A pilot investigation of low-pass genome sequencing identifying site-specific variation in chromosomal mosaicisms by a multiple site sampling approach in first-trimester miscarriages

Abstract

Abstract STUDY QUESTION Can multiple-site low-pass genome sequencing (GS) of products of conception (POCs) improve the detection of genetic abnormalities, especially heterogeneously distributed mosaicism and homogeneously distributed mosaicism in first-trimester miscarriage? SUMMARY ANSWER Multiple-site sampling combined with low-pass GS significantly increased genetic diagnostic yield (77.0%, 127/165) of first-trimester miscarriages, with mosaicisms accounting for 17.0% (28/165), especially heterogeneously distributed mosaicisms (75%, 21/28) that are currently underappreciated. WHAT IS KNOWN ALREADY Aneuploidies are well known to cause first-trimester miscarriage, which are detectable by conventional karyotyping and next-generation sequencing (NGS) on a single-site sampling basis. However, there are limited studies demonstrating the implications of mosaic genetic abnormalities in first-trimester miscarriages, especially when genetic heterogeneity is present in POCs. STUDY DESIGN, SIZE, DURATION This is a cross-sectional cohort study carried out at a university-affiliated public hospital. One hundred seventy-four patients diagnosed with first-trimester miscarriage from December 2018 to November 2021 were offered ultrasound-guided manual vacuum aspiration (USG-MVA) treatment. Products of conception were subjected to multiple-site low-pass GS for the detection of chromosomal imbalances. PARTICIPANTS/MATERIALS, SETTING, METHODS For each POC, multiple sites of villi (three sites on average) were biopsied for low-pass GS. Samples with maternal cell contamination (MCC) and polyploidy were excluded based on the quantitative fluorescence polymerase chain reaction (QF-PCR) results. The spectrum of chromosomal abnormalities, including mosaicism (heterogeneously distributed and homogeneously distributed) and constitutional abnormalities was investigated. Chromosomal microarray analysis and additional DNA fingerprinting were used for validation and MCC exclusion. A cross-platform comparison between conventional karyotyping and our multiple-site approach was also performed. MAIN RESULTS AND THE ROLE OF CHANCE One hundred sixty-five POCs (corresponding to 490 DNA samples) were subjected to low-pass GS. Genetic abnormalities were detected in 77.0% (127/165) of POCs by our novel approach. Specifically, 17.0% (28/165) of cases had either heterogeneously distributed mosaicism (12.7%, 21/165) or homogeneously distributed mosaicism (6.1%, 10/165) (three cases had both types of mosaicism). The remaining 60.0% (99/165) of cases had constitutional abnormalities. In addition, in the 71 cases with karyotyping performed in parallel, 26.8% (19/71) of the results could be revised by our approach. LIMITATIONS, REASONS FOR CAUTION Lack of a normal gestational week-matched cohort might hinder the establishment of a causative link between mosaicisms and first-trimester miscarriage. WIDER IMPLICATIONS OF THE FINDINGS Low-pass GS with multiple-site sampling increased the detection of chromosomal mosaicisms in first-trimester miscarriage POCs. This innovative multiple-site low-pass GS approach enabled the novel discovery of heterogeneously distributed mosaicism, which was prevalent in first-trimester miscarriage POCs and frequently observed in preimplantation embryos, but is currently unappreciated by conventional single-site cytogenetic investigations. STUDY FUNDING/COMPETING INTEREST(S) This work was supported partly by Research Grant Council Collaborative Research Fund (C4062-21GF to K.W.C), Science and Technology Projects in Guangzhou (202102010005 to K.W.C), Guangdong-Hong Kong Technology Cooperation Funding Scheme (TCFS), Innovation and Technology Fund (GHP/117/19GD to K.W.C), HKOG Direct Grant (2019.050 to J.P.W.C), and Hong Kong Health and Medical Research Fund (05160406 to J.P.W.C). The authors have no competing interests to declare. TRIAL REGISTRATION NUMBER N/A.