Efficacy of second-trimester selective termination for fetal abnormalities: international collaborative experience among the world's largest centers.

OBJECTIVE

Our goal was to develop the most comprehensive database possible to counsel patients about selective termination for fetal abnormalities, because no one center has sufficient data to assess much more than crude loss rates.

STUDY DESIGN

A total of 183 completed cases of selective termination from 9 centers in 4 countries were combined (169 twins, 11 triplets, 3 quadruplets). Variables included indications, methods, (potassium chloride, exsanguination, air embolus), gestational age at procedure, pregnancies lost (< or = 24 weeks), gestational age at delivery, and neonatal outcome.

RESULTS

Indications for selective termination were 96 chromosomal, 76 structural, and 11 mendelian. Selective termination was technically successful in 100% of cases. In 23 of 183 (12.6%) miscarriage occurred before 24 weeks; 2 of 37 (5.4%) occurred when the procedure done at < or = 16 weeks and 21 of 146 (14.4%) when it was done thereafter. Air embolization had a higher loss rate: 10 of 24 (41.7%) compared with 13 of 156 (8.3%) by potassium chloride (chi 2 = 117, p < 0.0001). Three cases of selective termination performed in monochorionic pregnancies all resulted in pregnancy loss. Among 183 potentially viable deliveries, 7 occurred before 28 weeks, 19 at 29 to 32 weeks, 41 at 33 to 36 weeks, and 93 at > or = 37 weeks. Gestational age at delivery was not influenced by the technique used or the indication but was negatively correlated with gestational age at the time of selective termination. No coagulopathy or ischemic damage was observed in survivors. There was no maternal morbidity.

CONCLUSIONS

(1) Selective termination in experienced hands for a dizygotic abnormal twin is safe and effective when done with potassium chloride. A total of 83.8% of viable deliveries occurred after 33 weeks and only 4.3% at 25 to 28 weeks. (2) Gestational age at the procedure correlated positively with loss rate and inversely with gestational age at delivery; this emphasizes the need for early diagnosis in multifetal pregnancies. (3) Coagulopathy tests are probably unnecessary.