin in terms of systemicexposure. For a considerable number of active substances (INN),systemic exposures in rodents as determined in toxicokineticassessments were found to be sufficiently in excess of exposurelevels at human therapeutic doses (Table 12). Therefore, it wasconsidered that no regulatory actions are required for these medic-inal products. Similar statements were included in the SPC Sec-tion 5.3 to indicate the large differences in exposure betweenrodents and humans at therapeutic doses and the low relevance



Table 6Active substances (INN) with negative carcinogenicity data.

Active substance

Negative in long-term carcinogenicity studies in mice and ratsOrlistatSaxagliptinNateglinideSapropterinClopidogrelIloprostIvabradineAmbrisentanTolvaptanIrbesartanTelmisartanAmlodipine besylateValsartanNicotinic acidEflornithineRaloxifene hydrochlorideFesoterodineSildenafilTadalafilVardenafilCinacalcetSaquinavirLamivudineAdefovir dipivoxilEmtricitabineRaltegravirMycophenolate mofetilThalidomideIbandronic acidAlendronate sodiumLevetiracetamZonisamideLacosamideZaleplonRivastigmineDuloxetineMemantineRiluzoleFluticasone furoateTravoprostBimatoprostEmedastineOlopatadine

Negative in a long-term carcinogenicity studyin rats and a transgenic mouse study

AliskirenDeferasirox

Negative in a long-term carcinogenicity study in miceImiquimodCapecitabineDesloratadine and Pseudoephedrine

Negative in a long-term carcinogenicity study in ratsLevodopa/Carbidopa

Negative in a transgenic mouse studyCapsaicin

Table 7Active substances (INN) with positive or equivocal genotoxicity findings.

Active substance Carcinogenic potential

Lamivudine Negative in long-term rodentcarcinogenicity studiesAdefovir dipivoxil

Mycophenolate mofetilCapecitabineRiluzoleSapropterin

Entacapone Rat-specific mechanism of carcinogenicityidentified

Ribavirin Positive in long-term rodentcarcinogenicity studiesCidofovir

StravudineAbacavir sulphateTenofovir disoproxilEntecavirZidovudineDexrazoxaneRasagilineHydrochlororothiazideTimolol


of the tumours observed in carcinogenicity studies (for details seeTable 10).

3.3.2.7. Product specific assessments of carcinogenic potential. As sta-ted in the ICH S6 guideline, standard rodent carcinogenicity studiesare generally inappropriate for biotechnology-derived pharmaceu-ticals. However, an assessment of the carcinogenic potential maybe needed depending on the duration of treatment and the biolog-ical activity of the products, e.g. monoclonal antibodies targetingimmune functions for the treatment of chronic diseases like rheu-matoid arthritis or psoriasis.

When there is a concern about the carcinogenic potential ofsuch products, a variety of approaches may be considered to eval-


uate the risk. Examples include carcinogenicity testing with a mur-ine analogue of a protein or the treatment of immunocompromisedor tumour bearing animals to assess the tumour promoting poten-tial of a product (Table 13).

If no product specific testing is feasible, monitoring of malig-nancies will be a part of the comprehensive risk management planof such products.

4. Discussion

Carcinogenicity data of medicinal products for human use thathave been authorised via the CP between 1995 and 2009 wereevaluated using EPARs and SPCs. In terms of reference medicinalproducts, carcinogenicity data have been generated for 159 outof 375 centrally authorised products (42%). In terms of active sub-stances (INN), carcinogenicity data, either from long-term rodentcarcinogenicity studies, transgenic mouse studies or repeat-dosetoxicity studies, were available for 144 individual compounds(see Table 1). Out of the 144 compounds tested for carcinogenicity,18 (12.5%) exhibited potential genotoxic properties.

With the exception of compounds for which DNA interactionsare not expected, e.g. biotechnology-derived pharmaceuticals,genotoxicity studies are routinely required for all medicinal prod-ucts for human use. Due to the established relationship betweenexposure to genotoxic compounds and carcinogenesis in man,genotoxicity testing is a fundamental part of carcinogenic riskassessment and compounds with proven genotoxic properties areusually not approvable for human use (ICH S2B guideline). There-fore, it is not surprising that the majority of centrally authorisedproducts with carcinogenicity data exhibited negative genotoxicityfindings either in a standard or a more comprehensive test battery.For most of the 18 compounds with potential genotoxic properties,the genotoxic potential was considered to be weak and an expo-sure threshold with regard to the induction of genotoxic effects ap-peared to exist. Thus, the relevance of the genotoxic properties tohumans was either considered to be low due to a high safety mar-gin in terms of exposure or the clinical benefit was considered tooutweigh the potential carcinogenic risk (see Table 8).

Out of the 144 compounds tested for carcinogenicity, 50 (35%)yielded negative results and 94 (65%) were positive in at leastone carcinogenicity study or in repeat-dose toxicity studies (Ta-ble 2). Based on the negative genotoxicity findings of the majorityof compounds, an epigenetic mechanism of carcinogenicity was



Table 8Active substances (INN) for which a genotoxic mechanism of carcinogenicity is plausible.

Active substance Type of tumour Species Relevance to humans according to EPAR/SPC

Cidofovir Mammary adenocarcinoma Rat Potential human carcinogenZymbal gland carcinoma

Stravudine Liver tumours Mouse and rat No relevant risk due to high safety marginUrinary bladder carcinoma Rat

Abacavir sulphate Tumours of the praeputial and clitoral glands Mouse and rat Safety margin established Clinical benefitoutweighs potential carcinogenic riskLiver tumours Rat (f)

Urinary bladder tumoursLymph node tumoursSubcutis haemangiosarcoma

Zidovudine Vaginal tumours Mouse and rat Relevance uncertain, but clinical benefit outweighspotential carcinogenic risk

Entecavir Lung tumours Mouse Relevance uncertain, but clinical benefit outweighspotential carcinogenic riskVascular tumours Rat

Salivary gland tumours (f)Liver tumours (m)Brain tumours (glioma) (m)Pancreas tumoursSkin fibroma (f)Zymbal gland carcinoma (f)Liver tumours (f)Uterine haemangiosarcoma (f)

Tenofovir disoproxil Lipoma (m) Rat Relevance uncertain, but clinical benefit outweighspotential carcinogenic riskUterus polyps (f) Mouse

Duodenal tumoursLiver adenoma (f)

Dexrazoxane Haematopoietic neoplasms Mouse (f) Known relevance, but clinical benefit outweighspotential carcinogenic riskUterine adenoarcinoma Rat (f)

Rasagiline Lung adenoma and carcinomaAQ Mouse No relevant risk due to high safety marginTimolol Adrenal pheochromocytoma Rat (m) No relevant risk due to high safety margin

Pulmonary tumours Mouse (f)Uterine polypsMammary gland tumours

m: Males; f: Females.

Table 9Active substances (INN) with a species- or rodent-specific mechanism of carcinogenicity.

Active substance Tumour findings Species Mechanism of carcinogenicity*

Pantoprazole Liver adenoma and carcinoma Rodent Hepatic microsomal enzyme inductionAprepitantRepaglinidePrasugrelBosentanVoriconazoleRitonavirFosamprenavirTipranavirDarunavirLopinavir/RiponavirNevirapineEtravirineEslicarbazepine acetateStiripentolAgomelatide

Pantoprazole Thyroid gland follicular cell adenomaand carcinoma

Rat Hepatic microsomal enzyme inductionFaster clearance of thyroxin andsubsequent TSH elevation

AprepitantRepaglinideBosentanVoriconazoleIndinavirNelfinavirFosamprenavirTipranavirDarunavirMaravirocMelatonin

Vildagliptin Mammary gland tumours Rat Disturbance of the hypothalamic-pituitary–gonadal axis

Olanazapine Mammary gland tumours Rodent Disturbance of the hypothalamic-pituitary–gonadal axisIncreased prolactin

TimololDronedarone



Please cite this article in press as: Friedrich, A., Olejniczak, K. Evaluation of carcinogenicity studies of medicinal products for human use authorised via theEuropean centralised procedure (1995–2009). Regul. Toxicol. Pharmacol. (2011), doi:10.1016/j.yrtph.2011.04.001


Table 9 (continued)


Aripiprazole Mammary gland tumoursPituitary adenoma

Rodent

Paliperidone Mammary gland tumoursPancreas adenomaPituitary gland adenoma

Rodent

Norelgestromin/Ethinylestradiol Mammary gland tumours Rat Oestrogenic effect in predisposed strains

Insulin aspart Mammary gland tumours Rat Mitogenic and growth promoting action ofinsulinInsulin human (rDNA)

Lasofoxifene Ovarian granulosa cell tumoursUterine polypsLeydig cell tumoursAdrenal gland cortical tumours

Rodent Disturbance of the hypothalamic-pituitary–gonadal axisIncreased LH

Fulvestant Ovarian granulosa cell tumoursLeydig cell tumours

Rat

Sirolimus Leydig cell adenoma RatPramipexole Leydig cell adenoma Rat Disturbance of the hypothalamic-

pituitary–gonadal axisInhibition of prolactin

Rotogotine Uterus carcinomaLeydig cell adenoma

Rat

Tolcapone Uterus carcinoma Rat Disturbance of the hypothalamic-pituitary–gonadal axisBuprenorphine/ Leydig cell adenoma Rat

Naloxone hydrochloride

Active substance Tumour findings Species Mechanism of carcinogenicity⁄Palonosetron Pituitary gland tumours

Pancreas tumoursMammary gland tumours

Rat Link between the serotonin and dopaminesystemsInhibition of dopamine release

Bazedoxifene Ovarian granulosa cell tumours Rodent Stimulation of ovarian follicle growthClass effect of SERMs

Toremifene Ovary tumoursTestes tumoursOsteosarcoma

Mouse Oestrogenic effect of anti-oestrogens

Posaconazole Adrenal gland cortical and medullarytumours

Rat Interruption of steroidogenesis withconsequent increased secretion of ACTH,leading to cortical cell proliferationAdrenal medullary tumours as aconsequence of altered calciumhomeostasisClass effect of azole antifungals

Insulin glargine Malignant fibrous histiocytoma Rodent Injection of solutions with non-neutral pHAtosiban Fibroma and fibrosarcoma Rodent Injection of irritating materialsDegarelix Sarcoma RodentEptotermin alfa Sarcoma Rodent Implantation of solid material

Teriparatide Oseosarcoma Rat Bone anabolic effect of parathyroidhormoneParathyroid hormone (rDNA)

Rufinamide Osteoma Mouse Activation of a mouse-specificpolyomavirus and retrovirus by fluorideions

ExanatideLiraglutide

Thyroid gland C cell adenoma Rat Persistent activation of C cellGLP-1 receptors

Entacapone Kidney adenoma and carcinoma Male rat a2l globulin nephropathyIndacaterol Mesovarian leiomyoma Rat ß adrenergic stimulationPregabalin Haemangiosarcoma Mouse Platelet changes and associated

endothelial cell proliferationBrinzolamide Urinary bladder leiomyosarcoma Mouse Histomorphically unique smooth muscle

tumour

TSH: Thyroid stimulating hormone.LH: Luteinizing hormone.ACTH: Adrenocorticotropic hormone.SERM: Selctive oestrogen receptor modulator.GLP-1: Glucagon-like peptide-1.rDNA: recombinant DNA.* According to Greaves, 2007.


likely for most of the compounds with positive carcinogenicityfindings.

Two long-term rodent carcinogenicity studies were availablefor 116 compounds (81%), 44 of which were negative and 32 ofwhich were positive in both mice and rats. Twenty-two com-pounds were positive in rats only and 18 were positive exclusivelyin mice (see Table 3). Data from one long-term carcinogenicitystudy in rats and a transgenic mouse model were available for


eight compounds (6%). Two compounds showed negative resultsin one long-term carcinogenicity study in rats and a transgenicmouse study, while five were positive in the long-term rat studyand negative in the transgenic mouse study. Only one compoundwas positive in both the long-term rat study and the transgenicmouse study (see Table 3). For 13 compounds (9%), carcinogenicitydata were available in only one rodent species. Within this group,three compounds were negative in mice and one compound in rats,



Table 10Overview of active substances (INN) with tumour findings in carcinogenicity and repeat-dose toxicity studies.

Active substance Tumour findings Mechanism ofcarcinogenicity

Relevance for humansaccording to EPAR/SPC

SPC section 5.3

ATC code A Drugs acting on the gastrointestinal system and on the metabolismATC code A02 Drugs for acid related disordersPantoprazole Mouse: Liver

Rat:Stomach (gastricfundus)LiverThyroid gland

Rodent-specificSecondary to elevated gastrinRodent-specificRodent-specific

Liver and thyroid glandtumours not relevantStomach tumours notrelevantdue to high safety margin

Stomach tumours due to chronic hightreatment of ratsLiver tumours due to high metabolicrate of pantoprazole in the liverThyroid gland tumours due to breakdownof thyroxine in the liver

ATC code A03Drugs for functional gastrointestinal disordersPrucalopride Mouse: Mammary gland

Rat: Liver (adenoma)Thyroid gland(follicular cells)Mammary gland(benign)Pituitary glandPancreas (islet cells)Adrenal medulla(phaeochromocytoma)

No strong support for a non-genotoxic mechanism frommechanistic studiesGenotoxic mechanism cannot beruled out

Genotoxic mechanismrelevantto humans cannot be ruledout

No special hazard for humans

ATC code A04 Anti-emetics and anti-nauseantsPalonosetron Rat:Pancreas

PituitaryMammary glandAdrenalsThyroid glandLiverSkinTail

Mainly tumours of the endocrinesystem to which rodents areparticularly susceptibleStimulation of dopamine release

Not relevant due to highsafety margin

Tumour findings not considered relevantsince high doses were employed in ratsand single administration is to be used in humans

Aprepitant Rat:LiverThyroid gland (follicu-lar cells)

Rodent-specificRat-specific

Not relevant No special hazard for humans

ATC code A10 Drug used in diabetesInsulin glargine Mouse: Fibrous

histiocytomaRat: Fibrous histiocytoma

Rodent-specific Not relevant No special hazard for humans

Insulin aspart Rat: Mammary gland Rat-specific Not relevant No special hazard for humansInsulin human

(rDNA)Rat: Mammary gland Rat-specific Not relevant No special hazard for humans

Metformin Rat: Uterus (benignpolyps)

No tumour promoting effectknown from the literature

Unlikely to be relevant No special hazard for humans

Glimepiride Mouse:

Pancreas (islet cells)Lung (adenoma)Rat: Uterus

Chronic pancreatic stimulationUnknownUnknown

Not relevant due tohigh safety margin

Carcinogenic effects observed only at exposuressufficiently in excess of human exposure or caused bya pharmacodynamic effect

Rosiglitazone Rat: Adipose tissue(lipoma)

Overstimulation of adipocytes Relevant in principleNo safety margin

Increased colon tumours In an animal modelfor familial adenomatous polyposisRelevance of this finding is unknown, however, noevidence of colon tumours in lifetime studies inrodents

Pioglitazone Rat: Urinary bladder Irritation due to urinary calculiformation

Relevance unknown Relevance of tumour finding is unknown

Sitagliptin Rat: Liver Secondary to hepatic toxicity Not relevant due tohigh safety margin

Liver tumours secondary to hepatic toxicity notconsidered relevant due of high safety margin

Vildagliptin Mouse:

Mammary glandBlood vessels(haemangiosarcoma)

Rodent-specificPromoting effect on tumourcommonly observed in mice


No significant risk due to high safety margin

Repaglinide Rat:LiverThyroid gland(follicular cells)



Exenatide Rat:Thyroid gland(C cells) (f)

Rat-specific Not relevant Tumours at plasma exposure 130-fold thehuman clinical exposureIncidence not statistically significant whenadjusted for survival

Liraglutide Mouse:

Thyroid gland (C cells)Uterus (leioma/leiosarcoma)Skin (sarcoma)

C cell tumours rodent-specificNo dose–response relationshipfor leioma/leiosarcoma and micevery sensitive to this tumourSkin tumours situated around themicrochipNo relevant dose related effect for

Thyroid C cell tumours notrelevantUterus leioma/leiosarcomaand skin sarcoma were notconsidered relatedPituitary carcinoma anduterine stromal polyps

Thyroid C cell tumours in mice and rats werecaused by a specific GLP-1 receptor mediatedmechanism to which rodents are particularly sensitiveThe relevance for humans is likely to be low,but cannot completely ruled out





Table 10 (continued)



SPC section 5.3

Rat:Thyroidgland (C cells)Pituitary gland (f)Uterus (polyps)

pituitary carcinoma and uterinepolyps

unlikely to present a risk forhumans

ATC code A16 Other alimentary drugsMiglustat Mouse: Large intestine

Rat: Testes (Leydig cells)UnknownRat-specific mechanismestablished

Relevance of largeintestine tumours unknown

A safety margin was establishedThe relevance of carcinoma of the largeintestine for humans cannot be completely ruled out

ATC code B Blood and blood forming organsATC code B01 Anti-thrombotic agentsPrasugrel Mouse: Liver Rodent-specific Not relevant Liver tumours considered secondary

to hepatic enzyme inductionThe increase in liver tumours in mice isnot considered a relevant human risk

ATC code C Cardiovascular systemATC code C01 Cardiac therapyRanolazine Rat:Adrenal cortex

Adrenal medullaThyroid glandTestesSkin (sarcoma)

Genotoxic metabolite Carcinogenic metabolitenot found in humans

No relevant increases in the incidence of anytumour types were seen in carcinogenicitystudies in mice and rats

ATC code C02 Anti-hypertensivesBosentan Mouse: Liver (m)

Rat: Thyroid gland (m)Rodent-specificRat-specific

Not relevant There was evidence for a mild thyroid hormonalimbalance induced in rats, but no evidence ofbosentan affecting thyroid function in humans

Sitaxentansodium

Rat:Adrenal medulla (phe-ochromcytoma) (m)Skin tumours (m)

Tumours probably not related Tumours not consideredrelated based on historicalcontrol data and statisticalanalyses

Sixatexan sodium was not carcinogenic

ATC code C09Agents acting on the renin-angiotensin systemHydro-

chlorothiazideMouse: Liver (m) Unknown Unlikely to be relevant Extensive human experience with hydrochlorothiazide

has failed to show an association between its useand an increase in neoplasms

ATC code C10 Lipid modifying agentsColesevelam Rat:Pancreas (islet cells)

(m)Thyroid gland (C cells)

Not clinically relevantSlight increase consideredincidental since thyroid adenomaare common in old rats


Carcinogenic effects observed only at exposuressufficiently in excess of maximum human exposureindicating little relevance to clinical use

Laropiprant Mouse: Testes Tumours not considered relateddue to atypical absence ofspontaneous tumours in controls

Not relatedHigh safety margin

Laropiprant was not carcinogenic

ATC code D DermatologicalsATC code D11 Other dermatological preparationsTacrolimus

(topical)Mouse:

LymphomaHairless mouse photo-carcino-genicity: Skintumours

Systemic immunosuppressiveeffect

Potential for localimmunosuppressionunknown

In a dermal carcinogenicity study in mice, lymphomawere observed in association with highsystemic exposure In a photocarcinogenicity study inhairless mice, a reduction in time to skin tumour andan in crease in the number of tumours was observed Arisk for humans cannot be completely ruled out as thepotential for local immuno-suppression with long-term use of tacrolimus ointment is unknown

ATC code G Genito-urinary system and sex hormonesATC code G02 Other gynecologicalsAtosiban Rat: Injection site

(fibroma/fibrosarcoma)Rodent-specific Not relevant Atosiban was not carcinogenic

Norelgestromin/Ethinylestradiol

Rat: Mammary gland Rat-specific Not relevant No special hazard for humans

Lasofoxifene Mouse:Adrenal cortexOvary (granulose cells)Uterus (polyps)Testes (Leydigcells)Rat:

OvaryKidney (m)

Rodent-specificMale rat specific expression ofERa versus ERß

Not relevantRelevance unknown

Although all of the observed tumours are believed tobe the result of rodent-specific hormonal mechanisms,the relevance for humans is currently unknown

Bazedoxifene TgHras2 mouse model:Ovary (granulosa cells)Rat: Ovary (granulosacells)

Rodent-specific Not relevant Ovary tumours are a class effect of SERMs, related to itspharmacology in rodents when treated during theirreproductive lives, when their ovaries are functionaland responsive to hormonal stimulation







SPC section 5.3

ATC code G04 UrologicalsDarifenacin Rat:Adrenal cortex (f)

Blood vessels(haemangiosarcoma)(m)

Tumours not considered related Tumour notconsidered related


ATC code H Systemic hormonal preparations, excluding sex hormones and insulinsATC code H01 Pituitary and hypothalamic hormones and analoguesMecasermin Rat:Adrenal medulla

(phaeochromocytoma)Skin(keratoacanthoma) (m)Mammary gland(m + f)

Promotion of a commonspontaneous tumour due tomitogenic/anti-apoptotic effectIndirect effect due to effect oncalcium metabolism, bloodglucose, body weight and foodconsumption

Relevance unknown An increased incidence of pheochromo-cytoma,keratoacanthoma in the skin and mammarygland carcinoma were observed in therat carcinogenicity study

ATC code H05 Calcium homeostasisTeriparatide Rat: Bone (osteosarcoma) Rat-specific Not relevant Due to the differences in bone physiology in

rats and humans, the clinical relevance ofbone tumours is probably minor

Parathyroidhormone(rDNA)

Rat: Bone (osteoma/osteosarcoma)

Rat-specific Not relevant Due to the differences in bone physiology inrats and humans, the clinical relevance ofbone tumours is probably minor

ATC code J Anti-infectives for systemic useATC code J02 Anti-mycotics for systemic useVoriconazole Mouse: Liver

Rat: LiverRodent-specific Not relevant No special hazard for humans

Posaconazole Rat:Adrenal cortexAdrenal medulla(pheochromocytoma)

Rat-specific Not relevant No special hazard for humans

Micafungin Rat: Liver Secondary to hepatotoxicity Relevant since no safetymargin could be established

SPC Sections 4.4 and 5.3:Foci of altered hepatocytes wereobserved in rat repeat-dose toxicity studiesIncreased tumour rates were observed atthe end of a 12-month recovery periodA reliable safety margin could be establishedThe relevance of the tumour finding for thetherapeutic use cannot be excludedLiver function should be carefully monitoredduring treatmentEarly discontinuation in the presence ofelevated AST/ALT is recommended

ATC code J05 Antivirals for systemic useCidofovir Rat:Mammary gland

Zymbal’s glandGenotoxic Relevant Tumours were observed in rats at subtherapeutic

plasma levels and within3 months of treatmentSPC section 4.4:Cidofovir should be considered a potentialcarcinogen in humans

Ribavirin Rat: Thyroid gland (C cells)(f)

Increased tumour incidence infemales most likely due toincreased survival rate

Not relevant Carcinogenic potential in humans is unlikely

Indinavir Rat: Thyroid gland Rat-specific Not relevant Tumours probably related to an increase in release ofTSH secondary to an increase in thyroxin clearanceThe relevance of the finding is likely to be limited

Ritonavir Mouse: Liver (m) Rodent-specific Not relevant Species-specific tumourigenic potential, which isregarded as of no relevance for humans

Nelfinavir Rat: Thyroid gland Rat-specific Not relevant Treatment of rats with nelfinavir produced effectsconsistent with enzyme induction, which predisposedrats, but not humans, to thyroid neoplasmsThe weight of evidence indicates that nelfinavir isunlikely to be a carcinogen in humans

Amprenavir Mouse: Liver (adenoma)(m)Rat: Liver (adenoma) (m)

Secondary to hepatotoxicity Relevance unknown The mechanism for the tumour findings wasnot elucidatedThe increased incidence was reported with alow safety margin and the clinical relevance inhumans in unknownIt should be considered that liver changes were alsoseen in repeat-dose toxicity studies in rats and dogsThe liver can be regarded as a target organ for toxicity

Lopinavir/Ritonavir

Mouse: Liver Rodent-specific Not relevant Mitogenic induction of liver tumours, generallyconsidered to have little relevance to human risk

Fosamprenavir Mouse: LiverRat:Thyroid gland (follicu-

lar cells)

Rodent-specificRat-specificUnknown

Not relevantRelevance unknown

Liver findings are consistent with hepatic enzymeinduction, which predisposes rats tothyroid neoplasms








SPC section 5.3

Uterus The incidence of uterus tumours was slightlyincreased over concurrent controls, but waswithin background range for female ratsThe relevance of the uterus tumours forhumans is uncertainThere is no evidence to suggest that thetumour findings are of clinical significance

Atazanavirsulphate

Mouse: Liver (adenoma) (f) Secondary to hepatotoxicity Not relevant at therapeuticexposures

Tumours likely to be secondary to cytotoxicliver changes and considered to have norelevance for humans at therapeutic exposures

Tipranavir Mouse: LiverRat:Liver

Thyroid gland


Not relevant Species-specific tumourigenic potential,which is regarded as of no clinical relevance

Darunavir Mouse: LiverRat:Liver

Thyroid gland (m)


Not relevant The observed liver and thyroid gland tumours areconsidered to be of limited relevance to humans

Stravudine Mouse: LiverRat:Liver

Urinary bladder

Genotoxic Not relevant due to highsafety margin

Tumours were observed at very high exposure levelssuggesting an insignificant carcinogenic potentialin clinical therapy

Abacavirsulphate

Mouse:Preputial glandClitoris glandRat:Preputial glandClitoris glandThyroid gland (m)Liver (m)Urinary bladder (m)Lymph nodes (m)Subcutis (m)

Genotoxic Safety margin establishedClinical benefit outweighscarcinogenic risk

Systemic exposure at the no effect level was equivalentto 3–7 times the human systemic exposureduring therapyWhile the carcinogenic potential in humans isunknown, these data suggest that a carcinogenic risk isoutweighed by the potential clinical benefit

Tenofovirdisoproxil

Mouse:Duodenum

Liver (adenoma) (f)Rat:

Adipose tissue(lipoma) (m)Uterus (polyps)

GenotoxicDuodenal tumours in mice due toformalaldehyde released fromtenofovir disoproxil

Tumours in rats notconsidered relatedTumours in mice notconsidered to present asignificant carcinogenic riskfor humans

Low incidence of duodenal tumours in mice,considered likely related to high concentrations oftenofovir disoproxil in the gastrointestinal tractFindings unlikely to be relevant to humans

Entecavir Mouse:

LungBlood vessels (f)Salivary gland (f)

Liver (m)Rat:

Brain (glioma)Pancreas (acinar cells)(m)Skin (fibroma) (f)Zymbal’s gland (f)Uterus

Genotoxic Key event in lung tumourdevelopment species-specificPredictivity of findings forhumans is unknown

Lung tumours in mice were preceded bypneumocyte proliferation which was notobserved in rats, dogs or monkeysIncreased incidences of other tumours wereseen only at high life-time exposures, however,the effect levels could not be precisely establishedPredictivity of the findings for humans is not known

Telbivudine Rat:Pancreas (acinar cells)Adrenal medulla(phaeochromocytoma)Mammary gland(fibroadenoma)

Size of the effects and lack ofdose response suggested thatfindings are incidental


Nevirapine Mouse: LiverRat: Liver

Rodent-specific Not relevant Liver tumours are most likely to nevirapine beinga strong inducer of liver enzymes

Efavirenz Mouse:

Liver (f)Lung (f)

Unknown Relevance unknown,but clinical benefitoutweighs potentialcarcinogenic risk

While the carcinogenic potential in humans isunknown, the data suggest that the clinical benefitoutweighs the potential carcinogenic risk to humans

Etravirine Mouse: Liver (f) Rodent-specific Not relevant Liver tumours generally considered to berodent-specific, associated with liver enzymeinduction, and of limited relevance to humans

Oseltamivir Rat:Blood vessels (hae-mangioma/haemangiosarcoma)Lymphoid system (m)Epithelia (f)

Typical tumours of rodent strainused

Findings in rats ofminor significance

Trend towards a dose-dependent increase in theincidence of some tumours that are typical for therodent strains usedConsidering the margin of exposure in relation to theexpected exposure in the human use, these findings donot change the risk–benefit in the adopted indications

Zidovudine Mouse: VaginaRat: Vagina

Genotoxic Relevance for humans isuncertain due to metabolicdifferences between rodentsand humans

An intravaginal carcinogenicity study comfirmed thatvaginal tumours were the results of long-termexposure of the vaginal epithelium to highconcentrations of unmetabolised zidovudine in urine

Maraviroc Rat:Thyroid gland Rat-specific Not relevant Thyroid tumours in rats were considered of low







SPC section 5.3

Bile duct (cholangi-oma/cholangiosarcoma)

Unknown Not relevant due to highsafety margin

human relevanceBile duct tumours in rats were reported at a systemicexposure at least 15 times the expected humanexposure

ATC code L Anti-neoplastic and immunomodulating agentsATC code L01 Anti-neoplastic agentsGefitinib Mouse: Liver (adenoma)

Rat:Liver (adenoma)Mesenteric lymphnodes (haemangiosar-coma) (f)

Unknown No safety margin establishedRelevance unknown

Liver adenoma and mesenteric lymph nodehaemagiosarcoma were observed in ratsLiver adenoma were also observed in miceThe clinical relevance of these findings is unknown

Everolimus Mouse: Leukaemia(granulocytic)

Unknown No safety margin established,but clinical benefit outweighspotential carcinogenic risk

Carcinogenicity studies in mice and rats didnot indicate any tumourigenic potential

Celecoxib Mouse:

Blood vessels(haemangiosarcoma)(f)Pituitary (f)Rat: Liver(m)

Tumour incidences appeared tobe with values of historicalcontrols

No safety margin establishedClinical benefit outweighscarcinogenic risk

No information presented

Anagrelide Rat:Adrenal medulla(phaeochromocytoma)Uterus (m)

Exaggerated pharmacologicaleffectHepatic enzyme induction


Adrenal medulla and uterus tumours wereobserved at high exposure levelsThere is no clinical evidence that thefindings are of relevance to human use

ATC code L02 Endocrine therapyToremifene Mouse:OvaryTestesBone Mouse-specific Not relevant Little relevance for the safety in man, where

toremifene acts mainly as an anti-oestrogenFulvestrant Rat:Ovary (granulosa cells)

Testes (Leydig cells)Rat-specific Not relevant Induction of tumours is consistent with

pharmacology-related endocrine feedback alterationsThese findings are not of clinical relevance for thetreatment of postmenopausal women with advancedbreast cancer

Degarelix Mouse:Liver (adenoma)Lung (adenoma) (f)Injection site (sar-coma)Rat: Lymphnodes (haemangiosar-coma) (f)

UnknownRodent-specificNot considered related

UnknownNot relevantProbably not relevant due tolow incidence and noconcurrent increase in males


ATC code L04 ImmunosuppressantsLeflunomide Mouse:

Lymphoma (m)Lung (f)

ImmunosuppressionUnknown

RelevantRelevance unknown

Malignant lymphoma observed in male mice wereconsidered to be due to the immuno-suppressiveactivityThe relevance of the lung tumours in femalemice is uncertain. SPC section 4.8:The risk of lymphoproliferative disorders is increasedwith use of some immuno-suppressive agents

Sirolimus Mouse:

LymphomaLeukaemia (f)Liver (m)Rat: Testes (Leydig cells)

ImmunosuppressionRat-specific

RelevantNot relevant

Lymphoma secondary to chronic use ofimmunosuppressive agent can occur andhave been reported in patients in rare instancesTestes tumours were considered to be due to aspecies-specific response to LH levels and oflimited clinical relevance

Abatacept Mouse:LymphomaMammary gland (f)

Decreased control of viralinfections due toimmunosuppression

Relevant Malignant lymphoma and mammary tumours inmice may be associated with decreased controlof murine leukaemia virus and mouse mammarytumour virus, respectively, in the presence oflong-term immunomodulationThe relevance of these findings to the clinicaluse of abatacept is unknownSPC section 4.4 and 4.8:Immunosuppression increases the susceptibilityto the development of lymphoma and othermalignancies

Tacrolimus(systemic)

Mouse (topicalapplication): Lymphoma

Systemic immunosuppressiveeffect

Relevant In mice, topical administration of tacrolimus wasassociated with high systemic exposure levelsresulting in the formation of lymphomaSPC section 4.4 and 4.8:Increased risk of malignancies secondaryto immunosuppression

ATC code M Musculo-skeletal systemATC code M05 Drugs for treatment of bone diseases








SPC section 5.3

Zoledronic acid Mouse: Harderian gland Unknown Tumour findings notconsidered relevant

Carcinogenicity studies did not provide anyevidence of a carcinogenic potential

Eptotermin alfa Rat: Implantation site(sarcoma)

Rodent-specific Not relevant Sarcoma in rats was associated with thelong-term presence of heterotopic boneSolid state carcinogenicity is frequentlyobserved in rats when solid materials areimplanted subcutaneouslyThere is evidence to suggest that heterotopicossification is not linked to sarcoma in humans

Strontiumranelate

Rat: Thyroid gland (C cells)(m)

Incidences within historicalcontrol range

Not related No special hazard for humans

ATC code N Nervous systemATC code N03 Anti-epilepticsRufinamide Mouse:

Bone (osteoma)Liver

Activation of a mouse-specificvirus by fluoride ions

Not relevant Osteomas were considered a result of activationof a mouse-specific virus by fluoride ion releasedduring oxidative metabolism of rufinamide

Eslicarbazepineacetate

Mouse: Liver Rodent-specific Not relevant Liver tumours consistent with an inductionof hepatic microsomal enzymes

Pregabalin Mouse: Blood vessels(haemangioma)

Mouse-specific Not relevant Platelet changes and associated endothelialproliferation were not present in rats or humansNo evidence to suggest an associated risk to humans

Stiripentol Mouse: Liver Rodent-specific Not relevant Special susceptibility of the mouse liver to tumourformation in the presence of hepatic enzyme inductionLiver tumours were not considered to indicate a risk oftumourigenicity in humans

ATC code N04 Anti-Parkinson drugsPramipexole Rat: Testes (Leydig cells) Rat-speciific Not relevant Tumours can be explained by a prolactin-inhibiting

effect of pramipexoleThe finding is not clinically relevant to man

Rotogotine Rat:Testes (Leydig cells)Uterus

Rat-specific Not relevant Tumours are well-known effects of dopamine-agonistsin rats and assessed as not relevant to man

Rasagiline Mouse: Lung Genotoxic Not relevant due to highsafety margin

Lung tumours were observed in mice atsystemic exposures 144–213 times the expectedplasma exposure in humans

Tolcapone Rat:KidneyUterus

Secondary to nephrotoxicitySpecies-specific

Safety factor was establishedNot relevant

Renal epithelial tumours were observed inthe mid- and high-dose groups in rats, however,there was no evidence of renal toxicity in thelow-dose groupUterine adenocarcinoma were observed inthe high-dose group in rats

Entacapone Rat: Kidney (m) Male rat-specific Not relevant No special hazard for humans

ATC code N05 PsycholepticsOlanzapine Mouse: Mammary gland (f)

Rat: Mammary gland (f)Rodent-specific Not relevant Based on the results of carcinogenicity studies

in mice and rats, it was concluded thatolanzapine is not carcinogenic

Aripiprazole Mouse:Mammary gland (f)Pituitary gland (f)Rat:

Mammary gland (f)Adrenal cortex (f)

Rodent-specificCytotoxicity due to increasedoxidative stress

Mammary and pituitarytumours not relevantSafety margin establishedfor adrenocortical tumours

Carcinogenic effects observed at doses/exposuressufficiently in excess of the maximum humandose/exposure, indicating that these effects wereof limited or no relevance to clinical use

Paliperidone Mouse:Mammary gland (f)Pituitary gland (f)Rat:

Mammary gland(m + f)Pancreas (islet cells)(m)

Rodent-specific Not relevant The observed tumours of the pituitary gland,endocrine pancreas and mammary gland can berelated to prolonged dopamine antagonismand hyperprolactinaemiaThe relevance of these tumours in terms ofhuman risk is unknown

Melatonin Rat:Pituitary gland (m)Thyroid gland (follicu-lar cells)

Pituitary tumours common inratsStatistical significance ofpituitary tumours below thevalue of triggering concernThyroid tumours rat-specific

Not relevant The carcinogenicity study in rats did notreveal any effect which may be relevant for humans

ATC code N06 PsychoanalepticsAgomelatine Mouse: Liver

Rat:Liver (m)Mammary gland(fibroadenoma) (m + f)

Liver tumours rodent-specificMammary gland tumoursunknown

Not relevant due to highsafety margins

Liver tumours were most likely related toenzymeinduction specific to rodentsThe frequency of mammary fibroadenomain rats wasincreased with high exposures

ATC code N07 Other nervous system drugsVarenicline Rat: Brown fat Increased sympathetic In humans, brown fat is There was a dose-related increase in the







SPC section 5.3

tartrate (hibernoma) stimulation of brown adipocytes presentat birth, after which itsmetabolicactivity and thermogeniccapacitydecreases to minimal levels.Therefore, the risk for humansistheoretical and probably non-existent

incidence of hibernoma in male rats

Buprenorphine hydrochloride/naloxonehydrochloride

Rat: Testes (Leydig cells) Rat-specific Not relevant

Statisticallysignificantincreases inthe incidenceof Leydig celladenomawereobserved inrats at all dosegroups atexposuremultiples of3–75 times

Sodium oxybate Mouse (c-butyrolactone):Adrenal medulla(pheochromocytoma) (f)Rat (sodium oxybate):Pituitary gland (f)

Tumours slightly increased anddifficult to interpret due to highmortalityDoubtful statistical significanceand incidence was at the upperbound of historical controls

Unlikely relevant In a mouse study with c-butyrolactone,results were equivocal due to a slightincrease of pheochromocytoma, whichwas difficult to interpret due to high mortalityIn a rat study with sodium oxybate,no compound-related tumours were observed

ATC code S Sensory organsATC code S01 OphthalmologicalsBrinzolamide Mouse: Urinary bladder

(leiomyosarcoma) (f)Mouse-specific Not relevant Smooth muscle tumour was considered

unique to miceTimolol Mouse:

Lung (f)Uterus (benign polyps)Mammary glandRat:Adrenal medulla(pheochromcytoma)

UnknownUnknownRodent-specificUnknown



ATC code V VariousATC code V03 All other therapeutic productsSevelamer

(carbonate)Mouse: LymphomaRat: Urinary tract andbladder (m)

Not considered related totreatmentDue to crystalline deposits in theurine

Tumours not consideredrelatedRelevance unknown

In rats, there was an increased incidence of urinarybladder transitional cell papilloma in males of thehigh-dose group

Dexrazoxane Mouse: Haematopoietictumours (f)Rat: Uerus

Clastogenic Probably relevant Secondary malignancies in mice and ratsafter prolonged administration

Not yet classifiedDronedarone Mouse:

Haemolymphoreticularsystem (sarcoma)Mammary gland (f)Harderian gland (f)Rat:Mesenteric lymphnodes (haemangioma)

Mammary tumours rodent-specificHaemangiosarcoma unknown

Not relevant due highsafety margin

None of the tumour findings were consideredrelevant for humans

Indacaterol Rat: Ovary (leiomyoma) Rat-specific Incidence of tumour notincreased in women

following use of adrenergicagents

Similar findings in rats reported forother ß2 agonistsSafety margin in terms of exposure

m: Males; f: Females.ER: Oestrogen receptor.AST: Aspartate aminotransferase.ALT: Alanine aminotransferase.TSH: Thyroid stimulating hormone.LH: Luteinizing hormone.




Table 11Active substances (INN) with carcinogenic effects attributable to exaggerated pharmacodynamic effects or toxic damage.


Pantoprazole Stomach squamous papilloma and carcinoma Rat Secondary to massively elevated serum gastrin levelsGlimepiride Pancreatic islet cell adenoma Rat Chronic islet cell stimulationSitagliptin Liver tumours Rat Secondary to hepatotoxicityAtazanavir sulphateAmprenavir Mouse

Mouse and ratMicafunginRat

Tolcapone Renal epithelial tumours Rat Secondary to nephrotoxicityTacrolimus Lymphoma Mouse Due to immunosuppressive actionLeflunomideSirolimusAbataceptPioglitazone Urinary bladder transitional cell tumours Male rat Irritation due to urinary calculi formationSevelamer

* According to Greaves, 2007.

2 Tacrolimus was counted twice as it is approved as topical product for thetreatment of atopic dermatitis (ATC code D 11) and systemic treatment of transplantrejection (ATC code L04).


while two and seven compounds were positive in mice and rats,respectively (see Table 3). For one compound, the only carcinoge-nicity data available were from a transgenic mouse study, whichshowed negative results (see Table 3). Six compounds (4%) weretumourigenic in repeat-dose toxicity studies in rats (see Table 5).Out of the 94 compounds with positive findings in either carcino-genicity or repeat-dose toxicity studies, 33 compounds were posi-tive in both mice and rats (35%), 40 were positive in rats only (43%)and 21 were positive in mice only (22%). Thus, the majority of po-sitive carcinogenicity findings (78%) have been produced in rats.

A similar analysis of carcinogenicity data all pharmaceuticals(NCEs) that were submitted to the regulatory authorities from Ger-many and The Netherlands between 1980 and 1995 (includingthose which were withdrawn or not approved later on) was per-formed by Van Ousterhout et al. (1997). The analysis showed that181 out of 221 compounds (approximately 82%) were tested in tworodent long-term carcinogenicity studies and 40 compounds (18%)were tested in either mice or rats (or the second species was re-placed, e.g. by hamsters). When comparing the two evaluations,it is of note that the frequency of using two long-term carcinoge-nicity studies has not changed during the last 15 years and theacceptance of transgenic mouse models as a replacement of thesecond long-term study is considered to be poor.

In terms of positive tumour findings, Van Ousterhout et al.found that 106 out of 221 compounds (48%) were positive in atleast one long-term carcinogenicity study. Approximately 44%positive pharmaceuticals were found in an evaluation of carcinoge-nicity studies in the FDA database and NTP rodent carcinogenicitydatabase (Contrera et al., 1997).

Considerably more compounds with positive tumour findings,i.e. 94/144 (65%) were identified in the evaluation of the EPARs,although the classification of tumours was similar to that usedby Van Ousterhout et al. A particular reason for the higher numberof positive findings could not be determined.

In the evaluation by Van Ousterhout et al., 92 out the 106 com-pounds with positive carcinogenicity findings were positive in bothmice and rats (34) or in rats only (58). Thus, approximately 87% ofthe compounds showed positive results in rats and only 13% werepositive in mice, but not in rats. In the evaluation of the EPARs, theproportion of compounds that yielded positive results in rats was78% (21/94). The results are consistent with the view that the ratis more sensitive towards carcinogenic effects than the mouse(Smith, 1996; Van Ousterhout et al., 1997).

As summarised in Table 10, most of the tumour findings ob-served in carcinogenicity studies in mice and/or rats were consid-ered not to be relevant for humans. Among these were tumourfindings for 38 compounds (40%) which were classified as species-or rodent-specific (see Table 9). For all of these tumours, the plau-


sibility of a species- or rodent-specific mechanism of carcinogenic-ity has been demonstrated by additional mechanistic evaluationsincluding hormone measurements.

For 20 compounds (21%), high safety margins in terms of expo-sure between the NOAEL in rodents and the recommended thera-peutic exposures in humans were established (see Table 12). Thisindicates the particular importance of toxicokinetic measurementwith regard to carcinogenic risk assessment. The systemic expo-sures determined in rodents were at least five times and up to sev-eral thousand times higher than the clinical exposures achieved atmaximum therapeutic doses. For most of these compounds, tu-mour findings in rodent carcinogenicity study were considerednot to be relevant for the clinical situation due to demonstratedhigh exposure differences between rodents and humans (see Ta-ble 10 for SPC wordings). However, for certain compounds, e.g.the potentially clastogenic compound abacavir sulphate, a lowersafety margin was considered to be acceptable based on the overallrisk–benefit evaluation for the medicinal product (see Table 10 forSPC wording for abacavir sulphate).

For 11 compounds (11%), tumours observed in rodent carcino-genicity studies were either considered not related to treatmentor were thought not to be relevant for humans. A number of tu-mours were considered incidental because they fell within therange of historical control data, due to a small effect size and lackof dose–response relationship or due to the fact that the tumourswere typically observed in the rodents strains used. Some tumourswere considered not relevant for humans based on available liter-ature and clinical data that did not indicate a carcinogenic risk forhumans, or based on likely differences in metabolism/concentra-tions between rodents and humans (for details see Table 10).

Tumours observed for 14 compounds (15%) were considered tobe of unknown relevance for humans. For many of the compounds,tumour findings were described in the SPC Section 5.3 and it was sta-ted that the relevance for humans is unknown. However, none of thefindings triggered any further regulatory actions (see Table 10).

A potential carcinogenic risk for humans was established foreleven compounds (11%). These compounds have neverthelessbeen approved for clinical use based on a positive risk benefitassessment. Based on carcinogenic effects observed in repeat-dosetoxicity studies in rats, the anti-retroviral agent cidofovir was la-belled as a potential human carcinogen in the SPC section 4.4.The immunosuppressive compounds leflunomide, sirolimus,abatacept and tacrolimus2 caused lymphoma in mouse carcinoge-nicity studies which are thought to be virus-related. It is well-



Table 12Active substances (INN) with a high safety margin in terms of systemic exposure.

Active substance Potential mechanism of carcinogenicity

Stravudine GenotoxicityAbacavir sulphateRasagilineTimolol

Pantoprazole Exaggerated pharmacodynamic effect or toxic damageGlimepirideSitagliptinAtazanavir sulphateTolcaponeAnagrelide

Palonosetron UnknownMarovirocDronedaroneAgomelatineColesevelamLaropiprantMiglustatDarifenacinVildagliptinAripiprazole

Table 13Biotechnology-derived products tested for carcinogenicity.


Internationalnonproprietaryname (INN)

Results of mechanisticstudies

NeoRecormon Epoetin beta Negative in a long-term mousestudy with murine epoetinNegative in a rat study withimplanted tumours andcyclophosphamide treatment

Remicade Infliximab Negative in a repeat-dosetoxicity study withanti-mouse TNFa


documented that immunosuppressive agent can cause the develop-ment of virus-induced malignancies. Corresponding wordings wereincluded in SPC section 4.4 and 4.8 for these compounds (seeTable 10). The antifungal agent micafungin induced liver tumoursin rats at exposure levels similar to those seen in humans. Thecarcinogenic effects of dexrazoxane in rodents were related to itsclastogenic activity (see Table 10). For the tumours observed in ro-dent carcinogenicity studies with the anti-retroviral compoundefavirenz and the anti-neoplastic compounds everolimus andcelecoxib, the EPAR/SPC indicated that a potential carcinogenic riskis outweighed by their clinical benefit (see Table 10).

Trans-species carcinogenicity findings which are usually con-sidered to pose a relatively greater risk to humans (Contreraet al., 1997) were rarely noted. Trans-species findings includedvaginal tumours observed in carcinogenicity studies with zidovi-dine and liver adenoma observed in carcinogenicity studies withgefitinib. Vaginal tumours were attributed to high concentrationsof unmetabolised zidovidine in the urine. The relevance to humanswas however considered to be uncertain due to metabolic differ-ences between rodents and humans. The relevance of liver ade-noma caused by gefitinib for humans was unknown and acorresponding statement was included in the SPC section 5.3.

In summary, the present evaluation indicated that a high num-ber of compounds contained in centrally approved medicinal prod-ucts produced positive tumour findings in rodent carcinogenicitystudies (94/144, 65%). The majority of the rodent carcinogenicityfindings were considered not to be of relevance for humans (69/94, 73%). Genotoxicity, toxicokinetic and mechanistic studies pro-


vided important information with respect to the interpretation ofrodent tumours findings and their potential relevance for humans.

The necessity for routinely conducting two long-term rodentcarcinogenicity studies has been under discussion for many years.The relative individual contribution of mouse and rat carcinogenic-ity studies and whether the use of rats alone would result in a sig-nificant loss of information on carcinogenicity relevant to humanrisk assessment has been addressed by a number of surveys of datafor human pharmaceuticals (ICH S1B guideline; Smith, 1996; VanOusterhout et al., 1997; Contrera et al., 1997). The analyses ledto the recommendation by ICH of conducting one long-term car-cinogenicity study in rats which should be supplemented by atransgenic mouse assay or the neonatal rodent tumourigenicitymodel (ICH S1B guideline).

However, the present evaluation indicated that this approachhas basically failed since the carcinogenic potential of the majorityof compounds were still evaluated using two long-term carcinoge-nicity studies (see Table 1).

The evaluation of the EPARs confirmed that the rat was moresensitive than the mouse towards carcinogenic effects with themajority of compounds being positive in both mice and rats or ratsalone (73/94, 78%). Although 21 compounds produced carcinogeniceffects in mice only, most of the findings were considered not to berelevant for humans and consequently did not trigger any regula-tory actions. For the development of lymphoma under treatmentwith immunosuppressive agents, the mouse was shown to be moresensitive than the rat. However, this is a well-documented phe-nomenon and considered to be virus-related (see Table 10).

5. Conclusions

The evaluation of carcinogenicity data of centrally authorisedmedicinal products revealed that for the majority of productstwo long-term rodent carcinogenicity studies were used for assess-ment of carcinogenic potential and that the acceptance of trans-genic mouse models was low.

The evaluation showed that the majority (69/94, 73%) of posi-tive carcinogenicity findings in rodents were considered not tobe relevant for humans. The findings confirmed the results of pre-vious similar reviews of carcinogenicity studies (Monro, 1996; VanOusterhout et al., 1997).

It can be concluded from the current evaluation that tumourfindings in rodents were largely not predictive for human use, par-ticularly for compounds with hepatic enzyme inducing and com-pounds inducing hormonal disturbances. In addition,carcinogenicity studies were redundant for compounds withimmunosuppressive properties due to their known carcinogenicpotential in humans. This is consistent with findings from previousreviews (Monro, 1996).

Furthermore, the current analysis revealed that positive carcin-ogenicity findings in either mice or rats did not trigger any regula-tory actions if there was no supportive evidence of their potentialrelevance for humans from genotoxicity, toxicokinetic or mecha-nistic studies.

In the current evaluation, a potential relevance for the humansituation was established for only 11 compounds (11%) with posi-tive carcinogenicity findings. These compounds were authorisedbased on a positive risk benefit assessment. It is of note that fourof these compounds produced tumour findings in repeat-dose tox-icity studies in rats (see Table 5). Four compounds were subjectedto carcinogenicity testing despite their immunosuppressive prop-erties which are known to allow the development of virus-relatedmalignancies in both mice and humans (see Table 10).

The lack of accuracy of long-term rodent carcinogenicity studiesfor predicting human cancer risk has been criticised in the past




(Monro, 1996; Ennever and Lave, 2003). The current evaluation ofthe EPARs confirmed the poor predictivity of long-term rodent car-cinogenicity studies for the human situation. Therefore, a revisionof the current carcinogenicity testing strategy for pharmaceuticalsis warranted.

As suggested by previous reviews of carcinogenicity studies(Alden et al., 1996; Van Ousterhout et al., 1997) and by the ICH S1Bguideline, long-term carcinogenicity studies in the mouse are unli-kely to add any significant value and should not be requested inthe future. As suggested by the ICH S1B guideline, the long-termmouse study may be replaced by a transgenic mouse model.The acceptability of this strategy may be improved by a closercollaboration between regulatory agencies and pharmaceuticalcompanies.

A recent compilation of rat chronic toxicity and 2-year carcino-genicity study data of both marketed and non-marketed com-pounds from a collaboration of 13 pharmaceutical companiesdemonstrated that rat chronic toxicity studies are good predictorsof the negative outcome in long-term rat carcinogenicity studiesprovided that genotoxicity studies are negative and no preneoplas-tic changes or hormonal disturbances were observed in chronic ratstudies. The evaluation of chronic rat toxicity studies has the poten-tial to eliminate approximately 40% of the long-term rat carcinoge-nicity studies based on their predictivity for a negative outcome forrat tumour development (Reddy et al., 2010; Sistare, 2010).

Two-year rodent carcinogenicity studies are currently the mostexpensive and time-consuming animal tests required for pharma-ceutical carcinogenicity assessment. Since these studies are largelynot predictive of human cancer risk, both pharmaceutical compa-nies and regulatory agencies should aim at their replacement.The collaborative assessment of existing carcinogenicity data frompharmaceutical companies utilising new approaches to identifypotentially human relevant or irrelevant mechanisms of carcinoge-nicity including genetically modified animal models and in vitrocarcinogenicity screening assays based on gene expression profil-ing (Vinken et al., 2008; Bercu et al., 2010) are likely to improvethe current carcinogenicity testing paradigm without the need oflong-term rodent carcinogenicity studies.

Conflict of interest

The authors confirm that they have no conflicts of interests thatcould inappropriately influence, or be perceived to influence, theirwork.


Guidelines

ICH S1A: The need for carcinogenicity studies of pharmaceuticals. CPMP/ICH/140/95, July 1996.

ICH S1B: Carcinogenicity: testing for carcinogenicity of pharmaceuticals. CPMP/ICH/299/95, March 1998.

ICH S2B: Genotoxicity: A standard battery for genotoxicity testing of pharma-ceuticals. CPMP/ICH/174/95.

ICH S3A: Toxicokinetics: A Guidance for Assessing Systemic Exposure in ToxicologyStudies CPMP/ICH/384/95, June 1995.

ICH S6: Preclinical Safety Evaluation of Biotechnology-Derived Pharmaceuticals.CPMP/ICH/302/95, March 1998.

ICH S9: Nonclinical Evaluation for Anticancer Pharmaceuticals. EMEA/CHMP/ICH/646107/2008, November 2009.

Note for guidance on carcinogenic potential. CPMP/SWP/2877/00, July 2002.Note for guidance on photosafety testing. CPMP/SWP/398/01, June 2002.Guideline on carcinogenicity evaluation of medicinal products for the treatment of

HIV infection. EMEA/CHMP/SWP/194898/2006, December 2007.Guideline on the clinical development of medicinal products for the treatment of

HIV infection. Doc. Ref. EMEA/CPMP/EWP/633/02, Revision 2, November2008.

Literature References

Alden, S.L., Smith, P.F., Piper, C.E., Brei, L., 1996. A critical appraisal of the value ofthe mouse cancer bioassay in safety assessment. Toxicol. Pathol. 24, 722–725.

Bercu, J.P., Jolly, R.A., Flagella, K.M., Baker, T.K., Romero, P., Stevens, J.L., 2010.Toxicogenomics and cancer risk assessment: a framework for key event analysisand dose-response assessment for nongenotoxic carcinogens. Regul. Toxicol.Pharmacol. 58, 369–381.

Contrera, J.F., Jacobs, A.C., DeGeorge, J.J., 1997. Carcinogenicity testing and theevaluation of regulatory requirements for pharmaceuticals. Regul. Toxicol.Pharmacol. 25, 130–145.

Ennever, F.K., Lave, L.B., 2003. Implications of the lack of accuracy of the lifetimerodent bioassay for predicting human carcinogenicity. Regul. Toxicol.Pharmacol. 38, 52–57.

Greaves, P., 2007. Histopathology of preclinical toxicity studies, third ed. ElsevierInc., New York, London.

Monro, A., 1996. Are lifespan rodent carcinogenicity studies defensible forpharmaceutical agents? Exp. Toxicol. Pathol. 48, 155–166.

Reddy, M.V., Sistare, F.D., Christensen, J.S., DeLuca, J.G., Wollenberg, G.K., DeGeorge,J.J., 2010. An evaluation of chronic 6- and 12-month rat toxicology studies aspredictors of 2-year tumor outcome. Vet. Pathol. 47, 614–629.

Sistare F.D. (2010). Lessons learned from an analysis of pharmaceutical experiencewith decades of rat carcinogenicity testing. Abstract, Workshop presentation atthe 49th Annual Meeting of the Society of Toxicology, Salt Lake City, Utah,USA.

Smith, P.F., 1996. Third international conference on harmonisation of technicalrequirements for registration of pharmaceutical for human use - a toxicologist’sperspective. Toxicol. Pathol. 24, 519–528.

Van Ousterhout, J.P.J., van der Laan, J.W., de Waal, E.J., Olenjniczak, K., Hilgenfeld,M., Schmidt, V., Bass, R., 1997. The utility of two rodent species in carcinogenicrisk assessment of pharmaceuticals in Europe. Regul. Toxicol. Pharmacol. 25, 6–17.

Vinken, M., Doktorova, T., Ellinger-Ziegelbauer, H., et al., 2008. ThecarcinoGENOMICS project: critical selection of model compounds for thedevelopment of omics-based in vitro carcinogenicity screening assays. Mutat.Res. 659, 202–210.



regulatory toxicology and pharmacology€¦ · a pharmaceutical is tumourigenic in animals and...

Documents