Prophylactic and perioperative replacement therapy for acquired factor XIII deficiency

Abstract

Acquired factor (F)XIII deficiency is a rare coagulation disorder with limited information on using laboratory assays to guide management. We encountered unique monitoring and therapeutic challenges while managing a 73-year-old man with acquired FXIII deficiency. He presented with an extensive leg hematoma after receiving low-molecular-weight heparin for a suspected calf vein thrombosis. Two years earlier, he had developed a leg hematoma while on warfarin, which was discontinued. His medical history included inactive rheumatoid arthritis, stable angina, and paroxysmal atrial fibrillation on amiodarone. There was no family or prior bleeding history. He had previously undergone tonsillectomy, cholecystectomy and dental extractions without incident. Coagulation tests indicated an isolated deficiency of FXIII, with no FXIII activity detectable by the urea clot solubility assay [1, 2] or by Laurell rocket immunoassays [3], undetectable FXIII subunit A (FXIIIA) antigen [< 0.05 U mL−1, reference interval (RI) 0.67–1.39 U mL−1] and modestly reduced FXIII subunit B antigen (0.43 U mL−1, RI 0.73–1.17 U mL−1). A FXIII inhibitor could not be detected with the urea clot solubility assay, but a low-titer inhibitor [< 2 Bethesda Units (BU) by the Nijmegen procedure [4]] was subsequently detected with the Berichrom assay (Dade Behring, Mississauga, ON, Canada). The patient declined immunosuppressive therapy, was given a dose of FXIII concentrate (Fibrogammin-P, Aventis Behring, Ottawa, ON, Canada; 1250 U, 21 U kg−1; recommended prophylactic therapy for congenital FXIII deficiency, 10 U kg−1 every 4 weeks) and 3 weeks later he had a spontaneous iliopsoas hemorrhage that resolved with increased FXIII replacement (1250 U daily for 5 days). He was noted to have accelerated clearance of infused FXIII as 96 h after receiving 1250 U FXIII he had no FXIII detectable by urea clot solubility and FXIIIA antigen assays. He had a few minor bleeds until his replacement was increased to ensure residual FXIII activity, detectable by the urea clot solubility assay, in predose samples (2500 U, 42 U kg−1, twice weekly). A portacath was placed for home replacement therapy. Amiodarone was discontinued due to concerns about a drug-induced FXIII inhibitor. Six months later, the patient developed increasing angina. Angiography revealed a 90% stenosis of the left anterior descending coronary artery and a single coronary artery bypass graft was recommended. His FXIII inhibitor titer remained low (< 2 BU) and preoperative FXIII survival studies confirmed accelerated FXIII clearance, with more rapid decay of FXIII activity compared with FXIIIA antigen (Fig. 1A). Bypass surgery was done using off-pump, beating heart surgery to minimize risks of a postoperative coagulopathy. Perioperatively, FXIII levels were monitored using the Berichrom assay, as no other rapid commercial assays were available. He was given 2500 U FXIII (42 U kg−1) immediately before and after surgery, and again on postoperative days 1–3, 5–7 and 9, before resuming his usual prophylaxis (2500 U twice weekly), which normalized his FXIII activity during surgery and maintained levels above 0.40 U mL−1 for 7 days postoperatively (Fig. 1B). No abnormal bleeding occurred and he recovered fully without complications. Over the subsequent 38 months, his FXIII inhibitor has remained < 2 BU and he has been maintained on lower doses of FXIII prophylaxis and aspirin 80 mg daily, without further angina or serious bleeding. He had a few minor bleeds when his FXIII prophylaxis was reduced to the point that no predose FXIII activity was detectable by urea clot solubility assays, but this resolved when his prophylaxis was adjusted to provide residual FXIII activity (1750 U, 29 U kg−1 weekly). Factor (F)XIII activity and FXIIIA antigen levels in a patient with acquired FXIII deficiency, illustrating discrepancies between FXIII activity (Berichrom assay) and FXIIIA antigen levels, and accelerated FXIII clearance while on prophylactic (A; following infusion of 1250 U at time 0) and perioperative replacement therapy (B) for coronary artery bypass surgery. In B, levels are shown on the day of surgery (baseline, preop, postop) and on postoperative days 1–3, 5 and 7, and arrows indicate when FXIII replacement (2500 U, indicated by arrows) was given. Reference ranges: FXIII activity: 0.60–1.69 U mL−1; FXIIIA antigen: 0.67–1.39 U mL−1. The measured ratio of FXIII activity/FXIIIA antigen for the FXIII (Fibrogammin) concentrate, preinfusion, was 0.5. FXIII is important for stabilizing fibrin and making clots less susceptible to plasmin lysis [5-7]. Inhibitors arising from replacement therapy in congenital FXIII deficiency are rare, and inhibitors causing acquired FXIII deficiency are 10 times as rare as congenital FXIII deficiency [8]. Acquired FXIII deficiency has been reported with autoimmune disorders [9], and medications including phenytoin [10], isoniazid [11, 12], penicillin [13] and procainamide [14]. FXIII inhibitors complicating amiodarone therapy have not been reported, and in our patient with inactive rheumatoid arthritis, the inhibitor persisted after discontinuing amiodarone. Presently, there are no guidelines on how to adjust and monitor prophylactic and perioperative therapy for FXIII deficiency. In clinical laboratories, FXIII function is commonly evaluated by clot solubility tests and the Berichrom assay [15], and clot solubility assays have the disadvantage of being able to detect only severe FXIII deficiency (0–3%) and longer turnaround times [16]. We had success using the Berichrom assay to monitor our patient during bypass surgery and achieved excellent hemostasis with FXIII replacement, adjusted to maintain normal FXIII levels (>0.60 U mL−1) during surgery and > 0.40 U mL−1 for the next 7 days. Multiple assays were used to follow our patient's response to prophylaxis and we observed no bleeding when he had basal (preprophylactic dose) FXIIIA activity detectable by the urea clot solubility assay. Unexpectedly, the Berichrom assay consistently detected FXIII activity (0.08–0.14 U mL−1) in his plasma when no FXIII was detectable by urea clot solubility and FXIIIA antigen assays, and it similarly detected FXIII activity (0.11–0.14 U mL−1) in severe congenital FXIII-deficient plasma that had undetectable FXIII in the other assays (n = 3 individuals; confirmed data on two samples, drawn before prophylaxis). As similar problems were identified in a recent proficiency survey [15], the Berichrom assay may not be as accurate in the low range as previously implied [17]. Until more is known about the low-range accuracy of the Berichrom, and other quantitative FXIII assays, it may be prudent to perform these assays in conjunction with urea clot solubility assays. Before FXIII replacement therapy, many individuals with congenital FXIII deficiency died from intracranial hemorrhage [18]. Although acquired FXIII deficiency is often associated with a very poor prognosis [8], our patient has done well on long-term (57 months) dose-adjusted prophylactic and perioperative FXIII replacement. Nonetheless, more information is needed on the optimal ways of managing FXIII deficiency, with or without inhibitors. The contributions of the specialized coagulation staff of the Hamilton Regional Laboratory Medicine Program to providing data are gratefully acknowledged.