Use of Aminocaproic Acid With Bivalirudin for Hemostatic Management of Abdominal Surgery for Neonate on Extracorporeal Support

Case Presentation A preterm neonate (34 weeks of gestation) weighing 2.5 kg with a history of fetal supraventricular tachycardia (SVT, successfully treated with maternal flecainide, though fetal development of ascites was noted) was transferred to our institution at 10 days of life. Pretransfer electrocardiogram revealed preexcitation, with intermittent orthodromic narrow complex tachycardia at a rate of 220 bpm (Figure 1). Before transfer, antiarrhythmic treatment strategies included propranolol, flecainide, amiodarone, and electrical cardioversion. On the day of transfer, the patient developed incessant SVT, with onset of emesis, hematochezia, abdominal distension, and pneumatosis on abdominal x-ray concerning for necrotizing enterocolitis (NEC).Figure 1.: A: Baseline echocardiogram showing sinus bradycardia at a rate of 90 bpm, left axis deviation, low voltage, and preexcitation. B: Orthodromic reentrant tachycardia at a rate of 220 bpm with delayed atrioventricular conduction secondary to recent Flecainide loading. aVF, augmented vector foot; aVL, augmented vector left; aVR, augmented vector right.Initial assessment by clinical exam, laboratory markers of oxygen delivery, and echocardiogram showed evidence of preserved cardiac output. An infusion of procainamide was attempted, resulting in modestly improved rate control without conversion to sinus rhythm. Mild lactatemia soon developed. Given concerns that NEC was a manifestation of compromised oxygen delivery and that the SVT duration was approaching 30 hours, we “electively” cannulated on to venoarterial (VA) extracorporeal membrane oxygenation (ECMO) to maintain end-organ perfusion while facilitating further antiarrhythmic management. Cannulation was uneventful using 8 and 10 French arterial and venous cannulas and an Affinity CP centrifugal pump (Medtronic, Minneapolis, MN). ECMO flow was established at 0.4 L/min. Systemic anticoagulation with bivalirudin was started per institutional guideline1 at a dose of 0.1 mg/kg/h, followed by amiodarone loading and successful electrical cardioversion over the following 12 hours. Despite ECMO support, the patient deteriorated with distributive shock requiring new vasopressor support and frequent episodes of nonsustained SVT, lasting for minutes. Leukocytosis and thrombocytopenia developed, and his abdominal exam became increasingly concerning for intestinal perforation. Without the need for systemic anticoagulation, surgical intervention would typically be pursued, but it was considered high risk due to potentially life-threatening hemorrhage. The next several hours showed rising lactatemia and worsening frequency of breakthrough SVT. The heightened concern for perforation and bowel necrosis shifted the risk-benefit of surgical intervention toward performing a bedside laparotomy while on VA-ECMO on the second day of admission. The ECMO anticoagulation requirements along with the need for abdominal surgery were challenging. Before the procedure, platelets were administered; bivalirudin infusion was stopped; resuscitative blood products, factor VIIa, and aminocaproic acid (ACA) were made available at bedside. One hour after bivalirudin discontinuation, the laparotomy started. Severe hemorrhaging was quickly noted, requiring 125 ml/kg of packed red blood cells, platelets, and fresh frozen plasma in addition to two units of cryoprecipitate for resuscitation. Medication management included two doses of 0.08 mg/kg of factor VIIa and an ACA infusion of 20 mg/kg/h for 1 hour before hemostasis was achieved. This was further complicated by frequent adenosine-responsive breakthrough episodes of SVT, which worsened in burden during the procedure. At the time of operation, enteric contents were evacuated, and a right colon perforation secondary to full thickness necrosis was identified. In addition, there was necrosis of the entire colon extending to just proximal to the peritoneal reflection. A subtotal colectomy was performed with the bowel left in discontinuity and abdomen left open with plans for a second-look surgery. We elected to hold bivalirudin and continue ACA infusion due to high risk of hemorrhage with open abdomen, a strategy that proved effective. The postsurgical course was complicated by postsurgical inflammatory response. Combined with the massive product resuscitation, this led to progressive third spacing, interfering with ECMO flows. Marked vasodilation with frequent recurrent nonsustained SVT persisted. As such, catecholaminergic medications that may worsen the arrhythmia were avoided. Instead, vasopressin was used at the cost of potentially limiting splanchnic circulation and further intestinal injury. The ACA infusion continued for 24 hours postoperatively without additional anticoagulation. There was gradual improvement in his hemodynamics without further hemorrhage or substantial worsening of ECMO clot burden for the next 72 hours. A second-look laparotomy was performed on day 5 of admission to assess intestinal viability, and an additional 36 cm of ilium was resected. An ACA infusion was restarted at procedure onset, and no significant hemorrhage occurred intra- or postoperatively. Four hours after the procedure, bivalirudin was restarted at 0.05 mg/kg/h with more conservative anticoagulation/hematologic goals (partial thromboplastin time of 40–50 sec, fibrinogen of >150 mg/dL, platelet of >30 K/cumm, and hemoglobin >10 g/dL). The ACA infusion was continued, overlapping with bivalirudin for additional 24 hours. Three days later, a third abdominal exploration was required to assess intestinal viability without the need for further intestinal resection. The same hemostatic strategy was repeated and again proved to be effective. Removal of necrotic bowel, relief of abdominal pressure, and antibiotic therapy allowed for resolution of the distributive shock and improved hemodynamics. SVT burden decreased as well but did not resolve, with nonsustained episodes occurring one to five times daily. The hemodynamic stability allowed the patient to undergo ablation 12 days after initial presentation, where successful cryoablation of a right posteroseptal accessory pathway was achieved while supported on VA-ECMO. Two day postprocedure, patient was successfully decannulated from ECMO. Gastrointestinal rehabilitation continued slowly. There were no further episodes of SVT. Three more abdominal operations were performed for washout, to establish ileocolic anastomosis and to close the abdominal wall 120 days after presentation. At the time of this report, patient was discharged home on total parenteral nutrition with no other reliance on technology. Discussion This is an example of a critically ill child in whom “elective” ECMO use can prevent deterioration of oxygen delivery and irreversible organ injury. There is a time window during cardiopulmonary deterioration, when the risks associated with impaired oxygen delivery outweigh those related to ECMO. Despite advances in ECMO technology, objective criteria for timing elective ECMO utilization are lacking, and determining the optimal time before irreversible organ deterioration or cardiopulmonary decompensation remains challenging. Here, “elective” ECMO was decided without the full knowledge of the degree of NEC and systemic involvement, but with concern for uncontrolled arrhythmia leading to compromised end-organ function and potential cardiovascular collapse. Nonetheless, the patient developed distributive shock compounded by later intraoperative hemorrhagic shock. To our knowledge, the management of surgical NEC on ECMO has not been previously reported. Reports of abdominal sepsis requiring surgery2 and gastric rupture3 in pediatric ECMO have been described and, as seen in this case, resulted in life-threatening hemorrhage. Abdominal surgery here was necessary to remove necrotic bowel that would lead to overwhelming sepsis with bacterial translocation, multiorgan dysfunction, and death even with VA-ECMO support. Without relief of abdominal pressure, abdominal compartment syndrome may have progressed and impeded venous return to the ECMO circuit, compromising ECMO support.4 Hemostatic management under these circumstances was challenging. Institutionally, we have recently transitioned to bivalirudin as our primary anticoagulant for patients outside of the neonatal intensive care unit with similar outcomes to our previous heparin experience.5 During the intestinal resection, massive bleeding occurred requiring 1-to-1-to-1 blood product resuscitation of four times the approximated blood volume of this neonate in addition to ACA and factor VIIa. ACA binds competitively to plasminogen blocking its binding to fibrin and subsequent conversion to plasmin, resulting in inhibition of fibrinolysis.6 ACA has previously been reported to alleviate surgical bleeding on ECMO with heparin anticoagulation.6–9 These reports describe hemostatic strategies for congenital diaphragmatic hernia repair on ECMO involving ACA loading preoperatively with continuous infusion for at least 24 hour postoperatively and at times up to ECMO decannulation. We utilized a similar approach with bivalirudin anticoagulation because of the inherent risk of thrombosis in a neonate limited by low-flow ECMO of 0.4 L/min. To our knowledge, this strategy has not previously been described. Bivalirudin directly inhibits both circulating and clot-bound thrombin, preventing thrombin-mediated cleavage of fibrinogen to fibrin.5 We chose to overlap ACA and bivalirudin to manipulate the balance of unwanted ECMO-related fibrin deposition and necessary fibrin formation in sites of surgical bleeding. This required close monitoring for bleeding and conservative anticoagulation goals (Figure 2), and after the initial operative hemorrhage, ACA was used without bivalirudin. It is important to note that this patient did not suffer any significant thromboembolic events. The combination of continuous bivalirudin infusion with ACA may be a safe approach to prevent bleeding and clotting during and immediately after procedural interventions while on ECMO.Figure 2.: Recommended approach for a high-risk-for-bleeding/bleeding patient on extracorporeal membrane oxygenation requiring invasive procedure. Prior to procedure initiation, obtain bedside resuscitative blood products and medications including factor VIIa and ACA (aminocaporic acid). Discontinue bivalirudin 1 hour prior to procedure start time. At procedure start time, initiate both platelet and ACA infusions (20–30 mg/kg/hr) with consideration for an ACA loading dose (100 mg/kg). If hemorrhaging is encountered during or after the procedure, consider an ACA loading dose (if not previously given) and/or factor VIIa replacement. Moving forward, continue blood product resuscitation as needed with plan for continued ACA infusion post-procedurally for approximately 24 hours (or at least 6 hours if hemostasis achieved). If clinical evidence of thrombosis appears during this time period, consider overlapping the ACA infusion with a bivalirudin infusion. We recommend a conservative bivlairudin starting dose of 0.05 mg/kg/hr with conservative laboratory monitoring (partial thromboplastin time, 40–50 sec; Fibrinogen, >200 mg/dL; platelets, >50 K/cumm; hemoglobin, >10 g/dL; thromboelastography (TEG) reaction time, 8–10 min) with reliance on TEG to guide product resuscitation. ACA, aminocaporic acid.Management of the refractory SVT was challenging. Catheter ablation on ECMO was a reasonable approach even in such a complex patient. Safety profiles for ablation have been described in refractory cases for infants of this weight range and on ECMO support as early as 20 years ago.10 Here, ablation was performed on the patient having an open abdomen and recovering from distributive shock and intestinal resection. A recent multi-institutional review of the management of pediatric tachyarrhythmias on ECMO described the risk of ablation as no more than ablation without ECMO.11 In summary, early recognition, communication, and multidisciplinary shared understanding of the disease process were crucial in achieving a successful outcome. Critical care and cardiology subspecialty management of the underlying arrhythmia and cardiopulmonary dynamics, combined with early use of ECMO, created a safer environment to treat the arrhythmia (medically and transcatheter ablation) and to allow pediatric surgery to adequately approach surgical NEC. A thoughtful hemostatic strategy with aid from clinical pharmacists allowed for an outcome that may not have been possible under other circumstances.