Strengths and limitations
The major strength of this study was the consecutive observational approach and prospective collection of blood samples. Previously, several studies have reported the disturbance of NIPT by residual cfDNA of twin demise [11, 21, 22]. However, very few studies demonstrate the longitudinal effect of residual cfDNA after twin demise. We successfully established the consecutive profiles of residual cfDNA changes after twin demise by densely testing blood samples until birth, which provides the robust evidence to elucidate an important question: how long the placenta of a deceased co-twin can release cfDNA into the maternal circulation? Another strength of this study was a combination of three different methods of calculating fetal fractions. This helps compare the total fetal fraction measured by FF-QuantSC, the fetal fraction of the demised co-twin measured by relative coverage of the trisomy chromosome, and in two cases (patient 3 and 4) the fetal fraction of the survived euploid fetuses measured by Y-chromosome.
The limitations of this study included a small sample size and the restriction of selecting DCDA twins with a trisomy co-twin to simulate a natural twin demise. Due to the small sample size, statistical analysis could not be conducted. For the ease of analyzing the individual fetal fraction of different co-twins, we only used dizygotic twins in our study. It is not clear if the same observations can be replicated in monozygotic twins, especially monochorionic twins who have a significantly higher risk of fetal demise [23, 24]. Moreover, the results obtained using the reduced trisomy co-twin might not fully represent natural twin demise, which can be caused by many other reasons with or without chromosomal abnormalities [7, 25]. Besides, in our study, women opted for reduction at 14 to 17 gestational weeks, thus missing the opportunities to observe residual cfDNA at early pregnancy when most vanishing twins occur. Lastly, because of the difficulties during clinical practice, the time points of sample collection in each woman were not identical. In one case (patient 2), the blood sample before the reduction was missing.