382

Amniotic fluid : a report of Thomas C. Mainprize MD, J. Roger Mattby MB B CHIR FFARCSFRCPC four probable cases

Four probable cases of embolism (AFE) Amniotic fluid embolism (AFE) is a dramatic and are reviewed. The outcome appeared to be determined by frequently fatal complication of and is a the severity of the insult, and possibly the gestation of the cause of which remains unpredict- pregnancy, rather than the management of the AFE. Two able and largely unpreventable, t The diagnosis is cases occurred during early labour; neither patient confirmed by the histological identification of fetal recovered consciousness. One died two weeks later and squames in the lungs. However, a presumptive the other suffered severe permanent cerebral damage. clinical diagnosis may be made when a patient The other two cases occurred during dilatation and collapses suddenly, usually after rupture of the curettage, one for therapeutic at fourteen weeks membranes, during violent uterine contractions. gestation and the other for missed abortion at twenty The collapse may be accompanied by a convulsion weeks gestation. Both patients made full recoveries. or muscular twitching, usually with dyspnoea and Disseminated intravascular coagulation (DIC) was a cyanosis. Coagulation defects may be associated feature of all four cases. In the patients in labour it with postpartum haemorrhage and bleeding from occurred almost immediately. In those undergoing dila- other sites. 2 tation and curettage it occurred after the patients had Not all cases are typical. The condition may apparently recovered but were under observation in the occur during apparently normal labour and also intensive care unit. during second trimester dilatation and curettage. Amniotic fluid embolism can occur during an appar- During the years 1974-84 four cases occurred in the ently uneventful labour, It should also be suspected when Foothills Hospital, Calgary. These demonstrate the unexplained collapse occurs d,tring second trimester diversity of both the presentation and outcome of dilatation and curettage. Because severe DIC may AFE. follow, such patients should be transferred immediately to a centre with full haemotology services. Case reports

Case 1 A thirty-year-old female, G4 P3, weighing 78 kg, was admitted to hospital at 37 weeks gestation in early labour following spontaneous rupture of the membranes. An oxytocin infusion was started at a rate of four Key words milli-units per minute to augment mild uterine COMPLICATIONS:amniotic fluid embolism; ANAES- contractions. Fifteen minutes later, without any THESIA: obstetric. warning, and with still only mild contractions, the patient had a grand mal convulsion, became cya- From the Department of and Gynaecology and notic and, within two minutes, pulseless. Cardio- The Department of Anaesthesia, Foothills Hospital at pulmonary resuscitation was immediately insti- the University of Calgary, Calgary, Alberta. tuted. Five minutes late a was Address correspondence to: Dr. J. Roger Maltby, De- commenced while external cardiac massage was partment of Anaesthesia, Foothills Hospital at the continued. A live female infant weighing 2,600 gm University of Calgary, 1403-29th Street NW, Calgary, was delivered with Apgar scores of 1 and 7. Alberta, Canada T2N 2T9. Within one hour of the initial collapse bleeding

CAN ANAESTH SOC J 1986 / 33:3 / pp382-7 Mainprize and Maltby: AMNIOTIC FLUID EMBOLISM 383 appeared from the incision and intravenous sites. The patient was admitted to the ICU for further Coagulation tests confirmed disseminated intravas- observation. cular coagulation (DIC) which was successfully Within two hours of the initial collapse vaginal treated. bleeding increased. Laboratory tests confirmed Mechanical ventilation was continued postopera- DIC. This was successfully treated and the patient tively but there was no recovery of cerebral func- made a full recovery. tion. After seven days the EEG showed no brain activity and life support was discontinued. Case 4 A thirty-one-year old female, G3 P2, weighing Case 2 60 kg, was admitted to hospital two weeks after An eighteen-year-old female, G1 P0, weighing intrauterine death had occurred at 20 weeks gesta- 67kg, was admitted at 34 weeks gestation in tion. The uterus was 12-14 weeks size. Laboratory premature labour with ruptured membranes. tests for coagulation were normal on two occasions At 1930 hrs an isoxuprine infusion (400 mg-L -~) and she was scheduled for dilatation and curettage. was commenced at a rate of 9.6 mg.hr-t to inhibit No premedication was given. Anaesthesia was uterine contractions. Ten minutes later she com- induced with thiopentone and maintained with plained of nausea and almost immediately had a nitrous oxide 50 per cent in oxygen. A considerable grand mal convulsion followed by cardiac arrest. amount of necrotic tissue was removed and, be- Cardiopulmonary resuscitation was started immedi- cause the uterus was bulky, ergometrine 0.25 mg ately and was successful. However, she did not was given intravenously. About one minute later recover consciousness and was transferred to the the patient coughed, became deeply cyanotic, with intensive care unit (ICU). marked congestion of the neck and face and stopped At 2110 hrs she vomited blood and began breathing. Her pulse was barely palpable, blood bleeding from intravenous puncture sites, vagina pressure was unrecordable and the EKG showed a and urinary tract. Laboratory tests confirmed DIC tachycardia with marked ST depression. Her pupils which was successfully treated. A premature infant, were fixed and dilated. An endotracheal tube was delivered vaginally with forceps, died within one immediately passed and she was ventilated with 100 hour. per cent oxygen. She did not recover consciousness and required As in case 3 the diagnosis was not immediately ventilation for the next two weeks. After this she apparent. Her condition did not deteriorate further was weaned from the ventilator and was transferred and within twenty minutes her colour improved, her to her home town for long-term care. pupils began to react and the EKG returned to normal. Fifteen minutes later she started coughing on the endotracheal tube. Soon she was awake and Case 3 breathing spontaneously. The endotracheal tube A healthy eighteen-year-old female, G2 P1, weigh- was removed and she was admitted to ICU. DIC ing 59 kg, was admitted as a day-care patient at 14 supervened but this was successfully treated and she weeks gestation for therapeutic abortion. Anaesthe- made a full recovery. sia was induced using thiopentone 250 mg and The clinical characteristics of each case are continued with fentanyl 100 p.g and nitrous oxide shown in Table I and the coagulation defects in 75 per cent in oxygen at a flow rate of eight litres per Table II. minute. During the dilatation and curettage the patient suddenly became cyanotic but continued to breathe normally with good air entry and clear lung Discussion fields. Nitrous oxide was discontinued and 100 per The clinical characteristics of the four cases in this cent oxygen given. Blood pressure became uurecor- series illustrate different presentations of amniotic dable although she still had a weak palpable pulse. fluid embolism. In neither of the two patients near Cardiac arrest did not occur. Further treatment was term was the labour tumultuous. Apart from nausea being considered when her colour improved, the in one of the patients, the onset was dramatic with a blood pressure rose, and consciousness returned. grand mal convulsion rapidly followed by cardiac 384 CANADIAN ANAESTHETISTS' SOCIETY JOURNAL

TABLE I Clinical characteristicsof cases

Weeks Pre-existing Time of Presenting syrup- Onset Case Age gestation disease presentation tams andsigns of DIC Treatment Outcome

1 30 37 Remote atrial Early labour Grandreal seizure, < I hr O2, CPR Mother: death septal defect cardio-vascular Caesarean section after 6 days; repair; mild collapse, coma PRC 29 units live infant painless bleeding FFP 35 units 33 weeks

2 18 34 Threatened Premature Nausea,grand mal < 1 br Diazepam, 02, Mother:permanent abortion labour, on seizure, intubation, CPR cerebraldamage; 10 weeks isoxuprine CV collapse Vaginal delivery babydied within drip coma in ICU 1 hr

3 18 14 Nil During Cyanosis, < 2 hr Oz Recovery from D & C for CV collapse, PRC 8 units collapse30 rains; therapeutic lungsclear FFP 6 units recovery from abortion DIC 18 hrs

4 31 22 Missed abortion During Cough, < 1 hr O~, intubation, Recoveryfrom 20 weeks D & C CV collapse, IPPV 30 mins collapse45 mins; for missed cyanosis, PRC 3 units recoveryfrom abortion lungs clear FFP 4 units DIC 24 hrs Platelets 6 units Cryoppt. 10 units CV = cardiovascular. CPR = cardio-pulmonary resuscitation. PRC = packed red blood cells. FFP = fresh frozen plasma. D & C = Dilatation & curettage. arrest. Immediate intensive resuscitation did not AFE are: respiratory failure and cyanosis (51 per prevent permanent brain damage. Severe coagulo- cent), cardiovascular collapse (27 per cent), con- pathy occurred within one hour and was successfully vulsions (10 per cent) and haemorrhage (12 per managed with fresh frozen plasma and packed red cent). 3 Of patients surviving the initial insult for cells. more than one hour, 40 per cent will develop DIC. 4 In the two patients undergoing dilatation and Any condition which may cause the acute onset of curettage, cyanosis and profound oc- these symptoms must be considered. The differen- curred without warning. One patient continued to tial diagnosis 3'5 therefore includes abruptio placen- breathe normally, the other became apnoeic. The tae, ruptured or inverted uterus; pulmonary critical condition continued, despite 100 per cent thromboembolism; cerebral insults including oxygen, for about 20 minutes before rapid sponta- , epilepsy and ruptured intracranial neous improvement occurred. Profound cardiovas- aneurysm; idiosyncratic or anaphylactic drug reac- cular collapse and cyanosis, not proceeding to tions; and misconnected gas lines. cardiac arrest, were the only features. The aetiology By a process of exclusion a diagnosis of probable was not immediately obvious. In one case the AFE can be made with reasonable confidence on possibility of crossed gas lines was considered and clinical grounds alone. 3'5'7 Even in patients with the oxygen supply changed from pipeline to the confirmed AFE the investigations of chest x-ray, cylinder on the anaesthetic machine. The intrave- EKG and pulmonary angiography are non-specific. nous fluid, which might have been contaminated, They have been reported as abnormal by some was also changed. Both patients started improving authors s-ll and normal by others.7" 2,~3 Diagnosis before a firm diagnosis was made. Each was is usually confirmed at autopsy,t4'i5 although there admitted to the 1CU for observation with a provi- are now several reports of confirmed AFE in sional diagnosis of AFE. patients who survived. Confirmation is obtained by The presenting features in patients who develop histological demonstration of fetal squames (Nile Mainprize and Maltby: AMNIOTIC FLUID EMBOLISM 385

TABLE lI Coagulationfunction tests - maximally abnormal values

Time from PT (sec) PTT (see) Fibrinogen (g'L- ~) FDP (rag.L- ~) Case onset(hrs) (11-13) (36-55) (1.5~1.5) (less than 40)

1 5 19. I 200 0.29 640-1280 2 2 50 421 0.15 160 3 6 no clot no clot 0.59" 320-640* 4 1 50 200 none 1280- 2560

PT = prothrombin time. PTT = partial thromboplastin time. FDP = fibrin degradation products. Normal values at Foothills Hospital, Calgary in parentheses ( ). *Values six hours after treatment with fresh frozen plasma had been started. blue 3 or Papanicolaou stainl6), keratin (Ayoub- For AFE to occur, the chorioamniotic membrane Shklar stain 1T), fat (oil red 0 stain~7), lanugo and/or must rupture, exposing amniotic fluid to a venous hair in blood collected from the pulmonary artery channel either at the endocervical veins, beneath the via a Swan-Ganz catheter, ~2'17'18 pulmonary an- placenta or at the site of uterine trauma. 2s This may giography, 1~ or from sputum. 9'19 Haematoxylin occur with normal labour, abruptio placentae, or and eosin stain is reported as unreliable. 2~ Garland any uterine manipulations such as Caesarean sec- and Thompson zj believe their antiserum to human tion, manual removal of placenta, artificial rupture keratin is more sensitive in diagnosing amniotic of membranes or insertion of an intrauterine pres- fluid debris. sure catheter. 3 Abruption of the placenta due to fetal In the cases reported here the diagnosis was only cord entanglement and a tear in the membranes confirmed histologically in the patient who died. At producing fatal AFE is reported by Corridan etal. 29 autopsy fetal squames were present in her lungs. However, the volume of amniotic fluid required to Mayer is usually credited with the first reported cause maternal symptoms and/or death, is not case of amniotic fluid embolism, published in known. Plauch63~ claims to have identified am- 1926. 22 However, Attwood has recently sug- niotic fluid debris in two patients who showed no gested 23 that a postmortem description by Baillie in clinical or haemodynamic changes of AFE, but 1825 may represent the first recorded case. Sparr and Pritchard31 were unable to demonstrate The reported incidence varies from one in that amniotic fluid escapes into the maternal circula- 8,00024 to one in 80,0002~ live births. Four pro- tion during labour in 28 patients. If Plauch6 is bable cases occurred during 34,366 deliveries at the correct the possibility of anaphylaxis 3 as an aetio- Foothills Hospital, giving an incidence of one in logic agent by virtue of prior sensitization is feasible 8600 live births. Although three of these four although this remains an unconfirmed hypothesis. women survived, a review of the English literature Once in the maternal circulation, amniotic fluid is up to 1979 showed a survival rate of only 14 per transported to the lungs and causes obstruction of cent. 3 the vasculature. This obstruction is accompanied by In Morgan's review 3 of 272 cases, 90 per cent of pulmonary hypertension, ventilation-perfusion in- cases occurred during labour; 28 per cent in equality and cardiac failure. 3,12,32 The release of a tumultuous labour and 22 per cent in oxytocin Factor X activator3 may worsen the obstructive stimulated labour. A small number complicated changes. The subsequent activation of lysis factors intrauterine deaths. The mean age of occurrence completes the vicious cycle of DIC and hae- was 32 years with a range of 15 to 48 years. The morrhage. association with intrauterine contraceptive devices Although AFE is a rare event it carries a remains unclear. 2 AFE are also reported after high mortality and case reports continue to appear second trimester intra-amniotic installations, 8.j4,26 in the literature, z1'27'29'33 The maternal mortality where the increased intrauterine pressure may play rate from sepsis and haemorrhage has declined a role, 3 and after dilatation and curettage. 5'6'9'27 and recent reports of survivals from AFE sug- 386 CANADIAN ANAESTHETISTS' SOCIETY JOURNAL gest there may be a similar trend in this con- References dition.S, 10, l l. 17,32,34.35 Even if this is so, the reason 1 Shnider SM, Moya F. Amniotic fluid embolism. is not clear, although it could be due to a greater Anesthesiology 1961; 22: 108-19. awareness of APE with better management of 2 Department of Health and Social Security. Report complications. on confidential enquiries into maternal deaths in AFE complications vary. Haematological England and Wales 1976-78; 96-100. London: Her changes of DIC occur in most patients although Majesty's Stationery Office 1982. bleeding may not be observed. Twenty-four per 3 Morgan M. Amniotic fluid embolism. Anaesthesia cent of patients develop pulmonary oedema, pro- 1979; 34: 20-32. bably related to excessive fluid administration 4 Anderson DG. Arrmiotic fluid embolism. A re- during resuscitation. Acute renal failure secondary evaluation. Am J Obstet Gynecol 1967; 98: 336-48. to prolonged hypotension, and cerebral damage 5 Cromey MG, Taylor PJ, Cummings DC. Probable from anoxia are uncommon. The overall death rate amniotic fluid embolism after first trimester preg- is 86 per cent and, of these, 24 per cent die in the nancy termination - a case report. J Reproductive first hour) Causes of death include haemorrhage Med 1983; 28: 209-11. from DIC, circulatory collapse, and right heart 6 Lees DE, Shin Y. MacNamara TE. Probable failure with pulmonary hypertension, secondary to amniotic fluid embolism during curettage for a missed massive cardiac tamponade, lZ absorption: a case report. Anesth Analg 1977; 56: Since there is no specific therapy for AFE general 739-42. supportive measures are imperative. 3 These include 7 Lumley J, Owen R, Morgan M. Amniotic fluid cardiopulmonary resuscitation, correction of the embolism. A report of three cases. Anaesthesia bleeding diathesis, and blood volume replacement. 1979; 34: 33-6. Central venous pressure or pulmonary artery pres- 8 Ballas S, Lessing JB, Michowitz M. Amniotic fluid sure monitoring are advisable. 36 The value of embolism and disseminated intravascular coagula- digoxin, diuretics, vasodilators, dextran, cortico- tion complicating hypertonic saline-induced absorp- steroids, and antiprostaglandins remains un- tion. Postgrad Med J 1983; 59: 127-9. clear. 3'36 It is possible that the outcome, as in our 9 Stromme WB, Fromke VL. Amniotic fluid embo- cases, is determined more by the severity of the lism and disseminated intravaseular coagulation after initial insult or the gestational age at occurrence evacuation of missed abortion. Obstet Gynec 1978; than the subsequent management. 52: 76s-80s. The cases described demonstrate three important 10 Dolyniuk M, Orfei E, Vania H, Karlman R, Tomich points about amniotic fluid embolism: (1) the P. Rapid diagnosis of amniotic fluid embolism. Ob- condition may occur when contractions are mild stet Gynecol 1983; 61: 28s-30s. rather than during tumultuous labour, (2) it may 11 Taenaka N, Shirnada Y, Kawai M, Yoshiya 1, Kosaki occur during second trimester dilatation and curet- G. Survival from DIC following amniotic fluid em- tage and, (3) despite apparent recovery, patients bolism. Anaesthesia 1981; 36: 389-93. may later develop DIC. When AFE is suspected in a 12 Schaerf RH, de Campo T, Civetta JM. Hemo- smaller hospital the patient should be transferred dynamic alterations and rapid diagnosis in a case of immediately to a centre with full haematology amniotic fluid embolus. Anesthesiology 1977; 46: services without waiting to see if DIC is going to 155-7. Occur. 13 Sterner S, Campbell B, Davies S. Amniotic flu~:l embolism. Ann Emerg Med 1984; 13: 343-5. 14 Grimes DA. Cares W Jr. Fatal amniotic fluid embo- Acknowledgements lism during induced abortion, 1972-1975. South The authors thank those members of the Depart- Med J 1977; 30: 1325-6. ments of Anaesthesia, Obstetrics and Gynaecology, 15 AttwoodHD. Amniotic fluid embolism. Pathol Intensive Care and Haematology who treated these Annu 1972; 7: 145-72. patients; Dr. George Carson for helpful criticism of 16 Masson RG, Rugieri J, Siddiqui MM. Amniotic the manuscript; and Mrs. Caroline Harris for fluid embolism: definitive diagnosis in a survivor. secretarial assistance. Am Rev Respir Dis 1979; 120: 187-92. Malnprize and Maltby: AMNIOTIC FLUID EMBOLISM 387

17 DuffP, EngelsgjerdB, Zingery LW, HuffRW, Mon- 33 Barrows JJ. A documented case of amniotic fluid tiel MM. Hemodynamic observations in a patient embolism presenting as acute . Am J with intrapartum fluid embolism. Am J Obstet Gyne- Obstet Gynecol 1982; 143: 599-600. col 1983; 146: 112-5. 34 Rodgers GP, Heymach GJ. Cryoprecipitate therapy 18 Botero SD, Holmquist ND. Cytologic diagnosis of in amniotic fluid embolization. Am J Med 1984; 76: arrmiotic fluid embolus - report of a case. Acta Cytol 916-20. (Baltimore) 1979; 23: 465-6. 35 Kern SB, DuffP. Localized amniotic fluid embo- 19 Tuck CS. Amniotic fluid embolus. Proc Royal Soc lism presenting as ovarian vein thrombosis and re- Med 1972; 65: 94-5. fractory postoperative fever. Am J Clin Pathol 1981; 20 Roche WD, Norris HJ. Detection and significance 76: 476-80. of maternal amniotic fluid embolism. Obstet Gyne- 36 Editorial. Amniotic-fluid embolism. Lancet 1979; 2: col 1974; 43: 729-31. 398-400. 21 Garland IWC, Thompson WD. Diagnosis of am- niotic fluid embolism using an antiserum to human Rrsum6 keratin. J Clin Pathol 1983; 36: 625-7. Quatre cas probable d'embolie de liquide amiotique 22 Mayer JR. Embolis pulmonar-caseosa. Brazil Me- (AFE) sont revus. Le rdsultat final est apparamment dico 1926; 2: 301-3. d~termind par la sdvdriM de l'insulte et possiblement le 23 AttwoodHD. Matthew Baillie - a possible early temps de la grossesse plus que par la conduite th3rapeuti- description of amniotic fluid embolism. Aust NZ J que. Deux cas sont survenus trt lors du travail; accune Obstet Gynaecol 1979; 19: 176-7. patiente n'a retrouv~ connaissance. Une est d~c~d~e une 24 Steiner PE, Lushbaugh CC. Maternal pulmonary semaine plus tard alors que I' autre a souffert de dom- embolism by amniotic fluid. JAMA 1941; 117: mage cdrdbral permanent sdv~re. Les deux autres cas 1340-5. sont survenus lors de la dilatation et curetage, une pour 25 Lewis TLT. Progress in clinical obstetrics and gyne- avortement th3rapeutique d la quatorzi~me semaine de la cology. 2nd ed. London: Churchill Livingstone gestation et l'autre pour avortement ~ la vingtidme 1964, p 48. semaine de gestation. Ces deux patientes n' ont accusd 26 Guidotti R J, Grimes DA, Cates W Jr. Fatal amnio- aucun dommage aprds l" accident. tic fluid embolism during legally induced abortion, La coagulation intravasculaire diss~minde (DIC) ~tait United States, 1972 to 1978. Am J Obstet Gynecol pr3sente pour les quatre cas. Pour les patientes au travail 1981; 141: 257-61. ce syndrome est survenu presqu'immddiatement. Pour 27 Cates W Jr, Boyd C, Halvorson-Boyd G, Holck S, celles subissant une dilatation et curetage ce syndrome Gilchrist TF. Death from amniotic fluid embolism est survenu aprds que les patientes se sont r~tablies mais and disseminated intravascular coagulation after a dtaient en observation aux soins intensifs. curettage abortion. Am J Obstet Gynecol 1981; 141 : L' embolie de liquide amiotique peut survenir apparem- 346-8. ment lors d'un travail sans probl~me. Elle doit #tre 28 Turner R, Gusack M. Massive amniotic fluid embo- suspecMe lorsqu' un coUapsus non expliqud survient lors lism. Ann Emerg Med 1984; 13: 359-61. d'une dilatation et curetage au second trimestre. Parce 29 Corridan M, Kendall ED, Begg JD. Cord entan- que une coagulation intravasculaire diss~minde peut ~tre glement causing premature placenta/separation and provoqu3e, les patientes doivent ~tre transfdrdes dz un arnniotic fluid embolism. Br J Obstct Gynccol 1980; centre fournissant des services d' hdmatologie. 87: 935-40. 30 Plauchd WC. Anmiotic fluid embolism (letter to the editor). Am J Obstet Gynecol 1983; 147: 982. 31 SparrRA, PritchardJA. Studies to detect the es- cape of amniotic fluid into the maternal circulation during parturition. Surg Gynecol Obstet 1958; 107: 560-4. 32 Moore PG, James OF, Saltos N. Severe amniotic fluid embolism: case report with hemodynamic findings. Anaesth Intensive Care 1982; 10: 40-4.