The split-membrane sign: A new sonographic marker to characterize chorionicity in twin gestation

OBJECTIVE: Several sonographic studies have documented a high degree of accuracy in determining twin chorionicity in either the first or second trimester. That accuracv drops significantly in the ihird trimester due to difficult visualization and to thinning of the dividing membrane. The pin -pose of this study is to describe our prospective ongoing experience with a previously undescribcd sonographic marker of dichorionicity in the third trimester; ihe split-membrane sign. STUDY DESIGN: Sixty diamniotic twin pregnancies with a single placental disc were prospective!}' followed with serial ultrasounds every four to six weeks. The insertion of the- dividing membrane into the placental disc was examined throughout its length. Ihe twin peak sign was noted to be present <M absent, the membrane was subjectively described as thick or thin, and the presence or absence of a splitting or separation in the dividing membrane was recorded. All placentas were examined postnatally to determine chorionicity. Fisher's exact test was performed. RESULTS: In none of the posterior placentas was the membrane insertion vi.suali/ed after 2(i weeks. However, in all of the anterior placentas, the membrane insertion was seen. The split sign was never seen before 28 weeks, and its width, when present, varied from '2 to 14 millimeters. 63% of the dichorionic anterior/lateral placentas had a split-membrane sign whereas none of the monochorionic an terior/lateral placenta had a split membrane (P = .007). The split membrane sign was seen more frequently after :W weeks than between 28-32 weeks (P = .07). CONCLUSION: The split-membrane sign may be a useful determinant ol twin chorioniciiv in the third trimester. After 28 weeks gestation, it appears to ha\e moderate sensitivity (63%) and high positive predictive \alue ( 100(/r). bui it is onlv useful for the evaluation of anterior or anlerolateral placental discs. '.