Fetoplacental histology as a predictor of karyotype: a controlled study of spontaneous first trimester abortions.

It has been suggested that inferences about fetal karyotype can be made from examination of placental and decidual histology in early, spontaneous abortions (SABs). We assessed the reproducibility and predictive value of histologic features in 75 karyotyped, first trimester SABs; 32% (24 of 75) had normal male karyotypes (46,XY) and 68% (51 of 75) were cytogenetically abnormal (29 trisomy, 12 triploidy, eight monosomy X, and two tetraploidy). Three pathologists independently assessed 17 fetal, placental, and decidual histological findings and made predictions about the karyotype (normal, abnormal, or uncertain). Good to excellent interobserver and intraobserver reproducibility (kappa > 0.58) was achieved for the identification of five histological features: villous cavitation, anucleate fetal erythrocytes, amnion, umbilical cord, and fetal tissue. When histology and karyotype were compared using Fisher's exact test, no histological feature was associated with "any abnormal karyotype," two features (anucleate, fetal erythrocytes and umbilical cord) were associated with a normal karyotype, two features (villous dysmorphism and cisterns) were associated with triploidy, and four features (villous hydrops, no umbilical cord, no fetal tissue, and no anucleate erythrocytes) were associated with trisomy. Despite these significant histological-cytogenetic associations, the positive predictive values of each of these histological features with their corresponding karyotypes were low, ranging from 0.41 to 0.73 (mean, 0.53). Our data suggest that certain histological features in first trimester SABs are associated with the SAB's karyotype and are reproducible; however, such histological features did not perform as well as diagnostic tests for predicting the likelihood of normal versus abnormal karyotype.