RASS SOT Webinar Are Nonmonotonic Dose Response Curves (NMDRCs) Common after or Androgen Signaling Pathway Disruption:

Fact or Falderal?

L. Earl Gray Jr., PhD

This presentation does not necessarily reflect USEPA policy, but rather represents the author’s current view on the state of the science

USEPAUSEPA scientistscientist grapplesgrapples withwith difficultdifficult environmentalenvironmental issuesissues Estrogen and Androgen signaling pathway toxicant literature review

Questions Shape of the dose response curve over a broad range of doses

Sensitivity of Endpoints to low dose effects

Impact on EDC screening and testing for E and A disruption

EDC Chemicals reviewed for the Shape of the Dose Response In the “Low” Dose Range. Threshold, LNT, or NMDR? Androgen signaling pathway Estrogen signaling pathway   AR antagonists  Flutamide  Ethinyl , Estradiol,  Vinclozolin  Procymidone  hormone synthesis Inhibitors  and , DBP and DEHP  Octylphenol,  Finasteride – inhibition of DHT synthesis   Unknown EDC mechanism, if any  Kepone ()  Semicarbazide – an EDC with a  NMDRC?  that disrupt the Androgen  Selective signaling pathway via multiple agonists mechanisms of toxicity  , ,  , ,  Linuron FC1271a,  Androgen agonists  Aromatase inhibitors  Trenbolone  –  Selective Androgen Receptor  Exemestane,, Anastrazole, Agonists Fadrazole, Letrozole

Studies included in the review

 Measured multiple endpoints affected related to disruption of the estrogen or androgen signaling pathways  Primarily, Reproductive, one or multigenerational studies (if any)  Primarily, Oral administration – diet or gavage  Included a broad range of dosage levels from “low” to “high”  Definitions of “Low Dose” used in the review  ng/kg for chemicals like EE2 and E2, µg/kg for pesticides and toxic substances, or  A dose below the reported NOEL  Preferred – 6 or more dosage levels, but no less than 4 dose levels (three treated groups and a control group)  Primarily rodent studies also includes some porcine, primate and human studies  Review includes about in vivo 200 studies with  >70 of which had 6 or more dose levels  >40 for the Androgen signaling pathway  >30 for the Estrogen signaling pathway 17- concentration in the diet causes sex reversal in Nile talapia. Phelps and Okoko (2011)

“ROBUST” NMRDC 100 80 60

Effect of Tamoxifen in adult female 40 rats in the OECD 407 assay 40 Not so “ROBUST”

20 30 NMRDC Percent Male 0 20 10 1 weight cervix 10 100 1000

Relative Uterius plus Uterius Relative 0 0 5 30 200 MT Dosemg/kgg/kg/day diet Question

Do all the effects of EDCs display a threshold?

 No, it does not appear so.

OECD Hershberger assay validation studies with and Flutamide given for ten days sc to castrate immature male rats: No apparent threshold

Testosterone Propionate stimulates Flutamide reduces androgen dependent Androgen-dependent organ weights in castrate immature, tissue growth in castrate male rats androgen-treated male rats

VP 100 LABC SV 100 VP 80 LABC 75 SV 60 GLANS GP COWS 50

40 Maximum

20 25 PERCENTOF

0 Percentcontrol of 0 0.0 0.2 0.4 0.6 0.8 1.0 0 1 2 3 4 5 6 7 8 9 10 TP mg/kg/d sc mg/kg/d •Hormone dependent endpoints

•Anogenital distance at birth •Nipple/ areolar numbers in infants •Reproductive Malformations •Undescended testes 100 Hypospadias •Gubernacular abnormalities Ectopic Testis •Epididymal agenesis 75 AGD •Ventral prostate agenesis Nipples 50 Areolae •Seminal vesicle agenesis Ventral Prostate •Vas deferens agenesis 25 •Nipples •Hypospadias PercentAffected 0 1 10 100 •Vaginal pouch Vinclozolin mg/kg/d GD 14 to PND 3 •Reproductive Organ Weights ED50 HillSlope •Glans penis Areolae 11.91 1.208 •Ventral prostate Ventral Prostate 36.39 1.125 •Seminal vesicle Nipples 36.72 7.451 •Testes AGD 44.46 1.29 •Epididymides Hypospadias 50.28 36.39 •Levator ani bulbocavernosus Ectopic Testis 197.4 2.28 •Cowper’s glands •Testis and epididymal histopathology

Question Do EDCs induce non-monotonic effects in vitro? Yes “Low dose” hypothesis: EDCs produce nonmonotonic responses in vitro and in vivo. Toxicology testing studies are conducted at high dosage levels at the right side of the curves whereas the relevant “low dose” studies are conducted on the left side of the curve and can see the opposite effects from the high dose studies.

Toxicologists ??? Data examined to date on E and A in vitro gene expression assays clearly X show that current toxicology studies on Endocrinologists ??? EDCS are here EE2 T47D KBLUC High concs irrelevant in vivo

100 Serum EE2 level in 80 rats treated w ith a very high oral dose of 60 EE2 - 1mg/kg

40 Conc in w hole lake causing

20 population crash Percent of 0.1nMPercent of E2 0 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 10 10 10 10 10 10 10 10 10 10 10 10 10 Concentration ppt Trenbolone High concs NMDRC irrelevant in vivo 15 Cytotoxicity

10

5

Fold Induction 0 1 10 0.1 100 0.01 1000 0.001 10000 100000 1000000 Concentration (ppb)

TB conc in tissue associated with in the fathead minnow TB conc in amniotic fluid in female rat fetus resulting in reproductive tract malformations What about NMDRCs in vivo?

Do EDCs induce robust, reproducible non-monotonic effects, if so

• At low dosage levels?

• Is the effect clearly adverse or causally linked to an adverse effect?

• Would it alter risk assessment?

My classification of in vivo NMDRCs and an estimate of their prevalance  “Robust”, reproducible, biologically plausible  “Trivial”, frequency of one effect declines as a more severe one develops  “False positive”, multiple comparisons or invalid analyses  “Confounded”, NMDRC is reported at dosage levels well below background levels  “Imaginary”, no group differs significantly from control but the data are interpreted as displaying a NMDRC  “Oblivious”, interpretation of high dose NMDRC ignores overt toxicity or adverse effects at much lower doses Dose-response database with Estrogens: number of studies reviewed (About 70 E and A studies with six or more dose groups (studies with 6 or more dose groups/total studies examined)

 Estradiol (2 / 9)  Ethinyl estradiol ( 7 / 10)  Genistein(2 /12)  BPA (4 / 14, and more on the way)  DES (1, 7)  Zearalenone (2, 8)  Octylphenol (1, 2)  Nonylphenol (2, 3)  Methoxychlor (3, 7)  Kepone (3, 5)  SERMs (2, 16)  Aromatase Inhibitors (3, 9)

vom saal PNAS 1997 vs Ashby 1999 Does an oral low dose of DES increase F1 male mouse prostate weight? 60 VOM SAAL 1997 Ashby 1999 40 Cagen 1999

20

CF-1 Male Mice CF-1 Prostate weight Prostate 0 0 2 0 0 20 200 200 200 2000 20000 200000 DES ng/kg Vom Saal PNAS 1997. Oral DES administration GD11-17 was reported to induce a NMDRC on F1 male mouse prostate weight. However, the dose of DES required to reduce prostate weight is clearly not a "low dose" effect Neonatal Doses 60 reducing fecundity Dose in mg/kg in tablets 55 prescribed for women 50 45 Dose that reduces F1 40 male mouse 35 prostate weight Dose range that 30 accelerates Dose that reduces

25 puberty in early embryo CF-1 Male Mice CF-1 Prostate weight Prostate female rats viability GD4-8 20 1 10 100 1000 10000 100000

DES ng/kg Dose-response database for Antiandrogens and Androgens (studies with 6 or more dose groups / total studies)

 Flutamide (2 / 9)  Vinclozolin (4 /8)  Procymidone (4 / 6)  DEHP (9 / 19)  DBP (3 / 6)  Finasteride (2 / 2)  Semicarbizide (1 / 3)  Prochloraz (3 / 5)  Linuron (0 / 6)  Testosterone (12 (4 rat, 8 men) / 13)  SARMS (1 – so far, more to come)

Low Dose Studies with Androgens and SARMS

 A few effects in rats display non-monotonic responses, but most do not  Testosterone Propionate (sc) to the pregnant rat (GD14-18)  Testosterone propionate (sc) in the adult male rat  None of these studies used the oral route of administration

 Dose response data in men do not show NMDRCs

 Extensive trenbolone data base unavailable to the public and general scientific community, but lots of low dose studies in many species Non-monotonic response of the adult male rat testis to TP implants. Many other effects showed monotonic responses

“ROBUST” NMDRC “ROBUST” NMRDC

F1 female hydrometrocolpos with vaginal agenesis 100

75

50

25

0 PercentAffected

10 100 1000 10000 Testosterone Propionate (sc) ng per pregnant rat GD 14-19 Low Dose Studies with Antiandrogens

 Antiandrogens  Finasteride  Rat studies – no evidence of non- monotonic effects  Flutamide  Rat studies – no evidence of non- monotonic effects  Dibutyl and DEHP  Rat studies – no reproducible evidence of non-monotonic effects for adverse effects at low dosage levels

Low to high-dose, dose response studies with the “antiandrogen” Finasteride administered orally to the dam during gestation at doses ranging from 0.0003 to 300 mg/kg/d. There were no nonmonotonic effects on the male offspring. Threshold effects of in utero flutamide on the incidence of male rat reproductive tract malformations later in life

F1 male rat malformations after in utero flutamide exposure Ten dose levels over 4 orders of magnitude Data combined from two studies(results of dose of 10 mg in one study combined with 12.5 mg in the other)

100 hypospadias epididymal agenesis 75 prostate agenesis Sv agenesis 50 Ectopic testis labc agenesis 25

0 Percentaffected 0.1 1 10 100 Flutamide mg/kg Low to high-dose, dose response studies. Data are from ten one generation studies with different phthalate esters that all disrupt male rat sexual differentiation via the same mode of action that also display similar potencies in short-term in vivo screening assays. Testis/Epididymal malformations and seminal vesicle weights are shown. None of the effects in these studies displayed any non-monotonic responses.

Testis-epididymal malformation data pooled from 10 studies

100 LogEC50 2.586 HillSlope 3.332 80 EC50 385.3 60 40 20

0 PercentAffected 0.01 0.1 1 10 100 1000 mg/kg/d PE DEHP DR Low dose study Andrade et al 2006 No non-monotonic low dose adverse effects

Body Weight TESTES WEIGHT Epididymal weight Seminal Vesicle 500 2.5 800 1000

400 2.0 600 800

300 1.5 600 400 200 1.0 400

100 0.5 200

WEIGHT (g) WEIGHT 200

WEIGHT (mg) WEIGHT

WEIGHT (mg) WEIGHT WEIGHT (mg) WEIGHT 0 0.0 0 0

0.0000.0150.0450.1350.4051.2155.000 0.0000.0150.0450.1350.4051.2155.000 15.00045.000 15.00045.000 0.0000.0150.0450.1350.4051.2155.000 0.0000.0150.0450.1350.4051.2155.000 135.000405.000 135.000405.000 15.00045.000 15.00045.000 135.000405.000 135.000405.000 Dose mg DEHP/kg/d Dose mg DEHP/kg/d Dose mg DEHP/kg/d Dose mg DEHP/kg/d

Percent without Phthalate Syndrome F1 Male Mating Behavior Prostate weight Reproductive Tract Malformations and Fertility Daily testis sperm production 600 150 150 50

40 400 100 100 30

20 200 50 50

(millions) 10

WEIGHT (mg) WEIGHT

PercentNormal PercentNormal

0 0 0 0 DSP unedited data DSPunedited

0.0000.0150.0450.1350.4051.2155.000 0.0000.0150.0450.1350.4051.2155.000 0.0000.0150.0450.1350.4051.2155.000 0.0000.0150.0450.1350.4051.2155.000 15.00045.000 15.00045.000 15.00045.000 15.00045.000 135.000405.000 135.000405.000 135.000405.000 135.000405.000 Dose mg DEHP/kg/d Dose mg DEHP/kg/d Dose mg DEHP/kg/d Hist Control DEHP DR Low dose study Andrade et al 2006 “non-monotonic “low dose effect

Range of phthalates in rodent chows Reported NMDRC for AGD Dietary phthalate levels estimated from another paper “Confounded and Imaginary” NMRDC ?? Range of Background 2.2 Dietary Phthalate Levels 2.0 1.8 1.6 1.4 1.2

1.0 Anogenital Distance (mm) Distance Anogenital 0.0 0.5 1.0 5.0 21.4 214.0 500.0 50000.0 500000.0 DEHP micrograms per kg per day PUP BIRTH WEIGHT NO STATISTICALLY SIGNIFICANT EFFECTS Imaginary NMDRC NOTED BY AUTHOR 7 Data from Grande et al., 2006, cited as examples of NMDRCs 6

induced in female rat offspring 5 from in utero exposure to DEHP Grams

4

0.0000.0150.0450.1350.4051.2155.000 15.00045.000 135.000405.000 Dose DEHP mg/kg/d AGE AT VAGINAL OPENING AGE AT FIRST ESTRUS * INDICATES STATISTICALLY SIGNIFICANT NO STATISTICALLY SIGNIFICANT EFFECTS NOTED BY AUTHOR EFFECTS NOTED BY AUTHOR

40 range of phthalates * in control * * 42 38 diets *

40 36

34 38 Age at "Puberty" at Age 32 "Puberty" at Age 36

0.0000.0150.0450.1350.4051.2155.000 0.0000.0150.0450.1350.4051.2155.000 15.00045.000 15.00045.000 135.000405.000 135.000405.000 Dose DEHP mg/kg/d Dose DEHP mg/kg/d My current conclusions on the shape of the dose response curves for EDCs-  EDCs appear to induce some effects that do not appear to display a threshold (apparent Linear No Threshold responses)  NMDRCs for EDCs  Biologically plausible  Occur frequently in vitro, but these are generally not relevant to in vivo effects and do not occur at low concentrations  It appears that NMRDCs are more common  in studies with short-term exposures and  On “upstream” mechanistic events versus “downstream” adverse phenotypic effects.  A few of the effects of androgens given sc are non-monotonic, but other effects in the same study occur at lower dosage levels and they display “normal” (monotonic) dose responses  A number of multigenerational studies of estrogens and antiandrogens have been reviewed. To date, these did not indicate that robust, reproducible NMDRCs were common events at low dosage levels.  Additional data needs to be examined from robust, multigenerational studies using a broad range of dosage levels for other pathways Impact of NMDRCs and LNT responses on EDC screening and testing Impact for EDC screening – NONE • Do the EDC screening assays fail to detect E or A activity? NO

Impact for multigenerational testing • Estimation of shape of the dose response curve in the low dose region could be enhanced by using more dose groups • For example, keep total N litters in a study as is, increase number of dose groups from 3 to 6 with half as many litters per dose group • Examining more than one animal per litter enhances endpoint sensitivity by increasing the statistical power to detect low dose categorical effects like malformations and histopathological lesions (Blystone et al., 2010; Hotchkiss et al., 2008).

Sensitivity of Endpoints to low dose effects in EDC testing • Recent studies conducted with E and A active chemicals have identified more sensitive endpoints that should be added on a case- by-case basis. • Many of these are not explicitly included in any current protocol