Cardiac markers: troponin, CK-MB and natriuretic peptides

Cardiac markers are blood tests that tell you something about whether heart muscle has been injured and how stressed it is. This page covers the four that a medical student meets again and again: cardiac troponin (the marker of myocardial injury), creatine kinase (CK) and its MB isoenzyme (older markers still met in practice and in questions), myoglobin (an early but non-specific marker), and B-type natriuretic peptide (BNP) with its N-terminal fragment NT-proBNP (markers of cardiac wall stress used mainly in heart failure).

By the end you should be able to:

1. Say why troponin is the marker of myocardial injury, and define the 99th percentile upper reference limit.
2. Interpret a 0-hour and 1-hour (or 0-hour and 3-hour) high-sensitivity troponin pathway using a published rule-in and rule-out algorithm, and state what a rise and fall means.
3. List the non-ischaemic causes of troponin elevation and explain why they matter for the diagnosis of myocardial infarction.
4. Explain what CK and CK-MB measure, why they are largely replaced by troponin, and why CK-MB measured without troponin can mislead.
5. Interpret BNP and NT-proBNP cut-offs and list the conditions and patient factors that push them up or down.
6. Read three sample result sheets end to end, reaching a one-sentence clinical conclusion with the next step for each.
7. Avoid the common errors that make a normal marker look abnormal, and vice versa.

A cardiac troponin is the test of choice for detecting myocardial injury. The Fourth (2018) and Fifth (2026) Universal Definitions of Myocardial Infarction define injury as a troponin above the sex-specific 99th percentile upper reference limit, and they define infarction as injury caused by myocardial ischaemia, alone or together with evidence such as symptoms, electrocardiographic change, imaging abnormality or intracoronary thrombus. ^[1] Troponin is released from cardiac myocytes, but not every elevation is infarction: injury is common in acute illness of almost any kind, and interpreting it demands the clinical picture, not the number alone. ^[1]

A BNP or NT-proBNP elevation behaves differently. It is not a marker of cell death; it is a marker of ventricular wall stretch. It is high in heart failure and in states that raise filling pressures, and it is low or normal when the problem is not cardiac. NICE uses NT-proBNP in the heart failure pathway, and the Universal Definition community and other guidelines use BNP or NT-proBNP to support or exclude heart failure. ^[4][6]

The older markers have a smaller role. Creatine kinase (CK) is a muscle enzyme and its MB isoenzyme was once the standard for myocardial infarction. With high-sensitivity troponin assays available, the Universal Definition prefers troponin, which has better sensitivity and specificity; CK-MB is reserved for when troponin is unavailable. ^[7][1] Myoglobin is released very early, but it is present in all muscle and is not cardiac-specific, so it no longer has a routine role. ^[8][1]

Troponin is the regulatory complex that controls calcium-mediated actin-myosin interaction in striated (cardiac and skeletal) muscle. Cardiac troponin T (cTnT) and cardiac troponin I (cTnI) are the components specific to cardiac myocytes. In the blood, small amounts circulate in healthy people; troponin measured by a high-sensitivity assay is detectable in most apparently healthy adults, so the reference interval is built around the 99th percentile of a healthy population rather than around a "zero" result. ^[1]

Released, measurable, raised. Troponin is soluble and bound to the contractile apparatus. Injury that damages sarcolemmal integrity releases it; the amount rises above the 99th percentile within a few hours of the onset of ischaemia in most infarcts, peaks around one to three days, and falls over several days for cTnI and up to two weeks for cTnT. ^[1] Laboratory methods differ: cTnT is made by one manufacturer worldwide; cTnI is made by many manufacturers, and each has its own antibody, calibrator and 99th percentile. You cannot compare cTnI results from two different assays numerically, and you should not switch assays mid-series without reporting the change. ^[1]

High-sensitivity assays. An assay qualifies as high-sensitivity when troponin is detectable above the limit of detection in at least 50 percent of healthy people and its coefficient of variation at the 99th percentile is below 10 percent. ^[1] Such assays are used with early-triage pathways (0/1 hour and 0/2 to 3 hour algorithms) because their low analytical noise allows small changes to be trusted. ^[2][1]

Creatine kinase (CK, also called CPK) is an enzyme that regenerates ATP in tissues that turn over energy fast, notably cardiac muscle, skeletal muscle and brain. Total CK in the blood rises when any of these tissues is injured, most often skeletal muscle after exercise, an injection, a fall or rhabdomyolysis. It is therefore not a cardiac marker unless you know which tissue was damaged. ^[7][8]

CK-MB is the isoenzyme made of one muscle-type (M) and one brain-type (B) subunit. It exists in both cardiac and skeletal muscle, but it is far more abundant in cardiac muscle, so an elevated CK-MB used to be read as cardiac injury. It is usually reported on its own as a mass concentration, and historically as a ratio to total CK, which was thought to point to a cardiac rather than a skeletal source. ^[7][1]

The Universal Definition now prefers cardiac troponin. Troponin has better sensitivity, rises for longer, and is more specific for myocardial injury; CK-MB is less sensitive and less specific, and some laboratories no longer offer it. ^[7][1] CK-MB by mass assay remains an acceptable alternative only if no cardiac troponin assay is available. ^[1] One specific modern use survives: if a person has a recent infarct and troponin is still elevated weeks later (cTnT often stays up), a new CK-MB rise can date a second event. ^[1]

Myoglobin is the oxygen-binding protein of muscle. It is found in cardiac and in skeletal muscle in roughly equal abundance, so an elevated myoglobin is a marker of muscle injury, never a marker of cardiac injury specifically. It appears in the blood earlier than troponin after ischaemia, but its lack of cardiac specificity, and the universal availability of high-sensitivity troponin, mean it now has no routine role in the diagnosis of myocardial infarction. ^[8][1][7]

Do not confuse myoglobin with troponin in a late presenter. A patient admitted days after chest pain may have a normal myoglobin but a raised troponin, because the early marker has already cleared.

B-type natriuretic peptide (BNP) is a hormone released by ventricular myocytes when the wall is stretched. It is synthesised as proBNP, cleaved to the active 32-amino-acid BNP and the inactive NT-proBNP (N-terminal pro-BNP), which has a longer half-life and greater plasma stability. ^[6] The active peptide causes natriuresis, diuresis and vasodilatation and also blocks renin and aldosterone release. ^[6]

BNP and NT-proBNP are not markers of cell death. That is the single most important distinction in this page. A high troponin says "myocardial injury"; a high BNP or NT-proBNP says "the ventricle is under stretch". In practice BNP and NT-proBNP are used to support or exclude heart failure, to gauge its severity, to monitor treatment, and to stratify risk in pulmonary embolism and acute coronary syndromes. ^[4][5][6]

NT-proBNP is cleared partly by the kidney and partly by neutral endopeptidase and receptor-mediated uptake. BNP is cleared by receptor binding and neutral endopeptidase. Because their clearance differs, the two analytes are not numerically interchangeable, and each assay and each guideline has its own cut-off. ^[6][4]

The values in the table are the adult decision limits quoted in the sources used for this page. They are not universal: high-sensitivity troponin varies by assay and sex, BNP and NT-proBNP vary by assay and age, and routine chemistry varies by method and laboratory. The interval printed by the laboratory that performed the test is the one to use, and when comparing serial samples the assay must be the same. ^[1][2][4]

A note on units. These decision limits are assay-specific for troponin and for the natriuretic peptides. A hs-cTnT of 14 ng/L means different things on an hs-cTnI platform. Always read the printed method and the local validated cut-off, and read serial samples on the same platform. ^[1][2] The ordinary unit for CK is units per litre; CK-MB by mass is usually reported in micrograms per litre or nanograms per millilitre. ^[1]

The course of a marker matters as much as its absolute level. Troponin changes over hours to days; a value that is rising, or that has risen from a prior documented normal, is more informative than a single raised value, especially when the patient presents early or has a condition that keeps the baseline high. ^[1] For BNP and NT-proBNP the issue is different: these peptides move with filling pressures and treatment, so a high value is interpreted with the clinical state, the heart failure score and the echocardiogram. ^[4][6]

Troponin elevation means myocardial injury. It is not the same as myocardial infarction, and this is where most clinical mistakes are made. ^[1] The table lists the usual causes of elevation for each marker. Low or undetectable values mostly tell you that the analyte is not in play: they have little clinical meaning on their own beyond allowing you to rule out the condition the test was taken to exclude.

The "causes of high values" list for troponin is deliberately long because an elevated troponin without infarction is common. The distinction between type 1 myocardial infarction (atherosclerotic plaque rupture or erosion with thrombus) and type 2 myocardial infarction (myocardial infarction from oxygen supply-demand imbalance without acute plaque rupture, for example severe anaemia, tachycardia or hypotension) is also in the Universal Definition, and it changes treatment. ^[1] A chronically elevated troponin without a rise-and-fall pattern is chronic myocardial injury, not infarction, and usually needs investigation of the underlying structural or systemic disease rather than antiplatelet therapy. ^[1]

Named patterns turn a list of results into a working diagnosis.

• No injury, normal ECG, low-risk presentation: troponin under the 99th percentile at 0 and 1 hour or 0 and 3 hour with a normal ECG supports ruling out infarction; the ECG and the time of onset still matter. ^[1][2]
• Acute injury with rise and fall plus ischaemic symptoms or ECG change: the Universal Definition's acute myocardial infarction, either type 1 or type 2 depending on mechanism. ^[1] Serial sampling is what shows the rise and fall. ^[1]
• Acute injury with rise and fall but no ischaemia: acute myocardial injury (myocarditis, sepsis, Takotsubo, procedural). This is not infarction, although it can still carry a serious prognosis. ^[1]
• Chronic elevation without a rise and fall: chronic myocardial injury, most often from coronary disease, left ventricular hypertrophy, valve disease or chronic kidney disease. ^[1]
• ESC 0/1-hour high-sensitivity troponin T pathway: 0-hour under 5 ng/L with symptoms over 3 hours before sampling rules out; 5 to 11 ng/L at 0 hours with less than 3 ng/L change at 1 hour rules out; 52 ng/L or more at 0 hours, or a change of 5 ng/L or more, rules in; the remainder go to an observation group. ^[2] The pathway is assay-specific: the same algorithm on an hs-cTnI assay uses its own 99th-percentile-based cut-offs, and the source of these numbers is the Roche hs-cTnT platform described in the cited implementation study. ^[2]
• ESC 0/3-hour pathway: 14 ng/L or less at 0 hours when symptoms began over 6 hours before sampling rules out; 14 ng/L or less with a 0 to 3 hour change of 7 ng/L or less rules out; a larger change or a raised 0-hour value needs further assessment. ^[2]
• Rule-out with BNP under 100 pg/mL and NT-proBNP under 300 pg/mL: heart failure is unlikely in a patient with dyspnoea, particularly in the emergency department. ^[5][6]
• BNP over 400 pg/mL, NT-proBNP over 900 pg/mL or over 2000 ng/L: supports significant heart failure or urgent specialist assessment, depending on the guideline. ^[4][5][6]
• Grey zone 100 to 400 pg/mL for BNP, or 300 to 900 pg/mL for NT-proBNP: the peptide alone does not decide. Add echocardiography, natriuresis status, renal function, obesity and the clinical syndrome. ^[5][6][4]
• CK-MB raised but troponin normal or minimally changed: the rise is almost certainly skeletal muscle in origin. Troponin has already ruled out significant myocardial injury. ^[7][1]
• Discordant methods: a hs-cTnT result from one month ago and a hs-cTnI from today cannot be trended against each other; use one assay, one laboratory and one reference interval. ^[1]

The Universal Definition also separates myocardial infarction into types: type 1 (atherosclerotic plaque disruption with thrombus), type 2 (supply-demand imbalance), type 3 (sudden cardiac death with evidence of ischaemia), type 4a and 4b (procedure-related or stent thrombosis) and type 5 (coronary artery bypass graft-related). ^[1] Procedure-related infarction uses much higher thresholds (for example over 5 times the 99th percentile after percutaneous intervention, or over 10 times with supporting features after bypass surgery), and these are different enough from the 0/1-hour thresholds that they should not be confused. ^[1]

The values below are invented for teaching and are labelled as illustrative. They are not real patient data and are not a diagnosis. The thresholds and rules applied to them are taken from the sources cited, and the ranges that are method-specific are reported as such.

1. hs-cTnT is under the sex-combined 99th percentile of 14 ng/L at 0 h and does not rise meaningfully at 1 h, so there is no acute myocardial injury. ^[1][2]
2. The 0/1-hour pathway classifies an under-5 plus 3-hour-symptoms result as rule-out; a 6 with under a 3 change in this clinical context is still in the rule-out band when symptoms pre-date sampling. ^[2]
3. BNP under 100 pg/mL makes acute heart failure unlikely, and CK-MB is not raised, so skeletal muscle and cardiac injury are not in play. ^[5][7]
4. The ECG is still required and is the third part of rule-out, alongside troponin and assessment.

Conclusion: This is a low-risk presentation with no current myocardial injury and no heart failure biomarker elevation, so the next step is observation or reassurance with written safety-netting and re-review if symptoms recur, not ACS treatment.

1. Troponin rises from 9 to 18 ng/L, a change of 9 ng/L, which is above the 5 ng/L rule-in change of the hs-cTnT 0/1-hour pathway, so this is acute myocardial injury. ^[2]
2. There is no ST elevation mentioned, but troponin-defined injury in a patient with ischaemic chest pain supports non-ST-elevation myocardial infarction until the mechanism (type 1 or type 2) is clarified. ^[1][3]
3. CK and CK-MB rise together with troponin; myoglobin is still normal, which illustrates that not all early markers move at the same time. ^[8][7]
4. BNP of 180 pg/mL is in the grey zone and should not be used alone; echocardiography and the overall syndrome matter more than this number. ^[5]
5. NICE reminds us that biochemical markers (troponin I or T) are part of the blood panel, measured alongside creatinine, glucose and haemoglobin, with the ECG and risk assessment. ^[3]

Conclusion: This pattern is acute myocardial injury that meets the rule-in change on a high-sensitivity troponin T pathway, so the next step is to treat as a possible non-ST-elevation acute coronary syndrome with anti-ischaemic and antithrombotic management and urgent cardiology review while the ECG and cause (type 1 versus type 2) are clarified.

1. Troponin is under the 99th percentile and stable, so there is no acute myocardial injury; the chronic elevation expected with renal impairment is not present, which argues for a preserved myocardium. ^[1]
2. BNP over 400 pg/mL supports significant heart failure, and NT-proBNP over 2000 ng/L places this in the NICE urgent referral range (over 2000 ng/L, 236 pmol/L). ^[4][5]
3. Creatinine is raised, which both lowers confidence in a single NT-proBNP value and argues for interpreting the natriuretic peptide with renal function and the echocardiogram rather than alone. ^[6]
4. A value this high carries an adverse prognosis, so cardiac assessment should not be delayed. ^[4]

Conclusion: This is probable decompensated heart failure with supportive natriuretic peptide elevation in the urgent range, so the next step is cardiology or heart-failure assessment with transthoracic echocardiography within two weeks, with review of renal function and diuretic therapy.

• Treating any troponin elevation as infarction. Most elevated troponins in hospital are not type 1 myocardial infarction. Myocarditis, sepsis, uncontrolled tachycardia, renal failure and procedural injury all raise it. Find the mechanism before you prescribe antithrombotics. ^[1]
• A single troponin. Acute myocardial injury is a rise and fall (or a fall). A single value can miss an early infarct or mislabel a chronic elevation. Serial samples at 0/1 or 0/3 hour are part of the diagnosis. ^[1][2]
• Applying the algorithm across different assays. The published cut-offs (5, 12, 52 ng/L) belong to the high-sensitivity troponin T assay. An hs-cTnI assay has its own cut-offs; do not transplant numbers. ^[2][1]
• Comparing serial values from two different methods or laboratories. Assay, platform and calibration all shift the number. Use one assay, one laboratory and the printed reference interval. ^[1]
• Ignoring sex. Women have lower 99th percentiles than men for both cTnT and cTnI. Using a male-biased combined cut-off under-recognises infarction in women. ^[1]
• Forgetting renal disease. Baseline troponin rises as glomerular filtration falls; interpret the change, not the absolute, and do not dismiss a rise in a patient with kidney disease. ^[1]
• Using CK-MB as the primary marker. CK-MB has lower sensitivity and specificity than troponin, and not all laboratories still measure it. Prefer troponin; reserve CK-MB for troponin-unavailable settings or for dating a second infarct. ^[1][7]
• Trusting myoglobin. Myoglobin is released from all muscle and is non-specific. An elevated value does not prove cardiac injury, and a normal one does not exclude infarction, because it clears early. ^[8][1]
• Reading BNP and NT-proBNP as a heart failure "diagnosis". They support or exclude heart failure in the right clinical context. Obesity lowers BNP (and peptides are cleared differently), renal impairment raises NT-proBNP, and high age and arrhythmia raise both. Interpret with the echocardiogram and the syndrome. ^[4][5][6]
• Mixing the natriuretic peptides. BNP and NT-proBNP have different half-lives, clearance, reference intervals and units. Compare like with like, and do not convert one into the other by imagination. ^[6]
• Treating 48 to 53 mmol/mol HbA1c style cut-offs as universal. Natriuretic peptide cut-offs in the NICE pathway are for heart failure diagnosis and referral, not for ischaemia, and are not individual treatment targets. ^[4]
• Waiting for biomarker results before acting on a sick patient. The ECG, the clinical judgement and the red-flag criteria come first. Troponin is a confirmatory, not therapeutic, decision. ^[3][1]

The ESC 0/1-hour algorithm quoted above derives from European cohorts and has been implemented in large emergency department studies. ^[2] In the cited implementation cohort the rule-out arm safely discharged most low-risk patients, and the rule-in arm concentrated risk, while the standard ESC 0/3-hour protocol remained an acceptable alternative. ^[2] The ESC acute coronary syndrome guidelines of recent years have adopted high-sensitivity troponin algorithms as the default diagnostic pathway, and NICE lists troponin among the key blood tests in the early assessment of acute coronary syndrome. ^[3] NICE does not give a single numeric troponin cut-off in its ACS guideline; it delegates the threshold to the assay and to the 99th percentile-based algorithms of the specialist bodies. ^[3][1]

The natriuretic peptide cut-offs used here come from two different traditions. The BNP review gives simple emergency-department thresholds: under 100 pg/mL to rule out and over 400 pg/mL to diagnose. ^[5] NICE uses NT-proBNP, which is more stable and more predictable in renal impairment: under 400 ng/L (47 pmol/L) makes heart failure less likely, 400 to 2000 ng/L (47 to 236 pmol/L) leads to referral within six weeks, and over 2000 ng/L (236 pmol/L) is an urgent referral. ^[4] A comparative study supports positivity at BNP over 100 pg/mL and NT-proBNP over 300 pg/mL, with a further elevation band above 900 pg/mL. ^[6] All of these numbers are platform and population specific and should be applied with the clinical context. ^[4][5][6]

• Cardiac troponin is the blood marker of myocardial injury, not of infarction. An infarct is troponin elevation due to myocardial ischaemia, with supporting evidence. ^[1]
• The threshold for injury is the 99th percentile upper reference limit, which is assay specific and, where available, sex specific. ^[1]
• High-sensitivity assays make troponin detectable in most healthy people and support 0/1-hour and 0/3-hour diagnostic algorithms. ^[1][2]
• A rise and fall pattern separates acute from chronic myocardial injury. ^[1]
• Troponin is raised by many non-ischaemic conditions: myocarditis, sepsis, Takotsubo, pulmonary embolism, tachyarrhythmia, stroke, drugs, and chronic kidney or structural heart disease. ^[1]
• Serial samples in one assay, in one laboratory, is the only safe way to interpret an early-attended presentation. ^[1][2]
• CK and CK-MB are older, less sensitive and less specific markers; troponin is preferred, and CK-MB is used only when troponin is unavailable or for dating a reinfarct. ^[1][7]
• CK is not cardiac-specific. Its MB isoenzyme is enriched in cardiac muscle but is present in skeletal muscle, so a normal troponin trumps a raised CK-MB. ^[7][1]
• Myoglobin is released early but is non-specific to cardiac muscle and no longer has a routine diagnostic role. ^[8][1]
• BNP and NT-proBNP are markers of ventricular wall stretch used to support or exclude heart failure, not markers of cell death. ^[4][5][6]
• BNP under 100 pg/mL rules out heart failure in the emergency department; BNP over 400 pg/mL supports the diagnosis; the 100 to 400 pg/mL range is a grey zone. ^[5]
• NT-proBNP under 300 pg/mL makes heart failure unlikely; NICE stratifies referral at under 400, 400 to 2000, and over 2000 ng/L with urgent action for the last. ^[4][6]
• Renal impairment strongly raises NT-proBNP; obesity lowers BNP; age, sex, LVH and arrhythmia move both. ^[4][5][6]
• The marker reports the biochemistry, not the whole patient. Combine the result with the history, the ECG, the renal function and, for BNP, the echocardiogram. ^[1][4]