Natriuretic Peptides and Troponins to Predict Cardiovascular Events in Patients Undergoing Major Non-Cardiac Surgery
Patients undergoing major surgery have a substantial risk of cardiovascular events during the perioperative period. Despite the introduction of several risk scores based on medical history, classical risk factors and non-invasive cardiac tests, the possibility of predicting cardiovascular events in patients undergoing non-cardiac surgery remains limited. The cardiac-specific biomarkers, natriuretic peptides (NPs) and cardiac troponins (cTn) have been proposed as additional tools for risk prediction in the perioperative period. This review paper aims to discuss the value of preoperative levels and perioperative changes in cardiac-specific biomarkers to predict adverse outcomes in patients undergoing major non-cardiac surgery. Based on several prospective observational studies and six meta-analyses, some guidelines recommended the measurement of NPs to refine perioperative cardiac risk estimation in patients undergoing non-cardiac surgery. More recently, several studies reported a higher mortality in surgical patients presenting an elevation in high-sensitivity cardiac troponin T and I, especially in elderly patients or those with comorbidities. This evidence should be considered in future international guidelines on the evaluation of perioperative risk in patients undergoing major non-cardiac surgery.
1. Introduction
Patients undergoing major surgery have a substantial risk of cardiovascular events during the perioperative period [1,2,3,4,5,6,7]. Although the rate of these events has declined over the past 30 years, they still represent a significant issue in patients undergoing non-cardiac surgery [8,9,10,11], with at least 167,000 cardiac complications of non-cardiac surgical procedures occurring annually in the European Union, 19,000 of which are life-threatening [8]. The 2014 European Society of Cardiology/European Society of Anaesthesiology (ESC/ESA) guidelines recommended that cardiac risk be carefully evaluated in patients undergoing non-cardiac surgery [8]. Despite the development of several risk scores based on medical history, classical cardiovascular risk factors (including sex, age, lipid profile and creatinine concentration), and some non-invasive cardiac tests (such as electrocardiogram, echocardiogram, stress tests), the possibility of predicting cardiovascular events remains limited [4,5,8,9]. This has prompted the assessment of cardiac biomarkers as additional tools for risk prediction [3,10,11,12]. The 2017 Canadian Cardiovascular Society guidelines have recommended the measurement of cardiac natriuretic peptides (NPs) (brain natriuretic peptide, BNP, or N-terminal fragment of proBNP, NT-proBNP) before surgery to refine perioperative cardiac risk estimation [13]. These recommendations were based on prospective observational studies and six meta-analyses evaluating the accuracy of NPs to predict major cardiovascular events after non-cardiac surgery [14,15,16,17,18,19,20,21].
Over the last 10 years, increases in cardiac troponin (cTn) have been associated with an increased short- and long-term risk of cardiac events in patients undergoing different types of non-cardiac surgery [10,11,12,13,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45]. Nonetheless, no specific recommendations on cTn measurement in the perioperative period have been issued so far.
This article aims to discuss the clinical value of preoperative values and perioperative changes in cardiac-specific biomarkers as tools to predict adverse outcomes in patients undergoing non-cardiac surgery.
2. Risk Evaluation Using Cardiac Biomarkers in Patients Undergoing Non-Cardiac Surgery
Half of perioperative cardiac deaths occur in patients with no history of heart disease [7,10], suggesting that our current protocols to screen for subclinical heart disease are far from optimal [10]. Risk prediction models based only on clinical criteria, also including classical risk factors and cardiac stress tests, were not shown to improve the accuracy of preoperative risk stratification and to reduce 30-day mortality after non-cardiac surgery [9,11,12]. In particular, a 2019 meta-analysis included six studies on the accuracy of cardiac stress test to predict 30-day mortality [11]. The authors concluded that, despite substantial research, the current body of evidence is insufficient to derive a definitive conclusion as to whether stress testing reduces perioperative mortality [11].
To improve risk stratification and prognostic accuracy in patients undergoing non-cardiac surgery, BNP and NT-proBNP have gained clinical consensus, especially for the detection of subclinical heart failure (HF) [46,47], and cTnI and cTnT for the identification of myocardial damage [43,44,45,46,47,48,49,50,51].
NPs are peptide hormones predominantly produced and secreted mostly by the human heart [43,46,51]. In particular, atrial natriuretic peptides are produced and secreted in the atria, while B-type natriuretic peptides (BNP and the related peptides) are produced and secreted in the ventricles, especially in patients with cardiac disease [43,46,51]. International guidelines recommend the measurement of BNP/NT-proBNP for the diagnosis, risk stratification and follow-up of patients with acute or chronic HF [47,52]. The 2014 ESC/ESA guidelines stated that routine pre-operative NP measurement for risk stratification is not recommended, but may be considered in high-risk patients [8].
More recently, the results of several prospective observational studies and six meta-analyses re-evaluated the prognostic accuracy of NT-proBNP and BNP to predict cardiovascular events after non-cardiac surgery [14,15,16,17,18,19,20,21]. Based on this evidence, the 2017 Canadian Cardiovascular Society Guidelines have strongly recommended BNP or NT-proBNP measurement before surgery to refine perioperative cardiac risk estimation in patients: (1) older than 65 years; (2) aged 45–64 years with cardiovascular disease; (3) with a Revised Cardiac Risk Index (RCRI) score > 1 [13]. The RCRI score includes ischemic heart disease, heart failure (HF), cerebrovascular disease, diabetes mellitus, increased serum creatinine (>177 mmol/L, corresponding to 2.0 mg/dL), and high-risk (major) non-cardiac surgery (defined as intraperitoneal, intrathoracic, or suprainguinal vascular surgery) [13].
The 2014 ESC/ESA Guidelines stated that cTnI and cTnT measurement may be considered in high-risk patients, either before or 48–72 h after major surgery, to detect a myocardial injury [8]. Furthermore, perioperative cTn increases are associated with increased short- and long-term risk of cardiac events in patients undergoing different types of non-cardiac surgery, especially when high-sensitivity (hs) immunoassay methods are used [10,11,12,13,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42].
In 2019, Humble et al. [35] conducted a systematic review and meta-analysis on the prognostic value of increased cTn levels above the cut-off level (usually the 99th percentile upper reference limit (URL)) in adult patients undergoing non-cardiac surgery. Adverse outcome was defined as short-term (in-hospital or <30 days) and long-term (>30 days) major adverse cardiovascular events (MACEs) and/or all-cause mortality non-cardiac [35]. This meta-analysis included 19 studies assessing preoperative cTn and 3 studies evaluating perioperative changes in cTn [35]. These studies assessed preoperative cTn values with a total sample size of 13,386 (range 33 to 4575). They were mainly prospective, single-center studies on patients undergoing a wide range of non-cardiac interventions (mostly intermediate- or high-risk procedures). In particular, hs-cTnT methods were used in just six studies, while the others employed non-hs-cTnI or TnT methods [35]. Preoperative cTn predicted short- (adjusted odds ratio (OR) 5.87, 95% confidence interval (CI) 3.24–10.65, p < 0.001) and long-term adverse outcome (adjusted hazard ratio (HR) 2.0, 95% CI 1.4–3.0, p < 0.001) [35]. Preoperative cTn predicted short- (adjusted odds ratio (OR) 5.87, 95% confidence interval (CI) 3.24–10.65, p < 0.001) and long-term adverse outcome (adjusted hazard ratio (HR) 2.0, 95% CI 1.4–3.0, p < 0.001) [35].
More recently, some studies evaluated the utility of perioperative cTn elevation as a prognostic indicator for mortality and cardiac morbidity in patients undergoing surgery for neck of femur fractures [39,53,54]. These fractures are rare in individuals aged <50 years in the absence of high-impact traumas, while in older patients they may be associated with low-velocity trauma because of reduced bone mineral density and frailty [39,53,54]. This meta-analysis included 11 studies with a total of 1363 patients (mean age 83 years, 351 men and 904 women) [39]. Seven studies measured cTnI, three studies measured cTnT and one study used a hs-cTnI assay. Overall, 497 patients (36.5%) experienced a cTn elevation following surgery. Perioperative troponin elevation was significantly associated with all-cause mortality (OR 2.6; 95% CI 1.5—4.6; p < 0.001) and cardiac complications (OR 7.4; 95% CI 3.5—15.8; p < 0.001) [39]. Increased cTn levels were associated to pre-existing coronary artery disease, cardiac failure, hypertension, previous stroke and previous myocardial infarction [39]. Therefore, perioperative troponin elevation is significantly associated with increased mortality and post-operative cardiac complications in patients operated for neck of femur fractures [39].
领英推荐
3. Analytical and Pathophysiological Correlates in Cardiac-Specific Biomarkers
Both NPs and cTn are then useful prognostic indicators in patients undergoing major non-cardiac surgery [10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42]. However, clinicians should interpret measured values based on the analytical performance of assay methods and the mechanisms of their production and release from the heart [55,56].
In healthy subjects, NP and cTn are present into the circulation in a range of concentrations (from about 3 to about 50 ng/L) [43,44,45] from 100 to 1000 times lower than other biomarkers, including C-Reactive Protein (CRP), creatinine, cholesterol, D-dimer, Neutral Gelatinase-Associated Lipocalin (NGAL) [43,56,57]. Some circulating proteins and peptides can directly affect the binding of NPs and cTn to specific antibodies of immunoassay methods, interfering with immunoassay methods. Some circulating proteins and peptides can directly affect the binding of NPs and cTn to specific antibodies of immunoassay, interfering with test methods. This interference becomes stronger as the molar concentrations of several substances, able to bind to the specific monoclonal antibodies utilized by immunometric systems, increases compared to that of NP and cTn, as discussed in detail elsewhere [43,44,45,58,59,60]. Furthermore, the measurement of cTnI and cTnT with immunoassay methods can be affected by binding these two cardiac troponins with troponin C, and also with some tissue or plasma proteins, and heterophile- or auto-antibodies to form macro-complexes [58,59]. Therefore, both the accuracy and clinical interpretation of the measured levels of cardiac biomarkers strongly depend on analytical characteristics and performance of assay methods [43,58,59,60,61].
3.1. NP Assay
NPs are key diagnostic and prognostic tools in patients with cardiac disease, and are released following every kind of cardiac damage able to activate the neuro-endocrine-immune system. NP increase does not provide any information on the mechanisms of damage acting in the individual patient [51,60,61,62]. In particular, NPs are highly sensitive in detecting cardiac stress in patients with risk factors and/or asymptomatic early vessel damage and/or cardiac dysfunction [51,60,61,62,63].
NPs are rapidly degraded in vivo. The active hormone BNP has a shorter plasma half-life (20–40 min) than the inactive peptide NT-proBNP (>60 min) [51,60,64]. Due to their rapid turnover rate, BNP shows larger intra- (from 40–50%) and inter-individual variability (from 50 to 60%) than those of NT-proBNP (intra 30–40%, inter 40–50%) [60,64]. A large variability means wide confidence intervals and large differences between serial measurements in the same individual, regardless of changes in the disease state [65].
Clinicians should take into consideration some critical issues for a proper interpretation of changes in circulating NPs. Most notably, values should be interpreted according to sex, age, body mass index, comorbidities and therapies [43,47,60,61,62]. In particular, kidney disease can significantly affect NP clearance, increasing BNP and NT-proBNP levels [60,61]. However, a meta-analysis [63] confirmed that NT-proBNP retains utility to diagnose acute HF also in patients with renal dysfunction (although with higher cut-off values) and holds prognostic significance regardless of renal function.
There are larger systematic differences among measured concentrations with the BNP methods than with NT-proBNP methods [60,66]. Due to its greater stability in vivo and in vitro and a smaller difference between assay methods, NT-proBNP is better suited to act as an indicator of disease and prognostic biomarker than BNP [60].
From a clinical perspective, it is important to note that international guidelines recommend cut-off values for the diagnosis and risk prediction in patients with HF, and not for risk stratification in patients undergoing major non-cardiac surgery [46,47]. Cut-off values for evaluation of perioperative risk in patients undergoing major non-cardiac surgery should be defined in specific clinical studies. Optimal cut-off values have been calculated for MACEs. There was wide variation in cut-off values reported in different studies. In 2009, a meta-analysis (including 15 studies with 4856 patients) reported that NT-proBNP cut-off values were higher than those for BNP (range 201–791 ng/L vs. 35–255 ng/L, respectively) [16]. More recently, Rodseth et al. [20] evaluated in an individual patient data meta-analysis the predictive value of preoperative BNP on cardiovascular events (defined as cardiovascular death and nonfatal myocardial infarction) and all-cause mortality during the first 30 days after vascular surgery. Using the receiver operating characteristic (ROC) statistics, the authors calculated the general optimal test cut values for BNP (116 ng/L), as the point that optimizes the rate of true-positive results while minimizing the rate the rate of false-positive results from five data sets [20]. Moreover, the authors proposed several pre-operative BNP cut-off values in predicting 30-day MACEs: 30 ng/L for screening (95% sensitivity, 44% specificity), 116 ng/L for optimal (highest accuracy point; 66% sensitivity, 82% specificity), and 372 ng/L for diagnostic (32% sensitivity, 95% specificity) [20].