The coagulation system undergoes significant change during pregnancy. The clinician caring for the parturient must understand these changes, particularly when the parturient has a pre-existing haematological condition. Because many haematological conditions are rare, there often is limited information to guide the obstetric and anaesthetic management of these parturients.
To limit blood loss after trauma it is essential to seal bleeding vessels without affecting blood flow permanently. Haemostasis, defined as the arrest of bleeding, comes from the Greek roots, haeme meaning blood and stasis meaning causing to stop. The process of haemostasis is a dynamic and delicate equilibrium between coagulation and fibrinolysis ( Fig. 1 ). Coagulation results from an interaction among vessel walls, platelets and coagulation factors. Following endothelial damage, platelets adhere to the subendothelium forming a platelet plug which then becomes permanent with fibrin deposition. Clot formation is limited by antithrombin (AT) and proteins C and S. The fibrinolytic system functions to maintain the fluid state through the breakdown of fibrin by plasmin. Plasmin is generated from plasminogen by the action of tissue plasminogen activator (t-PA).
Physiological changes to coagulation during pregnancy
Pregnancy is associated with changes in haemostasis, including an increase in the majority of clotting factors, a decrease in the quantity of natural anticoagulants and a reduction in fibrinolytic activity. These changes result in a state of hypercoagulability, are likely due to hormonal changes and increase the risk of thromboembolism.
The increase in clotting activity is greatest at the time of delivery with placental expulsion, releasing thromboplastic substances. These substances stimulate clot formation to stop maternal blood loss. As placental blood flow is up to 700 ml min −1 , considerable haemorrhage can occur if clotting fails. Coagulation and fibrinolysis generally return to pre-pregnant levels 3–4 weeks postpartum.
- a.
Platelets
The platelet count decreases in normal pregnancy possibly due to increased destruction and haemodilution with a maximal decrease in the third trimester.
- b.
Coagulation factors
Factors VIII (FVIII), von Willebrand factor (vWf), ristocetin cofactor (RCoA) and factors X (FX) and XII (FXII) increase during pregnancy. Levels of factor VII (FVII) increase gradually during pregnancy and reach very high levels (up to 1000%) by term. Fibrinogen also increases during pregnancy with levels at term 200% above pre-pregnant levels.
Other factors either remain at non-pregnant levels or decrease during pregnancy. Factor XIII (FXIII), which is responsible for stabilising fibrin, increases in the first trimester but by term it is 50% of non-pregnant levels. Factor V (FV) concentrations increase in early pregnancy then decrease and stabilize. Factor II (FII, prothrombin) levels may increase or not change in early pregnancy but are normal by term. There is debate about factor XI (FXI) levels with reports indicating increases or decreases. Similarly, FIX levels are reported as increasing, decreasing or remaining stable throughout pregnancy. In one study, 50% of carriers of FIX deficiency had FIX levels ≤50 IU dl −1 at term.
Protein C levels remain the same or are slightly increased during pregnancy while protein S decreases. AT levels remain normal during pregnancy.
- c.
Fibrinolysis
Fibrinolysis is reduced in pregnancy due to decreases in t-PA activity, which remains low until 1-h postpartum when activity returns to normal. This reduction is due to the gradual, eventually threefold, increase in plasminogen activator inhibitor-1 (PAI-1) and the increasing levels of plasminogen activator inhibitor-2 (PAI-2). The placenta produces PAI-1 and is the primary source of PAI-2. PAI-2 levels at term are 25 times that of normal plasma. Postpartum, t-PA levels quickly return to normal as PA-1 levels decrease; however, PA-2 levels remain elevated for a few days.
Thrombin-activatable fibrinolysis inhibitor (TAFI) (an antifibrinolytic which cleaves the C-terminal lysine in fibrin to render it resistant to cleavage by plasmin) levels are increased in the third trimester. D-Dimer levels increase in pregnancy but are not thought to indicate intravascular coagulation as fibrinolysis is depressed. These D-Dimers may originate from the uterus.
Tests of coagulation
Obstetric anaesthetists are concerned that providing neuraxial anaesthesia in parturients with coagulation abnormalities may cause bleeding in the epidural or subarachnoid space with neurological impairment (spinal/epidural/neuraxial haematoma). Neuraxial haematomas associated with neuraxial blocks have been reported rarely in parturients. In the non-pregnant population, spinal haematomas associated with neuraxial blocks often occur in patients with an underlying coagulopathy. Ideally, there would be a test that would predict the risk of bleeding in the epidural/spinal space.
- A.
Platelet count
Some anaesthetists rely on platelet count as a screening test for coagulation abnormalities in healthy parturients prior to neuraxial block. However, the American Society of Anesthesiologists Practice Guidelines for Obstetric Anesthesia states that a platelet count is not essential prior to neuraxial block in a healthy parturient with no risk factors for bleeding. Although there are some rare conditions that result in abnormal platelet function (e.g., May Hegglin anomaly) a normal platelet count will provide reassurance that coagulation is normal in a healthy parturient. A platelet count at term <70 × 10 9 l −1 may indicate the presence of HELLP (haemolysis, elevated liver enzymes, low platelets) syndrome, disseminated intravascular coagulation (DIC), immune thrombocytopaenia and other rare conditions co-existing with pregnancy. Questions have been raised about platelet function in pre-eclampsia, but as these studies used bleeding time, their results are not considered definitive. However, the concern remains and is not yet resolved.
- B.
Prothrombin time
As most coagulation factors increase in normal pregnancy, the prothrombin time (PT) and the activated partial thromboplastin time (APTT) are shortened. The PT and its derived measure, international normalised ratio (INR), test for factors such as FII, FV, FVII, FX and fibrinogen. Some nutritional deficiencies and/or liver disease will decrease these factors prolonging the PT. Furthermore, PT and APTT may be artificially prolonged due to the presence of an antiphospholipid antibody (APLA), such as lupus anticoagulant. In fact, patients with APLA are prothrombotic.
- C.
Activated partial thromboplastin time
The APTT is considered a good screening test for deficiencies of FVIII, FIX, FXI and FXII. The APTT may be prolonged by the presence of an APLA and/or unfractionated/standard heparin (SH). To differentiate between the presence of an APLA and a factor deficiency, the patient’s plasma is mixed 50:50 with normal plasma. If the APTT remains abnormal an APLA is present.
- D.
Bleeding time
Since its introduction, bleeding time was used to assess platelet function in parturients with thrombocytopaenia. A prolonged bleeding time was thought to predict the risk of bleeding, which theoretically could increase the risk of a neuraxial haematoma. However, bleeding time is used rarely now as it has a number of disadvantages: it is invasive, unreliable, highly operator dependent and not suitable to repeated tests. Bleeding time also is insensitive, especially to mild platelet defects, and not a good predictor of bleeding risk. Because of these problems, other tests are being studied.
- E.
Thrombo-elastography
The thromboelastograph (TEG) provides information about the various stages of coagulation and fibrinolysis. The maximum amplitude (MA) is thought to represent platelet function. TEG is used by some centres to predict the risk of bleeding from coagulation abnormalities; however, the sensitivity and specificity of TEG in pregnancy remain unproven. An abnormal test has not been shown to be predictive for development of a neuraxial haematoma, but most anaesthetists will not provide regional anaesthesia if the MA is abnormal.
- F.
Platelet function analyser ® (PFA)
This test measures the speed of formation of a platelet plug in vitro , expressed as closure time in seconds. Studies in parturients suggest that it is an effective bedside test of platelet function ; however, evidence is lacking to support its routine use.
Tests of coagulation
Obstetric anaesthetists are concerned that providing neuraxial anaesthesia in parturients with coagulation abnormalities may cause bleeding in the epidural or subarachnoid space with neurological impairment (spinal/epidural/neuraxial haematoma). Neuraxial haematomas associated with neuraxial blocks have been reported rarely in parturients. In the non-pregnant population, spinal haematomas associated with neuraxial blocks often occur in patients with an underlying coagulopathy. Ideally, there would be a test that would predict the risk of bleeding in the epidural/spinal space.
- A.
Platelet count
Some anaesthetists rely on platelet count as a screening test for coagulation abnormalities in healthy parturients prior to neuraxial block. However, the American Society of Anesthesiologists Practice Guidelines for Obstetric Anesthesia states that a platelet count is not essential prior to neuraxial block in a healthy parturient with no risk factors for bleeding. Although there are some rare conditions that result in abnormal platelet function (e.g., May Hegglin anomaly) a normal platelet count will provide reassurance that coagulation is normal in a healthy parturient. A platelet count at term <70 × 10 9 l −1 may indicate the presence of HELLP (haemolysis, elevated liver enzymes, low platelets) syndrome, disseminated intravascular coagulation (DIC), immune thrombocytopaenia and other rare conditions co-existing with pregnancy. Questions have been raised about platelet function in pre-eclampsia, but as these studies used bleeding time, their results are not considered definitive. However, the concern remains and is not yet resolved.
- B.
Prothrombin time
As most coagulation factors increase in normal pregnancy, the prothrombin time (PT) and the activated partial thromboplastin time (APTT) are shortened. The PT and its derived measure, international normalised ratio (INR), test for factors such as FII, FV, FVII, FX and fibrinogen. Some nutritional deficiencies and/or liver disease will decrease these factors prolonging the PT. Furthermore, PT and APTT may be artificially prolonged due to the presence of an antiphospholipid antibody (APLA), such as lupus anticoagulant. In fact, patients with APLA are prothrombotic.
- C.
Activated partial thromboplastin time
The APTT is considered a good screening test for deficiencies of FVIII, FIX, FXI and FXII. The APTT may be prolonged by the presence of an APLA and/or unfractionated/standard heparin (SH). To differentiate between the presence of an APLA and a factor deficiency, the patient’s plasma is mixed 50:50 with normal plasma. If the APTT remains abnormal an APLA is present.
- D.
Bleeding time
Since its introduction, bleeding time was used to assess platelet function in parturients with thrombocytopaenia. A prolonged bleeding time was thought to predict the risk of bleeding, which theoretically could increase the risk of a neuraxial haematoma. However, bleeding time is used rarely now as it has a number of disadvantages: it is invasive, unreliable, highly operator dependent and not suitable to repeated tests. Bleeding time also is insensitive, especially to mild platelet defects, and not a good predictor of bleeding risk. Because of these problems, other tests are being studied.
- E.
Thrombo-elastography
The thromboelastograph (TEG) provides information about the various stages of coagulation and fibrinolysis. The maximum amplitude (MA) is thought to represent platelet function. TEG is used by some centres to predict the risk of bleeding from coagulation abnormalities; however, the sensitivity and specificity of TEG in pregnancy remain unproven. An abnormal test has not been shown to be predictive for development of a neuraxial haematoma, but most anaesthetists will not provide regional anaesthesia if the MA is abnormal.
- F.
Platelet function analyser ® (PFA)
This test measures the speed of formation of a platelet plug in vitro , expressed as closure time in seconds. Studies in parturients suggest that it is an effective bedside test of platelet function ; however, evidence is lacking to support its routine use.
Disorders that affect coagulation
Thrombocytopaenia
Thrombocytopaenia affects 6–10% of all pregnancies. A decrease in platelet count is normal in pregnancy although most platelet counts remain within normal limits (≥150 × 10 9 l −1 ). A lower than physiological platelet count may occur in pregnancy for many reasons, ranging from the relatively benign, gestational thrombocytopaenia to more sinister conditions, such as HELLP syndrome.
Some pre-existing conditions that may cause thrombocytopaenia at term include: type 2b von Willebrand disease (vWD), idiopathic thrombocytopaenic purpura (ITP), lupus erythematosus and bone marrow disease. Pregnancy-related causes of thrombocytopaenia include gestational thrombocytopaenia, pre-eclampsia including HELLP syndrome, acute fatty liver of pregnancy, DIC and thrombocytopaenic purpura. Severe sepsis, some medications (e.g., SH) and viral infections may coincide with pregnancy producing thrombocytopaenia.
- a.
Gestational thrombocytopaenia
Gestational thrombocytopaenia is a benign condition that occurs during the third trimester with a platelet count that is generally ≥90 × 10 9 l −1 but may be as low as 70 × 10 9 l −1 . In one study, the incidence of thrombocytopaenia was 7.3%, of which 81% were gestational thrombocytopaenia. The diagnosis is made by exclusion of other disorders. Parturients with gestational thrombocytopaenia are asymptomatic, have a normal platelet count in early pregnancy with no history of previous thrombocytopaenia and no evidence of pre-eclampsia. These patients are not at increased risk of haemorrhage, and there is no contraindication to neuraxial anaesthesia.
- b.
Idiopathic thrombocytopaenic purpura
ITP may present for the first time in pregnancy and is the most common reason for isolated thrombocytopaenia in the first trimester. ITP is an autoimmune disorder which is associated with the production of anti-platelet immunoglobulin (IgG), resulting in platelet destruction in the reticulo-endothelial system. However, anti-platelet IgG is not always present, making the diagnosis problematic.
Laboratory findings include an isolated thrombocytopaenia presenting either pre-pregnancy or in early pregnancy with large, well-granulated platelets. Despite low platelet numbers, haemostasis is often normal. The British Committee for Standards in Haematology guidelines recommend a platelet count of 80 × 10 9 l −1 for epidural anaesthesia. However, many anaesthetists, such as the authors, would do a neuraxial block (especially spinal anaesthesia) in healthy, asymptomatic ITP patients with platelet counts ≥50 × 10 9 l −1 . This practice is based on the consideration that a platelet count ≥50 × 10 9 l −1 is considered sufficient for caesarean delivery.
The key issue in managing an ITP parturient is whether intervention is necessary to prevent haemorrhage. In patients with very severe, symptomatic thrombocytopaenia (platelet count ≤10 × 10 9 l −1 ), treatment is urgently required. One gram per kilogram per day of intravenous gammaglobulin (IVIg), administered over 2 days, will raise the platelet count in approximately 75% of ITP patients and the platelet count will remain elevated for 3–6 weeks. Further IVIg may be required later in pregnancy. Another option is high-dose corticosteroids (e.g., prednisone 1 mg kg −1 (pre-pregnancy weight) daily). There are no trials comparing IVIg to corticosteroids for effect. Administration of IVIg and/or corticosteroids usually will raise the platelet count to enable neuraxial anaesthesia. Occasionally, a splenectomy is required during pregnancy and this may be done laparascopically in the second trimester. Platelet transfusion is generally contraindicated, but, in the setting of acute haemorrhage and an extremely low platelet count, it may be life-saving.
Anaesthetic implications of thrombocytopaenia
There is considerable debate about administering neuraxial anaesthesia in parturients with thrombocytopaenia. A platelet count from 70 × 10 9 l −1 to 100 × 10 9 l −1 in an otherwise healthy parturient should not contraindicate regional anaesthesia. Most American anesthesiologists would insert an epidural in a healthy parturient with a platelet count ≥80 × 10 9 l −1 . A Canadian survey reported that 16.2% of university-based anaesthetists would place an epidural if the platelet count was ≥50 × 10 9 l −1 in an otherwise healthy parturient.
The situation is more controversial in the setting of pre-eclampsia, including HELLP syndrome. In this situation, most anaesthetists consider the platelet count, the clinical picture (i.e., haemorrhage risk and evidence of coagulopathy) and whether the thrombocytopaenia is stable or decreasing. In one report of HELLP syndrome, 12 parturients with a platelet count <50 × 10 9 l −1 received uneventful epidural anaesthesia. In some cases a platelet transfusion was given immediately prior to neuraxial anaesthesia. In contrast, there is a report of an epidural haematoma in two cases of HELLP syndrome: one after spinal anaesthesia and the other after epidural catheter removal. Both had signs of coagulopathy. Studies using TEG in women with pre-eclampsia suggest that coagulation is normal with a platelet count >75 × 10 9 l −1 .
Factor deficiencies
The most common factor deficiencies encountered in pregnancy are von Willebrand disease and haemophilia carrier states.
- a.
von Willebrand disease (vWD)
von Willebrand disease is the most frequent inherited bleeding disorder, affecting 1% of the population. vWD is due to quantitative (types 1 and 3) or qualitative (type 2) defects of vWf with autosomal dominant transmission. vWf protects circulating FVIII from proteolysis and is required for normal platelet adhesion to the injury site. In vWD the marked reduction in FVIII activity produces the clinical syndrome.
There are three types of vWD: types 1, 2 and 3 with subtypes in type 2 (2A, 2B, 2M, 2N). Type 1, the most frequent with an incidence of 75%, is usually mild and is characterised by a deficiency of vWf. Type 2 and its subtypes involve abnormalities of binding to glycoproteins and FVIII (qualitative defect). Type 3 is characterised by the complete absence of vWf and is usually severe. Thrombocytopaenia occurs with type 2b.
The diagnosis of vWD is made clinically and in the laboratory. The most common symptom of Type 1 vWD in women is menorrhagia, but there may be a history of bruising, epistaxis or other mucosal bleeding, and bleeding after dental extractions or surgery. Laboratory diagnosis involves measuring vWf, RCoA and FVIII levels.
Women with vWD have reasonably good pregnancy outcomes. The increases in fibrinogen, FVII, FVIII, FX and vWf during pregnancy are considered protective. Levels of vWf, RCoA and FVIII should be checked in early pregnancy and in the third trimester to ensure adequate levels for delivery. Pregnant patients with type 3 vWD, type 2 vWD or type 1 vWD with FVIII ≤50 IU dl −1 , vWf:RCo ≤50 IU dl −1 or a history of severe bleeding should be referred to a centre with appropriate consultants, laboratory and blood bank facilities. Postpartum haemorrhage (PPH) is a common complication in parturients with vWD. Levels of vWf begin decreasing 6 h postpartum, returning to pre-pregnancy levels by 7–20 days.
Desmopressin (DDAVP) is considered safe for parturients with type 1 vWD. The normal dose is 0.3 μg kg −1 , up to a maximum of 20 μg kg −1 . Desmopressin immediately increases vWf and FVIII:RCoA by 200–300%. As hyponatraemia is a complication of repeated doses of DDAVP, some recommend a minimum of 12 hourly intervals and restriction of fluids. Desmopressin is contraindicated or the response is variable in type 2 and ineffective in type 3 vWD.
Anaesthetic implications of vWD
Based on case reports and retrospective studies, neuraxial anaesthesia appears safe in type 1 vWD parturients with normal third trimester coagulation (FVIII ≥50 IU dl −1 ). In a study of 64 vWD parturients, 15 (17 deliveries) received uneventful, epidural anaesthesia without prophylactic DDAVP or factor concentrate. However, eight had a PPH, coinciding with the postpartum fall in FVIII and vWf levels. Seven of the eight with PPH had type 1 vWD whilst the eighth had type 2A. An epidural catheter should not be removed if coagulation is abnormal.
Most anaesthetists consider neuraxial block contraindicated in types 2 and 3 vWD but there is a report of spinal anaesthesia in a parturient with type–2M vWD whose coagulation was normal except for prolonged platelet adhesion and aggregation.
- b.
Factor VIII Deficiency
Also known as haemophilia A, FVIII deficiency is rare in females as it is an X-linked, recessive condition. Rarely, lyonisation of the X chromosome occurs and a woman has low FVIII levels. Haemophilia A is diagnosed when FVIII activity is ≤35%. Recombinant FVIII (rFVIII) is the treatment of choice, if required, in pregnant haemophilia A carriers. Although FVIII levels usually normalise during pregnancy, there are a few case reports of severe FVIII deficiency so check FVIII levels early in pregnancy and in the third trimester. In two cases of severe FVIII deficiency, the parturients had successful epidural analgesia for labour after rFVIII. One parturient had a spontaneous vaginal delivery and the other a caesarean delivery. The latter parturient had a brachial vein thrombosis 10 days postpartum.
- c.
Factor IX deficiency
Similar to haemophilia A, FIX deficiency (haemophilia B, Christmas disease) is X linked, recessive and hence rare in females. As FIX levels may decrease during pregnancy these women may have inadequate coagulation. Known carriers of haemophilia B should have FIX levels checked in the third trimester to ensure adequate coagulation for delivery. Recombinant FIX is the treatment of choice.
Anaesthetic implications of haemophilia carriers
Neuraxial anaesthesia is not contraindicated provided:
- (i)
Factor VIII level is ≥50 IU dl −1 at term;
- (ii)
PT and APTT are normal;
- (iii)
There is no evidence of bleeding or bruising;
- (iv)
Replacement therapy has raised the level to ≥50 IU dl −1 with normal coagulation in women whose FVIII levels were ≤50 IU dl −1 .
In 53 haemophilia A carriers and 12 haemophilia B carriers who had 90 pregnancies, FVIII and FIX levels increased during pregnancy but FVIII levels increased more. In a subset, 8% of haemophilia A carriers and 50% of haemophilia B carriers had FVIII or FIX levels ≤50 IU dl −1 at term. Regional anaesthesia was used in 25 deliveries; 20 had normal factor levels (≥50 IU dl −1 ) and normal coagulation screens. Five other parturients received regional anaesthesia; one after prophylactic recombinant factor and in the remaining four the carrier status was unknown. There were no complications.
- d.
Rarer coagulation disorders and their anaesthetic implications ( Table 1 )
