Objective: To test:he hypothesis that a range of severity of placental
vascular lesions underlies preeclampsia and that the likelihood of it
s clinical diagnosis increases with the extent and severity of uteropl
acental vascular lesions. Methods: Four hundred sixty-five consecutive
placentas of singleton, nonanomalous live-born infants born before 32
weeks' gestation were examined prospectively, and uteroplacental Vasc
ular and related villous lesions were assigned a semiquantitative lesi
on score based on severity and extent of lesions. The summed scores of
individual lesions yielded a total uteroplacental vascular lesion sca
re, ranging from 0 to 21, that was correlated with the odds of a clini
cal diagnosis of preeclampsia, as well as with potential confounders,
including maternal age, race, gestational age at delivery, and birth w
eight centile. Statistical analysis was performed using contingency ta
bles, one-way analysis of variance, multiple logistic regression, and
receiver operating characteristic curve. P < .05 was considered signif
icant. Results: A clinical diagnosis of preeclampsia was present in 78
of 465 (17%) cases. Logistic regression demonstrated that the total u
teroplacental vascular lesion score related significantly to the diagn
osis of preeclampsia (odds ratio 1.43, 95% confidence interval 1.31, 1
,57) and this association was independent of gestational age at delive
ry and birth weight centile. Preeclampsia was diagnosed in 12 of 284 (
4%) cases with no or minimal histologic evidence of placental vascular
injury (total score less than 4). Conversely, the diagnosis was not m
ade in 4% of cases despite the presence of extensive placental vascula
r injury (total score at least 14). Conclusion: The likelihood of clin
ical diagnosis of preeclampsia before 32 weeks increases with progress
ive impairment of the uteroplacental circulation. Histopathologic exam
ination of the placenta can be used to confirm the diagnosis of preecl
ampsia. (C) 1997 by The American College of Obstetricians and Gynecolo
gists.