Pathophysiology

It is unclear as to the exact aetiology of this disease process. Several hypotheses have been proposed, although none have been proven. Previous experimental animal models (usually the injection of amniotic fluid from a different non-human species into animal pulmonary vasculature) have not been very helpful in providing clinical information for pregnant women.^5,6 Two phases of clinical presentation have been suggested. The early phase may manifest with acute right-heart failure and significant pulmonary hypertension within 15 to 20 minutes after the onset of AFES. The later presentation is characterized by acute left-heart failure with pulmonary oedema, but without pulmonary hypertension.

Early researchers speculated that the mechanism was due to mechanical obstruction of pulmonary veins by intra-amniotic components such as fluid, fetal squames, fetal hair, etc., although it was later suggested that it may be a vasospastic disease process as a result of high concentrations of endothelin, proteolytic enzymes, histamine, serotonin, and leukotrienes in the composition of the amniotic fluid and debris.⁷ The clinical similarities with anaphylaxis have led to some experts proposing the alternate name ‘anaphylactoid syndrome of pregnancy’.^8,9 However, amniotic fluid components are commonly found in the maternal circulation of women who do not demonstrate any evidence of AFES symptoms. Amniotic fluid in the maternal circulation can result in vasospastic, proinflammatory, humoral and/or immunologic mechanisms.¹⁰–¹³ These mechanisms probably result initially from a breach in the maternal–fetal interface, most likely at the level of the placenta or the uterus. The proinflammatory responses seen in AFES are similar to those seen with the systemic inflammatory changes associated with sepsis and septic shock.^14,15 Disseminated intravascular coagulopathy, very commonly seen in AFES, is most likely a result of activation of factor VII and platelets, and release of inflammatory and proinflammatory mediators that subsequently activate coagulation.

Clinical Features

AFES can present with a myriad of clinical signs and symptoms, although it most frequently presents precipitously, inexplicably and unpredictably with acute cardiopulmonary compromise and collapse, and haemodynamic instability and acute hypoxic respiratory failure (Table 17.1). This sudden collapse often is seen as a cardiopulmonary arrest either during labour and delivery or in the immediate post-partum period. Less frequently, disseminated intravascular coagulopathy (DIC) and/or refractory seizures can be associated with the cardiopulmonary collapse, but can also be the isolated initial presenting symptoms. The symptoms most often present at the time of labour and delivery but have been reported to occur, less frequently, with amniocentesis, abortions, miscarriages, caesarean deliveries, difficult and long labours, assisted vaginal deliveries and labour induction.⁸ Post-partum presentations have also been reported as late as 24 hours post-partum, although symptoms occur more commonly from 30 minutes to 4 hours post-partum. It is unclear why some women tolerate amniotic fluid introduced into the maternal circulation and others do not, instead experiencing significant clinical symptoms. It has been suggested that this may be dependent on the antigenic load of the amniotic fluid in the maternal circulation.^10,13,16

Cardiovascular, pulmonary, neurologic and haematologic changes are most commonly associated with AFES. Haemodynamic compromise, including arrhythmias, cardiac arrest, pulmonary oedema of both cardiogenic and non-cardiogenic (acute respiratory distress syndrome or ARDS) aetiologies, DIC, and seizures are most commonly seen. Evidence for non-cardiogenic pulmonary oedema has been supported by the finding of high protein concentrations in the oedema fluid of these patients.

The cardiac failure seen in AFES is believed to be secondary to either direct myocardial depression by the antigenic response induced by the amniotic fluid or from myocardial ischaemia.¹⁸ Respiratory symptoms primarily from ventilation-perfusion mismatching can range from mild dyspnoea to significant hypoxia with severe dyspnoea.^8,19 Respiratory failure and acute pulmonary oedema associated with AFES is most likely secondary to acute left heart failure initially, and subsequently to later development of non-cardiogenic pulmonary oedema that often progresses to ARDS. The majority of mortality occurs within 5 hours, related to haemodynamic collapse and hypoxaemia, although this timing may be prolonged by resuscitative efforts.⁸

DIC has been frequently cited as occurring both early and late in the presentation of AFES, developing in as many as 80% of all patients with AFES.^8,9,19 It usually manifests after the severe cardiopulmonary effects are seen, but can present as an isolated event. DIC is often quite dramatic in its appearance, with significant maternal haemorrhage requiring aggressive blood product replacement. DIC is believed to be a result of both procoagulant and coagulant factors in the amniotic fluid. It is unclear if the coagulopathy is primarily a result of a consumptive process with an excess of thromboplastin or results from early-onset massive hyperfibrinolysis with increased levels of urokinase-like plasmin activator, thrombin–antithrombin complexes and plasminogen activator inhibitor-1.²⁰

Neurologic symptoms are relatively rare initial symptoms, although they may present later in the course of the disease. They can manifest as restlessness, anxiety, tingling and paresthesias, decreased level of consciousness, acute confusion, delirium, generalized tonic–clonic seizures, stroke, coma, vegetative state and brain death.^8,21 Unfortunately, for many of the survivors of AFES, neurologic sequelae often related to hypoxaemia and hypotension, are not uncommon, occurring in up to 60%.^3,8,21

Some authors have advocated that there may be a less dramatic and clinically recoverable presentation and course of AFES, with improved morbidity and mortality rates.²² However, no clear data or diagnostic criteria are uniformly accepted for this proposed hypothesis.

It is unclear if certain risk factors are directly related to the presence of AFE, but those listed in Table 17.2 have been most commonly cited in the literature.²³ Some women with AFES have no identifiable risk factors.¹

Diagnosis

The diagnosis of AFES is primarily clinical since there are no specific laboratory or radiographic tests that are pathognomonic or specific to the diagnosis of AFES. The practitioner must have a high index of suspicion in a pregnant woman with a constellation of signs and symptoms as described, presenting during labour or immediately post-partum. There is significant heterogeneity in the diagnosis of AFES. It is essentially a diagnosis of exclusion. Other potential diagnoses of cardiopulmonary failure and DIC must be excluded. The consumptive coagulopathy is not secondary to uterine or vaginal lacerations or uterine atony.

Severe haemodynamic compromise with and without cardiac arrest, respiratory failure, and a fulminant consumptive coagulopathy during the labour and delivery process or in the immediate post-partum period strongly suggest the diagnosis. Other serious clinical events such as acute myocardial infarction, sepsis, pulmonary embolism, fat or air embolism and anaphylaxis must be ruled out. Diagnostic criteria can be confusing since even autopsy results may not confirm the diagnosis. Women with classic symptoms consistent with AFES may have no histologic findings of amniotic debris in the pulmonary vasculature on autopsy. Yet, we now know that healthy women with no evidence of clinical symptoms usually have evidence of amniotic components in the maternal circulation.²⁵ Amniotic debris in the maternal lungs is now no longer considered pathognomonic of AFES as previously believed.

Routine laboratory examinations should include a basal metabolic panel with electrolytes, complete blood count with a differential, coagulation studies and liver function tests (Table 17.3). The most significant laboratory abnormalities are those confirming a diagnosis of DIC, i.e. elevated PT, PTT, decreased fibrinogen levels (often <200 mg/dl), and thrombocytopenia.²⁶ Acute anaemia will be present if there is maternal haemorrhage, usually associated with fulminant DIC. An increased A–a gradient will be present if there is significant hypoxia. B-type natriuretic peptide (BNP) levels are increased in the presence of left-heart failure with cardiogenic pulmonary oedema. The chest X-ray may be normal initially, but later in the course will often show evidence of pulmonary oedema with bilateral alveolar and interstitial infiltrates. Respiratory failure is often seen, presenting as cardiogenic or non-cardiogenic (ARDS) pulmonary oedema. Extreme hypoxaemia is often a cardinal feature and is responsible for the anoxic injury of many patients. The majority of these women will require assisted mechanical ventilation with aggressive measures including sedation, paralysis and prone positioning for severe ARDS and refractory hypoxia. Troponin-I levels are increased in the setting of demand ischaemia or from myocardial ischaemia and infarction. Although none of these tests are specific or sensitive to confirm the diagnosis of the AFES, they can be useful in diagnosing and treating various electrolyte and haematologic/coagulation abnormalities that are associated with critical illness.

Previously, pulmonary artery catheterization was helpful in determining if acute left-heart failure was present. However, diagnostic bedside echocardiography has proven useful in quickly diagnosing acute left-heart failure with significant left ventricular decompensation, although these changes are not specific for patients with AFES. A bedside transthoracic or transoesophageal echocardiogram is not usually performed within the first 30 minutes of presentation, so it is unlikely that evidence of acute right-heart failure with pulmonary hypertension will be seen. More commonly after a period of time, left-heart failure with either segmental wall motion or global left ventricular dysfunction is seen with echocardiography.

To aid in the standardization of the diagnosis of AFES, the Society for Maternal-Fetal Medicine and the Amniotic Fluid Embolism Foundation proposed a clinical definition of AFES using a set of clinical criteria consistent with most cases of AFES, based on four specific diagnostic criteria and symptoms (Table 17.4).^15,27 All of the criteria must be present to be consistent with the diagnosis. These criteria were proposed to provide some semblance of standardization for research in this field, but may be applicable clinically. Many practitioners and experts in the field consider that these criteria reflect a probable diagnosis of AFES. However, expert review of cases identified with AFES using these set criteria found that not all the patients with presumed AFES met the criteria proposed for the AFES diagnosis.²⁸ In addition, other authors have proposed different criteria. Not all of these proposed criteria include normothermia as a condition and others have included cases occurring up to 48 hours post-partum as qualifying for an AFES diagnosis; it is unclear if these patients represent atypical presentations of AFES.^11,29,30

Although there are no specific early or reliable radiographic or laboratory tests that are diagnostic of AFES, several newer inflammatory biomarkers have been associated with AFES. None of these tests have been validated. Decreased complement levels (C3, C3a, C4 and C1 esterase inhibitor (C1-INH)) secondary to increased complement activation has been shown to occur in AFES.^11,32,33 The degree of C esterase inhibitor decrease may also correlate with the severity of the disease and lower levels of C1-INH have been seen in cases of fatal AFES compared to non-fatal AFES.³⁴ There are case reports of administration of C1-INH concentrate with improved outcomes.³⁵ This has been suggested as a possible inciting mechanism in this syndrome, but no further evidence is available. Elevated tryptase levels and increased pulmonary mast cell activity (markers of anaphylaxis) have also been shown to occur in some cases of AFES, but have not been seen in all patients.³³ Increased levels of zinc coproporphyrin (Zn-CPI) and sialyl Tn (STN) have also been implicated in the pathogenesis of the disease process, but similarly to other newer tests, they have not yet been validated in the diagnosis of AFES.³⁶

Neurologic events are not uncommon in AFES, including generalized tonic–clonic seizures, cerebral infarction and hypoxic cerebral encephalopathy. Devastating hypoxic injury that is often associated with AFES early in the pathogenesis of the disease process is responsible for unfortunate outcomes such as refractory and uncontrolled generalized seizures. Neurological effects have been reported as a result of hypoxaemia, hypotension and direct arterial embolization via an intracardiac shunt.²¹

The diagnosis of AFES by autopsy and histologic evidence is consistent with the finding of amniotic debris in the pulmonary vasculature.³⁷–⁴⁰ However, it is important to note that many documented clinical cases of AFES have not shown any histologic abnormalities on autopsy, making the diagnosis problematic. The presence of fetal cells, lanugo and squames in the maternal pulmonary vasculature is no longer considered pathognomonic of the disease, since most pregnant women who die from other causes have similar findings on autopsy. Autopsy specimens should be taken from all the pulmonary lobes and segments to increase the probability of success in viewing the amniotic fluid. Embolic material in the pulmonary vasculature (arterioles and capillaries), has been identified as early as 2 hours within the initial clinical presentation, in women who have died presumptively from AFES.³⁷

Misdiagnosis of this disease is not unusual and any pregnant woman who dies with acute symptoms of cardiorespiratory failure, and DIC and maternal haemorrhage should have an autopsy performed to aid in the histologic diagnosis. The diagnosis has often been confused in those women who have died from acute haemorrhagic shock. Some evidence indicates that up to 40% of patients with maternal haemorrhagic shock showed evidence of amniotic emboli in their lungs on autopsy.⁴¹ Authors have suggested that if there is no histologic evidence of amniotic fluid within three days of exhibiting signs and symptoms consistent with AFES, then most likely the diagnosis of AFES is incorrect.

If the fetus has not been delivered during the period of acute maternal haemodynamic and respiratory compromise, fetal compromise will most likely result from decreased utero-placental perfusion with extreme fetal hypoxia further aggravated by severe maternal acidosis. Absent fetal heart rate variability with late decelerations and terminal bradycardia can be seen.