Author + information
- Received March 9, 2000
- Revision received August 23, 2000
- Accepted September 19, 2000
- Published online December 1, 2000.
- Randall C Starling, MD, MPH, FACC∗,* (, )
- Patrick M McCarthy, MD†,
- Tiffany Buda, RN†,
- James Wong, MB, BS, PhD∗,
- Marlene Goormastic, MPH‡,
- Nicholas G Smedira, MD†,
- James D Thomas, MD, FACC∗,
- Eugene H Blackstone, MD, FACC† and
- James B Young, MD, FACC∗
- ↵*Reprint requests and correspondence: Randall C. Starling, Department of Cardiology, The Cleveland Clinic Foundation, 9500 Euclid Avenue, Desk F25, Cleveland, Ohio 44195
The study was done to prospectively measure the echocardiographic, hemodynamic and clinical outcomes after partial left ventriculectomy (PLV).
Although PLV can improve symptoms of advanced heart failure, immediate postoperative echocardiographic findings remain abnormal.
Fifty-nine patients with cardiomyopathy and advanced heart failure underwent PLV and concomitant mitral valve surgery between May 1996 and December 1997. Thirty-nine percent were on inotropic therapy. All were New York Heart Association (NYHA) functional class III or IV. Mechanical circulatory support (LVAD) and transplant were provided for rescue therapy when hemodynamic compromise occurred. Patients were followed for a mean of 405 ± 168 days, and clinical, echocardiographic and hemodynamic measures were obtained preoperatively, immediately postoperatively, and at 3 and 12 months prospectively.
Comparing preoperative and 12-month postoperative values in event-free survivors, we found: NYHA functional class improved from 3.6 to 2.1, p < 0.0001; peak oxygen consumption increased from 10.8 to 16.0 ml/kg/min, p < 0.0001; LV ejection fraction increased from 13 ± 6.0% to 24 ± 6.9%, p < 0.0001; LV end diastolic diameter decreased from 8.2 ± 1.03 to 6.2 ± 0.64 cm, p < 0.0001, and volume was reduced from 167 ± 60 to 105 ± 38 ml/m2, p = 0.02. Central hemodynamics did not normalize after surgery.
Partial left ventriculectomy can provide structural remodeling of the heart that may result in temporary improvement in clinical compensation. However, perioperative failures and the return of heart failure limit the propriety of this procedure.
Heart failure is one of the most important public health problems in cardiovascular medicine (1,2). Its incidence is growing most rapidly in the population aged 65 years or older (3). Ten to fifteen percent of patients with heart failure have advanced symptoms despite optimized medical therapy and experience the highest 6- to 12-month mortality. Heart transplantation is the only treatment that makes a substantial long-term impact. However, there is an inadequate supply of donor hearts.
Motivated by the inordinate demand versus supply, partial left ventriculectomy (PLV), also known as “heart reduction surgery” or the Batista procedure, was developed to reduce the size and reshape the heart to improve mechanical function. Based on the Law of La Place, Batista proposed that wall stress would be reduced by surgical reduction in left ventricular radius and that the left ventricular volume-to-mass relationship would be normalized (4,5). The procedure was enthusiastically embraced internationally by patients and physicians, hopeful that it represented an alternative to cardiac transplantation (6). Despite encouraging short-term results, long-term outcomes remain undefined (7–11). Therefore, this prospective analysis characterizes the patients selected for PLV and describes the echocardiographic, exercise, and hemodynamic results at 3 and 12 months postoperatively.
From May 23, 1996, until December 3, 1997, a total of 59 patients underwent PLV at The Cleveland Clinic Foundation, Cleveland, Ohio. This group was selected from over 2,000 patients referred with advanced heart failure. An a priori hypothesis formulated our selection criteria: patients with dilated cardiomyopathy with intact ventricular muscle mass devoid of areas of extensive scarring will derive the greatest benefit from the operation. Thus, we chose to exclude patients with defined coronary artery disease and evidence of prior myocardial infarction. Patients were chosen who were diagnosed with dilated cardiomyopathy on the basis of a recent coronary angiogram (within three years from date of surgery) and no history of anthracycline use or other known cardiotoxic agents. A left ventricular end diastolic diameter of >7.0 cm determined by echocardiographic measurements was required. Patients were not excluded on the basis of severity of illness, requirement for inotropic therapy or pulmonary hypertension.
All patients were evaluated and considered suitable candidates for cardiac transplantation except for three elderly individuals with comorbidities. This strategy served two purposes: 1) it assured that all patients referred for PLV had advanced refractory heart failure without other available treatment options, and 2) patients listed for transplant were advised that mechanical circulatory support and cardiac transplantation would be made available as needed in the event that PLV was unsuccessful acutely or chronically. Informed consent was obtained, and all patients eligible for transplantation met standard listing criteria approved by The Cleveland Clinic Foundation Heart Transplantation Selection Committee and the Ohio Solid Organ Transplantation Consortium Heart Transplant Committee (12).
See Table 1and previous descriptions (10,11). One patient had been supported by the HeartMate implantable left ventricular assist device (LVAD) (Thermo Cardiosystems, Woburn, Massachusetts) for 88 days but had a device infection and evidence of improved ventricular function. Thirty-nine percent (23/59) were Status I awaiting cardiac transplantation according to the criteria of the United Network of Organ Sharing (UNOS), and 59% (35/59) were New York Heart Association (NYHA) functional class IV. Medications at the time of surgery and doses are listed in Table 2. All patients referred for surgery had been diagnosed with congestive heart failure at least six months before surgery (57.6 ± 49.2, range 6 to 240 months). All patients were deemed to have an optimized medical regimen by an experienced heart failure cardiologist from The Cleveland Clinic Foundation at the time of the preoperative evaluation. Neurohormones were measured in the fasting state with the patient supine on the morning of surgery. Samples were obtained in chilled EDTA specimen tubes and assayed using commercially available reagents and laboratories. Samples for tumor necrosis factor-alpha and intereleukin-6 were processed as previously described in the laboratory of Dr. D.L. Mann, Houston, Texas (13).
Baseline and follow-up measures
Echocardiographic measures, hemodynamics, metabolic stress testing and clinical assessments were performed preoperatively (generally <30 days before surgery), immediately postoperatively for most tests, and at 3 and 12 months postoperatively in survivors who returned for follow-up. Patients had transthoracic echocardiograms performed using an HP Sonos 2500 echocardiography machine to acquire M-mode and two-dimensional (2-D) images. Left ventricular volumes and ejection fractions were quantified from the apical four and long axis two chamber images using the biplane Simpson’s method. Left ventricular cavity dimensions were measured at end systole from 2-D images. The amount of mitral regurgitation was assessed qualitatively using a categorical grading from 0 (none) to 4+ (severe). Due to the characteristic structural changes in the mitral valve and the papillary muscles after surgery, the echocardiographic review could not be blinded. The echocardiographers were, however, blinded to the clinical outcomes of the patients. Right heart catheterization was performed with a triple lumen thermodilution 7.5F pulmonary artery balloon tip flow-directed catheter to measure right heart pressure. Cardiac output was measured by the Fick and thermodilution techniques. Metabolic stress testing was performed using a modified Naughton protocol (upright treadmill) that was symptom limited, while respiratory gas exchange was analyzed. Peak oxygen consumption, anaerobic threshold, exercise duration and respiratory exchange ratio were tabulated. Peak oxygen consumption and anaerobic threshold could not always be reliably determined. A respiratory quotient >1.1 at peak exercise was sought. Clinical follow-up was performed at The Cleveland Clinic when possible, and the heart-failure cardiologist performed a physical examination and assigned the NYHA functional class. Patients unable to return for follow-up were contacted by telephone, and clinical events and functional capacity were determined. Findings were corroborated by treating physicians when possible. Patients were treated postoperatively with angiotensin-converting enzyme (ACE) inhibitors, diuretics, digoxin and amiodarone when possible. Anticoagulation with warfarin was initiated before hospital discharge and maintained long-term with a target international normalized ratio of 2.0. Initially, electrophysiologic testing was event driven, and implantable cardiac defibrillators were used when indicated. After observation of the potential for increased sudden cardiac death after PLV, electrophysiologic testing was adopted as standard clinical practice.
Detailed surgical descriptions have been previously reported (10,11). The concomitant operative procedures are summarized in Table 3. Two patients required immediate mitral valve replacement because of residual 2+ mitral regurgitation observed intraoperatively by transesophageal echocardiography.
Data are presented as mean ± standard deviation. Comparisons of preoperative and immediate postoperative results were compared using a paired Student t test. To ascertain the stability of these early results, we utilized longitudinal data analysis techniques (PROC MIXED using SAS 8) and the best transformation of time interval from surgery to measurements that maximized goodness-of-fit (14). Each analysis included as fixed effects each patient’s baseline value and time of measurements, and as the random effect, time interval to measurement. Longitudinal data analysis methods were used in part because repeated measurements were, at times, missing values, and were also terminated by death or by placement of an assist device or transplantation, factors that can be accommodated by this methodology. We thus present information of the effect of preoperative level and postoperative trends and the time-related trend itself.
Perioperative and short-term clinical results have been previously described (10,11). Baseline neurohormones were obtained preoperatively the day of surgery: norepinephrine 906 ± 99 pg/ml (normal 80 to 520 pg/ml); atrial natriuretic peptide 370 ± 28 pg/ml (normal 25 to 77 pg/ml); TNF-alpha 6.3 ± 0.46 pg/ml (normal 0.75 ± 0.05 pg/ml); IL-6 19.0 ± 6.1 pg/ml (normal 1.8 ± 0.5 pg/ml).
The current results represent the complete analysis and follow-up of 59 consecutive patients operated on between May 1996 and December 1997 and followed until April 1998.
New York Heart Association (NYHA) functional class was assessed in 40 patients at three months postoperatively and in 26 patients at one year after surgery. The mean NYHA functional class at baseline was 3.6, fell to 2.2 at 3 months (p < 0.001) and was 2.1 at 12-month follow-up (p < 0.001 baseline vs. 12 months). Data points for the peak oxygen consumption is shown at baseline, at 3 months and at 12 months in Figure 1. Twenty-five patients were available for repeat testing at 3 months and 19 patients at 12 months. The metabolic exercise data are summarized in Table 4. Peak oxygen consumption, anaerobic threshold, and exercise duration increased postoperatively.
Suitable images for analyses were available for 58 patients at baseline, 35 patients at 3 months and 17 patients at 12 months. Serial echocardiographic measures and statistical values are shown in Table 5. Ejection fraction rose substantially after surgery; however, between then and 3 to 12 months, it decreased somewhat and reached a new plateau, still considerably above the preoperative value. Left ventricular internal diameter at end-diastole was decreased by operation, and this measurement on an individual patient basis only slightly declined in the 3- and 12-month assessments. Similarly, left ventricular end-diastolic volume index (LVEDVI) fell considerably with surgery and then increased somewhat by 3 and 12 months. Postoperative measures of mitral regurgitation at 3 months showed a marked decline (mean grade +0.63 ± 0.81, p < 0.001), which was maintained at 12 months (+0.53 ± 0.7).
Twenty-nine and 18 patients, respectively, underwent repeat assessment of hemodynamics at 3 and 12 months postoperatively. Systemic blood pressure rose after operation, and rose further from this level at 3 and 12 months (Table 6). The higher the baseline pressure, the more pronounced was this change (p = 0.008). Heart rate rose initially, but then fell to levels below baseline, ostensibly related to amiodarone therapy. Cardiac index rose somewhat initially, but then fell back to baseline levels. Pulmonary artery pressures fell after surgery, then rose back to their baseline levels.
Early follow-up from four institutions has shown an improvement in functional class in survivors of partial left ventriculectomy and short-term survival that ranged from 63% to 82% (7–11,15). Improvements in subjective clinical status (NYHA functional class) and objective functional capacity (peak oxygen consumption) were observed in event-free survivors. These patients have seemingly benefited from a surgical treatment that has improved their clinical status.
Serial assessment of multiple variables is a unique and an important aspect of this investigation and provides pathophysiological insights. Although some patients improved clinically, echocardiographic parameters remained abnormal, and there was minimal beneficial change in the hemodynamics. Thus, the physiologic basis for the clinical improvement observed in this observational study and, parenthetically, in other reports is not apparent from the results of this investigation. Cardiac transplantation will normalize hemodynamics and ejection fraction. It is well established that prognosis in patients with heart failure is related to structural, hemodynamic and neurohormonal variables. The lack of sustained benefit is understandable as PLV surgery does not normalize physiologic parameters (echocardiographic and hemodynamics). Optimized medical therapy and close follow-up in a specialized heart-failure clinic can independently yield improved outcomes (17). Clinical improvement in a nonrandomized surgical trial must be viewed with skepticism. The placebo effect, even in a surgical trial, must be considered (18). Partial left ventriculectomy appears to provide palliative benefit of limited duration with significant early morbidity (LVAD, transplant) and mortality. Indeed, our extended follow-up shows the three-year time-related freedom for failure (death, LVAD, transplantation, NYHA functional class IV heart failure) is 26% (16). Gorcsan et al. (19) demonstrated that surgical “unloading” by PLV can reduce wall stress immediately postoperatively, but they cautioned that the findings were heterogeneous and might not be sustained.
Role of mitral insufficiency
Conjecture has focused on the role of mitral valve repair in determining the outcomes with PLV. A European center reported 22 consecutive patients undergoing PLV with nonischemic dilated cardiomyopathy who were analyzed based on degree of preoperative mitral regurgitation (20). The improvement in ejection fraction was similar in both groups and not related to the correction of mitral regurgitation. Bolling et al. (21) reported beneficial ventricular remodeling with mitral valve repair alone in patients with dilated cardiomyopathy. Extended follow-up over three years from our cohort of patients concludes that the degree of preoperative mitral regurgitation did not correlate with clinical outcome, indicating effects of PLV are separate from mitral valve repair (16).
This investigation was an initial feasibility study, and we did not anticipate that PLV would be a definitive curative procedure. Hence, we decided it was important to combine aggressive medical unloading therapy (vasodilators) and antiarrhythmic treatment with surgical therapy. Beta-blocker therapy can have independent beneficial remodeling effects leading to structural and hemodynamic improvement (22). Only 15% of the patients were treated with beta-blockers in the follow-up period. The addition of amiodarone postoperatively may have contributed to the observations (e.g., increase in ejection fraction, reduced heart rate); however, the role of amiodarone in the treatment of dilated cardiomyopathy remains unresolved (23). Longitudinal data analysis methods were applied due to unvoidable censoring (death, insertion of LVAD, or transplantation). Mixed modeling, which included both fixed effects of preoperative baseline levels and time, and the random effect of time were both assessed. The resulting trends reflect the stability of measures across time and their relation to preoperative values.
We have observed short-term clinical improvement in event-free survivors after PLV. Physiologic variables (echocardiographic and hemodynamics) measured serially show that cardiac function and anatomy remain markedly abnormal. Partial left ventriculectomy does not provide normalization of cardiac structure or function.
We sincerely thank Douglas L. Mann, MD, for providing the tumor necrosis factor-alpha and interleukin-6 levels that were processed in his laboratory. We acknowledge Jennifer A. White, MS, for statistical review and analyses.
☆ No financial support was received to support this investigation.
- left ventricular assist device
- partial left ventriculectomy
- Received March 9, 2000.
- Revision received August 23, 2000.
- Accepted September 19, 2000.
- American College of Cardiology
- Ho K.K.L,
- Pinsky J.L,
- Kannel W.B,
- Levy D
- Kannel W.B,
- Ho K,
- Thom T
- Bocchi E.A,
- Bellotti G,
- Vilella de Moraes A,
- et al.
- McCarthy J.F,
- McCarthy P.M,
- Starling R.C,
- et al.
- Costanzo M.R,
- Augustine S,
- Bourge R,
- et al.
- Torre-Amione G,
- Kapadia S,
- Benedict C,
- Oral H,
- Young J.B,
- Mann D.L
- Diggle D.J,
- Liang K.Y,
- Zeger S.L
- ↵Franco-Cereceda A, McCarthy PM, Blackstone EH, et al. Part left ventriculectomy is not an alternative to transplantation in end-stage cardiac failure patients with idiopathic dilated cardiomyopathy. J Thorac Cardiovasc Surg. In Press.
- Fonarow G.C,
- Stevenson L.W,
- Walden J.A,
- et al.
- Leier C.V
- Gorcsan J 3rd.,
- Feldman A.M,
- Kormos R.L,
- Mandarino W.A,
- Demetris A.J,
- Batista R.J
- Popovic Z,
- Miric M,
- Gradinac S,
- et al.
- Doughty R.N,
- Whalley G.A,
- Gamble G,
- MacMahon S,
- Sharpe N
- Massie B.M,
- Fisher S.G,
- Deedwania P.C,
- Singh B.N,
- Fletcher R.D,
- Singh S.N