Amniotic fluid embolism is a serious complication of the peri-partum period that is associated with high maternal mortality rate, ranging from 20 to 40%.

The pathophysiology of amniotic fluid embolism remains poorly known at this time.

The definition and diagnosis of amniotic fluid embolism is only clinical, based on different scores, the most used of which is the score proposed by the United Kingdom Obstetric Surveillance System (UKOSS). This score supports the diagnosis of amniotic fluid embolism in the event of acute maternal collapse associated with one or more of the following features: cardiac rhythm problems, acute fetal compromise, respiratory distress, maternal hemorrhage, coagulopathy, convulsions, premonitory symptoms (restlessness, numbness, tingling, agitation, etc.). ) seizure, in the absence of any other clear cause. Amniotic fluid embolism is a differential diagnosis of maternal collapse.

Determining specific biological markers for this disease would be very useful in order to be able to affirm the diagnosis and refute other diagnostic hypotheses with certainty. The detection of amniotic cells in the blood or bronchoalveolar lavage fluid seems to be of little help. The assay of plasma tryptase does not confirm the diagnosis of amniotic embolism, but is useful for ruling out the diagnosis of anaphylactic shock. The dosage of the complement lacks sensitivity and specificity to be useful in the diagnosis of amniotic fluid embolism. The maternal plasma assay of IGFBP-1 ("Insulin Growth Factor Binding Protein" type 1) has been proposed for the biological diagnosis of amniotic embolism. IGFBP-1 is present in high concentration in amniotic fluid, while its concentration is low in maternal plasma. High levels of IGFBP-1 in maternal blood would therefore make it possible to establish the diagnosis of amniotic fluid embolism with excellent sensitivity and specificity according to previous data collected from 25 patients. However, no study has confirmed this result to date. Other markers have also been suggested (zinc-coproporphyrin in particular), but to date, no specific marker for this disease has been formally identified.

Metabolic phenotyping consists in the identification of subtle and coordinated metabolic variations associated with various pathophysiological stimuli. Different analytical methods, such as nuclear magnetic resonance, allow the simultaneous quantification of a large number of metabolites. Statistical analyses of these spectra thus lead to the discrimination between samples and the identification of a metabolic phenotype corresponding to the effect under study. This approach allows the extraction of candidate biomarkers and the recovery of perturbed metabolic networks, driving to the generation of biochemical hypotheses (pathophysiological mechanisms, diagnostic tests, therapeutic targets,...).. It is thus possible to obtain biochemical characterizations of human biological samples (also called "metabolic profile or signature") which can be compared in order to identify distinctive elements of certain pathophysiological states, establishing a metabolic phenotype of the pathological state studied. This analysis can be supplemented by a study of the proteome (proteomics), in order to identify one or more biological markers associated with a disease.

The aim of this study is to determine the metabolic and proteomic phenotyping in three groups of women: women for whom the diagnosis of amniotic fluid embolism was retained according to the UKOSS clinical criteria (Group AFE), women admitted for prophylactic elective cesarean section (Group Control 1), women presenting acute collapse or shock in the peri- partum for which the diagnosis of amniotic fluid embolism has been excluded (severe hemorrhage, SEPSIS, pulmonary embolism for example; Group Control 2)

Inclusion Criteria:

• Group AFE

  • adult women
  • presenting the diagnostic criteria for amniotic embolism according to the UKOSS initiative,
  • having no known notable medical history (class of the American Society of Anesthesiologists 1)
  • and whose samples were taken within the first 6 hours of collapse and were sent to the fetal-maternal biology laboratory of the Hospices Civils de Lyon (HCL).

• Group Control 1

  • term adult women
  • admitted for scheduled caesarean, term (> 38 weeks), without pathology or associated medical history (class of the American Society of Anesthesiologists 1),
  • biological sample planned before the cesarean section (search for irregular agglutinins).

• Group Control 2:

  • term adult women (> 38 weeks)
  • presenting a collapse (systolic blood pressure <80 mmHg twice and for> 5 minutes) or acute peri-partum shock, for which the diagnosis of amniotic embolism is not retained,
  • no known notable medical history (class of the American Society of Anesthesiologists 1),
  • sampling taken within the first 6 hours of the shock or collapse.

Exclusion Criteria:

• Refusal to participate
• birth as part of an intrauterine fetal death or a therapeutic termination of pregnancy
• preeclampsia, eclampsia
• Gestational Diabetes
• any known endocrine pathology (diabetes, dysthyroidism, adrenal insufficiency, pituitary and hypothalamic diseases).
• history of autoimmune disease
• hemoglobin rate <7 g / dl at the time of sampling (Control group 1 and 2)
• major protected (guardianship, curatorship) or placed under judicial protection

This study will focus on women with clinical criteria for amniotic embolism for whom the samples were sent to the fetal-maternal biology laboratory of Hospices Civils de Lyon (HCL), women with peri-partum collapse for which the diagnosis of amniotic embolism is excluded, and women admitted to term for a scheduled cesarean.