Ultrasound surveillance of the cervix during pregnancy in diethylstilbestrol-exposed offspring.

Twenty-three diethylstilbestrol (DES)-exposed patients were evaluated through 27 pregnancies to determine their eligibility for admission to a prospective protocol that combined serial ultrasound surveillance of the lower uterine segment-cervical complex with periodic pelvic examinations to diagnose cervical incompetency. Of these, 21 pregnant women, including seven vaginectomy patients, were matched to 84 low-risk controls to determine the following: 1) the effect of DES exposure on reproductive performance, 2) the efficacy of ultrasound selection of cerclage candidates, and 3) the influence of previous partial vaginectomy on reproductive outcome. Five DES-exposed patients were diagnosed as having cervical incompetency and had cerclages placed. There were no missed diagnoses of cervical incompetency. The DES-exposed patients delivered statistically earlier in gestation than did controls (268 +/- 13 versus 276 +/- 10 days). It was not evident that this difference was important clinically, as there were no neonatal deaths, very low birth weight infants, second-trimester losses, or deliveries before 252 days (36 weeks) among the study patients. Previous vaginectomy did not affect the frequency of the diagnosis of cervical failure or the neonatal outcome. After ultrasound surveillance and treatment for incompetent cervix, a majority of our patients delivered at term without cerclage placement. Therefore, routine cerclage placement is not recommended. Knowledge of the ultrasound criteria for diagnosing cervical incompetency is required.