Chapter 28. Surgical Treatment of Complications of Acute Myocardial Infarction: Postinfarction Ventricular Septal Defect and Free Wall Rupture

Rupture of the ventricular chamber (septum or free wall) after myocardial infarction is a relatively infrequent condition with high mortality. An acute postinfarction VSD is a perforation of the muscular ventricular septum occurring in an area of acutely infarcted myocardium. A ventricular septal rupture may be termed chronic when it has been present for more than 4 to 6 weeks. A postinfarction ventricular rupture is a perforation of the ventricular free wall occurring in an area of acutely infarcted myocardium. These conditions, resulting from transmural infarction, may cause rapid hemodynamic compromise and early death precluding surgical repair. Free wall rupture can result in tamponade and sudden cardiovascular collapse. In ventricular septal rupture, there is a variable amount of left-to-right shunting, but such defects typically lead to symptoms of heart failure. The clinical presentation ranges from an asymptomatic murmur to cardiogenic shock and sudden death.

The early evolution of successful surgical repair of an acute postinfarction ventricular septal rupture involved differentiating the surgical treatment of these acquired lesions from surgical approaches used to repair congenital ventricular septal defects (VSDs), which are for the most part not applicable. Initial success was achieved with methods involving infarctectomy and patching. Specific methods were developed for differing anatomical locations of postinfarction VSDs, including location of the cardiotomy and patch methodology. With experience, there was gradual appreciation of different clinical courses pursued by patients after postinfarction ventricular septal rupture, both in terms of location of the defect and the degree of right ventricular functional impairment, led to an increased urgency relative to the timing of surgical repair. An important paradigm shift had resulted from improved results utsing a technique of endocardial patching with infarct exclusion. Surgical management requires an understanding of the various approaches. The incorporation of specific anatomic concepts of surgical repair and a better understanding of the physiologic basis of the disease has led to an integrated approach to the patient that has improved salvage of patients suffering this catastrophic complication of acute myocardial infarction (AMI).

History

In 1845 Latham1 described a postinfarction ventricular septal rupture at autopsy, but it was not until 1923 that Brunn2 first made the diagnosis antemortem. Sager3 in 1934 added the 18th case to the world literature and established specific clinical criteria for diagnosis, stressing the association of postinfarction septal rupture with coronary artery disease (CAD).

The treatment of this entity was medical and strictly palliative until 1956, when Cooley and associates4 performed the first successful surgical repair in a patient 9 weeks after the diagnosis of septal rupture. These first patients who underwent similar repairs in the early 1960s usually presented with congestive heart failure (CHF), having survived for more than a month after acute septal perforation.5,6 The success of operation in these patients and the precipitous, acute course of other patients with this complication gave rise to the belief that operative repair should be limited to patients surviving for 1 month or longer.6,7 This purportedly allowed for scarring at the edges of the defect, which was thought to be crucial to the secure and long-lasting closure of the septal rupture.8,9

In the late 1960s, more rapid recognition of septal rupture after infarction led to the recommendation that operation be attempted earlier in patients who were hemodynamically deteriorating. Notable among these was a superb early study by Heimbecker and associates of infarctectomy and its clinical application to patients with postinfarction VSDs. The surgical management of these patients was further refined by the inclusion of infarctectomy and aneurysmectomy and the development of techniques to repair perforations in different areas of the septum.10–12 Improved surgical techniques, newer prosthetic materials, enhanced myocardial protection, and improved perioperative mechanical and pharmacologic support have led to more favorable results in the surgical management of patients with postinfarction septal rupture.

Incidence

Postinfarction ventricular septal defects complicate approximately 1 to 2% of cases of AMIs and account for about 5% of early deaths after MI.13,14 The average time from infarction to rupture has been reported to be between 2 and 4 days, but it may be as short as a few hours or as long as 2 weeks.15,16 These observations correlate well with the pathologic findings, which demonstrate that necrotic tissue is most abundant and ingrowth of blood vessels and connective tissue is only beginning 4 to 21 days after an MI.17 Postinfarction ventricular septal defects occur in men more often than women (3:2), but more women experience rupture than what would be expected from the incidence of CAD in women. The average age of patients with this complication is 62.5 years, although there is some evidence that more elderly patients are being seen in the recent era. The vast majority of patients who experience ventricular septal rupture do so after their initial infarction.18 The overall incidence of postinfarction ventricular septal rupture may have decreased slightly during the past decade as a result of aggressive pharmacologic treatment of ischemia and thrombolytic and interventional therapy in patients with evolving MI, as well as the prompt control of hypertension in these patients. The effect of widespread use of percutaneous angioplasty and stenting on the appearance of this complication of MI is not well documented, but several large centers, including ours, have had a decreased number of post-MI VSD patients in the past 10 years.

Angiographic evaluation of patients with postinfarction ventricular rupture indicates that septal rupture is usually associated with complete occlusion rather than severe stenosis of a coronary artery.19 On average, these patients have slightly less extensive CAD, as well as less developed septal collaterals than do other patients with CAD.20 The lack of collateral flow noted acutely may be secondary to anatomical configuration, edema, or associated arterial disease. Hill and associates,21 in reviewing 19 cases of postinfarction ventricular septal rupture, found single-vessel disease in 64%, double-vessel disease in 7%, and triple-vessel disease in 29%. However, the frequency of single-, double-, and triple-vessel CAD is more evenly distributed in other series.16,22

Postinfarction ventricular septal defects are most commonly located in the anteroapical septum as the result of a full-thickness anterior infarction (in approximately 60% of cases). These anterior septal ruptures are caused by anteroseptal MI after occlusion of the left anterior descending (LAD) artery. In about 40% of patients, the rupture occurs in the posterior septum after an inferoseptal infarction, which is usually owing to occlusion of a dominant right coronary artery, or less frequently, a dominant circumflex artery. Thus, ventricular septal perforations occur most frequently in 65-year-old men with single-vessel coronary disease and poor collateral flow who present 2 to 4 days after their first anterior MI.

Pathogenesis

The infarct associated with septal rupture is transmural and generally quite extensive, involving, on average, 26% of the left ventricular wall in hearts with septal rupture, compared with only 15% in other acute infarctions.14 In an autopsy study, Cummings and colleagues23 found that in patients with acute anterior or inferior infarctions, the amount of right ventricular infarction was much greater in the hearts with septal ruptures as compared with those without septal defects. Likewise, hearts with posterior septal rupture had more extensive left ventricular necrosis than did hearts with inferior infarctions and no septal defects.

Why certain hearts rupture and others do not is not fully understood. Slippage of myocytes during infarct expansion may allow blood to dissect through the necrotic myocardium and enter either the right ventricle or pericardial space. Hyaline degeneration of cardiomyocytes with subsequent fragmentation and enzymatic digestion may allow fissures to form, predisposing to rupture.24–26

There are two types of rupture: simple, consisting of a direct through-and-through defect usually located anteriorly; and complex, consisting of a serpiginous dissection tract remote from the primary septal defect, which is usually located inferiorly. Multiple defects, which may develop within several days of each other, occur in 5 to 11% of cases and are probably caused by infarct extension. Because a successful surgical outcome is related to adequacy of closure of septal defects, multiple defects must be sought preoperatively if possible, and certainly at the time of operative repair.

Of the small number of patients who survive the early period of ventricular septal rupture, 35 to 68% go on to develop ventricular aneurysms through the process of ventricular remodeling.27 This compares with an approximately 12% incidence of aneurysm formation in patients suffering an infarction but no septal rupture,28 and probably relates to the size and transmural nature of the infarction associated with septal rupture. Postinfarction septal rupture, especially in the posterior septum, may be accompanied by mitral valve regurgitation resulting from papillary muscle infarction or dysfunction. In approximately one-third of cases of septal rupture, there is a degree of mitral insufficiency, usually functional in nature, secondary to left ventricular (LV) dysfunction with mitral annular dilation, which usually resolves with repair of the defect.20

Pathophysiology

The most important determinant of early outcome after postinfarction ventricular septal rupture is the development of heart failure (left, right, or both). The associated cardiogenic shock leads to end-organ malperfusion, which may be irreversible.

The degree to which heart failure develops depends on the size of the ventricular infarction and the magnitude of the left-to-right shunt. Left ventricular dysfunction resulting from extensive necrosis of the left ventricle is the primary determinant of CHF and cardiogenic shock in patients with anterior septal rupture, whereas right ventricular dysfunction secondary to extensive infarction of the right ventricle is the principal determinant of heart failure and cardiogenic shock in patients with posterior septal rupture. However, the development of CHF and cardiogenic shock in a patient with postinfarction VSDs is not explained solely by the degree of damage sustained by the ventricle.

The magnitude of the left-to-right shunt is the other key variable in the development of hemodynamic compromise. With the opening of a VSD, the heart is challenged by an increase in pulmonary blood flow, and a decrease in systemic blood flow, as a portion of each stroke volume is diverted to the pulmonary circuit. As a consequence of the sudden increase in hemodynamic load imposed on a heart already compromised by acute infarction, and possibly by a ventricular aneurysm, mitral valve dysfunction, or a combination of these problems, a severe low cardiac output state results. The normally compliant right ventricle is especially susceptible to failure in this circumstance.29,30 Patients with posterior ventricular septal rupture and right ventricular dysfunction may display shunt reversal during diastole because the end-diastolic pressure in the right ventricle can be higher than in the left. Ultimately, persistence of a low cardiac output state results in peripheral organ failure.

Diagnosis

The typical presentation of a ventricular septal rupture is that of a patient who has suffered an AMI, and who after convalescing for a few days develops a new systolic murmur, recurrent chest pain, and an abrupt deterioration in hemodynamics. The development of a loud systolic murmur, usually within the first week after an AMI, is the most consistent physical finding of postinfarction ventricular septal rupture (present in greater than 90% of patients). The murmur is usually harsh, pansystolic, and best heard at the left lower sternal border. The murmur is often associated with a palpable thrill. Depending on the location of the septal defect, the murmur may radiate to the left axilla, thereby mimicking mitral regurgitation. Up to one-half of these patients experience postinfarction chest pain in association with the appearance of the murmur.14 Coincident with the onset of the murmur, there is usually an abrupt decline in the patient's clinical course, with the onset of congestive failure and often cardiogenic shock. The findings of cardiac failure that occur acutely in these patients are primarily the result of right-sided heart failure, with pulmonary edema being less prominent than that occurring in patients with acute mitral regurgitation caused by a ruptured papillary muscle.32

The electrocardiographic (ECG) findings in patients with acute septal rupture relate to the changes associated with antecedent anterior, inferior, posterior, or septal infarction. The localization of infarction by ECG correlates highly with the location of the associated septal perforation. In our review18 of 55 patients with postinfarction septal rupture, the location of the defect corresponded to the territory of transmural infarction as determined by ECG in all but three patients. Up to one-third of patients develop some degree of atrioventricular conduction block (usually transient) that may precede rupture,33 but there is no pathognomonic prognostic indicator of impending perforation. The chest radiograph usually shows increased pulmonary vascularity consistent with pulmonary venous hypertension.

It is important to realize that the sudden appearance of a systolic murmur and hemodynamic deterioration after infarction may also result from acute mitral regurgitation caused by a ruptured papillary muscle. Distinguishing these two lesions clinically is difficult, and an urgent echocardiogram should be obtained. A number of points may help with the initial evaluation. First, the systolic murmur associated with a septal rupture is more prominent at the left sternal border, whereas the murmur resulting from a ruptured papillary muscle is best heard at the apex. Second, the murmur associated with septal perforation is loud and associated with a thrill (in greater than 50% of patients), whereas the murmur of acute mitral regurgitation is softer and has no associated thrill. Third, septal rupture is often associated with anterior infarctions and conduction abnormalities, whereas papillary muscle rupture is commonly associated with an inferior infarction and no conduction defects. Finally, it should be noted that septal rupture and papillary muscle rupture may coexist after infarction.34,35

Historically, the mainstay of differentiating septal rupture from mitral valve dysfunction has been right heart catheterization using the Swan-Ganz catheter.36 With septal rupture, there is an oxygen saturation step-up between the right atrium and pulmonary artery. Step-up in oxygen saturation greater than 9% between the right atrium and pulmonary artery confirms the presence of a shunt. The pulmonary-to-systemic flow ratios (Qp:Qs) obtained from oxygen saturation samples range from 1.4:1 to greater than 8:1 and roughly correlate with the size of the defect. In contrast, with acute mitral regurgitation secondary to papillary muscle rupture, there are classic giant V waves in the pulmonary artery wedge pressure trace. It should be noted, however, that up to one-third of patients with septal rupture also have mild mitral regurgitation secondary to LV dysfunction.37

Advances in transthoracic and transesophageal echocardiography, especially color flow Doppler mapping, have revolutionized the diagnosis of both the presence and site of septal rupture. Echocardiography can detect the defect, localize its site and size, determine right and left ventricular function, assess pulmonary artery and right ventricular pressures, and exclude coexisting mitral regurgitation or free wall rupture. Twenty years ago, Smyllie and associates38 reported 100% specificity and 100% sensitivity when color flow Doppler mapping was used to differentiate ventricular septal rupture from acute severe mitral regurgitation following AMI. It also correctly demonstrated the site of septal rupture in 41 of 42 patients. Widespread use of this technology has made this imaging the primary method of diagnosis. Indeed, the trend in the past two decades toward early surgical referral and prompt operative repair is at least partially explained by the routine use of color Doppler echocardiography for diagnosis in peripheral centers.

The necessity of preoperative left heart catheterization with coronary angiography has been a matter of debate. On one hand, left heart catheterization provides important information concerning associated CAD, left ventricular wall motion, and specifics of valvular dysfunction, which are all important in planning operative correction of postinfarction septal rupture. In most series, greater than 60% of patients with septal rupture have significant involvement of at least one vessel other than the one supplying the infarcted area. Arguably, bypassing associated CAD may increase long-term survival when compared with patients with unbypassed CAD.39 However, left heart catheterization has disadvantages—it is time consuming, requires use of nephrotoxic dye, and may contribute to both the mortality and morbidity of these already compromised patients. Thus, some centers do not carry out preoperative left heart catheterization. Others use it selectively, avoiding invasive studies in patients with septal rupture caused by anterior wall infarction, which is associated with a much lower incidence of multiple-vessel disease than septal defects resulting from posterior infarctions. The issue of concomitant coronary bypassing is discussed in greater detail below.

Natural History

Reviews by Oyamada and Queen,40 and Kirklin and coworkers41 reveal that nearly 25% of patients with postinfarction septal rupture and no surgical intervention died within the first 24 hours, 50% died within 1 week, 65% within 2 weeks, and 80% within 4 weeks; only 7% lived longer than 1 year. Lemery and associates42 reported that of 25 patients with postinfarction VSDs treated medically, 19 died within 1 month. Thus, the risk of death after postinfarction VSD is highest immediately after infarction and septal rupture, and then gradually declines. Interestingly, there are reports of spontaneous closure of small defects, although this is so rare that it would be unreasonable to manage a patient with the expectation of closure.

Despite the many advances in the nonoperative treatment of CHF and cardiogenic shock, including the intra-aortic balloon pump and a multitude of new inotropic agents and vasodilators, these do not supplant the need for operative intervention in these critically ill patients.

Management

It has become clear that the early practice of waiting for several weeks after ventricular septal rupture before proceeding with intervention only selects out the small minority of patients in whom the hemodynamic insult is less severe and is better tolerated.22,24 Likewise, it has also become clear that to manage most patients supportively, in hopes of deferring intervention, is to deprive the great majority of those with postinfarction ventricular septal rupture of the benefits of definitive surgery before irreversible damage resulting from peripheral organ ischemia has occurred.43

In addition to definitive surgical closure, early intervention can include mechanical support and device closure. The routine use of the intra-aortic balloon pump (IABP), whenever technically feasible, frequently results in transient reversal of the hemodynamic deterioration. This period of stability often makes it possible to complete left heart catheterization before proceeding to operation, but should not significantly delay definitive surgical treatment. In general, an IABP should be placed once the diagnosis is made, unless more aggressive mechanical support is immediately planned. In patients who are deemed to be at unacceptable risk for definitive surgery, there may be a role for left or biventricular support to bridge the patient to surgery (see the following). The high-risk patient may also represent a category in whom catheter-based device placement may greatly improve hemodynamic function and lead to stabilization, with a planned operation after recovery from the acute illness (see the following).

Although we as well as others have advocated early intervention since the middle of the 1970s, some continue to prefer to defer operation in patients who are easily supported and exhibit no further hemodynamic deterioration. Persistence of CHF or marginal stabilization with rising blood urea nitrogen and borderline urine output necessitate aggressive therapy and prompt operation. Patients with septal rupture rarely die of cardiac failure per se, but rather of end-organ failure as a consequence of shock.

Our experience and the experience of others suggest that patients in cardiogenic shock represent a true surgical emergency requiring immediate intervention, which may include surgery, mechanical support, or catheter options. Because deaths in these patients result from multisystem failure secondary to organ hypoperfusion, delay in operative repair (or mechanical support) for patients in cardiogenic shock represents a failed therapeutic strategy. Those few patients who are completely stable, with no clinical deterioration, and who require no hemodynamic support, can undergo operative repair when convenient during that hospitalization. The large group of patients who are in an intermediate position between those with shock and those in stable condition should have intervention early (usually within 12 to 24 hours) after appropriate preoperative evaluation. Because the group of patients in stable condition constitutes 5% or less of the total population of patients with postinfarction ventricular septal rupture, the overwhelming majority of patients require prompt treatment.

Rarely, because of a delayed referral, a patient will be seen for surgical therapy who is already in a state of multisystem failure or has developed septic complications. Such a patient is unlikely to survive an emergency operation and thus may benefit from prolonged support with an IABP before an attempted operative repair. We have found it necessary to treat a small number of patients (3 of 92) in this fashion. Baillot and colleagues44 have reported individual successes with such an approach, which we consider the exception rather than the rule.

Preoperative Management

Preoperative management is directed toward stabilization of the hemodynamic condition so that peripheral organ perfusion can be best maintained while any further diagnostic studies are obtained and while deciding on the optimal time for surgical intervention. Although the early clinical course of patients with postinfarction ventricular septal rupture can be quite variable, 50 to 60% present with severe CHF and a low cardiac output state requiring intensive therapy.45,74

The goals of preoperative management are to: (1) reduce the systemic vascular resistance, and thus the left-to-right shunt; (2) maintain cardiac output and arterial pressure to ensure peripheral organ perfusion; and (3) maintain or improve coronary artery blood flow. This is best accomplished by the IABP. Counterpulsation reduces left ventricular afterload, thereby increasing cardiac output and decreasing the left-to-right shunt, as reported by Gold and associates in 1973.46 In addition, IABP support is associated with decreased myocardial oxygen consumption, as well as improved myocardial and peripheral organ perfusion. Although counterpulsation produces an overall improvement in the patient's condition, a complete correction of the hemodynamic picture cannot be obtained.47 Peak improvement occurs within 24 hours and no further benefit has been observed with prolonged balloon pumping.48

Pharmacologic therapy with inotropic agents and diuretics should be instituted promptly. The addition of vasodilators (ie, sodium nitroprusside or intravenous nitroglycerine) makes good theoretical sense, because it can decrease the left-to-right shunting associated with the mechanical defect, and thus increase cardiac output. However, these effects are often associated with a marked fall in mean arterial blood pressure and reduced coronary perfusion, both poorly tolerated in these critically ill patients. It must be stressed that pharmacologic therapy is intended primarily to support the patient in preparation for operation and should not in any way delay urgent operation in the critically ill patient. We now admit patients with postinfarction septal rupture directly to the surgical intensive care unit rather than to the coronary care or medical intensive care unit.

Other techniques that have been tried in an effort to improve the hemodynamics of patients with interventricular septal rupture include left ventricular mechanical support, venoarterial extracorporeal membrane oxygenation, and inflation of a balloon in the right ventricular outflow tract to decrease the left-to-right shunt. At the present, we believe these techniques should be reserved for patients with severe end-organ malperfusion deemed unlikely to survive surgery. To avoid shunting across the lesion (right to left at the ventricular level), atrial cannulation is necessary. Use of a catheter-mounted axial flow pump (Hemopump, Impella) in stabilizing these patients is controversial because of the risk of acute pump failure resulting from catheter blockage from pieces of necrotic tissue.49

Operative Techniques

Historical Development

The first repair by Cooley and colleagues of an acquired VSD was accomplished using an approach through the right ventricle with incision of the right ventricular outflow tract.4 This approach, which was adapted from surgical techniques for closure of congenital VSDs, proved to be disadvantageous for many reasons. Exposure of the defect was frequently less than optimal, particularly for defects located in the apical septum. It involved unnecessary injury to normal right ventricular muscle and interruption of collaterals from the right coronary artery. Finally, it failed to eliminate the paradoxic bulging segment of infarcted left ventricular wall. Subsequently, Heimbecker and associates9 introduced a left-sided approach (left ventriculotomy) with incision through the area of infarction. Such an approach frequently incorporates infarctectomy and aneurysmectomy, together with repair of septal rupture.50

Overview

For the purpose of planning an operative repair, there are three locations of defects: apical, anterior, and posterior/inferior. The apical defect can be considered as a subset of the anterior location, and provides the surgeon the possibility of a straightforward repair, similar to an aneurysmectomy. There are two very different surgical paradigms for repair, one using infarctectomy, and the other infarct exclusion. Posterior lesions are the most challenging with the infarctectomy technique, and there is now a general trend to approach those lesions with exclusion methods. As described in the following, the surgical approach to lesion in various locations requires specific considerations, but certain general principles apply (Tables 28-1A and 28-1B).

Anesthesia and Perfusion

Patients are anesthetized using a fentanyl-based regimen. Pancuronium is selected as the muscle relaxant so as to prevent bradycardia. Pulmonary bed vasodilators such as dobutamine are avoided to minimize the left-to-right shunt fraction. Preoperative antibiotics include both cefazolin and vancomycin, given the fact that prosthetic material may be left in the patient.

Cardiopulmonary bypass is accomplished with bicaval venous drainage. Standard techniques of myocardial protection for injured hearts are employed. Systemic cooling is begun to 25 to 30°C. Although a number of myocardial protection strategies are currently available, including warm continuous blood, we continue to use cold oxygenated, dilute blood cardioplegia to protect the heart during surgical correction of a VSD.51 A total of 1200 to 2000 mL of cardioplegia solution is delivered, depending on the size of the heart and the degree of hypertrophy. Although we have not employed warm cardioplegic induction, we do administer warm reperfusion cardioplegia just before removing the aortic cross-clamp.52 Patients with multivessel coronary disease and critical coronary stenoses are revascularized before opening the heart to optimize myocardial protection. In most of these patients, the saphenous vein rather than the left internal mammary artery is used.

Repair of Apical Septal Rupture

The technique of apical amputation was described by Daggett and colleagues in 1970.11 An incision is made through the infarcted apex of the left ventricle. Excision of the necrotic myocardium back to healthy muscle results in amputation of the apical portion of the left ventricle, right ventricle, and septum (Figs. 28-1A and B). The remaining apical portions of the left and right ventricle free walls are then approximated to the apical septum. This is accomplished by means of a row of interrupted mattress sutures of 1-0 Tevdek that are passed sequentially through a buttressing strip of Teflon felt, the left ventricular wall, a second strip of felt, the interventricular septum, a third strip of felt, the right ventricular wall, and a fourth strip of felt (Fig. 28-2A and B). After all sutures have been tied, the closure is reinforced with an additional over-and-over suture, as in ventricular aneurysm repair, to ensure hemostasis of the ventriculotomy closure.

Anterior Repair with Infarctectomy

The approach to these defects is by a left ventricular transinfarct incision with infarctectomy (Fig. 28-3). Small defects beneath anterior infarcts can be closed by the technique of plication as suggested by Shumacker.53 This involves approximation of the free anterior edge of the septum to the right ventricular free wall using mattress sutures of 1-0 Tevdek over strips of felt (Fig. 28-4A). The transinfarct incision is then closed with a second row of mattress sutures buttressed with strips of Teflon felt (see Figs. 28-4 B–D). An over-and-over running suture completes the ventriculotomy closure.

Most anterior defects require closure with a prosthetic patch (DeBakey Elastic Dacron fabric, USCI Division of C.R. Bard, Inc., Billerica, MA) or pericardium to avoid tension that could lead to disruption of the repair (Fig. 28-5). After debridement of necrotic septum and left ventricular muscle, a series of pledgeted interrupted mattress sutures are placed around the perimeter of the defect (see Fig. 28-5A). Along the posterior aspect of the defect, sutures are passed through the septum from right to left. Along the anterior edge of the defect, sutures are passed from the epicardial surface of the right ventricle to the endocardial surface. All sutures are placed before the patch is inserted, and then passed through the edge of a synthetic patch, which is seated on the left side of the septum (see Fig. 28-5B). Each suture is then passed through an additional pledget and all are tied. We use additional pledgets on the left ventricular side overlying the patch (see Fig. 28-5C) to cushion each suture as it is tied down to prevent cutting through the friable muscle. The edges of the ventriculotomy are then approximated by a two-layer closure consisting of interrupted mattress sutures passed through buttressing strips of Teflon felt (or glutaraldehyde-preserved bovine pericardium) and a final over-and-over running suture.

Posterior/Inferior Repair with Infarctectomy

Closure of inferoposterior septal defects, which result from transmural infarction in the distribution of the posterior descending artery, has posed the greatest technical challenge.54,55 Because of the difficulty in surgical management of these defects by the method of infarctectomy, many surgeons now feel that these defects may be better suited to repair using exclusion techniques, which are described in detail in the next section.

Early attempts at primary closure of these posterior/inferior defects by simple plication techniques similar to those used in the repair of anterior defects were frequently unsuccessful because of the sutures tearing out of soft, friable myocardium that had been closed under tension. This resulted in either reopening of the defect or catastrophic disruption of the infarctectomy closure. It was, in large part, the analysis of such early results that led to the evolution of the operative principles enumerated in Table 28-1A.

Use of the following techniques has been associated with an improved operative survival. After the establishment of bypass with bicaval cannulation, the left side of the heart is vented via the right superior pulmonary vein. The heart is retracted out of the pericardial well as for bypass to the posterior descending coronary artery. The margins of the defect may involve the inferior aspects of both ventricles, or of the left ventricle only (Fig. 28-6A). A transinfarct incision is made in the left ventricle, and the left ventricular portion of the infarct is excised (see Fig. 28-6B), exposing the septal defect. The left ventricular papillary muscles are inspected. Only if there is frank papillary muscle rupture is mitral valve replacement performed. When it is indicated, we prefer to perform mitral valve replacement through a separate conventional left atrial incision, to avoid trauma to the friable ventricular muscle. After all infarcted left ventricular muscle has been excised, a less aggressive debridement of the right ventricle is accomplished, with the goal of resecting only as much muscle as is necessary to afford complete visualization of the defect(s). Using this technique, delayed rupture of the right ventricle has not been a problem. If the posterior septum has cracked or split from the adjacent ventricular free wall without loss of a great deal of septal tissue, then the septal rim of the posterior defect may be approximated to the edge of the diaphragmatic right ventricular free wall using mattress sutures buttressed with strips of Teflon felt or bovine pericardium (see Fig. 28-6C and D).

Larger posterior defects require patch closure (Fig. 28-7). Pledgeted mattress sutures are placed from the right side of the septum and from the epicardial side of the right ventricular free wall (see Fig. 28-7B). All sutures are passed through the perimeter of the patch and then through additional pledgets, and are then tied (see Fig. 28-7C). Thus, as in closure of large anterior defects, the patch is secured on the left ventricular side of the septum. Direct closure of the remaining infarctectomy is rarely possible because of tension required to pull together the edges of the gaping defect. A prosthetic patch is generally required. Originally, we cut an oval patch from a Cooley low-porosity woven Dacron tube graft (Meadox Medicals, Inc., Oakland, NJ). Currently, we cut this patch from a Hemashield woven double velour Dacron collagen impregnated graft (Meadox Medicals). Pledgeted mattress sutures are passed out through the margin of the infarctectomy (endocardium to epicardium) and then through the patch (see Fig. 28-7D), which is seated on the epicardial surface of the heart. After each suture is passed through an additional pledget, all sutures are tied (see Fig. 28-7E). The cross-sectional view of the completed repair (Fig. 28-8) illustrates the restoration of relatively normal ventricular geometry, which is accomplished by the use of appropriately sized prosthetic patches.

Repair of Anterior and Posterior Defects by Infarct Exclusion

The concept that the preservation of left ventricular geometry plays a crucial role in the preservation of left ventricular function has laid the groundwork for evolution in the surgical approach to postinfarction VSDs—the technique of endocardial patch repair of postinfarction VSDs described by David,50 Cooley,56 and then by Ross57 in the early 1990s. This operative technique, which is an application to ventricular septal rupture repair of Dor's technique of ventricular endoaneurysmorrhaphy,58 involves intracavitary placement of an endocardial patch to exclude infarcted myocardium while maintaining ventricular geometry. Thus, instead of closing the septal defect, it is simply excluded from the high-pressure zone of the left ventricle. Some institutions have reported impressive results using infarct exclusion, but results in for other centers have been mixed. We currently consider this a particular helpful technique in selected patients with posterior/inferior defects, given the complexity of a infarctectomy repair in this location. The description that follows is based in large part on the work of David.46

In patients with anterior septal rupture, the interventricular septum is exposed via a left ventriculotomy, which is made through the infarcted anterolateral wall starting at the apex and extending proximally parallel to, but 1 to 2 cm away from, the anterior descending artery (Fig. 28-9A). Stay sutures are passed through the margins of the ventriculotomy to aid in the exposure of the infarcted septum. The septal defect is located and the margins of the infarcted muscle identified. A glutaraldehyde-fixed bovine pericardial patch is tailored to the shape of the left ventricular infarction as seen from the endocardium but 1 to 2 cm larger. The patch is usually oval and measures approximately 4 × 6 cm in most patients. The pericardial patch is then sutured to healthy endocardium all around the infarct (see Fig. 28-9B). Suturing begins in the lowest and most proximal part of the noninfarcted endocardium of the septum with a continuous 3-0 polypropylene suture. Interrupted mattress sutures with felt pledgets may be used to reinforce the repair.57 The patch is also sutured to the noninfarcted endocardium of the anterolateral ventricular wall. The stitches should be inserted 5 to 7 mm deep in the muscle and 4 to 5 mm apart. The stitches in the patch should be at least 5 to 7 mm from its free margin so as to allow the patch to cover the area between the entrance and exit of the suture in the myocardium. This technique minimizes the risk of tearing muscle as the suture is pulled taut. If the infarct involves the base of the anterior papillary muscle, the suture is brought outside of the heart and buttressed on a strip of bovine pericardium or Teflon felt applied to the epicardial surface of the left ventricle. Once the patch is completely secured to the endocardium of the left ventricle, the left ventricular cavity becomes largely excluded from the infarcted myocardium. The ventriculotomy is closed in two layers over two strips of bovine pericardium or Teflon felt using 2-0 or 3-0 polypropylene sutures, as illustrated in Fig. 28-9C. No infarctectomy is performed unless the necrotic muscle along the ventriculotomy is sloughing at the time of its closure, and even then it is minimized, because infarcted muscle will not be exposed to left ventricular pressures when the heart begins to work (see Fig. 28-9D). Alternatively, sutures can be passed through the ventricular free wall and through a tailored external patch of Teflon or pericardium (Fig. 28-10).

In patients with posterior septal defects, an incision is made in the inferior wall of the left ventricle 1 or 2 mm from the posterior descending artery (Fig. 28-11A). This incision is started at the midportion of the inferior wall and extended proximally toward the mitral annulus and distally toward the apex of the ventricle. Care is taken to avoid damage to the posterolateral papillary muscle. Stay sutures are passed through the fat pad of the apex of the ventricle and margins of the ventriculotomy to facilitate exposure of the ventricular cavity. In most cases, the rupture is found in the proximal half of the posterior septum and the posteromedial papillary muscle is involved by the infarction. A bovine pericardial patch is tailored in a triangular shape of approximately 4 × 7 cm in most patients. The base of the triangular-shaped patch is sutured to the fibrous annulus of the mitral valve with a continuous 3-0 polypropylene suture starting at a point corresponding to the level of the posteromedial papillary muscle and moving medially toward the septum until the noninfarcted endocardium is reached (see Fig. 28-11B). At that level, the suture is interrupted and any excess patch material trimmed. The medial margin of the triangular-shaped patch is sewn to healthy septal endocardium with a continuous 3-0 or 4-0 polypropylene suture taking bites the same size as those described for anterior defects. In this area of the septum, reinforcing pledgeted sutures may be required. The lateral side of the patch is sutured to the posterior wall of the left ventricle along a line corresponding to the medial margin of the base of the posteromedial papillary muscle. Because the posterior wall of the left ventricle is infarcted, it is usually necessary to use full-thickness bites and anchor the sutures on a strip of pericardium or Teflon felt applied on the epicardial surface of the posterior wall of the left ventricle right at the level of the posteromedial papillary muscle insertion, as shown in Fig. 28-11B. Once the patch is completely sutured to the mitral valve annulus, the endocardium of the interventricular septum, and the full thickness of the posterior wall (see Fig. 28-11C), the ventriculotomy is closed in two layers of full-thickness sutures buttressed on strips of pericardium or Teflon felt (see Fig. 28-11D). The infarcted right ventricular wall is left undisturbed. If the posteromedial papillary muscle is ruptured, mitral valve replacement is necessary.

There are several theoretical advantages in the technique of infarct exclusion: (1) It does not require resection of myocardium; excessive resection results in depression of ventricular function and insufficient resection predisposes to recurrence of septal rupture; (2) it maintains ventricular geometry, which enhances ventricular function; and (3) it avoids tension on friable muscle, which may diminish postoperative bleeding.

Other Techniques

Most other operative techniques that have resulted in successful management of postinfarction of ventricular septal rupture have adhered to the same general principles described in the preceding. For example, Tashiro and colleagues59 described an extended endocardial repair in which a saccular patch of glutaraldehyde-fixed equine pericardium was used to exclude an anterior septal rupture. Usui and coworkers60 reported the successful repair of a posterior septal rupture using two sheets of equine pericardium to sandwich the infarcted myocardium, including the septal defect and ventriculotomy. Others have modified the exclusion technique by using multiple patches or use of tissue sealants to aid in the septal closure.61,62

Percutaneous Closure

Successful transcatheter closure of postinfarction ventricular septal rupture has been reported using several types of catheter-deployed devices. Early experience was with the CardioSEAL device, a nitinol double-umbrella prosthesis. The device consists of two attached and opposing umbrellas formed by hinged steel arms covered in a Dacron meshwork that theoretically promotes endothelialization. The arms are manually everted to allow the device to be passed through a narrow percutaneous deployment system. When extruded from the guiding catheter, the arms spring backward, resembling a clamshell. The device approaches the septum via the systemic veins and through the atrial septum (or alternatively via the arterial system through the aortic valve). As reported by Landzberg and Lock,63 the experience at Boston Children's Hospital and Brigham and Women's Hospital indicates that although the device can be routinely deployed in the setting of an acute infarction, the continued necrosis of septal tissue led to decompensation and death in four of seven patients. In contrast, they reported success in six of six patients treated for residual or recurrent septal defects discovered after primary operative repair. Other catheter devices have also been attempted, including the Amplatzer septal occluder and the Rashkind double umbrella.64

The recently developed Amplatzer VSD device allows closure of muscular and membranous VSDs, and it can be used for larger postinfarction defects, with the septal Amplatzer device favored for smaller lesions.65 The device features a longer waist (10 mm) connecting the two umbrella ends, which corresponds more closely to the thickness expected for the adult ventricular septum

Results with device closure have been mixed. In a recent report involving use of these devices as a primary treatment strategy, "procedural success" was reported as 86%, but 41% of patients had important complications that included left ventricular rupture, device embolization, and major residual shunting.66 The 30-day survival rate was only 35%. Others have also reported that residual shunts are the norm, although often greatly improved by the procedure. Several series noted a poor outcome when this approach was attempted early after infarction, when the defect margins are fragile, but after about 2 weeks the results were considered acceptable.67–69 Of course, surgical results after a delay are also considerably better, but many patients would die in the first 2 weeks without intervention.

The best use of such devices in an overall treatment strategy is unclear. As a primary treatment, data suggest that the devices have a high early failure rate, but their potential role in improving the risk of an unstable surgical patient is not yet well characterized. Currently, catheter approaches appear to be most effective in treatment of recurrent or residual defects,70 and we preferentially employ them for these conditions. Device development is an ongoing process, and the future undoubtedly will see use of new devices, especially in the high-risk patient with multisystem failure. An additional area of use for catheter devices is to reduce shunt fraction in an attempt to stabilize a patient for a planned delayed operation. We are employing this technique when we deem early surgery to be at extreme risk, with a planned operation either immediately if device closure is unsuccessful, or at approximately 6 to 8 weeks if stabilization is obtained.

Of interest, two centers have reported using a standard Swan-Ganz balloon catheter from the groin to abolish the shunt in unstable patients with postinfarction septal rupture.71,72 Hemodynamic improvement was immediate in both patients, who underwent subsequent surgical repair of the defect.

Role of Ventricular Assist Devices

In patients who present for operation with evidence of potentially reversible multiorgan dysfunction, or in patients who have intractable failure after repair, there may be a role for temporary mechanical heart support. There have been multiple anecdotal cases of unstable patients being successfully managed by mechanical support followed by definitive operation.

The theoretical advantages that make mechanical support attractive as an initial therapy in very sick patients with postinfarction VSD include: (1) the potential to reverse end-organ dysfunction; (2) maturation of the infarct leading to firmer tissue, making the closure less prone to technical failure; and (3) recovery of the stunned and energy-depleted myocardium. However, there are potential hazards with mechanical support that are specific to the patient with postinfarction VSD. High right-to-left shunting across the ventricular septum has been reported to cause hypoxic brain injury in a postinfarction VSD patient placed on a Heart-Mate left ventricular support device.107 This observation suggests that either partial left heart support or preferably biventricular support should be considered when using mechanical assistance in these patients. In a report using the Hemopump axial flow device, two of two patients supported experienced lethal pump failure. Examination of the device at autopsy disclosed necrotic material clogging the catheter system.73a

Simultaneous Myocardial Revascularization

There has been controversy in the literature concerning the advantages and disadvantages of concurrent coronary artery grafting in patients undergoing emergent repair of postinfarction ventricular septal rupture. The potential benefit of revascularization in a patient with significant lesions to living muscle is obvious: improved myocardial distribution, protection from postoperative ischemic, and reduced late ischemic events. However, in patients with postinfarction VSDs, there is already a completed myocardial injury and the involved territory is unlikely to benefit. For this reason, some have argued that revascularization provides no survival benefit and subjects patients to preoperative left heart catheterization, a time-consuming and potentially dangerous diagnostic procedure. Loisance and associates73b base their policy of not revascularizing patients with postinfarction septal ruptures on the fact that none of their 20 long-term survivors (five of whom were bypassed) had incapacitating angina or recurrent myocardial infarction.

Some groups use left heart catheterization and coronary bypassing selectively. Davies and colleagues74 found that of 60 long-term survivors (median 70 months; range 1 to 174 months), only five patients developed exertional angina during follow-up and none required revascularization. Their current policy is to avoid left heart catheterization of patients in whom an acquired septal defect is suspected to be a consequence of their first anterior infarction, provided that the patient has no history of angina or electrocardiographic evidence of previous infarction in another territory.

Weaning from Cardiopulmonary Bypass

The two most common problems encountered in separating from bypass after repair of a postinfarction VSD are low cardiac output and bleeding. Although the treatment of low cardiac output after cardiac surgery is beyond the scope of this chapter, a few agents and principles are worth mentioning. First, most of these patients will have had an IABP inserted before surgery. If not, one should be inserted in the operating room, especially if the low-output state is secondary to LV dysfunction. Also, an IABP may benefit patients with right ventricular failure by improving right coronary artery blood flow resulting from diastolic augmentation. We have found intravenous milrinone, a phosphodiesterase inhibitor, to be very effective in reversing low-output states secondary to LV dysfunction. Milrinone possesses a balance of inotropic and vasodilatory properties that together produce an increase in cardiac output and reduction in right and left filling pressures and systemic vascular resistance. It is less arrhythmogenic than dobutamine, causes less hypotension than amrinone, and is not associated with thrombocytopenia.

Posterior defects are commonly associated with mitral regurgitation and right heart dysfunction secondary to extensive right ventricular infarction. Management of right heart failure is aimed at reducing right ventricular afterload while maintaining systemic pressure. Initial steps to manage right ventricular dysfunction include volume loading, inotropic support, and correction of acidosis, hypoxemia, and hypercarbia. If patients remain unresponsive to these measures, we have successfully treated right ventricular failure with a prostaglandin E1 infusion (0.5 to 2.0 μg/min) into the right heart, counterbalanced with a norepinephrine infusion titrated into the left atrium.75 Inhaled nitric oxide (20 to 80 ppm), which selectively dilates the pulmonary circuit, has also proved efficacious in the treatment of right heart failure.76

If a patient cannot be weaned from bypass using conventional therapy, we consider using a ventricular assist device. Indications for a left ventricular assist device are a cardiac index less than 1.8 L/min/m2, a left atrial pressure above 18 to 25 mm Hg, a right atrial pressure below 15 mm Hg, and an aortic pressure below 90 mm Hg peak systolic. Indications for a right ventricular assist device are a cardiac index less than 1.8 L/min/m2, an aortic pressure below 90 mm Hg peak systolic, and a left atrial pressure less than 15 mm Hg despite volume loading to a right atrial pressure of 25 mm Hg with a competent tricuspid valve. Important points to remember when instituting ventricular assistance are:

1. Right ventricular failure may not become evident until left ventricular assistance is instituted.

2. Once refractory ventricular failure has been identified, delay in initiating support is associated with increased morbidity and mortality.

3. Closure of a patent foramen ovale is mandatory before left ventricular support.

4. Postoperative hemorrhage should be treated aggressively and completely controlled.

5. Residual septal defects may result in right-to-left shunting and severe hypoxia when only left heart support is used.

Hemostatic Measures

To prevent postpump coagulopathy, we begin antifibrinolytic with ϵ-aminocaproic acid (Amicar) before commencing cardiopulmonary bypass. Amicar is administered by loading patients with 10 g before commencing bypass and then adding another 10 g to the pump prime. During the procedure Amicar is continuously infused at 1 g/h for the duration of surgery. Aprotinin was previously used routinely for this condition, but concerns regarding early and late complication led this drug to be withdrawn from the US market.

Postpump suture line bleeding may be reduced by application of a fibrin sealant to the ventricular septum around the septal defect before formal repair.77 Biological glue may be effective in controlling bleeding suture lines after repair. As a last resort, Baldwin and Cooley78 have suggested insertion of a left ventricular assist device solely as an adjunct to the repair of friable or damaged myocardium to reduce left ventricular distention and thus control bleeding. Finally, for intractable bleeding, there may be role for Factor VII concentrate, although results and risks of administration in cardiac surgery are limited.

Highlights of Postoperative Care

Early postoperative diuresis and positive end-expiratory pressure ventilation are used to decrease the alveolar-arterial gradient induced by the increased extravascular pulmonary water associated with cardiopulmonary bypass. Once the patient has warmed, we commonly use an intravenous infusion of furosemide combined with mannitol or, if needed, continuous venovenous hemofiltration is employed postoperatively.

Intractable postoperative ventricular arrhythmias secondary to reperfusion injury are sometimes difficult to control using standard therapy. We have been impressed with the efficacy of intravenous amiodarone in these situations (10 to 20 mg/kg over 24 hours).

Operative Mortality and Risk Factors for Death

Table 28-2 summarizes recently reported experience from several centers. Operative mortality, defined as death before discharge or within 30 days of operation, ranged from 30 to 50%. In the Massachusetts General Hospital experience of 114 patients, operative mortality was 37% (Fig. 28-12A). The risk for death was found to be very high initially, but dropped rapidly (see Fig. 28-12B). We identified independent risk factors for early and late death using multivariate methods (Table 28-3). The most important predictor of operative mortality in our study, and in other reports, was preoperative hemodynamic instability. Patients in this group are usually in cardiogenic shock, are emergency cases, are on inotropic support, and usually have intra-aortic balloon pumps. Several variables are highly correlated with hemodynamic instability, and different multivariate models may use one or more of these indicators of severe hemodynamic failure in their final model.

Additional risk factors for early and late death include the presence of left main CAD, previous MI, renal dysfunction, and right heart failure (Fig. 28-13). Other factors have been found to increase the risk of early death. Posterior location of the septal rupture has been associated with an increased operative mortality.22,23,29,31 This has been attributed to a more technically difficult repair, the increased risk of associated mitral regurgitation, and associated right ventricular dysfunction that is an independent predictor of early mortality after posterior infarction. A short time interval between infarction and operation selects for sicker patients unable to be managed medically. Older patient age has also been associated with an increased early mortality, but in our analysis, we found that the impact of age was more pronounced in the high-risk patient. Thus, age alone should not be used as a reason for denying surgery in an otherwise low-risk candidate (Fig. 28-14).

Our review of the Massachusetts General Hospital experience underscored the large variability of risk to which patients could be segregated using a few clinical variables (Figs. 28-15 and 28-16), most notably indicators of hemodynamic instability (emergency surgery and use of inotropes). The result was that a small group of high-risk patients dramatically affected the overall mortality rate. We believe that this phenomenon makes it very difficult to compare mortality among institutions. A slight difference in practice patterns, such as a tendency of a surgeon or referring cardiologist to deny operation, could substantially affect results. Additionally, any difference in transport dynamics to certain centers could lead to loss of unstable patients, which could create another type of selection bias. In our opinion, these issues are by far the most important source of mortality differences in modern series.

Several centers have reported improved early results with an "exclusion" repair.87,95 Our group has not been able to replicate these results, with a 60% mortality in 10 patients (higher than the rate achieved historically with traditional techniques). This is likely to the result of our tendency to use this technique on the most challenging type of defects in whom we anticipate a long and complex repair, and continued use of infarctectomy in the lower risk patients with anterior or apical defects.

Regardless of the technique, the most common cause of death after repair of acute postinfarction VSD was low cardiac output syndrome (52%). Technical failures, most commonly recurrent or residual VSD, but including bleeding, were the second most common (23%). Other causes of death include sepsis (17%), recurrent infarction (9%), cerebrovascular complications (4%), and intractable ventricular arrhythmias.

Long-Term Results

Long-term results have been favorable with regard to both mortality risk and functional rehabilitation. Actuarial survival at 5 years for most recent series generally ranges between 40 and 60% (see Table 28-2). Because of the overall high risk of the operation, it is rewarding to note that hospital survivors enjoy excellent longevity, with 1-, 5-, and 10-year survival of 91, 70, and 37%, respectively. They also are quite functional—among 15 of our patients contacted during follow-up of long-term survivors, 75% were in New York Heart Association functional class I, and 12.5% were class II.18

Recurrent Ventricular Septal Defects

Recurrent or residual septal defects have been diagnosed by Doppler color flow mapping early or late postoperatively in 10 to 25% of patients.96 They may be caused by reopening of a closed defect, to the presence of an overlooked defect, or to the development of a new septal rupture during the early postoperative period. These recurrent defects should be closed when they cause symptoms or signs of heart failure or when the calculated shunt fraction (pulmonary-to-systemic flow ratio) is large (Qp:Qs >2.0). When they are small (Qp:Qs <2.0) and either asymptomatic or controlled with minimal diuretic therapy, a conservative approach is reasonable and late spontaneous closure can occur. Intervention in the catheterization laboratory may be useful in closing symptomatic residual or recurrent defects postoperatively.

Chronic Ventricular Septal Defects

In 1987 Rousou and associates reported successful closure of an acquired posterior VSD by means of a right transatrial approach.97 Filgueira and colleagues have used the transatrial approach for delayed repair of chronic acquired posterior septal defects.98 Approaching a postinfarction VSD through the tricuspid valve should not be used in acute cases because of the friability of the necrotic septum, poor exposure, and because this technique does not involve infarctectomy, and thus cannot achieve the hemodynamic advantages of elimination of a paradoxically bulging segment of ventricular wall. However, the right heart approach can be used in chronic postinfarction VSD when the septum is well scarred and the patch can be safely sutured to it from the right atrium. We emphasize that although the transatrial approach may be used selectively for the closure of chronic defects, it is unlikely to be an appropriate choice for the closure of acute defects, except perhaps in the rare circumstance when an infarct is localized to the septum with no evidence of necrosis of the free wall of the left ventricle.25

History

William Harvey first described rupture of the free wall of the heart after AMI in 1647.99 In 1765, Morgagni reported 11 cases of myocardial rupture found postmortem.100 Ironically, Morgagni later died of myocardial rupture.101 Hatcher and colleagues from Emory University reported the first successful operation for free wall rupture of the right ventricle in 1970.101 FitzGibbon and associates102 in 1971 and Montegut103 in 1972 reported the first successful repairs of left ventricular ruptures associated with ischemic heart disease.

Incidence

Autopsy studies reveal that ventricular free wall rupture occurs about 10 times more frequently than postinfarction ventricular septal rupture, occurring in about 11% of patients after AMI.104 The incidence has been found to be as high as 31% in autopsy studies of anterior MI.105 Ventricular rupture and cardiogenic shock are now the leading causes of death after AMI, and together account for greater than two-thirds of early deaths in patients suffering their first acute infarction. Postinfarction ventricular ruptures are more common in elderly women (mean age of 63 years) suffering their first infarction.106 In the prethrombolytic era, 90% of ruptures occurred within 2 weeks after infarction with the peak incidence at 5 days.107 In contrast, the time to cardiac rupture (not frequency of rupture) seems to be accelerated by thrombolysis and coronary reperfusion, sometimes occurring within hours from the onset of symptoms.108

Opinions differ as to the most common site of left ventricular rupture. The older literature suggests that the anterior wall is the most frequent site; however, more recent series have observed a preponderance of lateral and posterior wall ruptures.104,105 Lateral wall infarction may be more likely to rupture than an anterior or inferior injury, but because anterior infarctions are much more frequent than lateral infarctions, overall the most common site of rupture is the anterior wall. Like postinfarction ventricular septal rupture, free wall ruptures may be simple or complex. A simple rupture results from a straight through-and-through tear that is perpendicular to the endothelial and epicardial surfaces, whereas a complex rupture results from a more serpiginous tear, often oblique to the endocardial and epicardial surfaces. Batts and coworkers107 reported 100 consecutive cases of left ventricular free wall rupture and found that half were simple ruptures and the rest were complex.

Pathogenesis and Pathophysiology

Left ventricular free wall rupture can be divided into three clinicopathologic categories: acute, subacute, and chronic. An acute or "blow-out" rupture is characterized by sudden recurrent chest pain, electrical mechanical dissociation, profound shock, and death within a few minutes because of massive hemorrhage into the pericardial cavity. This type of rupture is probably not amenable to current management. A subacute rupture is characterized by a smaller tear, which may be temporarily sealed by clot or fibrinous pericardial adhesions. These usually present with the signs and symptoms of cardiac tamponade, and eventually cardiogenic shock. Subacute rupture may mimic other complications of AMI, such as infarct extension and right ventricular failure, and may be compatible with life for several hours or days or even longer.109 A chronic rupture with false aneurysm formation occurs when the leakage of blood is slow and when surrounding pressure on the epicardium temporarily controls the hemorrhage. Adhesions form between the epicardium and pericardium, which reinforce and contain the rupture. The most common clinical presentation of patients with false aneurysms of the left ventricle is CHF.110 A false aneurysm may also be an echocardiographic finding in an otherwise asymptomatic patient recovering from AMI. Angina, syncope, arrhythmias, and thromboembolic complications occur in a small percentage of patients. There are four major differences between a true and false aneurysm of the left ventricle:

1. The wall of a false aneurysm contains no myocardial cells.

2. False aneurysms are more likely to form posteriorly.

3. False aneurysms usually have a narrow neck.

4. False aneurysms have a great propensity for rupture.

Rupture of the free wall of the left ventricle may occur in isolation or with rupture of other ventricular structures such as the interventricular septum, papillary muscles, or right ventricle.110,111

The pathogenesis of cardiac rupture remains poorly understood. However, cardiac rupture occurs only with transmural MIs and infarction expansion appears to play an important role in its pathogenesis.24,27,112 Infarct expansion is an acute regional thinning and dilatation of the infarct zone, seen as early as 24 hours after acute transmural MI and not related to additional myocardial necrosis.113 This regional thinning and dilatation of the infarct zone is a consequence of slippage between muscle bundles, resulting in a reduction in the number of myocytes across the infarcted area. Infarct expansion increases the size of the ventricular cavity, with a consequent increase in wall tension (Laplace effect) that subjects the infarct zone to more tension and predisposes to endocardial tearing.109 Systemic hypertension aggravates the problem of thinning and dilatation of the infarct wall and increases the probability of rupture.114 Lack of collateral flow may also promote ventricular rupture.115

Because myocardial rupture occurs in regions of complete transmural myocardial necrosis, usually after extensive hemorrhagic transformation of the acute infarct, and because thrombolytic therapy is associated with the conversion of a bland infarct into a hemorrhagic infarct, there has been an ongoing concern that thrombolysis might increase the likelihood of ventricular rupture.117 Honan and associates116 performed a meta-analysis of four large clinical trials (1638 patients) in which streptokinase was used to treat AMIs and concluded that the risk of cardiac rupture was directly related to the timing of thrombolytic therapy. Early treatment (within 7 hours from the onset of symptoms) decreased the risk of cardiac rupture, whereas late treatment (after 17 hours) increased the risk of this complication even though, surprisingly, the overall mortality rate was diminished when streptokinase was given late after acute infarction. In a prospective ancillary study of 5711 patients, Late Assessment of Thrombolytic Efficacy, Becker and colleagues108 were unable to show an increased risk of cardiac rupture in patients treated with recombinant tissue plasminogen activator 6 to 24 hours after the onset of symptoms. Thus, there is general agreement that early successful thrombolysis decreases the overall risk of cardiac rupture, probably by limiting the extent of necrosis, resulting in a nontransmural instead of transmural infarct, but the impact of late thrombolytic therapy on cardiac rupture remains unclear.

Diagnosis

The clinical picture of a subacute ventricular rupture is primarily that of pericardial tamponade, with pulsus paradoxus, distended neck veins, and cardiogenic shock. Although 5 to 37% of patients with AMI but no rupture may develop a pericardial effusion, echocardiographic signs that increase the sensitivity and specificity for cardiac rupture include effusion thickness greater than 10 mm, echodense masses in the effusion, ventricular wall defects, and signs of tamponade (eg, right atrial and right ventricular early diastolic collapse and increased respiratory variation in transvalvular blood flow velocities).118,119 Pericardiocentesis and aspiration of uncoagulated blood have historically been considered the most reliable criteria of subacute ventricular rupture; however, false-positive and -negative diagnoses have been reported. The demonstration of a clear pericardial fluid on pericardiocentesis definitively excludes cardiac rupture.118 Pericardiocentesis is of therapeutic value in some patients, often providing short-term circulatory improvement.

In an attempt to define symptomatic, electrocardiographic, and hemodynamic markers that may permit the prospective identification of patients prone to rupture of the heart after AMI, Oliva and colleagues104 retrospectively studied 70 consecutive patients with rupture and 100 comparison patients with AMI but without rupture. They found a number of markers that were associated with a significant increase in the risk of rupture (Table 28-4). The presence of a lateral infarction, especially with associated inferior or posterior infarction, identified a subset of patients at increased risk for rupture. Persistent, progressive, or recurrent ST-segment elevation, and especially persistent T-wave changes after 48 to 72 hours, or the gradual reversal of initially inverted T waves, are associated with an increased risk of rupture. Finally, the development of pericarditis, repetitive emesis, or restlessness and agitation, particularly two or three of these symptoms, conveyed predictive value.105

Natural History

Acute rupture of the free wall of the left ventricle is invariably fatal, with death usually occurring within minutes of the onset of recurrent chest pain.107,109 In most of these cases, the sequence of events leading to death is so rapid that there is not enough time for surgical intervention. In contrast, patients with a subacute rupture usually survive hours or days, rarely weeks, following the myocardial tear. Pollack and colleagues119 found that in 24 patients with postinfarction subacute rupture, survival time (ie, time from critical event to death) varied between 45 minutes and 6.5 weeks, with a median survival of 8 hours. Núñez and coworkers120 found that in 29 patients with subacute rupture, 20 (69%) died within minutes of the onset of symptoms, and 9 (31%) lived several hours, allowing time for treatment. Subacute ventricular ruptures are generally considered to be less common than acute free wall ruptures. In recent studies, with high autopsy rates, 21 to 42% of all postinfarction free wall ruptures followed a subacute course.119,121,122

Because of its rarity, the natural history of false aneurysm of the left ventricle has not been well established. It is believed to have a poor prognosis because of its high probability of rupture;122,123 however, there are patients in whom the diagnosis was made many years after MI.124,125 The increasingly wide application of echocardiography after AMI gives promise of altering clinical outcome for many patients with the various forms of ventricular wall rupture.

Preoperative Management

Usually the sequence of events leading to death is so rapid in patients with acute rupture of the free wall of the left ventricle that there is not enough time for surgical intervention. These patients typically die within minutes of the onset of recurrent chest pain.120 However, a high index of suspicion, emergent surgery, perhaps combined with a novel technique of percutaneous intrapericardial infusion of fibrin glue immediately after pericardiocentesis,126 may afford at least a chance of survival in this surgically untreatable subgroup of patients.

In contrast, patients with subacute left ventricular rupture can be saved with surgery. Once the diagnosis of rupture of the free wall is established, the patient should be immediately transferred to the operating room. No time should be wasted attempting to perform coronary angiography.108,110,125 Inotropic agents and fluids should be started while preparing for surgery. Pericardiocentesis often improves hemodynamics temporarily, and insertion of an intra-aortic balloon pump may be beneficial, even though the principal problem is cardiac tamponade.

The timing of surgery after the diagnosis of false aneurysm of the left ventricle is dependent upon the age of the MI. When a false aneurysm is discovered within the first 2 to 3 months after coronary infarction, surgery is urgently recommended after coronary angiography and ventriculography because of the unpredictability of rupture.111,127 However, when the diagnosis is made several months or years after MI, the urgency of the operation is not determined so much by the risk of rupture, but rather by symptoms and the severity of the CAD.109,124

Operative Techniques

Subacute Rupture of the Free Wall

As soon as the diagnosis of rupture of the free wall of the left ventricle is confirmed by echocardiography, the patient should be transferred to the operating room. In patients with tamponade, severe hypotension may result during the induction of anesthesia. Therefore, we usually complete the sterile preparation and draping of the patient before inducing anesthesia, with preparations for rapid cannulation of the groin if needed. A median sternotomy is performed and upon decompressing the pericardium, the blood pressure commonly rises quickly. This should be anticipated and controlled because hypertension can cause ventricular bleeding to start again, or may even increase the size of the ventricular rent. In most cases, however, the ventricular tear is sealed off by clot, and there is no active bleeding.

Traditionally postinfarction rupture of the free wall of the left heart has been repaired on cardiopulmonary bypass; however, some surgeons have suggested that cardiopulmonary bypass is not necessary except perhaps in patients with posterior wall rupture, severe mitral regurgitation, ventricular septal rupture, or graftable CAD.109,127 Although the ventricular tear can be repaired without aortic cross-clamping, cardiac standstill and left ventricular decompression make the procedure easier when the rupture is in the posterior wall of the left ventricle.109

The term ventricular double rupture has been applied when there is rupture of two of the following three structures: the ventricular septum, ventricular free wall, and papillary muscle.

Four surgical techniques have been used to control ventricular rupture. The first technique involves closing the rent with large horizontal mattress sutures buttressed with two strips of Teflon felt.101 This method is not recommended because the sutures are placed into necrotic, friable myocardium that can easily tear. The second method combines infarct excision and closure of the defect with interrupted, pledgeted sutures102,128 or a Dacron patch.129 This method usually requires aortic cross-clamping and is probably best reserved for those patients who have an associated VSD. The third technique, described by Núñez and colleagues,120 involves closing the defect with horizontal mattress sutures buttressed with two strips of Teflon felt, and then covering the closed ventricular tear and surrounding infarcted myocardium with a Teflon patch sutured to healthy epicardium with a continuous polypropylene suture (Fig. 28-17). Good control of active ventricular hemorrhage has been achieved with this method. The fourth method consists of simply gluing a patch of either Teflon127 or autologous glutaraldehyde-preserved bovine pericardium to the ventricular tear and infarcted area using a biocompatible glue of either fibrin (Tissucol, Immuno AG, Vienna, Austria), butyl-2-cyanoacrylate monomer (Histoacryl Blue, B. Braun, Melsungen AG, Germany), or gelatin-resorcin-formaldehyde (Pharmacie Centrale, C.H.V. Henry Mondor, Créteil, France). This technique does not necessarily require institution of cardiopulmonary bypass and, because of its simplicity, may be the repair of choice when the ventricle is not actively bleeding.

False Aneurysm of the Left Ventricle

Acute false aneurysms are probably best repaired with an endocardial patch using the same methods as those used in repairing true ventricular aneurysms.109 Chronic anterior false aneurysms can usually be closed primarily if the neck is fibrotic. However, primary closure of the neck of a posterior false aneurysm may exacerbate mitral regurgitation, and it therefore probably should be reconstructed with a patch of Dacron graft or glutaraldehyde-fixed bovine pericardium.110,111

Results

Subacute Rupture of the Free Wall

The surgical experience with this entity is largely anecdotal. The single largest experience with surgical repair of postinfarction left ventricular free wall rupture was reported by Padró and colleagues.127 They treated 13 patients using a Teflon patch glued onto the ventricular tear and surrounding infarcted muscle, and used cardiopulmonary bypass in only one patient who presented with a posterior defect. All of their patients survived and were alive after a mean follow-up of 26 months. Eleven of them were asymptomatic and two had exertional angina. Núñez and associates120 operated on seven patients, four of whom survived. Recently the use of unsupported felt secured with cyanoacrylate glue has been described by several authors with encouraging results.130,131

In ventricular double rupture involving the free wall and the septum, a recent series achieved an operative survival of 60% in five patients.132 Four of the patients initially had only septal rupture but progressed to include free wall rupture, making this diagnosis a consideration in acute deterioration of medically managed patients with postinfarct VSD.

Although operative risk cannot be determined from these small numbers, it is likely that without surgery all these patients would have died, and patients who survive surgery tend to do well afterward.109

False Aneurysm of the Left Ventricle

Komeda and David110 treated 12 patients with postinfarction left ventricular false aneurysms; four of them also had mitral valve replacements, one had repair of a fistula between the false aneurysm and the right ventricle, and nine had coronary artery bypass surgery. There were three operative deaths, all in patients who needed mitral valve replacements. Of the eight patients who underwent isolated repair of false aneurysms, all were alive after a mean follow-up of 62 months. Seven patients were asymptomatic and one had angina pectoris. Overall, the literature suggests that patients who have isolated repair of false aneurysms of the left ventricle have low operative mortality.