Aneuploidy Screening for Embryo Selection
Aneuploidy Screening for Embryo Selection
Female meiosis consists of two separate divisions, meiosis I (MI) and II (MII) and two different stages where the maturing oocyte temporarily arrests. During MI, chromosomes condense, find their homologues, align as a pair on the meiotic spindle, and exchange material with the formation of chiasmata. The oocyte remains arrested at this stage, until its ovulation, which may not occur for several decades. After MI resumption, chiasmata are resolved, homologues separate and start moving toward the two opposite poles of the meiotic spindle. One set of 23 chromosomes (each comprised of two chromatids) enters the first PB while the other set remains in the much larger oocyte. Another arrest follows, and MII begins again only if the oocyte is fertilized. During MI, sister chromatids are held together, and it is only after fertilization and at the end of MII that they are finally separated, one of the chromatids passing into the second PB, the other remaining in the oocyte.
Molecular analysis of samples from chromosomally abnormal pregnancies and miscarriages has shown that most aneuploidies occur during female meiosis (reviewed in Hassold et al). The advent of IVF meant that oocytes that remained unfertilized after sperm exposure or were unused for other reasons became available for research. Studies on the chromosomes of such oocytes demonstrated the presence of two aneuploidy-causing mechanisms. The first involves the mal-segregation of entire chromosomes, whereas the second involves the premature division (predivision) of a chromosome into its two constituent chromatids, followed by their random segregation, upon completion of the first and/or second meiotic division.
Various classical and molecular cytogenetic methods, such as G- or R-banding, FISH, spectral karyotyping, and CGH, have been used to examine the chromosomes of a large number of human oocytes. Data obtained from these investigations have demonstrated a clear relationship between advancing maternal age and increasing aneuploidy rates. Specifically, the expected aneuploidy rate seen in oocytes generated by women <25 years of age is ~5%, increasing to 10 to 25% in the early 30s, and >50% for women of ≥40 years.
The biopsy and examination of first and possibly second PBs for PGS offers an attractive alternative to blastomere biopsy and analysis. Once extruded from the oocyte, PBs are believed to have fulfilled their role, and their continued presence is not required for oocyte competence. Indeed, if left in place the PBs spontaneously degrade over the next few days. For these reasons, PB biopsy is considered to be less invasive than blastomere biopsy, and it does not affect either further embryonic development or implantation potential. The use of CGH to examine oocytes and corresponding PBs in a research context has shown that gain or loss of chromosome material in PBs is accompanied by a reciprocal loss or gain in the corresponding oocyte. The cytogenetic analysis of PBs can therefore be used to infer whether an oocyte is normal or aneuploid. Because any chromosome imbalance detected in the oocyte will be present in every cell of the resultant embryo, the diagnosis obtained after PB analysis would not be confused by mosaicism. The main disadvantage of PB examination is its inability to detect chromosome errors of paternal and/or postzygotic origin.
As with cleavage-stage PGS, FISH was the principal method used for the examination of PBs for many years. Verlinsky and coworkers analyzed large numbers of first and second PBs biopsied from oocytes generated by women of ARA (≥37 years). They used a five-probe combination to test chromosomes 13, 16, 18, 21, and 22, and their findings provided an insight into the incidence and frequency of aneuploidy during both meiotic divisions. Specifically, results obtained from the analysis of >20,000 oocytes generated by reproductively older women showed that 46.8% carried chromosome errors. Chromosome losses were seen more frequently (53%) in PBs than chromosome gains (26%), indicating that the corresponding oocytes would have a tendency to produce more trisomic rather than monosomic embryos. Errors of MII origin predominated over MI (41.8% versus 37.3%, respectively), and single chromatid mal-segregation occurred more frequently compared with whole chromosome nondisjunction during MI (27.1% versus 2.4% in a population of 7103 oocytes). As far as clinical outcomes were concerned, it was postulated that PB screening improved implantation and pregnancy rates and led to a reduction in spontaneous abortion and improved take-home infant rates, but no specific percentages were provided, and these studies were not randomized. Moreover, the use of five-color FISH not only meant that 18 of the 23 oocyte chromosomes were not examined, but it also necessitated that both PBs were spread onto microscope slides, a process that can lead to artifactual loss of chromosome material. Additionally, because PBs are degenerating cells, they are often of poor morphology, and the scoring of FISH probe signals can be difficult. Both these factors could therefore elevate the misdiagnosis risk.
The issues associated with the use of FISH for the examination of the female gamete in the context of PGS led to an increasing interest into the development and optimization of more comprehensive molecular cytogenetic methods for PB analysis. Wells and colleagues were the first to describe the clinical application of an accelerated metaphase CGH protocol, compatible with an embryo transfer on day 3 of preimplantation development. They used this CGH protocol for the analysis of 10 first PBs biopsied from oocytes generated by a 40-year-old IVF patient and accurately identified chromosome abnormalities in 9 of them. The aneuploidies detected were also confirmed with the use of FISH applied to blastomeres from the resulting embryos. The embryo that was characterized as normal by CGH was transferred, but unfortunately, no pregnancy ensued.
Conventional CGH was further optimized and validated in a large number of MII oocyte-first PB pairs and was clinically applied again as part of a "double-factor" PGD approach. During this investigation, chromosomal screening of the first PB using CGH was followed by analysis of two different cystic fibrosis mutations in blastomeres using a PCR approach. Of the five oocytes and two resulting embryos investigated, only one was found to have normal copies of the cystic fibrosis gene and also be chromosomally normal. Transfer of this particular embryo resulted in a clinical pregnancy and a healthy live birth.
Our group has also used metaphase CGH clinically to screen first and second PBs biopsied from a total of 308 oocytes generated by 70 reproductively older and, in some cases, poor prognosis women (average age: 40.8 years). The data obtained were confirmatory of Verlinsky's and Kuliev's observations. Specifically, the total oocyte abnormality rate was 70%, and unbalanced predivision of sister chromatids was the main MI aneuploidy-causing mechanism. Overall, a similar incidence of errors occurring in MI and MII was observed, although for women >40 years of age, MII errors were the most common. This finding suggests that the biopsy and analysis of not only the first but also the second PB could substantially improve accuracy, compared with the examination of the first PB alone. In our group of investigated samples, we would have failed to detect 45% of aneuploid oocytes if we had examined the first PB alone. On the contrary, if we had just analyzed the second PB, we would have wrongly characterized only 19.4% of the aneuploid oocytes as normal haploid. This suggests that, for women about ≥40 years of age, second PBs may provide more valuable information than first PBs. However, this may not be the case for younger patients.
The use of CGH showed that meiotic aneuploidy affects all oocyte chromosomes, but the smaller ones (13 onward) were found to be associated with higher error rates. Additionally, and concordant with Verlinsky's and Kuliev's observations, we saw that chromosome losses were occurring more frequently than chromosome gains in PBs, especially during MI (2:1 ratio of losses to gains for MI versus 1.5:1 ratio of losses to gains for MII). Nineteen (27%) of the 70 infertile couples who underwent PGS using first and second PB screening received no transfer because chromosome abnormalities were detected for all examined oocytes. The average age of this subgroup of women was 42 years (age range: 34 to 47 years), and most of the patients had experienced multiple unsuccessful IVF treatments and/or miscarriages. Preferential transfer of chromosomally normal oocytes has so far taken place for 36 of the remaining 51 couples (average maternal age: 40.5 years), and a total of 67 embryos were transferred. This led to an implantation rate of 13.4% and a pregnancy rate of 25% per transfer and 16.3% per initiated cycle (unpublished data and Fragouli et al). The observed pregnancy rate remains relatively low but is nevertheless almost double the rate expected for such poor prognosis patients. However, because this was not a well-controlled study, no definite conclusions can be drawn at this time.
Both aCGH and SNP microarrays have been used clinically for the analysis of first and sometimes second PBs. [Table 3] summarizes the findings and clinical outcomes (where available) from all the previously mentioned studies. It should be noted that the European Society of Human Reproduction and Embryology (ESHRE) PGS Task Force has also been exploring the use of aCGH to examine first and second PBs, as well as the corresponding oocytes, allowing the true reliability of this approach to be evaluated. The initial data are encouraging, indicating an accuracy rate of 94%. This investigation is the first part of a planned multicenter randomized clinical trial that ESHRE is undertaking to assess the efficacy of PGS via PB analysis. At the time of writing, this RCT is about to begin, and we hope the obtained results will determine whether PB analysis is capable of enhancing pregnancy rates.
Chromosome Screening of the Female Gamete
Female meiosis consists of two separate divisions, meiosis I (MI) and II (MII) and two different stages where the maturing oocyte temporarily arrests. During MI, chromosomes condense, find their homologues, align as a pair on the meiotic spindle, and exchange material with the formation of chiasmata. The oocyte remains arrested at this stage, until its ovulation, which may not occur for several decades. After MI resumption, chiasmata are resolved, homologues separate and start moving toward the two opposite poles of the meiotic spindle. One set of 23 chromosomes (each comprised of two chromatids) enters the first PB while the other set remains in the much larger oocyte. Another arrest follows, and MII begins again only if the oocyte is fertilized. During MI, sister chromatids are held together, and it is only after fertilization and at the end of MII that they are finally separated, one of the chromatids passing into the second PB, the other remaining in the oocyte.
Molecular analysis of samples from chromosomally abnormal pregnancies and miscarriages has shown that most aneuploidies occur during female meiosis (reviewed in Hassold et al). The advent of IVF meant that oocytes that remained unfertilized after sperm exposure or were unused for other reasons became available for research. Studies on the chromosomes of such oocytes demonstrated the presence of two aneuploidy-causing mechanisms. The first involves the mal-segregation of entire chromosomes, whereas the second involves the premature division (predivision) of a chromosome into its two constituent chromatids, followed by their random segregation, upon completion of the first and/or second meiotic division.
Various classical and molecular cytogenetic methods, such as G- or R-banding, FISH, spectral karyotyping, and CGH, have been used to examine the chromosomes of a large number of human oocytes. Data obtained from these investigations have demonstrated a clear relationship between advancing maternal age and increasing aneuploidy rates. Specifically, the expected aneuploidy rate seen in oocytes generated by women <25 years of age is ~5%, increasing to 10 to 25% in the early 30s, and >50% for women of ≥40 years.
The biopsy and examination of first and possibly second PBs for PGS offers an attractive alternative to blastomere biopsy and analysis. Once extruded from the oocyte, PBs are believed to have fulfilled their role, and their continued presence is not required for oocyte competence. Indeed, if left in place the PBs spontaneously degrade over the next few days. For these reasons, PB biopsy is considered to be less invasive than blastomere biopsy, and it does not affect either further embryonic development or implantation potential. The use of CGH to examine oocytes and corresponding PBs in a research context has shown that gain or loss of chromosome material in PBs is accompanied by a reciprocal loss or gain in the corresponding oocyte. The cytogenetic analysis of PBs can therefore be used to infer whether an oocyte is normal or aneuploid. Because any chromosome imbalance detected in the oocyte will be present in every cell of the resultant embryo, the diagnosis obtained after PB analysis would not be confused by mosaicism. The main disadvantage of PB examination is its inability to detect chromosome errors of paternal and/or postzygotic origin.
As with cleavage-stage PGS, FISH was the principal method used for the examination of PBs for many years. Verlinsky and coworkers analyzed large numbers of first and second PBs biopsied from oocytes generated by women of ARA (≥37 years). They used a five-probe combination to test chromosomes 13, 16, 18, 21, and 22, and their findings provided an insight into the incidence and frequency of aneuploidy during both meiotic divisions. Specifically, results obtained from the analysis of >20,000 oocytes generated by reproductively older women showed that 46.8% carried chromosome errors. Chromosome losses were seen more frequently (53%) in PBs than chromosome gains (26%), indicating that the corresponding oocytes would have a tendency to produce more trisomic rather than monosomic embryos. Errors of MII origin predominated over MI (41.8% versus 37.3%, respectively), and single chromatid mal-segregation occurred more frequently compared with whole chromosome nondisjunction during MI (27.1% versus 2.4% in a population of 7103 oocytes). As far as clinical outcomes were concerned, it was postulated that PB screening improved implantation and pregnancy rates and led to a reduction in spontaneous abortion and improved take-home infant rates, but no specific percentages were provided, and these studies were not randomized. Moreover, the use of five-color FISH not only meant that 18 of the 23 oocyte chromosomes were not examined, but it also necessitated that both PBs were spread onto microscope slides, a process that can lead to artifactual loss of chromosome material. Additionally, because PBs are degenerating cells, they are often of poor morphology, and the scoring of FISH probe signals can be difficult. Both these factors could therefore elevate the misdiagnosis risk.
The issues associated with the use of FISH for the examination of the female gamete in the context of PGS led to an increasing interest into the development and optimization of more comprehensive molecular cytogenetic methods for PB analysis. Wells and colleagues were the first to describe the clinical application of an accelerated metaphase CGH protocol, compatible with an embryo transfer on day 3 of preimplantation development. They used this CGH protocol for the analysis of 10 first PBs biopsied from oocytes generated by a 40-year-old IVF patient and accurately identified chromosome abnormalities in 9 of them. The aneuploidies detected were also confirmed with the use of FISH applied to blastomeres from the resulting embryos. The embryo that was characterized as normal by CGH was transferred, but unfortunately, no pregnancy ensued.
Conventional CGH was further optimized and validated in a large number of MII oocyte-first PB pairs and was clinically applied again as part of a "double-factor" PGD approach. During this investigation, chromosomal screening of the first PB using CGH was followed by analysis of two different cystic fibrosis mutations in blastomeres using a PCR approach. Of the five oocytes and two resulting embryos investigated, only one was found to have normal copies of the cystic fibrosis gene and also be chromosomally normal. Transfer of this particular embryo resulted in a clinical pregnancy and a healthy live birth.
Our group has also used metaphase CGH clinically to screen first and second PBs biopsied from a total of 308 oocytes generated by 70 reproductively older and, in some cases, poor prognosis women (average age: 40.8 years). The data obtained were confirmatory of Verlinsky's and Kuliev's observations. Specifically, the total oocyte abnormality rate was 70%, and unbalanced predivision of sister chromatids was the main MI aneuploidy-causing mechanism. Overall, a similar incidence of errors occurring in MI and MII was observed, although for women >40 years of age, MII errors were the most common. This finding suggests that the biopsy and analysis of not only the first but also the second PB could substantially improve accuracy, compared with the examination of the first PB alone. In our group of investigated samples, we would have failed to detect 45% of aneuploid oocytes if we had examined the first PB alone. On the contrary, if we had just analyzed the second PB, we would have wrongly characterized only 19.4% of the aneuploid oocytes as normal haploid. This suggests that, for women about ≥40 years of age, second PBs may provide more valuable information than first PBs. However, this may not be the case for younger patients.
The use of CGH showed that meiotic aneuploidy affects all oocyte chromosomes, but the smaller ones (13 onward) were found to be associated with higher error rates. Additionally, and concordant with Verlinsky's and Kuliev's observations, we saw that chromosome losses were occurring more frequently than chromosome gains in PBs, especially during MI (2:1 ratio of losses to gains for MI versus 1.5:1 ratio of losses to gains for MII). Nineteen (27%) of the 70 infertile couples who underwent PGS using first and second PB screening received no transfer because chromosome abnormalities were detected for all examined oocytes. The average age of this subgroup of women was 42 years (age range: 34 to 47 years), and most of the patients had experienced multiple unsuccessful IVF treatments and/or miscarriages. Preferential transfer of chromosomally normal oocytes has so far taken place for 36 of the remaining 51 couples (average maternal age: 40.5 years), and a total of 67 embryos were transferred. This led to an implantation rate of 13.4% and a pregnancy rate of 25% per transfer and 16.3% per initiated cycle (unpublished data and Fragouli et al). The observed pregnancy rate remains relatively low but is nevertheless almost double the rate expected for such poor prognosis patients. However, because this was not a well-controlled study, no definite conclusions can be drawn at this time.
Both aCGH and SNP microarrays have been used clinically for the analysis of first and sometimes second PBs. [Table 3] summarizes the findings and clinical outcomes (where available) from all the previously mentioned studies. It should be noted that the European Society of Human Reproduction and Embryology (ESHRE) PGS Task Force has also been exploring the use of aCGH to examine first and second PBs, as well as the corresponding oocytes, allowing the true reliability of this approach to be evaluated. The initial data are encouraging, indicating an accuracy rate of 94%. This investigation is the first part of a planned multicenter randomized clinical trial that ESHRE is undertaking to assess the efficacy of PGS via PB analysis. At the time of writing, this RCT is about to begin, and we hope the obtained results will determine whether PB analysis is capable of enhancing pregnancy rates.
