WRITE UPS - CRITICAL ISSUES - Amniotic Fluid Embolism


Amniotic fluid embolism is a serious complication of pregnancy with a high mortality. Despite its rare occurrence, the syndrome of amniotic fluid embolism (AFE) is well known to obstetricians. Given its sudden and dramatic presentation and its often devastating consequences, practitioners given the responsibility of caring for the unfortunate woman with AFE remember the experience in great detail for a long time thereafter. A leading cause of maternal mortality in industrialized countries, AFE remains an enigma. Understanding of its etiology and pathophysiology is incomplete, and the criteria used to make its diagnosis are controversial. Furthermore, despite advances in the care of critically ill patients, no management interventions have been found to improve the survival or long-term outcome of women with AFE. Gregory Locksmith’s recent account of this scary condition gives vital insight in this matter and this article draws vital points from the same with due appreciation.

Meyer first described the syndrome of AFE in 1926.  It became an established clinical entity in 1941 after Steiner and Luschbaugh published a maternal mortality case series that included eight women who had squamous cells and mucin, presumably of fetal origin, within their pulmonary vasculature. The authors postulated that these histologic findings formed the basis of a clinical syndrome characterized by sudden shock and pulmonary edema during labor that ultimately resulted in maternal death.

The hundreds of case reports that have followed Steiner and Lush Baugh’s landmark series have marginally increased understanding of this syndrome. Because AFE is so uncommon, no single institution has sufficient experience to assess risk factors, determine the pathophysiology and clinical course, or evaluate management strategies. Clark and co-workers established a national registry for AFE to assist in understanding they syndrome and published the results in 1995. The strength of the report lies in its relatively large size (46 cases), its stringent entry criteria, and its analysis of 121 different clinical factors. The report confirmed the high mortality rates associated with AFE but challenged several previously held beliefs about the etiology, risk factors, and pathophysiology.


The incidence of AFE has been reported to range from 1 in 8000 to 1 in 80,000 deliveries. It is responsible for 10% of all maternal deaths in the United States. The syndrome typically occurs during labor, soon after vaginal or cesarean delivery, or during second-trimester dilatation and evacuation procedures. In the national registry 70% of the cases occurred during labor, 19% were recorded during cesarean delivery, and 11% occurred after vaginal delivery. All of the cases noted during cesarean section had their onset soon after delivery of the infant.

The analysis of the national registry did not find any maternal demographic risk factors that predisposed gravid as to AFE. Previous cases reports of AFE described the syndrome as commonly occurring after long, hard, and tumultuous labor. These descriptions led many investigators to believe that oxytocin use and antecedent uterine hyperstimulation increased the risk for AFE: however, the registry found that these two factors were no higher among women diagnosed with AFE than in the general population. Although uterine tetany often occurs concomitantly with the initial syndrome of AFE, it is more likely a response to profound tissue hypoxia rather than the cause of entry of amniotic fluid into the maternal blood stream. Pang and Watts recently studied the risk of CS done under spinal anesthesia and suggested that this could be one of the risk factors. They examined the sympathetic blockade as a basis of this risk. However this requires to be carefully validated.

A factor that is consistently related to the occurrence of AFE is a tear in the fetal membranes. Of the women included in the national registry, 78% had ruptured membranes. Two thirds of these women had artificial rupture and one third had spontaneous rupture. In 13% of cases, maternal collapse occurred within 3 minutes of amniotomy or insertion of an intrauterine pressure catheter. Placental abruption was confirmed in another 13% of the women. These findings suggest that certain conditions may permit exposure of fetal tissue to the maternal vasculature and may increase the risk for AFE.


Amniotic fluid embolism classically presents as the sudden onset of dyspnoea and hypotension followed shortly by cardio respiratory arrest. Women may exhibit cyanosis or mental status changes from extreme hypoxemia. Arterial blood gas analysis often demonstrates a dramatic gradient between the alveolar and arterial partial oxygen pressure (PO2).  In the analysis of the national registry, the most common presenting signs were seizure like activity (30%), dyspnoea (27%), fetal bradycardia (17%), and hypotension (13%). The initial episode, if survived, is usually followed by disseminated intravascular coagulation (DIC) that can result in exsanguination. Hypo-perfusion of the heart, lungs, and kidneys may start a vicious cycle culminating in multi-organ system failure. Primary lung injury can lead to adult respiratory distress syndrome (ARDS) and secondary oxygenation deficiency. Permanent brain dysfunction can occur. In any individual patient, the hemodynamic, pulmonary, or hematological disturbances can dominate the presentation or be entirely absent. Kim and colleagues have demonstrated occlusion of branch retinal arterioles in the process of embolism.

The physiologic disturbance responsible for the extreme hypoxia is not well characterized but may be an initial and transient pulmonary vasospasm. An increase in pulmonary vascular resistance has not been well documented in humans. In a goat model, Hankins and co-workers demonstrated an initial and transient rise in pulmonary and systemic vascular resistance along with myocardial depression. These disturbances were especially prominent when the injectate included meconium. Because this initial hemodynamic response is transient, it may resolve before instruments and monitors can be set in place to document it in clinical situations.

Systemic hypotension is the most prominent hemodynamic alteration documented in humans and results principally from severe left sided heart failure. The etiology of the depressed cardiac function is unclear. Proposed mechanisms include a direct depressant effect of amniotic fluid elements on the myocardium and myocardial ischemia resulting from coronary artery vasospasm or global hypoxia. In the report on the national registry, 40 of the patients (87%) sustained cardiac arrest, 12 of whom were successfully resuscitated. Four other women experienced serious dysrhythmias without frank arrest. 

The coagulopathy that is part of the AFE syndrome ranges from minor disturbances in laboratory coagulations studies to severe DIC. Any or all of the following hematologic laboratory abnormalities may be present: elevated fibrin split products of D-dimer products, decreased fibrinogen, thrombocytopenia, and prolonged partial thromboplastin and prothrombin times. The exact incidence of coagulopathy with AFE is unknown, but it is common among those who survive the initial event. On rare occasions, it is the only manifestation present. The incidence of coagulopathy in the analysis of the national registry was 83%.

The etiologic mechanism of the coagulopathy is obscure, In vitro studies and animal models have yield inconclusive results. The mechanism may be similar to that proposed for the coagulopathy related to severe placental abruption. In this model, the maternal immune system recognizes fetal antigens that have gained entrance to the maternal blood stream. The fetal antigens, possibly trophoblastic tissue, exert potent thromboplastin like effects, initiating the extrinsic pathway of the clotting cascade. Amniotic fluid contains pro-coagulants that may be capable of initiating intravascular clotting.

Maternal survival rates associated with AFE range from 20% to 40%. In the AFE registry, the overall survival rate was 39%, with 15% of patients surviving neurologically intact. A total of 28 patients in the finding were that the dismal maternal outcome seemed to be independent of the clinical setting or the treatment rendered. No difference in survival was noted between registries patients diagnosed in tertiary care centers versus rural hospitals. Even with recent advances in critical care, the prognosis is not substantially better than the reported years ago.  Most patients in the registry died of refractory cardiopulmonary arrest. In six patients, life support was withdrawn after the diagnosis of brain death. Other causes of death reported to be associated with AFE are exsanguination, ARDS, and multiple organ failure.

The results of the national registry suggest that cardiac arrest, meconium-stained amniotic fluid, and the presence of a dead fetus before the event may be associated with a relatively poor prognosis. Of the six women who did not sustain cardiac arrest, four (67%) survived and two (33%) maintained baseline neurologic function. Eight registrants had meconium staining noted before the event, two of whom also had a dead fetus. None of these women survived neurologically intact. A trend was also noted between meconium-stained fluid and a shorter time from initial symptoms to cardiac arrest.  Although none of these associations were statistically significant, they suggest that substances in meconium or occurring with a dead fetus may be particularly inclined to initiate a more severe reaction.

In the national registry, 22 to 28 patients (79%) who became symptomatic while the fetus was alive in utero delivered surviving neonates:  half of these fetuses were neurologically intact. Usually, the infant demonstrates a profound respiratory acidosis at birth. Arterial cord blood was available for analysis for 11 neonates in the registry who were delivered after maternal collapse. All samples demonstrated a pH of less than 7.0. The mean pH was 6.79 and the low was 6.4 Interestingly, all of the infants who had cord blood analysis performed survived neurologically intact despite severe acidosis at birth.

The fetal survival rate after maternal cardiac arrest was 68% in the registry, with half of survivors having normal neurologic outcomes. The registry showed a correlation between short cardiac arrest to delivery interval and neurologically intact survival, but the numbers were too small to demonstrate statistical significance.


Animal and laboratory experiments have improved understanding of the pathophysiology of AFE. In most of these investigations, observations have been made of the physiologic effects of amniotic fluid or other fetal substances in animal models. Although the results conflict, most of the series demonstrate adverse hemodynamic effects of this condition. Occasionally, amniotic fluid injection resulted in death of the animal. The results of these studies need to be interpreted with caution because the experimental conditions often varied from true clinical situations. The injections used in some of the studies contained varying degrees of particulate matter. Some injections were filtered, whereas others were particulate-enriched. On occasion, the amniotic fluid source came from a different species from the one being studied. Many of the models were nonpregnant.

Although the results from any animal model experiment can never be extrapolated directly to humans, the results of the animal studies and the apparent association among AFE and abruptio placentae, intrauterine pressure catheter placement, and amniotomy suggest that this syndrome is initiated by maternal exposure to fetal tissue. Although AFE was previously postulated to result from forcible entry of a large volume into the maternal blood stream under pressure, the majority of evidence does not support this contention. The presence of uterine contractions is also not a prerequisite for the occurrence of AFE. AFE has occurred during first-trimester suction curettage abortion, a time when the total volume of amniotic fluid is relatively low. Therefore, the AFE syndrome could result from simple exposure to even small volumes of fetal tissue, could result from simple exposure to even small volumes of fetal tissue, which under the right circumstances, trigger a severe physiologic reaction.

Tissue factor (TF) is known to be present in a high concentration in amniotic fluid. The main question of this study is whether tissue factor pathway inhibitor (TFPI), a natural inhibitor of TF, is present in amniotic fluid. Uszynski M and colleagues recently studied 28 women with normal pregnancy course, and at the first stage of termed labour. Fifteen non-pregnant women were the control group. TF and TFPI were studied by an immuno-enzymatic method (ELISA). The level of TFPI in amniotic fluid was 38.7% of that in blood plasma (16.81+/-5.34ng/ml versus 43.41+/-18.70ng/ml, P<0.001), while the level of TF in amniotic fluid was 44.8 times higher than in blood plasma (9995.93+/-8533.11pg/ml versus 252.66+/-28.84pg/ml, P<0.001). They concluded that (i) amniotic fluid contains TFPI (ii) It is reasonable to assume that the intrusion of amniotic fluid into the blood stream may influence the plasmatic TFPI-TF equilibrium resulting in intravascular blood coagulation. This study can have a far-reaching bearing on the understanding of the pathology of AFE.

Some evidence indicates that fetal tissue transfers to the maternal blood stream frequently in the absence of the clinical symptoms of AFE. Perhaps, AFE is initiated only after an immunologic barrier has been breached by fetal antigens to which the pregnant woman is susceptible, In fact, the pathogenic, hemodynamic, and laboratory manifestation of the AFE syndrome are similar to those of anaphylactic and septic shock. All of these disorders involve the entrance of a foreign substance into the circulation, the release of endogenous mediators, and profound myocardial depression resulting in hypotension and reduced cardiac output. Consumptive coagulopathy, a common consequence of septic shock, is one of the hallmarks of the AFE syndrome. Interestingly, Clark and co-workers noted that 41% of the patients in the AFE national registry had a history of either drug allergy or atopy. One of the conclusions drawn by the investigators in the national registry analysis was that the pathophysiologic events noted in the subjects were more consistent with septic shock and anaphylactic shock than with an embolic process, and it was proposed that the term amniotic fluid embolism be changed to anaphylactic syndrome of pregnancy.

Despite the similarities between these disorders, important differences exist. Fever, which is a hallmark of sepsis, is not a typical manifestation of AFE. An urticarial rash often develops in patients sustaining anaphylactic reactions, whereas cutaneous manifestations are not part of the AFE syndrome. Furthermore, bronchospasm and upper airway swelling are hallmarks of most anaphylactic reactions. Although five patients in the national registry were noted to be difficult to ventilate and although wheezes were auscultator in another patient, obstructive ventilatory defects are not considered to be part of the AFE syndrome. An anaphylactic reaction generally results from re- exposures to an antigen to which an individual has previously been sensitized, whereas AFE occurs in primigravidae as well as well as in multi-gravidas.

An important step in learning more about the pathogenesis of AFE is determining the roles of various cellular mediators. The release of histamine, bradykinins, cytokines, prostaglandins (PG), Leukotrines, thromboxane, and other mediators may trigger the hemodynamic and hematologic alterations. Clark has postulated that arachidonic acid metabolites may be responsible for the physiologic disturbances seen in humans during AFE. Azegami and colleagues injected rabbits with amniotic fluid containing leukotriene activity and found that the animals often died, whereas rabbits pretreated with an inhibitor of leukotriene synthesis survived. Work of Kitz, Miller and Lucas suggest that PGF2 which is detectable in amniotic fluid only during labor, has a prominent pathogenic role. They reproduced clinical manifestations of AFE in cats after systemic injection of PGF and amniotic fluid from women in labour, whereas the same effects were not noted when the animals were injected with fluid from women who were not noted when the animals were injected with fluid from women who were not in labor.

Benson and colleagues recently evaluated the potential role of immunologic mechanisms that involve mast cell degranulation (anaphylaxis) or complement activation in the mechanism of amniotic fluid embolism.  They found that serologic findings suggest a role for complement activation in the mechanism of amniotic fluid embolism. Laboratory data from this series did not implicate mast cell degranulation (anaphylaxis) in the pathophysiology of the disease.

Endothelia, which are associated with many different pathophysiologic processes, may also be responsible for the hemodynamic alterations seen in AFE. Endothelin is a potent constrictor of human coronary and pulmonary arteries and of human bronchi. Increased endothelia levels were noted in experiments performed on rabbits and human endothelial cell cultures after the administration of human amniotic fluid. This increase was most prominent with meconium-stained fluid and was not elicited in subjects injected with saline.


The finding of fetal squames in the maternal pulmonary circulation, once considered pathognomonic, is neither specific nor sensitive for the diagnosis of AFE. Squamous cells have been retrieved from the pulmonary vasculature of patients being monitored for conditions other than AFE. Moreover, the differentiation between maternal and fetal elements retrieved from the maternal circulation is problematic. Various stains such as Alcian blue, Atwood, Giemsa, Wright, oil red O, and Sudan black may provide greater sensitivity for detecting fetal debris than routine hematoxylin- eosin staining, however, data concerning, the accuracy of any of these methods are lacking. Of the 22 patients in the national registry on whom autopsy was performed, 16 (73%) demonstrated cellular debris of presumed fetal origin in the pulmonary vasculature. Histologic examination of pulmonary artery catheter blood was performed in eight patients, of whom four (50%) had fetal elements identified.

A Japanese group has identified fetal mucin in the maternal pulmonary vasculature using the monoclonal antibody TKH-2. This technique may improve the ability to differentiate between fetal and maternal elements; however, its accuracy has not been determined. One prospective autopsy study demonstrated that four of four patients with clinically diagnosed AFE had positive TKH-2 immunoassaying of pulmonary vasculature sections. Additionally, TKH-2 staining was negative in four of four women who died of other causes during labour. Diagnosing AFE with TKH-2 immunoassaying in living patients is possible using blood from the maternal pulmonary circulation, however the results from such specialized tests are unlikely to be available during the acute episode or useful for clinical management.

In all circumstances, the diagnosis of AFE should be made on the basis of clinical presentation and laboratory findings. Patients must fulfill four clinical criteria based on physiologic signs, laboratory values, and the clinical setting in the absence of any other explanation for the manifestations observed. These clinical criteria are as follows:

Acute hypotension or cardiac arrest

Acute hypoxia


Absence of other explanations for the clinical manifestations observed.

Onset during labour or within 30 minutes of delivery or surgical abortion.

Other diagnoses to consider in a patient presenting with signs and symptoms of AFE include hemorrhagic shock, placental abruptions, sepsis, pulmonary embolism, aspiration of gastric contents, and eclampsia. Less common conditions include anaphylaxis, toxic reactions to anesthetic agents, myocardial infarction, air embolism, and cerebral hemorrhage.


No form of therapy has been found to improve outcome consistently in women with AFE: therefore care is supportive. The initial management objective is to maintain adequate oxygenation and vital organ perfusion. Cardiopulmonary resuscitation is necessary for patients with cardiac arrest. Oxygen should be provided at a high concentration (100%). Patients who are unconscious require endotracheal intubation. Volume replacement with isotonic crystalloid solution is a first line therapy for maintaining blood pressure.

Patients who survive the initial insult are at high risk for heart failure, ARDS, and DIC. Patients who demonstrate persistent hypotension despite adequate volume expansion should be treated with inotropic agents such as dopamine or dobutamine. A rule or thumb for supporting critically ill obstetric patients is to maintain systolic blood pressure above 90mm Hg, the PaO2 above 60 mm Hg, the arterial oxygen saturation above 90%, and the urine output greater than 25 ml./hour. After the correction of hypotension, fluid therapy needs to be guided gently to achieve the optimal balance between maintaining blood pressure and preventing pulmonary edema and ARDS. Information derived from a pulmonary artery catheter usually is valuable in guiding hemodynamic management.

The therapy for coagulopathy causing active bleeding is replacement of blood components. Platelets, fresh frozen plasma, cryoprecipitate, and packed red blood cells may all be necessary to correct deficiencies. Occasionally, uterine atony develops and results in hemorrhage. Intravenous oxytocin and intramuscular PGF are the preferred therapeutic agents if this condition arises. Kaneko and colleagues have used hemodiafiltration for these cases getting good results.

The fetal condition often deteriorates in women in whom AFE develops during pregnancy. After 24 weeks gestation, continuous and watchful monitoring of fetal heart rate is essential. Profound fetal academia often develops with maternal collapse, and early delivery may improve the neonatal prognosis. The performance of cesarean delivery in an unstable patient is fraught with pitfalls therefore management must be individualized. Although the primary responsibility of the obstetricians is to ensure the health and life of the mother, intervention on behalf of the fetus is appropriate in certain instances. 

If the patient with AFE sustains cardiac arrest, her chance of neurologically intact survival is poor. Delivery may actually improve the likelihood of success of the resuscitation. Relieving uterine compression of the inferior vena cava may improve cardiac output by increasing preload. Although never demonstrated in practice, the theoretical maternal benefits and probable neonatal benefits of immediate delivery seem  to outweigh the risks.  Therefore, peri-mortem cesarean delivery is recommended in pregnant women with AFE and cardiac arrest.

Other therapeutic options have been proposed. Given the pathophysiologic similarity with anaphylaxis, immunosuppression and sympathetic nervous system augmentation have been proposed as a therapy for AFE. This therapeutic approach has not been tested, but given the likely catastrophic outcome, it seems reasonable to try. Clerk and co-workers recommend giving 500 mg of hydrocortisone sodium succinct intravenously every 6 hours until improvement of the patient or death occurs. Epinephrine may also be given. Because many women with AFE have cardiac arrest, they are likely to have received epinephrine previously during the resuscitation.

Recently Hsieh Y and colleagues used extra corporeal membrane oxygenation and intra-aortic balloon counter pulsation as life-saving therapy for a patient with amniotic fluid embolism. They are very happy with the results and enthusiastically suggest that extra corporeal membrane oxygenation and intra-aortic balloon counter pulsation should be considered to save the life of a patient with amniotic fluid embolism and left ventricular failure unresponsive to medical therapy.


Fortunately, AFE is rare. Because it occurs so infrequently and because studies in animal models cannot reproduce accurately the physiologic and clinical alterations noted in humans, its pathogenesis remains an enigma. Based on clinical observations, AFE is similar in presentation to septic and anaphylactic shock. From these similarities, a theory has been proposed that the clinical syndrome of AFE results when fetal antigens breach a maternal immunologic barrier in susceptible mothers. Maternal immunologic recognition subsequently triggers the release of endogenous mediators that are responsible for dramatic physiologic disturbances.

Most cases of AFE are associated with dismal maternal outcomes and poor fetal outcomes regardless of the quality of care rendered. Improved understanding of the molecular pathophysiology of AFE may lead to the development of preventive measures and more effective and specific treatment. In the meantime, its occurrence remains unpredictable and unpreventable and its effects, for the most part, untreatable.


      1.  Atrash HK, Koonin LM, Lawson HW, et al: Maternal   mortality in the United States 1979-1986. Obstet    Gynecol 76-1055, 1990.

2.  Azegami M, Mori N: Amniotic fluid embolism and leukotrienes. Am J Obstet Gynecol 155: 119, 1986.

3.  Benson MD, Kobayashi H, Silver R K, Oi H, Greenberger PA, Terao T: Immunologic studies in presumed amniotic fluid embolism: Obstet Gynecol 2001 Apr; 97(4): 510-4

4.  Clark SL: Arachidonic acid metabolites and the pathophysiology of amniotic fluid embolism. Semin. Reprod Endocrinol 3:253, 1985.

5.  Clark SL: Pavlov A Z, Green spoon J, et al: Squamous cells in the maternal pulmonary circulation. Am J. Obstet Gynecol 154:104, 1986.

6.  Clark SL: New concepts of amniotic fluid embolism: A review. Obstet Gynecol Surv-45: 360, 1990.

7.  Clark SL, Hankins GDV, Dudley DA, et al: Amniotic fluid embolism: Analysis of the national registry. Am J. Obstet Gynecol 172:1158, 1995.

8.  Courtney LD: Coagulation failure in pregnancy: BMJ 1: 691, 1970.

9.  EI Maradny E, Kkanayama N, Halim A, et al: Endothelin has a role in early pathogenesis of amniotic fluid embolism. Gynecol Obstet Invest 40:14, 1995.

10. Gregory L: Amniotic Fluid Embolism: Obstet Gyn. Clinics of America

11. Hankins GDV, Snyder RR, Clark SL, et al: Acute hemodynamic and respiratory effects of amniotic fluid embolism in the pregnant goat model. Am J. Obstet Gynecol 168: 1113, 1993.

12. Hsieh Y, Chang CC, Li PC, Tsai HD, Tsai CH: Successful application of extra corporeal membrane oxygenation and intra-aortic balloon counter pulsation as lifesaving therapy for a patient with amniotic fluid embolism: Am J Obstet Gynecol 2000 Aug; 183(2): 496-7

13. Kaneko Y, Ogihara T, Tajima H, Mochimaru F: Continuous hemodiafiltration for disseminated intravascular coagulation and shock due to amniotic fluid embolism: report of a dramatic response. Intern Med 2001 Sep; 40(9):945-7

14. Kobayashi H, Ohi H, Terao T: A simple, noninvasive, sensitive method for diagnosis of amniotic fluid   embolism by monoclonal antibody TKH-2 that recognizes Neu Ac alpha 2-6 Gal Nac. Am J Obstet Gynecol 168: 848, 1993.

15. Kobayashi H, Ooi H, Hayakawa H, et al: Histological diagnosis of amniotic fluid embolism by monoclonal antibody TKH-2 that recognizes Neu AC alpha 2-6 Gal Nac epitopes. Hum Pathol 28:428, 1997.

16. Kim IT, Choi J B: Occlusions of branch retinal arterioles following amniotic fluid embolism: Ophthalmologica 2000; 214(4): 305-8

17. Kitz miller J L, Lucas WE: Studies on a model of amniotic fluid embolism Obstet Gynecol 39-626, 1972.

18. Meyer JR: Emboli a pulmonary amnio caseosa. Bras Medico 2:301, 1926.

19. Morgan M: Amniotic fluid embolism. Anaesthesia 34:20, 1979

20. Pang and Watts R L: Amniotic fluid embolism during Caesarean section under spinal anaesthesia. Is sympathetic blockade a risk factor?: Aust N Z J Obstet Gynecol Suppl. 2001 Aug; 41(3):342-3.

21. Richards DS, Carter LS, Corke B, et al: The effect of human amniotic fluid on the isolated per fused rat heart. Am J Obstet Gynecol 158:210, 1988

22. Steiner PE, Luschbaugh CC: Maternal pulmonary embolus by amniotic fluid. JAMA 117:1245, 1941

23. Uszynski M, Zekanowska E, Uszynski W, Kuczynski J: Tissue factor (TF) and tissue factor pathway inhibitor (TFPI) in amniotic fluid and blood plasma: implications for the mechanism of amniotic fluid embolism: Eur J Obstet Gynecol Reprod Biol 2001 Apr; 95(2): 163-6



Click here to view other articles in this section