Keywords
Thrombosis - heart - antithrombin
Introduction
Percutaneous mitral valve repair with the Mitraclip system (Abbott, Abbott Park, Illinois,
United States) has recently emerged as a safe and effective procedure to treat patients
with degenerative mitral regurgitation (MR) with high surgical risk and a technically
favorable anatomy. The Mitraclip is being extensively used also to treat functional
MR in patients with persisting heart failure symptoms despite optimal medical treatment
and, when indicated, cardiac resynchronization therapy (CRT).[1] The guiding catheter is advanced from the femoral vein to the left atrium through
a transseptal approach. Then, guided by transesophageal echocardiography (TEE), one
or more clips are placed on the mitral valve leaflets through a steerable delivery
system, to permanently approximate the leaflets and create a double valve orifice.
Interventional cardiology procedures require full anticoagulation to prevent thrombus
formation on catheters and devices with potential development of embolic complications.
The Mitraclip procedure implies the long permanence of bulky devices into the left
atrium, a chamber characterized by low blood flow velocities, especially in patients
with atrial fibrillation and a dilated atrium. Hence, thrombotic risk is high and
intraprocedural antithrombotic management must be very accurate. In general, an initial
bolus of heparin is given after transseptal puncture and activated clotting time (ACT)
is monitored to administer additional doses and maintain an ACT value greater than
250 seconds.
Heparin-induced thrombocytopenia (HIT) is a potentially catastrophic immune-mediated
complication of heparin caused by antibodies to complexes of platelet factor-4 (PF4)
and heparin.[2] HIT predisposes to thrombosis because platelets release microparticles that activate
thrombin. In patients undergoing percutaneous coronary intervention (PCI) with a diagnosis
of HIT, intraprocedural administration of bivalirudin is recommended.
We report the challenging management of a patient with HIT who underwent Mitraclip
procedure with bivalirudin infusion.
Case Presentation
An 81-year-old woman with dilated cardiomyopathy was referred to our institution for
urgent Mitraclip procedure because of severe MR and heart failure refractory to full
medical treatment, including high-dose intravenous (IV) diuretics. Echocardiography
showed severe left ventricular dilatation, with 20% ejection fraction, severe left
and right atrial enlargement, and severe functional mitral and tricuspid regurgitation
([Fig. 1]). Comorbidity included type 2 HIT, chronic hepatitis C virus infection, chronic
anemia, chronic kidney disease (CKD) with glomerular filtration rate 37 mL/min, previous
mammary cancer followed by radiotherapy (40 years before), and previous operated colon
adenocarcinoma (10 years before). Baseline activated partial thromboplastin time (aPTT),
international normalized ratio (INR), and fibrinogen were within normal limits. The
patient had received CRT-D 4 years earlier. Ongoing therapy: furosemide 500 mg/die IV, carvedilol 6.25 mg bid, potassium canrenoate 100 mg, iron
supplement, pantoprazole, levothyroxine, and levosimendan.
The Mitraclip procedure was conducted with standard technique. Saline solution without
heparin was used for catheters' flushing. We administered bivalirudin bolus (0.75
mg/kg IV) immediately after transseptal puncture and at the same time a PCI-dose infusion
of 1.40 mg/kg per hour was started (recommended dose reduction for CKD).
After release of the first clip, during positioning of a second clip to improve procedural
result, a moving image suggestive for thrombus became evident on top of the clip ([Fig. 1] and [Video 1]). ACT measured 240 seconds. An additional bivalirudin bolus of 0.3 mg/kg was administered
and the infusion rate increased to 1.75 mg/kg/hour. The mass did not reduce and actually
tended to increase. Hence, the clip was carefully removed to avoid embolization. Once
outside the catheter, a large thrombus attached to the clip was evident ([Fig. 1], panel D). Then, a line for additional bivalirudin continuous infusion (and flushing)
was placed directly on the guiding catheter, at a rate of 0.2 mg/kg/hour. ACT then
measured 455 seconds. The procedure was successfully completed with reduction of MR
grade from 4+ to 1+ . Bivalirudin infusion was stopped and the guiding catheter removed
without complications. No clinical signs of embolization appeared, although we cannot
rule out small silent systemic embolizations. The patient was discharged home after
1 week. Antithrombotic regimen was lifelong aspirin and clopidogrel for 6 months.
Fig. 1 Intraprocedural echocardiography and clip thrombosis. (A) Intraprocedural transesophageal echocardiography (TEE) color Doppler at the left
ventricular outflow tract view showing severe mitral regurgitation (MR). (B) TEE at the intercommissural view showing the extension of severe MR to the entire
mitral valve orifice. (C) Thrombus adherent to the clip during the maneuvers to orient the second clip toward
the mitral orifice. LA, left atrium; LV, left ventricle. (D) Picture of the clip and the thrombus once retrieved from the patient. (E) Double orifice of the mitral valve at 3D echocardiography after implantation of
the two clips. (F) Final procedural result with residual trivial mitral regurgitation, X-plane view.
Discussion
Bivalirudin is a direct thrombin inhibitor with short half-life which has been widely
used for anticoagulation in the setting of PCI, and represents the first-line intraprocedural
anticoagulant for patients with HIT. It is administered with a 0.75-mg/kg bolus, followed
by infusion of 1.75 mg/kg/hour in patients with GFR greater than 60 mL/min, reduced
to 1.40 mg/kg/min in patients with GFR 30 to 59 mL/min. An infusion at 0.25 mg/kg/hour
can be prolonged 4 to 12 hours if clinically indicated. ACT can be used to monitor
effective bivalirudin administration, but it is not routinely recommended. Safety
and efficacy of bivalirudin in structural heart disease procedures have been poorly
investigated. The Effect of Bivalirudin on Aortic Valve Intervention Outcomes-3 (BRAVO-3)
randomized trial demonstrated noninferiority of bivalirudin compared with heparin
in rates of major bleeding or net adverse cardiovascular events during transcatheter
aortic valve implantation.[3] Yet, flow in the left ventricular outflow tract is torrential and thrombotic risk
is supposedly lower than that in the left atrium. There are a few reports about successful
bivalirudin use during ablation procedures in patients with HIT.[4]
[5] To the best of our knowledge, this is the first report of bivalirudin use in a patient
undergoing a Mitraclip procedure.
Despite the use of recommended bivalirudin dose, our patient developed a large thrombus
on the clip during the procedure. ACT was measured only at that time and was actually
lower than expected. The procedure was successfully completed giving an additional
0.3 mg/kg bolus, as per indications for use, and arbitrary continuous catheter flushing
with a low bivalirudin dose.
Limitations of ACT monitoring of bivalirudin therapy have been previously demonstrated,[6] and there is no guidance regarding what value of ACT would warrant an additional
bolus.
Accordingly, routine monitoring of ACT during PCI is not mandatory and is not usually
performed. However, the instructions for use suggest to check ACT 5 minutes after
the bolus and to administer additional 0.3 mg/kg bolus dose “if needed.” In our case,
this was necessary to achieve effective thromboprophylaxis. We also increased infusion
rate and continuously flushed the guiding catheter with low-dose bivalirudin.
Differently from heparin, bivalirudin binds directly to thrombin without the need
of a cofactor and therefore exhibits predictable and dose-dependent anticoagulation.
While bivalirudin activity is independent of antithrombin deficiency, increased heparin
clearance, and increased heparin binding proteins, which are among the causes of heparin
resistance, other resistance mechanisms such as heparin have been hypothesized, including
elevated factor VIII and fibrinogen.[7] Furthermore, because bivalirudin inhibits both free and clot-bound thrombin, in
the presence of large clot burden higher dosing might be required because of the larger
number of binding sites.[7] In our patient, fibrinogen was normal and there was no clinical evidence of vascular
thrombosis. Additional coagulation tests were not performed; so, we cannot rule out
abnormality of factor VIII level or the presence of unrecognized vascular thrombosis.
Interestingly, bivalirudin resistance (defined as ACT < 300 seconds) was reported
in 2.4% of patients undergoing PCI in a single-center study.[8] While bleeding complications were not more common in bivalirudin hyperresponders
(ACT > 800 seconds), thrombotic complications were numerically higher in bivalirudin
hyporesponders.[8] Overall, the therapeutic window seems wide enough to allow additional bivalirudin
doses in suspected bivalirudin hyporesponders, based on ACT values. Remarkably, however,
monitoring of direct thrombin inhibitor's efficacy would require more sensitive markers,
such as the chromogenic anti-Xa level for heparin, dilute thrombin time, Ecarin thrombin
time, and specific chromogenic substrate-based assays, although clinical applicability
is limited because of the lack of widely available commercial products.[7]
Conclusion
This case suggests that bivalirudin may be used in place of heparin in patients with
HIT undergoing Mitraclip procedure. However, bivalirudin might be subject to resistance
mechanisms similar to those previously described in patients receiving heparin, and
careful ACT monitoring should be mandatory to verify achievement and permanence in
the therapeutic range. When possible, additional coagulation tests should be used.
Flushing of guiding catheters and devices with low-dose bivalirudin may represent
a useful adjunctive measure. More data are needed to clarify if different ways of
monitoring bivalirudin therapy are required for transcatheter structural heart disease
interventions. Importantly, this is a single clinical case and our observations should
be interpreted cautiously.
