Hemodynamic Support During High-Risk PCI

Introduction

Coronary balloon angioplasty (PTCA), introduced in 1977, was initially limited to treatment of focal, noncalcified, proximal, concentric coronary lesions. Initially case selection was confined to patients with normal left ventricular function and single vessel coronary artery disease. Complex lesions (eccentric, bifurcation, chronic total occlusion) and patients with multivessel coronary disease simply could not be technically approached. Dramatic advances in guidewire technology, advances in balloon technology, new atherectomy devices, and the introduction of flexible, easily deployable stents have dramatically expanded the complexity of the lesions that can be treated successfully. As the age of the population and the life expectancy of patients with overt coronary artery disease (CAD) increase, clinicians are increasingly faced with elderly, highly symptomatic patients with multiple comorbidities that make PCI and CABG prohibitively risky. In this setting, the coronary interventionalist is asked to perform a PCI. To safely undertake these high-risk procedures, hemodynamic support may be required. Who requires hemodynamic support and what method of support is safest and most effective are the subjects of this chapter.

Rationale for Definition of High Risk

Through the 1990s and up to 2005, scores of studies were published regarding “high-risk” percutaneous coronary intervention (HRPCI). Studies were mainly single center registries or a few prospective multicenter registries. Wide variations in outcomes were reported. The consistent characteristic that appeared to enhance risk was severely impaired ventricular function.

Keelan has demonstrated a thirtyfold increase in risk of death for patients with ejection fractions less than 40% compared to patients with normal ejection fraction. There are multiple reasons why underlying impaired left ventricular (LV) function increases risk. First, in order to have ejection fractions less than 40%, patients have survived an extensive acute myocardial infarction (MI). This often results in the presence of a chronic total occlusion. As survivors of acute MI, these patients present later with recurrent symptoms related to worsening of another coronary artery. Thus patients are later into the natural history of their disease and usually have multiple vessel disease. As CAD patients age, other comorbidities such as diabetes, chronic kidney disease (CKD), and chronic obstructive lung disease (COPD) increase procedural risk. During the interventional procedure, patients with severely depressed LV function pose a special challenge. Typically the operator is treating a stenosis that subtends a large segment of the normally contracting left ventricle and that often subtends a distant myocardial segment that has blood flow through the collaterals. Repeated interruptions of blood flow that occur with contrast injections, balloon inflation, and stent manipulation can lead to ischemia of normally contracting myocardial segments. This results in marked elevation of LV filling pressure and can quickly lead to sustained hypotension or cardiovascular collapse. For this reason the two main prospective randomized trials, hemodynamic support with Impella 2.5 versus the intra-aortic balloon pump and the balloon pump-assisted Coronary Intervention Study, have both used severely impaired LV function as the primary risk segregator. In addition, the BCIS investigators chose an advanced jeopardy score ≥8/12 and the Protect II investigators added the presence of a last patent coronary vessel as inclusion criteria ( Table 5-1 ). These simple, concise criteria have in fact defined a patient population with a thirtyfold increase in mortality risk during hospitalization compared to routine elective multiple vessel angioplasty ( Table 5-2 ). Future clinical trials and prospective quality assurance endeavors should use these benchmark inclusion criteria and outcome data for comparative safety analysis.

Clinical Presentation

This chapter initially discusses management of elective or urgent patients who are treated outside the setting of acute MI or cardiogenic shock. These patients present late in the natural history of coronary atherosclerosis. Average age is >70 years. A large proportion of the population has diabetes mellitus, renal function is impaired, and they are highly symptomatic. NYHA symptoms III or IV were present in 60% of the patients in BCIS and 65% of Protect II. Symptoms could be worsening heart failure or angina. In fact, since a large segment of these patients have diabetes, it is often difficult to differentiate worsening heart failure from worsening ischemia. In the Protect II study, 26% of the patients had peripheral vascular disease and 64% were evaluated and turned down for CABG by the site surgeons. Thus these elderly, highly symptomatic patients have limited therapeutic options apart from high-risk PCI. Age, poor ventricular function, diabetes, CKD, and peripheral vascular disease all increase procedural risk. In addition, coronary angiography, IVUS interrogation, coronary balloon inflation, rotational atherectomy, or complex stent procedures (kiss, crush, coulotte) may compromise coronary blood flow to the vital myocardial segments. Without hemodynamic support, these coronary manipulations can result in cardiovascular collapse. To avoid cardiac collapse, clinicians may choose to perform unassisted procedures with minimal coronary angiography, lack of IVUS interrogation, brief balloon, stent inflation, and incomplete revascularization. Even with these precautions, in the BCIS study, 10% of the control, unsupported patients suffered from severe hypotension, and 12% of the patients required rescue intraaortic balloon pump (IABP). To optimize patient safety and long-term clinical effectiveness, a controlled, well-planned, supported PCI procedure must be considered.

Coronary Blood Flow and Myocardial Energetics During High-Risk PCI

Coronary blood flow physiology is well defined in mammalian models ( Figure 5-1 ). It is known that under conditions of maximal stress, a fourfold increase in resting blood flow can occur. Coronary occlusions begin to impact vasodilator reserve at 50% area occlusion and eliminate vasodilator capacity at 80% occlusion. After this, resting coronary blood flow dramatically decreases with small increases in occlusion severity. It is also well understood that in mammals, a perfusion gradient of 30 mm Hg to 50 mm Hg is required to drive blood through the myocardium ( Figure 5-2 ). In systole, when aortic and ventricular pressures are the same and when ventricular compression occurs, little forward blood flow occurs. Thus mammalian coronary blood flow is largely diastolic. As coronary occlusion worsens, the phasic nature diminishes and blood flow is largely driven by the mean arterial pressure with requirement for a 30 mm Hg to 50 mm Hg gradient for perfusion. How do these principles translate to this high-risk population? First, generally all three major coronary territories are affected with severe atherosclerosis. One or more coronaries may be chronically totally occluded from previous infarctions. Patients are presenting with worsening of non–infarct-related vessels with lesions severe enough to diminish vasodilatory reserve or decrease resting coronary blood flow. The remaining patent vessels provide collateral flow to the occluded territories and likely have severe lesions. The demand for increased coronary blood flow to supply collateral flow and the usual hyperkinesis of the noninfarct zones cause the patent coronaries to be at a maximal state of coronary vasodilation. In this setting, blood flow is entirely driven by perfusion gradient ( Figure 5-3 ). No incremental vasodilation is possible, since the patent vessels have severe underlying lesions. Thus the coronary tree has no capacity for augmentation of blood flow ( Figure 5-4 ). Coronary blood flow is driven entirely by the transmyocardial perfusion gradient and diastolic interval. When patients become distressed, tachycardia, elevated LVEDP, and systolic hypotension occur. These factors dramatically increase myocardial oxygen demand (MVO ₂ ) ( Figure 5-5 ) or decrease supply. At the same time, decreased perfusion gradients and coronary blood flow can lead to a rapid cycle of worsening ischemia and lowering of blood pressure, which leads to less coronary blood flow, worsening myocardial contracture, and irreversible shock. To prevent this catastrophe, measures to maximize the myocardial perfusion gradient, minimize MVO ₂ , and maintain adequate cardiac output for vital organ perfusion are required.

In addition to coronary blood flow physiology, an understanding of myocardial energetics is required to determine optimal support strategies for these patients. Myocardial energetics can be studied using pressure volume relationships. Suga and Burkhoff demonstrated that pressure volume loops are an ideal way to define cardiac performance throughout the cardiac cycle ( Figure 5-6 ). Ideally, interventions that decrease the end diastolic pressure-volume relationship improve myocardial energetics by decreasing ventricular wall tension and thus decreasing MVO ₂ . At the same time, interventions that increase the end systolic pressure-volume relationship increase the transmyocardial perfusion gradient and improve coronary blood flow. When the transmyocardial pressure gradient drops below 30 mm Hg due to a decrease in mean arterial pressure or an elevation of LVEDP, dangerous reductions in coronary blood flow will occur and cardiac collapse will soon follow. Thus the goal of hemodynamic support is threefold. First, maintaining a mean arterial pressure ≥60 mm Hg is required to maintain coronary flow. Second, maintaining a cardiac power ≥0.6 watts will prevent systemic hypoperfusion. Finally, decreasing ventricular diastolic pressure will enhance the coronary perfusion gradient and decrease LV wall tension. Burkhoff and Naidu summarized the impact of hemodynamic support strategies on these three variables using a circulation simulator.

Each hemodynamic support maneuver has been modeled ( Figure 5-7 ). First, inotropic agents appear to augment MAP and cardiac output but do so at the expense of increasing diastolic filling pressure (PCWP) and thus increasing MVO ₂ . Inotropes increase MVO ₂ by increasing contractility, heart rate, and wall tension. If LVEDP is increased proportionally more than MAP, myocardial perfusion pressure is lowered and coronary blood flow will drop. At high levels of alpha adrenergic tone, coronary vasoconstriction will occur and coronary blood flow will drop further. Thus isotropes alone are a poor strategy that provides only modest support of MAP at the expense of increased MVO ₂ and potentially less coronary blood flow.

The addition of an intraaortic balloon pump to an inotropic support provides more support of mean arterial pressure and cardiac output but does so at the expense of elevations of wall tension and MVO ₂ . It is likely that the increased MVO ₂ demand will not be matched by increased coronary blood flow and this will set up a substrate for myocardial ischemia. The impact of IABP on coronary blood flow has been of interest for three decades. Soon after its introduction, it became apparent that IABP could stabilize patients with recurrent, medically refractory unstable angina. Williams et al. measured LAD blood flow by placing thermodilution catheters into the great cardiac vein in six patients with refractory angina who had severe proximal LAD lesions. They found that the greater cardiac vein flow decreased from 78 ± 11 to 69 ± 8 mL/min, p = 0.048, when IABP support was initiated. In addition, systolic pressure was significantly reduced and augmented diastolic pressure was increased. They concluded that the mechanism of the relief of angina related to the decreased afterload and decreased MVO ₂ rather than increased coronary blood flow. Yoshanti measured coronary pressures using the Radi Wire (Radi Medical Systems, Uppsala, Sweden) in 16 patients with unstable angina. Pressures were measured distal to 16 culprit lesions and 5 normal vessels. The distal diastolic pressures were unchanged after initiation of IABP distal to the severe stenotic segments (42.8 ± 17 to 44 ± 21, p = NS). In the normal vessels, pressures increased with initiation of IABP (78 ± 9 to 97 ± 8 mm Hg, p < 0.05). Kern et al. measured coronary flow velocity with a Doppler tipped catheter (Millar Instruments). They found that the coronary flow velocity was unchanged distal to severe lesions after IABP was turned on. After PTCA of the lesion, coronary flow velocity significantly increased with IABP support. Thus IABP support may be of benefit when severe hypotension occurs and one or more coronary arteries are free of disease. In this circumstance, augmentation of diastolic pressure will enhance coronary blood flow. Since most patients under­going HRPCI have multivessel disease, IABP support will be of little value in maintaining or supporting coronary blood flow.

The TandemHeart Pump (Cardiac Assist Inc., Pittsburgh, Pennsylvania) substantially decreases filling pressure and augments mean arterial pressure as well as cardiac output. Since wall tension is reduced and mean arterial pressure is elevated, MVO ₂ is reduced and coronary blood flow should be increased. Of the hemodynamic support devices, Impella is the only one that directly drains the left ventricle and augments forward cardiac flow. As cardiac output is increased from 2.5 to 5 L, filling pressures are reduced, mean arterial pressure is elevated, and myocardial perfusion gradient is elevated. This device has the ideal properties of maintaining mean arterial pressure and forward cardiac output while decreasing MVO ₂ . The device seems well suited for short- and long-term hemodynamic support. Finally, Extracorporeal Membrane Oxygenator (ECMO), also known as Peripheral Cardiopulmonary Bypass (CPS), is unequivocally useful when both right and left heart failure occurs. It provides excellent support for cardiac output and MAP. Unfortunately, it does so at the expense of elevated filling pressure and elevated MVO ₂ . For patients with severe multivessel CAD, it is likely that a longer duration of the use of this device sets up a substrate for myocardial ischemia. In the short term, the device provides superb support of the cardiac output and mean aortic pressure. To summarize, the previous discussion of modeling provides the clinician with the theoretic basis to choose from a variety of support strategies. In addition to these factors, patient body surface area, status of peripheral circulation, urgency of the need for hemodynamic support, and operator training and experience using the individual devices play a major role in defining an optimal support strategy for individual patients. Ideally these complex interventions should be performed in referral centers with expertise and access to many types of support and with open communication either in-house or rapidly available LVAD and transplant referral support.

History of Development of LV Support Devices for HRPCI

The 1985 to 1986 National Heart Lung and Blood Institute (NHLBI) registry of coronary angioplasty documented outcomes for PCI in the pre-stent era. The presence of depressed LV function identified a group of patients who often had multivessel CAD. When the LV ejection fraction was less than or equal to 35%, there was a 3% mortality rate compared to a 0.1% mortality rate for patients with LV EF greater than 45%. Left ventricular support with IABP was first developed to improve outcomes. The IABP had been in use since 1968 and was widely used to support patients in cardiogenic shock. Voudris reported the first European experience and Kahn reported the U.S. experience with IABP to support HR PTCA. Both groups used depressed LV function and/or unprotected left main angioplasty as entry criteria.

History of Use of IABP During HRPCI

The Mid America Heart group led by Dr. Goeffrey Hartzler pioneered the use of balloon angioplasty in patients with multivessel coronary artery disease. Similarly, this group first reported on adjunctive use of IABP in high-risk cases. Kahn et al. reported on 28 patients with a mean ejection fraction of 24%, three vessel diseases in 93%, and LMCA in 255 patients who had balloon angioplasty performed with IABP support. They found that augmented diastolic pressure greater than 90 mm Hg occurred in all cases. During hospitalization, no deaths were reported. Soon after, Voudris reported the initial European experience with supported HRPCI. In a 1-year period, 27/1385 patients in Toulouse, France, had supported HRPCI. Ejection fraction less than 40% was present in 24/27 cases. No deaths or MI occurred during initial hospitalization. The impact of poor ventricular function on outcomes was brought into clear focus by the NHLBI report in 1993. Holmes et al. reported on 244/1802 patients treated in the 1985 to1986 NLHBI PTCA registry who had EF ≤ 45% 9. Angioplasty was less successful (76% vs. 84%, p < 0.01) and 4-year survival was lower than patients with normal LV function. With the advent of coronary stent implantation, PCI of unprotected left main intervention began in the early 1990s. Briguori et al. reported the Milan, Italy, experience with stent implantation in 219 consecutive patients. Of these patients, 69 had prophylactic IABP support and 150 had conventional unsupported procedures. Severe hemodynamic instability occurred in 8% of the unsupported group and none of the IABP group. Major adverse events were higher in the unsupported group (9.5% vs. 1.5%, p = 0.032). Thus both depressed LV function and unprotected left main stent therapy started to be considered “high-risk PCI.” More registry data continued to be reported concerning the value of IABP support. Mishra et al. summarized the Washington Heart Center experience with prophylactic versus rescue IABP support. During a period from January 2000 to December 2004, outcomes of 68 patients with prophylactic IABP support were compared to 46 patients with IABP support started as a rescue for hemodynamic compromise. At 6 months' follow-up, mortality was lower (8% vs. 29%, p < 0.01) and MACE events were lower (12% vs. 32%, p = 0.02) for patients with elective IABP support. While these single center reports suggested a safety advantage for elective IABP support, Curtis et al. found a more mixed picture in the National Cardiovascular Registry. They reported outcomes from January 2005 to December 2007 in 181,599 high-risk PCI procedures; 1170 had cardiogenic shock, 80% had ST-elevation acute MI, 2% underwent unprotected LMCA PCI, and 20% had EF less than 30%. An IABP was used in 44% of cardiogenic shock patients, 28% of patients with UPLM, and 14% of patients with EF less than 30%; overall IABP support was used in 10.5% of HRPCI cases and varied widely from hospital to hospital. The hospitals were divided into quartiles of IABP use (6.5%, 6.6% to 9.2%, 9% to 14%, >14%). No convincing evidence was found that IABP use improved outcomes. The wide variation in IABP use rates across U.S. hospitals suggested enormous operator preference and a lack of strong evidence for the elective use of prophylactic balloon pump therapy.

The IABP is constructed of a polyethylene balloon that is mounted onto a catheter that can be delivered through a sheath or sheathless to provide hemodynamic support in patients undergoing high-risk PCI. Although it may be inserted in the femoral artery without fluoroscopic guidance in an emergent situation, fluoroscopy is preferable when available to help with positioning. Prior to insertion of the balloon pump, a one-way valve is attached to the over the wire lumen and a large syringe may be used to apply negative pressure to downsize the balloon to facilitate passage through the sheath. Using the standard technique, a Cook needle is used to access the femoral artery. A small incision may be made with a scalpel in the subcutaneous tissue, and a sheath may then be delivered over a 0.035-inch wire. After the sheath is inserted, most kits contain a 0.018-inch wire that is inserted through the sheath into the ascending aorta. The balloon is advanced over this wire and delivered 2 cm distal to the left subclavian artery. After position is verified, the inflation-deflation of the balloon pump is normally timed with ECG monitoring of the console. The balloon can be filled with up to 50 cc of helium, which allows rapid inflation and deflation to provide both augmentation in coronary perfusion and unloading of the left ventricle during systole. In addition to placement through the femoral artery, the balloon pump may also be inserted into the axillary artery for those patients with severe peripheral arterial disease. In the future, balloon pump shafts small enough for 6-F access, in which case it is conceivable that radial access for balloon pump support may be feasible, at least for short-term use. Most balloon pump consoles may provide augmentation to heart rates up to 150 beats per minute and may operate on battery power for limited time periods (Maquet Quick Reference Guide IABP insertion/CS300 Operation, 2014 Maquet). When heart rates greater than 130 beats per minute occur, a 2 : 1 ratio of pumping allows optimal balloon pump inflation-deflation. Daily x-rays should be performed to assure that catheter migration does not occur. In addition, knee splints are useful to place on the leg with the balloon pump inserted so as to immobilize the leg and prevent catheter dislodgment. If incremental support is required, balloon pumps may be exchanged for larger sheath sizes.

To do this, a stiff 260-cm 0.018-inch guidewire is advanced through the lumen of the balloon pump shaft and the balloon pump is withdrawn. Negative pressure should be maintained with a large bore syringe while the balloon pump is removed. Once the balloon pump is removed, firm pressure is maintained at the puncture site while a 6F or 7F sheath dilator is advanced. The dilator remains in place, the 0.018 wire is removed, and a 0.035 guidewire can then be advanced for upsizing of the sheaths. If concern exists at the time of the initial balloon pump insertion, a regular 8F sheath can be used, the balloon pump can simply be removed, and a 0.035 guidewire can be advanced. The balloon pump may be inserted into the femoral or axillary artery to provide hemodynamic support.