N-terminal pro B-type natriuretic peptide is an independent predictor of postoperative myocardial injury in patients undergoing major vascular surgery
Article Outline
Objective
Myocardial ischemia and infarction after surgery remain leading causes of morbidity and mortality in patients undergoing major vascular surgery. B-type natriuretic peptide has been shown to predict early postoperative cardiac events in patients undergoing major noncardiac surgery. We aimed to determine if N-terminal pro B-type natriuretic peptide (NT-pro-BNP), with its longer half-life and greater plasma stability, can predict postoperative myocardial injury in vascular patients.
Methods
Recruited were 136 patients undergoing elective surgery for subcritical limb ischemia or abdominal aortic aneurysm (AAA) repair. Plasma NT-pro-BNP was measured preoperatively, and troponin-I was measured immediately after surgery and on postoperative days 1, 2, 3, and 5.
Results
Twenty-eight patients (20%) sustained postoperative myocardial injury (troponin-I rise of >0.1 ng/mL). The median NT-pro-BNP level of those with myocardial injury was significantly higher than those who did not (380 pg/mL [interquartile range (IQR), 223-967] vs 209 pg/mL [109-363]; P = .003). NT-pro-BNP predicted this outcome with an area under the receiver operating characteristic (ROC) curve of 68% (95% confidence interval [CI] 0.56%-0.78%). In a multivariate analysis, a NT-pro-BNP value of ≥308 pg/mL (the optimal ROC curve–derived cutoff) was associated with an increased incidence of myocardial injury (odds ratio, 3.4; 95% CI, 1.41-9.09, P =.01).
Conclusion
Elevated preoperative plasma NT-pro-BNP levels independently predict postoperative myocardial injury, which is associated with adverse outcome in the short- and long-term regardless of the presence of symptoms of acute coronary syndrome.
Major vascular surgery is associated with a high risk of early cardiovascular complications.1, 2, 3 Myocardial injury, as detected by a rise in the troponin-I concentration, has been shown to occur in up to 20% of patients undergoing elective aortic aneurysm repair and 38% of patients undergoing revascularization for critical limb ischemia.4, 5 Postoperative myocardial injury is asymptomatic in most patients.6, 7, 8 However, even early small postoperative rises in troponin-I have been shown to correlate with adverse short- and medium-term outcome.6, 9, 10, 11, 12, 13
A means of identifying patients at risk of a postoperative rise in troponin may facilitate targeted perioperative interventions and optimize the use of limited resources. Preoperative risk stratification guidelines for patients undergoing noncardiac surgery do not include important vascular prognostic markers such as C-reactive protein.14 Furthermore, the low predictive value of noninvasive testing of cardiac function, coupled with limited resources, highlights the need for alternative methods of predicting cardiac injury in patients who are scheduled for major vascular surgery.15
Recent interest has focused on the possible role of B-type natriuretic peptide (BNP) and its N-terminal fragment (NT-pro-BNP) as predictors of early postoperative adverse cardiac events in patients undergoing noncardiac surgery.16, 17, 18, 19, 20, 21, 22 The major source of BNP synthesis and secretion is the ventricular myocardium. In vascular patients NT-pro-BNP has been shown to be an independent predictor of 30-day postoperative cardiac events as assessed by a composite end point of symptoms, troponin-T levels, and electrocardiograph (ECG) recordings.21 In this study we aimed to determine if NT-pro-BNP can predict early postoperative myocardial injury as assessed by a rise in troponin-I in patients undergoing elective major vascular surgery.
Methods
Study population
Recruited were 136 patients who were scheduled to undergo elective surgery for revascularization for subcritical limb ischemia (n = 112) characterized by rest pain, ulcer, or gangrene, regardless of absolute pressure in the ankle, or open repair for infrarenal abdominal aortic aneurysm (AAA, n = 24) at the Vascular Unit of Aberdeen Royal Infirmary. During the 21-month recruitment period, 18 patients did not fulfil the criteria and were excluded, a further five patients agreed to participate in the study but did not proceed to surgery, and one patient withdrew. Ethical approval was obtained from the Grampian Regional Ethical Committee, and written informed consent was obtained from each patient.
Exclusion criteria were patients with acute limb ischemia, current active infections; those taking antiplatelet agents than aspirin, or other nonsteroidal anti-inflammatory drugs, or anticoagulants like warfarin; and those with symptomatic or ruptured AAA. Patients unable to give informed consent were also excluded.23
Data from the study patients have been reported previously.23, 24 All patients were receiving statin drugs before surgery and aspirin therapy (75 mg/d), which was continued in the perioperative period. Documented were the patients' comorbidities, cardiac medications, and Revised Cardiac Risk Index (RCRI),25 which includes the risk factors of type of surgery, history of ischemic heart disease, presence of signs of congestive cardiac failure, history of cerebrovascular disease, need for insulin therapy, and a preoperative serum creatinine level >2 mg/dL (177 mmol/L). Patients with 0, 1, 2, and 3 or more of these risk factors are assigned to classes I, II, III, and IV, respectively. Routine preoperative cardiac stress testing was not performed in these patients.
Study design
The study was a prospective observational study of patients undergoing major elective vascular surgery. Venous blood samples were collected from all patients at baseline before surgery, immediately after surgery, and on postoperative days 1, 2, 3, and 5 for measurement of troponin-I. The baseline blood sample for NT-pro-BNP assay, drawn the day before surgery, was centrifuged and plasma stored in aliquots at −80°C. Recordings of comorbidities, current smoking, cardiac medications, and clinical symptoms on a standardized data sheet were performed along with a daily 12-lead ECG from the day before surgery until 5 days after surgery. The ECGs were interpreted by a consultant cardiologist who was blinded to the troponin-I and NT-pro-BNP levels.
Cardiac troponin-I
Troponin-I was measured using the Bayer ADVIA Centaur Immunoassay analyzer (Bayer Diagnostics, Tarrytown, NY). This chemiluminometric sandwich immunoassay has intra-assay coefficients of variation (CV) of 1.1% and 2.2% at levels of 2.9 ng/mL and 7.1 ng/mL, respectively. A troponin-I level of >0.1 ng/mL on this assay was considered elevated.13, 19 The assays were performed by the Department of Clinical Biochemistry, NHS Grampian.
N-terminal pro B-type natriuretic peptide
The Roche Elecsys N-terminal pro B-type natriuretic peptide electrochemiluminescence sandwich immunoassay was performed on a Roche Elecsys 2010 automated immunoassay analyser platform (Roche Diagnostics, Basel, Switzerland). The assay has an analytical sensitivity (lower detection limit) of 5 pg/mL and an effective measuring range of 5 to 35,000 pg/mL. The within-run CV was 2.7% at a concentration of 175 pg/mL and 1.9% at 1068 pg/mL. Total analytical precision demonstrated a CV of 3.2% and 2.6% at concentrations of 175 and 1068 pg/mL, respectively.
Statistical analysis
Data are presented as medians with interquartile ranges (IQR), and significance is developed at the 5% probability level. Differences between two independent categoric values were tested with χ2 test, and differences between two continuous independent values were tested with the Mann-Whitney U test. To test the strength of the association between NT-pro-BNP and the other continuous variables, the Spearman test for correlation was used. A receiver operating characteristic (ROC) curve was plotted to assess the ability of NT-pro-BNP to predict postoperative myocardial injury, and the area under the curve, with 95% confidence interval (CIs) was calculated. Multivariate logistic regression analysis was performed to determine the independent predictive value of NT-pro-BNP and other univariate predictors. Analysis was done with SPSS 15 software (SPSS Inc, Chicago, Ill).
Power calculation
Our original study23 was powered on the basis of a projected elevation of cardiac troponin in 24% of our 136 patients. The log-transformed NT-pro-BNP data (parametric) were used for power calculation for this article. The power of the study was calculated to be 87% for the mean difference of 0.28 in the log NT-pro-BNP level to be statistically significant at 5% level between the 28 patients who sustained myocardial injury and the 108 patients without injury. This is equivalent to a change from 238 to 436 pg/mL in the original NT-pro-BNP values.
Results
The patient characteristics are reported in Table I. In all patients, the preoperative troponin-I level was ≤0.10 ng/mL. A postoperative myocardial injury, defined as a troponin-I >0.1 ng/mL, was noted in 28 patients (20%), which occurred ≤48 hours of surgery in 23 (82%). One patient died on postoperative day 2 of a cerebrovascular accident. None of the study patients had typical ECG changes of acute myocardial infarction in the postoperative period. Thirty-nine patients (28%) were taking β-adrenoceptor antagonists. The incidence of postoperative myocardial injury was similar in patients who were and were not taking β-blockers (P = .65).
Table I. Patient characteristics in the entire cohort and in patients with and without myocardial injury
| Variable No (%) or median (IQR) | Study population | Myocardial injury | P | |
|---|---|---|---|---|
| Yes | No | |||
| Patients | 136 | 28 | 108 | |
| Age, years | 69 (43-97) | 73(49-97) | 68(43-87) | .015 |
| Male | 92(68) | 18(64) | 74(69) | .65 |
| Hypertension | 85(63) | 18(64) | 67(62) | .56 |
| Current smokers | 38(28) | 9(32) | 29(27) | .37 |
| Diabetes mellitus | ||||
| None | 104(76) | 21(75) | 83(76) | |
| Diet controlled | 9(6) | 2(7) | 7(6) | |
| OHA treatment | 10(7) | 2(7) | 8(7) | .55 |
| Insulin treatment | 13(10) | 3(10) | 10(9) | |
| Ischemic heart disease | ||||
| None | 89(66) | 14(50) | 75(69) | |
| Angina/previous MI | 47(34) | 14(50) | 33(31) | |
| Previous cardiac revascularization | 16(12) | 4(14) | 12(11) | .056 |
| Cardiac medication | ||||
| β-blockers | 39(28) | 9(32) | 30(27) | .65 |
| ACE inhibitors | 38(28) | 11(39) | 28(26) | .16 |
| Calcium-channel blockers | 38(28) | 9(32) | 29(27) | .57 |
| Nitrates | 15(11) | 4(14) | 11(10) | .53 |
| Revised Cardiac Risk Index | ||||
| I | 42 | 3(7) | 39(93) | |
| II | 61 | 14(22) | 47(78) | |
| III | 28 | 9(32) | 19(67) | |
| IV | 5 | 2(40) | 3(60) | .04 |
| Creatinine, μmol/L | 94(84-109) | 94(80-108) | 97(67-117) | .94 |
| Creatinine >177 μmol/L | 5(3) | 1(3.5) | 4(3.7) | .97 |
| NT-pro-BNP, pg/mL | 277(117-561) | 436(275-1024) | 238(111-435) | .003 |
| NT-pro-BNP ≥308 pg/mL | 58(42) | 20(71) | 38(35) | .001 |
The median (IQR) NT-pro-BNP levels in patients who sustained myocardial injury were significantly higher than levels in those who did not (436 [275-1024] pg/mL vs 238 [111-435] pg/mL; P = .003; Fig 1). This finding was apparent in patients with subcritical limb ischemia (717.5 [366.35-1479] pg/mL vs 246 [120.57-506.82] pg/mL) and those with AAA (319 [87.5-441] pg/mL vs 126.85 [84.9-265.72] pg/mL).
Fig 1.
The data for log-transformed N-terminal pro B-type natriuretic peptide (NT-pro-BNP) levels in patients who have myocardial injury after surgery, compared with those who do not, are shown in box and whisker plots. The horizontal line in the center of the box represents the median; the top and bottom borders of the box represent the interquartile range, and the whiskers mark the range. The y axis represents the log-transformed NT-pro-BNP, which was used in the power calculation. Patients who sustained postoperative myocardial injury had a significantly higher preoperative NT-pro-BNP value than those who did not (P = .003).
The incidence of troponin-I elevation in the 28 patients increased with higher quartiles of NT-pro-BNP as follows: quartile 1 (<117 pg/mL) in 5; quartile 2 (117 to <280 pg/mL) in 2; quartile 3 (280 to <566 pg/mL) in 9; and quartile 4 (≥566 pg/mL) in 12 (P = .01). Preoperative NT-pro-BNP levels correlated with patient age (r = 0.4, P = .001) and peak postoperative troponin-I levels (r = 0.29, P = .001). No correlation was found between preoperative NT-pro BNP levels and the RCRI or preoperative ECG findings.
The ability of NT-pro-BNP to predict postoperative myocardial injury from the ROC curve was 68% (95% CI, 56%-78%, P = .005; Fig 2). An optimal cutoff point of 308 pg/mL was derived with a sensitivity of 71% and specificity of 65% for predicting postoperative myocardial injury (Fig 2). The positive-predictive value (PPV) was 33% and negative-predictive value (NPV) 90%.
Fig 2.
A receiver operating characteristic curve was calculated for the ability of the preoperative plasma N-terminal pro B-type natriuretic peptide level to predict postoperative cardiac events in the elective vascular surgical cohort. The horizontal axis represents the false-positive ratio (1 minus specificity), and the vertical axis represents the true-positive ratio (sensitivity). The area under curve, which measures discrimination (ie, the ability of the test to correctly classify those with and without myocardial injury), was 68% (95% confidence interval, 56%-78%, P = .005).
On univariate analysis, NT-pro-BNP levels as a continuous variable and at a cutoff value of 308 pg/mL, along with age and the RCRI, were significantly associated with an elevated troponin-I level (Table I). Age was included in the multivariate analysis despite its linear relationship to NT-pro-BNP because the correlation coefficient was only 0.4. On multivariate analysis, after adjustment for cardiac risk factors, only a preoperative NT-pro-BNP level ≥308 pg/mL remained an independent predictor of myocardial injury (Table II).
Table II. Multivariate analysis to assess independent predictors of myocardial injury
| Predictor | OR | 95% CI | P |
|---|---|---|---|
| Age, years | 1.04 | 0.98-1.10 | .12 |
| Revised Cardiac Risk Index | .075 | ||
| I | 1 | ||
| II | 3.41 | 0.89-13.71 | .07 |
| III | 6.57 | 1.47-29.33 | .01 |
| IV | 8.47 | 0.90-9.71 | .06 |
| NT-pro-BNP ≥308 pg/mL | 3.41 | 1.27-9.09 | .01 |
Discussion
This study demonstrates that the preoperative NT-pro-BNP level is an independent predictor of postoperative myocardial injury in patients undergoing major vascular surgery. An asymptomatic postoperative troponin-I level >0.10 ng/mL occurred in 20% of our patients. This was not accompanied by ECG changes. However, previous studies have shown that a rise in troponin-I of this magnitude, with or without accompanying symptoms or ECG changes, correlates with adverse short- and medium-term survival.9, 10, 11, 12, 13 In this study a cutoff value of NT-proBNP of 308 pg/mL identified patients with a greater than threefold increased risk of myocardial injury in the early postoperative period. Thus, preoperative NT-pro-BNP may have a role in identifying high-risk vascular patients who may require more extensive preoperative and postoperative cardiovascular optimization and might be used as a screening test for risk stratification in patients undergoing elective or semielective vascular surgery.
Our group has previously shown that BNP levels of >40 pg/mL are predictive of perioperative death or myocardial injury ≤72 hours of surgery in patients undergoing noncardiac procedures.19 This study included 20 patients undergoing major vascular surgery who were also included in the current study. We assessed NT-pro-BNP in the current cohort because the half-life of BNP is 20 minutes, whereas NT-pro-BNP has a half-life of 120 minutes.26 Furthermore, the stability of NT-pro-BNP in stored plasma is superior to that of BNP.27 NT-pro-BNP values are approximately six times higher than the BNP levels,26 and interestingly, the optimum cutoff value for NT-pro-BNP in the current study is just slightly more than six times the cutoff value obtained in our earlier work on a more heterogeneous cohort using BNP (40 pg/mL).18 In both studies we found a positive linear relationship between NT-pro-BNP and age.
Previous studies
A single prior study has assessed the early predictive value of NT-pro-BNP in vascular patients.21 This also concluded that, as a continuous variable, NT-pro-BNP was not an independent predictor of perioperative outcome.21 However, Feringa et al21 found that a value of 533 pg/mL was independently associated with an increased risk of early cardiac events, defined as a composite end point of two or more of the following: elevated troponin-T or creatine kinase (or both), ECG changes, and ischemic symptoms. In this cohort the median concentration of baseline NT-pro-BNP was 110 pg/mL, which is considerably lower than our study. The same investigators have also demonstrated that preoperative NT-pro-BNP levels predict medium-term mortality and major adverse cardiac events with an optimal cutoff of 319 pg/mL.28 More recently, Mahla et al22 detected baseline NT-pro-BNP levels similar to our study, and their patients had a cardiac event rate of 20% during a median follow-up of 826 days. They also found that a NT-pro-BNP value >280 pg/mL was associated with a fourfold relative risk of adverse cardiac events during medium-term follow-up.
Study limitations
In this study we have not shown that NT pro-BNP levels are associated with an adverse outcome, but that they are associated with a rise in troponin, which in turn is known to correlate with adverse outcome. We do intend to monitor this cohort of patients to determine their short- and medium-term outcome, but at this current time, these data are not available.
Furthermore in this study, we have not shown a correlation between NT-pro-BNP levels and various comorbidities such as diabetes, which is likely to be a reflection of study numbers. Currently, the use of β-blockers in the perioperative period in patients undergoing vascular surgery is not routine practice in the United Kingdom. Despite the perceived potential benefits of β-blockers, the significant incidence of bradycardia and hypotension has led to the recommendation that there is a need for a suitably powered randomized, controlled trial of perioperative β-blockers to definitively establish the benefits and risks.29 Results of the PeriOperative ISchemic Evaluation (POISE) trial, a randomized, controlled trial using metoprolol vs placebo in noncardiac surgical patients, are yet to be published.
Clinical implications
The main clinical utility of NT-pro-BNP in this setting appears to be a very high NPV for early postoperative events. In the current study we found a NPV of 90% using our optimal cutoff value. This accords with prior reports of a NPV of 100% in patients undergoing a variety of noncardiac surgical procedures16 and 99% in patients undergoing major vascular surgery.21 Likewise, Feringa et al21 found that NT-pro-BNP was superior to dobutamine stress echocardiography in identifying patients at risk of cardiac events,21 and in the current study we have shown it to be superior to the RCRI. No patients in this study were readmitted with a cardiac event in the period between hospital discharge and the first follow-up outpatient visit. Data on medium term follow-up is not yet available.
Conclusions
This study has shown that preoperative NT-pro-BNP (using a cutoff value of ≥308 pg/mL) is an independent predictor of postoperative myocardial injury in patients undergoing vascular surgery. An asymptomatic postoperative rise in troponin-I has been shown to correlate with adverse short- and medium-term outcome. There is no current consensus on the optimum cutoff value for NT-pro-BNP, and further work is required in this area. Ultimately, there is the potential to identify patients at increased risk who may benefit from more aggressive investigation and monitoring in the pre and postoperative period.
Author contributions
We received statistical advice from Dr Gordon Prescott and Mr Edwin Amalraj (Department of Public Health, University of Aberdeen). Dr Jane McNeilly (Department of Biochemistry) performed the NT-pro-BNP assays.
References
- Perioperative myocardial ischemia in patients undergoing noncardiac surgery–I: Incidence and severity during the 4 day perioperative period (The Study of Perioperative Ischemia (SPI) Research Group). J Am Coll Cardiol. 1991;17:843–850
- . Myocardial infarction after reconstruction of the abdominal aorta. Br J Surg. 1993;80:28–31
- Comparative early and late cardiac morbidity among patients requiring different vascular surgery procedures. J Vasc Surg. 1995;21:935–944
- . Comparison of cardiac troponin I and creatine kinase ratios in the detection of myocardial injury after aortic surgery. Br J Surg. 2001;88:1196–1200
- . Peri-operative myocardial injury in patients undergoing surgery for critical limb ischaemia. Eur J Vasc Endovasc Surg. 2005;29:301–304
- Importance of long-duration postoperative ST-segment depression in cardiac morbidity after vascular surgery. Lancet. 1993;341:715–719
- . Postoperative myocardial ischemia: etiology of cardiac morbidity or manifestation of underlying disease?. J Clin Anesth. 1995;7:97–102
- . Myocardial infarction after noncardiac surgery. Anesthesiology. 1998;88:572–578
- Cardiac troponin I predicts short-term mortality in vascular surgery patients. Circulation. 2002;106:2366–2371
- Association of cardiac troponin, CK-MB, and postoperative myocardial ischemia with long-term survival after major vascular surgery. J Am Coll Cardiol. 2003;42:1547–1554
- . Long-term prognostic value of asymptomatic cardiac troponin T elevations in patients after major vascular surgery. Eur J Vasc Endovasc Surg. 2004;28:59–66
- Early increases in cardiac troponin levels after major vascular surgery is associated with an increased frequency of delayed cardiac complications. J Clin Anesth. 2006;18:280–285
- Predictors and outcomes of a perioperative myocardial infarction following elective vascular surgery in patients with documented coronary artery disease: results of the CARP trial. Eur Heart J. 2008;29:394–401
- Endothelial dysfunction in peripheral arterial disease is related to increase in plasma markers of inflammation and severity of peripheral circulatory impairment but not to classic risk factors and atherosclerotic burden. J Vasc Surg. 2003;38:374–379
- . Routine measurement of radioisotope left ventricular ejection fraction prior to vascular surgery: is it worthwhile?. Eur J Vasc Endovasc Surg. 2004;27:227–238
- . Preoperative plasma N-terminal pro-brain natriuretic peptide as a marker of cardiac risk in patients undergoing elective non-cardiac surgery. Br J Surg. 2005;92:1041–1045
- . Assessment of cardiac risk before noncardiac surgery: brain natriuretic peptide in 1590 patients. Heart. 2006;92:1645–1650
- . B-type natriuretic peptide predicts cardiac morbidity and mortality after major surgery. Br J Surg. 2007;94:903–909
- Utility of B-type natriuretic peptide in predicting perioperative cardiac events in patients undergoing major non-cardiac surgery. Br J Anaesth. 2007;99:170–176
- . Utility of B-type natriuretic Peptide in predicting medium-term mortality in patients undergoing major non-cardiac surgery. Am J Cardiol. 2007;100:1310–1313
- Association of plasma N-terminal pro-B-type natriuretic peptide with postoperative cardiac events in patients undergoing surgery for abdominal aortic aneurysm or leg bypass. Am J Cardiol. 2006;98:111–115
- N-terminal pro-brain natriuretic peptide identifies patients at high risk for adverse cardiac outcome after vascular surgery. Anesthesiology. 2007;106:1088–1095
- Platelet activation, myocardial ischemic events and post-operative non-response to aspirin in patients undergoing major vascular surgery. J Thromb Haemost. 2007;5:2028–2035
- . Platelet activation increases with the severity of peripheral arterial disease: implications for clinical management. J Vasc Surg. 2007;46:485–490
- Derivation and prospective validation of a simple index for prediction of cardiac risk of major noncardiac surgery. Circulation. 1999;100:1043–1049
- . Role of B-type natriuretic peptide (BNP) and NT-proBNP in clinical routine. Heart. 2006;92:843–849
- . Long-term stability of endogenous B-type natriuretic peptide (BNP) and amino terminal proBNP (NT-proBNP) in frozen plasma samples. Clin Chem Lab Med. 2004;42:942–944
- Plasma N-terminal pro-B-type natriuretic peptide as long-term prognostic marker after major vascular surgery. Heart. 2007;93:226–231
- How strong is the evidence for the use of perioperative beta blockers in non-cardiac surgery? (Systematic review and meta-analysis of randomised controlled trials). BMJ. 2005;331:313–321
Sriram Rajagopalan was sponsored by a grant from the Scottish Chief Scientist Office.
The Health Services Research Unit is core funded by the Chief Scientists Office of the Scottish Executive Health Department. The views expressed in this article are entirely those of the authors.
Competition of interest: none.
PII: S0741-5214(08)00731-3
doi:10.1016/j.jvs.2008.05.015
© 2008 The Society for Vascular Surgery. Published by Elsevier Inc. All rights reserved.
