The study of non-invasive euploid embryo selection methods in assisted reproductive technology

Ph.D.Despite advances in Assisted Reproductive Technology (ART), success rates from in vitro fertilization (IVF) remain relatively low. The presence of chromosomal abnormalities has a substantial effect on IVF failure. Preimplantation genetic testing for aneuploidies (PGT-A) has been utilized to select euploid embryos. However, this has a potential detrimental effect on the embryo. Therefore, PGT-A is still not a universal test in IVF.Non-invasive PGT-A (NiPGT-A) on spent culture media (SCM) is another potential means of screening embryos for aneuploidy. Embryonic DNA is released during embryo growth. The culture media in which the embryo has grown is a valuable source of DNA for genetic screening. In this study, I hypothesized that using SCM as a non-invasive tool, the ploidy status can be correlated with SCM, trophectoderm (TE) biopsies, tissue from miscarriages, and live births.In order to test my hypotheses, I examined cell-free DNA (cfDNA) in SCM from patients undergoing IVF/PGT. A total of 168 blastocysts underwent TE biopsies and their SCM was collected. The PGT-A results of TE and SCM were compared. I found cfDNA in only 150 SCM samples, 116 samples of which produced interpretable results. SCM had a concordance rate of 62.1% for autosomes and 82.4% for sex chromosomes, providing a sensitivity of 81.6% and a specificity of 48.3%.The second part of my study was a metabolomics approach using Raman spectroscopy to evaluate the light-scattering patterns by SCM. The molecular components in embryo culture media such as DNA, metabolites, and protein products can be spectroscopically measured and classified according to the ploidy status of corresponding embryos. Preliminary algorithms were developed based on Raman spectra of 25 euploid and 43 aneuploid embryos using artificial intelligence and computer learning. Each algorithm was then tested with 62 blinded SCM of euploid and aneuploid embryos. Raman spectra of euploid and aneuploid SCM was not discriminatory, but competent embryos of ongoing pregnancies showed different spectral patterns in their SCM compared with incompetent embryos. Raman spectroscopy was able to detect viable euploid blastocysts at 100% specificity, and its sensitivity was 53.6% with a positive predictive value at 100%. The measurement is fast and inexpensive compared to PGT-A.The last stage of this study examined the relevance of non-invasive PGT-A in relation to reproductive outcomes. My primary analysis on a set of 45 blastocyst transfers without PGT-A showed that 24 embryos implanted, with 5 resulting in miscarriage. I found that non-invasive PGT-A using NGS on SCM allowed accurate prediction of pregnancy outcomes in 31 cases (73.8%). Now SCM collection, processing, and analyses are ongoing to further investigate the usefulness of SCM as an embryo selection tool.I concluded that non-invasive evaluation of embryos can be achieved through ploidy assessment on SCM. I established the basis for this non-invasive approach, showing positive correlation between SCM results and clinical outcomes. Non-invasive PGT-A is not yet ready for clinical use. This study provides the foundation to further explore how cfDNA in SCM can predict embryo growth and development, and its potential clinical application.儘管輔助生殖技術(ART)取得了進展，體外受精(IVF)的成功率仍然相對較低。染色體異常的問題對IVF成敗具有實質影響。植入前遺傳學檢測(PGT-A)已被用於選擇整倍體胚胎。然而，這對胚胎有潛在的不利影響。因此，PGT-A仍然不是IVF的常規測試。利用廢培養液(SCM)作非侵入性PGT-A(NiPGT-A)是一種可行的篩選胚胎方法。胚胎生長期間釋放胚胎DNA。胚胎使用過的培養液可以為遺傳篩選提供有價值的資料。在這項研究中，我假設利用SCM作非侵入性工具測試倍性狀態，SCM的測試結果會與TE活檢，流產組織和活產相關。為了驗證這假設，我檢查了IVF/PGT病人的SCM中的游離DNA(cfDNA)。收集了168個SCM並為相應的囊胚進行活檢。對比SCM和TE活檢的PGT-A結果。我發現cfDNA僅存在於150個SCM樣本，其中116個樣本有合理的結果。SCM在常染色體的一致率為62.1％，性染色體的一致率為82.4％，靈敏性為81.6％，特異性為48.3％。我的研究的第二部分是利用代謝組學方法以拉曼光譜法評估SCM的光散射模式。可以光譜測量胚胎培養液中的分子組分（例如DNA，代謝物和蛋白質），並根據相應胚胎的倍性狀態進行分類。利用人工智能和計算機學習，以25個整倍體和43個非整倍體胚胎的拉曼光譜開發了初步算法。然後再以62個正常和異常胚胎的SCM樣本盲法測試每種算法。整倍體和非整倍體SCM的拉曼光譜沒有區別性，但有能力持續妊娠的胚胎與失敗的胚胎相比其SCM顯示出不同的光譜模式。拉曼光譜能夠以100％的特異性檢測整倍體囊胚，其靈敏性為53.6％，陽性預測值為100％。與PGT-A相比，測量快速且價錢便宜。本研究的最後階段剖析了NiPGT-A與懷孕結果的相關性。我對45個沒有PGT-A的囊胚移植作初步分析，研究顯示24個胚胎成功着床，其中5個導致流產。我發現以SCM作NiPGT-A能準確預測到31例（73.8％）。現在仍然繼續進行SCM收集，處理和分析，以進一步研究SCM作為胚胎篩選工具的有用性。我的結論是透過SCM作非侵入性的倍性評估，可以實現胚胎篩查。我建立了這種非侵入性方法的基礎，顯示SCM結果與臨床結果之間存在正相關關係。NiPGT-A尚未可以臨床使用。然而，本研究為進一步探索以SCM中的游離DNA預測胚胎生長發育及其潛在的臨床應用提供了基礎。Yeung, Sum Yee Queenie.Thesis Ph.D. Chinese University of Hong Kong 2019.Includes bibliographical references (leaves 109-128).Abstracts also in Chinese; some appendixes also in Chinese.Title from PDF title page (viewed on 20, May 2021)