Mitral Valve Replacement and Limited Myectomy for Hypertrophic Obstructive Cardiomyopathy

Abstract

Hypertrophic obstructive cardiomyopathy is a dynamic obstruction of the left ventricular outflow tract caused by septal hypertrophy and systolic anterior motion of the mitral valve. When the condition cannot be controlled by medical therapy, the most frequently used surgical approach is left ventricular myotomy-myectomy. Mitral valve replacement (to correct another mechanism of obstruction) is another surgical option; however, its use for this condition is controversial. We review the long-term results of patients who underwent limited left ventricular myotomy-myectomy and mitral valve replacement at our institution.

Eighteen patients who had hypertrophic obstructive cardiomyopathy and severe mitral insufficiency underwent surgery between 1978 and 1983: 7 were men and 11 were women (mean age, 41.8 ± 10.5 years). Preoperatively, most of the patients (78.8%) were in New York Heart Association functional class III or IV. The operation consisted of a shallow myectomy of the hypertrophied septum and mitral valve replacement.

One patient died in the hospital (5.5%); 3 patients died later during follow-up. The remaining 14 patients are alive and in good condition (mean follow-up, 21.9 ± 1.7 years). Functional class improved postoperatively in all surviving patients. The mean left ventricular outflow tract gradient fell from 78.1 ± 20.9 mmHg preoperatively to 9.4 ± 5.2 mmHg postoperatively (P < 0.001).

At present, surgical treatment of hypertrophic obstructive cardiomyopathy does not include mitral valve replacement. However, our long-term results show that limited ventricular myectomy and mitral valve replacement predictably and consistently lower the left ventricular outflow tract gradient and resolve the mitral valve insufficiency.

Hypertrophic obstructive cardiomyopathy (HOCM) is characterized by obstruction of the left ventricular outflow tract (LVOT). This obstruction is caused by a hypertrophied interventricular septum and by abnormal motion of the anterior mitral valve leaflet, which further obstructs the LVOT and causes mitral valve regurgitation. Patients who have HOCM are usually treated medically with b-blockers or calcium channel antagonists; a small number of these patients fail to respond to medical therapy and become surgical candidates.

Surgery for HOCM, or septal myotomy, was first performed in the 1960s by Cleland ¹ and by Bigelow and colleagues. ² This technique, consisting of incision of the hypertrophied septum, was subsequently modified by Morrow and coworkers ³ and became the left ventricular myotomy-myectomy (LVMM). The LVMM is, at present, the most frequently used surgical approach for HOCM. ³ Mitral valve involvement in the genesis of the LVOT obstruction prompted Cooley and associates, ⁴ in the early 1970s, to include elective mitral valve replacement (MVR) in the treatment of HOCM.

Because of Cooley's innovation, and because of the unclear outcome of LVMM alone at that time—especially with regard to the resolution of mitral regurgitation —we decided to combine MVR with a limited LVMM in all patients who presented with HOCM and severe mitral valve regurgitation.

We analyzed the results of a small, specific group of such patients who had undergone MVR and a limited LVMM at our institution from 1978 through 1983. December 1983 was chosen as the data cut-off point to provide at least a 20-year period that we could use to evaluate the long-term outcome of MVR and its advantages and disadvantages in patients who have HOCM.

Patients and Methods

This study included 18 consecutive patients who had undergone surgery for HOCM in our institution from January 1978 through December 1983. There were 7 men and 11 women, with a mean age of 41.8 ± 10.5 years (range, 21 to 62 years). Most of the patients (78.8%) had severe symptoms (dyspnea, angina, palpitations, and syncope), were in New York Heart Association (NYHA) functional class III or IV, and were unresponsive or poorly responsive to medical therapy. Clinical data are listed in Table I. During the study period, echocardiography was not available on a routine basis; therefore, the diagnosis was made in the cardiac catheterization laboratory. All patients underwent preoperative cardiac cathe terization, which included right brachial cutdown. Measurements included left ventricular end-diastolic pressure, pulmonary capillary wedge pressure, cardiac index, and LVOT pressure gradient both at rest and during provocative maneuver. Patients who had a resting gradient greater than 100 mmHg did not undergo provocative testing (Table II). Left ventricular angiography and selective coronary angiography were performed routinely. Patients who survived the surgery underwent cardiac catheterization again before being discharged from the hospital (Table II).

TABLE I. Patient Characteristics

TABLE II. Pre- and Postoperative Hemodynamic Data

Surgical Technique

Standard cardiopulmonary bypass with bubble oxygenation was used for all patients. Ascending aorta and bicaval venous cannulation with moderate systemic hypothermia (28–32 °C) was used in all patients, as were myocardial protection with antegrade cold (4 °C) potassium crystalloid cardioplegic solution and topical cooling. An oblique aortotomy was performed, the aortic cusps were carefully retracted, and the interventricular septum was exposed. Two shallow (4–5 mm in depth) parallel incisions were made in the interventricular septum toward the left ventricular apex, and a superficial portion of the septum was resected. Then the left atrium was incised anterior to the right pulmonary veins, to expose the mitral valve. The valve was resected completely, and a mechanical mitral valve (Bjork-Shiley in 8 patients, St. Jude Medical in 4 patients) and a biological mitral valve (Liotta-Bioimplant in 6 patients) were implanted by means of pledget-reinforced horizontal mattress sutures of 2–0 Tevdek. The left atrium was closed with 3–0 Prolene suture. Concomitant aortic valve replacement was performed because of aortic regurgitation in 4 patients (Liotta in 3 patients, Bjork-Shiley in 1 patient) with use of the same technique. The aorta was closed with 5–0 Prolene suture. Air was carefully evacuated, and the aortic clamp was released.

Choice of Prosthesis

The choice of prosthesis is particularly important in patients who have HOCM. Because these patients have a small left ventricular cavity, only low-profile mitral mechanical prostheses can be used. We started with the Bjork-Shiley valve. Then in the 1980s, when we began using more biological valves, we chose the Liotta low-profile valve. The Liotta valve had the advantage of sitting mainly in the left atrium and did not obstruct the already obstructed LVOT; in fact, it contributed to enlarging the LVOT. ⁵

Patients who received a mechanical prosthesis were placed on a lifelong regimen of warfarin sodium, and prothrombin activity was maintained in the range of 25% to 35%. Only in the patients with a Liotta bioprosthesis in the mitral position, warfarin sodium was given for 3 months postoperatively and then withdrawn.

Follow-Up

Follow-up ranged from 19.6 years to 25.1 years (mean, 21.9 ± 1.7 years). All 14 surviving patients returned for an out-patient visit during the 1st half of 2003. During this visit, hemodynamic data were evaluated with echocardiography, because it is noninvasive, and because echocardiographic and hemodynamic data can be used interchangeably. ⁶ The systolic pressure gradient across the LVOT was assessed quantitatively by Doppler echocardiography.

Definitions

The definitions of events, their classification, and the analysis of data are expressed in accordance with the guidelines for reporting morbidity and mortality after cardiac prosthesis surgery. ⁷ Operative mortality is defined as death in the operating room, in the hospital, or within 1 month of the procedure. Late mortality is defined as death beyond 1 month after surgery.

Statistical Methods

Values are given as mean ± SD. Patient survival was determined by Kaplan-Meier actuarial analysis and expressed as a percentage of those who were event-free ± SE. Hospital deaths are included in the survival curves. Comparison of values was performed with the Student's t-test. A P value of 0.05 was considered significant. Calculations were performed with use of the Statistica 5.0 program (StatSoft, Inc.; Tulsa, Okla).

Results

One patient died in the hospital after surgery (5.5%). This patient had received a Bjork-Shiley mitral valve, was in NYHA class IV, had moderate pulmonary hypertension, and was in cardiac cachexia. Seventeen patients were discharged in good condition from the hospital with no heart block.

One patient with a mitral Bjork-Shiley valve died of cerebral hemorrhage 10 years after the operation. Another patient, with mitral and aortic Bjork-Shiley valves, had right brachial artery embolization 6 years postoperatively, was hospitalized, and was successfully treated. Eight years later, the patient died suddenly at home. Another patient with a Bjork-Shiley valve, 8 years after surgery, underwent reoperation for prosthesis malfunction. Examination of the valve revealed a thrombus and pannus on the left ventricular side, and the prosthesis was replaced with a St. Jude Medical valve. The patient is now in fairly good condition.

All 6 patients with biological valves underwent reoperation for valve failure within 6.3 to 9.3 years (mean, 7.9 ± 1.2 years) after the 1st operation. Freedom from reoperation is depicted in Figure 1. Four of these patients (1 with mitral and aortic bioprostheses and 3 with mitral bioprostheses) switched to mechanical valves (St. Jude). The other 2 patients (with mitral and aortic bioprostheses) preferred new biological valves (Carpentier). One of the 2 patients with biological mitral and aortic valves died 10 years later (19 years after the 1st operation) of congestive heart failure; the other is in good condition 12 years after reoperation (21 years after the 1st operation). Of the 4 patients who switched to a mitral mechanical valve at reoperation, 1 had a transient ischemic attack 5 years postoperatively; the other patient had a severe epistaxis 9 years postoperatively. Both of these patients survived. Therefore, late deaths occurred in 3 patients (17.6%). The mean follow-up period for all patients was 21.9 ± 1.7 years after the 1st operation (range, 19.6–25.1 yrs), resulting in a linearized mortality rate of 1.1%/pt-yr. In patients with mechanical valves, the linearized rate for bleeding was 0.7%/pt-yr (70% confidence interval, 0.2%–1.2%), and for thromboembolism was 1.0%/pt-yr (70% confidence interval, 0.4%–1.6%). The probability of survival with freedom from these complications is shown in Figure 2. Fourteen patients are alive and in good condition. In these patients, the mean NYHA functional class improved from 3.1 ± 0.7 preoperatively to 1.8 ± 1.0 postoperatively. The mean NYHA class was 2.1 ± 0.9 at late follow-up (Table II). Actuarial survival of patients who underwent MVR for obstructive HOCM is shown in Figure 3.

Fig. 1 Actuarial freedom from reoperation for mechanical and tissue valves, according to the Kaplan-Meier method.

Fig. 2 Actuarial freedom from thromboembolism and bleeding in mechanical prosthesis (at 1st operation and reoperation), according to the Kaplan-Meier method.

Fig. 3 Actuarial survival curve according to the Kaplan-Meier method. Hospital death is included.

The mean, preoperative, resting LVOT gradient was 78.1 ± 20.9 mmHg (after provocative testing it rose to 88.7 ± 20.7 mmHg). After surgery, the mean gradient fell to 9.4 ± 5.2 mmHg (P <0.001). At late follow-up with Doppler echocardiography, the LVOT gradient was substantially unchanged (Table II). In our small group of patients, replacement of the mitral valve in combination with LVMM seemed to bring predictable and consistent abolition of LVOT gradients.

Discussion

Hypertrophic obstructive cardiomyopathy is characterized by asymmetric septal hypertrophy that, in association with abnormal motion of the anterior mitral valve leaflet, causes a dynamic obstruction of the LVOT. Patients with this entity are usually managed medically; however, about 10% to 15%, despite early clinical improvement, later fail to respond to medical therapy and become candidates for surgery. ^8,9 Treatment of this latter group of patients includes LVMM, ³ MVR, ⁴ dual-chamber pacing with right ventricular preexcitation, ^10,11 and nonsurgical reduction of the septum. ¹²

Morrow and colleagues ³ perfected the technique of septal myotomy and operated on hundreds of patients. The operation was directed toward the hypertrophied part of the interventricular septum; Morrow was firmly convinced that the mitral valve should be left intact. ¹³ Particularly in centers where not many patients are seen annually, the LVMM operation is technically difficult, the results are not easily reproducible, and complications are relatively frequent (including iatrogenic ventricular septal defect, complete heart block, and incomplete relief of the obstruction). These problems are of particular concern in patients with this difficult cardiomyopathy, wherein interventricular septal thickness and structure are variable, and complex associated mitral valve deformities are often present. ¹⁴

Because the mechanism of the dynamic obstruction to the LVOT in HOCM involves the anterior leaflet of the mitral valve—which not only exacerbates the obstruction but also produces mitral incompetence—Cooley ⁴ advocated replacement of the mitral valve to eliminate the dynamic obstruction. In 1970, he performed the 1st such operation, replacing the mitral valve with a Bjork-Shiley prosthesis. In our center, from 1978 through 1983, we treated patients who had HOCM and severe mitral insufficiency by performing a limited septal myectomy and mitral valve replacement. The follow-up period of approximately 20 to 25 years is long enough to draw a firm conclusion.

Our operative mortality rate was 5.5%. Delahaye and coworkers, ¹⁵ in their series of patients who underwent MVR and LVMM, found an operative mortality of 8.3%, and Fighali's group ¹⁶ had no instances of operative death. The mortality rates of LVMM alone are variable, even in centers with considerable experience: Krajcer and associates reported 4.7%; ¹⁷ Robbins and Stinson, 2.3%; ¹⁸ and Heric and colleagues, 4% for LVMM alone but 8% with an associated valve procedure. ¹⁹ McCully's group had no deaths in patients who underwent LVMM alone; however, the early mortality rate was 4.6% when other procedures were added. ²⁰ Of note, these operations took place at a later date than did ours; therefore, the patients most likely benefited from improved myocardial protection techniques and better perioperative management.

In our patients, mitral regurgitation and the LVOT gradient were completely abolished after surgery. The NYHA functional class also improved in our surviving patients. This improvement may mean that such patients fare well and that their health is similar to that of “normal” patients who have undergone MVR, once the LVOT gradient has been abolished by MVR. However, for our results to be meaningful, they must be compared with the results of less invasive surgery (LVMM alone) in centers with wide experience. Cooper and coworkers, ²¹ reviewing the cases of 482 patients who had undergone LVMM, found a significant reduction in the average LVOT peak gradient in 78% of the patients, and an improvement in the severity of mitral regurgitation in 51%. McCully and co-authors ²⁰ observed that the mean reductions in intraoperative rest and provoked gradients after myectomy were 67.8 ± 33.3 mmHg and 71.6 ± 37 mmHg, respectively. Heric and associates, ¹⁹ observing their results in 178 patients, found that the mean maximal preoperative LVOT gradient measured in 170 patients at rest or with provocation was 93 mmHg and decreased to a mean of 21 mmHg in the 133 patients for whom measurements were recorded. Of the 60 patients who had preoperative mitral regurgitation grades 3+ to 4+, 16 required MVR. However, it should be noted that 32 patients were returned to cardiopulmonary bypass for correction of mitral regurgitation or a persistent LVOT gradient. Robbins and Stinson ¹⁸ analyzed the results of 153 patients who had undergone LVMM and found a significant decrease in LVOT gradient; no patients exhibited a substantial residual LVOT gradient that was persistent over time. Those authors concluded that LVMM is safe, reproducible, and effective.

There was no heart block in our patients. This complication occurs in 5% to 10% of patients operated on for LVMM and is reported to be a strong predictor of reduced late survival by multivariate analysis. ¹⁹

Late deaths in patients with HOCM are usually sudden. ²² In our study, 1 patient died suddenly (linearized mortality rate of 0.3%/pt-yr). This patient had an embolic accident 6 years after surgery. His late sudden death (8 years after the accident) was likely the result either of another more massive embolism (possibly cerebral) or of arrhythmia. Of the other 2 deaths, one was due to congestive heart failure, which can result from the natural progression of HOCM, and the other was due to cerebral hemorrhage, which is a valve-related complication.

At present, in our center, left ventricular myotomy-myectomy is the procedure of choice for patients with hypertrophic obstructive cardiomyopathy. Nevertheless, the experience in our small group of patients has shown that mitral valve replacement should remain an option: it can be useful when structural anomalies of the mitral valve are present and when patients fail to improve after left ventricular myotomy-myectomy alone.