How does one assess the risk of abnormalities from human in vitro fertilization?

Misjudging early embryo mortality in natural human reproduction

In 2002, in a judgment relating to the use of the morning-after pill, Mr Justice Munby held that pregnancy begins with the implantation of an embryo into the uterus of a woman. The case involved a large body of expert witness evidence including medical and physiological details of human reproduction. Munby J. emphasised one particular aspect of this evidence: namely, the developmental failure rate of human embryos after fertilisation. Under natural conditions, embryo loss is approximately 10-40% before implantation, and total loss from fertilisation to birth is 40-60% (Jarvis, 2016). By contrast, and based on expert witness testimony, Munby J. stated that not much more than 25% of successfully fertilised eggs reach the implantation stage, and that fewer than 15% of fertilised eggs result in a birth, figures that do not accurately represent scientific knowledge regarding human embryo mortality and pregnancy loss under natural conditions. Rather, these figures were derived from experimental laboratory data and clinical outcomes from in vitro fertilisation treatment. Testimony provided by other expert witnesses directly contradicted these specific numerical claims. In emphasising these figures, Munby J. gave the impression that human embryo mortality is substantially higher than available scientific evidence indicated. In this critique, all the scientific expert witness evidence is presented and reviewed, and an explanation provided for why the emphasised figures are wrong. Whether there are implications of Munby J.’s scientific misjudgment on the legal outcome is for others to consider.

Maternal methionine adenosyltransferase I/III deficiency: reproductive outcomes in a woman with four pregnancies

Four pregnancies in a women with moderately severe deficiency of methionine adenosyltransferase I/III (MAT I/III) activity are reported. She is an apparent homozygote for a point mutation in MAT1A, the gene that encodes the catalytically active subunit of MAT I/III. This mutation reduces the activity of her expressed enzyme to some 11% of wild-type. She was the first such individual identified in the United States, and these are the first pregnancies known in anyone with this extent of MAT I/III deficiency. No adverse effects were noted in the mother. Three normal babies resulted, but fetal arrest was detected in one embryo at 10-11 weeks gestation. Plasma methionine concentrations remained virtually constant at their elevated levels of 300-350 micromol/L throughout the pregnancies. Plasma free choline was below the reference range. In view of the evidence that maternal choline delivery to the fetus is important for brain development, it was suggested the patient ingest two eggs daily from gestation week 17. Plasma choline and phosphatidylcholine tended to rise during such supplementation. Plasma cystathionine concentrations rose progressively to far above normal during these pregnancies, but not during pregnancies in control women. This may be explained by delivery of excessive methionine to the fetus, with consequent increased cystathionine synthesis by fetal tissues. Because fetal tissues lack gamma-cystathionase, presumably cystathionine accumulated abnormally in the fetus and was transferred in abnormal amounts back to the mother. Plasma and urinary concentrations of methionine transamination metabolites rose during pregnancy for reasons that remain obscure.

Diagnóstico pré-natal

O presente artigo apresenta uma descrição de todos os métodos de diagnóstico pré-natal de anormalidades genéticas e cromossômicas, bem como dos relativos a doenças infecciosas na gravidez. O autor discute as diferenças entre eles, e as novas ferramentas da biologia molecular que são aplicadas nesses diagnósticos. Ao final é feita uma descrição da técnica de diagnóstico pré-implantação e de seu uso in vitro em laboratórios de fertilidade situados a mais de dois mil quilômetros um do outro.

Preimplantation genetics: chromosomes of fragmented human embryos

STUDY OBJECTIVE

To examine the chromosome makeup of fragmented human embryos.

DESIGN

Prospective.

SETTING

Assisted reproductive technology (ART) program.

PATIENTS

One hundred twenty-one poor-quality embryos from 58 patients 31 to 40 years of age admitted for an ART program were examined for chromosome makeup.

RESULTS

Chromosome anomalies were observed in 31.9% (29/91) of poor-quality embryos, 19.8% (18/91) displayed mosaicism (diploid/haploid, diploid/triploid, diploid/aneuploid), 5.5% (5/91) showed polyploidy, 2.2% (2/91) had pulverized chromosomes, 2.2% (2/91) revealed aneuploidy, 1.1% (1/91) had prematurely condensed chromosomes, and 1.1% (1/91) had structural rearrangements involving chromosome number 2. The mean age of patients showing anomalies (36.5 years) was not significantly higher than the mean for the entire group (35.5 years).

CONCLUSIONS

The incidence of chromosome anomalies in fragmented human embryos is high. These anomalies originate either in the gametes or through mitotic nondisjunction within the embryos. It is not advisable to replace such embryos into patients going through in vitro fertilization.

Effect of the degree of maturation of mouse oocytes at fertilization: a source of chromosome imbalance

It has been suggested that the abnormal maturation of the human oocyte during fertilization in vitro may result in chromosome imbalance and induce embryonic loss. Using a mouse model, we have studied the influence of the degree of oocyte maturation (either immaturity or overmaturity) on the chromosome characteristics of embryos at the first-cleavage division. Immature oocytes were obtained 2-3 h or 3-4 h before the expected ovulation time (b.o.). Overmaturation was induced by aging the newly ovulated oocytes in vitro for 3, 6, and 12 h. Our results show a significant decrease in the fertilization rate in the immature groups (65.53% at 2-3 h b.o. and 16.59% at 3-4 h b.o. vs. 78.22% at control) and after 12 h of in vitro aging (69.39%), while a significant increase of this parameter was found at 3 h of aging (82.59%) as compared to the other groups. No significant differences were found in the occurrence of aneuploidy or hyperhaploidy in embryos obtained from immature, newly ovulated, and overmature oocytes. Finally, an increased incidence of polyploidy was detected in immature, 2-3 h b.o. (31.20%), and overmature, 3 h (23.04%) and 6 h (31.61%), groups as compared to the control group (14.59%).

Development of spare human preimplantation embryos in vitro: an analysis of the correlations among gross morphology, cleavage rates, and development to the blastocyst

Following in vitro fertilization, the criteria commonly used to select human embryos for transfer are the cleavage rate and gross morphology, the contention being that those embryos which divide more rapidly and have regular, spherical blastomeres are more likely to lead to a pregnancy. In order to assess the validity of this assumption, the development in vitro of spare embryos was investigated. Eggs and embryos were cultured in Earle's balanced salt solution containing 10% heat-inactivated patient's serum, and insemination was performed at 40 hr post human chorionic gonadotropin (hCG). At 82-90 hr post hCG, up to four embryos were transferred. Any spare embryos were cultured in the same medium for up to 6 days and scored daily for cell number and morphology using a "quality" scale of 4-1 according to degree of fragmentation and shape of the blastomeres. Of 317 fertilized eggs, 55 (17%) developed to the fully expanded blastocyst stage. The remaining embryos ceased development at the one-cell (6; 2%), two-cell (49; 15%), four-cell (110; 35%), eight-cell (61; 19%), and cavitating morula (36; 11%) stages. The relationship between developmental arrest and gross morphology is discussed.

Quality control in the in vitro fertilization laboratory: comparison of bioassay systems for water quality

Three mouse embryo bioassays [mouse one-cell and two-cell embryos and in vitro fertilization (IVF)] were tested for their ability to discriminate between three sources of water for medium preparation: tap water, high-performance liquid chromatography (HPLC)-grade water, and Milli-Q purified water. No differences could be detected using these assays. The lack of sensitivity of the mouse bioassays could not be attributed to the protein source or medium type. The hamster sperm motility assay (HSMA) permitted quantitative discrimination between water sources (Milli-Q greater than HPLC greater than tap). Media prepared for use in human IVF using water that exceeded minimal HSMA quality standards resulted in pregnancy rates that were greater than those attained with a lot of HPLC water that did not meet these standards. The HSMA can serve as a basis for a quality-control program in the human IVF laboratory.

Chromosomal preparations of human triploid zygotes and embryos fertilized in vitro

Forty-eight zygotes with more than two pronuclei were identified after in vitro fertilization, representing 6.1% of all fertilized oocytes. The chromosome preparations from pronuclear stage to the cleaved human embryos were examined. Prophase was found in eight out of ten zygotes. The spreading of chromosomes allowed an adequate counting in only two cases. Six of the eight preparations displayed a late prophase. In this stage each haploid group of chromosomes can be analysed separately. Kariogamy usually occurred 4 to 5 h after the pronuclei had disappeared, and polyploid number of chromosomes were found in well-spread metaphases. The chromosomal preparations were made for eleven human embryos arising from zygotes with three pronuclei. Out of ten preparations, where the chromosomes could be counted, seven embryos (70%) contained hypodiploidic groups of chromosomes. In two of the cases, however, triploid metaphases were found, and in the last one a triploid/diploid mosaicism.

Chromosome analysis of preovulatory human oocytes and oocytes failing to cleave following insemination in vitro

These investigations were carried out within a clinical in vitro fertilization and egg replacement program. A gradual fixation method for chromosome preparations of animal oocytes, when modified and applied to human oocytes, gave numerical analyzability in about 80%. Of preovulatory oocytes recovered in stimulated cycles, 53% of oocytes with numbers of chromosomes in the haploid range had an apparently normal set of 23 chromosomes. Oocytes with a very low number of chromosomes or in first meiosis were also found. In a separate investigation the chromosome constitution of oocytes not showing clinical signs of fertilization 48 hours after insemination in vitro was also studied. Of the oocytes with numbers of chromosomes in the haploid range, 50% had a normal 23,X karyotype. A different distribution of abnormal chromosome preparations was found, compared with the preovulatory oocytes. Twelve oocytes (16%) carrying between 10 and 18 chromosomes were found among the inseminated oocytes, but none in the other study. Moreover, 15 preparations displayed no chromosomes in the inseminated oocytes, whereas in the uninseminated group, none lacked chromosomes completely. These results may indicate how unfertilized human oocytes in vitro undergo chromosome loss and degeneration and could represent different stages of degeneration. Among both preovulatory and unfertilized oocytes the distribution between hyper- and hypohaploid oocytes was of particular interest in that hypohaploid chromosome complements were in excess. This finding supports the hypothesis that aberration in the number of chromosomes in an oocyte is caused not only by nondisjunction but also by anaphase lag, since nondisjunction alone would result in a 1: 1 ratio between hyper- and hypohaploid complements. In conclusion, human oocytes recovered for in vitro fertilization seem to have an incidence of numerical chromosome abnormalities in the region of 40-50%. This is close to the figures estimated on the basis of observations made in spontaneous abortions and could be a major explanation for the low developmental potential of human cleavage stage ova.

Microsurgical enucleation of tripronuclear human zygotes

Polyspermic fertilization of human oocytes in vitro produces genetically abnormal embryos whose replacement in utero represents a potential obstetrical risk. Microsurgical removal of extra male pronuclei offers the possibility that normal ploidy can be restored in these zygotes. Pronuclear removal was attempted in three tripronuclear human oocytes fertilized in vitro. Male pronuclei were distinguished by their larger size and an associated sperm tail piece. Zygotes were pretreated with cytochalasin B and colcemid in phosphate-buffered saline before microsurgery. Enucleation was completed in all embryos; syngamy occurred in one embryo, but cleavage was not observed.

The genotoxic effect of beta-propiolactone on mammalian oocytes

beta-Propiolactone (beta PL) has been tested on preimplantation mouse embryos for possible genotoxic effects. Tests were performed at different stages of meiosis (late prophase I, diakinesis/metaphase I, anaphase I, telophase I/prophase II and metaphase II) by injecting females at various times after the induction of superovulation. Male and female derived chromosome complements from first-cleavage embryos were analysed before syngamy for cytogenetic abnormalities. A higher proportion of diploid oocytes, produced by the non-extrusion of the first or second polar body, was found after fertilization when the compound was administered immediately before metaphase I or II. No obvious effect was detected at any other time of beta PL exposure. Based on these results, several possible modes of action for beta PL are postulated.

Chromosome analysis of human oocytes recovered from preovulatory follicles in stimulated cycles

To investigate the incidence and types of abnormalities of chromosome number in oocytes, we recovered preovulatory oocytes from 17 women who were undergoing clomiphene stimulation and laparoscopy because of infertility. Twenty-three oocytes were recovered and studied after they had been fixed with a gradual-fixation method: 17 of the oocytes had numbers of chromosomes in the haploid range (19 to 25 second-metaphase chromosomes), 4 had only 1 to 5 chromosomes, 1 was not analyzable, and 1 had 23 chromosome bivalents in the first metaphase. Of the oocytes with chromosome numbers in the haploid range, nine had an apparently normal haploid set of 23 chromosomes. Two had 1 to 2 additional chromosomes, three lacked 2 to 4 chromosomes, and three had totals of chromosomes that were close to 23 but could not be determined with certainty. We conclude that infertile women undergoing clomiphene stimulation have a high proportion (nearly 50 percent) of oocytes with an abnormal karyotype. If this is also true of fertile women, including those not taking clomiphene, it may explain the high frequency of chromosome aberrations occurring in early spontaneous abortion and the low pregnancy rate after in vitro fertilization.

A formula for scoring human embryo growth rates in in vitro fertilization: its value in predicting pregnancy and in comparison with visual estimates of embryo quality

Two systems for measuring embryo development in vitro were evaluated. One was a 1-4 scale based on a subjective evaluation of embryo quality (EQ) from microscopic appearance. In addition, a formula for scoring embryo growth rate in vitro was developed. The embryo development rating (EDR) was based on the ratio between the time at which embryos were observed at a particular stage after insemination and the time at which they would be expected to reach that stage in a hypothetical "ideal" growth rate with a cell cycle length of 11.9 hr. Using this scoring system, "normally" growing embryos scored 100. This approach was aimed at partially normalizing the data and allowed all embryos to be analyzed similarly regardless of the time of observation. Analysis of 1539 embryo replacements resulting in 232 clinical pregnancies showed that both EDR and embryo-quality scores were of value in predicting success, with clinical pregnancy most likely to eventuate from a combination of moderate to good EQ scores (2-4) coupled with average or above-average growth rates (EDR scores from 90 to 129). Poor-quality and very slowly or very rapidly growing embryos were underrepresented in cycles that proceeded to pregnancy. These inferences were based on all embryos transferred (mean, 2.73 per transfer cycle), and they were substantiated by an analysis of 33 pregnancies resulting from replacement of a single embryo and from 18 pregnancies in which all embryos scored the same with both systems. EQ and EDR were significantly associated with each other and together provide a valuable guide in predicting pregnancy, in selecting embryos for freezing, and in monitoring day-to-day performance in the in vitro fertilization (IVF) program.

Gestation sac size in in-vitro fertilization pregnancies

The gestation sac size in pregnancies resulting from in-vitro fertilization (IVF) and embryo transfer have been compared with those in spontaneous pregnancies. Small-for-dates gestational sac sizes were found in 36% of the IVF pregnancies. This proportion held for both singleton and multiple pregnancies. With increasing gestation beyond 8 weeks the gestation sac volume increasingly approached normal. In contrast to spontaneous conceptions, IVF pregnancies had a low rate of pregnancy loss once fetal heart movements were demonstrated, when the gestation sac size was small-for-dates. Small sac size in an IVF pregnancy may lead to the misdiagnosis of a failed pregnancy.

Chromosomes of human sperm: variability among normal individuals

The chromosomal constitution of 2468 human sperm cells has been investigated by fusion of human sperm with hamster eggs. The overall frequency of cells with structural aberrations was 7.7%, ranging from 1.9% to 15.8%, and varying significantly among individuals. The highest frequency occurred in sperm from the oldest donor (49 years), who also had had a vasectomy reversal three years prior to sampling. The overall aneuploidy frequency was 1.7%, ranging from 0.6% to 3.1%. In nine out of ten donors from whom blood samples were available the frequency of sperm cells with structural aberrations was higher than that for lymphocytes. Two previously reported donors (Brandriff et al. 1984) were resampled after an interval of 14 and 16 months respectively, and were each found to have similar frequencies of sperm chromosome abnormalities at both sampling times. A father-son pair included in the study had several chromosome breakpoints in common, although no more frequently than unrelated individuals.

Chromosome analysis of multipronuclear human oocytes fertilized in vitro

In an in vitro fertilization and embryo transfer program composed mainly of patients with tubal infertility, the incidence of fertilized oocytes with multiple pronuclei was 4.3%. An attempt was made to fix such oocytes with supernumerary pronuclei in order to assess the chromosome constitution of the gametes. Nine multipronuclear oocytes containing a total of 29 pronuclei were successfully fixed before the first cleavage division, and another oocyte containing 3 pronuclei was fixed before the second cleavage division. On analysis, chromosome counts could be obtained for 28 of the 29 total pronuclei, and 24 gave informative results. Nineteen pronuclei had a normal haploid chromosome count, 2 pronuclei contained one extra chromosome, and 3 pronuclei had one chromosome missing. The presence of a Y chromosome in six pronuclei identified their paternal origin; two of the six sperm pronuclei had a 22,Y,-E chromosome constitution. Only three pronuclei could be conclusively ascertained to be maternal in origin, yet two of these were aneuploid; one pronucleus had a 24,X,+D karyotype and the second had only 22 chromosomes. Multipronuclear oocytes present ideal material for analyzing the chromosome constitution of those human gametes which can undergo fertilization in vitro.

External human fertilization: an evaluation of policy

In vitro fertilization, in its first 5 years of use, has met minimum standards for efficacy and safety, as judged by published clinical reports. It is becoming more widely available as an approach for overcoming sterility in married couples and appears also to be gaining social acceptance in that context. Several technical options presented by the procedure, particularly storage of frozen embryos and embryo transfers involving third-party contributions, are less fully evaluated clinically and raise social, ethical, and legal questions that go beyond the original medical model for therapeutic intervention. The clinical success of in vitro fertilization and the options it affords call for careful policy consideration. Estimates of costs and of potential demand for and supply of services are provided and the current status of relevant policy in the United States and abroad is discussed.