Non-Invasive Assessment of Viability in Human Embryos Fertilized in Vitro.

Human reproduction is a relatively inefficient process and therefore the number of infertile couples is high. Assisted reproductive technologies (ART) have facilitated the birth of over five million children worldwide. ART, however, superimposes its own relative inefficiency on the preexisting inefficiency of normal reproduction. The efficiency (expressed as pregnancy rate) is generally not more than 30%. Modern reproductive medicine is gradually moving from multiple embryo transfer to the transfer of a single embryo, mainly because of obvious and unwanted side effects of multiple embryo transfer (e.g. "epidemic" multiple pregnancies). This concept, however, requires a fast, professional selection of the most viable embryo during the first few days of ART. Thus the aim of a modern ART is the safe transfer of a healthy, viable, single embryo. Accurate and rapid methods of quantifying embryo viability are needed to reach this goal. Methodological advances have the potential to make an important contribution, and there has been a drive to develop alternative non-invasive methods to better meet clinical needs. Metabolic and genetic profiling of spent embryo culture (SEC) media should offer an exceptional opportunity for the assessment of embryo viability. The current review focuses on the latest non-invasive diagnostic approaches for pre-implantation viability assessment of in vitro fertilized embryos.

INTRODUCTION

Infertility has been recognized as a public health issue worldwide (1) leading to an increasing need to the use of assisted reproductive technologies (ART), including in vitro fertilization (IVF). After the first reported case of IVF in 1978(2) ART enabled millions of people to have their own children in cases when pregnancy did not occur under natural circumstances. ART has advanced significantly and became more and more widespread resulting in ca. 700,000 cycles a year in the USA and Europe together (3,4). Despite of evolving intracytoplasmatic sperm injection (ICSI) technique the rate of the successful embryo implantations is surprisingly low (5,6). A success rate of 25% and 28% has been reported in 2005(7) and 2008(8), respectively. Nowadays, this rate went up to 32% (9), which cannot be considered as a significant development. Earlier clinical protocols preferred multiple embryo transfer, but multiple gestations can result in the increased risk of preterm delivery (10-16). Other studies report that multiple gestations also increased the risk of low birth weight cerebral palsy (17). In the US alone, preterm births resulting from multiple pregnancies during IVF cause a 1 billion USD extra cost to the social insurance (18). In order to exclude the discussed risk factors, single embryo transfer becomes the standard of care for all. It is imperative, however, that accurate and economical methods should be developed to ensure that the most viable euploid embryo is selected for transfer. Ideally, such tests would be noninvasive, lessening the risks to the embryo and reducing costs and workload in the embryology laboratory (19). The biggest issue with pre-implantation viability assessment is that due to ethical reasons any assay should be completely non-invasive because no one can predict what kind of interference would be the unwanted result in the later embryonic development.

THE MORPHOLOGICAL APPROACH

The most apparent – and routinely applied - way of the assessment of viability is the morphological evaluation of in vitro fertilized embryos using microscopy. There are several morphological features described which could be used for viability assessment purposes, these are dependent on the time spent after fertilization. Right after fertilization in the 1-cell embryo the size and symmetry of the two pronuclei can be examined. The time of the first cell division is also a good predictor of later implantation potential, as zygotes that divide early tend to develop more frequently to the blastocyst stage. Criteria as cleavage rate and blastomere shape and symmetry, an adequate trophectoderm layer (TE) and an inner cell mass (ICM) is a morphological marker of the later stages (5,20). Not only can the morphology of the fertilized embryo be used for further prediction of implantation potential, but morphological defects of the retrieved oocyte as well. Fertilization and pregnancy rate correlates with the grade of cumulus-oocyte complexes, and embryos originating from dysmorphic oocytes show a larger grade of pregnancy loss (21-23). The cleavage stages of morulae and blastocysts or the symmetry and patterns of cell division are also notable and frequently used aspects, and are often examined during the prediction of embryo viability (23). The biggest issue of morphological assessment is that it is still a highly subjective method (20). The reason is partly due to the fact that the final decision is made by a clinician, and not by an objective test result, and secondly it is does matter how important are the individual morphological features in the final conclusion (24-26). To overcome the different practice of laboratories worldwide in 2011 an international consensus (Istanbul Consensus) has been reached on embryo viability assessment (27). The selected morphological markers of respective stage embryos, the weighing of individual features and a scoring system has been set up. The limitations due to static time-point observation, is now solved with the use of time-lapse microscopy (28,29). Time-lapse microscopy also enables the observation of dynamics of cytoplasmic movements and cytokinesis, reflecting the functionality of microtubule and actin cytoskeleton, which is critical for proper development. In our laboratory, we aimed to improve the success rate of implantation by adopting and further optimizing the Istanbul consensus. This score has been called as the optimized criteria system (OCS). According to this scoring, 3-Day old embryos were divided into two subgroups: the subgroup with low blastomere number (less than 7) and with high blastomere number (7 or more). Symmetric position of blastomeres indicates the rate of symmetry of holoblastic cleavage along the embryo axis. It was classified as good (full symmetry); fair (light asymmetry); or poor (evident asymmetry). The percent values of fragmentation are based on the ratio of fragmented to total cell numbers. As a further modification to the Istanbul consensus, the assessment of fragmentation was slightly changed. Embryos were considered as good if the fragmentation rate was <15% (instead of the original 10%). This shift from 10 to 15% was the result of our observation that a fairly high proportion of the embryos between 10-15% appeared to be viable. In summary the optimized criteria system (OCS) highlights 3 modified or new parameters: fragmentation (with a more permissive criterion of <15% in the “good” category); symmetry and the blastomere number. In addition, the blastomere size was evaluated according to the original Istanbul consensus. A scoring-map was created to facilitate the evaluation (Table 1) As far as the 5-Day old embryos are concerned, we modified the original Istanbul Consensus for blastocysts by leaving out the hatched stage from the evaluation. The Istanbul Consensus for the 5-day old embryos has a shortcoming, i.e. it does not express the viability of embryos with a single category (good, fair, poor). We tried to overcome this by using a scoring map (Table 2). In conclusion, we constructed a composite score for Day-3, as well as Day-5 old embryos, based on morphological parameters. As it is evident from the results, this composite score is sensitive to evaluate viability (Figure 1)

Table 1

The composite score of the “optimized scoring system” for Day-3 old embryos

ICCS for cleavage stage embryos
Good	Fair	Poor
<10% fragmentation	10–25% fragmentation	Severe fragmentation (>25%)
Stage-specific cell size	Stage-specific cell size for majority of cells	Cell size not stage specific
No multinucleation	No evidence of multinucleation	Evidence of multinucleation

Table 2

The composite score of the “optimized scoring system” for Day-5 old embryos

ICCS for blastocysts (Day-5)
Stage
1	Early blastocyst
2	Blastocyst
3	Expanded blastocyst
4	Hatched/hatching
ICM
Good	Prominent, easily discernible, with many cells that are compacted and tightly adhered together
Fair	Easily discernible, with many cells that are loosely grouped together
Poor	Difficult to discern, with few cells
TE
Good	Many cells forming a cohesive epithelium
Fair	Few cells forming a loose epithelium
Poor	Very few cells

Figure 1

The sensitive “optimized scoring system”

Day-3 old embryos (left panel): Pregnancy occurred only in the group evaluated as good quality embryo. No pregnancies occurred if the embryos were assigned to the fair or poor quality groups. It should also be noted the good quality embryo does not necessarily mean pregnancy (maternal causes of infertility may be present).

Day-5 old embryos (right panel): With the exception of a very few cases, most pregnancies occurred in the good quality group.

THE BIOCHEMICAL APPROACH

Another possibility for non-invasive embryo viability assessment is the metabolomic examination of the culture medium surrounding the in vitro fertilized embryo. Metabolomic, (proteomic) profiling of spent embryo culture (SEC) offers an exceptional, non-invasive opportunity for the assessment of embryo viability (30,31). The metabolomic profiling (32,33) of early embryo development might mean the analysis of the total metabolome by following the changes of several selected compounds, metabolomic analysis using unidentified, but significantly differing metabolomic changes, or by the analysis of a limited population of nutrients or end products. The common feature in all three concepts is that they are concentrating on the metabolomic alterations caused by differently developing embryos in the culture medium. Very simple idea is the monitoring of glucose consumption or pyruvate formation, since this would directly indicate the metabolism of the developing embryo and it is an obvious conclusion that a metabolically active embryo would have higher implantation potential. Some authors reported that the identification of these parameters resulted in successful prediction of embryo implantation potential, but other research groups describe contradictory results (34, 35). The amino acid profile of culture media is also used in the prediction of implantation potential, though not exclusively as an independent parameter, rather in combination with morphological features (36). The detection of unidentified metabolomic changes using near infra-red (NIR) or Raman spectroscopy (37, 38) is a very interesting and challenging possibility.

More complicated is the concept when unknown, new biomarker molecules of embryo viability are searched for, assuming that these biomarkers were secreted by the embryo. The difficulty of the concept is that only 4-8 cells are present in the culture medium; thus a very sensitive analytical tool is required. Mass spectrometry (MS) has the potential of specific and sensitive quantification in a wide spectrum of molecular mass ranges and therefore suites well the needs of metabolomic or proteomic fingerprinting and quantification. In parallel to the spreading of mass spectrometry, proteomics is also an emerging field in the understanding of embryo development (39,40). The analysis of the embryonic secretome (41,42) provides information of the total transcriptome of the developing embryos. Mass spectrometry can be used both in targeted and discovery analysis with accurate quantification of identified biomarkers after molecular identification by bottom-up or top-down proteomics using tandem or multiple MS (43-46).

In a recent publication from our laboratory (47) using liquid chromatography coupled mass spectrometry (LC-MS), a fragment of the human haptoglobin molecule was identified in the culture medium. Rather than analyzing the embryonic secretome, the aim this experiment was to use preexisting molecules present in the cell culture media as biomarkers. Haptoglobin - which was detected in the culture medium - is not a product of the developing embryo; the polypeptide is a contaminant of the human serum albumin standard used to supplement the culture medium (47,48). During the first three days of embryo development the formation of a subunit (alfa-1) of the human haptoglobin molecule was observed. This subunit similar to the total haptoglobin molecule was detectable in the blank control medium samples as well. The differentiation of the viable and non-viable embryos was done using the observation that compared to blank controls the samples of embryos which later did not resulted in pregnancy contained the alpha-1 subunit in a much larger quantity than the samples of embryos which did (Figure 2). 160 samples of 77 Day-3 old embryos were analyzed. Clinical statistical analysis of the results revealed that the specificity of the diagnostic test was 64%, while the sensitivity was 100%. It is more informative that the positive predictive value of the assay was 51% and – maybe more importantly – the negative predictive value was 100%.

Figure 2

The mean difference in the amount of the haptoglobin alpha-1 peptide fragment between viable and non-viable embryos

The mean difference in the amount of the haptoglobin alpha-1 peptide fragment between viable and non-viable embryos.

The quantification of the biomarker (n=160) was carried out by mass spectrometric detection following reverse-phase HPLC separation; the analyzed sample volume was 25 µl. In the samples of the non-viable embryos (no pregnancy) the fragment was present in a significantly larger (p<0.001) content than in the sample of the viable (live birth) embryos.

Receiver operating characteristic (ROC) analysis provides tools to select possibly optimal models and to discard suboptimal ones. ROC analysis is related in a direct and natural way to cost/benefit analysis of diagnostic decision-making. The ROC curve of the morphological versus metabolomic approach in relation to the correct prediction of pregnancy outcome is illustrated in Figure 3. It is obvious that our biochemical investigation method enables a selection of the embryos by sorting out the non-viable ones. The test selected with 100% potential the embryos, which did not lead to successful implantation at all.

Figure 3

Receiver operating characteristics (ROC) curve of the blinded, retrospective cohort study

Embryo viability prediction was made using the data of the mass spectrometric haptoglobin alpha-1 chain quantification. n=160, the area under the curve (AUC) is 0.906.

A ROC curve (blue line) visualizes a calculation by graphing the sensitivity on the y-axis (vertical) and the false positive rate (1-specificity) on the x-axis (horizontal) for all possible cut-off values of the diagnostic test. The green line is the reference line. The AUC is used as a measure of the performance of a diagnostic test against the ideal and may also be used to compare different tests.

One of the areas of collaboration between clinicians, the clinical laboratory and the research laboratory at the University of Pécs is related to the research of infertility. Since the clinical background gives the beauty and the medical importance of laboratory research, it was of outstanding importance for us to receive the EFLM-Abbott Diagnostics Award for Excellence in Outcomes Research in Laboratory Medicine (Paris, 2015), the award given to the best published paper (47), as judged by an independent panel of experts, which demonstrates improved outcomes arising out of the application or improved utilization of an in-vitro diagnostics test. This short review summarizes some of our recent findings and views on this field.