Reducing Major Adverse Cardiac Events and All-Cause Mortality in Noncardiac Surgery: Perioperative Strategies

Patient's Age

The risk of PMI and MINS is nearly doubled in patients older than 70 years, especially in men with cardiovascular (CV) risk factors. Mortality and death from CAD are strongly associated with age. As a consequence of the aging population, it is estimated that this problem will increase in future decades. Older patients are more frail, have multiple comorbidities, and exhibit more severe CAD. They also tend to present greater technical challenges during percutaneous coronary intervention (PCI) because of heavier coronary artery calcification, tortuous anatomy in coronary and peripheral arteries, increased risk of procedure-related complications (e.g., contrast-induced nephropathy, vascular or neurologic complications), and reduced tolerance to bleeding.

Cardiac Risk Indexes

Two clinical indexes are used to estimate patients' risk of perioperative cardiac complications. The Revised Cardiac Risk Index (RCRI) incorporates six independent variables that predict the risk of cardiac complications: history of ischemic heart disease, HF, cerebrovascular disease, diabetes mellitus, chronic kidney disease (serum creatinine >2 mg/dL), and major operations (suprainguinal vascular, intrathoracic, and intraperitoneal). Perioperative risk of both cardiac complications (e.g., nonfatal AMI and nonfatal cardiac arrest) and death increases with index scores. For example, in a large cohort study including 782,969 patients, the in-hospital mortality rates were 1.4% for RCRI of 0, 2.2% for RCRI of 1, 3.9% for RCRI of 2, 5.8% for RCRI of 3, and 7.4% for RCRI of 4 and greater.

The RCRI is currently the most widely used cardiac risk stratification tool. However, it has several limitations, including its relatively low discriminative ability. In fact, although the RCRI has a moderately good ability to discriminate patients who will develop cardiac events from those who will not after mixed noncardiac surgical procedures (area under the curve [AUC] 0.75), it is less accurate in patients undergoing vascular surgical procedures (AUC, 0.64), and it is less able to predict all-cause mortality (median AUC, 0.62).

To overcome these limitations of RCRI, the National Surgical Quality Improvement Program (NSQIP) score was developed and validated on 211,410 surgical patients. This model includes age, ASA class, functional status, abnormal serum creatinine, and a novel and more appropriate organ-based categorization of surgery. Risk may be quantified by a risk calculator on the Internet. The discriminative or predictive ability of the NSQIP score is significantly better as compared with RCRI (AUC, 0.88), and it works well also in vascular surgical patients.

Kidney Disease

The most important comorbidity associated with poor postoperative outcome is chronic kidney disease (CKD). The rate of adverse cardiac events and the length of hospital stay increase significantly in patients with impaired renal function, especially in those with CKD from stage 3b onward (estimated glomerular filtration rate <45 mL/min).

Most of the CV disease risk factors, such as older age, diabetes mellitus, systolic hypertension, and low levels of high-density lipoprotein cholesterol, in addition to an inflammatory and thrombogenic milieu, are highly prevalent in patients with CKD. CAD and valvular disease are more common and severe in these patients, with half of deaths resulting from cardiac causes. CKD-associated anemia also reduces myocardial oxygen supply and is associated with cardiomyopathy. LV hypertrophy increases myocardial demand and evolves toward diastolic dysfunction, impairing subendocardial perfusion, and may be complicated by diastolic HF (the stiff ventricle is more vulnerable to preload and afterload changes, tachycardia, and loss of atrial kick during atrial fibrillation or other arrhythmias).

Some precautions may be useful to reduce the risk of perioperative cardiac events in patients with CKD. Stress testing can identify patients with CAD. Discontinuation of angiotensin-converting enzyme (ACE) inhibitor therapy for at least 10 hours before general anesthesia is recommended to reduce the risk of postinduction hypotension. Anemia may require preoperative blood transfusion, supplementation with iron, or administration of erythropoietin. Patients with end-stage kidney disease should undergo dialysis the day before the operation.

The main goals during surgical procedures include a mean arterial pressure greater than 65 mm Hg (or higher for the uncontrolled hypertensive patient) and adequate volume status. Particular attention should be paid to analgesic requirements in the perioperative period . Opioids may accumulate in patients with CKD, with increased risk of respiratory depression, but nonsteroidal antiinflammatory drugs are not recommended because of the risk of worsening renal function.

In patients with renal impairment, it is appropriate to measure baseline values of troponin to compare them with the postoperative values. Troponin values may be elevated in the setting of even mild kidney disease, probably reflecting microinfarctions or LV hypertrophy.

Anemia and Blood Transfusion

The prevalence of preoperative anemia is increasing in the surgical population, especially in older patients. In a large (39,309 patients) European study, anemia (defined according to the WHO criteria, e.g., hemoglobin [Hb] <13 g/dL in men and <12 g/L in nonpregnant women) was found in 31% of men and 26% of women. Preoperative anemia is commonly associated with comorbidities such as kidney disease, CAD, HF, diabetes mellitus, and hepatic cirrhosis and is known to be associated with increased mortality rates. In fact, anemia reduces DO ₂ , increases heart rate, and may be complicated by hypotension.

After adjustment for major confounders including transfusion, preoperative anemia was strongly associated with a more than twofold increase in 90-day mortality rates, as well as increased postoperative intensive care unit (ICU) admission and greater use of ICU resources (hemodynamic monitoring, mechanical ventilation, inotropic and vasoactive agents). In particular, in-hospital mortality rates increase linearly with hematocrit reduction.

Although anemia is associated with mortality, transfusions may contribute to increased mortality rates (according to the “second hit” theory). However, recent data suggest that blood transfusions in the perioperative period may not necessarily be harmful and, particularly, that more liberal transfusion strategies are associated with reduced mortality rates in certain settings.

Patients at risk for anemia who are undergoing elective surgical procedures should be screened 4 to 8 weeks preoperatively, and the causes of anemia (e.g., blood loss, nutritional deficiencies, kidney disease, chronic or inflammatory diseases) should be identified and treated. Iron supplementation (oral or intravenous [IV], depending on iron status or tolerance and timing of the operation) is recommended (grade 1C recommendation) in patients with iron deficiency (serum ferritin <30 µg/L). The efficacy of iron supplementation in raising Hb concentration and decreasing perioperative transfusion rate is well demonstrated. If iron deficiency is ruled out, erythropoietin-stimulating agents administered up to an Hb concentration of 12 to 13 g/dL are suggested (grade 2A recommendation). The need for blood transfusions has been shown to be reduced by approximately 50% in patients treated with these drugs (data from pooled studies including mainly orthopedic surgical patients). The risk of thrombotic complications, particularly in patients with CAD, coronary stenting, or risk of venous thrombosis, should be considered.

Type of Surgical Procedure

The type of surgical procedure is a strong risk factor for MACE and death. Urgent or emergency operation has been well recognized as the strongest predictor of death, with an increase of more than three times in 30-day mortality rates. Unfortunately, this is a largely unmodifiable risk factor.

Perioperative myocardial infarction is more common in patients who are urgently hospitalized, particularly those undergoing vascular, thoracic, and noncardiac transplant surgery (which are all independent risk factors for PMI) compared with elective hospital admissions (adjusted odds ratio [OR], 2.38).

Vascular surgical procedures are associated with a two- to fourfold higher risk of adverse cardiac events (PMI, cardiac death) compared with other types of noncardiac operations. In fact, CAD is more common among patients undergoing vascular surgical procedures (with a prevalence ranging from 37% to 78%) than in other noncardiac surgical patients. Aortic cross-clamping and declamping, abrupt changes in systemic arterial pressure, fluid shifts, hypoxia induced by one-lung ventilation, acute anemia secondary to major bleeding, and inflammatory or hypercoagulable states induced by both surgical procedures and transfusions can trigger perioperative ischemia and MI, especially in patients with CAD, acute HF, or LV dysfunction .

In a recent retrospective investigation, surgical priority was found to be the only preoperative risk factor independently associated with PMI among patients undergoing major open vascular surgery (OR, 1.70). In this cohort of patients, the only postoperative variables associated with PMI were the nadir hematocrit and postoperative transfusion, thus suggesting that minimizing intraoperative blood loss and prioritizing early intraoperative transfusion may be potential ways for preventing myocardial damage.

The vascular procedure with the highest associated mortality rate is surgery for abdominal aortic aneurysmal rupture, followed by elective thoracoabdominal aortic replacement, lower extremity arterial bypass, and carotid endarterectomy. Patients requiring lower extremity amputation also have diffuse and severe CAD (up to 92% in a pathologic study). Accordingly, perioperative risk is high in these patients, with reported 30-day mortality rates of up to 17% and PMI as the leading cause of postprocedural death. Conversely, endovascular aortic repair (EVAR) procedures are associated with reduced myocardial stress and, accordingly, with a decreased incidence of perioperative myocardial damage. However, an increase in troponin levels after EVAR was associated with a higher long-term incidence of adverse cardiac events (49 vs. 15% in a follow-up period of 3 years).

Altered Preoperative Coagulation

A recent substudy of an international prospective cohort investigation of perioperative CV events in noncardiac surgery (VISION) found that the preoperative elevation of blood markers of hypercoagulability was associated with an increased risk of MINS in patients undergoing vascular surgery. In particular, as compared with patients with no myocardial injury, patients with MINS showed a significantly higher concentration of factor VIII (186 vs 155%; P = .006), von Willebrand factor activity (223 vs 160%; P < .001), von Willebrand factor concentration (317 vs 237%; P = .02), fibrinogen concentration (5.6 vs 4.2 g/L; P = .03), d -dimer (1680.0 vs. 1090.0 ng/mL; P = .04), plasmin-antiplasmin complex (747 vs 512 ng/mL; P = .002), and C-reactive protein (10 vs 4.5 mg/L; P = .02).

Preoperative Troponin

Cardiac troponin has high sensitivity for detection of small amounts of myocardial necrosis. Increased cTn levels indicate the presence of, but not the underlying reason for, myocardial injury. Besides AMI, troponin release may be associated with many other disorders, including HF, sepsis, and end-stage kidney disease ( Box 22.1 ). Regardless of the cause of cTn release, elevated cTn levels almost always imply a poor prognosis. Elevated preoperative cTn values are found in a variable proportion of patients undergoing vascular surgical procedures. In the largest trial available, the preoperative finding of increased cTn (high-sensitive troponin T, hsTnT) was present in up to 24% of patients, and it was independently associated with a significantly higher risk of PMI, cardiac death, and all-cause death. Moreover, hsTnT showed an additive value (AUC, 0.80) in association with cardiac risk index (AUC, 0.65) and natriuretic peptide levels (AUC, 0.76). A combined endpoint (including all-cause death, PMI, acute HF, and cardiac arrest) occurred in 9.4% of patients with hsTnT levels higher than 0.014 ng/mL compared with 1.9% in patients with hsTnT levels of up to 0.014 ng/mL ( P < .001). Possible causes of elevated cTn associated with adverse outcomes include silent myocardial ischemia or microinfarction, LV dysfunction, cerebrovascular disease, renal impairment, sepsis, pulmonary hypertension, and pulmonary embolism. The need to add cTn to routine preoperative tests performed in high-risk surgical patients is still debated. According to the 2014 European Society of Cardiology/European Society of Anaesthesiology (ESC/ESA) guidelines, the assessment of cTn in high-risk patients, both before and 48 to 72 hours after major surgical procedures, may be considered (class IIb, level B), even if the suboptimal specificity of this test should be taken into account. A practical approach in patients with preoperatively increased troponin levels involves a baseline transthoracic echocardiogram (primarily assessing ventricular function and regional wall motion), a cardiology consultation, and when feasible, deferral of surgery until the troponin levels fall ( Fig. 22.1 ). If it is not possible to postpone the procedure, a less-invasive surgical approach, targeted perioperative monitoring, and careful cardiac optimization should be recommended. Moreover, patients should be informed about the increased risk.

Box 22.1

Causes of Troponin Elevation in the Absence of Myocardial Ischemia

Cardiac Causes

• Heart failure

• Cardiac arrhythmias

• Cardioversion

• Implantable cardioverter-defibrillator shock

• Myocarditis

• Pericarditis

• Cardiac amyloidosis

Noncardiac Causes

• Sepsis and septic shock

• Pulmonary embolism

• Primary pulmonary hypertension

• Pulmonary edema

• Chronic kidney disease

• Stroke

• Subarachnoid hemorrhage

• High dose of chemotherapy

• Sympathomimetic drugs

Postoperative Troponin

Evaluation of peak cTn level during the first 3 days after noncardiac surgical procedures improves the ability to identify patients with myocardial damage, even in the absence of symptoms or ECG changes. Moreover, this value is an independent predictor of 30-day mortality. In a recent large international cohort study involving 15,065 patients aged 45 years or older from five continents, an abnormal value of TnT (≥0.04 ng/mL) was found in 8% of patients within 3 days after noncardiac surgical procedures and was an independent predictor of 30-day mortality rates (9.8% vs 1.1%; adjusted ratio, 4.82). In another cohort study including 2216 participants older than 60 years of age who were undergoing medium-risk to high-risk noncardiac surgical procedures, an elevation of cTnI (>0.06 ng/mL) was recorded in 19% of patients. The 30-day mortality rate in these patients was 8.6% compared with 2.2% in patients without cTnI elevation ( P < .001). The relative risk of death was 2.4 for patients with lower increases in cTnI (0.07–0.59 ng/mL) and 4.2 for patients with higher increases (≥0.60 ng/mL). The median time to death was 12 days.

A recent meta-analysis of 11 studies including, overall, 2193 patients undergoing noncardiac, nonvascular surgery, found that postoperative troponin elevation was strongly associated with MACE at 30 days (OR, 5.92) and 1 year after surgery (adjusted OR, 3.0) and was a predictor of 30-day mortality (OR, 3.52) and an independent predictor of 1-year mortality (adjusted OR, 2.53).

A strong association between postoperative cTn elevation and both short- and long-term mortality was confirmed in two recent large observational investigations. Among 21,842 patients who underwent noncardiac surgery, multivariate analysis showed that peak postoperative hsTnT levels were correlated with 30-day mortality; in particular, 30-day mortality rates were 3%, 9.1%, and 29.6% in patients with an hsTnT value of 20 to 64 ng/L (hazard ratio [HR], 23.63), 65 to 999 ng/L (HR, 70.34), and 1000 ng/L or greater (HR, 222.01), respectively. An absolute hsTnT change of 5 ng/L or higher was associated with an increased risk of 30-day mortality (HR, 4.69).

A gradual association between postoperative TnT elevation and both short- and long-term mortality rates was also found among 12,882 vascular surgery patients, with the greatest hazard ratio for mortality within the first 10 months after surgery.

Postoperative cTn surveillance is cost effective in patients older than 45 years of age. According to the abovementioned evidence, it seems to be useful for early identification of patients at increased risk of myocardial injury and death and may also allow prompt initiation of appropriate therapeutic interventions. Optimization of perioperative care, including prevention of hypotension, tachycardia, anemia, hypoxia, pain, hypoglycemia, and hypothermia, may prevent postoperative troponin elevation and major cardiac events and potentially reduce mortality rates.

B-Type (Brain) Natriuretic Peptides

B-type (brain) natriuretic peptides (BNPs) are released from myocardium in response to multiple physiologic stimuli, including ischemia, myocardial stretch, inflammation, and other neuroendocrine triggers. Preoperative BNP levels are strong independent predictors of adverse short-term CV outcome . The preoperative addition of BNPs to the widely used risk stratification systems (RCRI and functional capacity assessment) leads to a significantly improved risk discrimination (AUC from 65% to 80%). The predictive value of N-terminal pro–brain natriuretic peptide (NT-proBNP) seems to be higher as compared with BNP, probably because it is more indicative of baseline conditions and is less affected by transient fluctuations in concentrations, given its longer half-life.

The optimal cutoffs of BNPs to predict CV events after surgical procedures are approximately 20 to 30 pg/mL for BNP (with 95% sensitivity and 44% specificity) and approximately 125 pg/mL for NT-proBNP. In a relatively small prospective study in high-risk patients undergoing major noncardiac operations, a preoperative BNP level greater than 40 pg/mL allowed identification of patients with an almost sevenfold increased risk of cardiac events. In particular, each 100-pg/mL increase in BNP levels was associated with a 35% increase in the relative risk of death . The utility of BNP testing in patients with kidney disease is controversial.

Finally, the negative predictive value of normal BNP levels (<20 pg/mL) to indicate a favorable postoperative outcome is as high as 96%, a finding suggesting that patients with normal levels of BNPs may proceed directly to surgery with no additional preoperative cardiac testing.

Postoperative (days 1–3) measurement of BNPs in addition to preoperative values significantly improve the prediction of death or nonfatal MI at both 30 days (OR, 3.7) and more than 180 days. An individual patient data meta-analysis including 2051 patients demonstrated that patients with postoperative BNP values of 0 to 250 pg/mL, greater than 250 to 400 pg/mL, and greater than 400 pg/mL reached a composite endpoint, including 30-day death and nonfatal MI at a rate of 6.6%, 15.7%, and 29.5%, respectively.

No prospective, randomized, controlled trials (RCTs) investigated the use of BNP-guided management in perioperative medicine. Nevertheless, according to a meta-analysis of RCTs that showed a 48% reduction in all-cause mortality rates with BNP-guided therapy in nonsurgical patients with HF, the following approach seems reasonable ( Fig. 22.1 ). In the presence of clinical risk factors and/or reduced physical capacity, measurement of BNPs should be performed 4 to 5 weeks before a scheduled major operation. If BNP levels are lower than the optimal cutoff (20 pg/mL), the patient can proceed with the surgical procedure without the need for further testing. Conversely, if BNP levels are higher than this threshold, further testing, primarily echocardiography and (BNP-guided) optimization of medical therapy (e.g., fluid restriction, diuretics, ACE inhibitors, nitrates, β-blocking agents) may be recommended. At the same time, worsening of renal function and hypotension, sometimes also induced by ACE inhibitors themselves, must be prevented. Specific therapeutic interventions may be considered in selected cases, for example, cardiac resynchronization therapy in patients with symptomatic NYHA functional class III disease with an LV ejection fraction (LVEF) of less than 35% and a large QRS complex (>120 ms) or transcatheter mitral clip implantation in patients with severe functional mitral regurgitation. Repeating assessment of BNPs shortly before the surgical procedure may allow for adjustment of perioperative treatment strategies (e.g., choice of surgical and anesthetic techniques, perioperative monitoring, fluid, drugs, and management of devices).