Severe Thrombocytopenia, Type 2B von Willebrand Disease and Pregnancy

A 35-yr-old, 157-cm, 71-kg, uniparous woman, gravida 2 presented to our high-risk anesthesia consultation service at 36 weeks’ gestation. Although the patient reported no known drug allergies, the use of aspirin was associated with prolonged bleeding. Her history was notable for chronic hepatitis C, with no associated liver dysfunction, resulting from a blood transfusion and type 2B vWD, which was diagnosed as a child because of a strong family history (grandmother, mother, uncle, and brother) of type 2B vWD. Her obstetric history included a normal spontaneous vaginal delivery at 39 weeks’ gestation 2 yr previously without neuraxial analgesia. The platelet count at the time of her first delivery was 20,000 giga/l, and the patient was given antihemophilic factor (factor VIII) (Humate-P; Aventis, Bridgewater, NJ) 3,000 ristocetin cofactor units intravenously on becoming fully dilated and per day for 48 h. Five days postpartum, the patient was readmitted for a perineal bleeding from a laceration, which was treated successfully with vaginal packing, four units of packed red blood cells, factor VIII 3,000 ristocetin cofactor units intravenously, and aminocaproic acid (Amicar; Immunex, Seattle, WA) 5 g per os  loading dose followed by 3 g per os  every 3 h.

The current pregnancy had a breech presentation, for which the patient was scheduled for a cesarean delivery. The platelet count at the time of consultation was 46,000 giga/l and the international normalized ratio 1.0; consultation with the hematologist included a plan to give factor VIII 4,000–4,500 ristocetin cofactor units followed by a unit of pheresis platelets starting 1 h before the procedure. On the day of the delivery, 2 weeks later, the initial platelet count was 36,000 giga/l and factor VIII and platelets were given. Within 30 min of the transfusion and before a planned spinal anesthetic technique, a repeat platelet count was drawn; the results demonstrated a platelet count of 30,000 giga/l. An additional unit of platelets was given in anticipation of bleeding during the cesarean delivery and the anesthetic plan was changed to a general anesthetic. Anesthesia was induced via  rapid sequence with 250 mg sodium thiopental, 100 mg of succinylcholine, and 100 μg of fentanyl. A male infant with Apgar scores of 81and 95was delivered 3 min after the anesthetic induction. On closing the abdomen, no excessive intraoperative bleeding was observed, the total blood loss was estimated at 1,000 ml, and the patient had received a total of 2,800 ml of lactated Ringer’s solution. After tracheal extubation, an intraoperative sample sent for a platelet count was reported as 10,000 giga/l. The hematology consultant recommended additional factor VIII (5,000 units every 12 h for 3 days) to be commenced immediately. The obstetric team followed the recommendation and gave an additional unit of pheresis platelets as well. Six hours after delivery, serial platelet counts revealed progressive increases from 20,000 giga/l to 31,000 giga/l at 24 h, 42,000 giga/l at 48 h, and 53,000 giga/l at the time of discharge on postoperative day four. No postpartum hemorrhagic complications occurred and no additional thrombogenic products were needed. At time of hospital discharge, vWF ristocetin cofactor activity (vWF:RCo) and factor VIII concentrations were 454% and 258%, respectively.

Previously defined by multimer analysis, the current classification scheme recognizes three general types of vWD based on molecular mechanisms.7Types 1 and 3 represent partial and severe quantitative deficiencies of vWF and type 2 exhibits a qualitative deficiency. A characteristic gain-of-function phenotype, subtype 2B, may be the result of single amino acid substitutions within the A1 domain of vWF.4,8,9These changes result in an increased interaction between vWF and the glycoprotein Ib-IX-V complex in platelets10and an acceleration in binding and clearance of vWF and platelets; this interaction can be exacerbated by an increased production of vWF that can be produced with stressful clinical situations, of which pregnancy is well recognized.2,11,12The resulting thrombocytopenia can be impressive, as observed in our case, and of sufficient concern as to allow some hematologists to support vWD testing for all women with thrombocytopenia in pregnancy.5 

Neonatal thrombocytopenia may also occur with parturients with type 2B vWD,1,4and in our case, the initial platelet count in the newborn was 35,000 giga/l. Antenatal testing for fetal vWD is possible but not often recommended; fetal morbidity and mortality may be associated with testing and the results are not always predictive of neonatal bleeding. The collection of cord blood at the time of delivery for vWD testing is preferable, and in our case, although the neonatal plasma vWF quantity was normal, the vWF activity and the amount of plasma factor VIII were markedly decreased. The possibility of neonatal vWD should not, however, prompt a cesarean delivery, as increased maternal bleeding may occur.13Instead, a vaginal delivery with the avoidance of an episiotomy or instrumented delivery is preferable.2The decision for cesarean delivery in our case was based on the breech presentation and an appropriate obstetric reluctance to perform either an external cephalic version or a vaginal breech delivery. All intramuscular injections were avoided in the mother and neonate, and a planned circumcision in the male infant was postponed.

Although desmopressin, a vasopressin analog that raises plasma vWF, is a common therapy for vWD, in type 2B vWD it can increase the abnormal vWF, resulting in further binding of platelets, depletion of high molecular weight multimers, and potentially greater bleeding.10,14–16As such, the preferred therapy for this specific vWD subtype is factor VIII plasma concentrate (FVIII) (Humate-Por Alphanate; Grifols, Los Angeles, CA), which, although FDA approved for hemophilia and not vWD, provides the following corrective components: 2.5 IU vWF:RCo, 1 IU of FVIII, and a near-normal count of high molecular weight multimers.1,14,17,18Threshold values of vWF:RCo and FVIII greater than 50 IU/dl at the time of delivery and for 72–96 h after an uncomplicated vaginal delivery and for 96–120 h after a cesarean section have been proposed.18 

Although the treatment and goals for vWF:RCo and FVIII are relatively straightforward, the response to thrombocytopenia remains controversial. In part, this is because of the variation in thrombocytopenia observed in patients with type 2B vWD, which can be both intermittent and variable in expression, from mild to severe. In addition, the etiology of the thrombocytopenia is often unclear. Platelet aggregation has been confirmed by a number of investigations to occur spontaneously and with the use of desmopressin in patients with type 2 vWD.15,19However, recent work suggests that desmopressin with this subtype may cause a “false” thrombocytopenia, resulting from short-term agglutination, but not aggregation.20Moreover, the significance of the thrombocytopenia is unclear; some have suggested that the thrombocytopenia per se  observed in type 2B vWD is not a major risk factor for bleeding, representing sequestration and margination of platelets, with later availability, rather than consumption.5 

In our case, FVIII was given to increase the vWF:RCo and FVIII concentrations. The decision to transfuse platelets was prompted by the desire to have a platelet count greater than an arbitrary threshold of 50 giga/l for the spinal anesthesia and cesarean section. Although these platelets may have provided additional substrate for further aggregation and consumption of both vWF and platelets,4we hypothesized that the degree of thrombocytopenia would be further accentuated by the surgical stress. Moreover, therapy with platelet transfusions has been found useful in some patients with type 2B vWD.15Whether FVIII given simultaneously with platelet transfusions is more efficacious in treating type 2B vWD induced thrombocytopenia is currently unknown. What our case demonstrates, however, is that the quantitative response to FVIII and platelet transfusions is slow.

Ultimately, our decision to proceed with general anesthesia was based on a risk-benefit analysis, weighing the risk of neuraxial bleeding with a regional anesthetic versus  the more common failed airway scenario observed in the parturient.21–23Although difficult airways can not always be identified a priori ,24,25especially during pregnancy,22,26we elected the general anesthetic based on our patient having a Mallampati class 2 airway, a wide mouth opening, a 5 cm thyromental distance, and a reasonable body mass index. If a more challenging airway had been observed, a regional anesthetic would have been more strongly considered, however, not before adequate vWF:RCo, FVIII, and platelet concentrations were obtained. Had our parturient presented in labor with a vertex presentation, intravenous patient-controlled analgesia with opioids would have been offered; were the desire for additional analgesia requested and hematologic goals achieved, a regional labor analgesic would have been considered. Despite these care algorithms, which were consistent with the desire of the patient to remain awake during her cesarean delivery, the falling platelet concentration despite hematologic intervention resulted in a general anesthetic being used.

This report emphasizes the severity of thrombocytopenia that can result in parturients with type 2B vWD. Early consultation with all involved healthcare providers, including obstetricians, hematologists, and anesthesiologists, and appropriate considerations should be made. Experience with this subtype of vWD suggests a correlation between vWF:RCo and FVIII concentrations and normal hemostasis exists,1and therapy should be directed to these endpoints. The implications and treatment of thrombocytopenia deserve further investigation. Consideration of neuraxial analgesic or anesthetic techniques in these patients should weigh the risks and benefits of various approaches and the observation that certain therapies for type 2B vWD may be ineffective or slow.