Obstetric outcome of vanishing twins syndrome: a systematic review and meta-analysis

Due to the high number of multiple embryo transfers into the uterus performed in assisted reproductive technology treatment (ART), the incidences of twin pregnancy and of vanishing twin syndrome (VTS) are correspondingly high. A number of studies have described the obstetric outcomes of the remaining fetus produced after the other twin had vanished compared with a singleton at the start following ART, but the results are mixed and contradictory. We performed a systematic review of the existing studies to explore the actual obstetric outcome of VTS to allow physicians to adequately advise their patients.

A detailed search strategy was used to conduct electronic literature searches (spanning 1978-2015) on Medline, EMBASE, the Cochrane library and Web of Science. As randomized trials are not feasible in this aspect, we included observational (cohort and case-control) studies which compared the obstetric outcomes of the VTS group and singleton at the start control group after ART. The outcomes were evaluated by two aspects, the duration of pregnancy (gestational age, preterm delivery rate, extremely preterm delivery rate) and the birth weight of the fetus [mean birth weight, low birth weight rate, very low birth weight rate and small for gestational age (SGA)].

1271 publications were identified by the initial search. 499 studies were excluded following duplication checks. 760 were excluded after reviewing the abstracts. Of the remaining 12 articles, 7 were excluded after a detailed full-text review. Two case-control studies and three cohort studies were included in the final analysis. The pooled mean gestational age difference (95% confidence intervals) was -0.27 (-0.60, 0.06) and failed to demonstrate a difference between the two groups. A similar result was found in the preterm delivery rate, with a pooled risk ratio of 1.33 (0.91, 1.94). The prevalence of extremely preterm delivery rate was higher in the VTS group, with a pooled risk ratio of 3.5 (1.72, 7.12). The mean birth weight was lower in the VTS group, with a mean difference of -0.3 (-0.59, -0.01). No difference was found in low birth weight rate, very low birth weight rate and rate of small for gestational age, with risk ratio of 1.85 (0.88, 3.86), 4.86 (0.91, 25.91) and 1.29 (0.52, 3.18), correspondingly.

There is a slight adverse effect of VTS on the remaining fetus for birth weight and extremely preterm delivery rate, but sensitivity analysis shows these effects to be statistically unstable. It is too early to draw conclusions for adverse obstetric outcomes for VTS patients. It could reduce much of the anxiety of couples who experience early embryonic loss of one of their twins. More research with rigorously designed and standardized methodologies are required that include larger, better clinically defined populations. Studies that show no correlation should be published in the future to avoid any possible impact of publication bias. After that, patients can receive the most accurate information.