Case presentation

An active 76-year-old man underwent elective coronary angiography for evaluation of a several-month history of exertional chest pain. This identified a severe stenosis in the midportion of a large caliber right coronary artery and a normal appearing left coronary artery. He was referred for percutaneous coronary intervention of the right coronary artery. Past medical history was notable only for hypertension, gout, and paroxysmal atrial fibrillation, and his medications included aspirin, atenolol, atorvastatin, and amlodipine. Physical examination was unremarkable, and his baseline electrocardiogram was notable for sinus rhythm and bifascicular block (right bundle branch and left anterior fascicular block).

Cardiac catheterization

He was administered 300 mg of clopidogrel as a loading dose prior to the intervention, and procedural anticoagulation was accomplished with a bolus of 6000 units of heparin, followed by eptifibatide bolus and infusion. An activated clotting time measured 230 seconds prior to intervention. An 8 French, Judkins 4.0 guide catheter was positioned in the right coronary ostium and a 0.014 inch floppy-tipped guidewire placed distally in the posterior descending artery. After predilating the stenosis with a 4.0 mm diameter by 15 mm long balloon, a 5.0 mm diameter by 24 mm bare-metal stent was positioned and successfully deployed and postdilated with a 5.0 mm diameter by 22 mm long semicompliant balloon. The operator achieved an excellent angiographic result with no residual stenosis and no apparent angiographic complications. The patient felt well and was hemodynamically stable and transferred to a recovery room adjacent to the catheterization laboratory with the arterial sheath in place to await his hospital bed.

Shortly after arrival in the recovery room, the patient developed diaphoresis and nausea. His blood pressure fell to 52/30 mmHg with a heart rate of 60 beats per minute. Physical exam demonstrated marked elevation of his jugular veins. The monitor leads (lead II) showed no ST-segment abnormalities. An infusion of normal saline was initiated and additional venous access obtained in order to infuse dopamine. An emergent echocardiogram confirmed a large circumferential pericardial effusion and the patient was immediately transported back to the cardiac catheterization laboratory. The patient was given 30 mg of protamine intravenously and the eptifibatide infusion was discontinued. The operator performed an emergent pericardiocentesis and aspirated a large volume of blood from the pericardial space. This stabilized the patient’s hemodynamic status, with a rise in systolic blood pressure to 100 mmHg. The pigtail catheter remained in the pericardial space and blood was continually aspirated. Meanwhile, a blood sample was collected and sent to the blood bank for transfusion crossmatching.

The operator performed right coronary angiography to determine the source of the bleeding. This showed wide patency of the stent in the right coronary artery with no evidence of a perforation at the site of the stent (Figure 21-1); however, contrast was observed in the pericardial space (Video 21-1) and the pigtail catheter continued to drain blood. Additional views were performed; free-flowing contrast was apparent from the distal tip of the posterior descending artery in an anteroposterior view with cranial angulation (Video 21-2). A 2.5 mm balloon was inflated in the posterior descending artery proximal to the perforation and effectively stopped the bleeding (Figure 21-2). While the balloon remained inflated, the patient remained hemodynamically stable with no additional blood accumulation from the pericardial drain. A cardiothoracic surgeon was informed of the patient’s condition and alerted to the possible need for emergency surgery to correct the problem. Meanwhile, 12 units of platelets were rapidly infused and an infusion of packed red blood cells begun. After 10 minutes, the balloon in the posterior descending artery was deflated. Repeat angiography showed ongoing contrast extravasation from the distal posterior descending artery. The balloon was then reinflated for 20 minutes. Angiography after balloon deflation confirmed no further evidence of contrast extravasation (Video 21-3). The patient was observed in the cardiac catheterization laboratory and another angiogram performed 10 minutes later showed no further contrast extravasation. No additional blood was aspirated from the pericardial catheter. In total, 2.6 L of blood drained from the pericardial catheter and he received a total of 4 units of packed red blood cells in the cath lab.

FIGURE 21-1 Right coronary angiogram performed after pericardiocentesis showing wide patency of the coronary stent (arrow) placed earlier with no apparent perforation at the site of the stent. A pigtail catheter is present in the pericardial space.

FIGURE 21-2 This figure depicts the site of balloon inflation in the posterior descending artery proximal to the site of contrast extravasation, resulting in successful control of additional bleeding.

Postprocedural course

The patient left the cath lab off pressors with the pericardial drain secured in place. Except for pleuritic chest pain, he remained clinically stable. He was maintained on clopidogrel and aspirin. There was no rise in serial biomarkers and no electrocardiographic changes. After 24 hours, no additional fluid could be aspirated from the drain and no effusion was observed on echocardiography, prompting removal of the drain. The remainder of his hospital stay was uncomplicated and he was discharged on the second day postprocedure. He was seen at follow-up at 1 and 7 months later feeling well with no further angina.

Discussion

When a patient develops hypotension after a successful coronary intervention, the physician must work quickly to determine if it is caused by benign etiologies such as dehydration or procedural sedation, or from life-threatening etiologies such as retroperitoneal bleed, contrast-related anaphylaxis, abrupt vessel closure, and tamponade. In the present case, elevated neck veins and the absence of ischemic chest pain or ST-segment changes prompted the physician to immediately consider tamponade. An emergency echocardiogram confirmed this suspicion and led to rapid treatment of tamponade, identification and treatment of the bleeding source, and an excellent ultimate outcome.

Tamponade is a very rare complication of coronary intervention, with an incidence in one recent study of 0.12%.¹ Coronary perforation caused by the guidewire, balloon, stent, or other interventional device is responsible for nearly all cases of tamponade. Rarely, tamponade may occur from perforation from a temporary pacemaker lead used to support the intervention or from free-wall rupture complicating an acute myocardial infarction.

Overall, tamponade occurs in 12% to 25% of coronary perforations, but is dependent upon the type of perforation present.^2,3 Ellis and colleagues³ classified perforations by angiographic criteria as follows: Type I are limited to a crater extending outside of the lumen; Type II are characterized by the presence of a contrast blush in the pericardium or myocardium without an exit hole greater than 1 mm in diameter; and Type III perforations are present when there is free-flowing contrast through an exit hole greater than 1 mm. A variation of Type III is known as “Type III cavity-spilling,” defined as perforation into an anatomic chamber such as a ventricle or the coronary sinus. Tamponade complicates about 10% of Type I and II perforations, but occurs in about 60% of Type III perforations.³ The patient shown in this case had a Type II perforation, with contrast seen in the pericardium but with a very small exit hole caused by the guidewire. Guidewire perforations are more commonly seen when hydrophilic or stiff-tipped wires are used to cross complex lesions or chronic total occlusions. In this case, the tip of a conventional, floppy-tipped guidewire perforated a very distal vessel. This case emphasizes the importance of carefully monitoring the location of the wire tip during the course of an intervention to be sure the tip does not stray into a small branch where it may perforate if advanced inadvertently.

When a perforation is immediately recognized at the time of the intervention, steps may be promptly taken to prevent the development of tamponade. In this case, the wire perforation was not appreciated and hypotension from tamponade developed after he left the catheterization laboratory, resulting in a medical emergency. This is not unusual. Tamponade may result from a slow, steady, unrecognized bleeding caused by an occult wire perforation or from a deep dissection at the site of the intervention. This process may be fueled by powerful platelet antagonists such as the platelet glycoprotein IIb/IIIa receptor antagonists. Only about half of cases of tamponade develop in the cath lab. In the remaining cases, tamponade may not be manifest until many hours after the intervention.¹ Thus, tamponade from an occult perforation should be strongly suspected in a patient with progressive hypotension after a coronary intervention.

Once identified, the first priority is to correct the hemodynamic derangement of tamponade. Rapid infusion of saline along with pressor support can temporarily stabilize a critically ill patient while preparing for pericardiocentesis. In the event of a coronary perforation, rapid accumulation of only a small volume of blood in the pericardium may cause tamponade; this may be difficult to remove, particularly in an obese patient. In addition, rarely, some patients with coronary perforation develop an epicardial hematoma causing local tamponade. In such cases, pericardiocentesis will not relieve compression. However, in most patients, pericardiocentesis rapidly stabilizes the patient, allowing the physician to concentrate on correcting the underlying problem.

In the presence of tamponade due to coronary perforation (or any other life-threatening bleed, for that matter), it is crucial to promptly reverse the procedural anticoagulation. Unfractionated heparin may be reversed with protamine. Low-molecular-weight heparin will only be partially corrected by protamine. Bivalirudin, a direct thrombin inhibitor, cannot be easily reversed, requiring cryoprecipitate, fresh frozen plasma, or factor VIIa infusion for correction; fortunately, it has a short half-life when discontinued. Pharmacologic differences in the platelet glycoprotein IIb/IIIa inhibitors create problems when bleeding requires reversal. The high affinity and irreversible binding property of abciximab means that there is little free drug circulating; this agent can be reversed by replacing all of the patient’s circulating, inhibited platelets with transfused platelets (10 to 12 units). The smaller molecules (eptifibatide and tirofiban) are low affinity and bind reversibly to platelets and thus there is a great amount of circulating drug present; transfused platelets will also be affected thus limiting the benefit of platelet transfusion. Fortunately, these drugs are short-acting and their antiplatelet effect drops off rapidly once discontinued.

Repair of the perforation site requires precise identification of the location. In the event of a distal wire perforation as shown in this case, a covered stent is not an option. Proximal occlusion of the vessel with a balloon catheter successfully stops the hemorrhage. Prolonged stasis results in clotting of the exit hole; this may require 10 to 20 minutes of balloon occlusion to accomplish. If prolonged balloon inflation coupled with reversal of anticoagulation does not successfully seal the perforation, surgical repair should be considered. This is necessary in only about 10% of cases.