Successful Treatment of Left Ventricular Perforation Occurring After Insertion of an Impella Device

Abstract

The Impella device is a miniaturized ventricular assist device that is being increasingly used to increase the safety and efficacy of high-risk coronary interventions and to treat patients with acute myocardial infarction complicated by cardiogenic shock. The device has a miniaturized rotary pump mounted on a 9F catheter with a pigtail conformation. The pump draws blood from the left ventricular cavity and expels it into the ascending aorta and systemic circulation. We report a patient who, following insertion of an Impella device, developed angiographically documented left ventricular perforation with marked hemodynamic instability. Our successful management of this patient is described and potential mechanisms responsible for the perforation are discussed.

The Impella device (Abiomed Inc., Danvers, Massachusetts) is a miniaturized ventricular assist device that is being increasingly used to improve the safety and efficacy of high-risk coronary interventions (PCI) and to treat patients with acute myocardial infarction complicated by cardiogenic shock. ¹ The Impella CP device (the particular Impella model used in our reported patient) has a miniaturized rotary pump mounted on a 9F catheter with a pigtail conformation. ¹ The device is placed retrogradely across the aortic valve via a 14 F sheath in the femoral artery, using conventional catheterization techniques. The pump draws blood from the left ventricular (LV) cavity and expels it into the ascending aorta and systemic circulation, providing up to 4.0 L/minute forward flow. The device effectively unloads the LV, improves coronary perfusion and augments cardiac output. Most of the complications encountered with the Impella device relate to the necessary large bore femoral access sites, and include bleeding requiring transfusion, femoral artery occlusion or vascular damage requiring surgical intervention, and infections. ¹ Prompted by a recently published report of a patient who developed LV perforation following placement of an Impella device, ² we reviewed the records of all patients who underwent Impella implantation at our center through February 2018 (1,127 patients). Only one patient with this complication was identified in our series. We now report the clinical presentation and management of this patient.

Case Report

A 79-year-old man was referred to our center for angioplasty of a chronically occluded proximal left anterior descending coronary artery (LAD). (Our center currently performs > 50 chronic total coronary artery occlusion PCI procedures per year with a success rate> 80%). The patient had, during the preceding 3 years, suffered multiple myocardial infarctions. Despite optimal medical therapy and several PCI procedures, he had developed progressively impaired LV systolic function and, subsequently, received a dual chamber implantable cardioverter-defibrillator. At the time of presentation, he had Canadian Cardiac Society class 3 angina ³ and was on dual antiplatelet therapy (ticagrelor and aspirin). A recent nuclear stress test had demonstrated evidence of inducible ischemia in the anterior wall and the estimated left ventricular ejection on a preprocedure echocardiogram was 30%.

Given the patient's advanced age and severely decreased LV systolic function, the proposed PCI was judged to fall in to the high-risk category. ¹ In devising our PCI strategy for this patient, we were conscious that we might not be able to cross the lesion antegradely and instead might need to attempt a retrograde approach to open the occluded vessel. The latter puts the donor vessel at risk and if indeed injury to the donor vessel occurs there is the potential for hemodynamic collapse and potentially lethal rhythm disturbances. In such a situation, the hemodynamic support provided by the Impella device can be lifesaving. Accordingly, in anticipation of these possible events, we elected to place an Impella CP device prior to attempting PCI. After placing 6F sheaths in both femoral arteries and an 8F sheath in the left femoral vein, the left arterial sheath was changed out over a 0.35″ Wholey wire (Medtronic Inc., Minneapolis, MN) for a long arterial sheath. Then a 6F multipurpose diagnostic catheter (Medtronic) was advanced through the sheath to the LV. Next, the Impella 0.18” deployment wire was advanced through the multipurpose catheter to the LV and the multipurpose catheter then withdrawn. The 8F sheath was changed out for a 14F sheath and an Impella CP device advanced over the wire to the LV. The wire was withdrawn, and fluoroscopy showed the Impella catheter in the LV.

Shortly after entry of the Impella catheter in the LV, the patient reported chest pain and became progressively hypotensive (nadir systolic blood pressure of 55 mm Hg). Pressors (epinephrine and norepinephrine) were started. Given the concern for ventricular perforation with tamponade echocardiography was performed immediately. This showed a moderate sized pericardial effusion with tamponade physiology. Echocardiography-guided pericardiocentesis was then performed with aspiration of > 600 cm ³ of bloody fluid and with immediate improvement in blood pressure. A 5F pigtail catheter (Medtronic) was left in the pericardium to allow continuing drainage. Left ventriculography by hand injection ( Fig. 1 ) was then performed using a 5F pigtail catheter; this demonstrated communication between the LV and the pericardium. Continuing brisk bloody flow was seen in the pericardial drain.

Fig. 1.

Left ventriculography (hand injection) showing ventricular perforation.

The patient was then brought emergently to the operating room with the intention of achieving in expeditious fashion definitive treatment of the perforation – i.e., surgical closure of the catheter induced ventricular laceration. A sternotomy was performed. Next, a small opening was made in the pericardium, resulting in significant relief of the tamponade (systolic blood pressure increasing promptly from 80 to 150 mm Hg). On further exposure of the heart, the Impella catheter tip was seen protruding through the posterolateral left ventricular wall ( Fig. 2 ). Several pledgeted 3–0 Prolene (Ethicon, New Jersey) sutures were placed in the area of perforation and, as the catheter was being withdrawn, the sutures were laid down. Good hemostasis was achieved, and the patient was gradually weaned off pressors. Intraprocedurally, the patient (who was on long term aspirin and ticagrelor therapy) received 6 units of blood, 2 units of fresh frozen plasma, 5 units of cryoprecipitate, and 5 units of platelets. Following closure of the perforation, bypassing the LAD, using the left internal mammary artery as a graft, was considered. However, we were conscious that the patient was on dual antiplatelet therapy and that this surgery could result in significant additional bleeding problems.

Fig. 2.

Intraoperative image showing the Impella catheter perforating the posterolateral wall of the left ventricle (arrow).

Post procedurally, the patient had an uneventful course. We did not attempt to intervene again on his LAD lesion. He was discharged home from hospital one week later. At 21 months clinical follow-up, the patient remains symptomatically stable.

Discussion

Our reported elderly patient with established heart failure and severe coronary artery disease (on maintenance dual antiplatelet therapy) arrived in the operating room in extremis . He was in cardiogenic shock and had profuse ongoing bleeding from the cardiac perforation site. The observation that the Impella catheter had perforated through the heart mandated immediate removal of the catheter and suturing of the perforation site. Fortunately, in our patient this was successfully achieved, and the patient remains alive at 21-month follow-up.

There has only been one previous report of LV perforation associated with Impella use. Unfortunately, perforation was only recognized after more than 24 hours following Impella device implantation and the outcome was fatal. ² As in our case, the patient required surgery with removal of the Impella device and suturing of the perforation site. The authors postulate that gradual erosion of the left ventricle by the Impella catheter was the mechanism of cardiac perforation.

As we have previously noted, it would be surprising that the “friendly” pigtail design of the Impella catheter would directly perforate the ventricle. ⁴ In discussing the perforation in our reported case, it is important to remember that entry of the Impella catheter in the LV was facilitated by prior placement in the ventricle both of an exchange wire and of a multipurpose end-hole catheter. Both exchange wires and end-hole catheters have been recognized to cause ventricular perforation. Accordingly, one potential explanation for the perforation occurring in our reported patient is that the perforation was initially caused by the deployment wire. The wire having become embedded in the ventricle, and then served as a rail to the perforation site for the Impella catheter. An alternative potential mechanism is that it was the end-hole catheter that caused the initial perforation and that, when the exchange catheter was advanced through this catheter, it became embedded in the perforation site with subsequent direction of the Impella catheter to the site. However, regardless of the mechanism, Impella-induced cardiac perforation requires immediate surgical intervention.

Over the past 5 years, driven in part by recognition of the superior hemodynamic support provided by the Impella device when compared with the intraaortic balloon pump, there has been a dramatic increase in the use of Impella devices. Indeed, to date, over 50,000 Impella devices have been used in the United States. ⁵ Accordingly, cardiologists and cardiac surgeons should be aware of the described potentially lethal complication associated with this device.