Spent Blastocyst Media and Blastocoel Fluid are not reliable DNA sources for preimplantation genetic diagnosis of aneuploidies and monogenic disorders

Study question: Is DNA from Spent Blastocyst Media (SBM) and Blastocoel Fluid (BF) a reliable template for preimplantation aneuploidy testing (PGD-A) and/or PGD for monogenic disorders (PGD-M)? Summary answer: PGD-A/PGD-M analysis using BF and SBM-derived DNA is unreliable due to high diagnostic failure rate and inconsistency with results derived from trophectoderm biopsy. What is known already: BF has been suggested as a reliable source of embryonic DNA (eDNA) for the assessment of blastocyst chromosomal constitution. BF aspiration entails a smaller degree of invasiveness compared to trophectoderm (TE) biopsy. The literature on the use of BF as source of eDNA for PGD/PGA reports extremely conflicting results and recommendations. Recently, attempts to use eDNA collected in SBM as a template for PGD have been made with poor results. To add further data on both these novel applications, we performed a prospective study investigating the reliability of BF and SBM for clinical diagnostics in PGD cycles. Study design, size, duration: Prospective cohort blinded study performed between December 2015 and December 2016. Three groups of samples were analyzed: BF (n = 80), SBM (n = 42) and TE (n = 84). PGD-A was performed on 23 BF-TE pairs; PGD-M was performed on 38 BF-SBM-TE trios, 19 BF-TE pairs and 4 SBM-TE pairs. TE biopsy results were used as control to assess the reliability of BF- and SBM-based diagnosis. Participants/materials, setting, methods: SBM (25ul) and BF samples were collected from expanded blastocysts, PGD-A on BF was performed by next-generation sequencing (NGS) and compared to qPCR outcomes generated by the corresponding TE. PGD-M analysis included 10 different disorders. Genotyping analysis was performed for the three study groups using 59 TaqMan probes: 48 for informative SNP and 11 for direct mutation analysis. A total of 311 probes were tested for TE, 286 for BF and 242 or SBM. Main results and the role of chance: PGD-A in BF showed 60.9% amplification failure (AF) rate (n = 14/23;95%IC=38.5–80.3); discordance between BF and TE was 17.4% (n = 4/23;95%IC=4.9–38.8) and concordance 21.7% (n = 5/23; 95%IC=7.5–43.7). Sensitivity per chromosome was 66.7% (n = 4/6;95% IC=11.8–61.6) and specificity 91.1% (n = 185/203;95%IC=95%IC=86.6–94.7). For PGD-M genotyping assays, 69.6% (n = 199/286; 95%IC=63.9–74.9) and 10.7% (n = 26/242; 95%IC=7.1–15.3) of probes showed AF on BF and SBM, respectively. Discordant results in 14.3% of BF (n = 41/286; ,95%IC=10.5–18.9) and 35.5% of SBM (n = 86/242; ,95%IC=29.5–41.9) were due to allele drop-out (ADO) and artifacts/contamination. ADO occurred for 12.2% (n = 35/286; 95% IC=8.7–16.6) of probes in BF; artifacts/contamination occurred in 2.1% (n = 6/286; 95%IC=0.8–4.5). For amplified probes from SBM, ADO occurred in 17.4% (n = 48/242;95%IC=12.8–22.7) while artifacts/contamination occurred in 15.7% (n = 38/242;95%IC=11.4–20.9). TE analysis reported just 0.3% of ADO (n = 1/311;95%IC=0.01–1.8) and no amplification failure (n = 0/311). Overall, concordant results were obtained for 16.0% (n = 46/286;95%IC=12.0–20.9) and 53.7% (n = 130/242;95% IC=47.2–60.1) of probes on BF and SBM, respectively. Haplotype reconstruction for PGD-M has been successful and consistent compared to TE in only 3.5% of BF (n = 2/57;95%IC=0.4%-12.1%) and 21.4% of SBM (n = 9/42; 95% IC=10.3–36.8). SBM showed higher amplification and consistent results compared to BF (P < 0.05). Considering probes for direct mutation analysis, AF in SBM was 18.3% (n = 9/49; 95%IC=8.7–32.0) and concordance 51.0% (n = 25/49;95% IC=36.4–65.6); while in BF 70.7% (n = 41/58;95%IC=57.3–81.9) and 15.5% (n = 9/58;95%IC=7.3–27.4) respectively. Limitations, reasons for caution: It does not exist a standard protocol for BF and SBM samples’ retrieval. This may limit the reproducibility and reliability of the procedure across different studies. DNA in BF and SBM may originate from apoptotic cells and contamination and this hypothesis has not been tested in the study. Wider implications of the findings: BF and SBM cannot be considered reliable sources of eDNA for diagnostic purposes, especially due to significantly lower amplification rates compared to TE biopsy. However, BF and SBM can still be considered important specimens for research purposes, which could contain novel biomarkers of reproductive competence beyond chromosomal constitution. Trial registration number: None.