Sunteți pe pagina 1din 10

Molecular Human Reproduction, Vol.20, No.2 pp.

117–126, 2014
Advanced Access publication on November 1, 2013 doi:10.1093/molehr/gat073

ORIGINAL RESEARCH

Morphological and cytogenetic


assessment of cleavage and blastocyst
stage embryos
E. Fragouli 1,2,*, S. Alfarawati2, K. Spath 1, and D. Wells 1,2
1
Nuffield Department of Obstetrics and Gynaecology, University of Oxford, Oxford OX3 9DU, UK 2Reprogenetics UK, Oxford OX4 2HW, UK

*Correspondence address. E-mail: elpida.fragouli@obs-gyn.ox.ac.uk or elpida@reprogenetics.co.uk

Downloaded from http://molehr.oxfordjournals.org/ at Aston University on January 16, 2014


Submitted on August 11, 2013; resubmitted on October 10, 2013; accepted on October 15, 2013

abstract: Morphological assessments are the main way in which fertility clinics select in vitro generated embryo(s) for transfer to the uterus.
However, it is widely acknowledged that the microscopic appearance of an embryo is only weakly correlated with its viability. Furthermore, the
extent to which morphology is affected by aneuploidy, a genetic defect common in human preimplantation embryos, remains unclear. Aneuploidy
is of great relevance to embryo selection as it represents one of the most important causes of implantation failure and miscarriage. The current
study aimed to examine whether morphological appearance can assist in identifying embryos at risk of aneuploidy. Additionally, the data produced
sheds light on how chromosomal anomalies impact development from the cleavage to the blastocyst stage. A total of 1213 embryos were exam-
ined. Comprehensive chromosome analysis was combined with well-established criteria for the assessment of embryo morphology. At the cleav-
age stage, chromosome abnormalities were common even amongst embryos assigned the best morphological scores, indicating that aneuploidy
has little effect on microscopic appearance at fixed time points up until Day 3 of development. However, at the blastocyst stage aneuploidies were
found to be significantly less common among embryos of optimal morphological quality, while such abnormalities were overrepresented amongst
embryos considered to be of poor morphology. Despite the link between aneuploidy and blastocyst appearance, many chromosomally abnormal
embryos were able to achieve the highest morphological scores. In particular, blastocysts affected by forms of aneuploidy with the greatest cap-
acity to produce clinical pregnancies (e.g. trisomy 21) were indistinguishable from euploid embryos. The sex ratio was seen to be equal throughout
preimplantation development. Interestingly, however, females were overrepresented amongst the fastest growing cleavage-stage embryos,
whereas a sex-related skew in the opposite direction was noted for the most rapidly developing blastocysts. In summary, this study confirms
that, at the cleavage stage, chromosome abnormalities have little if any effect on morphological scores assigned using traditional criteria.
At the blastocyst stage some forms of aneuploidy begin to affect microscopic appearance, but in most instances the impact is subtle. In the
case of the most clinically relevant aneuploidies (those capable of forming a pregnancy) there was no detectable effect on morphology at any
preimplantation stage.
Key words: embryo morphology / aneuploidy / cleavage stage / blastocyst stage / aCGH

the correlation between embryo morphology and implantation potential


Introduction is relatively weak (Moyaeri et al., 2008; Alfarawati et al., 2011). Even
In the vast majority of IVF laboratories around the world, embryo selec- embryos considered to be morphologically perfect do not always
tion is still based, for the most part, on morphological evaluation. Visual succeed in implanting in the uterus. Conversely, an appreciable fraction
analyses are typically carried out at three distinct developmental stages: of embryos of suboptimal microscopic appearance do produce healthy
just after fertilization, at the cleavage stage and at the blastocyst stage (if babies. In order to overcome deficiencies in embryo viability assessment,
undertaking extended embryo culture). The morphological criteria con- many IVF clinics worldwide opt to transfer more than one embryo in
sidered include the following: the number of pronuclei and polar bodies some cycles. Although this increases the probability of transferring a
(zygotes); cell number, evenness of mitotic divisions, presence of micro- competent embryo, the risk of multiple pregnancies and associated
nuclei and amount of cellular fragmentation (cleavage embryos); extent medical complications is inevitably increased (Bromer and Seli, 2008).
of blastocoel expansion and quality of inner cells mass (ICM) and troph- Aneuploidy is the principal genetic factor influencing human repro-
ectoderm (TE) (blastocysts) (e.g. Cummins et al., 1986; Puissant et al., ductive success. Prior to implantation, chromosome abnormality is ex-
1987; Van Royen et al., 1999; Gardner and Schoolcraft, 1999; Gianaroli tremely common, affecting more than half of all embryos produced by
et al., 2007; Stensen et al., 2010). It is generally accepted, however, that women over 35 years of age (Wells and Delhanty, 2000; Fragouli et al.,

& The Author 2013. Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved.
For Permissions, please email: journals.permissions@oup.com
118 Fragouli et al.

2011). In most cases embryos affected by such abnormalities undergo 39.4 years, age range 27 – 45 years) and 858 blastocysts from 161 patients
developmental arrest or fail to implant (Scott et al., 2012). Of those an- (average female age 38.6 years, age range 29 –47 years). There was no signifi-
euploid embryos that succeed in establishing a pregnancy, the vast major- cant difference in maternal age or any other clinical parameters for which data
ity eventually miscarry, explaining why approximately two-thirds of were available. The couples who generated these embryos were all undergo-
ing PGS for various indications in several IVF clinics. Participating clinics had
spontaneous abortions are found to be chromosomally abnormal
both ethical approval and appropriate clinical treatment licences in place. It
(reviewed in Nagaoka et al., 2012).
should be noted that only viable embryos were examined during the
Many studies have considered the relationship between embryo
course of this study. Embryos generated by these patients which either
morphology and aneuploidy. One such investigation examined the then arrested, or were not deemed suitable for biopsy, were not available
number of cells at the cleavage stage and correlated it with embryonic for cytogenetic analysis. The ovarian stimulation protocols used were
chromosome constitution (Magli et al., 1998). It was evident from the chosen by the physicians treating the patients, depending on clinical charac-
data obtained that embryos with delayed development (3– 4 cells on teristics, but did not differ significantly from strategies considered to be
Day 3) or those growing unusually fast (9 cells or more on Day 3) routine.
were at an elevated risk of being aneuploid (Magli et al., 1998). Other
investigations carried out at the cleavage stage have revealed associations Embryo scoring
between chromosome errors and the presence of cellular fragmenta- Embryo morphology was assessed and recorded on Day-3 and -5 post-
tion, asymmetric divisions and multinucleated blastomeres (Hardarson fertilization. The referring IVF clinics assessed embryos using standard mor-

Downloaded from http://molehr.oxfordjournals.org/ at Aston University on January 16, 2014


et al., 2003; Munne et al., 2006). More recently, Eaton et al. (2009) exam- phological criteria (Puissant et al., 1987; Steer et al., 1992; Gardner and
ined the relationship between embryo morphology on Day 3, aneuploidy Schoolcraft, 1999, Hardarson et al., 2001). These criteria are outlined in
and development to the blastocyst stage. They observed that, in general, Tables I and II.
embryos of superior morphological quality were more likely to be
euploid for certain chromosomes, including X/Y, 8, 15, 16, 18 and 22, Comprehensive chromosome analysis using
whereas other chromosomes such as 13, 20, or 21 did not show any microarray comparative genomic
obvious effect on either cleavage stage morphology or blastocyst devel- hybridization
opment (Eaton et al., 2009).
The 24 SureTM Cytochip (BlueGnome Ltd, Cambridge, UK) microarray plat-
The aforementioned studies succeeded in proving that chromosomal form was employed for the detailed assessment of chromosomes in cells
abnormality and morphological grade are not entirely independent, al- from the embryos analysed during the course of this study. The protocol
though it was also clear that the relationship between the two is not ab- was as reported in Fragouli et al. (2013). Briefly, the procedure included
solute: embryos receiving the highest morphological scores were lysis and whole genome amplification of the embryonic material, followed
sometimes found to be abnormal, while those achieving poor grades by their Cy3 or Cy5 fluorescent labelling, hybridization along with normal
were often characterized as euploid. Although these investigations male (46,XY) and female (46,XX) reference DNA samples, post-
yielded valuable data, they all utilized fluorescent in situ hybridization hybridization washes, scanning and analysis of images. All these occurred
(FISH) to assess aneuploidy, a method that only permits a partial with the use of reagents and as suggested by the manufacturer (BlueGnome
chromosomal assessment. Even the most detailed FISH-based studies Ltd, Cambridge, UK).
evaluated fewer than half of the chromosomes in each embryo. As
only limited numbers of chromosomes were examined, it is very likely Image analysis and interpretation
that some of the embryos classified ‘normal’ were in fact aneuploid for A laser scanner (InnoScan 710, Innopsys, Carbonne, France) was used to
chromosomes that were not tested, weakening the ability to detect asso- obtain images from the microarray slides. This scanner was capable of excit-
ciations between chromosomal anomalies and morphology. ing the Cy3 and Cy5 fluorophores, and of reading and storing the resulting
Over the last few years, methods allowing simultaneous testing of all hybridization images. Image analysis took place via a combination of the
MAPIX software (Innopsys, Carbonne, France) and the BlueFuse Multi soft-
24 types of chromosome in human embryos have become available
ware (BlueGnome Ltd, Cambridge, UK). Using criteria and algorithms
and are increasingly widely applied for the purpose of preimplantation
recommended by the manufacturers of the software (Bluegnome Ltd,
genetic screening (PGS) (e.g. Fragouli et al., 2010; Yang et al., 2012;
Rubio et al., 2013; Scott et al., 2013). The current investigation made
use of one of these techniques to provide a sensitive evaluation of the Table I Morphological criteria employed during Day 3
impact of aneuploidy on embryo morphology at the cleavage (Day 3) embryo scoring.
and blastocyst (Day 5) stages. Particular attention was paid to the morph-
ology of embryos carrying chromosome abnormalities of types that can Embryo grade/ Interpretation
quality
be observed during later stages of gestation (i.e. potentially compatible ........................................................................................
with implantation, formation of a clinical pregnancy and in some cases 1/very good Even-sized blastomeres, no visible fragmentation,
a live birth, e.g. trisomy 21). light homogeneous cytoplasm
2/good Less than 10–20% fragmentation, even or uneven
blastomeres, and/or non-homogeneous cytoplasm
Materials and Methods 3/average More than 20% fragmentation, even or uneven
blastomeres
Details of analysed embryos 4/poor More than 50% fragmentation, even or uneven
blastomeres
One thousand two hundred and thirteen embryos were included in this
study: 355 cleavage-stage embryos from 64 patients (average female age
Aneuploidy and embryo morphology 119

Table II Morphological criteria employed during Day-5 embryo scoring (Gardner and Schoolcraft, 1999).

Blastocyst Interpretation ICM Interpretation TE Interpretation


expansion grade grade grade
.............................................................................................................................................................................................
6 Hatched blastocyst: the blastocyst has completely A Tightly packed with A Many cells forming a
escaped the zona pellucida many cells cohesive epithelium
5 Hatching blastocyst: trophectoderm (TE) cells have begun B Loosely grouped with B Few cells forming a loose
to herniate through the zona pellucida several cells epithelium
4 Expanded blastocyst: the blastocoel is now larger than the C Very few cells C Very few large cells.
early embryo, and the zona pellucida has begun to thin
3 Full blastocyst: the blastocoel fills the embryo completely
2 Blastocyst: the blastocoel occupies more than one-half of
the volume of the embryo
1 Early blastocyst: the blastocele accounts for less than
one-half of the volume of the embryo

Downloaded from http://molehr.oxfordjournals.org/ at Aston University on January 16, 2014


Cambridge, UK) it was possible to assign molecular karyotypes to the embry- the younger and older groups of women was highly statistically significant
onic samples under investigation, as well as to classify them as normal or an- (P ¼ 0.006).
euploid. There are 3000 DNA clones spotted on 24 sure arrays and following
processing and smoothing with BlueFuse Multi software, the effective detect-
able resolution of the array is 10 Mb. The effect of aneuploidy on embryo
development to the cleavage stage
Statistical analysis A total of 273 (77%) of Day-3 embryos were classified as ‘good’ quality.
The data obtained during the course of this study were analysed with the use Of the remaining embryos, 61 (17%) were classified as ‘average’ and 21
of Fisher’s exact test. (6%) were considered to be of ‘poor’ quality. The morphological criteria
used for this classification are outlined in Table I, whereas the correlation
between cleavage stage morphology and embryo chromosome status is
Results described in Table III.
Of the good quality embryos that were found to be aneuploid, 119/
The incidence of aneuploidy at the cleavage 273 (43.4%) were characterized as complex abnormal, while 63/273
stage (23%) were affected by a single error and 43 (16%) carried 2 errors.
Of the 355 cleavage stage embryos assessed, 58 (16%) were character- The obtained data clearly show that embryos achieving the best morpho-
ized as chromosomally normal and 297 (84%) were aneuploid. The an- logical scores, the types preferred for transfer to the uterus, remain at
euploid embryos were further categorized according to the number of high risk of aneuploidy. It was also evident from our data that complex
chromosomes affected by malsegregation. Hence, 86 (29%) carried a chromosome errors were present in embryos of all morphological
single aneuploidy, 60 (20%) carried two aneuploidies and 151 (51%) grades. These findings are summarized in Table III and are illustrated in
carried 3 or more errors and were considered to be ‘complex’ abnormal. Fig. 1.
A total of 1262 individual chromosome abnormalities were scored in an- A normal sex chromosome complement was scored for a total of 302
euploid cleavage stage embryos. Among them, 652 (52%) were losses cleavage stage embryos (euploid and aneuploid). Of these 177 (59%)
and the remaining 610 were gains (48%). The vast majority (1188, were female and 125 were male (41%). Interestingly, there was a ten-
94%) of these errors affected entire chromosomes, but 74 (6%) dency for female embryos to develop faster than their male counter-
were segmental imbalances, involving loss or gain of chromosomal parts: 11.3% (20) of the female embryos consisted of 10 or more cells
fragments. at the time of morphological evaluation, whereas 5.6% (7) male
The embryos were also assessed in relation to female age. Specifically, embryos were at the same stage. This difference approached statistical
the samples were grouped into two different categories, according to the significance (P ¼ 0.065). There was no increase in the incidence of aneu-
age of the women who generated them. The first category consisted of a ploidy related to a specific sex. These findings are summarized in
total of 82 cleavage stage embryos generated by 13 reproductively Table IV.
younger women (average age: 33.6 years, age range: 27–36 years), The morphology of aneuploid embryos carrying types of chromosome
and the second consisted of the remaining 273 embryos which were gen- error most often associated with the formation of a clinical pregnancy
erated by 51 reproductively older women (average age: 41 years, age (i.e. trisomies 13, 16, 18, 21 and 22, and sex chromosome abnormalities)
range: 37–45 years). Of the 82 embryos analysed for the younger was also evaluated. Specifically, comparisons took place between 18
patient group, 22 (27%) were seen to be euploid and the remaining 60 embryos with such aneuploidies, 58 euploid embryos and 100
(73%) were aneuploid. For the older group, 36 (13%) of all examined embryos carrying single trisomies affecting other chromosomes or
embryos were euploid, whereas chromosome errors were scored for single monosomies. Statistical analysis of the data indicated that morpho-
the other 237 (87%). The difference seen in aneuploidy rates between logical scores were no different, regardless of whether a cleavage stage
120 Fragouli et al.

Table III The chromosome constitution of embryos in Table V The incidence of viable and non-viable
relation to quality at the cleavage stage. aneuploidies in relation to quality at the cleavage stage.

Chromosome Good Average Poor Chromosome Good Average Poor


constitution quality (%) quality (%) quality (%) constitution quality (%) quality (%) quality (%)
........................................................................................ ........................................................................................
Normal 48 (17.6) 8 (13) 2 (9.5) Normal 48 (83) 8 (14) 2 (3)
Aneuploid 225 (82.4) 53 (87) 19 (90.5) Viable single trisomies 13 (72) 5 (28) 0
Aneuploid/1 error 63 (23) 15 (24.5) 8 (38) Non-viable single 72 (72) 18 (18) 10 (10)
Aneuploid/2 errors 43 (16) 13 (21.3) 4 (19) abnormalities
Complex aneuploid 119 (43.4) 25 (40.9) 7 (33)

Downloaded from http://molehr.oxfordjournals.org/ at Aston University on January 16, 2014


Figure 2 Comparison of morphology/quality among euploid cleav-
age stage embryos, aneuploid ones carrying single possibly viable triso-
Figure 1 Embryo morphology/quality and chromosome status at the mies, and those aneuploid ones affected by single non-viable errors.
cleavage stage. Statistical analysis of the data indicated that morphological scores
were no different, regardless of chromosome status.

7.9% of errors (69) involved segmental abnormalities. Similar to the


Table IV Embryo sex and speed of development at the
cleavage stage, 449 (52%) of all abnormalities were chromosome
cleavage stage.
losses and 421 (48%) were gains.
Embryo sex 5 cells or 6– 9 cells (%) 10 cells or The incidence of aneuploidy at the blastocyst stage was also evaluated
less (%) more (%) in relation to female age. A total of 286 embryos were generated by 44
........................................................................................ reproductively younger women (average age: 33.5 years, age range: 29 –
Female 14 (8) 143 (80.7) 20 (11.3) 36 years). Among them, 156 (55%) were characterized as normal,
Male 11 (8.8) 107 (85.6) 7 (5.6) whereas chromosome malsegregation errors were scored for 130
(45%). Five hundred and seventy-two blastocysts were generated by
117 reproductively older women (average age: 40.5 years, range:
embryo was euploid, affected by a type of aneuploidy capable of preg- 37– 47 years). Fewer blastocysts were euploid (217, 38%) for this
nancy formation, or carrying a single aneuploidy of a type never seen patient group, with the majority being chromosomally abnormal (355,
during pregnancy (P ¼ 0.14 and P ¼ 0.6, respectively). This set of data 62%). As with the cleavage stage, the difference seen in aneuploidy
is shown in Table V and Fig. 2. rates between blastocysts from younger and older women was highly
statistically significant (P ¼ 0.0016).
The incidence of aneuploidy at the blastocyst
stage The effect of aneuploidy on embryo
A total of 858 blastocysts were also assessed. A normal chromosome development to the blastocyst stage
complement was identified for 373 (44%) of them, while abnormalities A total of 122 (14.2%) of the embryos assessed at the blastocyst stage
were detected for the remaining 485 (56%). Unlike the cleavage stage, were well developed (Grade 5 or 6) on Day 5, 642 (74.8%) were consid-
only a small minority of aneuploid blastocysts (79, 16%) were character- ered to be average (Grade 3 or 4) and 94 (11%) were early (i.e. blastocoel
ized as complex abnormal. 59% (284) were affected by a single chromo- cavity not fully expanded, Grade 1 or 2). The morphological criteria used
some error and 25% (122) had two errors. A total of 870 abnormalities to classify these embryos, according to the extent of development and
were detected. Most of these (801) affected entire chromosomes, but degree of blastocoel expansion, are summarized in Table II. Cytogenetic
Aneuploidy and embryo morphology 121

analysis showed that 62/122 (51%) of the fully expanded (Grade 5 or 6) greater chance of being euploid than those of poorer grades. Specifically,
blastocysts were chromosomally normal, whereas only 33% (31/94) of 137/337 (41%) of normal blastocysts had a Grade A TE, compared with
the early (Grade 1 or 2) blastocysts were characterized as euploid. It was 115/406 (28%) of those with errors, and this difference was statistically
therefore evident that euploid embryos tend to show a faster progres- significant (P ¼ 0.002). Additionally, decreasing TE quality was asso-
sion to the most advanced expansion stages when compared with aneu- ciated with an increase in the incidence of complex aneuploidy (ratio
ploid embryos (P ¼ 0.012). euploid:complex: Grade A, 15:1; Grade B, 4:1; Grade C, 4:1). A more
Interestingly, the incidence of multiple errors was twice as high in the detailed description of the ICM and TE data is shown in Tables VII
early blastocysts (affecting 14%), compared with those that had achieved and VIII.
full expansion (6%). A summary of these data is shown in Table VI and is Of the 795 blastocysts with a normal sex chromosome complement,
illustrated in Fig. 3. males and females were observed in equal incidence; 392 (49.3%) were
Morphologic evaluation of the ICM revealed that it tended to be male and 403 (50.7%) were female. However, a small excess of male
affected by the presence of chromosome errors in the embryo, although embryos over females (56 versus 44%, respectively) was observed for
this did not reach statistical significance (P ¼ 0.3). A tightly packed ICM the most developed blastocysts (Grades 5 and 6). On the contrary, in
with many cells (Grade A) was observed in 51% (172 out of 337 with ICM the case of slower growing (Grade ≤ 3) blastocysts, female embryos
grading) of normal embryos compared with 45% (181 out of 406 with predominated over males (52 versus 48%). It should be noted that
ICM grading) of aneuploid embryos. The data obtained also indicated neither of these gender differences were statistically significant (P ¼

Downloaded from http://molehr.oxfordjournals.org/ at Aston University on January 16, 2014


that ICM quality was highly dependent on the number of chromosome 0.15 and 0.9, respectively). No correlation was seen between the
abnormalities, with the ratio of euploid:complex aneuploid decreasing gender of the blastocyst and the incidence of aneuploidy. A summary
with declining ICM quality (Grade A, 11:1; Grade B, 4:1; Grade C, of these data is shown in Table IX.
3:1). In other words, complex aneuploid blastocysts generally had The morphology and growth rate of 87 aneuploid blastocysts carrying
ICMs of poorer morphology, compared with those of euploid embryos. varieties of abnormality with the greatest potential for implantation and
Similar to the ICM observations, blastocysts achieving the best quality clinical pregnancy (i.e. trisomies 13, 16, 18, 21 and 22, and sex chromo-
TE scores (Grade A—many cells forming a cohesive epithelium) had a some abnormalities) were compared with those of the 373 euploid blas-
tocysts and with 190 embryos carrying a single chromosome error not
capable of leading to clinically recognized pregnancies (e.g. affecting
larger chromosomes). Among the faster growing embryos (expansion
Table VI The chromosome constitution of embryos in Grades 5 and 6), blastocysts carrying an aneuploidy compatible with
relation to expansion at the blastocyst stage.

Chromosome Grades 5 and Grades 3 and Early


constitution 6 (%) 4 (%) (%) Table VII The chromosome constitution of embryos in
........................................................................................
Normal 62 (51) 280 (44) 31 (33)
relation to ICM quality.
Aneuploid 60 (49) 362 (56) 63 (67) Chromosome ICM ICM ICM Total
Aneuploid/1 error 43 (35) 208 (32) 34 (36) constitution Grade A Grade B Grade C
Aneuploid/2 errors 10 (8) 95 (15) 16 (17 (%) (%) (%)
........................................................................................
Complex aneuploid 7 (6) 59 (9) 13 (14) Normal 172 (51) 149 (44) 16 (5) 337 (100)
Aneuploid 181 (45) 202 (50) 23 (5) 406 (100)
Aneuploid/1 error 116 (29) 114 (28) 13 (3)
Aneuploid/2 errors 49 (12) 48 (12) 5 (1)
Complex aneuploid 16 (4) 40 (10) 5 (1)

Table VIII The chromosome constitution of embryos in


relation to TE quality.

Chromosome TE TE TE Total
constitution Grade A Grade B Grade C
(%) (%) (%)
........................................................................................
Normal 137 (41) 169 (50) 31 (9) 337 (100)
Aneuploid 115 (28) 242 (60) 49 (12) 406 (100)
Aneuploid/1 error 82 (20) 132 (33) 25 (6)
Figure 3 Rate of expansion and the chromosome status of blastocyst Aneuploid/2 errors 24 (6) 62 (15) 17 (4)
stage embryos. The incidence of multiple errors was twice as high in
Complex aneuploid 9 (2) 48 (12) 7 (2)
early blastocysts, compared with those that had achieved full expansion.
122 Fragouli et al.

clinical pregnancy occurred at approximately the same frequency as


those that were chromosomally normal (19.5 and 17%, respectively). Table XII The incidence of viable and non-viable
aneuploidies in relation to ICM morphology blastocyst
Aneuploid embryos affected by a single anomaly, incompatible with a
stage.
clinical pregnancy, formed blastocysts of top morphological quality at a
lower rate (13%), although the difference was not statistically significant. Chromosome constitution Grades A Grade
When the morphology of the ICM and the TE were compared for the and B (%) C/early (%)
three groups of embryos, it was evident that blastocysts carrying ........................................................................................
Normal 306 (82) 67 (18)
errors which could lead to clinical pregnancies were indistinguishable
from euploid blastocysts (ICM A/B euploid versus aneuploid, 86 Viable single trisomies 72 (83) 15 (17)
versus 90%, respectively; ICM C/early, 14 versus 10%; TE A/B, 82 Non-viable single abnormalities 138 (72) 52 (28)
versus 83%; TE C/early, 18 versus 17%). However, embryos with
chromosome abnormalities incompatible with clinical pregnancy gener-
ally had ICM and TE of poorer quality, compared with the other two
groups (ICM A/B, 76%; ICM C/early, 24%; TE A/B, 72%; TE C/early,
Discussion
28%). These differences were statistically significant (P ¼ 0.003 for During the first few days of life, the genome is essentially silent, the
ICM and P ¼ 0.01 for TE). These results are shown in Tables X, XI embryo surviving on stored mRNA and proteins derived from the

Downloaded from http://molehr.oxfordjournals.org/ at Aston University on January 16, 2014


and XII and Fig. 4. oocyte, and consequently genetic abnormalities may remain hidden.
Once the resources provided by the oocyte become depleted, the
embryo must begin expressing its own genes if it is to progress any
further. Genome activation usually takes place when the embryo consists
Table IX Embryo sex and expansion rate at the of 4– 8 cells (Braude et al., 1988). As the embryonic genome begins to
blastocyst stage. take control, the effects of any genetic abnormalities are likely to be felt
Embryo sex Grades 5 Grades 3 Early (%) Total increasingly strongly and may potentially start to influence aspects of
and 6 (%) and 4 (%) embryo morphology. In theory, key developmental processes occurring
........................................................................................ after genome activation, such as compaction, cellular differentiation
Female 51 (6) 308 (39) 44 (5.5) 403 (50.5) leading to formation of the ICM and TE and coordination of cells to
Male 64 (8) 284 (36) 44 (5.5) 392 (49.5) form an ordered blastocyst, could be perturbed by the presence of
Total 115 (14) 591 (75) 88 (11) 795 (100) genetic abnormalities. For this reason a morphological analysis after
genome activation at later preimplantation stages (e.g. blastocysts)
may be particularly valuable.
A previous study, examining the effect of chromosome abnormalities
on preimplantation development, postulated that aneuploid or mosaic
embryos have an increased likelihood of undergoing developmental
Table X The incidence of viable and non-viable
arrest before compaction occurs (Rubio et al., 2007). Recent investiga-
aneuploidies in relation to expansion rate at the
blastocyst stage. tions by our group and others have confirmed that the incidence of
chromosome abnormalities does decline during progression to the
Chromosome Grades 5 Grades 3 Early Total blastocyst stage, most likely due to demise of aneuploid embryos (Fra-
constitution and 6 (%) and 4 (%) (%) gouli et al., 2013). However, it is clear that only a minority of chromoso-
........................................................................................
mally abnormal embryos suffer this fate (reviewed in Fragouli and Wells,
Normal 62 (17) 280 (75) 31 (8) 373 (100)
2011). In the current study, the aneuploidy rate dropped significantly
Viable single trisomies 17 (19.5) 66 (76) 4 (4.5) 87 (100)
from 84% at the cleavage stage to 56% at the blastocyst stage (P ,
Non-viable single 25 (13) 143 (75) 22 (12) 190 (100)
0.0001). It should be noted that the well-documented effect of maternal
abnormalities
age on aneuploidy rates was clearly evident for both of the developmen-
tal stages examined, with abnormality rates steeply increasing in embryos
generated by older women. The effect of age was particularly apparent
among blastocysts, with aneuploidy increasing from 45% for an average
maternal age of 33.5 to 62% at an average age of 40.5 years (P ¼ 0.0016).
Table XI The incidence of viable and non-viable
At the cleavage stage, aneuploidy rates were remarkably high, even for
aneuploidies in relation to ICM morphology blastocyst
reproductively younger women (73% for average female age 33.6 years)
stage.
and reached 87% for the older group of patients (averaging 41 years of
Chromosome constitution Grades A Grade age). High aneuploidy rates have previously been observed in other
and B (%) C/early (%) investigations involving comprehensive cytogenetic analysis of cleavage
........................................................................................
stage embryos (Wells and Delhanty, 2000; Voullaire et al., 2000; North-
Normal 321 (86) 52 (14)
rop et al., 2010; Gutierrez-Mateo et al., 2011; Fragouli et al., 2013; Mert-
Viable single trisomies 78 (90) 9 (10)
zanidou et al., 2013). Considering that chromosome abnormalities occur
Non-viable single abnormalities 144 (76) 46 (24)
relatively infrequently in sperm (affecting 5%) (Carrell et al., 2003; Petit
et al., 2005), and as the possible false-positive rate for the aCGH
Aneuploidy and embryo morphology 123

methodology does not exceed 2.7% (Mir et al., 2013), the high preva- having good morphology almost as often as those that were chromoso-
lence of aneuploidy at the cleavage stage is likely due to a combination mally normal (72 versus 83% at the cleavage stage). This highlights the fact
of oocyte and post-zygotic malsegregation errors (Fragouli et al., that the type of embryos preferred for transfer to the uterus based upon
2013). In the younger patient group, between one-fifth and one-third their morphological appearance are often abnormal. The remarkable
of oocytes would be expected to harbour a meiotic chromosome abnor- ability of cleavage stage embryos to tolerate aneuploidy was further
mality and thus approximately half of all of the aneuploidies seen at the emphasized by the fact that more than half of the complex abnormal
cleavage stage are likely to be of mitotic origin. This finding is concordant embryos (having three or more chromosome errors) also achieved
with numerous studies reporting a high rate of chromosomal mosaicism good morphological grades. Only embryos of the very poorest morph-
amongst human cleavage stage embryos, indicative of mitotic errors oc- ology had any noticeable difference in the aneuploidy rate (P ¼
curring during the first few divisions following fertilization. Various inves- 0.0019). However, poor quality embryos were sometimes found to be
tigations using FISH to examine aneuploidy and morphology at the euploid, demonstrating that even at the extreme end of the morphologic-
cleavage stage have suggested a weak correlation between the two al scale the association between microscopic appearance and chromo-
(e.g. Hardarson et al., 2003; Eaton et al., 2009). However, such studies somal status remains weak.
were incapable of categorically defining embryos as euploid or chromo- Blastocyst morphology showed a slightly stronger link to aneuploidy.
somally abnormal due to the limitations of the FISH method. The current Approximately 50% of embryos with fully expanded blastocoel cavities
study employed molecular cytogenetic methodology. This meant that were chromosomally normal, compared with about one-third of early

Downloaded from http://molehr.oxfordjournals.org/ at Aston University on January 16, 2014


we were capable not only of examining the entire chromosome comple- blastocysts. The presence of chromosome abnormalities not only
ment, but distinguishing between normal and aneuploid embryos with affected the rate of expansion on Day 5, but the quality of the ICM and
much more certainty. Yet despite the improved methodology, the TE as well. Both these components of the blastocyst showed a declining
results obtained were broadly in agreement with the earlier studies. An- quality in relation to aneuploidy, although this was much more evident for
euploidy was found to have little effect on the morphology of the majority the TE. These findings are in agreement with our previous investigation
of cleavage stage embryos. Hence, aneuploid embryos were classified as (Alfarawati et al., 2011) and confirm the adverse effect of chromosome

Figure 4 (a). Comparison of expansion rates among euploid blastocysts, aneuploid ones carrying possibly viable single trisomies and those carrying single
non-viable errors. (b). Comparison of ICM morphology among euploid blastocysts, aneuploid ones carrying possibly viable single trisomies and those carry-
ing single non-viable errors. (c). Comparison of ICM morphology among euploid blastocysts, aneuploid ones carrying possibly viable single trisomies and
those carrying single non-viable errors. No morphological differences were observed between euploid blastocysts and those carrying possibly viable single
trisomies. However, blastocysts carrying single non-viable errors were as a whole of poorer quality, compared with the other two groups. These differences
were statistically significant for ICM (P ¼ 0.003) and TE (P ¼ 0.01).
124 Fragouli et al.

errors especially after activation of the embryonic genome. A direct re- grading, and it seems likely that they may be difficult if not impossible
lationship between TE quality and clinical outcome has been described to detect using time-lapse methods.
recently (Honnma et al., 2012). Specifically, blastocysts with TE of An interesting observation was that at the cleavage stage, females
Grades B and C were at a higher risk of not implanting or miscarrying, were overrepresented among the fastest growing embryos (10 or
compared with those with TE of Grade A. Our findings suggest more cells), while at the blastocyst stage the trend was reversed, with
that increased aneuploidy risk may be the underlying reason for this ob- a small excess of male embryos amongst the most expanded blastocysts.
servation. The reason behind these sex-related differences at the two developmen-
It has been shown previously that the timing of cleavage divisions is of tal stages is not clear at this time. Data obtained during a study carried out
crucial importance and that faster development could be associated with by Burgoyne (1993) suggested that the presence of a Y chromosome
aneuploidy (Magli et al., 2007). Recently, new systems have been devel- could have a positive influence on development due to the expression
oped which enable a detailed analysis of embryonic cleavage divisions of growth factor genes taking place after embryonic genome activation.
with the use of time-lapse imaging. Careful measurement of morphoki- It has also been postulated that the process of X chromosome inactiva-
netic data has allowed the creation of novel algorithms predictive of de- tion occurring in female embryos could influence their growth relative to
velopment to the blastocyst stage, behaviour after biopsy and clinical males (Van den Berg et al., 2009). Differences in developmental rates
outcome (Meseguer et al., 2011; Cruz et al., 2012; Kirkegaard et al., related to the sex of the embryo may also have important implications
2012; Chavez et al., 2013; Wong et al., 2013). on the assessment of embryo competence based upon timing of the

Downloaded from http://molehr.oxfordjournals.org/ at Aston University on January 16, 2014


An important question is whether or not time-lapse systems will be phases of the cell cycle, potentially complicating time-lapse analysis.
capable of providing an insight into aneuploidy. Can the soft association In summary, the use of methods providing a comprehensive analysis of
between static morphological scoring and aneuploidy be hardened into a chromosomes at the cleavage and blastocyst stages has confirmed that
clinically useful diagnostic via the development of specially designed algo- aneuploidy has relatively little impact on morphology. Generally, only
rithms? A recent study carried out by Campbell et al. (2013) defined a the most severe forms of aneuploidy (multiple and/or large gene-rich
time-lapse algorithm which may be able to identify some types of chromosomes) have a significant influence on the microscopic appear-
chromosome abnormalities. However, the results on time-lapse detec- ance of preimplantation embryos. Transferring cleavage or blastocyst
tion of aneuploidy were not definitive with 30% of the embryos consid- stage embryos of high morphological quality can lead to a slight reduction
ered to have ‘low’ aneuploidy risk subsequently found to be abnormal in aneuploidy risk, helping to avoid complex abnormal embryos in par-
using aCGH and 30% of embryos with ‘medium’ risk identified as ticular. However, the benefits are relatively small and morphological se-
euploid (Campbell et al., 2013). In virtually all somatic cells, errors that lection is completely ineffective for avoiding the types of aneuploidy
have the potential to produce aneuploidy, such as misalignment of chro- capable of producing a clinical pregnancy.
mosomes at metaphase or abnormalities of the mitotic spindle, induce a It is possible that in the future time-lapse methods of embryo evalu-
delay to the cell cycle. Since cleavage stage embryos often fail to segre- ation may lead to enhanced detection of aneuploidy during preimplanta-
gate their chromosomes correctly during mitosis, leading to chromo- tion development. However, this remains to be proven and it seems
somal mosaicism, alterations in the duration of specific phases of the likely that many forms of meiotic abnormality will remain invisible to
cell cycle might be expected in some cases. It seems particularly likely such analyses. Currently, the only reliable way to avoid the transfer of an-
that complex abnormal embryos, which are usually highly mosaic, euploid embryos remains biopsy, ideally performed at the blastocyst
often suffering from chaotic division of chromosomes, would suffer sig- stage, followed by analysis with the use of comprehensive chromosome
nificant distortions to the cell cycle. However, it is not clear whether screening methods. The clinical application of such methods for the pur-
time-lapse analysis could identify embryos with more subtle abnormal- poses of PGS has recently led to promising clinical results in randomized
ities (e.g. single aneuploidies), especially those arising during meiosis. controlled trials (Yang et al., 2012; Scott et al., 2013; Forman et al., 2013).
Meiotic aneuploidies, involving single chromosomes, are the most
common type of anomaly seen in miscarriages and prenatal samples
(reviewed in Nagaoka et al., 2012) and may therefore be considered of Acknowledgements
the greatest clinical importance, yet the abnormal cell divisions that The authors would like to thank the embryologists of the referring IVF
produce such errors occur prior to the time when most time-lapse eva- clinics for providing details about their scoring criteria.
luations have been undertaken.
While complex abnormal embryos were confirmed to have a greatly
reduced probability of reaching the most expanded blastocyst stages Authors’ roles
compared with euploid ones (P , 0.0001), embryos affected by aneu-
ploidies of chromosomes 13, 16, 18, 21 and 22 were morphologically in- E.F. designed the study, carried out experimental work, analysed scien-
distinguishable from those of normal embryos, during both examined tific data and wrote manuscript. S.A. carried out experimental work
stages. It seems that these particular chromosome abnormalities are and analysed scientific data. K.S. carried our experimental work. D.W.
not selected against during preimplantation development. This finding designed the study, analysed scientific data and co-wrote the manuscript.
is consistent with the fact that conceptions with aneuploidy affecting
one of these chromosomes are often capable of implanting and produ-
cing clinical pregnancies, usually ending in miscarriage (Menasha et al.,
Funding
2005). Aneuploidies compatible with implantation do not seem to Dagan Wells is supported by the NIHR Biomedical Research Centre
disrupt early development, have no effect on standard morphological Oxford.
Aneuploidy and embryo morphology 125

Gutierrez-Mateo C, Colls P, Sanchez-Garcıa J, Escudero T, Prates R,


Conflict of interest Ketterson K, Wells D, Munne S. Validation of microarray comparative
None declared. genomic hybridization for comprehensive chromosome analysis of
embryos. Fertil Steril 2011;95:953 – 958.
Hardarson T, Caisander G, Sjogren A, Hanson C, Hamberger L, Lundin K. A
morphological and chromosomal study of blastocyst developing from
References morphologically suboptimal pre-embryos compared with control
Alfarawati S, Fragouli E, Colls P, Stevens J, Gutiérrez-Mateo C, blastocyst. Hum Reprod 2003;18:399 – 407.
Schoolcraft WB, Katz-Jaffe MG, Wells D. The relationship between Honnma H, Baba T, Sasaki M, Hashiba Y, Ohno H, Fukunaga T, Endo T,
blastocyst morphology, chromosomal abnormality, and embryo gender. Saito T, Asada Y. Trophectoderm morphology significantly affects the
Fertil Steril 2011;95:520 –524. rates of ongoing pregnancy and miscarriage in frozen-thawed
Braude P, Bolton V, Moore S. Human gene expression first occurs between single-blastocyst transfer cycle in vitro fertilization. Fertil Steril 2012;
the four- and eight-cell qstages of preimplantation development. Nature 98:361– 367.
1998;332:459 – 461. Kirkegaard K, Hindkjaer JJ, Ingerslev HJ. Human embryonic development
Bromer JG, Seli E. Assessment of embryo viability in assisted reproductive after blastomere removal: a time-lapse analysis. Hum Reprod. 2012;
technology: shortcomings of current approaches and the emerging role 27:97– 105.
of metabolomics. Curr Opin Obstet Gynecol 2008;20:234 – 241. Magli MC, Gianaroli L, Munne S, Ferraretti P. Incidence of chromosomal

Downloaded from http://molehr.oxfordjournals.org/ at Aston University on January 16, 2014


Burgoyne PS. A Y-chromosomal effect on blastocyst cell number in mice. abnormalities from a morphologically normal cohort of embryos in
Development 1993;117:341– 345. poor-prognosis patients. JARG 1998;15:297 – 301.
Campbell A, Fishel S, Bowman N, Duffy S, Sedler M, Hickman CF. Modelling a Magli MC, Gianaroli L, Ferraretti AP, Lappi M, Ruberti A, Farfalli V. Embryo
risk classification of aneuploidy in human embryos using non-invasive morphology and development are dependent on the chromosomal
moprhokinetics. RBM Online 2013;26:477 – 485. complement. Fertil Steril 2007;87:534 – 541.
Carrell DT, Wilcox AL, Lowy L, Peterson CM, Jones KP, Erickson L, Menasha J, Levy B, Hirschhorn K, Kardon NB. Incidence and spectrum of
Campbell B, Branch DW, Hatasaka HH. Elevated sperm chromosome chromosome abnormalities in spontaneous abortions: new insights from
aneuploidy and apoptosis in patients with unexplained recurrent a 12-year study. Genet Med 2005;7:251 – 263.
pregnancy loss. Obstet Gynecol 2003;101:1229– 1235. Mertzanidou A, Spits C, Nguyen HT, Van de Velde H, Sermon K. Evolution of
Chavez SL, Loewke KE, Han J, Moussavi F, Colls P, Munne S, Behr B, Reijo aneuploidy up to Day 4 of human preimplantation development. Hum
Pera RA. Dynamic blastomere behaviour reflects human embryo ploidy Reprod 2013;28:1716 – 1724.
by the four-cell stage. Nat Commun. 2012;3:1251. Meseguer M, Herrero J, Tejera A, Hilligsøe KM, Ramsing NB, Remohı́ J. The
Cruz M, Garrido N, Herrero J, Pérez-Cano I, Muñoz M, Meseguer M. Timing use of morphokinetics as a predictor of embryo implantation. Hum Reprod
of cell division in human cleavage-stage embryos is linked with blastocyst 2011;26:2658 – 2671.
formation and quality. RBM Online 2012;25:371– 381. Mir P, Rodrigo L, Mercader A, Buendı́a P, Mateu E, Milán-Sánchez M,
Cummins JM, Breen TM, Harrison KL, Shaw JM, Wilson LM, Hennessey JF. A Peinado V, Pellicer A, Remohı́ J, Simón C et al. False positive rate of an
formula for scoring human embryo growth rates in in vitro fertilization: its arrayCGH platform for single-cell preimplantation genetic screening and
value in predicting pregnancy and in comparison with visual estimates of subsequent clinical application on day-3. J Assist Reprod Genet 2013;
embryo quality. J In Vitro Fert Embryo Transf 1986;3:284 – 295. 30:143– 149.
Eaton JL, Hacker MR, Harris D, Thornton KL, Penzias AS. Assessment of Moayeri SE, Allen RB, Brewster WR, Kim MH, Porto M, Werlin LB. Day-3
day-3 morphology and euploidy for individual chromosomes in embryos embryo morphology predicts euploidy among older subjects. Fertil Steril
that develop to the blastocyst stage. Fertil Steril 2009;91:2432 – 2436. 2008;89:118 – 123.
Forman EJ, Hong KH, Ferry KM, Tao X, Taylor D, Levy B, Treff NR, Scott RT Munne S. Chromosomal abnormalities and their relationship to morphology
Jr. In vitro fertilisation with single euploid blastocyst transfer: a randomised and development of human embryos. RBM Online 2006;12:234 – 253.
controlled trial. Fertil Steril 2013;100:100 – 107. Nagaoka SI, Hassold TJ, Hunt PA. Human aneuploidy: mechanisms and new
Fragouli E, Wells D. Aneuploidy in the human blastocyst. Cytogenet Genome insights into an age-old problem. Nat Rev Genet 2012;13:493 –504.
Res. 2011;133:149 – 159. Northrop LE, Treff NR, Levy B, Scott RT. SNP Microarray based 24
Fragouli E, Katz-Jaffe M, Alfarawati S, Stevens J, Colls P, Goodall NN, chromosome aneuploidy screening demonstrates that cleavage stage
Tormasi S, Gutierrez-Mateo C, Prates R, Schoolcraft WB et al. FISH poorly predicts aneuploidy in embryos that develop to
Comprehensive chromosome screening of polar bodies and blastocysts morphologically normal blastocysts. Mol Hum Reprod 2010;16:590– 600.
from couples experiencing repeated implantation failure. Fertil Steril Petit FM, Frydman N, Benkhalifa M, Le Du A, Aboura A, Fanchin R,
2010;94:875 – 887. Frydman R, Tachdjian G. Could sperm aneuploidy rate determination be
Fragouli E, Alfarawati S, Daphnis DD, Goodall NN, Mania A, Griffiths T, used as a predictive test before intracytoplasmic sperm injection?.
Gordon A, Wells D. Cytogenetic analysis of human blastocysts with the J Androl 2005;26:235– 241.
use of FISH, CGH and aCGH: scientific data and technical evaluation. Puissant F, Van Rysselberge M, Barlow P, Deweze J, Leroy F. Embryo
Hum Reprod 2011;26:480 – 490. scoring as a prognostic tool in IVF treatment. Hum Reprod 1987;
Fragouli E, Alfarawati S, Spath K, Jaroudi S, Sarasa J, Enciso M, Wells D. The 2:705 – 708.
origin and impact of embryonic aneuploidy. Hum Genet 2013; Rubio C, Rodrigo L, Mercader A, Mateu E, Buendı́a P, Pehlivan T, Viloria T, De
132:1001– 1013. los Santos MJ, Simón C, Remohı́ J et al. Impact of chromosomal
Gardner DK, Schoolcraft WB. In vitro culture of human blastocysts. In: abnormalities on preimplantation embryo development. Prenat Diagn
Jansen R, Mortimer D (eds). Toward Reproductive Certainty: Fertility and 2007;27:748 – 756.
Genetics Beyond. Carnforth, UK: Parthenon Publishing, 1999, 378 –388. Rubio C, Rodrigo L, Mir P, Mateu E, Peinado V, Milán M, Al-Asmar N,
Gianaroli L, Magli MC, Ferraretti AP, Lappi M, Borghi E, Ermini B. Oocyte Campos-Galindo I, Garcia S, Simón C. Use of array comparative
euploidy, pronuclear zygote morphology and embryo chromosomal genomic hybridization (array-CGH) for embryo assessment: clinical
complement. Hum Reprod 2007;22:241 – 249. results. Fertil Steril 2013;99:1044– 1048.
126 Fragouli et al.

Scott RT Jr, Ferry K, Su J, Tao X, Scott K, Treff NR. Comprehensive Van Royen E, Mangelschots K, De Neubourg D, Valkenburg M, Van de
chromosome screening is highly predictive of the reproductive potential Meerssche M, Ryckaert G, Eestermans W, Gerris J. Characterization of
of human embryos: a prospective, blinded, nonselection study. Fertil a top quality embryo, a step toward single-embryo transfer. Hum Reprod
Steril 2012;97:870 – 875. 1999;14:2345 – 2349.
Scott RT Jr, Upham KM, Forman EJ, Hong KH, Scott KL, Taylor D, Tao X, Voullaire L, Slater H, Williamson R, Wilton L. Chromosome analysis of
Treff NR. Blastocyst biopsy with comprehensive chromosome screening blastomeres from human embryos by using comparative genomic
and fresh embryo transfer significantly increases in vitro fertilization hybridization. Hum Genet 2000;106:210– 217.
implantation and delivery rates: a randomized controlled trial. Fertil Steril Wells D, Delhanty JD. Comprehensive chromosomal analysis of human
2013;100:697 – 703. preimplantation embryos using whole genome amplification and single cell
Stensen MH, Tanbo T, Storeng R, Byholm T, Fèdorcsak P. Routine comparative genomic hybridization. Mol Hum Reprod 2000;6:1055–1062.
morphological scoring systems in assisted reproduction treatment fail to Wong C, Chen AA, Behr B, Shen S. Time-lapse microscopy and image
reflect age-related impairment of oocyte and embryo quality. RBM analysis in basic and clinical embryo development research. RBM Online
Online 2010;21:118 – 125. 2013;26:120 – 129.
Van den Berg IM, Laven JS, Stevens M, Jonkers I, Galjaard RJ, Gribnau J, Yang Z, Liu J, Collins GS, Salem SA, Liu X, Lyle SS, Peck AC, Sills ES, Salem RD.
van Doorninck JH. X chromosome inactivation is initiated in Selection of single blastocysts for fresh transfer via standard morphology
human preimplantation embryos. Am J Hum Genet. 2009;84: assessment alone and with array CGH for good prognosis IVF patients:
771 – 779. results from a randomized pilot study. Molecular Cytogenetics 2012;5:24.

Downloaded from http://molehr.oxfordjournals.org/ at Aston University on January 16, 2014