A recent study suggests that preimplantation genetic diagnosis – PGD may be ineffective as a means for increasing the live birth rate in older women with infertility and, in fact, may actually decrease the live birth rate. A potential cause of the low pregnancy rates in women of advanced maternal age undergoing in vitro fertilization – IVF is the increased likelihood of embryos containing embryos with chromosomal abnormalities. The most common type of abnormality occurs when an embryos has too many or too few chromosomes. The abnormalities are called aneuploidies. Numerous investigators have demonstrated a higher percentage of abnormal embryos as women age.
It is thought that most of these chromosomally abnormal embryos do not implant in the uterus or are miscarried. Abnormal embryos, which develop to term, give rise to babies with birth defects such as Down’s Syndrome.
Consequently, preimplantation genetic diagnosis has been proposed as a way to increase live-birth rates in older women. In preimplantation genetic diagnosis, embryos that are identified as abnormal are discarded, and embryos with a normal genetic constitution are selected for transfer. Several studies have shown that embryos that are selected on this basis appear to have a higher potential for implantation, and a lower chance for miscarriage. Different centers have data indicating that the chances for pregnancy with IVF may be improved by screening embryos for abnormalities using PGD.
New PGD Study from the Netherlands
Women from 35 through 41 years of age who were scheduled for IVF and who had no previous failed IVF cycles were eligible to participate in this study. The women were randomly assigned to undergo three cycles of IVF. In one group, embryo selection was based preimplantation genetic diagnosis of the available embryos. In the control group, embryo selection was based only on the appearance of the embryos microscopically.
In the PGD group, testing was begun on the third day after the egg retrieval. On that day, if the embryos had divided into at least four cells, then one cell was removed for testing. This is called a blastomere biopsy. In order to remove a cell, the hard shell surrounding the embryo, known as the zona pellucida must first be opened. In this study, this was accomplished using a laser. If the cell that was removed did not contain a nucleus suitable for PGD testing, then a second cell was removed, but only if the remaining embryo contained at least four cells.
PGD testing for chromosome number was performed using the FISH (fluorescent in-situ hybridization) technique. With this technique, chromosomes can be labeled with probes of a certain color. Scientists can then count the number of each type of chromosome that was present in the cell. In this study, PGD was performed for a total of eight chromosomes. The chromosomes analyzed were 1, 13, 16, 17, 18, 21, X, and Y.
On the basis of the results of PGD analysis, embryos were categorized as normal if two copies of each chromosome were present. The possible categories of embryos were normal, abnormal or undetermined.
Up to two chromosomally normal embryos with the best microscopic features were selected for embryo transfer into the female’s uterus on the fourth day after egg retrieval. If no chromosomally normal embryos with good morphologic features were available for transfer, undetermined embryos with good morphologic features were selected for transfer.
In the control group, the selection of embryos for transfer was based solely on features visible under the microscope. In either group, if there were extra embryos that met the criteria for embryo transfer, they were frozen. If pregnancy did not occur in the initial IVF cycle, the frozen embryos were thawed and transferred to the woman’s uterus before a new IVF cycle was attempted.
PGD Study Results
Seventy-seven of the 206 women assigned to preimplantation genetic diagnosis and 71 of the 202 control women did not complete three cycles of IVF nor did they achieve an ongoing pregnancy. This was primarily due to poor ovarian response to the fertility drugs used in the study or what the study authors called the “burden of treatment”.
As compared with women in the control group, women in the preimplantation genetic diagnosis group had a lower live-birth rate. The live birth rate in the PGD group was 24% (49 live births out of 206 women who were originally assigned to this group). The live birth rate in the control group was 35% (71 deliveries out of 202 women originally assigned to this group). Statistical analysis determined that it was unlikely these results occurred by chance.
The miscarriage rates did not differ between the groups.
A total of 836 cycles of IVF with egg retrieval were performed (434 cycles with and 402 cycles without PGD). The ongoing-pregnancy rates in the two groups were not different when data were analyzed for first cycles only, for second cycles only, or for third cycles only.
In the PGD group, 75 of 642 transferred embryos (11.7%) implanted (as determined by the presence of a gestational sac seen in the uterus on transvaginal ultrasound). In the control group, 99 of 673 transferred embryos (14.7%) implanted. However, further analyses of the results from the PGD group showed that in cycles in which only two embryos with normal FISH results were transferred, the implantation rate was 16.8% (53 of 316), and that in cycles in which two undetermined embryos were transferred, the implantation rate was 6.0% (6 of 100).
There were two babies found to have chromosomal abnormalities. In the control group, one baby had three copies of chromosome 18 (Trisomy 18 or Edward’s Syndrome) and was terminated by the parents. In the PGD group, one woman conceived a spontaneous pregnancy after she was randomized but before she started treatment. She also had a baby with Edward’s syndrome and underwent elective termination of that pregnancy. There were no babies with chromosomal abnormalities in the PGD group who conceived as a result of treatment.
The authors of this study should be applauded for their efforts and trying to obtain high quality medical data to answer the question of whether or not PGD improves the live birth rate in women of advanced maternal age. That being said, however, there were several problems in the study design that severely limits interpretation of the results.
Age of the patients
Most previous published studies that showed a benefit to the use of PGD on pregnancy rates in women undergoing IVF were only able to find a benefit in women who were 37 years of age or over. This study included women who were 35 or 36 years of age. This group of women would not have been expected to benefit from PGD. Their inclusion would dilute potentially beneficial results.
Improper selection of embryos
The stage at which an embryo is biopsied is an important factor that determines the future viability of the embryo. Most centers that perform PGD will only biopsy embryos that have reached the 6 to 8 cell stage. In this study, the centers biopsied embryos if the embryos had divided into 4 or more cells. This would have the effect of compromising the viability of the embryos and contributed to lower pregnancy rates in the PGD group.
Furthermore, the researchers opted to transfer embryos that may have had a poor potential for pregnancy. These included embryos which had cells with no nucleus, cells which had to be biopsied multiple times (which are known to have a lower implantation rate), and cells which, after multiple attempts, were classified as “undetermined” because no chromosome results could be reported. In our center and most others, these embryos would not have been considered acceptable for embryo transfer.
In fact, the best method for determining the viability of embryos is by looking at the implantation rate. In this study, in embryos with complete chromosome results suggesting a normal embryo, the implantation rate was nearly three times higher than in the “undetermined” embryos. (16.8% vs 6%).
Normally, all of the cells in an embryo should be identical. However, as cells are dividing, sometimes mistakes are made and some cells may end up being different than the remaining cells in the embryo. This is known as mosaicism. Mosaicism can limit the effectiveness of PGD if the cell biopsied is different than the rest of the embryo. Performing multiple biopsies can reduce the number of errors made due to mosaicism. However, as stated above, removing two cells from a dividing embryo can affect its potential development. One solution, which is used by some centers with greater degrees of expertise, is to use a combination of polar body biopsy and blastomere biopsy.
Outcome of babies
It was reported that there were two babies born with chromosome abnormalities. One was born as a result of an IVF cycle without PGD. The other was from a spontaneous pregnancy that occurred during the study. Since there were 71 deliveries in the IVF alone group, this is a rate of 1.4% (1/71). In the total group of deliveries (IVF alone, PGD, and spontaneous) the rate was 1.7% (2/120). This is a significantly higher rate than is reported in a population of women aged 35-39 (1/4800) or 40-44 (1/1600).
Several studies have suggested that couples with infertility may be at greater risk for complications of pregnancy including birth defects and chromosome abnormalities. The study failed to adequately address the fact that preimplantation genetic diagnosis in this population appeared to eliminate the risk of babies with chromosome abnormalities. Even if subsequent data shows no improvement in the livebirth rate, if there are fewer babies born with birth defects it may still justify the use of preimplantation genetic diagnosis in these patients.
A study was conducted in the Netherlands to determine whether the routine use of preimplantation genetic diagnosis in in-vitro fertilization patients over age 34 would be of benefit in terms of improving the live birth rate. Due to methodologic problems with the study, the ability to draw firm conclusions was severely limited. Preimplantation genetic diagnosis was able to detect apparently abnormal embryos. When only normal embryos were transferred to the uterus compared to unknown or untested embryos, the likelihood of implantation was significantly improved. No babies with birth defects due to chromosome abnormalities were born after preimplantation genetic diagnosis. The rate in in vitro fertilization patients was between 1-2% of deliveries.