Elder K. Human preimplantation embryo selection

Подождите немного. Документ загружается.

HUMAN PREIMPLANTATION EMBRYO SELECTION

All calculations were re-analyzed for chromosome

abnormalities in multinucleated embryos, exclud-

ing aneuploidy and the influence of maternal age.

However, if an embryo analyzed as aneuploid also

possessed another chromosome abnormality (i.e.

ANE-MOS or ANE-POL), these were included in

the calculations. Rates of mosaicism were classified

according to parameters previously published.

Embryos from translocation patients were not

included in this study.

CHROMOSOMAL STATUS OF NON-MNE

As a control reference, a group of supernumerary

embryos that had not displayed multinucleation on

day 2 or day 3 (non-MNE) were analyzed. These

embryos were either previously rejected for transfer

or cryopreservation, or had previously been diag-

nosed as abnormal during preimplantation genetic

diagnosis (PGD) cycles. Excluding aneuploidy,

38.5% of all non-MNE were chromosomally nor-

mal (Table 5.6), and 27% were MOS-HI (majority

of the cells of the embryo were abnormal). The total

percentage of non-MNE exhibiting mosaicism was

46.3%, which was the lowest incidence in all of the

types of embryo analyzed in this study (Table 5.7).

The low percentage of normal embryos observed in

non-MNE (38.5% excluding aneuploidy) was not

surprising, considering their source as detailed

above. Furthermore, the number of chromosomal



Table 5.6 Incidence (%) and severity of multinucleation in relation to chromosomal abnormalities (excluding

aneuploidy)

Total no. Total Total Total

MNB type embryos Normal Abnormal MOS POL HAP ANE-MOS MOS-HI MOS-LO ANE-POL POL

2 cell day 2 MNE 2wⱖ2 89 44 56 46 5 6 8 34 4 1 3

All non-MNE 1630 39 61 46 11 4 11 27 9 1 10

2 cell, day 2 MNE 186 36 64 54 5 4 8 42 4 1 5

Day 2 MNE 356 32 68 49 14 6 6 39 3 1 14

All MNE 498 28 72 50 18 4 7 39 4 1 16

2 cell, day 2 MNE 1w2 45 27 73 64 7 2 7 56 2 0 7

All MNE 1w3 72 26 74 57 15 2 10 44 3 0 15

All MNE 1w2 118 25 75 59 15 1 8 46 5 2 14

Day 3 MNE 215 21 79 55 22 2 9 41 5 2 20

MNB, multinucleated blastomere; MOS, mosaic; POL, polyploidy; HAP, haploid; HI, high rate; LO, low rate; ANE, aneuploid.

Table 5.7 Incidence (%) and severity of multinucleation in relation to chromosomal abnormalities (excluding

aneuploidy), sorted by highest rate of total mosaicism (MOS)

Total no. Total Total Total

MNB type embryos Normal Abnormal MOS POL HAP ANE-MOS MOS-HI MOS-LO ANE-POL POL

2 cell, day 2 MNE 1w2 45 27 73 64 7 2 7 56 2 0 7

All MNE 1w2 118 25 75 59 15 1 8 46 5 2 14

All MNE 1w3 72 26 74 57 15 2 10 44 3 0 15

Day 3 MNE 215 21 79 55 22 2 9 41 5 2 20

2 cell, day 2 MNE 186 36 64 54 5 4 8 42 4 1 5

All MNE 498 28 72 50 18 4 7 39 4 1 16

Day 2 MNE 356 32 68 49 14 6 6 39 3 1 14

2 cell day 2 MNE 2wⱖ2 89 44 56 46 5 6 8 34 4 1 3

All non-MNE 1630 39 61 46 11 4 11 27 9 1 10

MNB, multinucleated blastomere; MOS, mosaic; POL, polyploidy; HAP, haploid; HI, high rate; LO, low rate; ANE, aneuploid.

HPE_Chapter05.qxp 7/13/2007 4:45 PM Page 46

anomalies for re-analyzed embryos of PGD cycles is

expected to be artificially high, since normally

dividing euploid embryos are transferred or cryo-

preserved and not subjected to further analysis.

CHROMOSOMAL STATUS OF ALL MNE (REGARDLESS

OF SEVERITY/DAY OF MULTINUCLEATION)

The chromosomal status of embryos that had dis-

played MNB on day 2 or day 3 was determined by

fixation and FISH on days 3, 4 or 5. Excluding

aneuploidy, only 28.1% of MNE were chromosomally

normal, compared to 38.5% for non-MNE (Table 5.6;

pⱕ 0.001). The total percentage of MNE exhibiting

mosaicism was 50.1%, which was not significantly

different from the 46.3% incidence of mosaicism

observed for non-MNE. However, a significant dif-

ference was seen between the number of MOS-HI

embryos found in MNE and non-MNE (Table 5.7):

38.9% MOS-HI for MNE vs. 27% for non-MNE

(p ⱕ 0.001). Notably higher rates (p ⱕ 0.001) of

total polyploidy were observed for MNE (17.5%)

when compared to rates of total polyploidy for

non-MNE (11%; Table 5.8).

CHROMOSOMAL STATUS OF DAY 2 VERSUS DAY

3 MNE (REGARDLESS OF SEVERITY OF

MULTINUCLEATION)

Day 2 and day 3 MNE were assessed for chromosomal

status. Excluding aneuploidy, only 20.5% of day 3

MNE were normal compared to 31.5% of day 2

MNE (Table 5.6; p ⱕ 0.002). Virtually the same rate

of MOS-HI (40.2%) was observed for day 3 MNE as

for day 2 MNE (39.0%; Table 5.7). The rate of total

mosaicism did not differ significantly between day 3

MNE (54.9%) and day 2 MNE (50.1%). There was

an interesting disparity in rates of total polyploidy

between day 2 and day 3 embryos, with a signifi-

cantly higher rate (p ⱕ 0.01) of total polyploidy

(22.3%) observed for day 3 MNE compared to the

rate for day 2 MNE (14.4%; Table 5.8).

CHROMOSOMAL STATUS OF MNE WITH ONE

BINUCLEATED VERSUS ONE MICRONUCLEATED

BLASTOMERE (REGARDLESS OF THE DAY ON WHICH

MULTINUCLEATION WAS FIRST OBSERVED)

MNE with only one binucleated blastomere (i.e. 1w2),

or only one micronucleated blastomere (i.e. 1ⱖ3),

irrespective of the day on which these were observed,

were assessed for their chromosomal status. All of

the other cells in these embryos had either a single

nucleus, or no nucleus visible. Excluding aneuploidy,

the total number of embryos that were chromo-

somally normal did not differ significantly between

binucleated embryos (25.4%) and micronucleated

embryos (26.3%; Table 5.6). The percentage of bi-

nucleated and micronucleated embryos exhibiting

extensive mosaicism was 45.7% and 44.4%, respec-

tively, and the percentage of embryos that were com-

pletely mosaic were comparable, 58.4% and 56.9%

MULTINUCLEATION AND MOSAICISM IN THE HUMAN PREIMPLANTATION EMBRYO



Table 5.8 Incidence (%) and severity of multinucleation in relation to chromosomal abnormalities (excluding

aneuploidy), sorted by highest rate of total polyploidy

Total no. Total Total Total

MNB type embryos Normal Abnormal MOS POL HAP ANE-MOS MOS-HI MOS-LO ANE-POL POL

Day 3 MNE 215 21 79 55 22 2 9 41 5 2 20

All MNE 498 28 72 50 18 4 7 39 4 1 16

All MNE 1w2 118 25 75 59 15 1 8 46 5 2 14

All MNE 1w3 72 26 74 57 15 2 10 44 3 0 15

Day 2 MNE 356 32 68 49 14 6 6 39 3 1 14

All non-MNE 1630 39 61 46 11 4 11 27 9 1 10

2 cell, day 2 MNE 1w2 45 27 73 64 7 2 7 56 2 0 7

2 cell, day 2 MNE 186 36 64 54 5 4 8 42 4 1 5

2 cell, day 2 MNE 2wⱖ2 89 44 56 46 5 6 8 34 4 1 3

MNB, multinucleated blastomere; MOS, mosaic; POL, polyploidy; HAP, haploid; HI, high rate; LO, low rate; ANE, aneuploid.

HPE_Chapter05.qxp 7/13/2007 4:45 PM Page 47

HUMAN PREIMPLANTATION EMBRYO SELECTION

(Table 5.7). Approximately 15% of binucleate or

micronucleated embryos were polyploid (Table 5.8).

CHROMOSOMAL STATUS OF TWO-CELL DAY 2 MNE

WITH BOTH CELLS MULTINUCLEATED

Non-arrested two-cell embryos with both cells

multinucleated (i.e. 2wⱖ2) on day 2 were assessed

for chromosomal status after further cleavage on

days 3, 4 or 5 (Figure 5.1D); when aneuploidy was

excluded, 43.8% of the 89 embryos analyzed were

normal (Table 5.6). Interestingly, this category of

embryos had the highest proportion of chromo-

somally normal embryos, comparable to the 38.5% of

chromosomally normal embryos observed for non-

MNE; however, it is important to note that these

results may be due to the relatively small dataset of

this particular subgroup of MNE. Results further

demonstrated that two-cell day 2 MNE with both

cells multinucleated had the lowest proportion of

MOS-HI (33.7%), total MOS (46.1%) and total-POL

(4.5%) embryos among all types of multinucleation

analyzed (Tables 5.7 and 5.8). These embryos had

the same rates of Tot-MOS (46.1%) as non-MNE

and were second to non-MNE in terms of lowest

rates of extensive mosaicism (MOS-HI) (Table 5.7).

These findings support the conclusion that MNB,

despite showing multiple nuclear membranes, may

still be capable of undergoing normal spindle forma-

tion and cleavage divisions. Unambiguous evidence

of this phenomenon is the birth of a healthy baby

derived from a completely multinucleated embryo

where all cells were multinucleated on day 2.

IMPACT OF MULTINUCLEATION ON

CLINICAL OUTCOMES

Comprehensive studies have reported inferior clinical

outcomes after transfer of MNE compared to transfer

of non-MNE.

4,8–10

Implantation and pregnancy

rates at IRMS after transfer of MNE exclusively were

approximately half of the rates after transfer of non-

MNE exclusively (Figure 5.2). However, in analyzing

cycle outcomes when only two embryos were trans-

ferred, the higher implantation and pregnancy rates

for controls are to some extent exaggerated, because



the majority of these transfers were carried out in

younger patients, in whom only two embryos are

transferred as a routine, in order to reduce the like-

lihood of multiple pregnancy. In contrast, transfer

of MNE exclusively occurs only in cycles with few

embryos or poor overall embryo development.

Embryos were selected for transfer and/or cryo-

preservation based on their assessment as having

the best prognosis for establishing viable pregnan-

cies.

3,5,26,39,40

Assisted hatching was performed as

indicated, based on criteria published by Cohen

et al.

Embryo transfers were usually performed

on day 3 of development (76–79 hours post oocyte

retrieval). Clinical pregnancy was defined as fetal

heart activity in the presence of an appropriate rise

in serum ␤hCG levels. A positive rise in serum

␤hCG level with a gestational sac, but where fetal

heart was absent, was classified as a blighted ovum

and not considered as a clinical pregnancy for this

large study of MNE.

Exclusive transfers of embryos where multinu-

cleation was observed first on day 2 had much higher

implantation rates than exclusive transfers of embryos

where multinucleation was first observed on day 3

(Figures 5.2 and 5.3). This would indicate that multi-

nucleation that occurs initially on day 2 is less

detrimental to embryos than when it is initiated on

Non-MNE (n = 290)

All MNE (n = 60)

Day 2 MNE (n = 39)

Day 3 MNE (n = 21)

Cell #

Fragmentation (%)

Implantation (%)

Pregnancy (%)

Figure 5.3 Comparison of clinical outcomes for transfer of

non-MNE exclusively, all MNE, day 2 MNE and day 3 MNE

exclusively, limited to transfer of one embryo only.

HPE_Chapter05.qxp 7/13/2007 4:45 PM Page 48

MULTINUCLEATION AND MOSAICISM IN THE HUMAN PREIMPLANTATION EMBRYO

day 3. Indeed, when chromosome constitutions of

day 2 and day 3 MNE were investigated and com-

pared, there were higher percentages of normal

embryos from day 2 (32%) than day 3 (21%), which

supports the above clinical findings (i.e. higher

implantation and pregnancy rates for exclusive trans-

fers of embryos whose multinucleation was first

observed on day 2 when compared to exclusive trans-

fers of embryos whose multinucleation was first

observed on day 3). Since a greater percentage of the

embryo is affected by multinucleated cells on day 2

than on day 3, these observations are unexpected,

and suggest the possibility that the process and con-

sequence of creating multiple nuclei within a blas-

tomere may differ according to the developmental

stage of the embryo.

Implantation and pregnancy rates (18% and 36%,

respectively) for exclusive transfers of binucleated

embryos (regardless of the day on which multinu-

cleation was first observed) were the highest for any

type of multinucleation patterns analyzed (Table 5.5).

It was surprising, therefore, to find that only 25% of

binucleated MNE were normal after chromosome

analysis. Although they had better positive clinical

outcomes, they also had one of the highest rates of

chromosome abnormality (Table 5.6). In contrast,

the embryos considered to contain the most severe

type of multinucleation (i.e. both cells multinucle-

ated on day 2) had the highest percentage of chromo-

somally normal embryos (44%); however, this may

again be simply a function of different sample sizes

when comparing the groups.

CONCLUSION

Embryos with multinucleated blastomeres are com-

monly observed during the development of pre-

implantation embryos cultured in vitro. Factors that

have been found to correlate with the prevalence of

multinucleation include an aggressive and/or high

response to ovarian stimulation, and features of

in vitro culture systems. Although the implantation

and pregnancy rates for these dysmorphic embryos

are lower than those for non-MNE, nonetheless

they still have the potential to lead to a healthy live

birth. Multinucleation is most easily observed on

day 2, and can also be seen, to a lesser extent, on day 3.

Embryos that display multinucleation first on day 2

have a better prognosis for implantation than do

those that develop this feature on day 3. MNE were

noted to have higher rates of extensive mosaicism,

haploidy and polyploidy than non-MNE, with mosai-

cism as the most common chromosomal abnormality

observed. Rates of mosaicism within the MNE

themselves appear to depend on the type and sever-

ity of multinucleation. After transfer of MNE exclu-

sively, those with binucleated cells were associated

with a better clinical outcome, although these embryos

overall had the highest rates of chromosome abnor-

malities. In contrast, embryos considered to contain

the most severe type of multinucleation (i.e. all cells

of the embryo were multinucleated on day 2) had

the highest percentage of normal embryos (44%)

when FISH was performed at later cleavage stages,

which indicates that multinucleated cells may undergo

some mechanism which lends itself to normal

spindle function and cleavage division (despite

multiple nuclear membranes) and eventually give

rise to diploid daughter cells.

This comprehensive review of multinucleated

cleavage stage human embryos has yielded some

unexpected results, which give us a better under-

standing regarding the incidence and effects of

multinucleation. This knowledge will hopefully

result in more effective embryo selection for uterine

transfer and/or cryopreservation during IVF cycles.

ACKNOWLEDGMENTS

Many thanks to Giles Tomkin for his aid with

Eggcyte™ database queries as well as Santiago

Munne, Jacques Cohen, John Garrisi, and Mina

Alikani for their help in formulating and interpret-

ing the data. Thanks to the embryologists at IRMS:

Toni Congedo, Kathleen Ferry, Sheri Klein, Elena

Kissin, Adrienne Reing, and Kenny Smalls.

REFERENCES

1. Sandalinas M, Sadowy S et al. Developmental ability of chromosomally

abnormal human embryos to develop to the blastocyst stage. Hum

Reprod 2001; 16: 1954–8.



HPE_Chapter05.qxp 7/13/2007 4:45 PM Page 49

HUMAN PREIMPLANTATION EMBRYO SELECTION

2. Munne S, Cohen J. Unsuitability of multinucleated human

blastomeres for preimplantation genetic diagnosis. Hum Reprod 1993;

8: 1120–25.

3. Munne S, Alikani M et al. Embryo morphology, developmental rates

and maternal age are correlated with chromosome abnormalities.

Fertil Steril 1995; 64: 382–91.

4. Jackson K, Ginsburg E et al. Multinucleation in normally fertilized

embryos is associated with an accelerated ovulation induction

response and lower implantation and pregnancy rates in in vitro fertil-

ization-embryo transfer cycles. Fertil Steril 1998; 70: 60–6.

5. Alikani M, Calderon G et al. Cleavage anomalies in early human

embryos and survival after prolonged culture in vitro. Hum Reprod

2000; 15: 2634-43.

6. Hardy K, Winston RML, Handyside AH. Binucleate blastomeres in

preimplantation human embryos in vitro: failure of cytokinesis during

early cleavage. J Reprod Fertil 1993; 98: 549–59.

7. Mohr LR, Trounson AO et al. Evaluation of normal and abnormal

human embryo development during procedures in vitro 1983. In:

Beier HM, Lindner HR, eds. Fertilization of the Human Egg In-Vitro.

Heidelberg: Springer Verlag 1983; 211–22.

8. Balakier H and Cadesky K. The frequency and developmental capabil-

ity of human embryos containing multinucleated blastomeres. Hum

Reprod 1997; 12: 800–4.

9. Pelinck MJ, De Vos M et al. Embryos cultured in vitro with multinu-

cleated blastomeres have poor implantation potential in human in-vitro

fertilization and intracytoplasmic sperm injection. Hum Reprod 1998;

13: 960–3.

10. Van Royen E, Mangelschots K,Vercruyssen M et al. Multinucleation in

cleavage stage embryos. Hum Reprod 2003; 18: 1062–9.

11. Handyside AH, Pattinson JK et al. Biopsy of human pre-implantation

embryos and sexing by DNA amplification. Lancet 1989; 1: 347-9.

12. Grifo JA, Boyle A et al. Preimplantation biopsy and analysis of blas-

tomeres by in situ hybridization. Am J Obstet Gynecol 1990; 163:

2013–19.

13. Kligman I, Benadiva C et al. The presence of multinucleated blas-

tomeres in human embryos is correlated with chromosomal abnor-

malities. Hum Reprod 1996; 11: 1492–8.

14. Staessen C, Van Steirteghem AC et al. The genetic constitution of

multinucleated blastomeres and their derivative daughter blastomeres.

Hum Reprod 1998; 13: 1625–31.

15. Magli MC, Gianaroli L, Ferrareti AP. Chromosomal abnormalities in

embryos. Mol Cell Endocrinol 2001; 22(Suppl 1): S29–34.

16. Meriano J, Clark C, Cadesky K et al. Binucleated and micronucleated

blastomeres in embryos derived from human assisted reproduction

cycles. RBM online 2004; 9: 511–20.

17. Munne S, Magli, C et al. Treatment-related chromosome abnormalities

in human embryos. Hum Reprod 1997; 12: 780–4.

18. Tesarik J,Kopecny V et al.Ultrastructural and autoradiographic obser-

vations on multinucleated blastomeres of human cleaving embryos

obtained by in-vitro fertilization. Hum Reprod 1987; 2: 127–36.

19. Munne S, Cohen J. Chromosome abnormalities in human embryos.

Hum Reprod Update 1998; 4: 842–55.

20. Van Blerkom J, Freytag M et al. The developmental potential of the

human oocyte is related to the dissolved oxygen content of follicular

fluid: association with vascular endothelial growth factor levels and peri-

follicular blood flow characteristics. Hum Reprod 1997; 12: 1047–55.

21. Nogueira D, Staessen C et al. Nuclear status and cytogenetics of

embryos derived from in-vitro matured oocytes. Fertil and Steril

(2000); 74: 295–8.

22. Winston NJ, Braude PR et al. The incidence of abnormal morphology

and nucleocytoplasmic ratios in 2-, 3-, and 5-day human pre-embryos.

Hum Reprod 1991; 6: 17–44.

23. Pickering SJ, Taylor A et al. An analysis of multinucleated blas-

tomere formation in human embryos. Mol Hum Reprod 1995; 10:

1912-22.

24. Angell RR, Sumner AT, West JD et al. Post-fertilization polyploidy

in human preimplantation embryos fertilized in vitro. Hum Reprod

1996; 2: 721–7.

25. Pickering SJ, Braude PR et al. Transient cooling to room temperature

can cause irreversible disruption of the meiotic spindle in the human

oocyte. Fertil Steril 1990; 54: 102–8.

26. Puissant F, Van Rysselberge M et al. Embryo scoring as a prognostic

tool in IVF treatment. Hum Reprod 1987; 2: 705–8.

27. Giorgetti C, Terriou P et al. Embryo score to predict implantation after

in vitro fertilization: based on 957 single embryo transfers. Hum

Reprod 1995; 10: 2427–31.

28. Ziebe S, Petersen K et al. Embryo morphology or cleavage stage: how

to select the best embryos for transfer after in vitro fertilization. Hum

Reprod 1997; 12: 1545–49.

29. Alikani M, Cohen J et al. Human embryo fragmentation in-vitro and

its implications for pregnancy and implantation. Fertil Steril 1999;

71:836–42.

30. Alikani M, Cohen J. Patterns of cell fragmentation in the human

embryo in vitro. J Assist Reprod Genet 1995; 12(Supp 1): 28s.

31. Hardarson T, Hanson C et al. Human embryos with unevenly

sized blastomeres have lower pregnancy and implantation rates: indi-

cations for aneuploidy and multinucleation. Hum Reprod 2001; 16:

313–8.

32. Laverge H, Sutter P et al. Triple colour fluorescent in-situ hybridiza-

tion for chromosomes X, Y and 1 on spare human embryos. Hum

Reprod 1997; 12: 809–14.

33. Almeida P, Bolton V. The relationship between chromosomal abnor-

mality in the human preimplantation embryo and development in-vitro.

Reprod Fertil Dev 1996; 8: 235–41.

34. Marquez C, Sandalinas M et al. Chromosome abnormalities in 1255

cleavage-stage human embryos. RBM Online 2000; 1: 17–27.

35. Munne S, Grifo J et al. Chromosome abnormalities in human arrested

preimplantation embryos: a multiple-probe FISH study. Am J Hum

Genet 1994A; 55: 150–9.

36. Artley J, Braude P, Johnson M. Gene activity and cleavage arrest in

human pre-embryos. Hum Reprod 1992; 7: 1014–21.

37. Munne S, Weier U et al. Reduction in signal overlap results in

increased FISH efficiency: implications for preimplantation genetic

diagnosis. J Assist Reprod Genet 1996; 13: 149–56.

38. Velilla E, Escudero T et al. Blastomere fixation techniques and risk of

misdiagnosis for preimplantation genetic diagnosis of aneuploidy.

RBM Online 2002; 4: No 3.

39. Cohen J, Inge K et al. Video-cinematography of fresh and cryopre-

served embryos: a retrospective analysis of embryonic morphology

and implantation. Fertil Steril 1989; 51: 820.

40. Bolton VN, Hawes SM et al. Development of spare human preimplan-

tation embryos in vitro: an analysis of the correlations among

gross morphology, cleavage rates, and development to the blastocyst.

J In Vitro Embryo Transf 1989; 7: 186.

41. Cohen J, Alikani M et al. Rescuing abnormally developing embryos by

assisted hatching. In: Mori T, Aono T, Tominage T, Hiroi M, eds.

Frontiers in Endocrinology, Perspectives on Assisted Reproduction.

Rome: Ares Serono Symposia, 1994; 4: 536–44.



HPE_Chapter05.qxp 7/13/2007 4:45 PM Page 50

INTRODUCTION

The typical course of development following normal

fertilization of the human egg in vitro begins with

completion of second meiosis, extrusion of a polar

body (PB2), and the emergence of one male and one

female pronucleus some 9–11 hours later (day 1).

Development continues with three to four mitotic

divisions in the following 48 hours, producing an

embryo with 2–4 (on day 2), and 8–10 (on day 3)

roughly equal-sized mononucleated cells; division

continues followed by compaction of the cells on

day 4, and cavitation and differentiation of the

embryo into two distinct cell types – inner cell mass

and trophectoderm – between days 4 and 6 of

development (Figure 6.1).

However, in vitro fertilization (IVF) and culture

of human eggs often leads to atypical development.

A recent survey of the EggCyte clinical embryology

database

indicates that roughly three-quarters of

over 79 000 human embryos generated in our labo-

ratory were affected by cell fragmentation, that is,

they contained non-nucleated cell fragments of

varying size and number along with nucleated cells.

The survey also revealed that fragmentation is often

accompanied by one or more other abnormalities,

including blastomere multinucleation, size discrep-

ancy, and disorganization, as well as reduced cell–cell

adhesion and abnormally thick zonae pellucidae

(Figure 6.1).

Fragmentation was described in human IVF

embryos as early as 1970,

but the phenomenon is

neither unique to IVF, nor to human embryos.

Fragments have been found in in vivo developed

embryos in the human

3–6

and in in vivo and in

vitro produced embryos of practically all other

mammalian species studied so far.

7–9

It has long been known that human embryo

fragmentation is associated with decreased embryo

viability

and given its prevalence, embryo frag-

mentation presents a formidable challenge to the

success and efficiency of human IVF.

It is also clear that fragmentation hampers the

efficiency of somatic cell nuclear transplantation

in mammals since many eggs fragment following

enucleation or after reconstitution with a donor

nucleus.

11–13

While many of the early studies on fragmenta-

tion in the human viewed this ‘anomaly’ in absolute

terms, i.e. whether it was present or absent, more

recently, detailed observations have led to a better

understanding of the relationship between frag-

mentation, developmental competence, and viability.

There is particular interest in determining the

causes and mechanisms of fragmentation. Earlier

studies identified a number of potentially contribut-

ing factors, including adverse follicular conditions,

in vitro culture conditions, suboptimal culture

medium composition, reactive oxygen species, and

chromosomal abnormalities.

14–18

More recently, a number of interesting observa-

tions have been made during detailed microscopic

examination of embryos at 8–12 hour intervals.

Many of these observations are relevant in the

context of the origins of fragmentation. For

instance, little change was noted in the volume and

distribution of fragments in embryos that were

mitotically inactive. This is at least suggestive that

fragmentation does not occur in arrested cells.

Over time, some fragments were seen to swell and

6. The origins and consequences of

fragmentation in mammalian

eggs and embryos

Mina Alikani



HPE_Chapter06.qxp 7/13/2007 4:46 PM Page 51

HUMAN PREIMPLANTATION EMBRYO SELECTION

lyze, raising the possibility of a ‘toxic’ microenvi-

ronment within the confines of the zona pellucida.

Many ‘fragments’ present at the 2–4-cell stage were

undetectable at the 8–10-cell stage and the absence

of residual cellular debris suggested resorption

rather than lysis of these structures. This was

reported to be accompanied by ‘turbulence’ or

movement in the underlying cytoplasm, and repre-

sentative of the ‘transient’ nature of fragments in

some cases.

Time-lapse video images

19,20

showed that the

initial fragmentation event occurred over a

30 minute interval in about 90% of the embryos. At

the same time, some embryos with extensive frag-

mentation during early cleavage stages were noted

to have returned to apparently normal morphology

by day 3 of development. In some embryos, frag-

mentation was described as ‘episodic,’ occurring at

both the 2- and 4-cell stages. This would imply that

fragmentation is not restricted to the first cleavage

division. The observations regarding ‘disappear-

ance’ of fragments following completion of division

point to normally occurring changes in cell shape

that are cell cycle specific. They highlight the

importance of distinguishing between the different

types of fragments and defining fragmentation in

relation to the cell cycle.

DEFINING FRAGMENTATION AND

DELINEATING ITS CONSEQUENCES

IN HUMAN EMBRYOS

The extent to which fragmentation occurs, along

with cell number, are primary criteria for selecting

embryos for transfer during clinical IVF, but the



Day 1

BC D

FGH I J

Day 2 Day 3 Day 4 Day 5

Figure 6.1 Human embryos showing normal (A)–(E) and abnormal (F)–(J) development on days 1 through 5 in culture. (A) At the

zygote stage, pronuclei are evenly sized, the cytoplasm is slightly contracted, and two polar bodies (pb) are visible. (B) On day 2, the

embryo has gone through two divisions and has four evenly sized cells, each with one nucleus. (C) On day 3 of development, the

embryo is at the 8-cell stage with evenly sized, mononucleate cells and minimal fragmentation (arrowhead). (D) On day 4, the

embryo has undergone full compaction, a prerequisite for normal blastulation. (E) On day 5, a blastocyst has formed with a distinct

inner cell mass (arrowhead) and a cohesive trophectoderm. Embryos in (F)–(J) show various morphological abnormalities, including

(F) uneven pronuclear size (pronuclei are outlined); (G) slow development and multinucleation of blastomeres (multiple nuclei are

outlined), and fragmentation (arrowhead); (H) extensive fragmentation (arrowheads) and disorganization of blastomeres;

(I) abnormal compaction with exclusion of cells and fragments (arrowheads); and (J) abnormal blastulation and a thick zona

pellucida (arrow).

HPE_Chapter06.qxp 7/13/2007 4:46 PM Page 52

ORIGINS AND CONSEQUENCES OF FRAGMENTATION

adoption of a universal and reproducible fragmen-

tation classification system has been slow. In some

laboratories, loosely defined morphological terms

(such as minor or major fragmentation) are still

being used for embryo characterization and incor-

porated into ‘embryo grading’ systems that fail to

convey adequate detail. Other laboratories includ-

ing our own use somewhat more specific, albeit still

subjective, descriptions of fragmentation in their

evaluation schemes. These descriptions entail a

close approximation of the volume of anucleate

fragments, expressed as a percentage of the total

volume of the embryo, and referred to as the ‘degree’

of fragmentation. Apart from fragmentation degree,

non-invasive morphological evaluation of large

numbers of embryos led us to define several distinct

patterns of fragmentation on day 3 of development

(Figure 6.2).

The patterns are distinguished on the basis of the

size and location of fragments relative to the size

and position of nucleated cells: few small fragments

typically associated with only one blastomere (type

I), many small and localized fragments associated

with one or more cells (type II), small and scattered

fragments associated with multiple cells (type III),

large and scattered fragments associated with several

unevenly sized cells (type IV), and fragments appear-

ing with a characteristic granularity associated with

cytoplasmic contraction in intact blastomeres

(type V). Some embryos, particularly those with

extensive fragmentation, do not show a specific pat-

tern, but a non-distinct combination of patterns.

These embryos were classified as having no distinct

pattern (NDP) of fragmentation. Type III is the most

frequently occurring pattern of fragmentation,

while type V is rare (Figure 6.3).

The patterns do not appear to be related to

maternal age, but the degree of fragmentation,

contrary to general belief, appears to decrease with

increasing maternal age.

The reason for this is not

immediately clear, particularly since, according to

our database, average cell number is not affected by

maternal age.

Our original study

also showed that embryos

with localized fragments (type II) had on average

fewer cells on day 2 than embryos with all other

patterns of fragmentation. On day 3, embryos with

both types II and IV fragmentation had fewer cells

than those with other patterns. The differences in

cell number are both interesting and important, and

are likely to be related to how the different patterns

arise. In particular, in the case of type II and IV,

reduced cell number suggests that these patterns

may result from loss of one or more cells, and/or



AA⬘

BB⬘

CC⬘

DD⬘

Figure 6.2 Human embryos showing different patterns of frag-

mentation, before (left panel) and after (right panel) microsur-

gical fragment removal. (A) and (B) Type II; (C) type III; (D)

type IV. Note the gross discrepancy in size of the blastomeres in

(D) and (D⬘).

HPE_Chapter06.qxp 7/13/2007 4:46 PM Page 53

HUMAN PREIMPLANTATION EMBRYO SELECTION

slow division or arrest of cells following loss of large

volumes of cytoplasm.

A similar fragmentation classification system was

developed by Antczak and Van Blerkom,

which

considered the temporal as well as spatial character-

istics of fragmentation: type I fragmentation was

defined as a monolayer carpet of very small fragments

that coated only a modest portion of the cell surface

with minor or no apparent reduction in cell size;

multiple layers of fragments involving a consider-

able portion of the cell surface and accompanied by

a significant reduction in cell size was designated as

type II; type III was defined as complete fragmenta-

tion of one blastomere in an embryo showing no

fragmentation on a previous day; type IV was defined

as occasional fragments of variable size scattered

over several blastomeres in otherwise normal appear-

ing and developmentally progressive embryos.

These patterns, in ascending order, are roughly

equivalent to fragmentation types III, IV, II, and I as

defined by Alikani et al.

Both the degree and the

pattern of fragmentation determine the viability of

fragmented embryos following intrauterine transfer

and their survival in extended culture to the blasto-

cyst stage.

THE IMPACT OF FRAGMENTATION ON

IMPLANTATION, PREGNANCY, AND

NEONATAL OUTCOME

The relationship between fragmentation and

embryo viability has been analyzed in a number of

studies. With the exception of one study

in which

neither implantation nor multiple birth correlated

with fragmentation, all these studies indicate that

cleavage stage embryos with no fragmentation

establish pregnancy more readily than those with

fragmentation.

21,22,24–28

Our own study

included a large number of

‘homogeneous’ cases – those in which more than

half of transferred embryos were in the same mor-

phological category, with respect to degree and/or

type of fragmentation. This was an important (and

unique) aspect of the study, and its aim was to elim-

inate or minimize correlative uncertainties when



Figure 6.3 The frequency of fragmentation patterns (bars, left Y axis) and the average fragmentation degree for each pattern (red

line, right Y axis). Types III and IV are the most commonly occurring patterns.

Percentage

Type I Type II Type III Type IV Type V NDP No

fragments

Fragmentation pattern frequency

Degree of fragmentation according to pattern

HPE_Chapter06.qxp 7/13/2007 4:46 PM Page 54

multiple embryos were transferred, but without

restricting the analyses to too few cases. Homo-

geneous transfers in this study included all single

embryo transfers, two-embryo transfers in which

both embryos were in the same category, three-

embryo transfers in which two or three of the

embryos were in the same category, etc. The majority

of transfers involved two to four embryos. The

study established a clear relationship between the

degree of fragmentation and embryo viability.

Fewer fetuses developed (85 fetuses/420 transferred

embryos; 20.2%) following transfer of embryos the

majority of which had ⬎15% fragmentation than

when the majority had ⱕ15% fragmentation

(1681/5496; 30.1%) at the time of evaluation, on the

morning of day 3 of development.

We recently reanalyzed 3322 transfers (regular

and oocyte recipient cycles) in which all embryos

(a total of 8064 transferred embryos) were in the

same category with respect to ‘fragmentation degree’

(Tables 6.1 and 6.2). We ran contingency table

analyses to test whether the proportions of embryos

implanted and of pregnancies achieved were the

same among the five fragmentation transfer groups.

The log-likelihood ␹

values generated in these

analyses were significant (p ⬍0.01) in all cases,

suggesting that the frequency of implantation and

pregnancy in fact differed among the embryo frag-

mentation categories. Because of this, we followed

up by testing ten pair-wise contrasts among the five

fragmentation groups and evaluated the probabili-

ties associated with the ␹

values from these contrasts

ORIGINS AND CONSEQUENCES OF FRAGMENTATION



Table 6.1 Implantation and pregnancy outcome in homogeneous transfer groups including oocyte recipients

Fragmentation

degree transfer No. of embryos Implantation

group No. of procedures No. of pregnancies Pregnancy rate

(%) transferred No. of FHB rate

(%)

(1) 0–5% 2295 1270 55.3 5864 1940 33.1

(2) 6–15% 745 341 45.8 1730 503 29.1

(3) 16–25% 172 47 27.3 307 54 17.6

(4) 26–35% 67 15 22.4 92 18 19.6

(5) over 35% 43 6 14.0 71 6 8.5

FHB, fetal heart beat.

Pair-wise contrasts (Bonferroni corrected) of pregnancy results in groups 1–5 show statistically significant differences between all except group 3 vs groups

4 and 5; and group 4 vs group 5.

Pair-wise contrasts of implantation results in groups 1–5 show statistically significant differences in all except group 2 vs 4; group 3 vs groups 4 and 5; and

group 4 vs group 5.

Table 6.2 Other characteristics of the embryos in homogeneous fragmentation degree transfer groups

Average Average day-3 Average day-3

Fragmentation day-3 cell Average day-3 cell number

fragmentation

degree transfer number

fragmentation

(transferred (transferred

group (all embryos) ⫾SD (all embryos) (%) ⫾SD embryos) ⫾SD embryos) (%) ⫾SD

(1) 0–5% 6.7 1.2 8.4 6.8 7.6 1.2 2.1 2.0

(2) 6–15% 6.0 1.3 19.6 9.6 7.2 1.4 12.3 2.0

(3) 16–25% 5.3 1.5 29.0 11.7 6.4 1.5 22.1 2.2

(4) 26–35% 4.9 1.3 38.9 11.7 6.2 1.2 33.0 2.3

(5) over 35% 4.3 1.3 47.8 14.2 5.6 1.0 45.8 7.2

Differences among the groups are significant at the 0.05 significance level (Newman–Kuels multiple comparisons test).

HPE_Chapter06.qxp 7/13/2007 4:46 PM Page 55