| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2001;71:S86-S91
© 2001 The Society of Thoracic Surgeons
a Division of Cardiothoracic Surgery, Department of Surgery, College of Physicians and Surgeons, Columbia University, New York, New York, USA
Address reprint requests to Dr Oz, MHB 7-435, 177 Ft. Washington Ave, New York, NY 10032
e-mail: mco2{at}columbia.edu
Presented at the Fifth International Conference on Circulatory Support Devices for Severe Cardiac Failure, New York, NY, Sept 15–17, 2000.
Abstract
Cardiac assist devices have become an important component of transplantation programs as they successfully bridge unsalvageable patients who would otherwise die. The indications for a device can still be classified into short-term and long-term situations. The short-term indications have expanded into areas such as postcardiotomy failure, high-risk cardiac operations, and acute myocardial infarction with results that were not previously possible in the absence of a well-established mechanical assistance program. Appropriate patient selection remains challenging and perhaps the most important attribute of a successful ventricular assist program. Although few exact criteria can define patients who are not eligible, several considerations and screening scales can help determine a particular patient’s suitability. Specific attention must be given to right heart function, neurologic status, existing infections, renal function, and transplantation suitability. The future of this field will not only be in technological advances with devices but in optimization of patient selection and expanding indications such as permanent replacement therapy.
Left ventricular assist devices (LVADs) have become a well-accepted treatment to bridge patients with severe heart failure to transplantation. The indications for this modality have expanded beyond just transplantation candidates who deteriorate while awaiting a heart. Since 1995, we have seen indications previously thought to prohibit LVADs become predominant sources of referrals. We also anticipate that the future indications will include heart failure patients who are not candidates for transplantation.
The next decade promises several technological advances and techniques to optimize this therapy; however, the success of the field will still be dictated by appropriate patient selection. Indeed, appropriate patient and device selection is equally as important as creating smaller, infection-resistant, and durable devices. To this end, several authors have created unique scoring systems to improve patient selection, and through these scales the intricacies of the preoperative profile have been better elucidated. Unfortunately, despite much better understanding in this arena, there are still no absolute criteria that can guide decisions on patient suitability. Ultimately, the decision is best made by an experienced team who is able to appreciate all the nuances of a particular patient.
This article is intended to provide a review of eligibility criteria and indications for receiving ventricular assistance. Particular focus will be paid to the criteria we follow at Columbia University. Although much of the future of this technology will be in permanent replacement therapy, we will focus on its use as a bridge to transplantation.
Indications
The current generic indication for device therapy is the presence of heart failure in patients approved for undergoing transplantation. However, this simple definition is misleading. The initial rationale for device use was to bridge patients who would have died awaiting a transplant. Thus, the chronic heart failure patients were treated. In our experience, devices have actually enabled an equally large population of patients with acute heart failure to survive until transplantation. In fact, support of patients with acute heart failure indications has not only bridged waiting list patients to transplantation but also enlarged the recipient pool. We analyzed 115 patients during 6 years (August 1993 to September 1999) who received devices as a bridge to transplantation. We specifically looked at patients who received LVADs within 24 hours of being listed for cardiac transplantation and defined them as receiving urgent LVADs. Of the whole population, 63% (n = 73) received urgent LVADs for failed coronary artery bypass grafting or percutaneous transluminal coronary angioplasty (n = 26), acute decompensation of chronic congestive heart failure (n = 26), acute myocardial infarction (n = 13), acute myocarditis (n = 7), or failed coarctation repair (n = 1). In this group of patients we could not demonstrate a statistically significant lower survival than the nonurgent patients (unpublished data). However, a group in Germany using a smaller cohort of patients was able to demonstrate improved survival in patients who did not require devices for emergent indications [1].
Postcardiotomy failure
We have addressed this issue by creating an LVAD bridge to transplant network for failed cardiotomy [2]. With the rationale that most centers have the capacity for short-term support but will require a specialized transplantation and LVAD program for optimal longer term management of these patients, this was accomplished by contacting all cardiac surgery centers within 300 miles. These centers were encouraged to contact us early in the course of postcardiotomy failure. Initial evaluation was performed that included maximization of short-term LVAD support, cardiac enzymes, cardiac function, and issues such as neurologic and renal function that may ultimately play a role in patients’ eligibility for receiving a transplant. We encouraged centers to place short-term devices rather than further damage the myocardium by multiple attempts at weaning. Provided significant bleeding was controlled, transfer to our center was performed within 72 hours of the initial surgical procedure. Upon arrival, repeat assessment was made to determine further therapy in terms of device weaning or bridging to transplant. Suitability for the device is described in both the algorithm (Fig 1) and in the patient selection below. A key point in managing these patients is both rapid transfer to the hub facility and quick implantation of an LVAD, if required.
|
Acute decompensation of chronic failure
Patients with long-standing heart failure who deteriorate are also candidates for mechanical support. Often these patients have been followed by a heart failure or transplant team but have not yet fit criteria as a transplant candidate. Several factors can initiate their deterioration and include new ischemic injury, arrhythmia, or infection.
Myocarditis
Acute myocarditis, especially in young patients, represents the one area that is more reliably a bridge to recovery than a bridge to transplant. Nonetheless, one should approach these patients with the understanding that a bridge to transplant may be required, and indeed, sometimes it is difficult to predict which patients will recover function [4]. Because of the greater likelihood of recovery in this population, short-term devices are often warranted early during support. Once this device is in place, the patient can follow a similar algorithm as the postcardiotomy population in deciding further therapy.
Chronic heart failure in transplantation candidates
This tends to be the most straightforward indication for device placement with generally very good results. Devices in this population are especially helpful, as the patient has a period of recovery for any heart failure–related end-organ impairment, thus significantly reducing the risk of transplantation [5].
Ventricular arrhythmias
Intractable ventricular arrhythmias represent an interesting and unique population for assist devices. These patients do not fit classic criteria because, in the absence of their arrhythmia, they have acceptable heart function. Patients should obviously be treated with cardiac defibrillators and aggressive pharmacologic regimens before considering this option. This device indication has been successfully reported in several patients [6–8]. In this population, biventricular support should be strongly considered owing to the loss of right ventricular function during the arrhythmia.
High-risk reparative cardiac operation
There are some patients with ischemic and valvular cardiomyopathies that may be spared transplantation if their underlying cause is surgically corrected. Unfortunately, these patients represent a population at very high risk for a cardiac surgical procedure. In these situations we routinely perform the procedure with LVAD backup, in case performing the intended procedure is not feasible or they are unable to be weaned off bypass. Before performing the operation the patient has been screened by the transplantation service so they may be listed to receive a donor heart should they require an LVAD.
Patient selection
Inappropriate patient selection is the most common cause of a high perioperative LVAD mortality. On the other hand, too conservative an approach may afford a better survival rate, but may be at the expense of some patients who may have benefited from device implantation.
Appropriate device selection also has societal implications. Left ventricular assist devices, although not as precious a resource as a donor heart, still use significant financial resources and must be used judiciously to justify their use to hospitals and insurance providers. Appropriately selected patients are able to reduce heart failure–related end-organ impairment, reduce the risk of transplantation, and ultimately improve utilization of scarce donor organs.
Several general criteria for implantable device insertion exist. Approval as a transplant candidate is required by the U.S. Food and Drug Administration, although this decision is challenging in the setting of acute cardiac failure in which one is not afforded a thorough evaluation. The generally advocated hemodynamic criteria for device placement includes systolic blood pressure less than 80 mm Hg (or mean arterial pressure < 65 mm Hg), pulmonary capillary wedge pressure more than 20 mm Hg, systemic vascular resistance more than 2100 dynes · s · cm-5, urine output less than 20 mL/h (in adults) despite diuretics, and a cardiac index less than 2 L · min-1 · m-2 despite maximal inotropic or intraaortic balloon pump support [9]. Some centers are more specific about the inotropic requirements and require the use of at least two agents at specified high doses [10]. In practice, several other factors must be taken into account (Table 1). We prefer to evaluate patients using a systematic approach that can be broken down into both cardiac and extracardiac factors.
|
Valve disease
Valve disease should play a role in patient selection, but fortunately, most problems can be addressed at the time of device implantation. Patients with congestive heart failure and aortic stenosis should first have their valve repaired in hopes of myocardial recovery. However, in the case of severe heart failure, an assist device should be available. The same can be said for aortic insufficiency, but special considerations also exist. Placement of a device in the setting of severe aortic insufficiency is challenging, as the device raises the mean arterial blood pressure while lowering the ventricular pressure. This, in turn, increases the regurgitant flow. The resulting circulation of blood negates the benefit of a device. If necessary, this problem can be addressed at the time of implant by simply suturing the valve leaflets together. Patients with a mechanical prosthetic aortic valve can also receive devices, but we recommend excluding the prosthetic valve from the systemic circulation by sewing graft material above the valve. This should reduce systemic embolism from the valve. Mitral stenosis is not a contraindication but should be addressed to maximize device filling. Mitral regurgitation should be insignificant during device support, but repair should be considered in case weaning is attempted, or a device malfunction occurs. This can most easily be accomplished with a simple edge-to-edge apposition of the mitral leaflets [14]. We previously repaired tricuspid disease, but because of poor results, we no longer advocate this practice. Tricuspid regurgitation should improve as pulmonary hypertension resolves and the heart is unloaded. Although never a contraindication, all patients need a thorough evaluation with a bubble study to assess foramen ovale patency. This is because hypoxia can result from a right to left shunt owing to left-sided unloading.
Coronary artery disease
There are several reasons to evaluate the extent of coronary artery disease in patients who may need ventricular assistance. First, at the time of implantation, coronary artery bypass grafting, especially to the right system, may be required to maximize function. This may not be true on the left side as there is early evidence of significantly decreased coronary blood flow on positron emission spectroscopy scanning and perhaps even an increased incidence of left-sided graft closure (unpublished data).
Another reason to address untreated coronary lesions is to help the patient’s quality of life by reducing episodes of angina. Finally, optimizing coronary flow may reduce the incidence of ischemic arrhythmias. In general, we do not revascularize patients unless left ventricular recovery is feasible. If grafts are previously present, we take special care to preserve them.
Arrhythmias
As mentioned previously, arrhythmias can represent a unique indication for device placement. If a patient has unresponsive recurrent ventricular arrhythmias, biventricular support should be implemented [15, 16].
Cardiac reoperation
Reoperation is not only a factor because of the obvious technical difficulties but is also a risk factor for mortality. In one large study of patients receiving the Thoratec device, reoperation greater than 30 days before LVAD was found to be one of only two risk factors for mortality by univariate analysis [17]. Our own experience also revealed reoperation to be a significant factor for mortality, with a relative risk of 1.8 [18]. Other series have not found reoperation to be a significant risk for mortality [10, 19]. Overall, although certainly not the most important criterion, reoperation appears to incur a significant risk in the perioperative period.
Neurologic
Evaluation of each patient’s neurologic status is one of the most important but difficult factors in deciding the appropriateness of LVAD support, especially as an increasing number of patients with short-term causes are receiving devices. Many of these patients are transferred to a specialized center either after operation or after a recent cardiac arrest, and neurologic evaluations are sometimes not feasible. In these cases it is ethically permissible to discontinue support if it is found after device placement that patients are unable to regain neurologic function. Ideally, patients will also have a psychiatric evaluation to determine how well they will tolerate mechanical support.
Infection
If right heart failure is the most important cause of morbidity in the perioperative period, then infection is the most important issue in the postoperative period. Increased chance of infection is an integral risk for every type of device. Ideally, patients will have negative results from blood cultures for the week before device implantation. Because of the functional T-cell deficiency in these patients, this is particularly true of fungal blood cultures. Any documented infections should be well controlled and ideally eliminated before implant. Although we do firmly believe these criteria need to be closely adhered to, we have been unable to identify either fever or an elevated white blood cell count as a risk factor [18]. Other studies have also been unable to demonstrate systemic manifestation of infection as being a risk factor [17]. However, two studies showed an elevated white blood cell count to be a risk for mortality, despite no association with fever [10, 20]. Lack of a correlation with outcomes in these populations is likely related to the strict adherence of not placing devices in patients with ongoing infections.
Renal
As with cardiac transplant recipients, patients bridged with devices should not be dependent on dialysis. In our experience renal failure is the strongest predictor for mortality. In general, we avoid implanting devices into patients whose creatinine is greater than 5 mg/dL. However, in acutely ill patients, creatinine is not the best indicator as it takes time for an elevation to occur, and thus the true status of one’s renal function can be misrepresented. Blood urea nitrogen has been shown to correlate with survival in one study [17]. These investigators demonstrated a 46% survival rate if blood urea nitrogen was more than 40 mg/dL, whereas the survival was 76% if the blood urea nitrogen was less than 20 mg/dL. We have found that urine output is a more immediate and direct indicator of renal function, and urine output of less than 30 mL/h despite diuretic use has been our most important risk factor. Of all the risk factors we studied, this was the single biggest risk factor, with a relative risk of 3.9 [18]. Fortunately, despite general renal impairment in this population, recovery of function during the support period is excellent [21–23].
Hepatic
Like the kidneys, liver function is frequently abnormal as a result of end-organ damage caused by heart failure. This is most problematic when hepatic function is altered enough to cause a deficiency in clotting factors as evidenced by a prolonged prothrombin time of greater than 16 seconds. Patients with coagulopathy do particularly poorly when given devices. This is primarily owing to their increased requirement for perioperative blood products. An increased transfusion requirement translates directly to increased right heart dysfunction. This can partially be alleviated by use of a serine protease inhibitor, but it is best addressed by aggressively correcting the coagulopathy with intravenous vitamin K and fresh frozen plasma before device implantation. Some authors have proposed that liver function, primarily evidenced by bilirubin, is the most predictive factor for patient survival [19]. These authors demonstrated decreasing survival with increasing bilirubin levels. Bilirubin as a predictor for survival did not reach significance in a larger multicenter trial that included many of the same patients as the previous study [17], although the smaller study had much lower average bilirubin levels then the larger study. Although excessively elevated bilirubin levels are associated with a higher mortality, the LVAD population as a whole does not frequently present with excessively high levels.
Pulmonary
Any severe pulmonary condition such as adult respiratory distress syndrome or obstructive or restrictive disease is a contraindication to mechanical support. Most patients will have pulmonary edema with subsequent suboptimal oxygenation, but this is expected to resolve once a normal cardiac output is restored. We have found ventilator dependence to be a risk factor associated with mortality, but this is not a contraindication.
Other factors
Other criteria that come into play include body surface area that, depending on the type of device available, may be a contraindication. Conditions such as malignancy that would make the patient an unacceptable transplantation candidate or limit life expectancy to less than 2 years should be considered a contraindication to mechanical support. Other comorbid factors including diabetes and peripheral vascular disease are not strict contraindications provided that the patient is still a transplantation candidate.
Screening systems
In an attempt to simplify the selection process, we and others have developed various criteria to aid in determining which patients will benefit most from mechanical support. No screening scale and virtually no single risk factor should serve as absolute criteria delineating a patient’s candidacy for ventricular assist.
The group in St. Louis has created a scoring system usng 21 clinical variables that have each been assigned a relative weight depending on the severity in relation to survival [10]. In their early experience they found that patients in the lower scoring group did not suffer a single mortality whereas those in the highest scoring group suffered the highest mortality. Their scale appears to be useful in helping predict the success of device support. However, no nondevice patients were examined, so it does not help direct a decision about which patients may not need a device.
Gracin and colleagues [21] have used Acute Physiology and Chronic Health Evaluation II (APACHE II) scores as a selection tool, and they applied this system to both patients receiving LVADs and those evaluated for but not receiving LVADs. They found that patients with a score between 11 and 20 were most likely to benefit from this intervention. Additionally, incremental increases in the APACHE II score were associated with increasing mortality.
We have also developed a screening scale but have chosen to focus on a few easily determined variables for rapid determination. Like the St. Louis group, our score helps predict a particular patient’s outcome with device support but does not help predict which patients do not need this intervention. We looked at seven factors and assigned them each a relative risk, which was then provided a score for each variable (Table 2) [18]. Two of the variables were not found to be helpful in predicting mortality. In our experience, patients scoring greater than 5 points were more likely to suffer a mortality.
|
Sometimes an error is made in trying too many interventions before an assist device. For example, intraaortic balloon pumps are often placed, although patients with severe nonischemic heart failure are highly unlikely to respond to this therapy. Once a patient has been determined to be an appropriate candidate, the team should move ahead within 12 hours to implant the device.
Conclusions
The success of an assist program is largely dependent on appropriate patient selection. This is best performed by an experienced team of surgeons, cardiologists, consultants, and nurses. Screening criteria will continue to evolve to aid in this difficult decision.
As devices have become more prominent and as centers have gained more experience, we have seen an expansion of the patient populations that can be successfully treated with an assist device. Ultimately, we expect to see a paradigm shift from a bridge to transplant to a bridge to recovery, or permanent replacement therapy.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
