Lifelong exposure to di-(2-ethylhexyl)-phthalate induces tumors in liver and testes of...

Toxicology 206 (2005) 359–371

Lifelong exposure to di-(2-ethylhexyl)-phthalate inducestumors in liver and testes of Sprague–Dawley rats

Cristina Vossa, Heide Zerbanb, Peter Bannaschb, Martin R. Bergera,∗

a Unit of Toxicology and Chemotherapy, Deutsches Krebsforschungszentrum Heidelberg, Im Neuenheimer Feld 280,69120 Heidelberg, Germany

b Division of Cell Pathology, Deutsches Krebsforschungszentrum Heidelberg, Im Neuenheimer Feld 280, 69120 Heidelberg, Germany

Received 24 June 2004; received in revised form 21 July 2004; accepted 21 July 2004Available online 11 September 2004

Abstract

The plasticizer di-(2-ethylhexyl)-phthalate (DEHP) is the most important phthalate with respect to its production, use andoccurrence in the environment. In standard carcinogenicity experiments with F344 rats and B6C3F1 mice, DEHP has beenshown to induce hepatocellular tumors. Moreover, DEHP is strongly suspected to be a developmental and reproductive toxicant.The present study aimed at determining the long-term toxic effects of lifetime exposure to low concentrations of DEHP inSprague–Dawley rat strain. Seven hundred and thirty male rats, stratified into four groups, received DEHP with the diet, resultingin dosages of 300, 95, 30 and 0 mg/kg per day for up to 159 weeks and were only sacrificed when moribund. All organs of the deada 300 mg/kgp r)w developede ed to theh wsf exposure.T an healthr ce in theo©

K posure;T

0

nd sacrificed animals were histopathologically examined. Significantly increased tumor incidences after exposure toer day DEHP (P= 0.04 for testes and 0.05 for liver) and a significant dose-related trend (PTrend= 0.02 for testes and 0.03 for liveere detected in both organs liver and testes. Time to tumor analysis revealed that DEHP-induced testicular tumorsarlier in lifetime than hepatocellular neoplasias, and their multiplicity increased with time. In addition, animals exposighest DEHP dose showed a significantly increased rate of testicular tubular atrophy (P< 0.01). In conclusion, this study sho

or the first time that the rat testes are a target organ of DEHP carcinogenicity in Sprague–Dawley rats upon lifetimehis new finding indicates the importance of evaluating the effects of lifetime exposure in assessing the potential hum

isks of DEHP. In addition, the carcinogenicity should be evaluated in rat strains with low spontaneous tumor incidenrgans known as target of DEHP toxicity.2004 Elsevier Ireland Ltd. All rights reserved.

eywords: Di-(2-ethylhexyl)-phthalate; DEHP; Leydig cell tumors of the testes; Epigenetic carcinogenicity; Rat liver; Lifelong exoxicity

∗ Corresponding author. Tel.: +49 6221 423310; fax: +49 6221 423313.E-mail addresses:[email protected], [email protected] (M.R. Berger).

300-483X/$ – see front matter © 2004 Elsevier Ireland Ltd. All rights reserved.doi:10.1016/j.tox.2004.07.016

360 C. Voss et al. / Toxicology 206 (2005) 359–371

1. Introduction

The phthalate ester di-(2-ethylhexyl)-phthalate(DEHP, CAS No. 117-81-7) is the commercially mostimportant plasticizer for PVC plastics (Mersiowsky etal., 2001). According to the European Council for Plas-ticizer and Intermediates, DEHP accounts for 45% ofall plasticizer usage in Western Europe. DEHP is ubiq-uitously distributed in the environment and poses a po-tential risk to human health, since it can leach fromDEHP-containing products and wastes and may be re-leased by burning of plastics. The health risks for hu-mans related to DEHP exposure have been repeatedlyinvestigated (Tickner et al., 2001). Recently, the me-dian daily DEHP intake of the general population inGermany was estimated to be 13.8�g/kg body weight,a value much higher than previously believed (Koch etal., 2003). This value is in agreement with the rangeof DEHP intake estimated by an expert panel of theUS NTP Center (3–30�g/kg body weight per day).Furthermore, especially critically ill neonates and in-fants could be exposed to considerably higher DEHPdoses as a result of intensive medical care, since medi-cal equipment is frequently made of DEHP-containingPVC (Kavlock et al., 2002).

In a study of the US National Toxicology Program,DEHP was shown to be a hepatocarcinogen in ratsand mice (Kluwe et al., 1982). Consequently, the USEnvironmental Protection Agency (US-EPA) classifiedDEHP in 1992 as a Group B2 (probable human) car-c tsa m inv tor(f tedta gensi tor-a era-t rateo eralg ncebH ly top fac-t rmsi ARt

For this reason, the International Agency for Researchon Cancer (IARC) reclassified DEHP from category2B (possible carcinogenic to humans) to category 3(cannot be classified as to its carcinogenity to humans;IARC, 2000).

In addition to the carcinogenic risk, several stud-ies (reviewed, e.g. byTickner et al., 2001) revealedthat DEHP exposure of animals and humans can leadto toxic effects in organs including liver, male and fe-male reproductive tracts, kidneys, lungs and heart. Thefocus, however, is lately shifting towards the toxic ef-fects to the reproductive system, which is believed tobe more critical. DEHP is a well-known developmen-tal and reproductive toxicant in rats (Kavlock et al.,2002). WhenShaffer et al. (1945)published the firstreport on testicular injury in an animal model, DEHPhas been shown to cause apoptosis and degeneration ofseminiferous tubules with loss of spermatogenic cellsin male rats, resulting in testicular atrophy (Gray andGangolli, 1986; Poon et al., 1997; Gray et al., 2000;Park et al., 2002). Additionally,Akingbemi et al. (2004)found an increased proliferative activity of Leydig cellsin the testes of DEHP-treated rats, indicating inductionof Leydig cell hyperplasia.

In an attempt to study the long-term toxic effectsof DEHP, we administrated low dosages of DEHP tomale Sprague–Dawley rats in a lifetime experiment. Inthis paper we report the observed lesions in responseto this treatment, emphasizing the liver and testes car-cinogenicity of DEHP in this rat strain.

2

2

meP ny( fac-t 9%.

2

ley( er-B 10d gen-f ge

inogen (US-EPA, 2000). However, DEHP neither acs a direct genotoxic agent in standard short-teritro or in vivo assays, nor as an in vivo tumor initiafor review, seeBudroe and Williams, 1993). The ef-ect of DEHP on the rodent liver is thought to be relao its function as a peroxisome proliferator (Klaunig etl., 2003). These substances act as hepatocarcino

n species expressing active peroxisome proliferactivated receptor (PPAR) and peroxisome prolif

or response elements (PPREs), by reducing thef apoptosis and stimulating the activities of sevrowth-regulatory genes, thereby altering the balaetween cell growth and apoptosis (Corton et al., 2000).owever, humans and rodents respond differenteroxisome proliferators, due to species-specific

ors, such as lower concentration of PPAR isofon tissues and an inability of activated human PPo trigger downstream events (Klaunig et al., 2003).

. Materials und methods

.1. Chemicals

Di-(2-ethylhexyl)-phthalate (DEHP; trade naalatinol AH) was obtained from BASF Compa

Ludwigshafen, Germany). According to the manuurer, the purity of the chemical was greater than 9

.2. Animals

Seven hundred and thirty male Sprague–DawSD-CD) rats were obtained from Charles-Rivreeding, Sulzfeld, Germany, at an age of 100 +ays. The animals were kept under specific patho

ree conditions: two animals per Makrolon III ca

C. Voss et al. / Toxicology 206 (2005) 359–371 361

with granulated sawdust as bedding. The relative airhumidity (55 + 10%), room temperature (23 + 2◦C)and light/darkness rhythm (12 h) were regulated, as wasthe turnover of filtered air (20 fold/h). The cages wereinspected twice daily. All animals received autoclavedtap water ad lib and were fed a nitrosamine-free diet(1320 N Altromin, Lage, Germany).

For the study, the rats were stratified into fourgroups. DEHP was mixed into the food at 6000, 1897and 600 mg/kg diet 1320 N. DEHP-treated animalswere fed 5 g DEHP-diet/100 g rat/day for 6 days/weekand received DEHP-free food on the seventh day onlyafter they had consumed the rest of their DEHP-diet.On this basis, the DEHP-treated rats were exposedto 300 (n = 60), 95 (n = 100) and 30 mg DEHP/kg(n = 180) respectively. The animals in the controlgroup (n = 390) were fed an equicaloric DEHP-freediet. The different group sizes were based on es-timated tumor incidences, to enhance the statisticalpower for detecting effects in the medium and low dosegroups.

The DEHP-treatment was applied over the entirelifetime of the animals (up to 159 weeks). All ratswere inspected twice daily and weighed weekly (first3 months) or monthly (after 3 months). They were ob-served for their entire lifespan and sacrificed only whenmoribund, according to the intention of a lifetime study.Sacrificed animals totaled to subgroups ofn= 167 (con-trol), 84 (low DEHP group), 53 (medium DEHP group),and 31 (high DEHP group). Spontaneously deceaseda d form thel

2

2thy-

r opi-c rou-t ineg sec-t nde forH nta-t oft

2.3.2. Additional detailed histopathologicalexamination of the liver

In order to assess pre-neoplastic changes, the liverof sacrificed animals was additionally subjected to adetailed liver-specific histopathological examination.Liver slices were fixed in Carnoy’s fluid. Paraffin sec-tions were stained with haematoxylin and eosin (H&E)or treated with the periodic acid-Schiff (PAS) reagentto detect glycogen and counterstained with orange Gand iron haematoxylin. Pre-neoplastic foci of alteredhepatocytes were classified as described previously:clear/acidophilic cell foci, basophilic/mixed cell foci,tigroid/X-cell foci and amphophilic cell foci (Bannaschet al., 1997). In addition, other types of pre-neoplasticand all neoplastic liver lesions (hepatocellular ade-noma and carcinoma; cholangiofibrosis and cholan-giofibroma; peliosis hepatis, hemangioendotheliomaand hemangiosarcoma; spongiosis hepatis and pericy-toma; cirrhosis, fatty liver degeneration and infiltratesfrom systemic diseases) were distinguished accord-ing to published criteria (Bannasch and Zerban, 1997;Jones et al., 1997).

2.4. Statistics

Survival times of treated and untreated rats werecompared using the Kruskal–Wallis test (Kalbfleischand Prentice, 1980). Tumor rates of all organs werestatistically analyzed by comparing the observed anda ng toP ichs val-u rel-e rcino-m blad-d lat-e ap-p lif-e tiono roupsw c-c rea es, asi -t lcu-l ilityo

nd sacrificed animals were dissected and checkeacroscopically manifest pathologic lesions and

iver was routinely weighed.

.3. Histology

.3.1. General histopathological examinationThe organs brain, liver, adrenal glands, testes,

oid gland, lungs and spleen, as well as all macroscally detectable lesions and atypical findings wereinely fixed in 7% formalin and processed for a routeneral histopathological examination. Paraffin

ions (5�m thick) were stained with hematoxylin aosin and examined by Dr. D. Komitowski (Dept.istodiagnostics and Pathomorphologic Docume

ion, DKFZ, Heidelberg) according to the criteriahe IARC (Turusov and Mohr, 1990).

ge-adjusted expected overall incidences accordieto et al. (1980). For the organs liver and testis, whhowed increased overall tumor rates at the initial eation, a statistical analysis was performed onvant diagnoses. These included adenomas, caas, spongiosis hepatis, peliosis hepatis and galler proliferates for the liver as well as uni- and biral occurrence of Leydig cell tumors, multifocalearance of Leydig cell tumors, interstitial cell prorates and tubular atrophy for the testes. Distribuf these diagnoses between treated and control gere compared by theχ2-test. In addition, the test aording toPeto et al. (1980)was applied to compage-adjusted expected versus observed incidenc

ndicated inTables 2–5. The reportedP-values are twoailed except for those specifically indicated as caated by using a dose-related trend-test. A probabf 0.05 was used to determine significance.


The results of the detailed histopathological anal-ysis of the liver collected from the sacrificed animalswere evaluated independently. In addition to the cate-gories of the general liver evaluation (adenomas, car-cinomas, spongiosis hepatis, peliosis hepatis and gall-bladder proliferates), special attention was given to theoccurrence of liver pre-neoplastic foci as well as toother liver diagnoses (cirrhosis, fatty liver degenera-tion and infiltrates resulting from systemic diseases).The diagnoses were statistically analyzed according toPeto et al. (1980), as well as by theχ2-test, as describedabove.

A time to tumor analysis was performed by splittingthe total observation time into three periods: 0–750,750–950 and 950–1150 days. The respective inci-dences for each time period were compared by theχ2-test and a linear test for trend was applied over the threeconcentrations.

3. Results

The animals receiving the DEHP-diet did not showany signs of systemic toxicity. The weight gain of thetreated rats was similar to that of the control group, ex-cept for a short period around day 300 after the startof the experiment, when the body weight of DEHP-fedanimals was lower than that of the control. However,this weight loss was quickly compensated during thefollowing period (Fig. 1). Treated and untreated ani-m sur-v of

F o untre than 10%)w betwee

DEHP-fed rats during the second part of the experimentwas not significant (Fig. 2).

Cumulated incidences of all types of malignant andbenign tumors did not differ significantly between con-trol and DEHP-treated rats (Table 1).

Analysis of organ-specific tumor incidences showedfor the majority of organs no modulation of tumor in-cidences by DEHP, specifically not for those that arecommon in aging male Sprague–Dawley rats, like pi-tuitary gland and adrenal gland tumors and for tu-mors of the liver originating from endothelial, peri-cytic and cholangiocellular cells (Table 2). However,significant differences between DEHP-treated and con-trol groups were observed for hepatocellular lesionsof the liver and Leydig cell tumors of the testes (seebelow).

Besides neoplastic changes, the DEHP-treatedgroups showed dose-dependent slight increases inmean liver weight of 18.2± 7.6 g (108%, 300 mg/kggroup), 17.5± 6.5 g (104%, 95 mg/kg group), and 17.2± 7.8 g (102%, 30 mg/kg group) over control (16.8±5.8 g), which were not significant. A slightly but notsignificantly increased incidence of spongiosis hep-atis was observed following treatment with 300 mg/kgDEHP only (29% versus 22% for the control), whenthe overall tumor distributions were compared. Inter-estingly, significantly increased incidences of spongio-sis hepatis were observed in the time to tumor analy-sis for rats receiving the highest and medium DEHPdose for more than 950 days (15 and 13% for 300 and9 ,r

als did not show any difference in their medianival (Table 1). The slightly prolonged survival time

ig. 1. Body weight of rats treated with DEHP in comparison tas omitted for clarity. No significant difference was observed

ated control animals. Plotting of the standard deviation (lessn all experimental groups.

5 mg/kg versus 9% in the control;P = 0.03 and 0.04espectively).


Table 1Median survival time and overall tumor rates

Group Na Median survival time Malignant tumors Benign tumors Malignant tumorsper ratb

Benign tumorsper ratb

Days Pc nd % nd %

Control 390 946 (928–1062)e – 136 35 311 80 1.1± 0.3 2.0± 1.0DEHP 30 mg/kg 180 948 (902–1001)e >0.1 66 37 144 80 1.1± 0.2 1.9± 0.9DEHP 95 mg/kg 100 970 (901–1027)e >0.1 33 33 76 76 1.1± 0.2 1.9± 0.9DEHP 300 mg/kg 60 974 (859–1041)e > 0.1 24 40 44 73 1.0± 0.0 1.8± 0.8

a Number of animals per group.b Number of tumors± S.D. per tumor bearing animal.c Kruskal–Wallis test.d Number of animals with tumors.e 95% confidence interval.

In addition, significantly increased tubular atrophyof the testis (reduced spermatogenesis) was observed(see below). All other diagnoses were equally dis-tributed between treated and control groups.

3.1. Hepatocellular lesions of the liver

The general histopathological examination of theliver (Table 3) revealed a slight increase in the per-centage of rats with hepatocellular carcinomas follow-

F to DEH HP did noti

ing treatment with the highest DEHP concentration(300 mg/kg) as compared to the control group (6.7%versus 4.9%), but this difference did not reach signifi-cance. However, the detailed histopathological exami-nation of the liver, which was confined to sacrificed an-imals only, revealed a significant (P = 0.005) increasein the incidence of all neoplasias (benign and malig-nant, 29.0%) in the highest DEHP concentration groupcompared with the control group (9.0%,Table 4). Thetime to tumor analysis showed that this effect resulted

ig. 2. Survival curves of male Sprague–Dawley rats exposednfluence the survival of treated rats.

P compared to untreated control animals. The exposure to DE


Table 2Organ-specific tumor incidences in Sprague–Dawley rats after lifetime exposure to DEHP, as revealed by the general histopathologicalexamination

Diagnosisa Control DEHP30 mg/kg

DEHP95 mg/kg

DEHP300 mg/kg

nb % nb % nb % nb %

Animals per group 390 180 100 60

Liver Endothelial cell tumors 1 1 0 0 0 0 2 1Cholangiocellular tumors 72 40 47 47 20 33 170 43Pericytic cell tumors 5 3 4 4 3 5 13 3

Urinary tract Tumors of the urinary bladder 0 0 1 1 2 3 7 2Tumors of the kidneys 0 0 1 1 1 2 4 1All tumorsc 0 0 2 2 3 5 11 3

Gastro-intestinal tract Tumors of the oral cavity 0 0 0 0 0 0 1 0Tumors of the esophagus 0 0 0 0 0 0 0 0Tumors of the stomach 2 1 0 0 1 2 4 1Tumors of the small and large bowel 0 0 0 0 0 0 1 0All tumorsc 2 1 0 0 1 2 6 2

Hematopoietic andlymphatic system

Leukemias 11 6 5 5 6 10 17 4Lymphomas 6 3 3 3 0 0 6 2Histiozytic tumors 8 4 4 4 0 0 23 6All tumorsc 25 14 12 12 6 10 44 11

Other organs Neurogenic tumors 9 5 5 5 2 3 17 4Tumors of the mammary gland 15 8 5 5 1 2 23 6Skin tumors 9 5 3 3 1 2 19 5Osteosarcomas 1 1 1 1 2 3 4 1Tumors of the peritoneal cavity 5 3 3 3 1 2 7 2Adrenal gland tumors 84 47 49 49 28 47 203 52Pituitary gland tumors 80 44 47 47 24 40 180 46Thyroid gland tumors 7 4 4 4 1 2 20 5Tumors from lymph nodes and spleen 1 1 0 0 0 0 3 1Tumors of the heart 0 0 1 1 0 0 0 0Tumors of the lung 0 0 0 0 0 0 4 1Pancreatic acinar cell tumors 3 2 1 1 1 2 11 3Tumors of the external auditory canal 1 1 1 1 0 0 6 2Tumors of the parathyroid gland 14 8 6 6 5 8 45 12

a Differences between DEHP-treated and control groups (according toPeto et al., 1980) were not significant.b Number of animals with tumors.c For two or more tumors only one tumor per animal was counted.

from an increased tumor incidence which occurred inthe second and third time periods but became signifi-cant only late in the lifetime of the animals (Fig. 3). Inaddition, the overall incidences of hepatocellular ade-nomas (19.4% versus 7.8% in the control) and carci-nomas (9.7% versus 1.2% in the control) were signif-icantly higher in the highest concentration group (P =0.045 for adenomas,P = 0.006 for carcinomas). Nosignificant differences in the rates of neoplasias were

observed between the lower DEHP groups (95 and30 mg/kg) and the control animals. Nevertheless, a sig-nificantly positive dose-related trend over all three dosegroups was observed for all neoplasias, when the resultsof the detailed histopathological liver examination ofall three dose groups were analyzed (PTrend = 0.001;Table 4).

The overall incidence of foci did not differ be-tween treated and control groups (Table 5). In-


Table 3Incidences of hepatocellular lesions in Sprague–Dawley rats following lifetime exposure to DEHP, as revealed by the general histopathologicalexamination of the liver

Liver tumors Control Exposure to DEHP Dose-relatedtrend,PTrend

na % 30 mg/kgb 95 mg/kgb 300 mg/kg

na % na % na % P

Total no. of rats 390 100 180 100 100 100 60 100Adenomas 16 4.1 5 2.8 2 2.0 1 1.7 >0.05c

Carcinomas 19 4.9 11 6.1 3 3.0 4 6.7 >0.05c

All neoplasiasd 35 9.0 16 8.9 5 5.0 5 8.3 >0.05e >0.05f

a Number of animals with tumors.b Differences to control were not significant (two-tailedχ2-test and test according toPeto et al., 1980).c Compared with control (two-tailed,χ2-test).d Hepatocellular adenomas and carcinomas.e Compared with control according toPeto et al. (1980)(two-tailed, time-dependent).f Test for dose-related trend according toPeto et al. (1980)(one-tailed, time-dependent).

Table 4Incidences of hepatocellular lesions in Sprague–Dawley rats for the subgroups of sacrificed animals, following lifetime exposure to DEHP, asrevealed by the detailed histopathological examination of the liver

Liver tumors Control Exposure to DEHP Dose-relatedtrend,PTrend


na % na % na % P

Total no. of rats 167 100 84 100 53 100 31 100Adenomas 13 7.8 3 3.6 4 7.5 6 19.4 0.045c

Carcinomas 2 1.2 3 3.6 0 0.0 3 9.7 0.00c

All neoplasiasd 15 9.0 6 7.1 4 7.5 9 29.0 0.005e 0.001f

a Number of animals with tumors.b Differences to control were not significant (two-tailedχ2-test and test according toPeto et al., 1980).c Compared with control (two-tailed,χ2-test).d Hepatocellular adenomas and carcinomas.e Compared with control according toPeto et al. (1980)(two-tailed, time-dependent).f Test for dose-related trend according toPeto et al. (1980)(one-tailed, time-dependent).

Table 5Incidences of hepatocellular foci in Sprague–Dawley rats following lifetime exposure to DEHP

Group No. of rats Rats with liverfocia

Phenotypes of focib

Clear/acidophilic Basophilic/mixed Tigroid/X-cell Amphophilic

Control 167 nb 88 76 12 8 2% 52.7 45.5 7.2 4.8 1.2

DEHP 30 mg/kg 84 nb 43 39 5 6 1% 51.2 46.4 6.0 7.1 1.2

DEHP 95 mg/kg 53 nb 23 19 2 4 4% 43.4 35.8 3.8 7.5 7.5

DEHP 300 mg/kg 31 nb 16 14 2 3 1% 51.6 45.2 6.5 9.7 3.2

a Differences in the incidence of liver foci between DEHP-exposed and control rats were not significant (χ2-test).b Number of animals with foci (rats showing two or more phenotypes of liver foci were counted only once).


Fig. 3. Incidence of hepatocellular tumors as revealed by time to tumor analysis. The percentage of rats with hepatocellular tumors within therespective time period was determined from the total number of animals sacrificed during the respective time period. A significant increase(marked by an asterisk) was observed for the highest dose group in the third time period.

terestingly, there was a dose-dependent tendencyfor foci of the tigroid/X-cell and amphophilic typeto show increased incidences after treatment withDEHP.

3.2. Lesions of the testes

In addition to the liver, rat testes were found tobe most affected by the DEHP treatment (Table 6).The percentage of benign Leydig cell tumors in thehighest DEHP group (300 mg/kg) was almost twiceas high as the percentage in the control group (28.3%versus 16.4%,P = 0.038). Over all three DEHP con-centrations, a significant dose-related trend of Leydigcell tumor formation was observed (PTrend = 0.019).An independent evaluation of the uni- and bilateralLeydig cell tumor incidences revealed an additionaltrend. The highest tumor incidence (20.0% unilat-eral, 8.3% bilateral tumors) was observed followingthe highest DEHP exposure; concentration-dependentlower incidences were detected following the lower ex-posures (17.0% unilateral, 4.0% bilateral and 16.7%unilateral, 2.2% bilateral tumors following 95 and30 mg DEHP/kg, respectively). The lowest incidences

(13.1% unilateral, 3.3% bilateral tumors) were ob-served in the control group. Although comparisonsbetween single DEHP and control groups were notsignificant when analyzed with theχ2-test, a signif-icant trend over all groups was observed for unilat-eral Leydig cell tumors (PTrend = 0.046). In addi-tion to uni- and bilateral Leydig cell tumor formation,the multifocal appearance within one testis was alsoevaluated. Increased rates of multifocal Leydig celltumors were observed with increasing DEHP doses(16.7, 5.0 and 7.8% following 300, 95 and 30 mg/kgDEHP, respectively), compared with the control group(4.1%, Table 6). The difference between the high-est DEHP and control groups was significant (P <0.001) and a dose-related trend according toPetoet al. (1980)was observed (PTrend < 0.001).

The incidence of testicular neoplastic lesions in-creased with age in all experimental groups. In the high-est DEHP group, Leydig cell tumors were observedonly after an age of 700 days (Fig. 4). The time to tumoranalysis showed a significantly increased incidence ofLeydig cell neoplasias in the highest dose group and aclear dose-related trend only for the second time period(750–950 days lifetime,P = 0.018 andPTrend = 0.015,


Table 6Incidences of testicular lesions in Sprague–Dawley rats following lifetime exposure to DEHP

Testicular lesions Control Exposure to DEHP Dose-relatedtrend,PTrend


na % na % na % P

Total no. of rats 390 100 180 100 100 100 60 100

Leydig cell tumorsAll 64 16.4 34 18.9 21 21.0 17 28.3 0.038c 0.019d

Unilateral 51 13.1 30 16.7 17 17.0 12 20.0 >0.05e 0.046f

Bilateral 13 3.3 4 2.2 4 4.0 5 8.3 >0.05e >0.05f

Multifocal 16 4.1 14 7.8 5 5.0 10 16.7 <0.001c <0.001d

Leydig cell proliferates 54 13.8 32 17.8 15 15.0 10 16.7 >0.05c >0.05d

Tubular atrophy 176 45.1 75 41.7 45 45.0 43 71.7 <0.001c <0.001d

a Number of animals with testicular lesions.b Differences to control were not significant (two-tailedχ2-test and test according toPeto et al., 1980).c Compared with control according toPeto et al. (1980)(two-tailed, time-dependent).d Test for dose-related trend according toPeto et al. (1980)(one-tailed, time-dependent).e Compared with control (two-tailed,χ2-test).f χ2-test for trend.

Fig. 5a). Remarkably, grading of the Leydig cell tu-mors by their unilateral versus bilateral and multifocaloccurrence showed a clear time-dependent increase inthe multiplicity of tumors with time and dose. Specif-ically, the incidence of unilateral tumors was signifi-

Fig. 4. Occurrence of Leydig cell tumors related to survival following exposure to DEHP. Rats treated with the highest DEHP dose (300 mg/kg)developed Leydig cell tumors only after 700 days of life.

cantly increased in animals of the highest dose groupfor the second time period (P= 0.008), and a significantdose-related trend was seen over all dose groups (PTrend= 0.008,Fig. 5b). In the third time period, a comparablyhigh incidence was observed for unilateral tumors over


Fig. 5. Incidence of Leydig cell tumors as revealed by time to tumor analysis. The percentage of rats with Leydig cell tumors within the respectivetime period was determined from the total number of animals during the respective time period. Significant increases are marked by an asterisk.The incidence of all Leydig cell tumors (a) was grouped into unilateral Leydig cell tumors (b) and bilateral Leydig cell tumors (c), as well asmultifocal Leydig cell tumors (d).

all dose and control groups, but a significant increasein the incidence of bilateral tumors (P= 0.030,Fig. 5c)became obvious in the highest dose group, with a sig-nificant dose-related trend over all dose groups (PTrend= 0.0002). Similarly, the incidences of multifocal Ley-dig cell tumors were significantly increased in the sec-ond (P = 0.007) and third (P = 0.009) time periodsfor the highest dose group, with corresponding dose-related trends over all dose groups (PTrend = 0.014 and0.046, respectively,Fig. 5d).

Interstitial (Leydig) cell proliferates showed no sta-tistically significant differences (Table 6). Significantlyincreased tubular atrophy (reduced spermatogenesis)was observed in the highest DEHP group when com-pared to the control group (71.7% versus 45.1%,P< 0.001). In addition, a significant trend over thethree DEHP concentrations was detected (P < 0,001,Table 6).

4. Discussion

This lifetime study with 730 male Sprague–Dawleyrats demonstrated that prolonged treatment with DEHPcaused an increase in the incidence of hepatocellularand Leydig cell tumors as compared with the untreatedcontrol. The control group was sufficiently large to al-low a reliable differentiation between DEHP-specificand aging effects. The differences were significant forthe highest DEHP dose (300 mg/kg per day), and overall dosage groups a significant dose-related trend wasobserved.

The peroxisome proliferator DEHP has been shownto be a hepatocarcinogen in rats and mice (Kluwe etal., 1982; David et al., 1999). The postulated mode ofaction of DEHP in rat liver involves hydrolysis of thedi-ester to its mono-ester metabolite MEHP and sub-sequent activation of the PPAR. The activated receptor


triggers DNA transcription of genes regulating perox-isomal enzymes, cell proliferation and apoptosis thatlead to selective clonal expansion of cells resulting inadenomas and carcinomas (Klaunig et al., 2003).

The increased incidence of hepatocellular tumorsobserved in this study is therefore consistent with theliterature. However, at variance with Fischer 344 rats,which showed significantly increased incidences ofhepatocellular tumors in response to 140 mg DEHP/kgper day (2500 ppm,David et al., 1999) in a 2-yearbioassay, the Sprague–Dawley rats used in this life-time study developed more liver tumors than the con-trol only following a two-fold higher DEHP dose of300 mg/kg per day and at the very end of their life-time (>950 days,Fig. 3). In line with this somewhatreduced sensitivity, there was no significant differencebetween the overall incidence of spongiosis hepatis inthe DEHP-exposed and control Sprague–Dawley rats,but only in rats treated with the highest DEHP dosefor more than 950 days. This contrasts to observationsin Fischer 344 rats, where a significant increase in therate of spongiosis hepatis was observed at DEHP dosesof 150 mg/kg per day after 2 years. Also, there was noincrease in the incidence of pancreatic acinar cell ade-noma, as observed for Fischer rats (David et al., 2000).

This lifetime study revealed also that chronic DEHPexposure can induce Leydig cell tumors in the testes ofadult rats. This finding was proven not only by a sig-nificant increase in the incidence of Leydig cell tumorsin the high dose group and a significant dose-relatedt asedm o-c theh ose-r

cellt cea , i.e.e lti-p n-c canti sec-o int pe-r ors,D tifo-c riods(

The DEHP carcinogenicity studies available so farwere performed in Fischer 344 rats. This rat strain isknown to spontaneously develop Leydig cell tumors ata high rate (Prentice and Meikle, 1995), making it dif-ficult if not impossible to detect a xenobiotic-inducedtesticular tumor in this strain. Compared to the resultswith Fischer rats, Sprague–Dawley rats in this studyshowed a distinctly lower incidence of spontaneousLeydig cell tumors (16%). In addition, the DEHP-induced Leydig cell tumors developed only late in thelifetime of Sprague–Dawley rats, at ages when standardcarcinogenicity studies have already been terminated.

Although new, our finding on Leydig cell tumorsis not surprising. A number of other peroxisomalproliferators (DINP, gemfibrozyl, HCFC-123, methyl-clofanepate, perchlorethylene, perfluorooctanoic acid,trichloroethylene and WY 14,643) have been reportedto induce testicular tumors as well (Klaunig et al.,2003). However, the underlying mechanism is differ-ent from that in hepatocytes since steroid hormoneeffects are involved. PPAR agonists are reported toinduce Leydig cell tumors by inhibiting testosteronebiosynthesis and/or by inducing aromatase, therebyincreasing estradiol levels (Klaunig et al., 2003). Inline with this, DEHP was found to induce high levelsof gonadotropin-luteinizing hormone and to increaseserum concentrations of estradiol and testosterone inLong-Evans rats exposed to 10 and 100 mg/kg per dayDEHP for up to 100 days. Also, Leydig cell hyperpla-sia was demonstrated, which explained the increaseds teroneb

y–t on-c tes.R anelr anti sd ophya reat-mep cher3(e e toi nice

rend over all dose groups, but also by the increultiplicity of these tumors: the incidence of multif

al testicular lesions was significantly increased inigh dose group and a corresponding significant delated trend was observed over all groups.

Our time-to-tumor analysis showed that Leydigumors occurred with a significantly higher incidenlready in the second time period (750–950 days)arlier than hepatocellular tumors. In addition, mulicity of the DEHP-induced Leydig cell tumors ireased with time, as indicated by the early signifincrease in the rate of unilateral tumors during thend time period, followed by a significant increase

he rate of bilateral tumors during the third timeiod. In contrast to spontaneous Leydig cell tumEHP-induced tumors were characterized by mulal appearance in the second and third time peFig. 5).

erum testosterone level despite a reduced testosiosynthesis per Leydig cell (Akingbemi et al., 2004).

The disruption of the hypothalamic–pituitaresticular hormone axis is also responsible for narcinogenic toxic effects of DEHP on the rat tesecently, the US NTP Center’s phthalates expert p

eport concluded that DEHP is a reproductive toxicn male rats (Kavlock et al., 2002). Several reportescribe an increased occurrence of tubular atrnd reduced spermatogenesis following DEHP tent of adult and developing rats (reviewed byTicknert al., 2001; Kavlock et al., 2002). Consistent withrevious observations in DEHP-exposed adult Fis44 (Kluwe et al., 1982; David et al., 2000), WistarDalgaard et al., 2000) and Sprague–Dawley rats (Poont al., 1997), our data demonstrate that DEHP is abl

nduce tubular atrophy in adult rats following chroxposure.


The NOAEL for liver carcinogenic effects of DEHPhas been proposed to be 29–36 mg/kg derived froma study in Fischer rats (David et al., 2000). InSprague–Dawley rats from this study, the respectiveNOAEL was found to be 95 mg/kg. The same NOAELvalue was found for carcinogenic effects in SD rattestes. However, the per se non-significant (two-tailedtest) increases in Leydig cell tumor incidence follow-ing 30 and 95 mg/kg DEHP were significant in a one-tailed test for trend. These results are consistent withthe observation ofGanning et al. (1991), who noted tes-ticular lesions but no liver tumors in Sprague–Dawleyrats after 2 years treatment with up to 2% DEHP inthe diet. Derived from a subchronic oral toxicity study,Poon et al. (1997)proposed a relatively low NOAELof 3.7 mg DEHP/kg per day for Sprague–Dawley rats,which reflects the ability of DEHP to induce Sertoli cellvacuolization, as an early indicator of testicular toxi-city. In this study, the NOAEL for tubular atrophy, asa marker for non-carcinogenic testicular toxicity, wasfound to be 95 mg/kg. However, the high rate of spon-taneous tubular atrophy (45%) in 2–3 years old rats ofthis study may have impeded the discrimination of alower NOAEL.

As a consequence of its testicular carcinogenicityin Sprague–Dawley rats, potential implications of hu-man exposure to DEHP need to be considered. Theeffect of DEHP on the rodent liver is thought to berelated to its function as a PPAR agonist but the avail-able data lead to the conclusion that a carcinogenicr hism cellt vi-da xi-cA per-o althe

firstt er,b ter-e tu-m ose,w rcent-a n att ob-a ent,

since induction of Leydig cell tumors in the rat testesare to be considered of concern, as long as the mode ofinduction by DEHP is not fully understood. The newaspect of DEHP toxicity indicates that more researchon the risks of long-term and lifetime DEHP exposureis needed.

References

Akingbemi, B.T., Ge, R., Klinefelter, G.R., Zirkin, B.R., Hardy,M.P., 2004. Phthalate-induced Leydig cell hyperplasia is asso-ciated with multiple endocrine disturbances. Proc. Natl. Acad.Sci. U.S.A. 101, 775–780.

Bannasch, P., Zerban, H., Hacker, H.J., 1997. Foci of altered hepa-tocytes rat. In: Jones, T.C., Popp, J.A., Mohr, U. (Eds.), Mono-graphs on Pathology of Laboratory Animals: Digestive System,2nd ed. Springer, Heidelberg, pp. 3–37.

Bannasch, P., Zerban, H., 1997. Experimental chemical hepatogen-esis. In: Okuda, K., Tabor, E. (Eds.), Liver Cancer. Livingstone,New York, pp. 213–253.

Budroe, J.D., Williams, G.M., 1993. Genotoxicity studies of peroxi-some proliferators. In: Gibson, G., Lake, B. (Eds.), Peroxisome:Biology and Importance in Toxicology and Medicine. Taylor &Francis, London, pp. 525–568.

Corton, J.C., Lapinskas, P.J., Gonzalez, F.J., 2000. Central role ofPPARalpha in the mechanism of action of hepatocarcinogenicperoxisome proliferators. Mutat. Res. 448, 139–151.

Dalgaard, M., Ostergaard, G., Lam, H.R., Hansen, E.V., Ladefoged,O., 2000. Toxicity study of di(2-ethylhexyl)phthalate (DEHP)in combination with acetone in rats. Pharmacol. Toxicol. 86,92–100.

David, R.M., Moore, M.R., Cifone, M.A., Finney, D.C., Guest,galydi(2-Sci.

D nic5,

G ts ofacol.

G D.N.,BBP,nti-

G ty of

I RCans.n.

J s oned.

esponse is not likely to occur in human liver via techanism. However, the mechanism for Leydig

umor induction by PPAR agonists is not clearly eenced and might be relevant for humans (Klaunig etl., 2003). The same is true for the reproductive toity of DEHP observed in rats (Kavlock et al., 2002).s a result, the focus is shifting from cancer andxisome proliferation to reproductive and other heffects.

In conclusion, this study demonstrates for theime that DEHP is a carcinogen not only to the livut also for the testes of Sprague–Dawley rats. Instingly, in the dose range investigated, the liveror incidence was increased only at the highest dhereas testes tumors occurred at increased peges (significant in a one-tailed test for trend) eve

he lower DEHP concentrations. This finding is prbly of relevance for human cancer risk assessm

D., 1999. Chronic peroxisome proliferation and hepatomeassociated with the hepatocellular tumorigenesis ofethylhexyl)phthalate and the effects of recovery. Toxicol.50, 195–205.

avid, R.M., Moore, M.R., Finney, D.C., Guest, D., 2000. Chrotoxicity of di(2-ethylhexyl)phthalate in rats. Toxicol. Sci. 5433–443.

anning, A.E., Olsson, M.J., Brunk, U., Dallner, G., 1991. Effecprolonged treatment with phthalate ester on rat liver. PharmToxicol. 68, 392–401.

ray Jr., L.E., Ostby, J., Furr, J., Price, M., Veeramachaneni,Parks, L., 2000. Perinatal exposure to the phthalates DEHP,and DINP, but not DEP, DMP, or DOTP, alters sexual differeation of the male rat. Toxicol. Sci. 58, 350–365.

ray, T.J., Gangolli, S.D., 1986. Aspects of the testicular toxiciphthalate esters. Environ. Health Perspect. 65, 229–235.

nternational Agency for Research on Cancer (IARC), 2000. IAMonographs on the Evaluation of Carcinogenic Risks to HumSome Industrial Chemicals. World Health Organization, Lyo

ones, T.C., Popp, J.A., Mohr, U. (Eds.), 1997. MonographPathology of Laboratory Animals: Digestive System, 2ndSpringer-Verlag, Heidelberg.


Kalbfleisch, L.D., Prentice, R.L., 1980. The Statistical Analysis ofFailure Time Data. Wiley, New York.

Kavlock, R., Boekelheide, K., Chapin, R., Cunningham, M., Faust-man, E., Foster, P., Golub, M., Henderson, R., Hinberg, I., Lit-tle, R., Seed, J., Shea, K., Tabacova, S., Tyl, R., Williams, P.,Zacharewski, T., 2002. NTP Center for the Evaluation of Risksto Human Reproduction: phthalates expert panel report on thereproductive and developmental toxicity of di(2-ethylhexyl) ph-thalate. Reprod. Toxicol. 16, 529–653.

Klaunig, J.E., Babich, M.A., Baetcke, K.P., Cook, J.C., Corton, J.C.,David, R.M., DeLuca, J.G., Lai, D.Y., McKee, R.H., Peters, J.M.,Roberts, R.A., Fenner-Crisp, P.A., 2003. PPARalpha agonist-induced rodent tumors: modes of action and human relevance.Crit Rev. Toxicol. 33, 655–780.

Kluwe, W.M., Haseman, J.K., Douglas, J.F., Huff, J.E., 1982. Thecarcinogenicity of dietary di(2-ethylhexyl) phthalate (DEHP) inFischer 344 rats and B6C3F1 mice. J. Toxicol. Environ. Health10, 797–815.

Koch, H.M., Drexler, H., Angerer, J., 2003. An estimation of the dailyintake of di(2-ethylhexyl)phthalate (DEHP) and other phtha-lates in the general population. Int. J. Hyg. Environ. Health 206,77–83.

Mersiowsky, I., Weller, M., Ejlertsson, J., 2001. Fate of plasti-cised PVC products under landfill conditions: a laboratory-scale landfill simulation reactor study. Water Res. 35, 3063–3070.

Park, J.D., Habeebu, S.M., Klaassen, C.D., 2002. Testicular toxic-ity of di-(2-ethylhexyl)phthalate in young Sprague–Dawley rats.Toxicology 171, 105–115.

Peto, R., Pike, M.C., Day, N.E., Gray, R.G., Lee, P.N., Paris, S., Peto,J., Richards, S., Wahrendorf, J., 1980. Guidelines for simple,sensitive significance tests for carcinogenic effects in long termanimal experiments. In: IARC Monographs on the Evaluationof Carcinogenic Risk of Chemicals to Humans. Suppl. 2. Longterm and Short term Screening Assay for Carcinogens, A CriticalAppraisal. IARC, Lyon.

Poon, R., Lecavalier, P., Mueller, R., Valli, V.E., Procter, B.G., Chu,I., 1997. Subchronic oral toxicity of di-n-octyl phthalate anddi(2-ethylhexyl) phthalate in the rat. Food Chem. Toxicol. 35,225–239.

Prentice, D.E., Meikle, A.W., 1995. A review of drug-induced Leydigcell hyperplasia and neoplasia in the rat and some comparisonswith man. Hum. Exp. Toxicol. 14, 562–572.

Shaffer, C.B., Carpenter, C.P., Smith, H.F., 1945. Acute and sub-acute toxicity of di-(2-ethylhexyl)phthalate with note upon itsmetabolism. J. Ind. Hyg. Toxicol. 27, 130–135.

Tickner, J.A., Schettler, T., Guidotti, T., McCally, M., Rossi, M.,2001. Health risks posed by use of di-2-ethylhexyl phthalate(DEHP) in PVC medical devices: a critical review. Am. J. Ind.Med. 39, 100–111.

Turusov, V.S., Mohr, U. (Eds.), 1990. Pathology of Tumors in Labo-ratory Animals, vol. 1: Tumors of the Rat. IARC Scientific Pub-lishers, Lyon.

US Environmental protection Agency (US-EPA), 2000. EPA’s Of-fice of Research and Development, National Center for En-vironmental Assessment Integrated Risk Information System(IRIS). http://www.epa.gov/iriswebp/iris/subst/0014.htm(ac-cessed March 2002).

http://www.epa.gov/iriswebp/iris/subst/0014.htm

Lifelong exposure to di-(2-ethylhexyl)-phthalate induces tumors in liver and testes of...

Documents

Transcript of Lifelong exposure to di-(2-ethylhexyl)-phthalate induces tumors in liver and testes of...