Documentation of amniotic fluid embolism via lung histopathology: Fact or fiction?

OBJECTIVE: To evaluate pulmonary histopathology for confirming amniotic fluid embolism. STUDY DESIGN: The Capra hircus (goat) model with fresh, homologous amniotic fluid was used. Raw fluid (n = 8), fluid filtered through a 5-μm filter (n=14) and meconium-stained fluid with 1-7% solid debris (n=7) were injected. Three hours after embolization the animals were euthanized and specimens collected. Three to five areas of lung were sampled based on the most abnormal areas visually. Traditional and special stains were utilized. The study protocol was approved by the institutional review board and animal use and care committee. Statistical analysis was by X² with Yates correction. Significance was defined as P < .05. RESULTS: Amniotic fluid debris (fetal squames, mucin or foreign pigments) was found in 10 of 29 animals (34.5%). Debris was found in 7/7 (100%) of the meconium group, 2/8 (25%) of the raw fluid group and 1/14 of the filtered group (7%). The likelihood of finding debris in amniotic fluid embolism with meconium-stained fiuid was greater than with raw (P < .017) or filtered amniotic fiuid (P < .001). CONCLUSION: In this animal model, histopathologic confirmation of amniotic fluid embolism was an unreliable marker of the event except in cases of amniotic fluid embolism involving meconium-stained fluid.