- Email: [email protected]
The in vivo and in vitro efficiency and efficacy of PGD for aneuploidy
Molecular and Cellular Endocrinology 183 (2001) S13– S18 www.elsevier.com/locate/mce
The in vivo and in vitro efficiency and efficacy of PGD for aneuploidy Luca Gianaroli *, M.C. Magli, A.P. Ferraretti S.I.S.ME.R., Reproducti6e Medicine Unit, Via Mazzini 12, 40138 Bologna, Italy
Abstract Preimplantation genetic diagnosis for aneuploidy was implemented on 1782 morphologically normal embryos generated in vitro by patients with a poor prognosis of pregnancy. Only 592 of them (34%) were diagnosed as chromosomally normal. Embryo transfer was accomplished in 240 cycles resulting in 79 clinical pregnancies (33%) and an implantation rate of 22.6%. The in vitro efficiency of the procedure was established by analysing all the blastomeres obtained from 311 non transferrable embryos and resulted to be 97.1%. The in vivo efficiency of the technique was calculated with the data derived from the prenatal diagnoses by examination of the infants at birth and was 97.8%. In consideration of the reported inaccuracy rate, patients are still recommended to undergo prenatal diagnosis. The transfer of PGD selected embryos in women of advanced reproductive age reduces by half the risk of having a trisomic pregnancy. © 2001 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Aneuploidy; Implantation rate; Multicolour FISH; Poor prognosis patients; Preimplantation genetic diagnosis
1. Introduction The clinical application of Preimplantation Genetic Diagnosis (PGD) is aimed to provide couples at high reproductive risk a better chance of conceiving an unaffected child by identifying those embryos which carry genetic or chromosomal abnormalities and preventing them to be transferred. The first healthy pregnancy was reported in 1990 (Handyside et al., 1990) and originated a strong incentive to the more and more frequent application of PGD in in vitro fertilisation (IVF) techniques. A special application of PGD is represented by the screening of aneuploidy in embryos generated by patients with a poor prognosis of full term pregnancy. The technique is recommended to couples at risk of developing chromosomally abnormal embryos with the aim of improving their chances of conceiving after IVF. Identification of the categories of patients for which this approach should be routinely proposed is still under investigation. However, the results obtained in the recent years permitted to derive valuable information and draw some conclusions. * Corresponding author. Tel.: + 39-051-307307; fax: +39-051302933. E-mail address: [email protected] (L. Gianaroli).
Advanced maternal age, an altered karyotype, and recurrent abortions are indications to PGD of aneuploidy for which a clinical advantage has been demonstrated. In these cases, the selection and transfer of chromosomally normal embryos yields a higher implantation rate and a reduced incidence of spontaneous abortions compared to the controls (Gianaroli et al., 1997a; Simon et al., 1998; Gianaroli et al., 1999a; Munne´ et al., 1999). Still under investigation are: patients with unexplained, repeated IVF failures, and MESA-TESE patients (Gianaroli et al., 2000). According to the preliminary results the transfer of PGD selected embryos does not increase substantially the clinical outcome in these patients despite the high incidence of chromosomal abnormalities in their embryos (Gianaroli et al., 1999a). This finding suggests that other concomitant factors may be involved in this poor prognosis condition; among them the possibility that other chromosomes, different from those currently investigated, play a role in determining blockage of development in these categories of embryos (Bahc¸e et al., 1999; Gianaroli et al., 1999a). Alternatively, alterations in the mechanisms entering mitotic divisions could be involved, as suggested by the finding of mosaicism and haploidy/polyploidy as the most frequent abnormalities observed, possibly due to numerical or functional centriolar defects (Gianaroli et al., 1997b).
0303-7207/01/$ - see front matter © 2001 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 3 0 3 - 7 2 0 7 ( 0 1 ) 0 0 5 7 0 - 6
S14
L. Gianaroli et al. / Molecular and Cellular Endocrinology 183 (2001) S13–S18
All these studies have been performed in concomitance with the continuos effort of maximising the efficiency inherent to the chromosomal diagnosis performed on a single cell. A system of quality control has been implemented combined to the design of adequate scoring criteria aimed at reducing the inaccuracy rate to the lowest level (Magli et al., 1998; Munne´ et al., 1998a; Gianaroli et al., 1999b). In the present study, the results obtained by the clinical application of PGD for aneuploidy during 3 years and a half of activity are presented and discussed. Special emphasis is dedicated at verifying the in vitro and the in vivo efficiency of the technique.
2. Materials and methods
2.1. Patients From September 1996 to April 2000, 255 patients underwent induction of multiple follicular growth for infertility with PGD for aneuploidy at the S.I.S.ME.R. Reproductive Medicine Unit in Bologna. In all they performed 343 PGD cycles. Indications for the study were: (a) maternal age]36 years (176 cycles); (b)] 3 previous IVF failures (60 cycles); (c) an altered karyotype detected in peripheral blood (47 cycles) due to gonosomal mosaicism or balanced translocations; (d) MESA-TESE patients with]1 IVF failure (31 cycles); (e) other indications (29 cycles). Controlled ovarian hyperstimulation was accomplished by exogenous gonadotropin administration after a desensitisation protocol with long-acting GnRH analogues (Ferraretti et al., 1996). At approximately 36 h following the injection of HCG, oocytes were retrieved transvaginally by ultrasound guidance and incubated in Earle’s balanced salt solution (EBSS) supplemented with 10% heat inactivated maternal serum (MS), in a 5% CO2 humidified atmosphere at 37 °C. Oocytes were fertilised by ICSI or by brief exposure to spermatozoa for IVF, depending on sperm sample requirements (Gianaroli et al., 1996).
2.2. Fertilisation and embryo de6elopment assessment Oocytes were checked at 14– 18 h after insemination for the presence of pronuclei and polar bodies. Regularly fertilised oocytes were cultured individually in EBSS 15% MS and scored at 40 h after insemination for the number and appearance of nuclei and blastomeres, and the degree of fragmentation. Day 3 embryos with]4 cells and5 50% fragmentation were selected for embryo biopsy that was performed at 62– 64 h after insemination. After blastomere removal, embryos were transferred to blastocyst culture medium (Scandinavian IVF Sc., Goteborg, Sweden) and cul-
tured for 1 or 2 additional days. Only embryos diagnosed as chromosomally normal were transferred. Transfers were performed into the uterine cavity generally on day 4 (Gianaroli et al., 1999b). For 15 cycles the transfer was delayed to day 5 to select among the euploid embryos those with the best morphology. This approach is adopted at S.I.S.ME.R. when the number of transferrable embryos exceeds two on day 4, with the additional advantage of reducing the need of cryopreserving spare embryos (Magli et al., 1999). Clinical pregnancies were established by the presence of a gestational sac with fetal heartbeat revealed by ultrasound analysis. The implantation rate defined the ratio between the number of gestational sacs with fetal heartbeat and the total number of embryos transferred. When this calculation is done for pregnant patients (the denominator is represented by the total number of embryos transferred in the pregnant patients) the percentage obtained represents the implantation rate per pregnant patient (Gianaroli et al., 1999a).
2.3. Blastomere biopsy and fluorescence in situ hybridisation (FISH) Day 3 embryos selected for FISH analysis were manipulated individually in HEPES-buffered medium overlaid with pre-equilibrated mineral oil. A breach of 20–22 mm was opened in the zona pellucida with acidic Tyrode’s solution and a nucleated blastomere was gently aspirated by a polished glass needle (40 mm diameter). The biopsied embryo was immediately washed and returned to culture. The removed blastomeres was transferred to a hypotonic solution (1% sodium citrate), the nucleus was fixed on a glass slide using methanol– acetic acid 3:1, and dehydrated sequentially in 70, 85 and 100% ethanol series. Multicolour FISH was performed in one step (486 embryos, of which 141 were analysed for XY, 13, 18, 21 and 345 for XY, 13, 16, 18, 21) or two step protocol (1134 embryos, all analysed for XY, 13, 15, 16, 18, 21 and 22; in 447 embryos chromosome 14 was also screened; in 158 embryos the probe for chromosome 14 was replaced by the probe specific for chromosome 17, and by the one detecting chromosome 1 in 391 embryos; in the last 138 embryos no extra chromosomes were tested as another probe specific for chromosome 21 was added in the second panel to have this chromosome screened twice). The selection of chromosomes to be screened was based on the results of spontaneous abortions and live births (Munne´ et al., 1998b). Other chromosomes that do not seem to have clinical relevance were also analysed with the aim of verifying whether this is also true during the first embryonic stages. After co-denaturation of probes and nuclear DNA, hybridisation for approximately 4 h at 37 °C in a moist chamber, and washing of unbound probe, antifade
L. Gianaroli et al. / Molecular and Cellular Endocrinology 183 (2001) S13–S18 Table 1 Overall FISH and clinical results in PGD cycles
S15
3. Results
solution was added and fluorescence was evaluated with a Olympus microscope (Olympus, Tokyo, Japan) at 60 × magnification. In the two-step protocol, the procedure was repeated as already described.
Table 1 depicts the overall outcome of the 343 treatment cycles included in the study. In all, 2096 embryos were generated, and 1782 were selected on the basis of their morphological appearance to undergo FISH analysis. A diagnosis was obtained for 1758 embryos (99%). Only 592 of them were classified as chromosomally normal (34%) whereas the remaining 1166 (66%) exhibited chromosomal abnormalities which were not compatible with either implantation or an healthy pregnancy. Consequently, embryo transfer was cancelled in 103 of the started PGD cycles (30%) where no euploid embryos were detected. An average of 2.090.8 embryos were transferred to the remaining 240 cycles resulting in 79 clinical pregnancies (33%) and an overall implantation rate of 22.6%. Spontaneous abortion occurred in seven pregnancies establishing an abortion rate after PGD of 9%.
2.4. Embryo spreading
3.1. In 6itro efficiency
The cells constituting 311 non transferrable embryos were spread after removing the zona pellucida by a short incubation in acidic Tyrode’s solution. The nuclei were fixed on a glass slide and hybridised with the same probes used for the one-cell diagnosis following the same protocol as above. The results obtained were compared to the original diagnosis to evaluate the inaccuracy rate of the technique.
As represented in Table 2, 311 non transferrable embryos were spread to FISH analyse all their blastomeres. Most of these embryos (271) were chromosomally abnormal based on the result obtained from one-cell analysis. The remaining 40 embryos had been diagnosed as euploid; however they were not transferred or cryopreserved due to blockage of development or increased fragmentation in the 2 days in culture after blastomere biopsy. Confirmation of the one-cell diagnosis was demonstrated for 185 embryos (92%). Of the remaining 26 embryos, 17 had been classified by PGD on day 3 as chromosomally abnormal; the analysis of all their blastomeres revealed that they carried an abnormality that was different from the one diagnosed by
No. of of cycles Age (mean 9SD, years) No. of of generated embryos No. of FISH analysed embryos No. of FISH diagnosed embryos FISH normal (%) FISH abnormal (%) No. of embryos transferred (mean 9 SD) No. of cycles transferred No. of clinical pregnancies (%) Aborted (%) Implantation rate (%) a
343 36.4 94.3 2096 1782 1758 592 (34) 1166 (66) 2.0 90.8 240 (70) 79 (33) 10a 28.4
One ectopic; one after amniocentesis; one therapeutic.
2.5. Statistical analysis Results were evaluated by Student’s t-test and by chi-square analysis applying the Yates’ correction, 2× 2 contingency tables. Table 2 Embryo spreading and FISH re-analysis of all the blastomeres generated PGD
No. of embryos
Re-analysis Confirmed
Normal Monosomic One chromosome Two chromosomes Three chromosomes Trisomic One chromosome Two chromosomes Three chromosomes Haploid Polyploid Complex abnormalities Total
40 85 60 21 4 85 75 9 1 12 20 69 311
40 (100) 72 52 19 1 79 70 9 – 10 17 67 285 (92)
Normal
Other abnormality
– 3 1 2 – 3 3 – – – 1 2
– 10 7 – 3 3 2 1 2 2 –
9 (3)
17 (5)
The results are compared to those obtained by PGD on one cell. Values in parentheses are percentages.
S16
L. Gianaroli et al. / Molecular and Cellular Endocrinology 183 (2001) S13–S18
Table 3 Delivery outcome Deli6eries Singleton Twin Triplet Gestational weeks (M 9SD) Singleton Twin Triplet Deli6ery mode Cesarian section Vaginal
57 39 16 2 37.5 9 3 37.9 9 2.5 36.89 1.8 37 46 11
one-cell analysis. This represents a 5% inaccuracy that does not have any clinical relevance, as these embryos had been classified not suitable for transfer. Conversely, nine embryos (3%) were assessed as abnormal by PGD but were found to be chromosomally normal after re-analysis of all the blastomeres. According to these data, the in vitro inaccuracy of the technique is 3%.
3.2. Pregnancy outcome The outcome of the 79 pregnancies is the following: 57 women already delivered, and 12 have a regularly ongoing pregnancy. In ten cases the pregnancy was interrupted: spontaneous abortion occurred in seven cases, one pregnancy was ectopic, one aborted after amniocentesis (the fetal karyotype was normal), and one underwent therapeutic abortion because at prenatal diagnosis the foetus was diagnosed to have trisomy 21. Finally, two couples decided to undergo selective fetal reduction, and already delivered healthy babies. Table 3 summarises the main characteristics of the 57 deliveries, with 39 singleton, 16 twins, and two triplet. Cesarian section was the most frequent mode of delivery (46 vs 11).
3.3. In 6i6o efficiency All the pregnant patients were recommended to undergo conventional prenatal diagnosis. Sixteen out of 79 (20%) decided not to follow the clinicians’ recommendation for the following reasons: 12 couples claimed to be afraid of the risk of abortion inherent to the procedure; two were against the possibility of therapeutic abortion; and two adduced other reasons involving ethical and moral implications. The collection of the information derived by prenatal diagnosis permitted to verify the exactness of PGD, and to establish the in vivo efficiency of the technique. The results and conclusions are reported in Tables 4 and 5. The karyotype of 91 foetuses was assessed by either chorionic villus sampling (CVS) or amniocentesis, or by examination of the infant at birth. Confirmation of the
diagnosis performed by PGD was obtained in 89 cases with an in vivo efficiency corresponding to 97.8%. In the remaining two foetuses, trisomy 21 was detected: in one case the patient, screened for the translocation 13:14, underwent therapeutic abortion, whereas in the other case the infant was delivered as the couple decided not to perform any prenatal diagnosis. In consideration of the results obtained and based on the general data from prenatal diagnosis, a theoretical calculation was made on the risk of trisomic foetuses (Table 6). This was made with the purpose of estimating the possible advantage of transferring FISH selected embryos. The calculation was based on the number of trisomic embryos detected in the pregnant patients that could develop to the blastocyst stage (seven and 13 depending on the age of the patient). According to the implantation rate per pregnant patient, 4.9 and 7.8 trisomic embryos, respectively were expected to implant. If these figures are multiplied by a correction factor derived by the general data from CVS, 0.2 and 4.0 are the numbers of trisomic embryos expected at CVS in the absence of PGD in the two age categories, compared to zero and two trisomic foetuses, respectively generated after PGD.
4. Discussion PGD for aneuploidy was integrated in IVF with the main purpose of increasing the chances of full term pregnancy in poor prognosis patients, mainly women in advanced reproductive age (Munne´ et al., 1993; Verlinsky and Kuliev, 1996; Gianaroli et al., 1997b). It was a big challenge, because such an invasive and expensive procedure would be implemented based on the hope of overcoming the age effect by selecting for transfer only those embryos having a normal chromosomal complement. The results obtained after several years of intense work done by a few centers worldwide, have permitted to establish that a positive outcome is associated to the transfer of FISH selected embryos in poor prognosis patients (Gianaroli et al., 1999a; Munne´ et al., 1999). Concomitantly, special attention has been dedicated to monitor the quality of the results obtained by multicolour FISH on a single cell. The efficiency of the technique has been continuously evaluated by reTable 4 PGD of aneuploidy: in vivo efficiency No. of gestational sacs Aborted Selective reduction Ongoing On term CVS, chorionic villus sampling; AC, amniocentesis.
110 17 4 12 77
L. Gianaroli et al. / Molecular and Cellular Endocrinology 183 (2001) S13–S18
S17
Table 5 PGD of aneuploidy: in vivo efficiency Confirmed: 89 (97.8%)
Not confirmed: (2.2%)
Unknown
At CVS/AC
At birth
At CVS/AC
At birth
–
Ectopic Misacrriages Late abortion Ongoing On term
– – 5a 8 53
– – – – 23
– – 1 – –
– – – – 1
1 14 – 4b –
Total
66
23
1
1
19
CVS, chorionic villus sampling; AC, amniocentesis. a Four sacs reduced; one normal foetus lost after AC. b Waiting for prenatal diagnosis results.
analysing as many embryos as possible among those that are classified as non transferrable on the basis of the FISH results or their morphology on day 4 or 5. The data obtained permit not only to establish the current inaccuracy rate of the technique, but also to define and propose solutions to the most common situations that may represent a source of incorrect diagnosis. As represented in Table 2, most of the discrepancies involve false monosomies. Loss of micronuclei during fixation, failed hybridisation and signal overlapping represent common causes of misdiagnosis. A significant reduction in the loss of micronuclei was derived through a modification in the fixation protocol that substantially improved the performance of the technique (Munne´ et al., 1998a). In addition, a more accurate fixation of the nuclei and the use of pure fluorochromes, instead of a combination of them, has proven to minimise the risk of signal overlapping (Munne´ et al., 1998b). Failed hybridisation is probably less common; however, due to its clinical relevance, a different probe specific for chromosome 21 is now added also in the second round of hybridisation, in order to make it tested twice. This strategy has been recently adopted and its performance is still under evaluation. The data yielded by embryo spreading and FISH re-analysis also enabled the formulation of scoring criteria for the correct interpretation of FISH signals. The most confounding situation is when the target splits into two signals producing false positive results. The least error-prone criterion states that two signals represent two homologue chromosomes when their distance apart is at least two domains diameters (Munne´ et al., 1998a). Some probes have a higher tendency to split (i.e. the probes for chromosomes 18 and Y) and this must be taken into consideration for the interpretation of the correspondent signals. Finally, mosaicism that is very common in in vitro human embryos, did not seem to represent a frequent source of incorrect diagnosis maybe due to the fact that
the use of multiple probes enabled the identification of complex abnormalities. Interestingly, all the 40 embryos diagnosed as normal by PGD were confirmed after re-analysis; although this figure is relatively small, this finding suggests that the frequency of mosaic embryos with euploid cells could be very low. In consideration of the current efficiency of FISH performed on single cell, all the patients entering a PGD cycle at S.I.S.ME.R. are thoroughly informed of the need to undergo conventional prenatal diagnosis in case of pregnancy. This approach permits not only to verify and complete the diagnosis done on preimplantation embryos in patients that are exposed at a very high risk of generating aneuploid embryos, but also to establish the in vivo efficiency of PGD. This parameter is extremely useful for the team, since it represents the true evaluation of the efficacy of the technique. According to data presented in this study, the in vitro efficiency is 97.1%; a 97.8% in vivo efficiency represents the expected figure within the limits of the technique. In addition, it is important to remind that with an inaccuTable 6 Number of trisomic foetuses after PGD compared to the expected in the absence of PGD: in vivo efficiency of the technique 535
]36
No. of diagnosed embryos in pregnant patients 207 No. of trisomic embryos 24 Blastocyst development correction factora 0.3 No. of expected trisomic blastocysts 7 IRPP correction factorb 70.4 No. of expected implantations 4.9 CVS correction factorc 0.3 No. of expected trisomic foetuses at CVS 0.2 No. of observed trisomic foetuses at CVS after 0 PGD
284 45 0.3 13 59.8 7.8 5.1 4.0 2
Age (years)
CVS, chorionic villus sampling; IRPP, implantation rate per pregnant patient. a Magli et al. (2000). b Gianaroli et al. (1999). c Snijders et al. (1994).
S18
L. Gianaroli et al. / Molecular and Cellular Endocrinology 183 (2001) S13–S18
racy rate of 3%, PGD for aneuploidy still enables to significantly reduce the risk of trisomic foetuses in women of advanced age by identify 97% of the generated trisomies. As postulated in Table 6, the number of expected trisomic foetus at CVS would be four in the absence of PGD; this number was cut exactly by half with embryos transferred on the basis of FISH diagnoses. In conclusion, the experience gathered during these years in the field of PGD for aneuploidy has contributed some important points: patients with advanced reproductive age, an altered karyotype, and (possibly) recurrent abortions derive a benefit in terms of increased implantation rate and lower incidence of spontaneous abortions; for women in advanced reproductive age the risk of a trisomic pregnancy is significantly reduced by half. Nevertheless, in view of the patients’ health as the main priority, recommendation of prenatal diagnosis is still mandatory. A systematic monitoring of the pregnancies and the follow up of the babies born and their development are needed for a comprehensive evaluation of the efficacy of the technique.
References Bahc¸ e, M., Cohen, J., Munne´ , S., 1999. PGD of aneuploidy: were we looking at the wrong chromosomes? J. Assist. Reprod. Genet. 16, 176 – 181. Ferraretti, A.P., Magli, M.C., Feliciani, E., et al., 1996. Relationship of timing agonist administration in the cycle phase to the ovarian response to gonadotropins in the long-down regulation protocols for assisted reproductive technologies. Fertil. Steril. 65, 114 – 121. Gianaroli, L., Fiorentino, A., Magli, M.C., et al., 1996. Prolonged sperm – oocyte exposure and high sperm concentration affect human embryo viability and pregnancy rate. Hum. Reprod. 1, 2507 – 2511. Gianaroli, L., Magli, M.C., Ferraretti, A.P., et al., 1997a. Preimplantation genetic diagnosis increases the implantation rate in human in vitro fertilisation by avoiding the transfer of chromosomally abnormal embryos. Fertil. Steril. 68, 1128 – 1131. Gianaroli, L., Magli, M.C., Munne´ , S., et al., 1997b. Will preimplantation genetic diagnosis assist patients with a poor prognosis to achieve pregnancy? Hum. Reprod. 12, 1762 – 1767.
Gianaroli, L., Magli, M.C., Ferraretti, A.P., et al., 1999a. Preimplantation diagnosis for aneuploidies in patients undergoing in vitro fertilisation with a poor prognosis: identification of the categories for which it should be proposed. Fertil. Steril. 72, 837 –844. Gianaroli, L., Magli, M.C., Munne´ , S., et al., 1999b. Advantages of day four embryo transfer in patients undergoing preimplantation genetic diagnosis of aneuploidy. J. Assist. Reprod. Genet. 16, 170 – 175. Gianaroli, L., Magli, M.C., Ferraretti, A.P., et al., 2000. Preimplantation diagnosis after assisted reproduction techniques for genetically determined male infertility. J. Endocrinol. Invest. 23, 711 – 716. Handyside, A.H., Kontogianni, E.H., Hardy, K., et al., 1990. Pregnancies from biopsied human preimplantation embryos sexed by Y-specified DNA amplification. Nature (London) 344, 768 –770. Magli, M.C., Gianaroli, L., Munne´ , S., et al., 1998. Incidence of chromosomal abnormalities from a morphologically normal cohort of embryos in poor prognosis patients. J. Assist. Reprod. Genet. 15, 297 – 301. Magli, M.C., Gianaroli, L., Fortini, D., et al., 1999. Impact of blastomere biopsy and cryopreservation techniques on human embryo viability. Hum. Reprod. 14, 770 – 773. Magli, M.C., Jones, G., Gras, L., et al., 2000. Chromosome mosaicism in day 3 aneuploid embryos that develop to morphologically normal blastocysts in vitro. Hum. Reprod. 15, 1781 – 1786. Munne´ , S., Lee, A., Rosemwaks, Z., et al., 1993. Diagnosis of major chromosome aneuploidies in human preimplantation embryos. Hum. Reprod. 8, 2185 – 2191. Munne´ , S., Marquez, C., Magli, M.C., et al., 1998a. Scoring criteria for preimplantation genetic diagnosis of numerical abnormalities for chromosomes X, Y, 13, 16, 18, and 21. Molec. Hum. Reprod. 4, 863 – 870. Munne´ , S., Magli, M.C., Bahc¸ e, M., et al., 1998b. Preimplantation diagnosis of the aneuploidies most commonly found in spontaneous abortions and live births: XY, 13, 14, 15, 16, 18, 21, 22. Prenat. Diagn. 18, 1459 – 1466. Munne´ , S., Magli, M.C., Cohen, J., et al., 1999. Positive outcome after preimplantation diagnosis of human embryos. Hum. Reprod. 14, 2191 – 2199. Simon, C., Rubio, C., Vidal, F., et al., 1998. Increased chromosomal abnormalities in human preimplantation embryos after in vitro fertilisation in patients with recurrent miscarriage. Reprod. Fertil. Dev. 10, 87 – 92. Snijders, R.J.M., Holzgreve, W., Cuckle, H., et al., 1994. Maternal age specific risks for trisomies at 9 – 14 weeks’ gestation. Prenat. Diagn. 14, 543 – 552. Verlinsky, Y., Kuliev, A., 1996. Preimplantation diagnosis of common aneuploidies in fertile couples of advanced maternal age. Hum. Reprod. 11, 2076 – 2077.
The in vivo and in vitro efficiency and efficacy of PGD for aneuploidy Luca Gianaroli *, M.C. Magli, A.P. Ferraretti S.I.S.ME.R., Reproducti6e Medicine Unit, Via Mazzini 12, 40138 Bologna, Italy
Abstract Preimplantation genetic diagnosis for aneuploidy was implemented on 1782 morphologically normal embryos generated in vitro by patients with a poor prognosis of pregnancy. Only 592 of them (34%) were diagnosed as chromosomally normal. Embryo transfer was accomplished in 240 cycles resulting in 79 clinical pregnancies (33%) and an implantation rate of 22.6%. The in vitro efficiency of the procedure was established by analysing all the blastomeres obtained from 311 non transferrable embryos and resulted to be 97.1%. The in vivo efficiency of the technique was calculated with the data derived from the prenatal diagnoses by examination of the infants at birth and was 97.8%. In consideration of the reported inaccuracy rate, patients are still recommended to undergo prenatal diagnosis. The transfer of PGD selected embryos in women of advanced reproductive age reduces by half the risk of having a trisomic pregnancy. © 2001 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Aneuploidy; Implantation rate; Multicolour FISH; Poor prognosis patients; Preimplantation genetic diagnosis
1. Introduction The clinical application of Preimplantation Genetic Diagnosis (PGD) is aimed to provide couples at high reproductive risk a better chance of conceiving an unaffected child by identifying those embryos which carry genetic or chromosomal abnormalities and preventing them to be transferred. The first healthy pregnancy was reported in 1990 (Handyside et al., 1990) and originated a strong incentive to the more and more frequent application of PGD in in vitro fertilisation (IVF) techniques. A special application of PGD is represented by the screening of aneuploidy in embryos generated by patients with a poor prognosis of full term pregnancy. The technique is recommended to couples at risk of developing chromosomally abnormal embryos with the aim of improving their chances of conceiving after IVF. Identification of the categories of patients for which this approach should be routinely proposed is still under investigation. However, the results obtained in the recent years permitted to derive valuable information and draw some conclusions. * Corresponding author. Tel.: + 39-051-307307; fax: +39-051302933. E-mail address: [email protected] (L. Gianaroli).
Advanced maternal age, an altered karyotype, and recurrent abortions are indications to PGD of aneuploidy for which a clinical advantage has been demonstrated. In these cases, the selection and transfer of chromosomally normal embryos yields a higher implantation rate and a reduced incidence of spontaneous abortions compared to the controls (Gianaroli et al., 1997a; Simon et al., 1998; Gianaroli et al., 1999a; Munne´ et al., 1999). Still under investigation are: patients with unexplained, repeated IVF failures, and MESA-TESE patients (Gianaroli et al., 2000). According to the preliminary results the transfer of PGD selected embryos does not increase substantially the clinical outcome in these patients despite the high incidence of chromosomal abnormalities in their embryos (Gianaroli et al., 1999a). This finding suggests that other concomitant factors may be involved in this poor prognosis condition; among them the possibility that other chromosomes, different from those currently investigated, play a role in determining blockage of development in these categories of embryos (Bahc¸e et al., 1999; Gianaroli et al., 1999a). Alternatively, alterations in the mechanisms entering mitotic divisions could be involved, as suggested by the finding of mosaicism and haploidy/polyploidy as the most frequent abnormalities observed, possibly due to numerical or functional centriolar defects (Gianaroli et al., 1997b).
0303-7207/01/$ - see front matter © 2001 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 3 0 3 - 7 2 0 7 ( 0 1 ) 0 0 5 7 0 - 6
S14
L. Gianaroli et al. / Molecular and Cellular Endocrinology 183 (2001) S13–S18
All these studies have been performed in concomitance with the continuos effort of maximising the efficiency inherent to the chromosomal diagnosis performed on a single cell. A system of quality control has been implemented combined to the design of adequate scoring criteria aimed at reducing the inaccuracy rate to the lowest level (Magli et al., 1998; Munne´ et al., 1998a; Gianaroli et al., 1999b). In the present study, the results obtained by the clinical application of PGD for aneuploidy during 3 years and a half of activity are presented and discussed. Special emphasis is dedicated at verifying the in vitro and the in vivo efficiency of the technique.
2. Materials and methods
2.1. Patients From September 1996 to April 2000, 255 patients underwent induction of multiple follicular growth for infertility with PGD for aneuploidy at the S.I.S.ME.R. Reproductive Medicine Unit in Bologna. In all they performed 343 PGD cycles. Indications for the study were: (a) maternal age]36 years (176 cycles); (b)] 3 previous IVF failures (60 cycles); (c) an altered karyotype detected in peripheral blood (47 cycles) due to gonosomal mosaicism or balanced translocations; (d) MESA-TESE patients with]1 IVF failure (31 cycles); (e) other indications (29 cycles). Controlled ovarian hyperstimulation was accomplished by exogenous gonadotropin administration after a desensitisation protocol with long-acting GnRH analogues (Ferraretti et al., 1996). At approximately 36 h following the injection of HCG, oocytes were retrieved transvaginally by ultrasound guidance and incubated in Earle’s balanced salt solution (EBSS) supplemented with 10% heat inactivated maternal serum (MS), in a 5% CO2 humidified atmosphere at 37 °C. Oocytes were fertilised by ICSI or by brief exposure to spermatozoa for IVF, depending on sperm sample requirements (Gianaroli et al., 1996).
2.2. Fertilisation and embryo de6elopment assessment Oocytes were checked at 14– 18 h after insemination for the presence of pronuclei and polar bodies. Regularly fertilised oocytes were cultured individually in EBSS 15% MS and scored at 40 h after insemination for the number and appearance of nuclei and blastomeres, and the degree of fragmentation. Day 3 embryos with]4 cells and5 50% fragmentation were selected for embryo biopsy that was performed at 62– 64 h after insemination. After blastomere removal, embryos were transferred to blastocyst culture medium (Scandinavian IVF Sc., Goteborg, Sweden) and cul-
tured for 1 or 2 additional days. Only embryos diagnosed as chromosomally normal were transferred. Transfers were performed into the uterine cavity generally on day 4 (Gianaroli et al., 1999b). For 15 cycles the transfer was delayed to day 5 to select among the euploid embryos those with the best morphology. This approach is adopted at S.I.S.ME.R. when the number of transferrable embryos exceeds two on day 4, with the additional advantage of reducing the need of cryopreserving spare embryos (Magli et al., 1999). Clinical pregnancies were established by the presence of a gestational sac with fetal heartbeat revealed by ultrasound analysis. The implantation rate defined the ratio between the number of gestational sacs with fetal heartbeat and the total number of embryos transferred. When this calculation is done for pregnant patients (the denominator is represented by the total number of embryos transferred in the pregnant patients) the percentage obtained represents the implantation rate per pregnant patient (Gianaroli et al., 1999a).
2.3. Blastomere biopsy and fluorescence in situ hybridisation (FISH) Day 3 embryos selected for FISH analysis were manipulated individually in HEPES-buffered medium overlaid with pre-equilibrated mineral oil. A breach of 20–22 mm was opened in the zona pellucida with acidic Tyrode’s solution and a nucleated blastomere was gently aspirated by a polished glass needle (40 mm diameter). The biopsied embryo was immediately washed and returned to culture. The removed blastomeres was transferred to a hypotonic solution (1% sodium citrate), the nucleus was fixed on a glass slide using methanol– acetic acid 3:1, and dehydrated sequentially in 70, 85 and 100% ethanol series. Multicolour FISH was performed in one step (486 embryos, of which 141 were analysed for XY, 13, 18, 21 and 345 for XY, 13, 16, 18, 21) or two step protocol (1134 embryos, all analysed for XY, 13, 15, 16, 18, 21 and 22; in 447 embryos chromosome 14 was also screened; in 158 embryos the probe for chromosome 14 was replaced by the probe specific for chromosome 17, and by the one detecting chromosome 1 in 391 embryos; in the last 138 embryos no extra chromosomes were tested as another probe specific for chromosome 21 was added in the second panel to have this chromosome screened twice). The selection of chromosomes to be screened was based on the results of spontaneous abortions and live births (Munne´ et al., 1998b). Other chromosomes that do not seem to have clinical relevance were also analysed with the aim of verifying whether this is also true during the first embryonic stages. After co-denaturation of probes and nuclear DNA, hybridisation for approximately 4 h at 37 °C in a moist chamber, and washing of unbound probe, antifade
L. Gianaroli et al. / Molecular and Cellular Endocrinology 183 (2001) S13–S18 Table 1 Overall FISH and clinical results in PGD cycles
S15
3. Results
solution was added and fluorescence was evaluated with a Olympus microscope (Olympus, Tokyo, Japan) at 60 × magnification. In the two-step protocol, the procedure was repeated as already described.
Table 1 depicts the overall outcome of the 343 treatment cycles included in the study. In all, 2096 embryos were generated, and 1782 were selected on the basis of their morphological appearance to undergo FISH analysis. A diagnosis was obtained for 1758 embryos (99%). Only 592 of them were classified as chromosomally normal (34%) whereas the remaining 1166 (66%) exhibited chromosomal abnormalities which were not compatible with either implantation or an healthy pregnancy. Consequently, embryo transfer was cancelled in 103 of the started PGD cycles (30%) where no euploid embryos were detected. An average of 2.090.8 embryos were transferred to the remaining 240 cycles resulting in 79 clinical pregnancies (33%) and an overall implantation rate of 22.6%. Spontaneous abortion occurred in seven pregnancies establishing an abortion rate after PGD of 9%.
2.4. Embryo spreading
3.1. In 6itro efficiency
The cells constituting 311 non transferrable embryos were spread after removing the zona pellucida by a short incubation in acidic Tyrode’s solution. The nuclei were fixed on a glass slide and hybridised with the same probes used for the one-cell diagnosis following the same protocol as above. The results obtained were compared to the original diagnosis to evaluate the inaccuracy rate of the technique.
As represented in Table 2, 311 non transferrable embryos were spread to FISH analyse all their blastomeres. Most of these embryos (271) were chromosomally abnormal based on the result obtained from one-cell analysis. The remaining 40 embryos had been diagnosed as euploid; however they were not transferred or cryopreserved due to blockage of development or increased fragmentation in the 2 days in culture after blastomere biopsy. Confirmation of the one-cell diagnosis was demonstrated for 185 embryos (92%). Of the remaining 26 embryos, 17 had been classified by PGD on day 3 as chromosomally abnormal; the analysis of all their blastomeres revealed that they carried an abnormality that was different from the one diagnosed by
No. of of cycles Age (mean 9SD, years) No. of of generated embryos No. of FISH analysed embryos No. of FISH diagnosed embryos FISH normal (%) FISH abnormal (%) No. of embryos transferred (mean 9 SD) No. of cycles transferred No. of clinical pregnancies (%) Aborted (%) Implantation rate (%) a
343 36.4 94.3 2096 1782 1758 592 (34) 1166 (66) 2.0 90.8 240 (70) 79 (33) 10a 28.4
One ectopic; one after amniocentesis; one therapeutic.
2.5. Statistical analysis Results were evaluated by Student’s t-test and by chi-square analysis applying the Yates’ correction, 2× 2 contingency tables. Table 2 Embryo spreading and FISH re-analysis of all the blastomeres generated PGD
No. of embryos
Re-analysis Confirmed
Normal Monosomic One chromosome Two chromosomes Three chromosomes Trisomic One chromosome Two chromosomes Three chromosomes Haploid Polyploid Complex abnormalities Total
40 85 60 21 4 85 75 9 1 12 20 69 311
40 (100) 72 52 19 1 79 70 9 – 10 17 67 285 (92)
Normal
Other abnormality
– 3 1 2 – 3 3 – – – 1 2
– 10 7 – 3 3 2 1 2 2 –
9 (3)
17 (5)
The results are compared to those obtained by PGD on one cell. Values in parentheses are percentages.
S16
L. Gianaroli et al. / Molecular and Cellular Endocrinology 183 (2001) S13–S18
Table 3 Delivery outcome Deli6eries Singleton Twin Triplet Gestational weeks (M 9SD) Singleton Twin Triplet Deli6ery mode Cesarian section Vaginal
57 39 16 2 37.5 9 3 37.9 9 2.5 36.89 1.8 37 46 11
one-cell analysis. This represents a 5% inaccuracy that does not have any clinical relevance, as these embryos had been classified not suitable for transfer. Conversely, nine embryos (3%) were assessed as abnormal by PGD but were found to be chromosomally normal after re-analysis of all the blastomeres. According to these data, the in vitro inaccuracy of the technique is 3%.
3.2. Pregnancy outcome The outcome of the 79 pregnancies is the following: 57 women already delivered, and 12 have a regularly ongoing pregnancy. In ten cases the pregnancy was interrupted: spontaneous abortion occurred in seven cases, one pregnancy was ectopic, one aborted after amniocentesis (the fetal karyotype was normal), and one underwent therapeutic abortion because at prenatal diagnosis the foetus was diagnosed to have trisomy 21. Finally, two couples decided to undergo selective fetal reduction, and already delivered healthy babies. Table 3 summarises the main characteristics of the 57 deliveries, with 39 singleton, 16 twins, and two triplet. Cesarian section was the most frequent mode of delivery (46 vs 11).
3.3. In 6i6o efficiency All the pregnant patients were recommended to undergo conventional prenatal diagnosis. Sixteen out of 79 (20%) decided not to follow the clinicians’ recommendation for the following reasons: 12 couples claimed to be afraid of the risk of abortion inherent to the procedure; two were against the possibility of therapeutic abortion; and two adduced other reasons involving ethical and moral implications. The collection of the information derived by prenatal diagnosis permitted to verify the exactness of PGD, and to establish the in vivo efficiency of the technique. The results and conclusions are reported in Tables 4 and 5. The karyotype of 91 foetuses was assessed by either chorionic villus sampling (CVS) or amniocentesis, or by examination of the infant at birth. Confirmation of the
diagnosis performed by PGD was obtained in 89 cases with an in vivo efficiency corresponding to 97.8%. In the remaining two foetuses, trisomy 21 was detected: in one case the patient, screened for the translocation 13:14, underwent therapeutic abortion, whereas in the other case the infant was delivered as the couple decided not to perform any prenatal diagnosis. In consideration of the results obtained and based on the general data from prenatal diagnosis, a theoretical calculation was made on the risk of trisomic foetuses (Table 6). This was made with the purpose of estimating the possible advantage of transferring FISH selected embryos. The calculation was based on the number of trisomic embryos detected in the pregnant patients that could develop to the blastocyst stage (seven and 13 depending on the age of the patient). According to the implantation rate per pregnant patient, 4.9 and 7.8 trisomic embryos, respectively were expected to implant. If these figures are multiplied by a correction factor derived by the general data from CVS, 0.2 and 4.0 are the numbers of trisomic embryos expected at CVS in the absence of PGD in the two age categories, compared to zero and two trisomic foetuses, respectively generated after PGD.
4. Discussion PGD for aneuploidy was integrated in IVF with the main purpose of increasing the chances of full term pregnancy in poor prognosis patients, mainly women in advanced reproductive age (Munne´ et al., 1993; Verlinsky and Kuliev, 1996; Gianaroli et al., 1997b). It was a big challenge, because such an invasive and expensive procedure would be implemented based on the hope of overcoming the age effect by selecting for transfer only those embryos having a normal chromosomal complement. The results obtained after several years of intense work done by a few centers worldwide, have permitted to establish that a positive outcome is associated to the transfer of FISH selected embryos in poor prognosis patients (Gianaroli et al., 1999a; Munne´ et al., 1999). Concomitantly, special attention has been dedicated to monitor the quality of the results obtained by multicolour FISH on a single cell. The efficiency of the technique has been continuously evaluated by reTable 4 PGD of aneuploidy: in vivo efficiency No. of gestational sacs Aborted Selective reduction Ongoing On term CVS, chorionic villus sampling; AC, amniocentesis.
110 17 4 12 77
L. Gianaroli et al. / Molecular and Cellular Endocrinology 183 (2001) S13–S18
S17
Table 5 PGD of aneuploidy: in vivo efficiency Confirmed: 89 (97.8%)
Not confirmed: (2.2%)
Unknown
At CVS/AC
At birth
At CVS/AC
At birth
–
Ectopic Misacrriages Late abortion Ongoing On term
– – 5a 8 53
– – – – 23
– – 1 – –
– – – – 1
1 14 – 4b –
Total
66
23
1
1
19
CVS, chorionic villus sampling; AC, amniocentesis. a Four sacs reduced; one normal foetus lost after AC. b Waiting for prenatal diagnosis results.
analysing as many embryos as possible among those that are classified as non transferrable on the basis of the FISH results or their morphology on day 4 or 5. The data obtained permit not only to establish the current inaccuracy rate of the technique, but also to define and propose solutions to the most common situations that may represent a source of incorrect diagnosis. As represented in Table 2, most of the discrepancies involve false monosomies. Loss of micronuclei during fixation, failed hybridisation and signal overlapping represent common causes of misdiagnosis. A significant reduction in the loss of micronuclei was derived through a modification in the fixation protocol that substantially improved the performance of the technique (Munne´ et al., 1998a). In addition, a more accurate fixation of the nuclei and the use of pure fluorochromes, instead of a combination of them, has proven to minimise the risk of signal overlapping (Munne´ et al., 1998b). Failed hybridisation is probably less common; however, due to its clinical relevance, a different probe specific for chromosome 21 is now added also in the second round of hybridisation, in order to make it tested twice. This strategy has been recently adopted and its performance is still under evaluation. The data yielded by embryo spreading and FISH re-analysis also enabled the formulation of scoring criteria for the correct interpretation of FISH signals. The most confounding situation is when the target splits into two signals producing false positive results. The least error-prone criterion states that two signals represent two homologue chromosomes when their distance apart is at least two domains diameters (Munne´ et al., 1998a). Some probes have a higher tendency to split (i.e. the probes for chromosomes 18 and Y) and this must be taken into consideration for the interpretation of the correspondent signals. Finally, mosaicism that is very common in in vitro human embryos, did not seem to represent a frequent source of incorrect diagnosis maybe due to the fact that
the use of multiple probes enabled the identification of complex abnormalities. Interestingly, all the 40 embryos diagnosed as normal by PGD were confirmed after re-analysis; although this figure is relatively small, this finding suggests that the frequency of mosaic embryos with euploid cells could be very low. In consideration of the current efficiency of FISH performed on single cell, all the patients entering a PGD cycle at S.I.S.ME.R. are thoroughly informed of the need to undergo conventional prenatal diagnosis in case of pregnancy. This approach permits not only to verify and complete the diagnosis done on preimplantation embryos in patients that are exposed at a very high risk of generating aneuploid embryos, but also to establish the in vivo efficiency of PGD. This parameter is extremely useful for the team, since it represents the true evaluation of the efficacy of the technique. According to data presented in this study, the in vitro efficiency is 97.1%; a 97.8% in vivo efficiency represents the expected figure within the limits of the technique. In addition, it is important to remind that with an inaccuTable 6 Number of trisomic foetuses after PGD compared to the expected in the absence of PGD: in vivo efficiency of the technique 535
]36
No. of diagnosed embryos in pregnant patients 207 No. of trisomic embryos 24 Blastocyst development correction factora 0.3 No. of expected trisomic blastocysts 7 IRPP correction factorb 70.4 No. of expected implantations 4.9 CVS correction factorc 0.3 No. of expected trisomic foetuses at CVS 0.2 No. of observed trisomic foetuses at CVS after 0 PGD
284 45 0.3 13 59.8 7.8 5.1 4.0 2
Age (years)
CVS, chorionic villus sampling; IRPP, implantation rate per pregnant patient. a Magli et al. (2000). b Gianaroli et al. (1999). c Snijders et al. (1994).
S18
L. Gianaroli et al. / Molecular and Cellular Endocrinology 183 (2001) S13–S18
racy rate of 3%, PGD for aneuploidy still enables to significantly reduce the risk of trisomic foetuses in women of advanced age by identify 97% of the generated trisomies. As postulated in Table 6, the number of expected trisomic foetus at CVS would be four in the absence of PGD; this number was cut exactly by half with embryos transferred on the basis of FISH diagnoses. In conclusion, the experience gathered during these years in the field of PGD for aneuploidy has contributed some important points: patients with advanced reproductive age, an altered karyotype, and (possibly) recurrent abortions derive a benefit in terms of increased implantation rate and lower incidence of spontaneous abortions; for women in advanced reproductive age the risk of a trisomic pregnancy is significantly reduced by half. Nevertheless, in view of the patients’ health as the main priority, recommendation of prenatal diagnosis is still mandatory. A systematic monitoring of the pregnancies and the follow up of the babies born and their development are needed for a comprehensive evaluation of the efficacy of the technique.
References Bahc¸ e, M., Cohen, J., Munne´ , S., 1999. PGD of aneuploidy: were we looking at the wrong chromosomes? J. Assist. Reprod. Genet. 16, 176 – 181. Ferraretti, A.P., Magli, M.C., Feliciani, E., et al., 1996. Relationship of timing agonist administration in the cycle phase to the ovarian response to gonadotropins in the long-down regulation protocols for assisted reproductive technologies. Fertil. Steril. 65, 114 – 121. Gianaroli, L., Fiorentino, A., Magli, M.C., et al., 1996. Prolonged sperm – oocyte exposure and high sperm concentration affect human embryo viability and pregnancy rate. Hum. Reprod. 1, 2507 – 2511. Gianaroli, L., Magli, M.C., Ferraretti, A.P., et al., 1997a. Preimplantation genetic diagnosis increases the implantation rate in human in vitro fertilisation by avoiding the transfer of chromosomally abnormal embryos. Fertil. Steril. 68, 1128 – 1131. Gianaroli, L., Magli, M.C., Munne´ , S., et al., 1997b. Will preimplantation genetic diagnosis assist patients with a poor prognosis to achieve pregnancy? Hum. Reprod. 12, 1762 – 1767.
Gianaroli, L., Magli, M.C., Ferraretti, A.P., et al., 1999a. Preimplantation diagnosis for aneuploidies in patients undergoing in vitro fertilisation with a poor prognosis: identification of the categories for which it should be proposed. Fertil. Steril. 72, 837 –844. Gianaroli, L., Magli, M.C., Munne´ , S., et al., 1999b. Advantages of day four embryo transfer in patients undergoing preimplantation genetic diagnosis of aneuploidy. J. Assist. Reprod. Genet. 16, 170 – 175. Gianaroli, L., Magli, M.C., Ferraretti, A.P., et al., 2000. Preimplantation diagnosis after assisted reproduction techniques for genetically determined male infertility. J. Endocrinol. Invest. 23, 711 – 716. Handyside, A.H., Kontogianni, E.H., Hardy, K., et al., 1990. Pregnancies from biopsied human preimplantation embryos sexed by Y-specified DNA amplification. Nature (London) 344, 768 –770. Magli, M.C., Gianaroli, L., Munne´ , S., et al., 1998. Incidence of chromosomal abnormalities from a morphologically normal cohort of embryos in poor prognosis patients. J. Assist. Reprod. Genet. 15, 297 – 301. Magli, M.C., Gianaroli, L., Fortini, D., et al., 1999. Impact of blastomere biopsy and cryopreservation techniques on human embryo viability. Hum. Reprod. 14, 770 – 773. Magli, M.C., Jones, G., Gras, L., et al., 2000. Chromosome mosaicism in day 3 aneuploid embryos that develop to morphologically normal blastocysts in vitro. Hum. Reprod. 15, 1781 – 1786. Munne´ , S., Lee, A., Rosemwaks, Z., et al., 1993. Diagnosis of major chromosome aneuploidies in human preimplantation embryos. Hum. Reprod. 8, 2185 – 2191. Munne´ , S., Marquez, C., Magli, M.C., et al., 1998a. Scoring criteria for preimplantation genetic diagnosis of numerical abnormalities for chromosomes X, Y, 13, 16, 18, and 21. Molec. Hum. Reprod. 4, 863 – 870. Munne´ , S., Magli, M.C., Bahc¸ e, M., et al., 1998b. Preimplantation diagnosis of the aneuploidies most commonly found in spontaneous abortions and live births: XY, 13, 14, 15, 16, 18, 21, 22. Prenat. Diagn. 18, 1459 – 1466. Munne´ , S., Magli, M.C., Cohen, J., et al., 1999. Positive outcome after preimplantation diagnosis of human embryos. Hum. Reprod. 14, 2191 – 2199. Simon, C., Rubio, C., Vidal, F., et al., 1998. Increased chromosomal abnormalities in human preimplantation embryos after in vitro fertilisation in patients with recurrent miscarriage. Reprod. Fertil. Dev. 10, 87 – 92. Snijders, R.J.M., Holzgreve, W., Cuckle, H., et al., 1994. Maternal age specific risks for trisomies at 9 – 14 weeks’ gestation. Prenat. Diagn. 14, 543 – 552. Verlinsky, Y., Kuliev, A., 1996. Preimplantation diagnosis of common aneuploidies in fertile couples of advanced maternal age. Hum. Reprod. 11, 2076 – 2077.