Keywords
complication - myocardial ischemia - hematoma - coronary artery bypass grafting -
CABG
Introduction
Anastomosis-related complications after coronary artery bypass grafting (CABG) usually
comprise anastomosis narrowing and leakage. Narrowing of an anastomosis can impair
bypass function and run-off into the target vessel and is described in as many as
34% of patients undergoing reangiography due to symptomatic graft failure up to 1 year
after CABG.[1] Anastomosis leakage occurs in 2 to 6% of CABG procedures, manifests in the immediate
perioperative period, and results in acute bleeding, pericardial tamponade, and the
need for re-exploration.[2] The healing process of coronary anastomoses, which includes formation of a neointima
and the covering by an epicardial layer, is completed after 30 to 50 days.[3] Thereafter, spontaneous bleeding from the anastomosis is unlikely; should it occur,
however, it would result in a subepicardial hematoma. Formation of a subepicardial
hematoma following perforation of a bypass graft during percutaneous coronary intervention
of the graft has been described in two cases.[4]
[5] Spontaneous subepicardial hematoma due to delayed anastomosis leakage has not been
described thus far. Here, we present a case of the formation of an anastomosis-related
subepicardial hematoma 7 weeks after CABG that compromised the downstream coronary
artery and resulted in myocardial ischemia and acute coronary syndrome.
Case Description
A 63-year-old male patient underwent aortic valve replacement with a mechanical valve
prosthesis (ATS Medical, Open Pivot 29 mm) for severe aortic regurgitation and CABG
(left internal mammary artery [LIMA] to obtuse marginal branch 2 (OM2). Total postoperative
pericardial and pleural drainage volume was 350 mL. Due to intermittent atrioventricular
block III postoperatively, a permanent dual chamber pacemaker was implanted on postoperative
day (POD) 5. After an uneventful postoperative period, the patient was discharged
to cardiovascular rehabilitation on POD 10 in functional New York Heart Association
class I without any signs of angina pectoris. He received phenprocoumon with a target
international normalized ratio (INR) of 2.0 to 2.5 and acetylsalicylic acid (100 mg/d).
Seven weeks postoperatively, the patient was readmitted to our hospital due to new-onset
angina pectoris Canadian Cardiovascular Society IV. The electrocardiogram at admission
showed no abnormalities. Initial serum testing for troponin-I was negative (0.02 µg/L;
upper reference limit 0.05 µg/L) but became slightly elevated after 3 hours (0.32
µg/L). Transthoracic echocardiography confirmed regular function of the aortic valve
prosthesis, and regional wall motion abnormalities were ruled out. Coronary angiography
showed no new relevant stenoses in the native coronary system. However, a leakage
of contrast medium from the LIMA-OM2 anastomosis into a heretofore indefinable spherical
paraventricular/para-arterial space and thrombolysis in myocardial infarction (TIMI)
1 flow in the distal OM2 was documented. The reduced coronary flow in OM2 was attributed
to impaired inflow due to anastomosis leakage ([Fig. 1]). The etiology and anatomic architecture of this spherical mass was not distinguishable
by angiography alone. Since a connection to the heart chambers was suspected, cardiac
computed tomography (CT) was performed. CT showed a hematoma (measuring 5.5 × 3.1 × 5.0
cm) connected to the LIMA-OM2 anastomosis and thereby compressing the OM2 by several
centimeters. Any further connection to either cardiac cavities or the pericardium
was excluded. The hematoma appeared to localize subepicardially and not intrapericardially
([Fig. 2]).
Fig. 1 Coronary angiogram (A) RAO view of native coronary vessels. The OM2 shows a long-segment discontinuity
(black arrow) around the bypass anastomosis. (B) RAO view of the LIMA-OM2 bypass: from the anastomosis, there is retrograde flow
into the OM2, TIMI 2 flow into the distal OM2, and a large spherical space is contrast
enhanced. (C) LAO view of the LIMA-OM2 bypass anastomosis shows an apparent leak of contrast medium
at the region of the anastomosis into a paraventricular spherical space (white arrow).
LAO, left anterior oblique; LIMA, left internal mammary artery; OM2, obtuse marginal
branch 2; RAO, right anterior oblique.
Fig. 2 Multiplanar reconstruction of the cardiac CT (ECG-triggered mid-diastolic recording).
(A) Two-chamber view of the left ventricle and the left atrium. (B) Four-chamber view. (C) Short-axis view. Imaging shows a spherical space (white arrows) adjacent to the
basal lateral wall of the left ventricle (diameter 5.5 × 3.1 × 5.0 cm). There is slight
contrast medium enhancement in this space. The space can be anatomically localized
outside the myocardium and appears not to be connected to the pericardium (asterisks),
thus being bordered by the epicardium. CT, computed tomography; ECG, electrocardiogram.
As the patient suffered from angina pectoris and myocardial ischemia of the distal
OM2 territory, the heart team decided to perform surgical re-exploration to remove
the hematoma and restore the integrity of the anastomosis. During the course of the
operation, we observed that the intrapericardial space was well cicatrized and free
from any hematoma. As excessive scar tissue was present at this time point after the
index operation and the localization of the hematoma required extensive mobilization
of the heart, extracorporeal circulation was used. The subepicardial hematoma was
observed to be without connection to the pericardial space ([Fig. 3A]) at the lateral cardiac wall. After evacuation of the hematoma, bleeding from a
small anastomosis suture leak was identified and closed with a single suture by adapting
the adventitial tissue (Prolene 8/0) ([Fig. 3C, D]). As we observed some diffuse bleeding from the preparation region around the anastomosis,
a hemostatic patch (Veriset) was applied ([Fig. 3E]). The epicardial layer that had contained the hematoma was left open. Transit-time
flow measurement of the bypass revealed appropriate flow (19 mL/min).
Fig. 3 Surgical removal of the subepicardial hematoma and repair of the LIMA-OM2 bypass
anastomosis. View into the operation site from caudal-anterior perspective (course
of the bypass in (c)—(e): white asterisks). (A) The patient is on cardiopulmonary bypass and the heart elevated to the right. The
exposed hematoma on the left lateral wall is located subepicardially without connection
to the pericardial space. (B) After opening the epicardial cover of the hematoma, a partially organized clot is
removed. (C) Exposure of the anastomosis: a small defect with pulsating bleeding is identified.
(D) The defect is repaired with a single suture. (E) To further secure the site of anastomosis, a hemostatic patch (Veriset) is applied.
(F) Removed hematoma mass. LIMA, left internal mammary artery; OM2, obtuse marginal
branch 2.
Control coronary angiography performed on POD 4 showed a good function of the LIMA-OM2
bypass and TIMI 3 flow with a decompressed lumen of the OM2 ([Fig. 4]). On POD 7, a pericardial effusion (400 mL; exudative, hemoglobin 1.3 g/dL) was
removed by ultrasound-guided pericardiocentesis and immunosuppressive therapy with
prednisolone and ibuprofen for postcardiotomy syndrome was initiated. The patient
recovered well subsequently and was discharged from the hospital on POD 15.
Fig. 4 Coronary angiogram on POD 4. (A) The native left circumflex artery shows the previously known stenosis of OM2. The
OM2 distal to the bypass anastomosis shows no narrowing. (B) The LIMA-OM2 bypass shows good function without leakage of contrast medium and TIMI
3 flow in the distal OM2. LIMA, left internal mammary artery; OM2, obtuse marginal
branch 2; POD, postoperative day; TIMI, thrombolysis in myocardial infarction classification.
Discussion
To our knowledge, this is the first description of a spontaneous secondary subepicardial
hematoma during the very late postoperative period after CABG. We suspect bleeding
to have started after primary healing of the anastomosis was completed, resulting
in an entirely encapsulated hematoma within the epicardial layer. Previously published
case reports referred to subepicardial hematomas after attempted coronary intervention
in bypass grafts or anastomoses.[4]
[5]
[6] Furthermore, usage of heart positioning devices for off-pump CABG can induce epicardial
trauma and hematomas.[7] Another potential pathomechanism includes trauma with sudden tension leading to
dissection of the anastomosis. None of these effectors was relevant to the current
case. As the patient was set on vitamin K antagonists and acetylsalicylic acid, potentially
excessive anticoagulation could have induced ongoing bleeding. However, from the patients'
records with INR controls twice weekly, no phases of hyperanticoagulation are evident.
However, intermittent hyperanticoagulation cannot be excluded and bleeding is likely
to be supported by the anticoagulation therapy.
The exact localization of the hematoma was not possible using coronary angiography
alone, but CT provided excellent accuracy.
Due to persistent symptoms and myocardial ischemia, treatment options included interventional
or surgical therapy. We considered treatment with percutaneous coronary intervention
of the native OM2 and application of a polytetrafluoroethylene (PTFE)-covered coronary
stent within the region of the LIMA-OM2 anastomosis. Several factors, however, spoke
against using this method: first, the hematoma-induced compression of the distal OM2
would not have been resolved; second, the bleeding from the anastomosis would not
definitively have been stopped; and third, a high risk of perforation of the anastomosis
with consecutive severe bleeding would have been expected. In addition, long-term
patency of PTFE-covered coronary stents might be inferior as restenosis rates of up
to 21% 10 months after treatment of coronary artery aneurysms have been reported.[8] In the treatment of saphenous vein graft stenosis, the use of PTFE-covered stents
compared with bare metal stents led to a higher rate of restenosis (68.3 vs. 51.8%,
p = 0.007) after 5 years.[9] Although the surgical approach is more invasive, the complete removal of the hematoma
was achievable and relief of hematoma-induced compression of the distal coronary artery
was provided. Furthermore, direct exploration and sophisticated repair of the anastomosis
can be safely performed using a surgical approach with reasonable risk profiles. Due
to the location of the hematoma and excessive pericardial adhesions, an on-pump approach
was chosen during the reoperation. An off-pump strategy would have been considerable
in situations with easier exposure of the hematoma.
In summary, we present for the first time a case of a rare complication after CABG
in the late postoperative period. Efficient diagnostic and therapeutic management
was achieved by a heart team approach involving interventional cardiologists, cardiac
imaging specialists, and cardiac surgeons.