Author + information
- Received April 30, 1998
- Revision received August 11, 1998
- Accepted September 10, 1998
- Published online January 1, 1999.
- Naser M Ammash, MD, FACCa,* (, )
- Heidi M Connolly, MD, FACCa,
- Martin D Abel, MDa and
- Carole A Warnes, MD, MRCP, FACCa
- ↵*Address for correspondence: Dr. Naser M. Ammash, Senior Associate Consultant, Division of Cardiovascular Diseases and Internal Medicine, West 16B, Mayo Clinic, 200 First Street SW, Rochester, Minnesota 55905
Objectives. The purpose of this study was to evaluate the morbidity and mortality associated with noncardiac surgery (NCS) in patients (pts) with Eisenmenger syndrome.
Background. Noncardiac surgery in pts with Eisenmenger syndrome is associated with increased cardiovascular complications.
Methods. Fifty-eight pts with Eisenmenger syndrome (17M, 41F aged 18 to 69 years (mean 41 years) who had been followed for up to 41.5 years (mean 9.3 years) were retrospectively evaluated for any NCS done at ≥17 years of age.
Results. Twenty-four pts had a total of 28 NCSs at an age of 17 to 55 years (mean 29 years) including 9 tubal ligations, 3 neurosurgeries, 3 cholecystectomies, 3 hysterectomies, 3 vasectomies, and 1 each spinal fusion, appendectomy, eye enucleation, hernia repair, hand surgery, tonsillectomy and therapeutic abortion. There were two deaths (7%), one following spinal fusion and the other following appendectomy at another institution.
Fourteen of these NCSs were performed at our institution, including 11 under general anesthesia. The duration of anesthesia varied from 75 to 525 min (mean 165 min). All pts remained in sinus rhythm. The lowest systolic blood pressure (BP) ranged from 78 to 125 mm Hg. Of those 11 pts, 9 were extubated immediately after surgery and 2 needed dopamine. Ten patients were discharged without any complications, including 3 within 1 day of surgery. One death occurred 10 days following spinal fusion. This pt had the longest anesthesia (525 min) and an intraoperative systolic BP as low as 78 mm Hg. She also needed the largest fluid administration (6,475 cc) in addition to postoperative mechanical ventilation and dopamine.
Conclusions. Adult pts with Eisenmenger syndrome are at increased risk with NCS, but with current/modern techniques, the risk of death is less than previously thought. In the vast majority of cases, NCS can be undertaken without substantial morbidity, and early extubation is achievable. However, even with relatively minor surgery, significant complications, including death, can occur. Referral to major centers with expertise in the care of pts with Eisenmenger syndrome is advisable.
A growing number of adult patients (pts) with congenital heart disease are surviving into adulthood and some into the sixth and seventh decade (1). As a result, many face the prospect of requiring noncardiac surgery (NCS) (2). To date, only very few generalized guidelines exist for the preoperative evaluation and perioperative care of such pts when undergoing NCS (2–6). Patients with irreversible pulmonary vascular obstructive disease, secondary to an intracardiac shunt (Eisenmenger syndrome) (7), are known to be at particularly high risk when undergoing NCS (8). This risk of NCS may relate to 1) the extent of the surgical procedure, 2) severity of pulmonary hypertension, 3) severity of tricuspid regurgitation and right ventricular dysfunction, and 4) additional acquired cardiac and systemic diseases such as ischemic heart disease, and renal dysfunction. These pts are very vulnerable to alteration in hemodynamics induced by anesthetics or surgery. This includes a minor fall in systolic blood pressure that can increase right-to-left shunting and possibly potentiate cardiovascular collapse. Additional surgical risks include increased risk of bleeding due to platelet dysfunction, arterial and venous thrombosis secondary to increased blood viscosity, paradoxical embolus and arrhythmia, both ventricular and supraventricular.
The purpose of this report was to review our experience to determine, first, the operative risk in pts with Eisenmenger syndrome undergoing NCS; second, the possible determinants of morbid events or adverse outcome in relation to the congenital defect, type of surgery and type of anesthesia; and third, to provide guidelines for physicians involved in the care of these complex pts in an attempt to minimize perioperative morbidity and mortality.
We retrospectively reviewed all pts who attended the Adult Congenital Heart Disease Clinic between 1988 and 1996. Medical records of all pts with Eisenmenger syndrome who had had noncardiac surgery were reviewed. Particular attention was given to age at the time of operation, hematological profile (hemoglobin, hematocrit, mean corpuscular volume, platelet count), creatinine, arterial oxygen saturation, cardiac rhythm on 12-lead electrocardiogram (ECG) and history of arrhythmias. In addition, the echocardiogram performed closest to the time of surgery was reviewed to obtain information on left ventricular ejection fraction, right ventricular size and function, and severity of tricuspid regurgitation. The following information was collected from the anesthesia records: type of surgery, mode of anesthesia and anesthesia used, duration of operation, sedating medication used, and perioperative monitoring such as central venous pressure line, arterial line and pulmonary artery catheter. The lowest intraoperative systolic blood pressure recorded intraoperatively was noted in addition to the total amount of fluid administered. Postoperative records were also reviewed with respect to the need for postoperative mechanical ventilation and inotropic support, as well as final outcome in terms of dismissal and in-hospital mortality.
A total of 58 pts with Eisenmenger syndrome (17 men, 41 women) age 18 to 69 years (mean 41 years) had been followed for up to 41.5 years (mean 9.3 years). During that period, 24 pts (age 17 to 55, mean 29 years) underwent NCS at age ≥17 years (Table 1). Fourteen of these operations were done on 11 pts at the Mayo Clinic and the remaining 14 operations were done in 13 pts elsewhere (Table 2). The NCS included 9 tubal ligations, 3 neurosurgeries, 3 cholecystectomies, 3 vasectomies, 3 hysterectomies, and 1 each spinal fusion, appendectomy, eye enucleation, hernia repair, hand surgery, tonsillectomy and therapeutic abortion.
The hematologic, renal and echocardiographic characteristics of the 11 pts who underwent 14 NCSs at the Mayo Clinic are summarized in Table 3. The operative records are summarized in Table 4. Thirteen NCSs were uncomplicated. Patients were generally kept fasting following midnight of the night of the operation and started on intravenous fluids to prevent hypovolemia. The average duration of fasting was 10.2 h. None of the pts were hypotensive at entry into the operating room. Four had mild hypotension (90 mm Hg) during induction of anesthesia. This was transient in two but necessitated 10 mg of ephedrine in the other two. Two additional pts became hypotensive during surgery. This was transient and mild in one pt (90 mm Hg) while it was more persistent and serious in the other (78 mm Hg) necessitating dopamine.
Four pts needed ≥1 liter of intraoperative fluid administration, two needed postoperative dobutamine and one pt needed postoperative mechanical ventilation. One pt with a patent ductus arteriosus died 9 days following spinal fusion. She had disabling scoliosis and, despite our considerable reluctance to proceed with surgery, was willing to accept the high surgical risk. This pt had the longest anesthesia time (525 min), the lowest intraoperative systemic pressure (78 mm Hg), necessitating the largest fluid resuscitation needed (6,475 cc) and ultimately needing inotropic and mechanical ventilation support for 24 h following her surgery. Her postoperative course was remarkable for a drop in hemoglobin from 16.9 to 10.3 g/dl, necessitating three blood transfusions. She also had episodes of respiratory distress, agitation and variable level of consciousness associated with decreased arterial oxygen saturation to 72% that was felt to be due to possible pulmonary embolism, in situ thrombosis or neurologic events. On her 10th postoperative day, she suddenly lost consciousness associated with oxygen desaturation and bradycardia, and she could not be resuscitated.
Of the 14 NCSs done elsewhere, our medical records showed that 1 pt died following appendectomy.
Perioperative cardiac morbidity is one of the leading causes of perioperative death in cardiac pts undergoing NCS. The guidelines for preoperative cardiovascular evaluation have been recently reviewed by Eagle et al. (9); however, these guidelines were intended for adults with acquired rather than congenital heart disease. Eisenmenger syndrome is an uncommon complication of many congenital heart diseases, both simple, such as ventricular and atrial septal defect, or complex, such as single ventricle (7,10). Many pts do not survive beyond the third and fourth decades, but in our experience, as well as others, some pts may survive into the sixth and seventh decade (8,10–12). The longer their survival, the greater is the likelihood of needing NCS such as cholecystectomy because of the increased risk of calcium bilirubinate gallstone secondary to red blood cell turnover. Eisenmenger pts are known to be at particularly high risk when undergoing NCS, with mortality rate reaching 30% (8). The severity of cyanosis, tricuspid regurgitation and right ventricular dysfunction of the pts might play an important role in determining the perioperative risk. The primary congenital defect might also have some prognostic significance. Both reported fatalities occurred in pts with Eisenmenger syndrome secondary to patent ductus arteriosus (at our institution) and aorto-pulmonary window (elsewhere). Both conditions represent the most direct transmission of systemic arterial blood to the pulmonary arteriolar bed and therefore might be associated with more severe pulmonary vascular obstructive disease.
Furthermore, acquired cardiac and systemic diseases such as hypertension, ischemic heart disease, lung disease, renal dysfunction, hematological disorders and electrolyte imbalance pose additional significant problems (9,13–17). Only one of our pts had hypertension; another had diabetes mellitus. Three had a history of supraventricular tachycardia controlled by medication, and one had ectopic atrial rhythm. All survived their surgeries. However, the only pt with known severe lung disease secondary to severe kyphoscoliosis (forced expiratory volume, 39% of predicted) accounted for the only mortality in our group.
The underlying pathology for which NCS is needed, as well as the urgency, duration of operation, and anesthesia used, affects the perioperative risk. Hemodynamic variables such as profound alteration in heart rate, blood pressure, vascular volume, oxygenation, and neurohormonal activation and perturbation adds an extra stress on an already abnormal cardiopulmonary system (4,9,13,15,16). Our study confirms that lengthy operations, associated with hemodynamic instability necessitating large volume fluid resuscitation, are linked with increased risk for perioperative mortality, even in a tertiary care center.
Preoperative evaluation of Eisenmenger pts should be performed by physicians experienced in managing these complex cases. Both a detailed history and a physical examination are necessary in addition to appropriate preoperative testing. These tests should include a 12-lead ECG, chest X-ray film, complete blood counts, blood chemistry, clotting studies, drug level (if applicable) and probably an echocardiographic assessment of left ventricular function, right ventricular size and function and severity of pulmonary hypertension and tricuspid regurgitation. In our group of pts, the majority had normal left ventricular function and mostly mild to moderate tricuspid regurgitation with moderate right ventricular dysfunction. Although the predictive value of preoperative ECG is controversial in the general population (13,18), this should be done in the Eisenmenger pts to establish the baseline rhythm. All except one pt in our group were in normal sinus rhythm and none had unexpected findings on their chest X-ray film.
Obtaining a preoperative hematologic profile in Eisenmenger pts is even more crucial because of the associated secondary erythrocytosis and its potential complications including cerebrovascular events, decreased level of clotting factors, and bleeding diathesis (19–22). The latter is secondary to platelet dysfunction and less so to thrombocytopenia (3,8,19,23,24). In our series, as expected, the mean and median hemoglobin and hematocrit values were elevated, the platelet counts were in the lower range of normal and the arterial oxygen saturation was below normal. In addition, abnormalities of the coagulation system have been described in cyanotic pts (4,19,22,25,26). When these tests are performed care must be taken to assess the hematocrit by automated technique to prevent plasma trapping seen with microhematocrit tubes (4,2,27)and to correct for the erroneously elevated prothrombin and partial thromboplastin time observed in these pts secondary to the elevated hemoglobin (2,4). Perloff and others (3,4,8)suggested that preoperative therapeutic phlebotomy with fluid replacement is indicated in pts with hyperviscosity syndrome and a hematocrit of more than 65% as this will improve the hemostatic defect and reduce the risk of intraoperative bleeding. Blood removed during phlebotomy should be kept for potential autologous transfusion.
In our series, two pts had a preoperative hematocrit of more than 65%. Only one pt needed preoperative phlebotomy prior to craniotomy because of symptoms of hyperviscosity, with a hemoglobin of >20 g/dl and a hematocrit of >70%. The other pt had an asymptomatic hematocrit of 65.9% and therefore phlebotomy was not recommended. These two pts had an uneventful postoperative course. The result of this preoperative evaluation must be reviewed with the surgical team and the anesthesiologist (preferably cardiac) in a collaborative approach and the decision communicated to the pt. Patients who are felt to be at very high risk should have the NCS canceled or postponed depending on its urgency; conversely, pts who are felt to have an acceptable risk should have their NCS planned very carefully.
Patients taking antiarrhythmic medications should receive them the day of surgery, but preferably diuretics should be withheld. Endocarditis prophylaxis should also be administered depending on the type of operation and in accordance with the American Heart Association recommendations (28). For abdominal and pelvic procedures, a laparoscopic approach can be considered, but the risk of uncontrollable bleeding should be considered particularly in the setting of a clotting diathesis. The choice of anesthetics and intraoperative monitors is best left to the discretion of an experienced anesthesia care team (2,4,9). Numerous studies have suggested that there is no one best myocardial protective anesthetic technique in these complex pts (9,29–31). Eisenmenger pts pose a difficult challenge as they have lost the ability to adapt to sudden changes in hemodynamics because of fixed pulmonary vascular disease (22). As a result, a minor fall in systemic blood pressure can increase right to left shunting and potentiate cardiovascular collapse (3–5,8,10), so the basic principle of any anesthestic technique chosen needs to maintain both cardiac output and systemic vascular resistance (5,32). Although epidural anesthesia has been successfully employed for minor surgeries such as tubal ligation and cesarean section (33,34), the resulting sympathetic blockade and decrease in both preload and afterload may be very hazardous, so general anesthesia is preferable for these pts (5,32). This may be achieved by combining a short-acting, intravenous (IV) narcotic such as fentanyl, which is usually well tolerated, in addition to a low-dose induction agent such as sodium thiopental or ketamine or inhalational agent, for example, isoflurane (5,32). Atracurium and vecuronium can be used as muscle relaxants because of their minimal effects on the cardiovascular system (5).
In addition to the choice of anesthesia technique, additional important measures need to be taken. All IV lines are a potential source of systemic embolization and should have filters placed along them. Furthermore, a pulse oximeter can provide continuous monitoring of the degree of veno-arterial mixing. Periodic arterial blood gas determinations facilitate assessment of acidosis, hypercarbia and hypoxia, which can increase the pulmonary vascular resistance and increase right to left shunting.
Irrespective of the kind of intraoperative monitoring used, the goal is to detect sudden changes in hemodynamics very early so as to initiate appropriate treatment and prevent further complications. Whether or not the pt should have an arterial line, central venous line or pulmonary artery pressure recording remains to be determined. Continuous intra-arterial monitoring of the blood pressure is a safe and reliable means of early recognition of sudden alterations in intravascular volume and hemodynamics, and this enables periodic arterial blood gas determination if desired. The pulmonary artery catheter can also provide critical information, but the benefit should be balanced against the risk of complications from insertion and use (2–4,35). These include pulmonary artery rupture, ventricular arrhythmias, pulmonary embolization, plus the additional risk of infection and paradoxical embolism (5,35). In addition, cardiac output measurement by thermodilution technique can be spurious (5). Patients undergoing intermediate to high-risk procedures might warrant placement of a pulmonary artery catheter (with or without oximetry) (36). In our pt population, only two pts had a pulmonary artery catheter inserted, whereas three others had central venous pressure monitoring only and nine had arterial lines. None of these pts had catheter-related complications during insertion or intraoperatively. It is therefore recommended and probably safe to monitor Eisenmenger pts with a central venous pressure catheter and arterial line primarily for early detection of sudden changes in volume and pressure.
The postoperative care of cardiac pts undergoing NCS is as important as the preoperative evaluation and intraoperative management. Eisenmenger pts should be observed on a monitored bed because of their predisposition to develop ventricular and supraventricular tachycardia. In addition, depending on the co-morbid conditions, they might be prone to bradyarrhythmia and myocardial ischemia (2). Baum (4)and Perloff (3)recommended that Eisenmenger pts be observed in an intensive care unit overnight. Precautions must be taken to prevent venous stasis by early ambulation and by applying effective elastic stocking or periodic pneumatic compression. In addition, every attempt should be made to prevent hypovolemia; thus, meticulous attention to fluid balance is essential. Convalescent pts should also be advised to change position slowly until the risk of postoperative postural hypotension with secondary increase in right to left shunting abates (2–4). Furthermore, adequate pain management is crucial because the stress of postoperative pain can result in adverse hemodynamics and possibly hypercoagulable state (4,9). It has been suggested that patient-controlled analgesia techniques are associated with greater pt satisfaction and lower pain scores. However, most analgesics can cause a drop in afterload and, secondarily, more right to left shunting in pts with Eisenmenger syndrome. The use of such a delivery system possibly contributed to the unstable hemodynamics and the altered mental status in our only pt who died postoperatively and therefore might not be advisable.
Adult pts with Eisenmenger syndrome require special care while undergoing NCS. This is directed toward ameliorating their preoperative status and avoiding complications related to Eisenmenger syndrome and age-related cardiovascular and noncardiovascular diseases (4). Every attempt should be made to minimize the risk of perioperative morbidity by meticulously planning for the NCS, even minor procedures, far in advance and anticipating the problems inherent in these pts (3,37). This is best done in medical centers with expertise in the care of adult congenital heart disease.
These guidelines are intended to help physicians caring for these complex pts to make the appropriate judgments and decisions when facing the prospect of NCS. It is very important that the pts themselves are involved in the decision making. Our study does suggest that Eisenmenger pts with severe right ventricular enlargement and dysfunction facing a lengthy operation with anticipated large fluid shifts are at increased risk for perioperative complications. Conversely, the majority of minor surgeries such as tubal ligation and chlolecystectomy, which are usually done with a short period of anesthesia and without significant fluid shift, can be done safely with special care and minimal cardiovascular risks. Nonetheless, even supposedly low-risk surgery such as appendectomy can pose significant risk without meticulous care.
☆ No financial support was needed for this study.
- noncardiac surgery
- Received April 30, 1998.
- Revision received August 11, 1998.
- Accepted September 10, 1998.
- American College of Cardiology
- Perloff J.K.
- Webb G.D.,
- Burrows F.A.
- Perloff J.K.
- Mulhern K.M.,
- Mahoney L.T.,
- Bjornsen K.D.,
- Skorton D.J.
- Wood P.
- Liberthson R.R.
- Jamieson W.,
- Janusz M.,
- Miyagishima R.,
- Gerein A.
- Larsen S.,
- Olesen K.,
- Jacobsen E.,
- et al.
- Territo M.C.,
- Rosove M.H.,
- Perloff J.K.
- Ammash N.,
- Warnes C.
- Lumley J.,
- Whitwam J.G.,
- Morgan M.
- Asling J.,
- Fung D.