Aneuploidy Screening for Embryo Selection

Elpida Fragouli, Ph.D.; Dagan Wells, Ph.D.


Semin Reprod Med. 2012;30(4):289-301. 

In This Article

Chromosome Screening At the Cleavage Stage of Preimplantation Development

The cleavage stage of preimplantation development begins 1 day after oocyte and sperm syngamy when the newly formed zygote commits itself to a succession of mitotic divisions. These cleavage divisions lead to the formation of small spherical cells, known as blastomeres. During the first few mitotic divisions, that is, until day 3, the blastomeres are considered to be totipotent, but this capacity is progressively lost as the divisions become asynchronous. Perhaps the most important event that characterizes this period of development is the transition of control from the maternal genome to the embryonic, the latter of which is activated between the four- and eight-cell stages.

Numerous studies using FISH to examine the chromosome content of preimplantation embryos have shown that in addition to meiotically derived aneuploidy, mostly originating from the oocyte, the cleavage stage is characterized by a large number of postzygotic chromosome segregation errors.[7,11,20] Mitotic chromosome mal-segregation leads to a phenomenon called "mosaicism," which effectively can be defined as the presence of two or more karyotypically distinct cell lines within the same embryo. The degree of mosaicism tends to vary, with a minority of embryos displaying a mixture of normal and aneuploid cell lines, whereas most mosaic embryos are found to contain two or more abnormal cell lineages and an absence of any euploid cells. Some embryos appear to have suffered a total failure of the cellular mechanisms that usually ensure accurate chromosome segregation, becoming karyotypically unstable and highly mosaic, with abnormalities affecting multiple different chromosomes in each of their cells. Embryos of this type are termed chaotic mosaics.[7]

The true frequency and extent of chromosomal mosaicism during the cleavage stage of preimplantation development, the stage at which most PGS has traditionally been performed, became clear with the application of CGH. Two key studies used CGH to examine the copy number of every chromosome in all of the cells from 24 embryos.[8,21] Both investigations reported that approximately two thirds of the embryos assessed were mosaic. In most cases, the mosaic embryos were abnormal in every cell (51% of embryos tested contained no euploid cells). However, 24% contained a mixture of abnormal and normal cells, suggesting a potential problem for diagnosis based on analysis of a single cell. Only a quarter of the embryos tested were composed solely of normal cells.[8,21] Although the existence of cleavage-stage mosaicism has been acknowledged for 15 years, the vast majority of PGS cases have continued to be conducted on day 3. The reasons for the continued application of chromosome screening at the cleavage stage may be due to most embryologists having little experience with embryo biopsy at other stages, limiting exploration of alternatives, and due to publications providing reassurance that, despite the presence of mosaicism, the proportion of embryos misdiagnosed is low (~5%).[13]

Until recently, laboratories offering PGS services were only able to score a restricted number of chromosomes (between 5 and 12) in each embryo. This was due to the fact that only five distinct colors (fluorochromes) in the visible spectrum were available for probe labeling, and consequently the number of chromosomes that could be simultaneously assessed was limited to five. The selection of the probe combination used and set of chromosomes screened was based on data obtained from the cytogenetic analysis of spontaneous abortions and abnormal live births. Specifically, it was shown that 72% of abnormalities leading to spontaneous abortions affect chromosomes X, Y, 13, 16, 18, 21, and 22.[22] Hence most PGS FISH protocols examined these seven chromosomes, and some others including 15, 17, and 19, in sequential hybridization rounds performed on the same blastomere nucleus.[23]

Early data obtained with use of FISH protocols that examined up to 12 chromosomes in blastomeres generated by women of ARA (≥37 years) and/or recurrent miscarriages were encouraging, as far as the implantation and pregnancy rates were concerned.[24–29] However, no clear benefits were observed in cases of younger women or for couples who were being treated due to RIF. Moreover, most of these studies were either descriptive or observational, they examined a variable number of chromosomes, and the day of embryo transfer differed. Most importantly, although some studies were prospective, none took place in a randomized controlled manner.

Staessen and colleagues[30] were the first to report results after carrying out a randomized controlled trial (RCT) that compared clinical outcomes after blastocyst transfer with or without PGS. The investigated patient group included women of ARA (>37 years), two blastomeres from each embryo were biopsied, and seven chromosomes (13, 16, 18, 21, 22, X, and Y) were screened. The reported clinical outcomes included an implantation rate of 17.1% and a clinical pregnancy rate of 16.5% after PGS, compared with 11.5% and 10.4%, respectively, for the control group. Although PGS appeared to produce superior results, the study had insufficient statistical power to determine whether this was a truly significant difference. The conclusion was that PGS made no difference to outcome. As well as statistical considerations, it is also worth contemplating whether the relatively small number of chromosomes assessed or the decision to biopsy two cells could have reduced the efficacy of PGS. Although the analysis of two cells can potentially lead to small improvements in the accuracy, there is experimental evidence that the removal of such a large part of the embryo's volume has a detrimental effect on subsequent implantation.[31,32] Indeed, a more recent RCT performed by the same group proved that the biopsy of a single blastomere was associated with superior clinical outcome compared with the biopsy of two cells, suggesting that the earlier PGS study might have seen better outcomes if a single cell had been biopsied.[33]

The debate about the clinical efficacy of PGS using FISH gathered momentum with the publication of a multicenter double-blind RCT performed by Mastenbroek and coworkers.[34] The investigated patient group was again women of ARA, although on this occasion the threshold was set lower, at 35 years of age (range: 35 to 41 years). A single blastomere biopsy took place, removing concerns about the impact of two-cell biopsy, and a total of eight chromosomes (1, 13, 16, 17, 18, 21, X, and Y) were examined. Far from improving IVF success rates, this study found that PGS had a detrimental effect on the likelihood of establishing a clinical pregnancy, with an ongoing pregnancy rate of 25% seen in the test group, compared with 37% seen for the control group. Not surprisingly, the findings initially caused great alarm. However, subsequent in-depth analyses of the data and methodology used have led to heavy criticism of the study on multiple levels. Inappropriate patient selection, lack of embryo biopsy experience, incorrect chromosome selection, and an unusually high "no result" rate have come in for particular criticism.[31,35–37]

Although some have argued that the Mastenbroek study was so fundamentally flawed that no useful information on the efficacy of PGS could be deduced from it, it remains the case that eight more RCTs examining the use of FISH for PGS at the cleavage stage have followed, and none of them was able to demonstrate a benefit of chromosome screening.[38–45] [Table 1] summarizes these RCTs and their findings. Some experts in the field still assert that FISH applied to blastomeres can lead to improved IVF outcomes, and they point to technical deficiencies affecting all of the RCTs undertaken thus far. However, even if it is true that FISH-based analyses can be beneficial, the fact that so few laboratories are able to demonstrate any efficacy is indicative of a technology that is not sufficiently robust, leading to problems applying it in different laboratories.

It is widely agreed that a large proportion of the oocytes and embryos produced during IVF treatment are chromosomally abnormal and have little possibility of producing an infant. Why, then, has PGS been unsuccessful? Several factors could have contributed to the failure of PGS using FISH to improve implantation and pregnancy rates. Some problems may be related to the stage of analysis. As mentioned previously, the removal of even a single cell from a cleavage-stage embryo may cause a reduction in implantation potential. If the biopsy procedure is not performed by experienced practitioners, and according to published guidelines,[46] the reduction in implantation rate is likely to be larger still. Single-cell FISH is also hampered by several technical limitations. These include the fixation of the cell onto a microscope slide, a difficult procedure affected by temperature and humidity, which if performed incorrectly can lead to diagnostic problems and the interpretation of results, especially in the cases of splitting or overlapping signals. Additionally, because more than half of the chromosomes remain unexamined after FISH analysis, some of the embryos diagnosed "normal" will actually be abnormal, with aneuploidies affecting chromosomes that were not assessed. CGH examination of cleavage-stage embryos has shown that abnormalities can affect the entire chromosome complement.[8,21] The existence of mosaicism is also likely to have had some impact on accuracy and therefore outcomes.

Recent technological developments and validation of more comprehensive molecular cytogenetic methodologies such as CGH (metaphase or array) or SNP microarrays could potentially offer solutions to some of the problems just discussed. Wells and colleagues[47] were the first to combine a whole genome amplification approach, namely degenerate oligonucleotide primed polymerase chain reaction with CGH, which until then had been restricted to the cytogenetic analysis of large numbers of cells derived from tumors.[48] Validation and use of this more comprehensive method followed for the analysis of blastomeres, oocytes, and their corresponding PBs, and also TE samples.[8,15,16,21,49–53] aCGH evolved from metaphase CGH and has a similar technical approach. Both methods use a simultaneous hybridization of differentially labeled DNA samples (test sample: green; chromosomally normal reference: red) to normal metaphase chromosomes (metaphase CGH) or chromosome-specific DNA probes affixed on a microscope slide (aCGH). The ratio of green-to-red fluorescence along the length of each chromosome or on each chromosome-specific probe indicates whether there has been any gain or loss of chromosomal material in the test sample. Specifically, an excess of green fluorescence is indicative of a gain (e.g., a trisomy), whereas an excess of red fluorescence is indicative of a loss (e.g., monosomy). The advantage of aCGH over metaphase CGH is that it is more easily automated and requires a shorter time for hybridization and data analysis. It is therefore able to provide results much faster. SNP microarrays, in contrast, detect aneuploidy by interrogating 10,000 to 500,000 individual SNP sequences that are found throughout the human genome. Aneuploidy is detected either by an increase/decrease in fluorescence intensity for all of the probes on a specific chromosome or by using informative polymorphisms on each chromosome to track their inheritance from parents to embryos.[18,19]

Metaphase CGH has been used clinically to analyze blastomeres biopsied from embryos generated by couples who were undergoing PGS due to RIF.[49,53,54] Because the protocol required several days to complete, all biopsied embryos were frozen, and those found to be chromosomally normal were thawed and transferred in a subsequent cycle. Results suggested that couples experiencing RIF have a tendency to produce embryos that are highly abnormal, possibly explaining the patients' inability to achieve a pregnancy. As far as the clinical outcomes were concerned, both implantation and pregnancy rates seemed improved compared with those achieved using FISH for the purpose of PGS.[53,54] However, concerns were raised over the relatively low survival rates of the thawed embryos (46% did not survive the thawing process), and as a result CGH was not applied widely.[55–57]

aCGH was recently validated and applied for the analysis of biopsied blastomeres in a study performed by Gutiérrez-Mateo and colleagues.[58] During this investigation, a total of 795 embryos underwent a single blastomere biopsy. The obtained cell was examined with the use of aCGH, whereas the remainder of the embryos were spread onto microscope slides and analyzed by FISH for chromosomes 8, 13, 14, 15, 16, 17, 18, 20, 21, 22, X, and Y plus any others that aCGH had shown to be abnormal. The results obtained during the course of this study led to the conclusion that aCGH detects 42% more chromosome errors compared with FISH using probes for 12 chromosomes. Some embryos are affected by multiple chromosome anomalies and may therefore be correctly characterized as abnormal using FISH even if only one of the abnormalities present is detected. Nonetheless, aCGH was associated with the detection of 13% more abnormal embryos compared with FISH. Due to the high incidence of mosaicism observed (44% of embryos were mosaic), clinical validation was based on embryo classification (normal or abnormal), rather than confirmation of specific chromosome abnormalities after FISH reanalysis. The reported error rate ranged from 1.9% to 9%, depending on the method of whole genome amplification used.[58] The error rates for the optimal protocol were generally lower compared with those reported with the use of FISH.[13,59]

Treff and colleagues[19,60] validated a SNP microarray approach by examining the chromosome complement of 99 single cells derived from 9 aneuploid cell lines and 335 blastomeres obtained from a total of 235 cleavage-stage embryos. As with the aCGH methodology, SNP microarrays were shown to detect chromosome abnormalities accurately in both types of single cells (concurrence rate range: 96.5 to 99.8%). Also, similar to both CGH and aCGH findings, various types of mosaicism, including mitotic nondisjunction leading to reciprocal gains and losses of chromosomes in blastomeres from the same embryo, were identified.[60] The same research group compared the accuracy of their SNP microarray approach with that of FISH by examining a total of 160 blastomeres from 13 arrested cleavage-stage embryos with the use of one of the two methods.[19] This study was performed in a prospective randomized blinded manner and showed that FISH may have a tendency to overestimate aneuploidy compared with the SNP microarray.[19] [Table 2] summarizes the methodology and findings of studies that used more comprehensive cytogenetic strategies to examine the cleavage stage of preimplantation development.

It is evident that comprehensive cytogenetic methodologies have advantages over FISH in the clinical context of PGS. However, what was also obvious from the results obtained after CGH, aCGH, and SNP microarrays is that mosaicism significantly affects embryos at the cleavage stage and poses a serious risk of either misdiagnosing a mostly normal embryo as abnormal or vice versa. Concerns associated with this misdiagnosis risk, and generally about the efficacy of cleavage-stage PGS, have led to interest in testing during different stages, either earlier (PB sampling) or later (TE sampling).


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.