Cardiac function

Epidemiology

In the United States, CHF is the only CVD with an increasing incidence. The National Heart, Lung, and Blood Institute estimates current prevalence at 4.8 million Americans and 23 million worldwide with CHF. There are approximately 580,000 new cases each year, with approximately 1 million admissions to hospitals for CHF per year. CHF is the leading cause of hospitalization in individuals 65 years of age and older.

Therapeutic options for patients with HFPEF are more limited than for those who have systolic abnormalities. ^, Current prognosis depends on disease severity, but overall it is poor. Mortality at 5 years is approximately 50%, and 10-year mortality is 90%. These poor outcomes are not without substantial cost, estimated at $24 billion per year in the United States.

Currently, CHF patients are staged with the New York Heart Association (NYHA) functional classifications I to IV. Class I patients are generally considered asymptomatic, with no restrictions on physical activity; class IV patients are often symptomatic at rest, with severe limitations on physical activity. The problem with this classification system is that much of it is based on subjective criteria. Thus patients with comorbidities that reduce their activities are hard to classify. In addition, dyspnea, which is the primary symptom in many of these individuals, has many causes. Finally, many patients with ventricular dysfunction modify their activities to accomplish activities of daily living and thus lack overt symptoms until late in their disease. Therefore patients with CHF often go undiagnosed and untreated early in their disease or are misdiagnosed because of conditions such as pulmonary disease. Initiating treatment in the more advanced disease state (higher degree of irreversible cardiac function and patient deconditioning) is challenging and more expensive (often requiring extended inpatient stay) and leaves patients with considerable morbidity on a daily basis. Obviously, misdiagnoses often lead to patient morbidity. That is the reason why natriuretic peptides (NPs) have been such an important advance in facilitating the diagnosis of HF ^, and if used properly help with treatment.

Precipitating factors

In many patients with AMI, no precipitating factor can be identified. Studies have noted the following patient activities at the onset of AMI: (1) heavy physical exertion, 13%; (2) modest or usual exertion, 18%; (3) surgical procedure, 6%; (4) rest, 51%; and (5) sleep, 8%. Exertion before infarction is somewhat more common among patients without preexisting angina than in those who have a history of angina.

Causes of infarction other than acute atherothrombotic coronary occlusion have been identified. For example, prolonged vasospasm can induce infarction, and spontaneous dissection is becoming more commonly appreciated in women, many of whom have fibromuscular dysplasia. In addition, it is now clear that some patients, particularly women, can have acute infarction with normal-appearing angiographic coronary arteries. Other conditions ( Box 48.1 ) can cause the death of cardiomyocytes, leading to a biochemical signal (such as increased circulating concentrations of cTns) of myocyte damage, but these conditions should not be confused with MI. Pulmonary embolism (PE) is another common cause of biochemical elevations that is secondary to right ventricular damage related to acute increases in wall stress and reduced subendocardial perfusion. ^,

Chronobiology

There is a pronounced periodicity for the time of onset of STE AMI. ^, Often an AMI occurs in the morning hours soon after rising; this is a period of (1) increasing adrenergic activity, (2) increased plasma fibrinogen levels, (3) increased inhibition of fibrinolysis, and (4) increased platelet adhesiveness. Studies have demonstrated that the early morning peak in MI parallels the peak incidence of death from ischemic heart disease, which occurs at approximately 8 am to 9 am. A second peak has been noted at approximately 5 pm. Diurnal differences affect many physiologic and biochemical parameters; the early morning hours are associated with rises in plasma catecholamines and cortisol and increases in platelet aggregability. Tissue plasminogen activator (t-PA) activity is low and plasminogen activator inhibitor (PAI) activity is high during the early morning hours. Thus it is possible that some cyclic aspects of combined vasospastic, prothrombotic, and fibrinolytic factors, in the setting of preexisting atherosclerosis, lead to AMI. NSTEMI does not exhibit this diurnal pattern.

Prognosis

STE and non-STE infarctions have distinctly different short-term prognoses. STE AMI is associated with higher early and in-hospital mortality. It is said that mortality associated with STE AMI can occur up to 6 months after the event, but the vast majority (at least two-thirds) occurs during the first 30 or 40 days. It is this risk that coronary recanalization seems to benefit. NSTE AMI is associated with lower acute mortality and complication rates but a longer period of vulnerability to reinfarction and death. As a result, 1- to 2-year survival rates are similar to those for STEMI. This is why intervention has been so effective in this group.

There is an additional subset of patients who have what are known as type 2 MI. Type 1 MI is the typical one due to acute atherothrombotic plaque disruption. Type 2 MI is due to supply-demand imbalance leading to ischemia. Coronary artery disease (CAD) may or may not be present but changes in oxygen demand and, usually concomitantly, myocardial supply can lead to acute ischemia and if severe enough to acute myocardial injury (an increased cTn value with a rising and/or falling pattern). An example might be severe anemia where there is reduced oxygen carrying capacity along with compensatory increases in systemic volume and with these, increased myocardial work. Type 2 MI events can have ST segment elevation, ST segment depression, or even normal ECGs. The prognoses of these patients in terms of mortality and cardiovascular events is at least as adverse as that of those with type 1 events. A similar process can occur with supply-demand imbalance or direct myocardial injury due to circulating toxins such as carbon monoxide which can cause the same pattern of cTn release. However, if ischemia is not present, the proper term for this situation is “acute myocardial injury.” Acute myocardial injury can also occur after coronary interventions and cardiac bypass surgery. Criteria for MI exist in both those situations. The key to being able to make the diagnosis is having a normal baseline cTn value.

Clinical history

The clinical history remains of substantial value. A prodromal history of angina is elicited in 40 to 50% of patients with AMI. Among patients with AMI who present with prodromal symptoms, approximately one-third have had symptoms from 1 to 4 weeks before hospitalization; in the remaining two-thirds, symptoms predate admission by a week or less, with one-third of patients having had symptoms for 24 hours or less. In most patients the pain of AMI is severe, but it is rarely intolerable. The pain may be prolonged, lasting up to 30 minutes. The discomfort is described as constricting, crushing, oppressing, or compressing; often the patient complains of something sitting on or squeezing the chest. Although usually described as a squeezing, choking, viselike, or heavy pain, it may be characterized as a stabbing, knifelike, boring, or burning discomfort. The pain is usually retrosternal in location, spreading frequently to both sides of the chest, and favoring the left side. Often the pain radiates down the left arm. Some patients note only a dull ache or numbness in the wrists in association with severe substernal discomfort. In some instances, the pain of AMI may begin in the epigastrium, simulating a variety of abdominal disorders; this often causes MI to be misdiagnosed as indigestion. In other patients, the discomfort of AMI radiates to the shoulders, upper extremities, neck, and jaw, again usually favoring the left side. In patients with preexisting angina, the pain of infarction usually resembles that of the angina pain with respect to features and location. However, it is generally much more severe, lasts longer, and/or is not relieved by rest and nitroglycerin.

Older individuals, patients with diabetes, and women often present atypically. For example, among individuals older than 80 years, less than 50% of those with AMI will have chest discomfort at the time of AMI. Sometimes these patients will present with shortness of breath, fatigue, or even confusion. The pain of AMI may have disappeared by the time a physician first encounters the patient (or the patient reaches the hospital), or it may persist for a few hours.

Myocardial changes after acute myocardial infarction

Fig. 48.5 shows the temporal sequence of early biochemical, histochemical, and histologic findings after the onset of AMI. On gross pathologic examination, AMI can be divided into subendocardial (nontransmural) infarctions and transmural infarctions. In the former, necrosis involves the endocardium, the intramural myocardium, or both without extending all the way through the ventricular wall to the epicardium. In the latter, myocardial necrosis involves the full thickness of the ventricular wall. The histologic pattern of necrosis may differ: contraction band injury occurs almost twice as often in nontransmural infarctions as in transmural infarctions. Unfortunately, the pathologic changes correlate poorly with clinical, ECG, and biochemical markers of necrosis, which is why those terms are no longer used clinically. Statistically, patients are more apt to have STE MI Q waves on the ECG and larger biochemical signals when the infarction is transmural pathologically.

Development and progression of atherosclerosis

Intrinsic to modern-day understanding of ischemic heart disease and to the intense interest in the development of markers of inflammation is the concept that atherosclerosis is a chronic inflammatory disease. The concept is that some event damages the endothelium of blood vessels, which facilitates the egress of lipid into the subendothelial space. Putative injurious stimuli include turbulent flow in a blood vessel, which could occur for example because of hypertension or a noxious metabolite from a lipid fraction. This damage tends to occur at branch points of blood vessels. Regardless of the initial stimulus, once damaged, low-density lipoprotein (LDL) can cross into the vessel wall more easily in a nicotinamide adenine dinucleotide phosphate (NADPH) oxidase–mediated fashion. Whether minimal oxidation facilitates that egress or it occurs once the LDL is within the vessel wall is unclear, but a minimal degree of oxidation once in the vessel wall facilitates the egress of smooth muscle cells from the media of the vessel and macrophages that ingest cholesterol, hence the rationale for the measurement of oxidized lipids in blood. The process of atherosclerosis progresses slowly, with involvement of lymphocytes, monocytes, macrophages, and smooth muscle cells. The dynamic within a given plaque may vary, but there clearly is an inflammatory milieu, in part mediated by substances such as CD40 ligand, which can be measured directly or indirectly as C-reactive protein (CRP). Interleukins (IL)-1, IL-6, IL-8, and IL-18 also participate to various extents as part of this chronic inflammatory process. This process involves adherence of white blood cells to the damaged endothelial surface, with subsequent degranulation and elaboration of myeloperoxidase (MPO). A procoagulant component is due predominantly to the presence of tissue factor, which is localized immediately under the cap of the plaque. Intermittent instability is noted because of inflammatory products within the plaque that release chemicals, such as metalloproteinases. Initially the plaque expands by stretching the adventitia through a process of small ruptures with release of procoagulant and proinflammatory materials and then remodeling over time as anti-inflammatory and anticoagulant and thrombolytic substances are elaborated. This process of stretching the adventitia preserves the lumen such that by the time luminal encroachment occurs, there is a very large plaque burden.

A categorization of plaques has been proposed to facilitate identification of those at risk for rupture that could lead to an acute event. It is acknowledged that the propensity for a plaque to rupture probably reflects a systemic predilection rather than a local one. Thus for a given patient at risk, there likely are many plaques that are metabolically at risk for rupture at any given time. ^, High-risk plaques have the following:

• 1.

  an active inflammatory environment that not only may be intrinsic but may be stimulated additionally by systemic infection;

• 2.

  a thin fibrous cap on the endothelial surface with a large lipid core that is filled with procoagulant substances, predominantly tissue factor;

• 3.

  endothelial denudation and fissuring caused by the elaboration of metalloproteinases;

• 4.

  local high shear stress, usually because they are severe, at branch points in the vessel.

Events likely occur because of superimposed thrombosis. This can be the result of erosions on the surface of the plaque or more often rupture of the plaque at its edges, where the cap is thinnest and most of the metalloproteinases reside. If rupture induces total thrombotic occlusion, the event is usually a STE AMI. If lesser degrees of occlusion occur, an NSTE AMI or UA may ensue. One of the causes that may participate in subtotal occlusive plaque rupture involves platelets and abnormal coronary vasomotion. It is known that diseased coronary arteries respond atypically to many stimuli, often constricting rather than dilating. Because the cross-sectional area of a vessel is related to the square of the radius, even modest amounts of constriction can markedly increase the extent of occlusion. Whether constriction occurs first, leading to changes of coronary flow and platelet aggregation on the plaque, or whether platelets stick and cause the aggregation, is not certain, but these processes reinforce one another. Platelets secrete vasoconstricting substances in response to a denuded area, which expresses cell adhesion molecule (CAM) receptors. This, in addition to stagnant blood flow, will cause platelets and white blood cells to adhere to the surfaces of vessels. It appears likely that platelets adhere and enhance vasoconstriction and then break off, causing small vessel emboli, sometimes in association with plaque debris and sometimes without. These processes, in addition to a reduction in flow, can lead to necrosis or at least recurrent ischemia. It is apparent that the process that eventually leads to acute events involves a systemic propensity to platelet aggregation and inflammation, because effluent flowing from the nonculprit vessel (distant from the putative coronary lesion causing the acute event) elaborates inflammatory mediators (e.g., MPO) similar to those observed from the affected vessel. This pathophysiology has recently been supported by the Cantos trial where inhibition of the interleukin one beta inflammasome was shown to reduce ischemic events. Of interest, the antibody therapy used also reduced the incidence of lung cancer. A separate trial of low-dose methotrexate was null. Finally, necrosis when present stimulates an acute-phase reaction and inflammation. Given this pathophysiology, many therapies are now oriented toward inhibition of thrombosis, fibrinolysis, platelet aggregation, and inflammation. Many inflammatory markers are used diagnostically and for assessment of therapeutic efficiency.

The more recent appreciation of the high frequency of type 2 MI has changed some of the thinking in regard to MI. Type 1 events which depend on acute atherosclerosis are benefited by invasive intracoronary interventions. Type 2 events due to supply-demand mismatch often do not require mechanical interventions even when coronary heart disease is present. In many instances, the coronary arteries may be normal or at least not badly diseased. This entity has also resulted in rethinking which has moved those studying atherogenesis in multiple additional directions.

Diagnosis of acute myocardial infarction

The diagnosis of AMI established by the World Health Organization in 1986 included biomarkers as an integral part of the disorder and required that at least two of the following criteria be met: (1) a history of chest pain, (2) evolutionary changes on the ECG, and/or (3) elevations of serial cardiac markers to a level two times the normal value. However, over time, it became rare for a diagnosis of AMI to be made in the absence of biochemical evidence of myocardial injury. A 2000 European Society of Cardiology (ESC)/ACC consensus conference updated in 2007 and 2012 (Global Task Force) ^, codified the role of markers by advocating that the diagnosis should be regarded as evidence of myocardial injury based on markers of cardiac damage in the appropriate clinical situation ( Box 48.2 ). The guidelines thus recognized the reality that neither the clinical presentation nor the ECG had adequate sensitivity and specificity. This guideline does not suggest that all elevations of these biomarkers should elicit a diagnosis of AMI—only those associated with appropriate clinical and ECG findings (see discussion later in this chapter). When elevations that are not caused by acute ischemia occur, the clinician is obligated to search for another cause for the elevation. In the 2007 revision of the guidelines, several types of AMI were recognized, including the spontaneous type, which is associated with plaque rupture or erosion, and the type associated with fixed or transient coronary abnormalities but not thrombotic occlusion. These are discussed in greater detail in the following paragraphs. It is also recognized that one can have a classic AMI and succumb before markers are obtained or become elevated, and cardiac injury can occur in association with cardiac procedures. In addition, criteria for different types of MI, including after coronary interventions and bypass surgery were suggested ( Box 48.3 ).

Guideline-supported recommendations

Consensus guidelines from the Fourth Universal Definition of Myocardial Infarction (2018), the joint laboratory medicine AACC Academy and International Federation of Clinical Chemistry (IFCC) C-CB, the ACC/AHA and Epidemiology groups, the ACC Foundation, and the European Clinical and Laboratory groups have recommended that, in patients who present with ischemic symptoms, a rising or falling serial pattern with at least one cTn concentration higher than the sex-specific 99th percentile URLs, during the first 24 hours after onset of symptoms indicates myocardial injury/necrosis/cell death. Fig. 48.11 shows a representative profile of the rise and fall pattern of cTn for type 1 MI, type 2 MI, and chronic myocardial injury patients. If this elevation occurs in the clinical setting of ischemia consistent with MI, that diagnosis should be made (see Box 48.2 ). It is recommended that cTn assays with appropriate QC and optimal total imprecision (CV ≤ 10%) at the 99th percentile URL are preferred. ^, Better imprecision at low cTn concentrations, based on high-sensitivity assays, appears to improve the value of interpreting cTn as an early rule out and rule in diagnostic tool and risk indicator (within 3 hours of baseline sampling), as will be discussed in the clinical section. Use of contemporary or POC cTn assays with intermediate imprecision (10 to 20% CV) at the 99th percentile, however, is deemed clinically acceptable and does not lead to patient misclassifications when serial cTn results are interpreted.

A challenge that needs to be considered as hs-cTn assays with improved analytical sensitivity become increasingly incorporated in laboratory practice is determining how these new assays compare with the older contemporary assays. In essence, as will be discussed below, the hs-assays improve clinical practice for diagnostics and risk assessment for more rapid (early) and improved patient management and care. Diagnostic clinical sensitivities using specimens collected serially from presentation for detection of MI have improved from 15 to 35% for the initial generations of cTn assays, to 50 to 75% for current contemporary assays, and to more than 80% for hs-assays, as shown in Fig. 48.12 . To exclude an AMI with contemporary assays, the Global Task Force originally recommended a 6-hour period for assessing the optimal negative predictive value (NPV) for ruling out AMI. With hs-cTn assays, the timing for ruling out an AMI with greater than 99.5% NPV and 99% clinical sensitivity has decreased to less than 2 to 3 hours from the time of baseline/first blood draw, better defining the clinical playing field for hs-assays used to assist in the diagnosis of MI, rule out MI, and better stratify patients for risk of adverse events. The basis for these early approaches was that these early signals anticipated that eventually the criteria proposed by the Universal Definition of MI would be met. Further, studies consistently now show by using hs-assays that in patients presenting with a low clinical likelihood of ACS, both hs-cTnI and hs-cTnT baseline concentrations below an assay’s LOD with a nonischemic ECG can be used to rule out AMI, as will be discussed later in the chapter. Measurable cTn concentrations but less than the sex-specific 99th percentile URL or less than overall 99th percentile URL and without a significant delta (change) value (assay dependent) 1, 2, or 3 hours after the baseline sample, can also provide an NPV greater than 99.5% for early rule out in a substantial number of patients, as will be discussed later in the clinical section that addresses the caveat of early presenters at less than 2 hours following the index event onset. As the use of hs-cTn assays grows in the USA, having already done so worldwide outside the USA, triage of patients in the emergency department (ED) will be significantly improved, allowing for more rapid triage to an appropriate level of care or earlier discharge home, with minimal risk for an adverse event, with substantial financial savings to the health care system/hospital. Ongoing analytical and clinical development is underway to validate hs-POC cTn assays.

Defining reference populations

To address the conundrum of concentration differences across assays that have been shown for contemporary cTn assays, appropriate-sized, population-based, direct comparisons of hs-cTn assays have been carried out. Only the IFCC C-CB and the AACC Academy have provided an international group of experts that have published their opinion or guideline that defines an “apparently healthy population.” This approach suggests the use of a common normal reference population within a focused, healthy, age-defined group similar in age to patients presenting with symptoms suggestive of ACS to establish 99th percentile URLs, with the use of surrogate biomarkers to exclude silent pathophysiologies. Multiple studies using first-generation, contemporary, and hs-cTn assays have demonstrated that the cTn 99th percentile strongly depends on selection of individuals to be included in the reference population. Examining multiple contemporary and POC assays, the percent of measurable concentrations below the 99th percentile concentration were all less than 50%. Further, the 99th percentile URL variability among assays is substantial, further exemplifying the lack of cTnI and cTnT assay standardization. hs-cTn assays in comparison have demonstrated near-Gaussian distributions (see Fig. 48.10 ; assay dependent), with sex-specific URL differences that are not observed with contemporary and POC assays because of their lack of analytical sensitivity.

Defining assays analytical characteristics

A two-tiered system of analysis using both 99th percentiles and imprecision values at the 99th percentile, based on a younger, healthy reference population that is diversified by sex, race, and ethnicity, has been proposed. This approach has been challenged because (1) it does not provide an age-matched normal cohort matching the ACS patient population that typically presents to rule out an AMI and (2) it is not based on clinical diagnostic or outcomes data at the 99th percentile URL. The scorecard approach was based on a published scorecard concept to capture the essence of which assays are acceptable for use in clinical practice and to facilitate the transition to hs-cTn assays; based on designations of the total imprecision (% CV) of each assay at the 99th percentile and how many specimens from normal individuals have cTn concentrations that are actually measurable below the 99th percentile. The ultimate goal is to have all assays be level 4 guideline acceptable. A point of controversy does exist regarding the way Roche has commercially designated their cTnT assay as high-sensitivity; the evidence-based literature does not support its designation as an hs-assay because the assay consistently measures less than 50% of normal female subjects above the LOD (see Fig. 48.9 ); as such they have designated their FDA-cleared assay as Gen 5 (even though it is the same assay designated high sensitivity for sale outside the USA). The likely clinical effects of using hs-assays rated by the scorecard as “guideline acceptable” (≤10% CV) or contemporary “clinically usable” (11 to 20% CV) include: (1) all providers will more accurately detect patients within the normal range and with minor myocardial injury, independent of the pathophysiologic mechanism; (2) emergency medicine physicians will achieve improvements in triage through earlier ruling out (improved NPV and sensitivity) and ruling in (improved positive predictive value [PPV] and specificity) of patients with MI; (3) cardiology, internal medicine (hospitalists), and family practice physicians will see improved outcomes for both inpatients (hospitalized, short-term risk) and outpatients (posthospitalization, long-term risk) because of the ability to detect injury earlier, and manage patients accordingly, compared to other diagnostic tools; (4) other medical specialty physicians will be better able to identify patients early at presentation, often without clinical symptoms, who may be at risk of cardiac-related adverse outcomes; and (5) clinical trial investigators will be able to identify appropriate and optimal patient reenrollment and outcome measures.

Optimal discrimination between a small amount of myocardial injury and analytical noise requires assays that have low limits of quantitation (LOQ; lowest concentration at 20% CV) and LOD, and require low imprecision at low cTn concentrations. Efforts to improve the imprecision of cTn assays are always warranted. Irrespective of how the testing is performed, whether in the central laboratory or at the bedside using POC assays, manufacturers need to define the imprecision profile (i.e., the scatter graph showing the %CV versus increasing cTn concentrations obtained using the Clinical and Laboratory Standards Institute [CLSI] EP5-A2 protocol) by assessing pools of human samples containing different cTn concentrations. In particular, at least two cTn concentrations that cover the range between the LOD and the 99th percentile URL of the assay are recommended to be included; one around the LOD and one around the female and or male 99th percentile URL. Imprecision characteristics, including the 10 and 20% CV concentrations, the LOD, and 99th percentile data from current commercially available assays for contemporary, POC, and hs-cTn assays need to be addressed. It is concerning in the context of clinical practice that there still is some variability for an individual patient’s cTn concentration caused by different lots of calibrators and different antibodies that do not recognize the same epitopes when the same sample is measured using the same manufacturer’s assay in different laboratories. Such discrepancies highlight the need for local analytical validation whenever feasible. Laboratories at a minimum should determine total imprecision at the LOD and at each sex-specific URL and should continue to monitor assay performance over time using internal QC materials that include the lowest concentration corresponding to the 10 and 20% CVs, as well as at the 99th percentile if different from the 10% CV concentration. Patient specimen comparisons such as regression analysis and bias assessment should be performed according to CLSI guidelines and as outlined by AACC Academy and IFCC C-CB.

Many organizations, including the AACC Academy, IFCC C-CB, Global Task Force, and Cardiology and Epidemiology Societies have outlined analytical recommendations for cTn (as well as for other biomarkers of ACS). The most current recommendations by the AACC Academy and IFCC C-CB for the appropriate implementation and utilization of cardiac biomarkers, specifically for hs-cTn assays, to aid in the early rule-in and rule-out diagnosis of MI, complementing the quality specifications for contemporary cTn assays previously published, follow (see Box 48.4 ). ^,

Recommendation 1: For hs-cTn assays, laboratories should measure at least three different concentrations of QC materials at least once per day. For contemporary cTn assays, at least two different concentrations of QC materials must be assessed at least once per day. Before patient testing can be initiated, the values for acceptable imprecision must be, at a minimum, consistent with those specified by the manufacturer.

• A.

  QC concentrations for hs-cTn assays (ng/L units with one decimal place):

  • (1)

    Concentration 1: a concentration between the LOD and the lowest sex-specific 99th percentile.

  • (2)

    Concentration 2: a concentration that is higher than but close (within 20%) to the highest sex-specific 99th percentile URL.

  • (3)

    Concentration 3: a concentration that challenges the upper analytical range of reportable cTn results (e.g., multiples above the 99th percentile concentration).

• B.

  QC concentrations for contemporary (pre-hs assays) cTn assays (ng/mL or μg/L units) with three decimal places:

  • (1)

    Concentration 1: a concentration at or close to (within 20%) the overall 99th percentile.

  • (2)

    Concentration 2: a concentration that challenges the upper analytical range of reportable cTn results (e.g., manufacturer).

Recommendation 2: During initiation of hs-cTn testing, clinical laboratories should validate the LoB, LOD outside the United States, or LoQ as applicable per FDA regulations in the United States. These analytical parameters should be validated minimally on an annual basis or more frequently as deemed necessary.

Recommendation 3: Report hs-cTn in whole numbers, using ng/L without decimal points. For reporting QC values, we recommend one decimal point. For contemporary cTn assays, units are reported in ug/L to two significant figures, with QC values reported to three significant figures.

Recommendation 4: Use a defined reference population to report 99th percentile concentrations according to sex-specific cutoffs for hs-cTn assays. This recommendation is not relevant for contemporary cTn assays. A minimum of 300 men and 300 women per group of healthy individuals is required for appropriate statistical determination for the 99th percentile (the robust statistical method is not recommended) of a normal reference URL for both cTn and CK-MB.

Recommendation 5: We recommend that assays unable to detect cTn at concentrations at or above the LOD in at least 50% of healthy men and women be labeled as contemporary cTn assays.

Recommendation 6: Laboratories should communicate with clinicians on the influence of preanalytic and analytic problems that confound hs-cTn assays. For institutions or health systems using 2 or more cTn assays, differences in the sensitivity of the various cTn assays should be explained to assist clinicians in understanding discrepancies when patients are transferred from other facilities

Recommendation 7: Authors of studies using cardiac biomarkers, including hs-cTn, should document preanalytical and analytical variables important to the study and be explicit concerning their postanalytical interpretative approaches.

Recommendation 8: Commutable materials should be developed for use in harmonizing and standardizing cTn measurements.

Recommendation 9: cTn results should be reported within 60 minutes or less of when a sample is received. There should be continued efforts to improve this to a time of 60 minutes from when the sample was collected.

Recommendation 10: The laboratory should help educate clinicians on the importance of specific metrics by which true clinical changes in cTn concentrations can be distinguished from analytical and biological variabilities.

History

cTns have been available for clinical use since 1995, when the first cTn T (cTnT) assay was regulatory approved outside the USA. Since that time, enhanced diagnostics and particularly improved sensitivity in females led to more frequent and more accurate diagnosis of cardiovascular abnormalities, especially with the growing implementation of high-sensitivity cTn assays. Initial use of these assays was influenced in large part by CK-MB. CK-MB is substantially less sensitive and less cardiac specific than cTn, but because of its long use, it became entrenched in the thinking of clinicians and laboratorians. For this reason, perhaps, the first set of guidelines by the National Academy of Clinical Biochemistry (NACB) attempted to compromise between the use of these CK-MB and cTn using two cutoffs. One cutoff was selected to be equivalent to values of CK-MB in the clinical identification of patients. This cutoff, known as the receiver operating characteristic (ROC) curve–derived cutoff value, considered CK-MB the standard. However, this cutoff did not take advantage of the increased analytical sensitivity of cTn values. A second cutoff was recommended for use at a lower level, the 97.5th percentile of a normal patient population, from subjects who were poorly clinically defined, above which patients were considered to have cardiac injury. But even patients with a classic history of AMI were designated as having “unstable angina with minimal myocardial necrosis.” The guidelines followed extensive research demonstrating large numbers of patients with chest pain (roughly an additional 33%) whose prognoses in terms of subsequent frequency and timing of events were identical, and who had increased cTn values but normal CK-MB values. These findings were documented despite the fact that the initial version of cTn assays was relatively insensitive compared with contemporary assays.

This concept of using prognosis to confirm diagnosis rapidly became the paradigm not only for cTn but for the validation of most other markers as well. It was not until 1999, when a task force empaneled by the ESC/ACC took issue with these guidelines, that the concept of using only one cutoff value was initiated. ESC/ACC guidelines suggested the use, not of the 97.5th percentile of a reference population but the 99th percentile, in recognition of increased sensitivity of cTn that led to many elevations that were difficult to explain. This had the potential benefit of reducing the overlap to 1% between those with disease and normal individuals. It was conceptually important because even by this time, it was clear that enhanced sensitivity of cTn was leading to the identification of a substantial number of patients in whom the pathophysiologic cause of cardiac injury (see later sections) could not be determined. This is often the case when one moves from a relatively insensitive measure to one that has substantially greater sensitivity (i.e., when one uncovers new diagnostically and prognostically important increases). Nonetheless, clinicians had difficulty understanding this because such elevations had not been described previously and the pathophysiology was obscure. This problem has been revisited today as hs-cTn assays become increasingly sensitive and replace contemporary cTn assays. The ESC/ACC biochemical group at that time also recommended that the 99th percentile should be measurable with excellent imprecision and suggested, not mandated, a criterion of a total 10% CV or less at the 99th percentile. The group did not, however, recommend that if assays did not reach that level of precision, the decision values should be raised above the 99th percentile value. Unfortunately, this was the extrapolation of some, and it led to even greater heterogeneity in how cTn values were interpreted.

Some individuals persisted in using the NACB recommendations, some used the ESC/ACC recommendations, and some because of fear of false-positive results caused by imprecision employed a value higher than the 99th percentile, usually at the level of the 10% CV value. Add to this the fact that local laboratories often felt the need (appropriately) to validate the assays in their own hands and come up with different values, and one can see how the use of cTn values became fragmented and idiosyncratic. This problem also led to tremendous heterogeneity in the published literature and confusion on the part of clinicians.

It is now very clear that normal hs-cTn values are substantially lower than concentrations that are being reported for healthy individuals by contemporary assays. The reported concentrations in healthy individuals often represent “noise” (<LOD) in the contemporary assays. hs-cTn assays are growing worldwide, including in the United States, to have the ability to measure healthy individuals’ low concentrations accurately. For the past 2 years, the IFCC C-CB has posted and revised quarterly updates provided by manufacturers, current hs-cTn assays, along with POC and contemporary assays. A wealth of clinical information has made it clear that for contemporary assays, the 99th percentile value is the key value to use when measuring cardiac cTn. ^, Any value above the 99th percentile should be considered abnormal, indicative of myocardial injury. Concerns about false-positive results induced by such criteria should be minimal because assuming adequate quality assurance of the assays, even reduced imprecision should not lead to false-positive results provided cTn is measured in two serial samples. Usually reduced imprecision simply impairs the sensitivity of the assay. Statistical modeling of this issue suggests that the clinical impact of using even an imprecise assay is negligible with regard to false-positive results.

For this reason, along with the clinical data, the criteria in all guideline papers now include use of the 99th percentile, revised in 2018 from an overall URL to the use of individual, sex-specific 99th percentile URLs for males and females. Although some studies have suggested a return to the use of the 97.5th percentile value for high-sensitivity cTn assays, this is unlikely to occur unless important clinical reasons for such a change can be demonstrated. ^{, ,} Once one recognizes that the 99th percentile is the only criterion for abnormality, one can progress to understanding how to use hs-cTn assays with sex-specific URLs in a variety of clinical situations for early rule out, early rule in, and risk assessment. Numerous publications have clearly demonstrated this point, which will be discussed in the following section.

Use of cardiac troponin for the diagnosis of acute myocardial infarction

AMI is a state characterized by abnormalities between nutrient perfusion and myocardial oxygen consumption. It is usually diagnosed when abnormalities in coronary flow at least in part contribute to the pathogenesis of cardiac injury. Because of the high sensitivity and specificity of cTn for the heart, cTn has become the cornerstone of the diagnosis of MI (see Boxes 48.2 and 48.3 ). As shown in Fig. 48.16 high-sensitivity cTn assays eliminate the noise of contemporary (and POC) assays, providing greater reliability of true measurement of cTn at the 99th percentile upper reference interval, and is now the standard assay of use in all clinical and research laboratories if available. An increased concentration of cTn greater than the 99th percentile URL is required in the appropriate clinical setting along with a rising or falling pattern for the diagnosis of AMI. With the growing use of high-sensitivity assays, earlier and more accurate diagnoses for ruling in and ruling out MI are being made within 1 to 3 hours of first blood draw compared to contemporary and POC assays. At A Glace 3 shows schematics of representative strategies to consider for early rule in and rule out based on implementation of hs-cTnI testing. Specifics of the MI definition are as follows.

The clinical substrate (usually risk factors for AMI in the patient) is key. An increased cTn value is not synonymous with the diagnosis of MI. First, risk factors and the presentation (most often chest discomfort) must be compatible with the diagnosis. At times, imaging, whether performed at the time of intervention by angiography or thereafter, may provide proof of the substrate. ^{, , ,} Many clinical situations can mimic AMI, the most common of which is myocarditis. Apical ballooning is another. Because of the increased sensitivity of cTn, clinicians need to be astute to the possibility that a given elevation in cTn, even with a rising pattern suggesting that it is acute, may not be due to ischemic heart disease (see Box 48.1 ). In addition, many groups such as women, the elderly, and diabetics can present atypically so interdigitating the biomarker information with the clinical is essential. This is conceptually difficult for some clinicians because when CK-MB, which was reasonably insensitive, was used, most substantive elevations were associated with coronary heart disease, albeit not all.

Acute events leading to cardiac injury (see Box 48.1 ) usually will manifest a changing pattern of values. ^, The rationale for this is that as assays have become more analytically sensitive, it has become clear that elevations of cTn can exist chronically. ^, The most overt case of this is seen in patients with renal failure. This is also the pattern manifest when artifactual increases occur as the result of analytical problems caused by interferences from human anti-mouse antibodies and heterophilic antibodies in the blood of some patients. ^, However, hs-cTn assays do not appear prone to as many analytical problems as early generations of assays. Nevertheless, a list of assay interferences involving hemolysis and biotin as potential interferents that are assay dependent is available on the IFCC C-CB website. Although patients with chronic elevations (not those with artifactual causes) are at long-term high risk, they are not necessarily at high risk over the short term. For this reason, the guidelines groups have advocated the need for a rising pattern of cTn values in patients who present early after the onset of symptoms, because such increases usually are indicative of an acute process. However, the acute process may not be AMI. Whether AMI is present is a clinical decision. When cTn values are not rising, one might seriously question whether acute disease of any kind is present or whether the elevations are of a more chronic nature. Care is necessary to ensure that one does not miss individuals who come in late after the onset of symptoms, whose cTn values may be near peak values, or whose values are on the long persistent tail of the time-concentration curve. They may appear not to have a changing pattern of values because they are near the peak (that can range from 12 to 48 hours) of the time-concentration curve or on the gradual downslope. This may occur in greater than 20% of patients.

With hs assays, there will be more AMIs that are not due to plaque rupture (so called supply-demand type 2 AMI, which has its own ICD-10 code) because these usually elaborate less cTn than plaque rupture events (type 1 AMI). ^, However, the criteria for a rising and/or falling pattern do not differ. These will occur more frequently in women than in men. There are good data that the use of sex-specific cutoff values with hs-cTnI assays (limited data for hs-cTnT assays) improve the diagnosis of these patients. While switching from a contemporary cTnT (fourth generation) to the hs-cTnT or Gen 5 cTnT assays is associated with an increased number of patients with the diagnosis of MI, this is not the case for replacing a contemporary to a hs-cTnI assay, as shown in Fig. 48.17 .

The guidelines distinguish five types of MI (see Box 48.3 ); two will be considered here because they both manifest with chest discomfort and increased biomarkers. One (type 1 MI) is the so-called spontaneous, or “wild type,” in which acute plaque rupture leads to some degree of thrombosis, or an episode in which platelet accretion occurs on a plaque, leading to thrombosis. ^, The second type (type 2 MI) includes coronary abnormalities with possible supply-demand imbalance, vasospasm, or endothelial dysfunction, which provides evidence of myocardial injury. ^,

Individuals with spontaneous AMI can have STEMI or non-STEMI as indicated by the ECG pattern they manifest. ^{, ,} Data indicate that immediate treatment aimed at opening the occluded artery is mandatory for patients with STEMI. This should be done based on the ECG pattern alone, even before biomarker values become available to assist with diagnosis. Opening of the artery is currently done with primary PCI and/or thrombolytic agents. The former is preferred when the two are available in similar time frames. With prompt coronary recanalization, the amount of myocardium lost is minimized and mortality is reduced. It should be noted that coronary recanalization increases the rapidity of biomarker release, and thus the rate of rise of the time-concentration curve is increased and the time to peak values is shortened.

A non-STEMI is less often associated with total coronary occlusion and usually is identified by ECG changes, which show not ST segment elevation but rather ST segment depression or T-wave changes. Given the sensitivity of hs-cTn for this diagnosis, the ECG is even at times totally normal. Patients who present with chest pain as a result of CAD should have risk factors, an appropriate presentation, or imaging evidence of this syndrome. Patients with an increased cTn, and the recommended 99th percentile URL are known to have more severe coronary heart disease than individuals without increased cTns. Likely for this reason they also have more procoagulant activity. Multiple intervention studies have shown that these patients benefit from aggressive anticoagulation, including heparin (the data for low-molecular-weight heparin are much stronger than the data for unfractionated heparin), glycoprotein IIb/IIIa antiplatelet agents, and an early invasive strategy consisting of PCI or coronary artery bypass grafting (CABG). Use of these strategies in patients without increased cTn values has been shown to be of no benefit and, in some trials, has actually proved detrimental. For these patients, expeditious but not necessarily immediate coronary interventions are suggested. Studies using hs-cTn assays have demonstrated that within the sex-specific reference interval for hs-cTn assays ( Fig. 48.18 A and B) as has been documented for patients above the 99th percentiles Fig. 48.19 , increasing levels of risk are associated with increasing cTn concentrations. However, given the increased frequency of type 2 MIs, it is unclear that invasive intervention will be necessary in all such patients. However, it is now well documented that among patients presenting to the EDs with an increased cTn, those with type 2 MI and myocardial injury from non-MI pathophysiology are at greater risk of adverse outcomes compared to type 1 MIs ( Fig. 48.20 ). Importantly, although patients with nonSTEMI who have an increased cTn value benefit from an invasive strategy, those who do not, as a group, do worse. ^,

Not all patients have easily identifiable culprit lesions in which to intervene, and this explains some apparent spontaneous AMIs without severe CAD. ^{, , ,} These patients, often women, may have endothelial dysfunction, coronary vasospasm, coronary dissection, or some other transient process that has resolved before investigation. These patients have type 2 MI according to the guidelines. This group might also include those individuals who have fixed but stable CAD but have some degree of damage as a result of excessive myocardial oxygen consumption, such as that caused by severe tachycardia, hypotension, or hypertension. Coronary arteries that are diseased tend to vasoconstrict rather than dilating in response to stimuli that would normally evoke vasodilation and increases in coronary flow. These patients are probably fundamentally different from individuals who have spontaneous MIs and likely are not a subgroup that has nearly as much procoagulant activity; it is unclear whether they benefit from interventional therapy. This may be one of the reasons why women who tend to have more severe endothelial dysfunction seem to have a different profile. ^, Nonetheless, depending on the specific study reviewed, 10 to 30% of patients may not have an identifiable lesion that has caused the event, leading to consideration of this alternative pathophysiology. From the perspective of diagnosis, however, the cTn criteria are identical.

Distinctions among the MI types have to be made on the basis of clinical characteristics and other diagnostic studies. It is now clear that women can have type 2 MI secondary to spontaneous coronary dissection. It used to be thought that this occurred almost exclusively during pregnancy, but that is clearly not the case. This diagnosis can be easily missed angiographically and is associated in at least half of patients with coronary tortuosity and evidence of fibromuscular dysplasia in other vascular beds. Preliminary data suggest that unless these patients are hemodynamically compromised or have total coronary occlusion, a strategy of watchful waiting rather than intervention may be preferred.

Other types of MIs are recognized. First, one type does not use biomarkers at all. This includes patients who may present classically with both chest pain and ECG changes but succumb either before blood tests can be obtained or before enough time has elapsed for the circulating concentration of cTn to exceed the 99th percentile. Two other types of MIs have been designated; these are related to PCI and CABG and will be covered later under “Special Situations.”

AT A GLANCE

Strategies for ruling in and ruling out acute MI with contemporary and high-sensitivity cTn assays.

After percutaneous coronary intervention

It has long been appreciated that interventions done on the coronary arteries can result in the release from heart muscle of biomarkers indicative of cardiac injury. ^, For this reason, a variety of criteria have been promulgated, starting years ago with CK-MB. It was shown initially that elevations of CK-MB that occurred after the procedure were highly predictive of adverse events over the long term. These findings were difficult to explain because the amount of injury involved often was very minor but was nonetheless easy to document with sophisticated techniques such as cardiac magnetic resonance imaging (cMRI). However, controversy about the mechanisms continued and a large number are being hypothesized. The advent of cTn biomarkers has rendered this issue far less complex. Patients who have elevations of circulating cTn and who present with acute coronary problems have much more severe abnormalities in coronary anatomy than those who do not have cTn elevations. In recent studies, the baseline cTn value has been added to the analysis of these post-PCI elevations. ^, This was done for many years, but not with assays that measured cTn concentrations even close to the 99th percentile. Accordingly, it is only recently that the data have confirmed several important conceptual issues.

The first concept is that a vast majority of patients with significant elevations (threefold to fivefold increase in CK-MB or cTn) that in the past were associated with an adverse prognosis are those with increased cTn values at baseline. This usually (although not always) is indicative of an acute presentation and therefore a rising pattern of values. In the setting with a rising pattern of values, it is hard to know whether one should attribute the increases to the PCI or to the initial insult. Nonetheless, when one incorporates the baseline value into the analysis, most of the prognostic significance of the post procedure cTn values is ablated. Data suggest that it is only at an elevation of cTn at baseline that marked post-procedure elevations in cTn are apt to be frequently found, suggesting that in most instances at least a blend of the two processes occurs. These issues have been confounded not only by failure to use appropriate cutoffs for cTn, but by the use of insensitive cTn assays. More recent guidelines clarify this circumstance substantially by indicating that if one has an increased cTn at baseline before PCI, the diagnosis of post-PCI injury is confounded and cannot be made. Thus a normal baseline cTn is required in almost all instances to identify procedure-related myocardial injury. If the baseline cTn is increased, subsequent increases may or may not be able to be attributable to the procedure itself and a diagnosis of post-PCI injury cannot be made with absolute confidence.

Marked elevations, as indicated previously, are uncommon but can occur. When they do occur, they are rarely of cardiovascular importance, in the sense that the more marked elevations are easily presaged by clinical information at the time of the procedure. Thus post-PCI cTn values add very little that is new. In addition, it is unclear whether patients have an adverse cardiovascular prognosis over time. In the most recent analysis done by using the most appropriate cutoff values and contemporary cTn assays, borderline statistical significance was attributed to post-PCI elevations. However, most events that did occur were noted during procedures in patients who had severe underlying noncardiac comorbidities and therefore underwent palliative procedures. The event rate, if patients with non–cardiac-related subsequent complications are excluded, would not have been of statistical importance, even if their inclusion were of only borderline significance. Thus it now appears that collection of this information after PCI is not necessary. Recent data suggest that values even modestly below the 99th percentile upper reference interval have important prognostic significance, presaging in our opinion the likelihood that with hs-cTn assays, this effect will be strengthened and the prognostic impact of post-PCI values attenuated still further.

Nonetheless, criteria still exist for these occasional patients who may have post-PCI injury. The appropriate cutoff cTn value is unclear, but it is clear that very few patients will reach the marked fivefold elevation previously advocated if the baseline cTn is not increased. ^, This is likely to change, however, with hs-cTn assays. If cTn is increased post-PCI more than fivefold, with a normal baseline cTn value assumed, the patient can be diagnosed as having had a periprocedural MI if they have had symptoms or have ECG changes and/or had complications of the procedure. No specific therapy is mandated. Regarding biomarker changes that occur after reperfusion of an occluded vessel after PCI or thrombolytic therapy in AMI patients, greater than twofold increases in biomarkers occur within 90 minutes of reperfusion, the rate of increase of biomarkers within the first 4 hours separates reperfused from nonreperfused patients, and after myocardial reperfusion in AMI patients, washouts of all biomarkers parallel each other. However, increased washout does not define the level of reperfusion and cannot be used to define a post-PCI AMI.

One unique subset needs to be considered, and that is the small group of patients who may have chronic preprocedural elevations in cTn. These individuals will manifest a rising pattern, albeit from an increased baseline, when they have events, including post-PCI events. Accordingly, what has been suggested is that the criteria used for reinfarction of an increased value of 20% or more should be used to define post-PCI infarction in this group.