Genomics for embryos selection

USA: Europe: Asia: South America:

Livingston, NJ Barcelona, Spain Kobe, Japan Lima, Peru

Los Angeles, CA Oxford, UK

Hamburg, Germany

Santiago Munné, Ph.D.

High Rates of Chromosome Abnormalities

embryos analyzed: 6054. Morphologically normal embryos: 3751. Source: Munné et al. 2007.

Similar results also found by Munne et al 1995, Marquez et al. 2000, Magli et al. 2007.

% chromosomally

abnormal embryos56%

Maternal age

Morphology:

The Majority of Embryos with „Good‟

Morphology are Chromosomally Abnormal

<35 35-37 38-40 41-42

- Ongoing implantation without PGD * 30% 20% 12% 6%

- Abnormal embryos (array CGH) ** 54% 67% 75% 82%

- Non-chromosomal implantation loss 16% 13% 13% 12%

*estimated from SART (2008), ** Reprogenetics

IVF Results (SART 2008) Compared

to Chromosome Abnormalities

Conflicting PGD Results Using Day 3 Biopsy and FISH

Positive effectGianaroli et al. 1999Munne et al 1999Gianaroli et al 2001aGianaroli et al. 2001bMunne et al. 2003Gianaroli et al. 2004Munne et al. 2005Munne et al 2006Verlinsky et al. 2005Colls et al. 2007Garrisi et al. 2009Rubio et al. 2009

No effectWerlin et al. 2003Staessen et al. 2004Platteau et al. 2005Jansen et al. 2008Mersereau et al. 2008Schoolcraft et al. 2009

Negative effect Mastenbroek et al. 2007Hardarson et al. 2008

Optimal PGD Questionable PGD

Biopsy media with aminoacids simple media

Biopsy time / embryo 1 min > 5 min

# cells biopsied one two

Fixation method Carnoy’s Tween 20

# chromosomes tested ≥8 6

# analysts / case 2 1

Reanalysis no results yes no

Large experience yes no

Error rate <10% 10-50%

Number of zygotes >5 5

Reviewed in Munné et al. (2010) Fertil. Steril. 94:408

Optimal PGD Methods for FISH

Analysis of remaining cells of embryos previously analyzed by PGD:

study technique error rate

Baart et al 2004 FISH 50.0%Li et al. 2005 FISH 40.0%Gleicher et al. 2009 FISH 15-20%Munne et al. 2002 FISH-9 7.2% Colls et al., 2007 FISH-9 4.7%Magli et al. 2007 FISH-9 3.7%Munne et al. 2010 array CGH 1.8%

Technical Errors Should be <10%

• 24 chromosomes analysis: array CGH, SNP arrays, qPCR

• Blastocyst biopsy

• vitrification

New developments

24-chromosme analysis:

• array CGH

• SNP arrays

• q PCR

aCGH vs. SNP arrays: Resolution

• Abnormalities smaller than 5MB are not well understood and difficult to counsel

• At 150 Kb resolution 50% of people have CNVs: it is difficult to differentiate a new polymorphism from an abnormality

More resolution, more counseling nightmares:

Array CGH*

Array CGH**

* Bac array, ** oligo array

Array CGH

Test DNA

Normal DNA

4000 probesMore resolution than karyotype

46,XY

46,XY

46,XX-10 +16

Array CGH Advantages

• All 24 chromosome type of aneuploidies detected

• Results in 24 hours; allows for PB or day 3 biopsy

• Parental DNA not required: ad hoc decisions possible for day 3 or 5 biopsies.

• ICSI not required.

• Used extensively in postnatal testing and soon in prenatal (NIH large study).

PGD Reanalysis

Diagnosis: Trisomy 18 Mosaic 100% abnormal

Confirmed?:

Exact abnormality: No

Abnormal embryo: Yes

Definition of error

Array CGH Validation Using Cell Lines

Analysis of single cells from euploid and aneuploid cell lines

Array CGH* SNP arrays**False negatives: 0% 2.1%False positives: 0% 3.9%

(This method does not account for mosaicism errors.)

*Mamas et al (submitted), **Johnson et al. (2010)

aCGH validation: reanalysis by FISH

- Reanalysis of the remainder cells of the embryo by FISH.- This detects mosaicism and intrinsic technical errors.

Tris 16

Mono 22

Day-3 Day-5PGD results not confirmed: 3% (3/104) 0% (0/18)

Gutierrez et al. Fertil Steril (2011) and Reprogenetics unpublished data

aCGH validation: Double cell aCGH analysis

- When IVF centers doubt about the presence of a nucleus they might biopsy two cells.

- This provide us with two cells analyzed / embryo:

Day-3Abnormal not confirmed: 2.5% (1/49)Normal not confirmed: 0.0% (0/32)Total error rate: 1.2%

Reprogenetics data, unpublished

aCGH validation:

biopsy stage

Biopsy Embryos Not confirmedstage reanalyzed (error rate)

ESHRE PGD consortium* PBs 156 11%

Reprogenetics ** day 3 222 2%

Reprogenetics ** day 5 18 0%

*Geraedts et al. (2011) Human Reprod, in press; ** Gutierrez-Mateo et al (2011) Fertil Steril., and unpublished data

aCGH vs. SNP arrays: Genome Coverage

# of probe genomeprobes size covered

aCGH 4,000 x 150,000 kb = 600.0 Mb (25%)

SNPs 300,000 x 50 kb = 1.5 Mb (>0.1%)

aCGH vs. SNPs: Parental “Support”

SNP arrays at the single cell level produce noisy and incomplete

results and need parental DNA (“parental support”) in order to

interpret the data.

aCGH does not require to test the parents before PGD

Disadvantages: 1) It prevents deciding on day 3 if to do PGD or not based

on the number of embryos.2) May result in unnecessary cancellation fees.3) Misleading: Parental Support simply supports the

software analysis of poor SNP calls, not the patients.

aCGH vs. SNPs: Origin of Aneuploidy

Meiotic abnormalities:Although aCGH cannot detect the meiotic origin of the aneuploidies, it is

well known that >90% of them are maternal in origin.

According to a GSN: - only < 4% of meiotic aneuploidies were paternal in origin.- 13% of all abnormalities were paternal. Thus 13-4 = 9% were mitotic.

Our opinion:- Mitotic abnormalities = mosaicism. Mosaicism involucrates

chromosomes at random and its parental origin is irrelevant.- If a patient has a high rate of paternal mitotic aneuploidy (mosaicism)

diagnosing a “male factor” is incorrect. Recommending sperm donation is not warranted.

aCGH vs. SNPs: Validation Pitfalls

Reporting Precision per Chromosomechromosome copy number probability

1 2 0.99

2 2 0.99

3 2 0.99

4 2 0.99

5 2 0.99

6 2 0.96

7 2 0.99

8 3 0.98

9 2 0.99

10 2 0.99

11 2 0.99

12 2 0.99

13 2 0.99

14 2 0.99

15 2 0.99

16 2 0.99

17 2 0.95

18 2 0.99

19 2 0.90

20 2 0.99

21 2 0.94

22 2 0.80

sex Chr XY 0.95

P overall = P1 * P2 * P3 …

To calculate the probability of each and every copy number being correct, you multiply the probabilities of each result being correct together

Probability of Assigning the correct Genetic diagnosis(Assume that those which are 99% are halfway between 99 and 100% = 99.5%)

Probability = 54%

Based on actual clinical case done via Qualitative SNP microarrayAdapted from R.T. Scott

Polyploidy cannot be detected by aCGH but the majority of them

have also other abnormalities:

• Abnormal embryos fully reanalyzed: 1,416

• polyploid or haploid by PGD and

polyploid or haploid after reanalysis: 140

• Of the above, that had no other aneuploidies: 27 (1.8%)

Missed total abnormalities: 1.8%

aCGH vs SNPs: polyploidy

aCGH vs SNPs: mitotic abnormalities

Qualitative SNP arrays cannot detect mitotic trisomies and

meiotic duplications without recombination:

• Abnormal embryos fully reanalyzed: 1,416

• Of the above, that were 100% mosaic with

only trisomies (mitotic trisomies): 53 (3%)

Not diagnosed by qualitative SNP array: 3%

aCGH vs. SNPs: Fingerprinting

Fingerprinting can be used to determine which embryo implanted

when more embryos are replaced than those implanting. It

requires testing the fetus or baby and comparing it with the

embryos.

SNP arrays obtains the information from the embryos during the

PGD test.

aCGH produces enough DNA so left over DNA is always stored and

available for fingerprinting when necessary.

Clinical ResultsArray CGH with Day 3 Biopsy

Day 3: Prognosis depending on age and ovarian response

egg

donors

<35

years old

35-39

years old

40-42

years old

>42

years old

100% 80% 46% 41% 20%50% 43% 18% 17% 6%

100% 89% 83% 60% 20%40% 38% 26% 15% 5%

100% 97% 91% 72% 60%65% 35% 28% 15% 8%

100% 100% 96% 93% 77%47% 47% 32% 16% 8%

(% normal embryos)

> 10

5-7

8-10

# day 3

embryos

1-4

Reprogenetics data to 2/2011: 609 cycles, 5101 embryos

Day 3 Biopsy, Array CGH with Day 5 Transfer

Cycles performed: 509Maternal age (av.) 37.1

Pregnancy Rate Ongoing Pregnancy RatePer Cycle Per ET Per Cycle Per ET

Control 38% 38% 31% 31%PGD 38% 54% 35% 52%

NS < 0.001 NS < 0.001

Data from all referring centers. Munné et al. (2010) ASRM, and unpublished data

Array CGH for Recurrent Pregnancy Loss

Indication: ≥3 previous miscarriages, <42 years oldProcedure: Biopsy on day 3, PGD by array CGHPopulation: average 36 years old, Av. 3.2 previous miscarriages

Expected:Miscarriage rate (Brigham et al. 1999) for this population: 30%

Observed: <35 years 35-41 yearsCycles 20 49Pregnancy rate / retrieval 45% 41%Pregnancy rate / transfer 47% 52%Miscarriage rate 0% 5%

Arrays and Blastocyst Biopsy

Array CGH on Blastocyst Biopsies: Why?

Advantages:

1) More DNA: More robust diagnosis

2) Blastocysts have less mosaicism

3) Lowest error rate (≈0%)

4) Reduced impact of embryo biopsy

5) Less embryos to process

6) Facilitates single embryo transfer

7) If frozen cycle: Uterine environment

may be optimized after thaw

Cycles Mat. Prev. embryos implant.age failed replaced (+ sac)

cycles

CGH : 45 37.7 2.4 2.0 72%

control : 113 37.1 1.2 2.7 46%

p=0.0003

Schoolcraft et al. (2010) Fertil. Steril. 94:1700

CGH, Blastocyst Biopsy and VitrificationClinical Results

Day 5: Prognosis depending on age and ovarian response

egg

donors

<35

years old

35-39

years old

40-42

years old

>42

years old

100% 63% 68% 40%50% n/a 50% 36% 20%

100% 100% 87% 56% 65%70% 72% 44% 38% 23%

100% 100% 100% 100%73% 57% 50% 36% n/a

100%58% n/a n/a n/a n/a

(% normal embryos)

> 10

4-6

7-10

# day 5

embryos

1-3

Reprogenetics data to 2/2011: 109 cycles, 595 embryos

Array CGH with Biopsy on Day 3 or Day 5

Results from All Referring Centers

day 3 day 5biopsy biopsy

Maternal age: 37.1 37.0# zygotes 8.6 4.6% normal embryos: 35% 48%cycles: 509 70

Implantation after PGD 41% 60%Pregnancies / pick up 40% 46%Pregnancies / transfer 54% 65%

Biopsy orThaw transferEffect

�

Scott et al., 2010 ASRM, O-5 (RCT data)Treff et al., 2009 ASRM, O-169 (qPCR methods)Treff et al., 2010 Fertil Steril. 94(6):2017-21 (SNP array methods)

Day 5 biopsy and

SNP array and qPCR PGD

Courtesy of Dr. Treff

Day 5 Blastocysts Biopsy and CGH Abrogate the Maternal Age Effect

Day 5 data: CGH and aCGH data, Day 3 data: from aCGH, *cycles with >4 embryos biopsied

0

10

20

30

40

50

60

70

80

90

30-34 35-38 39-42 43-45

implantation rate

per replaced

embryo

aneuploidy rate

cycles with all

embryos abnormal

Day 5 biopsy

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

<35 35-39 40-42 43-45

implantation rate

aneuploidy rate

cycles with all abnormal

Day 3 biopsy*

Arrays and vitrification

Effect of embryo vitrification:

Non-PGD results

Array CGH with Biopsy on Day 3 or Day 5

Results from All Referring Centers

day 3 day 5biopsy biopsy

Maternal age: 37.1 37.0# zygotes 8.6 4.6% normal embryos: 35% 48%cycles: 509 70

Implantation after PGD 41% 60%Pregnancies / pick up 40% 46%Pregnancies / transfer 54% 65%

Biopsy orThaw transferEffect

PGD Av. day day pregnancies

Test age biopsy replacement / transfer

FISH-9 39.6 3 day 5 29%

FISH-9 39.0 3 thaw cycle 47%

FISH-12 39.3 3 thaw cycle 46%

FISH-12 39.5 5 thaw cycle 40%

p<0.05

Effect of embryo vitrification:

FISH data

Reprogenetics-Japan, unpublished data

Effect of embryo vitrification:CGH + SNP array data

- PGD by CGH or SNP array- PGD and control both frozen cycles:

Control PGD *Implantation 35% 62.6% p<0.0001Pregnant 57% 81.4% p<0.001

Schlenker et al. (2010) Fertil Steril

Conclusions

A majority of embryos at any age have numerical

chromosome abnormalities (54% in <35 to 82% in

>40 y. o.).

Other factors may contribute to only 20% of lost

ongoing implantation potential

Conclusions: Chromosome abnormalities

- Mosaics are common but the majority of them have 100% abnormal cells once analyzed for all chromosomes.

- This results in less than 3% of errors.

- Error rates are laboratory-dependent and are the best parameter to evaluate a PGD lab.

Conclusions: Mosaicism not an issue

With day 3 biopsy:

- Only 3% error rate with aCGH

- If normal embryos are available ongoing implantation

rates improve.

With day 5 biopsy and vitrification:

- 0% error rate

- 60-80% implantation rate

- Optimal endometrium receptivity

Conclusions: arrays

Santiago Munné, PhD, DirectorJacques Cohen, PhD, Director

munne@reprogenetics.com www.reprogenetics.com

Pere Colls, Ph.DDagan Wells, Ph.DGeorge Pieczenik, Ph.DJorge Sanchez, PhDJohn Zheng, MDTomas Escudero, MSKelly Ketterson, MSJill Fischer, MSGary Harton, MS

Jessica Vega, MSTim SchimmelSasha SadowySophia TormasiN-neka GoodallRenata PratesPiedad GarzónLaurie FerraraBekka Sellon-WrightMaria Feldhaus

USA

SpainMireia Sandalinas, MSCarles Giménez, PhD

César Arjona, MSAna Jiménez, PhDElena Garcia, MS

JapanTetsuo Otani, MD

Muriel Roche, PhDMiho Mizuike

UKDagan Wells, PhD

Elpida Fragouli, PhDSamer Alfarawati, MS

Michalis Konstantinidis

South AmericaPaul Lopez

Luis Alberto Guzman

GermanyKarsten Held, MD