**2. cfDNA research based on Spent culture media (SCM)**

#### **2.1 The discovery of cfDNA in SCM**

Stigliani et al. [18] first demonstrated the presence of cfDNA in SCM in 2013, where genomic DNA (gDNA) was found in 63% (205/326) of the 326 SCMs collected, ranging from 41 pg to 1.8 ng. Subsequently, Hammond et al. [19] also detected mitochondrial DNA (mtDNA) and gDNA in SCMs. In addition, Stigliani et al. [22] reported that the ratio of mtDNA to gDNA in the SCM of D3 was positively correlated with the formation rate of blastocysts. These results suggest that SCM can serve as a sampling source for early embryonic DNA, laying the foundation for the development of the non-invasive method of PGT-A.

#### **2.2 Application of SCM in PGT-A**

In recent years, an increasing number of studies have been conducted to evaluate the feasibility of SCM-based non-invasive PGT-A approaches, and some of the results are encouraging. Many studies have found that PGT-A by trophectoderm (TE) biopsy or whole embryo of SCM has a high consistency rate (as shown in **Table 1**. Pro). Nevertheless, other research groups have reported relatively low consistency rates when compared with TE or whole embryo results (see **Table 1**. Con). These contradictory results may be related to factors such as the difference in SCM sampling methods and sampling time used in the studies, embryo treatment, and the definition of consistency. By improving the detection method, non-invasive PGT-A using SCM is expected to meet the requirements of clinical diagnosis.

#### **2.3 Factors affecting the accuracy of non-invasive PGT-A**

When comparing the ploidy consistency of cfDNA in SCM, all studies have reported a high success rate of cfDNA amplification and detection, ranging from 77.3% to 100% (as shown in **Figure 1**). However, its consistency to the conventional PGT-A results fluctuated widely, ranging from 33% to 100% [23, 26, 27, 39] (as shown in **Figure 2**). Hence, it is critical to recognise the factors influencing accuracy.


#### *Non-Invasive Chromosome Screening for Embryo Preimplantation Using Cell-Free DNA DOI: http://dx.doi.org/10.5772/intechopen.98416*


*Infertility and Assisted Reproduction*

**86**


*Non-Invasive Chromosome Screening for Embryo Preimplantation Using Cell-Free DNA DOI: http://dx.doi.org/10.5772/intechopen.98416*


 **1.** *Concordance*

 *rates between niPGT-A based on SCM and TE or WB.*

*Infertility and Assisted Reproduction*

**88**

*Non-Invasive Chromosome Screening for Embryo Preimplantation Using Cell-Free DNA DOI: http://dx.doi.org/10.5772/intechopen.98416*

**Figure 1.**

*Success rates of WGA amplification and CNV detection reported in different niPGT studies.*

#### **Figure 2.**

*CNV concordance rate reported in studies comparing SCM with TE or whole embryo samples.*

#### *2.3.1 Influence of sampling time on accuracy*

At present, there are two common types of commercial embryo culture systems: sequential culture and single culture systems. However, the quality of DNA in SCM degrades as a result of cfDNA degradation over time. Therefore, the time and opportunity for cfDNA degradation could be minimized via media changes. Additionally, the maternal DNA contamination introduced by residual cumulus cells could be reduced or eliminated by media change to a certain extent. The results of a few studies using a single culture system [36–38] and sequential culture medium [31, 39, 41] are listed in **Table 1**. Importantly, Rubio et al. showed no significant

difference in the influence of different culture systems on consistency when performing the same sampling process [29] (as shown in **Figure 3**).

As the fertilized egg develops, DNA is constantly released into the SCM, which is a dynamic process. The test results of SCM collected at different time points during embryo culture may vary. Using a single medium, Ho et al. [37] achieved detection success rates of 39% (16/41) and 80.4% (33/41) when SCMs were collected on D3 and D5, respectively. When comparing the SCM collected on D3-D5 and D4-D5, Lane et al. [43] found that the accuracy was higher in later samples, with >95% ploidy consistency and 100% consistency of the sex chromosome. The primary explanation for these observations may be that the number of embryonic cells increases exponentially with in line with embryonic development and that the concentration of cfDNA surges dramatically at later stages of the development. Rubio et al. [28] transferred embryos into new culture drops onD4 and collected SCM on D5 to D7. The SCM consistency of D4-D6/7 was significantly higher than that of D4-D5 (84.0% vs. 63.0%), and the level of maternal contamination was also reduced. A multicentre clinical study which supports the above assumption. conducted by the same team that compared SCM and TE samples using a large sample size of 1,301 embryos, also achieved a remarkable 78.2% (866/1108) concordance rate using D4-D6/7 SCM samples [29]. While the consistency rate jumped to 92.0% if the culture were last until D6.

The above studies have shown that to maximize the collection of cfDNA content and ensure the detection rate and accuracy of cfDNA, there is an optimal collection time for SCM on the premise of ensuring the blastocyst rate and excellent embryo rate. For the fresh ET cycle, D4 to the pre-cryopreservation blastocyst stage (D5/D6/D7) could be the optimal collection time. Several studies of ni-PGTA have also been conducted on frozen-thaw embryos, where the culture media were collected at later stages and the assay performance was slightly better compared to that of the fresh embryos (as shown in **Figure 4**).

#### *2.3.2 Fresh-frozen embryos yield better niPGT performance than fresh embryos*

In cryopreservation of embryos, assisted blastocyst shrinkage is usually performed during vitrification of the embryo. Shrinkage of the blastocyst cavity before vitrification can prevent the formation of ice crystals and improve the survival rate of embryos after cryopreservation [44]. This process leads to lysis of cell membrane, which increases the likelihood of that embryonic cells releasing cytoplasmic materials, including genomic DNA, into the SCM, thereby increasing

#### **Figure 3.**

*CNV concordance rates of SCM versus TE biopsies in using different brands of culture media. No statistical differences were observed in four major brands of media. Data from Rubio et al. [42] (Ref). Data from Rubio et al. Embryonic cell-free DNA released to the spent blastocyst media. Am J Obstet Gynecol 2020.*

*Non-Invasive Chromosome Screening for Embryo Preimplantation Using Cell-Free DNA DOI: http://dx.doi.org/10.5772/intechopen.98416*

#### **Figure 4.**

*Comparing SCM collection strategies and its impact on CNV concordance rate.*

the concentration of DNA templates available for the downstream whole genome amplification (WGA) of niPGT-A. Furthermore, the vitrification process may lead to a higher rate of cell apoptosis, which also potentially increases the amount of cfDNA in the media after thawing and recovery. Kuznyetsov et al. [23] compared the amplification outcomes of a mixture of SCM and blastocyst fluid collected before vitrification (19) and after vitrification (30). The average amount of DNA obtained from fresh embryo culture amplification (6.3 ng/μL 36.0 ng/μL) was lower than that obtained from vitrification-revival blastocysts (10.5 ng/μL 44.0 ng/μL). After thawing 52 donated frozen blastocysts with TE biopsy results, Huang et al. [24] found that 48 of them (92.3%) were successfully amplified, and the data could be used for the analysis, with a consistency of 93.8% (45/48) to the whole embryo results. Jiao et al. [25] assessed 41 frozen blastocysts donated by 22 couples with known chromosomal rearrangements and 21 frozen blastocysts donated by 8 couples with normal karyotypes for PGT-SR and PGT-A analysis, respectively. All BCM (a mixture of SCM and BF) samples (62/62, 100%) were amplified successfully, and the clinical consistency between the niPGT-A results and the whole embryo was 90.48% (19/21). For PGT-SR results, the clinical consistency of BCM in chromosomal rearrangement and its corresponding whole embryo sample was 100% (41/41). Li et al. [26], using similar sampling methods, assessed 41 embryos and 97.6% were successfully amplified, and the clinical consistency between the niPGT-A results and the whole embryo was 87.2%. In summary, the SCM of cryopreserved-and-thawed embryos may contain more cfDNA than fresh embryos, which would minimize amplification failures and improve the reliability of the test results. This finding may be beneficial for patients who have had miscarriages due to embryo chromosomal abnormalities in the previous ART cycle. Providing that patients have sufficient cryopreserved embryos, they may take advantage of niPGT-A to screen euploid embryos for transplantation prior to the next transplantation cycle with resuscitated media (as shown in **Figure 4**).

#### *2.3.3 Methods of whole-genome amplification*

Given the challenge of low-concentration cfDNA in SCM, a WGA method with high uniformity and fidelity is required to amplify a small amount of genetic

material to accurately detect the genetic status of embryos. Different WGA techniques discussed in the chapter include multiple displacement amplification (MDA), multiple annealing and looping-based amplification cycles (MALBA) and Sureplex/Picoplex (as shown in **Figure 5**).

MALBAC has a unique characteristic similar to linear amplification, which can reduce the sequence-dependent bias exacerbated by exponential amplification [34]. Several niPGT-A studies have used MALBAC to amplify cfDNA in SCM, and the success rates ranged from 90.9% to 100% [24, 31, 35, 38]. In 2019, Jiao et al. developed NICS-INST technology based on MALBAC, which incorporated WGA and library preparation in the same amplification step, achieving a 100% (41/41) detection rate in SCM [25].

Picoplex/Sureplex is currently the most widely used WGA methods for conventional PGT-A. One study used Picoplex for SCM amplification with a success rate of 97% [37]. In another study using the Sureplex amplification method, the amplification rate of 22 SCM samples reached 81.8%, and all the amplified samples produced PGT-A results by aCGH [36]. A study using Sureplex WGA in combination with blastocyst fluid (BF) and SCM found amplification rates of 100% in 28 samples, all of which produced PGT-A results [23]. In one SCM study with a large sample size of 168, 89.3% of the samples were successfully amplified by Sureplex, and 77.3% of the sequenced samples yielded PGT-A results [34]. Lledo et al. compared Veriseq (Illumina®) and NICS (Yikon®), which both achieved 92.4% amplification success rates. The consistency of the culture medium collected on D6 and TE biopsy was 92.0% and 86.5%, respectively [30].

MDA is an isothermal amplification reaction using Phi29 DNA polymerase [45]. Amplification bias due to nonlinear amplification remains a significant fault in the technique. Studies using MDA on SCM have shown an amplification success rate of 97%; however, only 2% of the amplified samples produced reliable PGT-A results [39]. Since the cfDNA of SCM is dominated by short fragments with a length of 160–220 bp [46], this would especially affect the MDA-based method, which requires longer DNA fragments to achieve optimal amplification.

The Phi29DNA polymerase used in MDA technology has high fidelity but is restricted by the starting amount of the DNA template. When the starting amount is very low, the coverage rate and accuracy of the amplified products can decrease

#### **Figure 5.**

*Comparing SCM amplification methods and the concordance of CNV in different study MALBAC and SurePlex, PicoPLEX and IonReproseq are the main methods, MDA is not applicable.*

dramatically. Malbac and Sureplex have the advantages of high genome coverage, low ADO rate, and a low number of starting DNA templates required. However, DNA polymerase has lower fidelity than Phi29 polymerase, which leads to an increased false positive rate compared to MDA.
