1 Institute for Reproductive Medicine and Science, Saint Barnabas, Livingston, NJ, 2 S.I.S.Me.R, Bologna, Italy, 3 Reproductive Biology Associates, Atlanta, GA, USA
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key words: in-vitro fertilization/preimplantation genetic diagnosis/trisomy 21, 18, 13, 16
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Because of the association between aneuploidy and implantation, it was postulated that selection of chromosomally normal embryos could reverse this trend (Munné et al., 1993). However, while some research groups felt that this was possible and desirable (Verlinsky and Kuliev, 1996
: Gianaroli et al., 1997
) others doubted its value or feasibility (Egozcue, 1996
; Reubinoff and Sushan, 1996
). Currently, removal of aneuploid embryos can only be done through PGD, after either polar body or blastomere analysis. Fluorescence in-situ hybridization (FISH) allows chromosome enumeration on interphase cell nuclei, i.e. without the need for culturing cells or preparing metaphase spreads. FISH has been applied to preimplantation genetic diagnosis (PGD) of common aneuploidies (at least XY, 13, 18, 21), testing either human blastomeres from cleavage-stage embryos or oocyte polar bodies (Munné et al., 1993
, 1995a
,Munné et al., b
, 1998a
,Munné et al., b
; Manor et al., 1996
; Verlinsky et al., 1995
Verlinsky et al., 1996b; Gianaroli et al., 1997
). More than 500 cycles have been performed using this technique, resulting in >100 chromosomally normal babies. However, an increase in implantation rate or a decrease in abnormal offspring has not been demonstrated, either for lack of data or appropriate controls. Here we present the first study including carefully matched retrospective controls to analyse whether PGD of aneuploidy has any effect on implantation rates, spontaneous abortions, and live births.
![]() |
Materials and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Patient matching
Patients 35 years or older undergoing in-vitro fertilization (IVF) treatment at the three centres participating in this study were offered PGD of aneuploidy. Randomization was initially offered but, because there was no data supporting a beneficial effect of PGD of aneuploidy, few patients agreed to the study and those committed to it rejected randomization. Therefore, the study was planned as a retrospective study where controls were matched to test patients blindly or previous pregnancy outcome was known. PGD cases for centres 2 and 3 were matched, as far as was possible, on the following variables which are listed in order of matching priority: maternal age (±1 year); concentration of oestradiol (with a difference between control and test of ±20%); month of retrieval; duration of stimulation; and number of mature follicles. The rationale for matching them according to these factors, in addition to maternal age, was to avoid changes in laboratory performance and to control for ovarian reserve and response (Scott and Hofmann, 1995). Patients from centre 1 were also matched according to the number of previous IVF attempts. This extra condition is included because in that centre implantation significantly decreases between the first and second attempt. This phenomenon was not observed in many other centres (see review Meldrum et al., 1998b)
The controls also fulfilled the requirements that at least one embryo was replaced and that pregnancy information was available at least up to the 20th week of pregnancy.
Biopsy, fixation and fish procedure
During day 3 of development, one or two cells per embryo were biopsied, and the embryos returned to culture as described elsewhere (Grifo, 1992). All of the embryos were at the 412-cell stage of development at the time of biopsy. Most embryos classified as normal after PGD were transferred to the uterus on the same day of analysis. All blastomeres were fixed individually following our protocol (Munné et al., 1996
).
The present study has taken 2 years, during which fluorescence in-situ hybridization (FISH) protocols for PGD of aneuploidy have evolved considerably. PGD cases included in this study were therefore performed using the protocols available at the time. Some cases used probes for simultaneous detection of chromosomes X, Y, 18 and the shared alpha-satellite region of chromosomes 13 and 21 (Munné et al., 1993) (n = 14). Later on, cases used specific probes for X, Y, 13, 18 and 21 (Munné et al., 1996
) (n = 22). Later still, a probe for chromosome 16 was added to the previous mixture (Munné et al., 1998a
) and used in a proportion of the cases (n = 50). And finally, a small fraction of cases (n = 31) benefited from having the biopsied cells analysed with the X, Y, 13, 16, 18, and 21 probe mixture (Munné et al., 1998b
), and re-analysed with a second probe mixture specific for chromosomes 14, 15 and 22. Thereby all the cases had their biopsied cells analysed for the trisomies with potential of arriving to term (X, Y, 13, 18, 21), and some for trisomies commonly found in spontaneous abortions, but which do not survive to term (14, 15, 16, 22). The protocols for these probe sets have been previously published (Munné et al., 1993
, 1996
, 1998a
,Munné et al., b
; Munné and Weier, 1996
) and a few of the cases here reported were also included in previous manuscripts (Munné et al., 1996
, 1998a
,Munné et al., b
).
The FISH error rate has already been evaluated in previous studies of probes for X, Y, 18, 13/21 (7%) (Munné et al., 1993); X, Y, 13, 18, 21 (13%) (Munné and Weier, 1996
); X, Y, 13, 16, 18, 21 (9%) (Munné et al., 1998a
); and X, Y, 13, 16, 18, 21, 14, 15, 22 (15%) (Munné et al., 1998b
). The scoring criteria (described in Munné and Weier, 1996) were applied in any cases that used at least probes specific for X, Y, 13, 18 and 21 chromosomes.
Statistical analysis
The patients in the two treatment groups were matched, primarily by age, but also to a lesser extent by oestradiol and cycles of IVF. This was to avoid any potential bias caused by these variables.
The data were analysed in the following way. For each patient the statistic of interest (say FHB/transfer) was calculated, and from that the mean values and standard errors were calculated. Thus the `error' reflects inter-patient variation. This is a more appropriate way to analyse the data than by counting embryos as units. Therefore, following the example of the variable `FHB/transfer', the mean for that variable would not be the sum of all FHB divided by the total number of embryos transferred. The variables of interest were generally proportions based on very small frequencies, so that great care was required in the statistical analysis.
Since the variables of interest were proportions, the main method of analysis was logistic regression. The variables contributing to the proportions of interest (for example number of FHB/transfer) were entered into the GLM (generalized linear modelling) analysis, which then tested for systematic differences caused by treatment and centre. Maternal age was also included as a covariate, but, because of the matching process described above, this had minimal impact. The analysis generated estimated mean values and standard errors for the various sub-groups, these figures being quoted in the tables. The statistical tests were carried out within the algorithm, and the findings are also quoted in the tables. The algorithm used to carry out the calculations was Genstat (1988).
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Pregnancy outcome
The overall results and statistical analysis are shown in Table I. Implantation rates, as FHB/embryo, were slightly higher in the PGD group (17.8%) but not statistically different from the control group (13.7%). However, spontaneous abortions (lost FHB/FHB detected) decreased 2.5-fold after PGD (P < 0.05), from 23% in control to 9% in the test group. The slight increase in implantation and the significant decrease in spontaneous abortions resulted in a significant increase (P < 0.05) in ongoing and delivered babies, from 10.5% (43/408) in controls to 16.1% (57/354) in the PGD group. When the experimental data were summarized as the number of patients becoming pregnant, or miscarrying, the trends were similar to those observed in Table I
, although the findings were not statistically significant. It would appear therefore that the additional information contained in variables such as `number of transfers' was necessary in order to highlight the treatment differences. Also the results derived from the rather complex GLM (general linear modelling) method of Table I
could often be replicated by a more simple approach. Thus if the ratio (ongoing/transfer) was calculated for each patient, the mean ± SD for the PGD group was 0.157 ± 0.024, and for the controls was 0.107 ± 0.021, values very similar to those quoted in Table I
. Furthermore, working with pairs of matched patients, we found that in 66 the same ratio was obtained, in 32 pairs the PGD ratio was greater than the control ratio, and in 19 pairs the control ratio was greater than the experimental group ratio. Therefore, of the 51 pairs where there was a non-zero difference in the ratios, the PGD ratio was the higher in 32 of those pairs. The departure from the expected value under the null hypothesis (25.5) was just significant at the 5% level, thus supporting the finding in Table I
. Thus by adopting a completely different, non-parametric approach of counting differences, we obtained precisely the same result as the GLM analysis used to produce Table I
.
|
|
Analysis during prenatal diagnosis and after birth showed no abnormalities in the test group for the chromosomes assessed. However, a spontaneous abortion was shown to be trisomy 21. The cells belonging to the transferred embryos of the case resulting in trisomy 21 were re-hybridized with a probe for chromosome 21 (already described in a previous study) (Munné et al., 1998b) and the same normal result was obtained.
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
From the present results, it can be postulated that aneuploidy determination prior to embryo transfer may reduce the incidence of miscarriage, and increase the rate of delivery. Because the chromosomes analysed in this study are involved in at least three-quarters of chromosomally abnormal miscarriages (Schmidt-Sarosi et al., 1998), a reduction in pregnancy loss should be expected, and emphasizes the accuracy of the proposed hypothesis. It is important to consider the conditions under which these results were obtained, emphasizing the need for further, more advanced studies. The results were evaluated retrospectively and obtained in three different centres, during a 3 year period. During this period, conditions of assisted reproduction changed, most notably in the use of follicular stimulation protocols, but also because other probes became available. An increase in ongoing pregnancies and live births was most notably due to a reduction of the occurrence of miscarriage.
In most IVF centres there is a reduction of ~3050% in implantation rate from patients 2534 to 3542 years old. If such decrease were solely produced by chromosome abnormalities, we would expect an increase in implantation after PGD of a similar magnitude, provided sufficient number of oocytes were available, which is not the case. The low PGD effect on implantation can be attributed to at least three factors. One could be that the chromosomes chosen for this test are those producing abnormalities compatible with further development. For instance, trisomy 1 is found in cleavage stage embryos but never detected in spontaneous abortions (Watt et al., 1987; Simpson 1990
; Delhanty et al., 1997
; Laverge et al., 1997
), and other chromosomes, such as 17, seldom found in spontaneous abortions, are quite common in cleavage-stage embryos (Simpson, 1990
; Bahcie et al., 1999
). The importance of the probes used is also evident when comparing PGD cases performed with different probe mixtures, with the one used to test eight chromosomes producing the highest implantation rate improvement and the highest reduction in spontaneous abortions.
A second factor would be that the working hypothesis is incorrect, namely, in opposition to previously held ideas, chromosomally abnormal embryos may in fact often implant, many of them dying shortly afterwards and being detected as empty sacs. This would also imply that embryos of women of advanced maternal age may be less likely to implant because of non-chromosomal causes.
The third factor affecting implantation might be the embryo biopsy procedure. Although a previous study on embryo biopsy did not show a negative effect on embryo development to blastocyst (Hardy et al., 1990), embryo transfer effects, if any, were not assessed. Moreover, the embryos studied were from donors and not from older patients or those with a poor prognosis. Evidence of embryo damage during transfer of zona-drilled embryos comes from assisted hatching studies. For instance, assisted hatching has been demonstrated to increase implantation rates (Cohen et al., 1990
), but its potential is related to the diameter of the opening in the zona. Openings of 4060 µm diameter commonly used for biopsy, instead of the recommended 2030 µm for assisted hatching, might eliminate any beneficial effect (B.Schoolcraft and T.Schlenker, personal communication), and cause embryo damage immediately after the procedure or during and after embryo transfer. Preliminary data from one of the centres reporting in this study, compared two embryologists performing embryo biopsies, one with zona openings 4050 µm in diameter and the other 3040 µm. The first obtained a 13% (22/166) implantation rate per embryo transferred compared to 21% (16/76) of the second embryologist. Although the differences were not significant, they point again to a very serious detrimental effect of large zona openings. Another effect of embryo biopsy could be the toxicity of acidified Tyrode's solution on the adjacent blastomeres, although the appearance of effective laser devices could make its use obsolete (Montag et al., 1998
). Other negative effects of embryo biopsy may be that it interferes with cell allocation and positioning during compaction and blastulation (Edwards and Beard, 1998
). However, embryos with 15% fragments and seven cells on day 3, have the same implantation rate as 8-cell embryos when the fragments have been removed (Alikani et al., 1999
), and these embryos could be compared to biopsied embryos in which one cell has been removed. Similarly, linear tracing experiments have demonstrated that any cell of an 8-cell embryo could produce ICM (Mottla et al., 1995
). A more clear negative effect of embryo biopsy could be the interference with compaction. For instance, data from centre 1 indicates that there is a 4% decrease in embryo implantation when the biopsy is performed 70 h or later after embryo retrieval compared to when it is performed 67 h or earlier. Clearly, more research is needed on the effect of embryo biopsy on implantation, when beneficial effects such as PGD are excluded.
The potential hazards of embryo biopsy could be avoided by using polar body instead of blastomere biopsy (Verlinsky et al., 1995, 1996
; Munné et al., 1996
). Polar body biopsy can be performed by mechanical or laser techniques and does not interfere with cell allocation. The hole in the zona is on average 1020 µm, about the size recommended for AHA. If the polar body approach is taken, the issue of which probes to use might be solved using spectral imaging techniques (Schröck et al., 1996
) to karyotype all 23 chromosomes of the polar body as already demonstrated (Márquez et al., 1998
). Although most aneuploidies are maternal in origin, polar body analysis would not detect other abnormalities such as polyploidy, haploidy and mosaicism involving most or all cells and accounting for at least 19% of embryonic chromosome abnormalities (Munné et al., 1995).
In conclusion, the present study demonstrates that PGD of aneuploidy can positively affect the outcome of embryos by reducing embryo wastage and increasing implantation rates. As PGD of aneuploidy now stands, patients of advanced maternal age undergoing IVF can benefit from this technique by avoiding the trauma of losing greatly desired pregnancies and giving themselves a greater chance of having a baby. Because the procedure appears beneficial, further research can continue in order to elucidate the causes behind the disappointing effects on implantation. We are currently investigating the use of other probes, spectral imaging, laser embryo biopsy and polar body biopsy in a randomized study to improve the procedure, hopefully to bring implantation rates in those women in line with the rates found in younger patients.
|
![]() |
Acknowledgments |
---|
![]() |
Notes |
---|
* Some of the data in this paper were presented as an abstract in the 54th Annual Meeting of The American Society of Reproductive Medicine (San Francisco, CA, USA, 1998), where it was designated the Prize Paper of the Society for Assisted Reproductive Technology.
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Antonorakis, S.E., Lewis, J.G., Adelsberg, P.A. et al. (1991) Parental origin of the extra chromosome in trisomy 21 revisited: DNA polymorphism analysis suggests maternal origin in 95% of cases. N. Engl. J. Med., 324, 872876.[Abstract]
Bahcie, M., Cohen, J. and Munné, S. (1999) PGD of aneuploidy: were we looking at the wrong chromosomes? J. Assist. Reprod. Genet., 16, 176181.[ISI][Medline]
Cohen, J., Elsner, C., Kort, H. et al. (1990) Impairment of the hatching process following IVF in the human and improvement of implantation by assisting hatching using micromanipulation. Hum. Reprod., 5, 713.[ISI][Medline]
Dailey, T., Dale, B., Cohen, J. and Munné, S. (1996) Association between non-disjunction and maternal age in meiosis-II human oocytes detected by FISH analysis. Am. J. Hum. Genet., 59, 176184.[ISI][Medline]
Delhanty, J.D.A., Harper, J.C., Ao, A. et al. (1997) Multicolour FISH detects frequent chromosomal mosaicism and chaotic division in normal preimplantation embryos from fertile patients. Hum. Genet., 99, 755760.[ISI][Medline]
Edwards, R.G. and Beard, H.K. (1998) Oocyte polarity and cell determination in early mammalian embryos. Mol. Hum. Reprod., 3, 863905.[Abstract]
Egozcue, J. (1996) Of course not. Hum. Reprod., 11, 20772078.[ISI][Medline]
Fisher, J.M., Harvey, J.F., Morton, N.E. and Jacobs, P.A. (1995) Trisomy 18: Studies of the parent and cell division of origin and the effect of aberrant recombination on nondisjunction. Am. J. Hum. Genet., 56, 669675.[ISI][Medline]
Garside, W.T., Loret de Mola, J.R., Bucci, J.A. et al. (1997) Sequential analysis of zona thickness during in vitro culture of human zygotes: correlation with embryo quality, age, and implantation. Mol. Reprod. Dev., 47, 99104.[ISI][Medline]
Genstat (1988) The Genstat 5 Reference Manual. Oxford, Clarendon Press.
Gianaroli, L., Magli, M.C., Munné, S. et al. (1997) Will implantation genetic diagnosis assist patients with a poor prognosis to achieve pregnancy? Hum. Reprod., 12, 17621767.[Abstract]
Grifo, J.A. (1992) Preconception and preimplantation genetic diagnosis: polar body, blastomere, and trophectoderm biopsy. In Cohen, J., Malter, H.E., Talansky, B.E. and Grifo, J. (eds), Micromanipulation of Gametes and Embryos. Raven Press, New York, pp. 223249.
Hardy, K., Martin, K.L., Leese, H.J. et al. (1990) Human preimplantation development in vitro is not adversely affected by biopsy at the 8-cell stage. Hum. Reprod., 5, 708714.[Abstract]
Hassold, T. and Chiu, D. (1985) Maternal age-specific rates of numerical chromosome abnormalities with special reference to trisomy. Hum. Genet., 70, 1117.[ISI][Medline]
Janny, L. and Ménézo, Y.J.R. (1996) Maternal age effect on early human embryonic development and blastocyst formation. Mol. Reprod. Dev., 45, 3137.[ISI][Medline]
Laverge, H., De Sutter, P., Verschraegen-Spae, M.R. et al. (1997) Triple colour fluorescent in-situ hybridization for chromosomes X, Y and 1 on spare human embryos. Hum. Reprod., 12, 809814.[Abstract]
Manor, D., Kol, S., Lewit, N. et al. (1996) Undocumented embryos: do not trash them, FISH them. Hum. Reprod., 11, 25022506.[Abstract]
Márquez, C., Cohen, J. and Munné, S. (1998) Chromosome identification on human oocytes and polar bodies by spectral karyotyping. Cytogenet. Cell Genet., 81, 254258.[ISI][Medline]
Meldrum, D.R., Wisot, A., Yee, B. et al. (1998a) Assisted hatching reduces the age-related decline in IVF outcome in women younger than age 43 without increasing miscarriage or monozygotic twinning. J. Assist. Reprod. Genet., 15, 418421.[ISI][Medline]
Meldrum, D.R., Silverberg, K.M., Bustillo, M. and Stokes, L. (1998b) Success rate with repeated cycles of in vitro fertilizationembryo transfer. Fertil. Steril., 69, 10051009.[ISI][Medline]
Montag, M., van der Ven, K., Delacretaz, G. et al. (1998) Laser-assisted microdissection of the zona pellucida facilitates polar body biopsy. Fertil. Steril., 69, 539542.[ISI][Medline]
Mottla, G.L., Adelman, M.R., Hall, J.L. et al. (1995) Lineage tracing demonstrates that blastomeres of early cleavage-stage human pre-embryos contribute to both trophectoderm and inner cell mass. Hum. Reprod., 10, 384391.[Abstract]
Munné, S. and Weier, H.U.G. (1996) Simultaneous enumeration of chromosomes 13, 18, 21, X and Y in interphase cells for preimplantation genetic diagnosis. Cytogenet. Cell Genet., 75, 263270.[ISI][Medline]
Munné, S., Lee, A., Rosenwaks, Z. et al. (1993) Diagnosis of major chromosome aneuploidies in human preimplantation embryos. Hum. Reprod., 8, 21852191.[Abstract]
Munné, S., Alikani, M., Tomkin, G. et al. (1995a) Embryo morphology, developmental rates and maternal age are correlated with chromosome abnormalities. Fertil. Steril., 64, 382391.[ISI][Medline]
Munné, S., Dailey, T., Sultan, K.M. et al. (1995b). The use of first polar bodies for preimplantation diagnosis of aneuploidy. Hum. Reprod., 10, 10151021.
Munné, S., Dailey, T., Finkelstein, M. and Weier, H.U.G. (1996) Reduction in signal overlap results in increased FISH efficiency: implications for preimplantation genetic diagnosis. J. Assist. Reprod. Genet., 13, 149156.[ISI][Medline]
Munné, S., Márquez, C., Magli, C. et al. (1998a) Scoring criteria for preimplantation genetic diagnosis of numerical abnormalities for chromosomes X, Y, 13, 16, 18 and 21. Mol. Hum. Reprod., 4, 863870.[Abstract]
Munné, S., Magli, C., Bahcie, 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, 14591466.[ISI][Medline]
Navot, D., Drews, M.R., Bergh, P.A. et al. (1994) Age related decline in female fertility is not due to diminished capacity of the uterus to sustain embryo implantation. Fertil. Steril., 61, 97101.[ISI][Medline]
Reubinoff, B.E. and Sushan, A. (1996) To biopsy or not to biopsy? Hum. Reprod., 11, 20712075.[ISI][Medline]
Scott, R.T. and Hofmann, G.E. (1995) Prognostic assessment of ovarian reserve. Fertil. Steril., 63, 111.[ISI][Medline]
Schmidt-Sarosi, C., Schwartz, L.B., Lublin, J. et al. (1998) Chromosomal analysis of early fetal losses in relation to transvaginal ultrasonographic detection of fetal heart motion after infertility. Fertil. Steril., 69, 274277.[ISI][Medline]
Simpson, J.L. (1990) Incidence and timing of pregnancy losses: relevance to evaluating safety of early prenatal diagnosis. Am. J. Med. Genet., 35, 165173.[ISI][Medline]
Schröck, E., du Manoir, S., Veldman, T. et al. (1996) Multicolor spectral karyotyping of human chromosomes. Science, 273, 494497.[Abstract]
Van Blerkom, J., Davis, P.W. and Lee, J. (1995) ATP content of human oocytes and developmental potential and outcome after in vitro fertilization and embryo transfer. Hum. Reprod., 10, 415424.[Abstract]
Verlinsky, Y. and Kuliev, A. (1996) Preimplantation diagnosis of common aneuploidies in fertile couples of advanced maternal age. Hum. Reprod., 11, 20762077.[ISI][Medline]
Verlinsky, Y., Cieslak, J., Frieidine, M. et al. (1995) Pregnancies following pre-conception diagnosis of common aneuploidies by fluorescence in-situ hybridization. Hum. Reprod., 10, 19231927.[Abstract]
Verlinsky, Y., Cieslak, J., Ivakhnenko, V. et al. (1996) Birth of healthy children after preimplantation diagnosis of common aneuploidies by polar body fluorescent in-situ hybridization analysis. Fertil. Steril., 66, 126129.[ISI][Medline]
Warburton, D., Kline, J., Stein, Z. and Strobino, B. (1986) Cytogenetic abnormalities in spontaneous abortions of recognized conceptions. In Porter, I.H. and Willey, A. (eds), Perinatal Genetics: Diagnosis and Treatment. Academic Press, New York, pp. 133148.
Watt, J.L., Templeton, A.A., Messinis, I. et al. (1987) Trisomy 1 in an eight cell human pre-embryo. J. Med. Genet., 24, 6064.[Abstract]
Submitted on March 15, 1999; accepted on June 10, 1999.