| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
From the *Department of Anesthesiology, the Division of Cardiovascular Disease, the Division of Cardiac Surgery, and the Section of Biostatistics, Mayo Clinic and Mayo Foundation, Rochester, Minnesota.
Address correspondence and reprint requests to Martin D. Abel, MD, Department of Anesthesiology, Mayo Clinic, 200 First St. SW, Rochester, MN 55905. Address e-mail to abel.martin{at}mayo.edu.
Abstract
BACKGROUND: Cardiac surgery for carcinoid heart disease is complicated by hemodynamic instability secondary to carcinoid crises, cardiovascular dysfunction, and blood loss. The safety of vasopressors and the benefit of aprotinin during concomitant octreotide administration are uncertain.
METHODS: We reviewed the effects of vasopressors and aprotinin on octreotide administration and mortality by univariate analysis in 100 consecutive cases of cardiac surgery for carcinoid heart disease from 1985 to 2003. Because mortality declines were temporally related to the introduction of aprotinin, bivariate analyses were performed to identify other factors associated with mortality.
RESULTS: Carcinoid symptoms and hypotension were treated with octreotide (n = 89) and/or vasopressors (n = 93). Vasopressors were not associated with increased octreotide administration. Patients requiring epinephrine had higher mortality but also had worse preoperative New York Heart Association class, higher urinary 5-hydroxyindoleacetic acid levels, and increased blood transfusion requirements. Aprotinin (n = 54) was associated with decreased blood transfusion requirements, increased octreotide administration, but not mortality. Overall mortality was 13%, declining from 28% between 1985 and 1994 to 6% between 1995 and 2003. Mortality was associated with greater blood transfusion requirements and longer duration of cardiopulmonary bypass.
CONCLUSIONS: Vasopressors may be used in conjunction with octreotide in carcinoid patients. The increased mortality associated with epinephrine likely reflects selection bias rather than a primary adverse effect. The improved survival over time in carcinoid patients is multifactorial and unrelated to aprotinin administration, suggesting further inhibition of the kallikrein–kinin system has little added benefit for this outcome in the presence of octreotide.
Carcinoid tumors arise from enterochromaffin cells found in the small bowel. Tumor cells secrete an array of substances, including vasoactive amines such as serotonin, prostaglandins, and vasoactive peptides. The carcinoid syndrome occurs when these substances are secreted into the systemic venous system from a tumor metastasized to the liver or originating in the ovaries, lungs, or thyroid. Carcinoid crisis, characterized by flushing, hypotension, and bronchospasm, can be provoked pharmacologically by administration of epinephrine (1), norepinephrine (2), dopamine (3), isoproterenol (4), and calcium (5). Treatment of metastatic carcinoid disease with hepatic arterial embolization and administration of the somatostatin analog, octreotide, have emerged as effective therapies for symptomatic management. Octreotide has been demonstrated to control intraoperative carcinoid symptoms (6).
Carcinoid valvular disease results from exposure of cardiac valves to chronically elevated plasma serotonin levels. Valve leaflets become thickened and chordae shortened (Figs. 1 and 2). These lesions are usually limited to right-sided valves, typically resulting in tricuspid and pulmonary valve regurgitation and stenosis (Figs. 3 and 4). Left-sided cardiac involvement occurs in the presence of pulmonary tumors, right-to-left intracardiac shunts, or in the presence of overwhelming disease (7).
|
|
|
|
In this study, we analyzed data from 100 consecutive patients who underwent valvular surgery for carcinoid heart disease. Our aim was to evaluate whether vasoactive medications used during cardiac surgery would provoke carcinoid activity resulting in an increased administration of octreotide. We further sought to evaluate whether the use of aprotinin would attenuate carcinoid activity as demonstrated by less frequent use of octreotide. Finally, we analyzed whether octreotide and/or aprotinin usage impacted mortality.
METHODS
With approval of the Mayo IRB, data from patients with symptomatic carcinoid heart disease who underwent cardiac surgery from 1985 to 2003 were entered into a database designed to track clinical outcomes. The primary indication for surgery was treatment of carcinoid valvular disease, intracardiac metastatic carcinoid tumor resection, or biopsy. Patients medicated with preoperative octreotide continued their usual dosing schedule until the time of surgery.
The anesthetic technique for each case was determined by the supervising anesthesiologist. Anesthetic management typically included an opioid (fentanyl, mean dose 30 ± 2 µg/kg, or sufentanil, mean dose 9 ± 2 µg/kg) supplemented with volatile anesthetic, benzodiazepines, and muscle relaxants. Monitoring included transesophageal echocardiography and measurements of direct arterial blood pressure and either central venous pressure or pulmonary artery pressure.
Octreotide was administered intermittently as discrete IV injections to patients with suspected carcinoid symptoms or a sudden unexplained decline in venous return during cardiopulmonary bypass (CPB) believed to be secondary to venous dilation resulting from tumor-released vasoactive substances. Patients received repeat doses of octreotide when clinically indicated. The dose of octreotide was decided by the supervising anesthesiologist but was typically 500 to 1000 µg, though on occasion a smaller dose (50–100 µg) was used. Vasoactive medications were used to support the circulation when hypotension did not respond to octreotide or for treating low vascular resistance during separation from CPB. This included phenylephrine in incremental doses (40–100 µg) or as an infusion (10–50 µg/min), ephedrine in incremental doses (5–10 mg), or calcium chloride in a single dose (500–1000 mg). Dopamine infusions (2–6 µg · kg–1 · min–1) were initiated when additional inotropic support was needed. Epinephrine infusions (0.01–0.2 µg · kg–1 · min–1) were reserved for continued poor ventricular performance and hypotension that was unresponsive to other vasoactive medications, octreotide, and intravascular volume resuscitation.
With the exception of one patient treated in 1994, the use of antifibrinolytics was implemented during 1995 to reduce perioperative blood loss. The decision to use tranexamic acid or aprotinin, and the timing of the initial loading dose relative to sternotomy, was left to the discretion of the anesthesiologist. Aprotinin was administered using either the "full dose" (280 kallikrein inhibitory units (KIU) loading dose over 20–30 min, 280 KIU in the CPB circuit prime, and an infusion of 70 KIU/h continued for 2 h postoperatively) or "half dose" regimen. Anticoagulation for CPB was achieved with 300 U/kg of heparin plus 10,000 U in the CPB prime with the goal of an activated clotting time 
500 s or 750 s when aprotinin was given. After discontinuation of CPB, heparin was neutralized with protamine sulfate dose (0.013 mg/U heparin administered).
Data Collection
Demographic, preoperative, intraoperative, and mortality data were prospectively collected via an electronic anesthesia database. Data abstraction was by a trained registered nurse who reviewed the chart within 60 days of the index procedure. Anesthesia records were reviewed by two investigators (T.N.W., M.D.A.) to ensure the accuracy and completeness of the data. Preoperative information included New York Heart Association (NYHA) class, the use of preoperative octreotide, and urinary 5-hydroxyindoleacetic acid (5-HIAA) level. Intraoperative information included anesthetic technique, intraoperative octreotide dose and frequency of administration, the use of aprotinin or other antifibrinolytic drugs and dose, the number of units of packed red blood cells (PRBC) and non-PRBC blood products (e.g., cryoprecipitate, platelets, and fresh frozen plasma) transfused during surgery and the subsequent 48 h, the use and dose of vasoactive medications, the duration of CPB, and the duration of aortic cross-clamping. Mortality occurring within 30 days of the index surgery was recorded.
Statistical Analysis
Univariate analyses were performed to assess the association between the administration of different vasoactive medications or aprotinin with total intraoperative octreotide dosage and mortality. Because a large decline in mortality was temporally related to the introduction of antifibrinolytics into clinical practice in 1995, a series of bivariate logistic regression analyses were performed to assess the association of other patient and procedural characteristics with mortality after adjusting for the time periods from 1985 to 1994 and 1995 to 2003. Data are summarized for each time period (1985–1994, 1995–2003) using the median and the 10th and the 90th percentiles for continuous variables and percentages for categorical variables. Only the index surgery is included for analysis. Patient and procedural characteristics were compared between time periods using the rank sum test or Fisher's exact test as appropriate. Characteristics that were continuous (i.e., urinary 5-HIAA) or ordinal (i.e., NYHA Classification) were treated as continuous variables in the logistic regression analysis. Characteristics that were categorical were modeled using a binary indicator variable to indicate the presence of the given characteristic. Because of the relatively small sample size and low frequency of adverse perioperative events, additional multivariate analyses of characteristics associated with adverse outcomes were not performed. In all cases, a two-tailed P 
0.05 was considered significant.
RESULTS
Over an 18-yr period, 100 consecutive patients underwent 104 surgeries for carcinoid heart disease. The medical record was incomplete in one patient and it was excluded from further analysis. Table 1 describes patient and procedural characteristics during the two time periods. Thirty-two patients underwent surgery in the 1985 to 1994 period and 68 underwent surgery from 1995 to 2003. There was an overall decline in mortality between the two time periods (28% vs 6%, P = 0.004).
|
Octreotide Administration
Ninety-three patients were taking octreotide preoperatively. Octreotide was used to treat carcinoid symptoms in 89 patients with a median intraoperative dose of 1500 µg (mean 3666 ± 6461 µg, range of 50–54,000 µg) and usually had a salutary effect. Patients receiving octreotide preoperatively received higher intraoperative octreotide doses (median 1500 vs 0 µg, P = 0.007) and were more likely to receive intraoperative octreotide (88 of 92 vs 2 of 7, P < 0.01).
Vasoactive Medications Administration
Ninety-three patients were treated with vasoactive medications with 65% of patients receiving calcium, 49% dopamine, 45% phenylephrine, 17% ephedrine, and 11% epinephrine. Sixty percent of patients received two or more vasoactive medications. Preoperative characteristics of patients who received different vasoactive medications did not differ with the exception of epinephrine. Patients treated with epinephrine infusions tended to have worse NYHA class scores (median 4 vs 3, P = 0.04), higher urinary 5-HIAA levels (median 266 vs 141 µg, P = 0.03), more PRBC units transfused (median 2 vs 4 U, P = 0.01), and more non-PRBC units transfused (median 7 vs 20 U, P = 0.05). The use of vasoactive medications was not associated with higher intraoperative doses of octreotide nor increased mortality, with the exception of epinephrine, which was associated with higher mortality (Table 2).
|
Aprotinin Administration and Blood Transfusion Requirements
Since December 1995 all patients except one received antifibrinolytic therapy (Table 1). The full dose of aprotinin was used in 35 patients and the half dose of aprotinin was used in 19 patients. The loading dose of aprotinin was administered before sternotomy in 18 cases and after sternotomy in 36 patients. Besides transfusion requirements, intraoperative characteristics did not vary with the exception of longer aortic cross-clamp time in patients who were not administered antifibrinolytics (median 47 vs 23 min, P = 0.01).
Patients receiving aprotinin required less PRBC units than those not receiving an antifibrinolytic (median 1 vs 4 U, P < 0.001). PRBC transfusions were negatively correlated with preoperative hemoglobin concentration (r = –0.29, P = 0.004) and positively correlated with the duration of CPB (r = 0.46, P < 0.001) and aortic cross-clamping (r = 0.29, P < 0.006). Similarly, non-PRBC products were transfused less in patients receiving aprotinin compared with no antifibrinolytic (median 0 vs 10 U, P < 0.001) and the number of non-PRBC units transfused was positively associated with CPB (r = 0.48, P < 0.001) and aortic cross-clamp times (r = 0.26, P = 0.01).
Patients receiving aprotinin had higher median intraoperative octreotide doses than patients who did not receive an antifibrinolytic (median 2000 vs 1000 µg, P = 0.04). Intraoperative octreotide administration was also greater in patients who received full dose compared with half dose aprotinin, but this did not reach significance (median 4000 vs 1500 µg, P = 0.07). Timing of the aprotinin initial loading dose compared with sternotomy was not associated with median intraoperative octreotide dose (1925 vs 2000 µg, P = 0.74). Five patients who received aprotinin died including the patient who received aprotinin in 1994. The overall mortality between patients who were administered aprotinin compared with those who did not receive an antifibrinolytic did not differ (8 of 36 vs 5 of 54, P = 0.13).
Mortality
There were 13 perioperative deaths. The cause of death was attributed to multiorgan failure in two patients, heart failure in 10 patients, and complications secondary to metastatic carcinoid disease in one patient. There were no differences in 30-day mortality between patients having pulmonary valvotomy compared with pulmonary valve replacement. There were no deaths the day of surgery. The median day of postoperative death increased from three during the 1984 to 1994 time period to 11.5 days since 1995 (P = 0.05). Mortality was more likely in patients with higher urinary 5-HIAA level, longer duration of CPB, higher number of blood products transfused, and use of epinephrine. Bivariate logistic regression analyses to adjust for the time period of surgery revealed that longer duration of CPB, a larger number of blood product units transfused, and epinephrine use were statistically associated with mortality (Table 2).
DISCUSSION
This study is a comprehensive analysis of 99 patients who underwent cardiac surgery for carcinoid heart disease. A substantial decline in mortality was noted after 1995 after implementation of changes in perioperative management that included the use of aprotinin, an increased frequency of pulmonary valve replacement rather than valvotomy, shorter duration of CPB and aortic cross-clamping, and fewer patients requiring aortic cross-clamp.
The clinical introduction of octreotide has provided the anesthesiologist with an effective tool to manage manifestations of carcinoid activity (6). Octreotide controls carcinoid symptoms by inhibiting carcinoid tumor cells from secreting vasoactive peptides and amines into the circulation. In this study, we made the assumption that the total intraoperative octreotide dose was a reflection of carcinoid activity. This assumption is supported by the observation that patients not administered octreotide preoperatively were less likely to be administered octreotide intraoperatively, suggesting less carcinoid activity. Currently, there is no other quantitative measure of intraoperative carcinoid activity. However, the intermittent octreotide dosages were left to the discretion of the anesthesiologist, introducing a source of bias. Also, total intraoperative octreotide dose cannot be used as a surrogate measure for hemodynamic instability. Hemodynamic instability in these patients is multifactorial, consisting of carcinoid activity, underlying cardiac disease, and cardiac functional changes associated with CPB. The fact that intraoperative octreotide administration did not vary with mortality suggests the cause of death was multifactorial.
The majority of patients in this cohort received vasopressor medication including epinephrine, dopamine, and calcium. Catecholamines trigger carcinoid tumors to release a variety of substances that are responsible for the carcinoid crises (1). In the 1950s and 1960s, researchers routinely provoked carcinoid crises by administering as little as 2 µg epinephrine or 10 µg norepinephrine (13,15,16). Intraoperative hypotension in these patients can be recalcitrant, or even paradoxically worsened by the administration of vasoactive medications, because these medications stimulate more carcinoid activity. Octreotide has become the therapy of choice for treating intraoperative hypotension in patients with carcinoid syndrome because it blocks the release of tumor-secreted substances (6,17). In our series, administration of catecholamine drugs was not associated with higher intraoperative octreotide dosages, suggesting that their use did not result in increased carcinoid secretion. Our finding that vasoactive medications can be given safely to patients with carcinoid heart disease in the presence of octreotide therapy is supported by three previous case reports (4,10,12).
There was an increased mortality associated with epinephrine usage during surgery, but not other vasoactive medications. This might have resulted from our practice of reserving epinephrine infusions when there was difficulty separating from CPB. Patients who required epinephrine had worse preoperative NYHA class, higher urinary 5-HIAA values, and greater number of PRBC and non-PRBC units transfused. These observations suggest that patients receiving epinephrine had more active carcinoid symptoms, advanced carcinoid heart disease, and were at higher risk regardless of vasoactive medication use.
We found that the use of aprotinin was associated with increased intraoperative octreotide dosages. Carcinoid tumors secrete kallikrein–kinin peptides that are inhibited by aprotinin in dosages achieved, even with the half-dose regimen used in some patients in this series (13,14). Aprotinin use has been reported in patients with carcinoid syndrome to treat carcinoid crises with variable results (9,14,18). Aprotinin failed to block carcinoid symptoms provoked by administration of epinephrine and norepinephrine (10,19,20). These observations, along with our findings, suggest that in the setting of octreotide administration, further kallikrein–kinin system inhibition may offer little additional benefit in mitigating carcinoid activity. We did find that aprotinin administration was associated with reduced amounts of blood products transfused, but it had no effect on mortality (21). During the latter time period, approximately 80% of patients, including all patients who died, received aprotinin suggesting a bias for the use of aprotinin.
Limitations
Our database was developed to track outcomes in relation to clinical management. The observational nature and the lack of standardized protocols make it difficult to establish formal causal relationships between measured variables and outcomes. Furthermore, changes in outcome may reflect changes in overall practice rather than a single particular variable. The small sample size limits the ability to develop multivariate models for identifying risks for mortality. Results of our bivariate logistic regression analyses should be interpreted with caution. Another limitation is the likelihood of a referral bias in the latter time period because patients were referred earlier in their disease and were more likely to survive valvular heart surgery.
CONCLUSIONS
In this review of the anesthetic management of 99 consecutive carcinoid heart disease patients undergoing cardiac surgery, we found that the use of aprotinin was associated with reduced blood transfusions, but not with differences in mortality. These results suggest that hypotension in patients with carcinoid heart disease should be treated with octreotide, but vasoactive medications can be safely administered for refractory hypotension in the presence of octreotide. We found an improvement in perioperative mortality from 1995 to 2003 in comparison to 1985 to 1994 that likely reflects changing perioperative practices and earlier referral of patients with carcinoid disease.
Footnotes
Accepted for publication July 24, 2007.
Supported by the Department of Anesthesiology, Mayo Clinic and Mayo Foundation, Rochester, Minnesota.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  P. L. Koyyalamudi, A. Alfirevic, and C. G. Koch An Unusual Presentation of Carcinoid Tumor Anesth. Analg., May 1, 2009; 108(5): 1463 - 1464. [Full Text] [PDF]
|
|
|
|
|
|
K. A. Mizuguchi, A. A. Fox, T. M. Burch, L. H. Cohn, and J. A. Fox Tricuspid and Mitral Valve Carcinoid Disease in the Setting of a Patent Foramen Ovale Anesth. Analg., December 1, 2008; 107(6): 1819 - 1821. [Full Text] [PDF]
|
|
|
|
 |
|
 J. G. Castillo, F. Filsoufi, D. H. Adams, J. Raikhelkar, B. Zaku, and G. W. Fischer Management of patients undergoing multivalvular surgery for carcinoid heart disease: the role of the anaesthetist Br. J. Anaesth., November 1, 2008; 101(5): 618 - 626. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
