Algorithms to minimize the effect of maternal contamination in non-invasive chromosome screening.

Research Question: Can algorithms be developed to assess maternal contamination in the embryo spent culture medium without parental samples and correct copy number variations (CNV)?

Design: An algorithm was developed to confirm the proportion of maternal contamination based on the mosaic ratios of X and Y chromosomes and an algorithm to correct copy number variation (CNV) in embryos with maternal contamination. Performance of algorithms was tested in 317 culture medium samples. A prospective clinical study was then conducted to validate the accuracy.

Results: Among 55 male embryos with maternal contamination, the consistency of the corrected CNV results with the whole embryo results was higher than that of the CNV results before correction (69.1% versus 10.9%, P < 0.001). Among the 317 embryo culture medium samples, maternal contamination was identified in 88 (27.8%) samples through the algorithm based on the mosaic ratios of X and Y chromosomes and quantitative parental contamination test. Upon re-assessment using the CNV correction algorithm, two and five of the 19 samples deemed as euploid initially were re-classified as mosaic and aneuploid, respectively; 23 and three of the 36 samples deemed as mosaic initially were re-classified as aneuploid and euploid, respectively. In the clinical study, the live birth rate was 44.7%. Accuracy of the algorithms was assessed in 25 donated embryos and 47 cases with prenatal testing. The overall concordance rate was 95.8% (69/72).

Conclusions: The CNV correction algorithm could be used to minimize the effect of maternal contamination in culture medium and enhance the accuracy of non-invasive chromosome screening.