Hum. Reprod. Advance Access originally published online on May 13, 2007
Human Reproduction 2007 22(7):1844-1853; doi:10.1093/humrep/dem102
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Variable aneuploidy mechanisms in embryos from couples with poor reproductive histories undergoing preimplantation genetic screening
1 Department of Obstetrics and Gynaecology, University College London Centre for Preimplantation Genetic Diagnosis, UCL, 86-96 Chenies Mews, London, WC1E 6HX, UK 2 The Assisted Conception Unit, University College London Hospitals Foundation Trust, London, UK
3 Correspondence address. E-mail: a.mantzouratou{at}ucl.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
BACKGROUND: Preimplantation genetic screening (PGS) is used to determine the chromosome status of human embryos from patients with advanced maternal age (AMA), recurrent miscarriage (RM) or repeated implantation failure (RIF).
METHODS: Embryos from 47 such couples were investigated for chromosomes 13, 15, 16, 18, 21 and 22 using fluorescence in situ hybridization with two rounds of hybridization. The investigation included parental lymphocyte work-up, the screening of blastomeres on day 3 and full follow-up on day 5/6 of untransferred embryos.
RESULTS: The outcome of 60 PGS cycles is described, in which 523 embryos were biopsied; 91% gave results, of which 18% were diploid for all the chromosomes tested and 82% were abnormal. The pregnancy rate per cycle that reached the biopsy stage was 27%, and 30% per embryo transfer. Satisfactory follow-up was obtained from 353 embryos; all those diagnosed as abnormal were confirmed as such, although two false-positives were detected in relation to specific chromosome abnormalities. Meiotic errors were identified in 16% of embryos. Between the RM, AMA and RIF groups, there was a significant difference in the distribution of embryos that were uniformly abnormal and of those with meiotic errors; with an almost 3-fold increase in meiotic errors in the first two groups compared with the RIF group.
CONCLUSIONS: This complete investigation has identified significant differences between referral groups concerning the origin of aneuploidy in their embryos.
Key words: PGS/FISH/aneuploidy mechanisms/recurrent miscarriage/implantation failure
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
Preimplantation genetic diagnosis (PGD) involves the removal of one or two cells from human in vitro fertilization (IVF) embryos, usually at the cleavage stage. Preimplantation genetic screening (PGS) is a derivation of PGD and is used in a more general context. It allows for the screening of specific chromosomes that are most commonly involved in aneuploidy, in blastomeres from embryos produced by certain high risk couples. PGS is mainly done using fluorescence in situ hybridization (FISH), and at present thousands of embryos are being screened around the world (Harper et al., 2006
), and a wealth of genetic information is being generated that should provide insight into the early human genetic processes, and their role in fertility and infertility.
Early studies on the outcome of using FISH for sexing embryos to avoid X-linked disease rapidly revealed that on average, at least 30% of embryos generated by IVF are chromosomally mosaic and that many are so highly abnormal that they are classified as chaotic mosaics (Delhanty et al., 1997). Later, the use of additional probes for PGS in high risk patient groups showed that at least 50–60% of embryos may be classed as mosaic or chaotic (Magli et al., 2000; Bielanska et al., 2002). Studies with comparative genomic hybridization (CGH) in human preimplantation embryos (Voullaire et al., 2000; Wells and Delhanty, 2000) confirmed the existence of widespread chromosomal abnormalities in this early stage of development. These data have changed our perception of the genetic processes in early human embryos. Follow-up studies on embryos generated for PGS, far from revealing mainly uniformly diploid or aneuploid embryos, showed that these two categories constituted the minority groups. The question as to why so many IVF generated human embryos have such a high level of post-zygotic abnormalities, and the possible cause, is still to be answered.
The first problem may lie with IVF procedures themselves. The conditions that the oocytes, sperm and embryos are subjected to do not resemble those of a natural cycle, so the environment necessary for IVF may play a vital role in this process (Gardner et al., 2005). However, since there is no control group of human embryos from natural cycles that can be directly compared with the IVF ones, our only information comes from indirect studies and most significantly from animal studies (Carrell et al., 2005; Redding et al., 2006). Some of them provide evidence that the IVF process can induce errors in early non-human embryos, particularly in mouse strains that are susceptible to aneuploidy (Bean et al., 2002). However, even in natural cycles, human reproduction is more error prone than that of other animals and an increased rate of chromosomal abnormalities in human development seems to provide at least part of the answer (Delhanty, 2001).
The problem with chromosomal screening in preimplantation embryos is that only a few chromosomes can be detected at any one time due to various technical limitations of the preferred PGS method, FISH. Results are usually based on one or two biopsied cells and follow-up of the untransferred embryos is not carried out in most centres. So for screening purposes in PGS, only those chromosomes that are deemed at high risk of error are being checked, for example, chromosomes 13, 16, 18, 21, 22 and X, Y. Different centres use different approaches and for PGS up to nine chromosomes are being screened for routinely, using FISH (reviewed in Wilton, 2002). However, most centres rely on the PB MultiVysion probe set developed by Vysis (Abbott) that simultaneously tests for five autosomes (13,16,18, 21 and 22).
We have evaluated this probe set using somatic cells and have not been satisfied with its degree of efficiency or accuracy. This led us to develop our own protocol which showed efficiencies of 89–96%, whereas the PB MultiVysion probe mix showed efficiency levels of 75% or lower. Problems with the latter included difficulty in locating embryonic nuclei without using blue counterstain and difficulty with overlapping signals. In addition there were wide variations in efficiency between different probes within or between experiments and bleed through of signals between filters. In this centre, six chromosomes (13, 15, 16, 18, 21 and 22), in two rounds of FISH, are screened for in embryos to detect aneuploidy with a high degree of accuracy. The sex chromosomes are not screened for unless there is a specific reason for an increased risk of sex chromosome aneuploidy, such as sex chromosome mosaicism.
The couples that are referred for PGS are usually those who have not been helped with conventional IVF or have a high risk of producing abnormal offspring but with no specific aetiology, apart from advanced maternal age (AMA). So these couples usually fall into three categories: (i) AMA (women over 40 years at our centre), (ii) repeated implantation failure (RIF), with three or more IVF attempts and (iii) recurrent miscarriage (RM), with three or more spontaneous abortions.
Studies on the outcome of PGS cycles have shown varied results. Gianaroli et al. (2005) found 33% of embryos to be normal and a 30% pregnancy rate per cycle with embryo transfer after PGS. They also found an increased implantation rate in those patients that had PGS compared with their previous reproductive history and concluded that PGS was beneficial to poor prognosis patients. In a wide variety of other studies, a high incidence of aneuploidy in poor prognosis IVF/PGS couples has been found with abnormalities in 55–70% of embryos tested for PGS (reviewed in Donoso et al., 2007).
In this report, we present the results of the first 60 cycles of PGS treatment at the UCL Centre for PGD, involving the investigation of over 500 embryos. The aim was to uncover the origins of embryonic aneuploidy in our investigated group of patients and gain information about the genetics of human preimplantation development and the mechanisms that govern it.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
Studies were performed on lymphocytes, blastomeres and untransferred embryos from IVF patients undergoing PGS for various reasons. First, patients who were referred for PGS had their lymphocytes cultured and analysed using FISH with the diagnostic probe set. This was done in order to investigate any possible chromosomal polymorphisms that might interfere with diagnosis in embryos, to exclude possible mosaicism and optimize FISH protocols for their use in embryonic cells during PGS. Secondly, the blastomeres that were biopsied during PGS were screened for chromosome anomalies, and after the PGS cycle, the non-transferred embryos were also analysed with the diagnostic probe set as standard practice.
Patient details and ethical consent
The patients are divided into three major groups according to their main indication for undergoing PGS. The first group (AMA) included seven patients undergoing PGS for AMA and in this study included females over 40 years of age. The second group (RM) included 10 couples that had experienced three or more spontaneous abortions mostly from natural cycles, and the third group (RIF) included 30 couples that had experienced failure of implantation after three or more routine IVF cycles. Treatment and research on embryos from PGS couples was carried out under licences from Human Fertilization and Embryology Authority (HFEA) of the UK. Informed written consent was obtained from all the couples.
Lymphocyte culture and counts
Karyotyping of all couples was performed by a clinical cytogenetics laboratory prior to the onset of treatment to exclude constitutional abnormality. For preparative FISH studies, lymphocyte cultures from both partners of PGS couples were carried out by standard methods. The efficiency of the FISH probe combination was calculated by counting the number of correct signals in 100–200 interphase nuclei from each sample.
IVF and stimulation protocol, embryos biopsy, blastomere and embryo spreading
Ultrasound guided vaginal oocyte collection was performed at 37 h post-hCG injection. IVF or IVF/ICSI was performed at 40 and 41 h post-hCG, respectively, and was dependent on semen parameters and past fertilization rates. Fertilization was evaluated at 18–20 h post-insemination. Embryos were cultured in IVF medium (GIII series, Vitrolife, UK). On day 3, embryos that reached at least the four-cell stage (the majority of embryos had reached the 6–8 cell stage on day 3) were biopsied in Ca2 + –Mg 2 + free biopsy medium (G-PGD, Vitrolife, UK). One cell was removed from most embryos, whereas a second cell was removed from a small number of embryos where the nucleus of the first cell was not clearly seen. Biopsied blastomeres for PGS were spread using the method described by Harper et al. (1994).
Fluorescence in situ hybridization
FISH experiments were previously undertaken in order to test and optimize conditions for all the probes in this study. Hybridization to probes for six pairs of chromosomes was undertaken in two separate rounds; probes for chromosomes 13,18 and 21 were used in the first round and those for chromosomes 15, 16 and 22 were used in the second round. FISH procedures were carried out essentially as described in Harper et al. (1994) with some modifications in hybridization times and post-washes in order to ensure maximum efficiency. The slides were examined under an epifluorescence Olympus microscope (Olympus BX 40) fitted with a Photometrics cooled CCD camera utilizing Smartcapture software (Digital Scientific, UK). DAPI stained nuclei were located using the blue filter. Using different colour filters, the scoring of signals for each of the probes to the nuclei on the slides was possible with a good degree of accuracy. All scoring decisions were made directly by viewing signals under the microscope and verified by at least two observers.
Scoring and classification criteria of embryos according to FISH results
Strict scoring criteria were applied in order to classify the studied nuclei and embryos correctly according to Hopman et al. (1988): (i) split signals: when a chromosome has two chromatids in interphase they may appear as doublets, which are equal in size and smaller than the normal signal. The split signals must be separated by less than the width of a normal signal in order to be classified as one chromosome. (ii) Stretched or diffused signals must not present any interruption in order to be classified as one chromosome. (iii) Nuclei with uniformly diploid signals were classified as normal. In addition: (i) Embryos with blastomeres showing the same abnormality in at least 90% of cells were classified as uniformly aneuploid due to meiotic error. (ii) Embryos with cell lines showing different abnormalities were classified as mosaic; reciprocal errors were recorded as due to mitotic non-disjunction; non-reciprocal errors in more than 20% of nuclei in an embryo were classified as chromosome loss or chromosome gain. (iii) Embryos with blastomeres showing different abnormalities in each nucleus, affecting at least three chromosome pairs, were classified as chaotic mosaics. (iv) Embryos were fully chaotic if all cells were affected; otherwise, they may be partially diploid or aneuploid and partially chaotic. In order to assess the mode of aneuploidy in embryos, only those with definitive cell lines were included, and we excluded those where there were doubts over the events that lead to aneuploidy. (v) Embryos were classified as diploid on follow-up if they contained at least 90% diploid cells.
Statistical analysis
The mean and standard deviation was calculated for the number of oocytes collected per cycle, embryos biopsied and embryos with diploid result. The chi-square distribution test was used to compare the distribution of various types of abnormalities in embryos from couples from the different referral or age groups.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
Lymphocyte and FISH efficiency results
Once the karyotypes were confirmed to be normal, lymphocyte slide preparations from both partners were analysed after FISH with the diagnostic probe set. The efficiency of the PGS protocol in both partners was calculated and the existence of any cytogenetic polymorphisms was investigated. The overall efficiency (percentage of cells with diploid signals) of FISH in lymphocyte slides ranged from 89% to 96%, and no polymorphisms were found in any of the PGS couples. There were however some exceptions. The lymphocyte preparations from three males that were referred from RM couples showed considerably increased aneuploidy compared with the controls for the same experiments. The normality rate was below 70% in all and as low as 58% in one case. All three couples had previous aneuploid conceptions and there were no abnormal sperm parameters. One of the couples proceeded to PGS and had a normal live birth with their first cycle; upon examination of the non-transferred embryos, it was observed that 50% of these embryos were aneuploid with errors that were meiotic in origin. Whether the aneuploidy of these embryos was connected with the apparently increased aneuploidy in the paternal lymphocytes is still under investigation. The other two couples did not proceed to PGS, so no follow-up data could be obtained.
Overall diagnostic data
Single biopsied cells from embryos generated for PGS were screened for chromosomal abnormalities on day 3 using FISH for six chromosomes (13, 15, 16, 18, 21 and 22). All untransferred embryos (apart from four that were frozen) were spread on slides and studied using the same probe set after embryo transfer on day 5. This provided us with valuable information about the accuracy of our PGS programme and the genetic make-up of the embryos.
Table 1 summarizes the diagnostic PGS data. In total, 60 PGS cycles were performed for the 47 couples, with average maternal age of 37. In the 60 cycles, 709 oocytes were collected, and 523 embryos were biopsied with an average of 8.7 embryos per cycle. From the biopsied embryos with results (91% of the total), 18% (85) were found to be diploid for the tested chromosomes and 82% (391) were abnormal. The overall pregnancy rate (27% per oocyte retrieval, 30% per embryo transfer and 26% delivered) is broken down according to the maternal age, the lowest being for females of 37 years and over, which was 17% per cycle with biopsy and 19% per embryo transfer, and the highest being for the 25–30 years age group, which was 50% per embryo transfer. There were 14 deliveries of healthy singletons, and two very early spontaneous abortions (12.5% spontaneous abortion rate). Both spontaneous abortions were detected at 7 weeks of gestation but no follow-up karyotype analysis was done. There were only four cycles out of 60 that had more than two normal embryos available for transfer, an important factor since two is the maximum number allowable for transfer in the UK.
|
Overall follow-up data
Table 2 summarizes the overall results from the follow-up of untransferred embryos. Follow-up results were obtained for 353 of the 391 embryos diagnosed as abnormal (84%). Among these 353 embryos, 57.8% were fully chaotic mosaic and 36.8% were classified as other mosaic types. The most prevalent of the other mosaic types were the aneuploid mosaics (33.8%) followed by those that were diploid/chaotic (21.5%) or aneuploid/chaotic (20.8%).
|
Biopsied embryos resulting from 37 0PN, 5 1PN and 8 3PN were also included in the follow-up studies, since the pronuclear classification does not always predict accurately the chromsomal status of subsequent embryos. Follow-up results were obtained from 19 0PN, 3 1PN and 5 3PN embryos. Their inclusion or exclusion has not affected the results to a significant degree (Table 3 shows follow-up results on these embryos separately).
|
Fifty mosaic embryos had diploid cells on follow-up and were diagnosed as abnormal on biopsy, which is not surprising since only in 5 out of these 50 embryos was there a majority of diploid cells. In these five embryos, the diploid cell lines constituted a range of 55–68% of the total. In all, only 5.4% of the embryos were uniformly abnormal where all the cells carried the same abnormalities; three quarters of these were aneuploid and the remaining were haploid. Parental origin errors (meiotic) were identified in 15.9% of all embryos.
From the embryos with no result on biopsy, follow-up was obtained in 26 out of 47, five had been transferred undiagnosed as there was no alternative but no pregnancy resulted from any undiagnosed transfers, and one was frozen as it reached blastocyst on day 5 and was of good quality. The remaining 15 could not be analysed further mostly due to being degenerate by days 5/6.
Although the exact abnormalities detected on biopsy were not necessarily present in the embryo after follow-up, in cases of reciprocal mitotic non-disjunction or of chaotic mosaicism, for example, it was considered that only two of the 353 embryos had false-positive results, where an error for one of the chromosomes tested did not show in the follow-up as expected. However, these were not clinically significant as there were detected abnormalities in other chromosomes on biopsy, which were also picked up on follow-up. This gives a false positive rate of 0.6% and in both cases after re-examination of the captured nuclei, it was thought that the errors were probably caused by overlapping or diffused signals. There were no known false negatives.
Maternal age and embryo abnormalities
Figure 1 shows a general trend of embryos with meiotic errors increasing with maternal age. Table 4 shows the meiotic errors found in embryos from women of different age groups, 25–30 years, 31–36 years, 37–39 years and 40 years and over, with the highest being 19% embryos in the oldest group with meiotic errors in contrast with 12%, 14% and 14% in the other three age groups. However, there was no statistically significant difference between the maternal age groups for the distribution of embryos with meiotic abnormalities. Additionally, no significant difference in the distribution of normal, mosaic, chaotic and uniformly abnormal types between the maternal age groups could be found, even when the two groups of younger women were combined. The very young women did show increased mosaicism (56%) compared with the other two groups (32% and 35.5%), but the numbers are not large enough for this group to show a statistical difference. The prevalent type of mosaicism differs slightly between the groups. Diploid/chaotic mosaics appear to be prevalent in the two younger groups (30% and 31%) but in the older group aneuploid mosaics were the majority (44%); again there were no significant differences.
|
|
Referral groups and embryo abnormalities
Table 5 summarizes the data from the AMA, RM and RIF referral groups. Average maternal ages were 42.4, 37.3 and 36, respectively. The highest pregnancy rate per embryo transfer was achieved in the RM group (33%), which also had the highest average number of embryos biopsied per cycle (9.2 ± 3.6). The lowest pregnancy rate per cycle was in the AMA group. The RIF group was the largest group with a 26% pregnancy rate per cycle (29% per embryo transfer). Two very early spontaneous abortions occurred in the RIF group bringing the ongoing pregnancy rate for RIF group to 27.6% per embryo transfer. Table 3 also shows the distribution of the various mosaic types and the uniformly abnormal and meiotic errors for each group. No significant difference was found between the distribution of normal, fully chaotic and other mosaics, in general among these groups. Fully chaotic embryos seem to occur irrespective of age and reproductive history in roughly the same proportion. The distribution of less severe mosaicism however appears to differ. In the AMA and RM groups, aneuploid mosaics dominate (36% and 41%, respectively), whereas in the RIF group, aneuploid mosaics and diploid/chaotic mosaics are in almost equal proportions. Within the mosaic types, there was no statistically significant difference between the distribution of aneuploid mosaics, but there was a significant difference in the distribution of diploid/chaotic (P < 0.05) embryos in the groups; there were more diploid/chaotics in the RIF group.
|
On the other hand, there was a statistical significant difference between the distribution of uniformly abnormal embryos (chi-square, P < 0.005) and embryos with meiotic abnormalities (chi-square P < 0.005) in these groups. An almost 3-fold increase in the percentage of embryos with meiotic errors is evident in the AMA and RM groups versus the RIF group, whereas more than 98% of embryos were mosaics in the RIF group versus 88% and 90.8% in the RM and AMA types. Also from Table 5, the striking similarities in all aspects of chromosomal errors in RM and AMA groups could be observed. Most RM patients were clearly able to conceive, since most of the previous spontaneous abortions were from natural cycles apart from one couple, with severe male factor infertility, who had experienced spontaneous abortions from IVF pregnancies; the AMA group on the other hand included women over 40 who were unable to conceive naturally and needed IVF.
Chromosomes and mechanisms of aneuploidy
The overall data for the chromosomal errors where a mechanism could be established are shown in Table 6. Meiotic abnormalities are the largest identifiable group, because the errors were universal in the embryos, and the aneuploidy was clearly seen to be of parental origin prior to fertilization. These errors most commonly affected chromosomes 21, 18 and 22, and trisomy appeared to be more frequent than monosomy. The mechanisms of mitotic abnormalities were less obvious in most cases, but overall, they seemed to affect chromosomes 16, 15 and 22 more often. Mitotic non-disjunction was the most easily identifiable mechanism of post-zygotic errors and hence appeared most prevalent, followed by chromosome loss and lastly chromosome gain. Figure 2 shows an example of mitotic non-disjunction.
|
|
Overall, the meiotic and mitotic chromosome error frequency was, in order, 22 (21%), 21 (18.5%), 16, 18 and 15 (17%) and 13 (12%). Meiotic trisomy 15 was not detected at all but monosomy 15 was present among the meiotic errors.
Breaking down the chromosomal errors for each referral group provided another insight into the mechanisms of aneuploidy. Very few of the errors in the RIF group appeared to originate from meiosis (9%), whereas in the other two groups, at least a quarter of embryos had errors that seem to have originated in meiosis. For the RIF group, the most frequent mitotic errors involved chromosome 16 (33%) and 22 (29%), whereas meiotic errors affected chromosomes 22 (29%) and 18 (25%) (2:1 trisomy:monosomy). Mitotic abnormalities where the mechanism could be identified were very few in the AMA and RM groups. Meiotic errors in the RM group involved mostly chromosomes 13 (25%) and 21 (25%) followed by 18 (20%). However, trisomy and monosomy was seen almost in a 1:1 ratio. In the AMA group, most meiotic errors involved chromosomes 21 (36%) and 18 (21%).
Both false positive results in this study involved chromosome 18. These were classified as scoring errors, since upon re-evaluation of the biopsied nuclei after follow-up it was concluded that split and diffused signals of varying size and distance had lead the observers to classify both as trisomy 18.
Insemination method and chromosomal abnormalities
The data were also investigated in relation to the insemination method for each cycle. There were 34 IVF and 36 ICSI cycles. There was a slightly higher percentage of normal embryos found in the ICSI group (20%) than the IVF group (13.5%). There was no significant difference in the distribution of normal, mosaic and chaotic embryos (P > 0.05) between these two groups.
Embryos with meiotic errors were 19% in the IVF cycles and 10% in the ICSI cycles (P < 0.05); however, average maternal age was higher in the IVF group (38.7) than in the ICSI group (35.4). Similarly, there was a significant difference in the distribution of uniformly abnormal embryos (P < 0.025) (16/225 in the IVF group versus 3/128 in the ICSI group), aneuploid mosaics (P < 0.025) and diploid/chaotic embryos (P < 0.005) with the latter two being higher in the ICSI group.
However, since the ICSI group consisted mostly of RIF couples (23/26 ICSI cycles were performed for RIF couples) to investigate whether insemination method had any effect, we decided to use the RIF group data only. Within the RIF group, there were 23 ICSI and 15 IVF cycles. No significant difference was found in any of the follow-up data between the IVF and ICSI cycles in the RIF group, which shows that the chromosomal abnormalities in these embryos are not related to the insemination method or poor sperm parameters.
Embryo classification and developmental potential
Overall, 89 out of 353 non-transferred embryos (25%) reached the blastocyst or morula stage by days 5/6 (on day 5/6, all untransferred embryos were spread irrespective of their developmental stage). Of these embryos, 36 were blastocyst and the rest (53) were morulas or cavitating morulas. Of the blastocysts, 19/36 (53%) had a diploid cell line as well as chaotic or aneuploid lines and the rest were either fully chaotic or had no diploid cell lines. Morulas had various abnormalities including some being uniformly abnormal, an observation that was not made for any of the blastocysts. For the AMA group, 13/65 (20%) embryos reached the blastocyst or morula stage (of the four blastocysts, one had a small number of diploid cells and the rest had meiotic errors with additional post-zygotic errors). In the RM group, 24/76 (31.6%) embryos reached the blastocyst or morula stage, of which 11 were blastocysts and 13 morulas. Four blastocysts had a diploid cell line and one of them had a 1:1 ratio of diploid:aneuploid cells; in the remaining three, the diploid cells were a minority. In the RIF group, 52/212 (24.5%) progressed to 21 blastocysts and 31 morulas. Ten of the 21 (48%) blastocysts had a diploid cell line with three of them having a diploid predominant cell line (in those three embryos, the diploid versus aneuploid cell ratio was 41:25, 31:24 and 31:23).
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
Results from 60 cycles of PGS, with over 500 embryos biopsied and with full follow-up studies were obtained using FISH to screen for chromosomes 13, 15, 16, 18, 21 and 22 in two rounds of hybridization. The use of different fluorochromes for each chromosome, rather than ratio labelling, increased the accuracy of the protocol. The choice of chromosomes was carefully considered in order to provide the maximum benefits for the couples undergoing PGS without compromising on the efficiency of the FISH technique. Although chromosome 15 is not routinely included in most PGS programmes, the decision to include it appears to be a valid one for this centre, as errors in chromosome 15 constitute 15% of all identifiable errors of mitotic and meiotic origin. Preliminary studies before the start of this PGS programme showed reduced FISH efficiency when more than three probes are used in a single hybridization. Given the choice between screening for more chromosomes but with higher error rate and screening for six chromosomes in two hybridization rounds with increased efficiency, we have chosen the latter. With efficient screening for six chromosomes, the proportion of embryos diagnosed as normal in this study is quite low (18%), and in only four of the 60 cycles were more than two normal embryos found.
At 82%, the rate of abnormality in this study is much higher than in most previously published studies on PGS diagnoses in embryos, where 50–70% of embryos appeared to be abnormal [Gianarolli et al., 1997, 2005; Rubio et al., 2005; Munne et al., 2005 (reviewed in Donoso et al., 2007)]. Bienlanska et al. (2002) reported an overall mosaicism rate in untransferred embryos of 48.1%, but found that mosaicism was increasing with the embryonic developmental stages and at the blastocyst stage the mosaicism rate reached 90.9% if tetraploidy was included. The high abnormality rate in this study could be a reflection of the selection of the couples in this centre where PGS is done as a last option and most couples present with a very poor reproductive history. Alternatively, the outcome could be attributed to a combination of the increased efficiency of this protocol as proved by the follow-up data, strict embryo scoring criteria and the patient selection procedure.
Despite the low normality rate (1.4 ± 1 embryos per cycle), the pregnancy rate of 30% per embryo transfer is above the average of the latest ESHRE data for PGS (25% per embryo transfer) (Harper et al., 2006). This suggests that efficient screening for six chromosomes is sufficient to detect the few embryos that are suitable for transfer in this cohort of patients.
However, even with efficient embryo screening, the majority of transferred embryos fail to implant and this is an area that requires further investigation. It might be that screening for all chromosomes will lead to an improved detection rate but this will result in even fewer embryos for transfer. Mosaicism as widespread as this study suggests must also contribute to the failure of implantation but this will remain a problem while reliance is placed upon single cell biopsy for diagnosis.
Initially, full follow-up studies for all couples in the PGS programme were carried out to validate the biopsy results. All the embryos that were found to be abnormal on diagnosis were found to have varying degrees of abnormality in the follow-up. Our very low false positive rate of 0.6% shows that the abnormalities observed in the biopsied cell are a true representation of at least some of the cells found in the embryos and that the abnormality rate reported in the study is a true finding. For comparison, Daphnis et al. (2005) found an average 5% FISH error rate in a careful study of human preimplantation embryos. A retrospective study of monosomic embryos at the biopsy stage and their follow-up (Cooper et al., 2006) found a false positive rate of monosomy of 3.8% and concluded that monosomy in biopsy results should be taken as a true representation of the status of the embryo.
Maternal age has an undisputed link with embryonic aneuploidy and the risk of various trisomies in the fetus. Women, who are older than 37 years, present reduced fertility due to their ageing oocytes being prone to various chromosomal errors (ESHRE Capri workshop group, 2005). This increased aneuploidy is observed in preimplantation embryos but to a much more severe degree than in prenatal studies (Munne et al., 2002; Munne, 2003). The impact of maternal age on the pregnancy rate in this study can be seen in Table 1 where the pregnancy rate drops to 19% for women 37 years and older. However, overall maternal age is not a major factor affecting the frequency of all abnormalities seen in preimplantation embryos and oocytes (Delhanty et al., 1997; Bielanska et al., 2002; Baart et al., 2006; Fragouli et al., 2006a). From our results, it can be seen that chaotic mosaic and other types of mosaic embryos seem to occur irrespective of age and reproductive history in roughly the same proportion for all groups. The frequency of meiotic abnormalities in embryos from older women was not significantly different from those in the younger age groups (Fig. 1). This shows that although older females in general have higher rates of meiotic abnormalities, some younger females (25–35 years of age) going through this PGS programme have an almost equally high chance of a meiotic chromosomal abnormality. That is probably an explanation of why the younger women that have resorted to PGS in this study present fertility problems beyond those that are encountered in routine IVF.
Of course, these meiotic abnormalities may come from either parent; however, it has been noted in another recent study of ovum donors that younger women can present high aneuploidy rate in their oocytes (Munne et al., 2006). We can conclude that maternal age alone in these couples is not enough of an indicative parameter on which to base a prognosis for genetic abnormalities in their embryos. Rather, there might be other parameters either related to IVF processes and/or in their genetic make up that predisposes them to an increased risk of aneuploidy in their gametes or embryos (Warren and Gorringe, 2006).
In contrast to maternal age, the different referral groups have some significant differences in the types of chromosomal abnormalities found in their embryos. The most striking differences are in the number of embryos with meiotic abnormalities and uniformly abnormal embryos found in the AMA, RM and RIF groups (Table 5). The distribution of these two types and of the diploid/chaotic mosaics is significantly different between the groups. Only 4 out of 212 embryos in the RIF referral group were uniformly abnormal (Table 5), pointing to the fact that whatever the genetic make-up of the zygote; in this group, in particular, post-zygotic errors are a major factor in their subsequent demise.
Although the cycles for the RIF group are much greater in number compared with the other two refferal groups, a satisfactory number of embryos were studied for all three groups. Additionally, the fact that only 20 out 212 embryos showed any identified meiotic errors for the RIF is a cause for further investigation in order to establish the cause of infertility in this group.
The few identifiable meiotic abnormalities in the RIF embryos is another feature that distinguishes this group from the other two. A study by Voullaire et al. (2002) also indicated that RIF is not associated with an increased level of meiotic aneuploidy. Again, this fact points to post-fertilization errors that are almost universal via a mechanism that may be independent of the outcome of parental meiosis but is probably inherited by the embryos at a molecular level. For example, in a study by Kay et al. (2006), it was suggested that a polymorphism (Pro72) of the p53 tumour suppressor gene is associated with RIF, by an unspecified mechanism. More research in this area will probably be very useful in identifying the couples at risk of RIF and may provide some answers as to the molecular mechanisms involved in this type of infertility.
Moreover, most of the total ICSI cycles in this study were performed for couples with RIF, since it was our biggest referral group. Comparing the ICSI and IVF cycles within the RIF group, it was found that there was no significant difference in any of the results. This suggests that the pattern of embryonic abnormalities highlighted in this study for RIF couples was not solely a consequence of the insemination method or of poor sperm parameters, but a characteristic of all couples in this group.
In contrast, the RM and AMA groups seem to have marked similarities in almost every aspect of the chromosomal abnormalities found in their embryos, although their reproductive history was completely different. Most RM patients were able to conceive, since most of the previous spontaneous abortions were from natural cycles, their problem was to achieve an ongoing pregnancy. The meiotic errors within the RM group were spread across all ages; the age group 29–36 years had an average of 29% of embryos with meiotic errors and the 37–42 age group had 24% on average. This compares with the AMA group that included women over 40 who were unable to conceive naturally and needed IVF, which had on average 25% of embryos with meiotic errors. In contrast, Baart et al. (2006) reported follow-up results from 83 embryos from IVF couples (average maternal age 33.1) with no indication for PGS and found an incidence of meiotic errors of 23%. The study of individual chromosomes in embryos with errors does suggest further similarities in the RM and AMA groups; the most frequent meiotic abnormalities were in chromosomes 21 and 18 in the AMA group and in 21, 13 and 18 in the RM group.
These results indicate an unidentified underlying common mechanism that links the infertility in these two groups (RM and AMA) and is worth investigating further. The RM group appear to be affected by an age independent predisposition to aneuploidy as detected by comprehensive studies of human oocytes (Fragouli et al., 2006b). Furthermore, studies on meiotic recombination in human and mouse gametes provide some evidence of a ‘genetic background’ effect in the causes of aneuploidy (reviewed in Lynn et al., 2004; Hunt, 2006). So, it might be that the similarities in these groups of patients observed in this study are the result of a genetic predisposition in the younger women.
The chromosome error rates in the preimplantation stage are still under investigation but findings in one study (Munne et al., 2004) show that monosomy may be more common than trisomy and the chromosomes most affected overall are 22, 16, 21 and 15. However, that study was unable to distinguish between errors due to meiosis and those of post-zygotic origin. In our study, meiotic errors most frequently affected chromosomes 21, 18 and 22, in that order, whereas overall, those most frequently affected were chromosomes 22, 21, 16, 18 and 15. We understand that the numbers of individual chromosome error rates are still too small to allow for any definitive conclusions but they do present us with a starting point on which to base further investigations. Few previous studies have been sufficiently detailed to provide precise information on meiotic errors; this highlights the need for more detailed investigations in a large number of human embryos in order to understand the causes of chromosomal errors in preimplantation development.
Another observation in this study is the presence of a majority of fully chaotic embryos in all groups of patients irrespective of maternal age or reproductive history. Delhanty et al. (1997), studying fertile patients undergoing PGD, found the extensive generation of fully chaotic embryos to be patient specific, an observation confirmed by Voullaire et al. (2002). From this observation, we can conclude that our cohort of couples undergoing PGS all have a predisposition to generating chaotic embryos. The developmental potential of the untransferred embryos was also briefly assessed, and from this and previous studies, it is obvious that varying degrees of mosaicism in embryos does not prohibit blastocyst or morula development (Magli et al., 2000; Ruangvutilert et al., 2000; Li et al., 2005). In this study, no blastocyst was uniformly abnormal and half of the mosaic blastocysts had a diploid cell line. Around 30% of embryos from the RM group reached the blastocyst or morula stage, a frequency higher than in the other two groups, which will fit in with the higher implantation potential of the embryos but the lack of progression in pregnancy.
|
Conclusions |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
To conclude, our study is one of the most detailed investigations to date into chromosomal abnormalities in embryos from PGS cycles for couples with poor reproductive history. Follow-up data proved a valuable source of information first for confidence in our screening programme and secondly for the analysis of mechanisms leading to the chromosomal abnormalities in the embryos. The efficient screening of six chromosomes appears to be highly effective and the low frequency of embryos diagnosed as normal has no adverse effect on the pregnancy rate. The differences between the types of chromosomal abnormalities in embryos from couples of the three major PGS referral groups (AMA, RM and RIF) were investigated in great detail. RM and AMA embryos show consistent similarities that would imply a common underlying mechanism in the causation of aneuploidy. Embryos from the RIF group show a significantly lower incidence of meiotic origin errors and almost universal post-zygotic errors implying that mitosis and not meiosis is more error prone in this group. Differences in error rates in individual chromosomes for each of the groups also suggest a pattern of different chromosomal susceptibilities in each group.
|
Acknowledgements |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
The authors would like to thank all the remaining members of the Human Genetics and Embryology Group at University College, London and all the Staff at the Assisted Conception Unit, UCLH, in particular Ozkan Ozturk and Iffat Khadum. We would also like to thank Prof. Charles Rodeck for all the help and encouragement, he has provided.
|
Footnotes |
|---|
4 Present address: Department of Obstetrics and Gynaecology, Yale University Medical School, New Haven, CT 06510, USA
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
Baart EB, Martini E, van den Berg I, Macklon NS, Galjaard RJ, Fauser BC, Van Opstal D. Preimplantation genetic screening reveals a high incidence of aneuploidy and mosaicism in embryos from young women undergoing IVF. Hum Reprod (2006) 21:223–233.
Bean CJ, Hassold TJ, Judis L, Hunt PA. Fertilization in vitro increases non-disjunction during early cleavage divisions in a mouse model system. Hum Reprod (2002) 17:2362–2367.
Bielanska M, Tan SL, Ao A. Chromosomal mosaicism throughout human preimplantation development in vitro: incidence, type and relevance to embryo outcome. Hum Reprod (2002) 17:413–419.
Carrell DT, Liu L, Huang I, Peterson CM. Comparison of maturation, meiotic competence, and chromosome aneuploidy of oocytes derived from two protocols for in vitro culture of mouse secondary follicles. J Assist Reprod Genet (2005) 22:347–354.[CrossRef][Web of Science][Medline]
Cooper ML, Darilek S, Wun WS, Angus SC, Mensing DE, Pursley AN, Dunn RC, Grunert GM, Cheung SW. A retrospective study of preimplantation embryos diagnosed with monosomy by fluorescence in situ hybridization (FISH). Cytogenet Genome Res (2006) 114:359–366.[CrossRef][Web of Science][Medline]
Daphnis DD, Delhanty JD, Jerkovic S, Geyer J, Craft I, Harper JC. Detailed FISH analysis of day 5 human embryos reveals the mechanisms leading to mosaic aneuploidy. Hum Reprod (2005) 20:129–137.
Delhanty JD. Preimplantation genetics: an explanation for poor human fertility? Ann Hum Genet (2001) 65(Pt 4):331–338.[CrossRef][Web of Science][Medline]
Delhanty JD, Harper JC, Ao A, Handyside AH, Winston RM. Multicolour FISH detects frequent chromosomal mosaicism and chaotic division in normal preimplantation embryos from fertile patients. Hum Genet (1997) 99:755–760.[CrossRef][Web of Science][Medline]
Donoso P, Staessen C, Fauser BC, Devroey P. Current value of preimplantation genetic aneuploidy screening in IVF. Hum Reprod Update (2007) 13:15–25.
ESHRE. Capri Workshop Group. Fertility and ageing. Hum Reprod Update (2005) 11:261–276.
Fragouli E, Wells D, Thornhill A, Serhal P, Faed MJ, Harper JC, Delhanty JD. Comparative genomic hybridization analysis of human oocytes and polar bodies. Hum Reprod (2006a) 21:2319–2328.
Fragouli E, Wells D, Whalley KM, Mills JA, Faed MJ, Delhanty JD. Increased susceptibility to maternal aneuploidy demonstrated by comparative genomic hybridization analysis of human MII oocytes and first polar bodies. Cytogenet Genome Res (2006b) 114:30–38.[CrossRef][Web of Science][Medline]
Gardner DK, Reed L, Linck D, Sheehan C, Lane M. Quality control in human in vitro fertilization. Semin Reprod Med (2005) 23:319–324.[CrossRef][Web of Science][Medline]
Gianaroli L, Magli MC, Ferraretti AP, Fiorentino A, Garrisi J, Munne S. Preimplantation genetic diagnosis increases the implantation rate in human in vitro fertilization by avoiding the transfer of chromosomally abnormal embryos. Fertil Steril (1997) 68:1128–1131.[CrossRef][Web of Science][Medline]
Gianaroli L, Magli MC, Ferraretti AP, Tabanelli C, Trengia V, Farfalli V, Cavallini G. The beneficial effects of preimplantation genetic diagnosis for aneuploidy support extensive clinical application. Reprod Biomed Online (2005) 10:633–640.[Web of Science][Medline]
Harper JC, Coonen E, Ramaekers FC, Delhanty JD, Handyside AH, Winston RM, Hopman AH. Identification of the sex of human preimplantation embryos in two hours using an improved spreading method and fluorescent in-situ hybridization (FISH) using directly labelled probes. Hum Reprod (1994) 9:721–724.
Harper JC, Boelaert K, Geraedts J, Harton G, Kearns WG, Moutou C, Muntjewerff N, Repping S, SenGupta S, Scriven PN, et al. ESHRE PGD Consortium data collection V: cycles from January to December 2002 with pregnancy follow-up to October 2003. Hum Reprod (2006) 21:3–21.
Hopman AH, Ramaekers FC, Raap AK, Beck JL, Devilee P, van der Ploeg M, Vooijs GP. In situ hybridization as a tool to study numerical chromosome aberrations in solid bladder tumors. Histochemistry (1988) 89:307–316.[CrossRef][Web of Science][Medline]
Hunt PA. Meiosis in mammals: recombination, non-disjunction and the enviroment. Biochem Soc Trans (2006) 34:574–577.[CrossRef][Web of Science][Medline]
Kay C, Jeyendran RS, Coulam CB. p53 tumour suppressor gene polymorphism is associated with recurrent implantation failure. Reprod Biomed Online (2006) 13:492–496.[Web of Science][Medline]
Li M, DeUgarte CM, Surrey M, Danzer H, DeCherney A, Hill DL. Fluorescence in situ hybridization reanalysis of day-6 human blastocysts diagnosed with aneuploidy on day 3. Fertil Steril (2005) 84:1395–1400.[CrossRef][Web of Science][Medline]
Lynn A, Ashley T, Hassold T. Variation in human meiotic recombination. Ann Rev Genomics Hum Genet (2004) 5:317–349.[CrossRef][Web of Science][Medline]
Magli MC, Jones GM, Gras L, Gianaroli L, Korman I, Trounson AO. Chromosome mosaicism in day 3 aneuploid embryos that develop to morphologically normal blastocysts in vitro. Hum Reprod (2000) 15:1781–1786.
Munne S. Preimplantation genetic diagnosis and human implantation—A review. Placenta (2003) 24::S70–S76.[CrossRef][Web of Science][Medline]
Munne S, Sandalinas M, Escudero T, Marquez C, Cohen J. Chromosome mosaicism in cleavage-stage human embryos: evidence of a maternal age effect. Reprod Biomed Online (2002) 4:223–232.[Medline]
Munne S, Bahce M, Sandalinas M, Escudero T, Marquez C, Velilla E, Colls P, Oter M, Alikani M, Cohen J. Differences in chromosome susceptibility to aneuploidy and survival to first trimester. Reprod Biomed Online (2004) 8:81–90.[Web of Science][Medline]
Munne S, Chen S, Fischer J, Colls P, Zheng X, Stevens J, Escudero T, Oter M, Schoolcraft B, Simpson JL, et al. Preimplantation genetic diagnosis reduces pregnancy loss in women aged 35 years and older with a history of recurrent miscarriages. Fertil Steril (2005) 84:331–335.[CrossRef][Web of Science][Medline]
Munne S, Ary J, Zouves C, Escudero T, Barnes F, Cinioglu C, Ary B, Cohen J. Wide range of chromosome abnormalities in the embryos of young egg donors. Reprod Biomed Online (2006) 12:340–346.[Web of Science][Medline]
Redding GP, Bronlund JE, Hart AL. The effects of IVF aspiration on the temperature, dissolved oxygen levels, and pH of follicular fluid. J Assist Reprod Genet (2006) 23:37–40.[CrossRef][Web of Science][Medline]
Ruangvutilert P, Delhanty JD, Serhal P, Simopoulou M, Rodeck CH, Harper JC. FISH analysis on day 5 post-insemination of human arrested and blastocyst stage embryos. Prenat Diagn (2000) 20:552–560.[CrossRef][Web of Science][Medline]
Rubio C, Rodrigo L, Perez-Cano I, Mercader A, Mateu E, Buendia P, Remohi J, Simon C, Pellicer A. FISH screening of aneuploidies in preimplantation embryos to improve IVF outcome. Reprod Biomed Online (2005) 11:497–506.[Web of Science][Medline]
Voullaire L, Slater H, Williamson R, Wilton L. Chromosome analysis of blastomeres from human embryos by using comparative genomic hybridization. Hum Genet (2000) 106:210–217.[CrossRef][Web of Science][Medline]
Voullaire I, Wilton L, McBain J, Callaghan T, Williamson R. Chromosome abnormalities identified by comparative genomic hybridization in embryos from women with repeated implantation failure. Mol Hum Reprod (2002) 8:1035–1041.
Warren WD, Gorringe KL. A molecular model for sporadic human aneuploidy. Trends Genet (2006) 22:218–224.[CrossRef][Web of Science][Medline]
Wells D, Delhanty JD. Comprehensive chromosomal analysis of human preimplantation embryos using whole genome amplification and single cell comparative genomic hybridization. Mol Hum Reprod (2000) 6:1055–1062.
Wilton L. Preimplantation genetic diagnosis for aneuploidy screening in early human embryos: a review. Prenat Diagn (2002) 22:512–518.[CrossRef][Web of Science][Medline]
Submitted on January 16, 2007; resubmitted on February 20, 2007; resubmitted on February 20, 2007; accepted on March 28, 2007.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  C. Hanson, T. Hardarson, K. Lundin, C. Bergh, T. Hillensjo, J. Stevic, C. Westin, U. Selleskog, L. Rogberg, and M. Wikland Re-analysis of 166 embryos not transferred after PGS with advanced reproductive maternal age as indication Hum. Reprod., November 1, 2009; 24(11): 2960 - 2964. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Fragouli, M. Lenzi, R. Ross, M. Katz-Jaffe, W.B. Schoolcraft, and D. Wells Comprehensive molecular cytogenetic analysis of the human blastocyst stage Hum. Reprod., November 1, 2008; 23(11): 2596 - 2608. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Munne, J. Cohen, J. L. Simpson, L. J. Wilton, A. H. Handyside, A. R. Thornhill, S. Mastenbroek, F. van der Veen, and S. Repping In Vitro Fertilization with Preimplantation Genetic Screening N. Engl. J. Med., October 25, 2007; 357(17): 1769 - 1771. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||