EMERGENCY MANAGEMENT of acute gastrointestinal (GI) bleeding in patients who reject allogeneic transfusion or blood products, such as Jehovah's Witnesses, presents clinicians with significant challenges. The clinical picture is further complicated when these patients have prescribed anticoagulants, as reversibility with established interventions, such as four-factor prothrombin complex concentrate (4F-PCC), recombinant factor VIIa, or fresh frozen plasma, is not feasible due to, but not limited to, patients' preference (Berend & Levi, 2009; Cheung & Leung, 2017).
Gastrointestinal bleeding is a serious complication of anticoagulation. Despite a favorable safety and efficacy profile, predictable dosing, and minimal monitoring requirements, direct oral anticoagulants, including dabigatran, apixaban, rivaroxaban, and edoxaban, carry a risk of bleeding (Cheung & Leung, 2017). Additional risk factors include advanced age, history of peptic ulcer disease, non-steroidal anti-inflammatory drugs, antiplatelet agents, and corticosteroids (Drini, 2017). The incidence of GI bleeding in patients older than 75 years is estimated to be roughly 375 patients per 100,000 per year, with mortality rates ranging between 5% and 14% (Barkun et al., 2010; Laine, Yang, Chang, & Datto, 2018). Typical management strategies for life-threatening GI bleeding include fluid resuscitation, blood transfusions, and reversal of anticoagulants (Christos & Naples, 2016). A significant challenge exists in managing patients when religious or personal beliefs preclude the use of blood products (Brooks, 2012).
Tranexamic acid (TXA) is a hemostatic agent traditionally used in cardiac surgeries. It is a synthetic derivative of the amino acid lysine that exerts its antifibrinolytic effect through the reversible blockade of lysine-binding sites on plasminogen molecules. As such, it is not a blood-derived product and is a treatment option for patients refusing allogeneic transfusions or blood products (Dunn & Goa, 1999). Herein, we present the case of a Jehovah's Witness with a life-threatening GI bleed complicated by hemodynamic instability who refused conventional therapy but was treated successfully with TXA.
CASE REPORT
An 80-year-old female Jehovah's Witness patient presented to the emergency department (ED) with rectal bleeding and a 2-day history of tarry stools. The patient reported that her bleeding started approximately 8 hr before the presentation to the ED. She denied abdominal pain, chest pain, or shortness of breath, as well as feeling lightheaded or dizzy. Her past medical history was significant for heart failure with reduced ejection fraction (ejection fraction 30%), peripheral vascular disease (PVD), hypertension (HTN), chronic kidney disease (Stage 3), severe aortic stenosis, coronary artery disease, dysphagia, and recent acute lower extremity deep vein thrombosis (DVT) 4 weeks prior to admission. Five weeks prior to presentation, a colonoscopy was performed, and six polyps were removed, including one in the duodenum. Her medication history included apixaban 5 mg twice daily (the last dose was taken approximately 12 hr before arrival at the ED), aspirin 81 mg daily, and clopidogrel 75 mg daily.
Upon arrival to the ED, the patient had systolic blood pressure (SBP) of 112 mmHg and diastolic blood pressure (DBP) of 89 mmHg, heart rate of 102 beats per minute, afebrile (99 [degrees]F), oxygen saturation of 100% on room air, respiratory rate of 15 breaths per minute, and a mean arterial pressure (MAP) of 96 mmHg. She received 1.5 L of 0.9% sodium chloride. Within 5 hr of arrival, she became progressively unstable, and her MAP dropped to a nadir of 63 mmHg (SBP of 70 mmHg and DBP of 59 mmHg), so she was started on norepinephrine 0.01 mcg/kg/min (weight: 81.8 kg), requiring titration to a maximum dose of 0.3 mcg/kg/min (a goal MAP of 65 mmHg). Other initial laboratory parameters were as follows: prothrombin time 35.5 s, international normalized ratio 3.0, serum creatinine 1.13 mg/dl, and lactate 1.4 mmol/L. Her complete blood count was as follows: white blood cells count 9.1 x 109/L, hemoglobin 8.8 g/dl, hematocrit 25.6%, and platelets 293 x 109/L.
As the patient became progressively more unstable, the reversal of anticoagulation was deemed necessary. Some of the options considered were 4F-PCC (derived from human blood); andexanet alfa (derived from human proteins); recombinant factor VIIa (produced in baby hamster kidney cell lines and free of human protein; arterial and venous thrombotic events reported); fibrinogen concentrate (made from pooled human plasma); desmopressin (a synthetic analog of the natural hormone arginine vasopressin, although sometimes used off-label in patients with moderate to severe renal impairment to treat hematuria was not selected because of patient's history of heart failure and heightened risk of renal water retention); fibrinogen concentrate (made from pooled human plasma), and cryoprecipitate (produced from human plasma). After extensive discussions with the patient, she refused any blood products, so TXA was administered (1 g intravenously [IV] for more than 10 min, followed by 1 g IV for more than 8 hr). Phytonadione (vitamin K1) 10 mg IV was also administered to stimulate coagulation factor production. She was admitted to the medical intensive care unit and GI was consulted. Six hours following the pharmacological treatment with TXA, the patient was stabilized hemodynamically (MAP of 78 mmHg); norepinephrine drip was turned off, and the patient was able to undergo esophagogastroduodenoscopy, which did not show any signs of bleeding. On the next day, her hemoglobin improved to 8.9 g/dl (the nadir before the pharmacological interventions including TXA was 7.6 g/dl) but the planned colonoscopy had to be postponed as the patient developed atrial fibrillation with a rapid ventricular response (see Figure 1). The patient's CHA2DS2VASC score was 8 (age, sex, CHF, HTN, DVT, PVD), and the HAS-BLED (Hypertension, Abnormal renal/liver function, Stroke, Bleeding history or predisposition, Labile international normalized ratio, Elderly [>65 years], Drugs/alcohol concomitantly) score was 5 (bleeding risk around 9%; Lip et al., 2010). Based on these scores, the patient met the criteria for anticoagulation, but with such a high HAS-BLED score, a careful assessment of risk versus benefit was warranted. After 1 day, the patient was sufficiently stable to transfer to the general floor. A heparin drip was initiated for atrial fibrillation and a history of DVT. A week into her hospital stay, a colonoscopy was completed, and it noted diverticulosis and hemorrhoids. The patient was transitioned off heparin drip back to her home regimen of apixaban 5 mg twice daily and was discharged on hospital Day 8 with a plan for capsule endoscopy as an outpatient. Upon discharge, home medications were restarted except clopidogrel, aspirin, levothyroxine, lisinopril, and metoprolol succinate.
DISCUSSION
This case describes an 80-year-old female Jehovah's Witness patient, anticoagulated with a direct factor Xa inhibitor, who presented with acute life-threatening GI bleeding and was successfully treated with TXA and vitamin K. The estimated Jehovah's Witnesses population accounts for just less than 1% of U.S. adults, with approximately 1,000 resultant deaths annually due to the prohibition of blood transfusions and blood-related products (Bock, 2012; Robb, 2005). As this unique group of patients is fast-growing, medical practitioners should be prepared for the management of acute life-threatening bleeding (Robb, 2006). The management of severe bleeding in this population should allow patients enough time to recover both hemoglobin and coagulation factors. But even within this population, their views on the matter of blood products are not entirely homogenous, so a careful patient-specific discussion should precede any interventions. In that respect, the use of TXA makes sense mechanistically. Nonallogeneic pharmacological interventions should be discussed and decided upon on a case-by-case basis using shared decision making and may include antifibrinolysis, erythropoiesis, and stimulating blood factor production, in addition to maintaining adequate organ perfusion, controlling blood loss, and maximizing hemostasis (Nishant & Kumari, 2014; Zeybek et al., 2016).
Tranexamic acid is an antifibrinolytic, which inhibits both plasminogen activation and plasmin activity, thus preventing clot breakdown rather than promoting new clot formation (Astedt, 1987). Although it has been used around the world to control bleeding since the 1960s safely, recent literature has explored its role in the treatment of life-threatening bleeding related to trauma, postpartum hemorrhage, and other bleeding conditions such as hemoptysis, epistaxis, and GI bleeding (CRASH-2 Collaborators et al., 2011; Ker, Beecher, & Roberts, 2013; Komura, Rodriguez, & Peabody, 2018; Prutsky, Domecq, Salazar, & Accinelli, 2016). The Clinical Randomization of an Antifibrinolytic in Significant Hemorrhage 2 (CRASH-2) trial enrolled more than 20,000 patients with trauma-related known or suspected hemorrhage and found that early administration of TXA was associated with a reduction in mortality at 4 weeks postinjury (14.5% vs. 16%, p value less than .01). This was most apparent within 3 hr of injury, emphasizing the importance of time from injury to TXA administration (Williams-Johnson, McDonald, Strachan, & Williams, 2010). Similar results were found in a study including combat-related trauma patients, where TXA was shown to reduce all-cause mortality (Morrison, Dubose, Rasmussen, & Midwinter, 2012). In addition to aiding in hemostasis, TXA mortality benefit has been attributed to a potential anti-inflammatory effect, and the improved coagulation profile of trauma patients receiving TXA, regardless of blood products used, supports its additional clot-stabilizing effect (Morrison et al., 2012; Williams-Johnson et al., 2010). Although the optimal effective dose remains to be determined, a dose of 500/1,000 mg every 6 hr IV, or 10 mg/kg 3-4 times per day, up to a maximum rate of 100 mg/min, has been suggested (Berend & Levi, 2009). Considering the complexity of our case, we administered a loading dose of 1 g for more than 10 min, followed by an infusion of 1 g for more than 8 hr, which has been reported to be safe and effective in trauma patients. To that end, it also has been hypothesized that TXA might have decreased efficacy and possible harm when given late, perhaps due to the prevalence of hypothermia and acidosis after the initial phase of traumatic injury (CRASH-2 Collaborators et al., 2011; Williams-Johnson et al., 2010).
The use of phytonadione (vitamin K) is not common in the setting of bleeding secondary to apixaban. Furthermore, phytonadione does not reverse the effect of apixaban per se but, in this case, was used to support the intrinsic production of K-dependent clotting factors and to ensure cofactor availability for the normal clotting processes thereafter.
After parenteral administration of phytonadione, increased concentrations of blood coagulation factors (K-dependent clotting factors II, VII, IX, and X, as well as regulatory proteins C and S) are evident within 1-2 hr, and hemorrhage is typically controlled within 3-6 hr. A normal prothrombin level may often be achieved in 12-14 hr (Hunt & Levi, 2018).
It is worth mentioning that in addition to the refusal of the patient to accept any blood-derived products and recombinant factor VIIa, the clinical picture was further complicated by the fact that she was on a direct factor Xa inhibitor as well as dual antiplatelet therapy (DAPT) with clopidogrel and aspirin. A good choice for reversing the apixaban effect on the coagulation cascade would have been the off-label use of 4F-PCC, had it not been the patient's wish not to pursue this agent. This is one of the reasons why we pursued the alternative of using TXA as salvage therapy. In addition, a small observational study (n = 40) explored the effect of TXA on platelet function in patients on DAPT undergoing cardiac surgery (Weber et al., 2011). To that end, the results indicated a potential benefit of TXA in correcting defects in arachidonic acid-induced and ADP-induced platelet aggregation imposed by antiplatelet therapy. Further studies are needed to explore this clinical use of TXA.
Tranexamic acid has also demonstrated a reduction in mortality due to bleeding in women with postpartum hemorrhage or cesarean delivery with no adverse effects, including thromboembolism. Similar to traumatic injury, the effect was stronger in women who received TXA within 3 hr of delivery (Morrison et al., 2012; Williams-Johnson et al., 2010). Successful management of refractory bleeding with TXA has also been reported in a liver failure patient with a subdural hematoma (Kodali et al., 2019). Additional utility for TXA has been demonstrated in the setting of hemoptysis and epistaxis. A small randomized placebo-controlled trial including 47 patients with hemoptysis showed significant resolution of hemoptysis within 5 days of admission in almost all patients treated with nebulized TXA (Wand et al., 2018). Furthermore, when used for epistaxis, TXA has been shown to reduce rates of rebleeding (Gottlieb, DeMott, & Peksa, 2019). Overall, TXA has many applications for facilitating hemostasis in the context of treating life-threatening bleeding.
Several studies have also acknowledged the usefulness of TXA in the setting of GI bleed. In a systematic review, compared with placebo, TXA reduced mortality in patients with upper GI bleeding (relative risk = 0.60, 95% confidence interval [CI]: 0.42-0.87; Bennett, Klingenberg, Langholz, & Gluud, 2014). However, in another clinical trial of 100 patients, no difference was found in terms of blood loss when using a drop in hemoglobin levels as a surrogate marker in patients with lower GI bleeding (Smith et al., 2018). Trending hemoglobin levels should be taken with a grain of salt as fluctuations could be affected by many factors. For example, the original drop in the patient's hemoglobin could be considered dilutional in nature because she received 1.5 L of normal saline before it was decided to resort to TXA. Considering the whole presentation of the case and the ostensible evidence of acute bleeding, a more likely explanation for the drop and the resultant improvement in the patient's hemoglobin could be the stabilization of the actual bleeding. Perhaps this could be a corollary of several factors. In a meta-analysis looking at the use of TXA to reduce the blood loss after total-knee arthroplasty (TKA), it was indicated that oral TXA was able to effectively reduce total blood loss (weighted mean difference [WMD] = -159.53, 95% CI: -256.95 to 62.11, p = .001) and hemoglobin decline after TKA (WMD = -0.492, 95% CI: -0.696 to -0.288, p = .000; Li, Bai, Li, & Fang, 2018).
Although there are data suggesting TXA utility in various patient populations with GI bleeds, no previous reports are addressing its use in the setting of a blood product-restrictive approach (Cap et al., 2011; Sabovic, Lavre, & Vujkovac, 2003). The HALT-IT trial is a large randomized controlled trial investigating TXA use for acute upper or lower GI bleeding that has been enrolling patients since 2012, which hopefully will bring more clarity to this important issue. Its estimated completion date was set for March 2019, but the study is still recruiting patients (Roberts et al., 2014). A recent study of 410 patients with acute GI bleed has shown that TXA use augmented the need for endoscopy from urgent to elective (Tavakoli et al., 2018). In the setting of limited published research, this case adds to the literature supporting the use of TXA for acute GI bleeds. In the presented patient case, TXA was administered in combination with vitamin K with no adverse events reported, which facilitated hemodynamic stabilization and ultimate recovery to baseline. Timely administration of TXA offers an alternative to blood products for patients with life-threatening GI bleeding.
Multidisciplinary teams have shown to improve patients' outcomes by streamlining the patient care process and often include nurses, nurse practitioners (NPS), physical therapists, social workers, anesthesiologists, attending physicians, and pharmacists. Advanced practice providers, such as NPS, have become an integral part of the complete patient care process (Hickey et al., 2016). They are well equipped to practice both independently and as extenders in a variety of situations and to collaborate on a multitude of levels. Pharmacists, on the other hand, have established their role in the patient care process as medication experts shortening the time to administration of medications such as analgesics, sedatives, and antibiotics (Amini et al., 2013). In this case, the timely administration of TXA could not have been achieved without the common effort from all members of the multidisciplinary team.
CONCLUSION
This case describes an 80-year-old female Jehovah's Witness patient, anticoagulated with apixaban, who presented with acute life-threatening GI bleeding and was successfully treated with TXA. Given the risk of morbidity and mortality in life-threatening GI bleeding, TXA should be considered when a blood product-restrictive approach is needed to manage life-threatening GI bleeding.
REFERENCES