Interpretation of Low-Incidence Findings in Developmental and Reproductive Toxicity Studies

Joseph F. Holson, Ph.D.WIL Research Laboratories, Inc.

Acknowledgements:Bennett J. VarshoJeffrey A. PittLewis E. Kaufman

ADExperience (formerly ADR, Adverse Drug Reports)

Based on assumptions that rare, latent or long-term effects that the premarketing testing regimen does not reveal will eventually show themselves given a wide enough distribution

FDA’s View

A serious Adverse Drug Experience is one that kills someone, has been shown to pose a threat to life, is seen to cause persistent or significant disability or incapacity, requires inpatient hospitalization or is diagnosed as a congenital anomaly, cancer or overdose.

“Serious” can also mean occasions in which a reaction to a product prolongs an existing hospitalization or causes other medical events doctors deem important.

Products deemed likeliest to have unexpected serious medical problem if they fail to produce their expected outcomes: FDA considers likeliest to fall in this category formulations approved within last 3 years

(with emphasis on lack-of-effect reports) new molecular entities products known/suspected of having

bioavailability or bioequivalence problems.

FDA’s View (continued)

Human Epidemiology Studies vs. Experimental Animal Assessments

Less rigorously monitoredGreater phenotypic heterogeneityLess amenable to dose-response

evaluationsRamifications of endpoints better

ascertainedRetrospective studies plagued with recall

bias

Observational Determinants of Anatomical/Functional Deviations

Degree of deviation from average

Incidence (prevalence)*Impact on salubrityCosmetic significance

*In humans convention of <4%; no medical or surgical significance has been used

*In experimental studies, no convention used – statistical

Animal:Human Concordance Studiesfor Prenatal Toxicity

Nisbet & Karch, 1983 Many chemicalsRelied on authors’ conclusionsEmphasis on fertilityNo measures of internal dose

Attributes

Interdisciplinary team Criteria for acceptance of data/conclusionsConcept of multiple developmental toxicology endpoints No measures of internal dose

Authors

Holson et al., 1981 (Tox Forum)Kimmel et al., 1984 (NCTR Report)

Attributes

Interspecies inhalatory doses adjustedRelied on authors’ conclusions23 occupational chemicals and mixtures No measures of internal dose

Provided detailed informationOnly 4 drugsEmphasis on morphologyFocus on NOAELsNo measures of internal dose

Many chemicalsRelied on authors’ conclusionsNo measures of internal dose

Authors

Hemminki & Vineis, 1985

Newman et al., 1993

Schardein, 1995

Animal:Human Concordance Studiesfor Prenatal Toxicity

Awareness of Developmental Toxicity of Selected Agents

Agent Year First Reported Species*

Alcohol(ism)

Aminopterin

Cigarette Smoking

Diethylstilbestrol

Heroin/Morphine

Ionizing Radiation

Methylmercury

Polychlorinated Biphenyls

Steroidal Hormones

Thalidomide

1957

1950

1941

1940

1969

1950

1953

1969

1943

1961

(gp), ch, hu, mo, rat

(mo & rat), ch, hu

(rab), hu, rat

(rat), hu, mi, mo

(rat), ha, hu, rab

(mo), ha, hu, rat, rab

(rat), ca, hu, mo

(hu), rat

(monk), ha, hu, mo, rat, rab

(hu), mo, monk, rab

*ca - cat, ch - chicken, ha - hamster, gp - guinea pig, hu - human, mi - mink, mo - mouse, monk - monkey, rat - rat, rab - rabbit

Effect-Levels for Teratogensin Humans and Test Species

Aminopterin Death/Malformations

Death/Malformations

Agent ResponseHuman

Rat

Species Dose0.1 mg/kg/da

0.1 mg/kg

Diethylstilbestrol Genital Tract Abnormalities/Death

Genital Tract Abnormalities/Death

Human

Mouse

0.8-1.0 mg/kg

1 mg/kg

Ionizing Radiation Malformations

Malformations

Human

Rat/Mouse

20 rads/da

10-20 rads/da

Cigarette Smoking Growth Retardation

Growth Retardation

Human

Rats

>20 cigarettes/da

>20 cigarettes/da

Thalidomide Malformations

Malformations

Malformations

Human

Monkey

Rabbit

0.8-1.7 mg/kg

5.0-45 mg/kg

150 mg/kg

Possible Inter-relationships of Developmental Toxicity Endpoints

Toxic Stimulus

Malformations

Functional Impairments

Growth Retardation

Death

Toxic Stimulus GrowthRetardation Death

Malformation

Functional Impairment

Maldevelopment vs. Tumorigenesis

Not more frequent with timeMyriad possible underlying mechanisms

Amniotic banding, oligohydramnios-skull dysgenesis Interference with signaling pathways (TGF-) Mutation

Maternal influences possibleMultiple endpoints interrelated

Weight alterations causing cleft palate and NTDsOccur early in life & hence greater

economic/social impact

A Ventral view illustrates:

the stomodeum (primitive oral

cavity), the heart in the

pericardial cavity,

the anterior intestinal portal leading to the

foregut,and the

posterior intestinal portal leading to the

hindgut.Kathleen Sulik (http://www.med.unc.edu/embryo_images/unit-welcome/welcome_htms/akgs.htm)

Selected Differences in Developmental vs. Oncogenic Endpoint Ascertainment

Smaller group sizes (25 vs. minimum 100/group) Macroscopic – histopathology very rare Physical constraints/difficulty Less standardized nomenclature No certification, controls over training, etc.

More impact because earlier in development due to women in clinical trials

Involves coapt organisms (dam & fetuses) Always potential for maternal influence, but goes both ways

Dynamic morphology & function ACE example

Animals evaluated in the midst of changing morphology No two points in development are the same

Exposure hourly and daily key to outcome An important aspect of human studies

A B C D E F

Premating to Conception

Conception to Implantation

Implantation to Closure of Hard Palate

Hard-Palate Closure to End of Pregnancy

Birth to Weaning Weaning to Sexual Maturity

Parturition Litter Size Landmarks of Sexual DevelopmentGestation Length Pup Viability Neurobehavioral Assessment F1 Mating and Fertility Pup Weight Acoustic Startle Response

Organ Weights Motor Activity Learning & Memory

ParturitionGestation Length Pup Viability Litter SizeLandmarks of Sexual Development Pup WeightNeurobehavioral Assessment Organ Weights Acoustic Startle Response F1 Mating and Fertility Motor Activity Hormonal Analyses Learning & Memory Ovarian QuantificationHistopathology Premature Senescence

Postimplantation Loss

Postimplantation LossViable FetusesMalformations & VariationsFetal Weight

Postimplantation LossViable FetusesMalformationsVariationsFetal Weight

Estrous Cyclicity Mating Corpora Lutea Fertility Implantation SitesPre-Implantation Loss Spermatogenesis

Estrous CyclicityMatingFertilityCorpora LuteaImplantation SitesPre-Implantation LossSpermatogenesis

Denotes Dosing Period

Single- and Multigenerational

Satellite Phase

OECD 415, OECD 416, OPPTS 870.3800, FDA Redbook I, NTP RACB

F1

F2 ????????????????

????????????????

Pre- and Postnatal Development

F1

ICH 4.1.2F0

????????????????

Prenatal DevelopmentICH 4.1.3 OECD 414

OPPTS 870.3600 870.3700

Fertility StudyICH 4.1.12W4W

CMAX

AUC

CMAX

AUC

10W

Standard Study Designs

Comparison of Study Scale/Size

0

500

1000

1500

2000

2500

3000

Acute OralToxicity

SubchronicToxicity

2-Year CancerBioassay

DevelopmentalToxicity

2-GenerationReproduction

No

. o

f A

nim

als/

Stu

dy

MARTA/MTA Database Fetal Abnormalities

Examination Type Individual Descriptors

External 123

Visceral 277

Skeletal 467

Combined 867

www.hcd.org

Freehand Section

Whole-Body Microdissection

½ Skeletal ½ Visceral 100%

Guideline minimum = 25%, 175 + 175 vs. 1400

4 Groups

(of 25 Dams)X

350 Fetuses =1400 Fetuses

½ & ½ Control and High Group (per guideline)

Size Comparison at Near Term

CR Length 75 mmCR Length 35 mm

3.6 grams 47 grams

CR Length 19 mm

1.3 grams

RabbitRatMouse

Taylor, 1986

FDA Definition

Rare Event – “an endpoint that occurs in less than 1 percent of the control animals in a study and in historical control animals”

Reviewer Guidance(Draft)

Integration of Study Results to AssessConcerns About Human Reproductive

And Developmental Toxicities

CDER, 10/2001Pharmacology/Toxicity

Rare Events (Low-Incidence Findings): Typical Reaction to, and Subsequent Scenario

Disbelief, rely on statistical insignificanceComparison to concurrent controlComparison to historical control (HC)Comparison to other HC databasesAsk experience/opinions of othersConstruct explanation to negateAgency rejectsRe-do study or label appropriately

Classes of Reproductive and Developmental Toxicity

Reproductive Developmental

Fertility Mortality

Parturition Dysmorphogenesis

Lactation Alterations to Growth

Functional Toxicities

Selected Reproductive Endpoints Exhibiting Strong Signals from Rare Events/Low Incidence

EndpointExamples from WIL Research Historical Control in Crl:CD(SD)IGS BR

Mean Viable Litter Size

13.9 1.02 decrease of 1

Mortality PND 4Mean = 96.2% Min/Max 91-95%

91%

Total Litter LossMean = 0.94% (10/1061)

1 is equivocal 2 is more significant signal

Newborn Pup Weights

Mean = 7.0g 0.23

range 6.5-7.4g n = 1100 litters

6.5g strong signal

Case Study: Dystocia, Extended Parturition and/or Pregnancy

2-generation with second mating phase of F1, vapor inhalation, used industrially, OTC pharmaceutically

PPM 0 70 300 500 700

F0 0 0 0 2/24 3/26

F1-1st 0 0 0 0 1/17

F1-2nd 0 0 1/21 1/18 0/12

HC then: 2/333 = 0.60% HC now: 4/1100 = 0.36%

Organogenesis (classically defined) is unaffected

Effects are severe

Risk is low

Caused by ACEinh that cross placenta

ACEinhFetal

Hypotension

RenalCompromise

(Anuria)Oligohydramnios

Calvarial Hypoplasia

Neonatal Anuria

IUGR

Death

Case Study: Functional Alteration Example ACE Inhibition-Induced Fetopathy (Human)

Case Study: Functional Alteration Example ACE Inhibition in Developing Rats

RAS (renin-angiotensin system) matures around GD17

No ‘apparent’ effect in initial reproductive studies Nonstatistically significant increase in postnatal mortality (~8%)

Subsequent postnatal studies with direct administration to pups Growth retardation

Renal alterations (anatomic and functional)

Mortality increased to more than 30%

Comparison of Overall Spontaneous Malformation Rates in Different Species

Species Mean % Range (%) N

Rat* 0.33 0-1.6 9643

Mouse 1.2 0-3 5207

Rabbit 3.2 0-10 4708

Dog 5.5 5.3-5.7 167

Human 4.0 3-9 Multiple Surveys

* Actual number in laboratory population ~44,000

Case Study: Malformation Example Topical Antibiotic for Oral Mucosa

Historical Control Data

Malformation TotalMean% PL

Min Max

Retroesophageal Aortic Arch

2/9643 0.02% 0.0% PL 0.3%PL

Rat Study Data

Malformation 1 2 3 4

Retroesophageal Aortic Arch

0 01

(0.3%PL)1

(0.3%PL)

Rare Events: Control vs. Treated Groups

3

1

3:1 Probability that spontaneous event will occur in treated group

Malformations which Have Occurred as Rare Events in Numerous Scenarios

MalformationIncidence (%PL)

Rat Rabbit

Ventricular Septal Defect 0 0.02

Cleft Lip/Palate 0.02 0.04

Abdominal Wall Defect Including Gastroschisis

0.04 0.06

Hydrocephaly 0.03 0.20

Spina Bifida 0 0.17

Renal Agenesis 0.01 0.02

Diaphragmatic Hernia 0 (2/39442) 0.04

Malformations can involve any tissue or structure and may constitute a rare event issue

Rare Event Manifestation Matrix

Control Low Mid High

0 0 0 1

0 1 0 0

0 0 1 0

0 0 1 1

1 0 0 1

Paradigm to Evaluate Rare Findings

Comparison to concurrent control Evaluate dose-responsiveness including TK, AUC/Cmax Compare to HC range and mean, consider other statistical tests,

including Monte-Carlo Analysis Evaluate signals of developmental toxicity among dose groups Compare to second species Compare to findings in the combined pre-/postnatal study Perform confirmatory study:

Increasing N Increasing number of concurrent controls Increasing dose (based on TK: AUC/Cmax) Consider unbalanced study design Delimited exposure regime Evaluate pharmacologic action relative to ontogeny of receptors, etc. and

reconcile with modified dosing regime Label and follow-up in birth defects registry

The Bottom Line

With rare events, the best practice for resolution of the relationship to treatment is through a large historical control database developed at the same laboratory, using consistent methodology and conditions in conjunction with appropriately designed confirmatory study.

Human risk assessment and management may require study in human registries.