HS troponin-T to rule out MI

Published

Bandstein N, et al. "Undetectable high-sensitivity cardiac troponin T level in the emergency department and risk of myocardial infarction". Journal of the American College of Cardiology. 2014. 63(23):2569-2578.
PubMed • Full text

Clinical Question

Among patients presenting to an ED with chest pain who had a high-sensitivity troponin-T <5 ng/L and no EKG changes, what is the 30 day incidence of MI?

Bottom Line

Among patients with chest pain presenting to one of two EDs in Sweden, 30 day incidence of MI was low (0.17%) for those with a high-sensitivity troponin-T <5 ng/L and no EKG changes.

Major Points

Acute chest pain is a common complaint in the emergency department but is caused by MI a minority of the time. Traditional troponin assays lag behind onset of necrosis-related symptoms in MI. As a result, these labs must be checked serially (eg, q6h x3) to 'rule out' individuals presenting with chest pain or other symptoms concerning for MI. A more rapid method to rule out ED patients with chest pain would result in savings for the health care system. The 2012 ROMICAT-II study found that use of coronary CT scanning in evaluating acute chest pain was associated with an 8 hour reduction in hospital stay but no cost savings. High-sensitivity (HS) troponin-T is thought to be an improvement over the traditional troponin study as it more rapidly detects myocardial damage. Prospective studies have found a high NPV when HS troponin-T was measured serially (at least twice) among patients presenting with acute chest pain. Whether a single measurement would be associated with a high NPV in all-comers was unclear.

This 2014 publication by Bandstein and colleagues retrospectively reviewed ED records from 2 Swedish hospitals. Patients >25 years old were included if they presented with chest pain and had >1 HS troponin-T measurement. Those with EKG changes an a low HS troponin-T measurement (<5 ng/L) were excluded. The incidence of MI at 30 days in individuals with low HS troponin-T measurement and no EKG changes was 0.17% and a NPV approaching 100%.

The trial has several limitations. As there was no reported duration of symptoms among patients, it's unclear if the assay was used rapidly to rule out patients as they presented to the ED as would be expected with individuals with suspicion of ACS. Comparisons between the three study arms is not equivalent as they varied on multiple baseline characteristics and those with abnormal EKGs were excluded from the low HS troponin-T group but not from the other two groups. The lack of active outcome ascertainment may have resulted in a falsely low incidence of the primary outcome. Finally, as this trial was not an RCT, discussion about outcomes like cost savings is purely speculative. This is particularly a concern as patients who may have been ruled out for ACS by the conventional troponin assay may have a positive HS troponin-T assay, which may increase utilization of high cost diagnostics (eg, angiography) without clinical benefit.

The FDA has yet to approve this assay for use in the US. While this publication suggests that there may be clinical utility of HS troponin-T in rapid rule-out of ED patients with chest pain, the true implications of the assay will not be known without an RCT.

Guidelines

As of March 2015, no guidelines have been published that reflect the results of this trial.

Design

• Two center retrospective cohort study
• N=14,636
  • HS troponin-T <5 ng/L (n=8,907)
  • HS troponin-T 5-14 ng/L (n=3,150)
  • HS troponin-T >14 ng/L (n=2,579)
• Setting: 2 hospitals in Sweden
• Enrollment: 2010-2012
• Follow-up: 1 year
• Primary outcome: MI by 30 days

Population

Inclusion Criteria

• ED patients age >25 seeking treatment for chest pain
• ≥1 HS troponin-T

Exclusion Criteria

• Excluded if EKG changes (only those with HS troponin-T <5)

Baseline Characteristics

Of all patients. Table 1 on pg. 2572 describes patients by HS troponin-T measurement. Those with higher measurements were older, had more chronic diseases, and were taking more medications than those in the <5 ng/L group.

• Demographics: Age 55 years, female 48%
• PMH: DM 10%, COPD 3%, MI 8%, CVA 5%, HF hospitalization 6%
• Medications: ASA 19%, beta-blocker 23%, ACE-inhibitor/ARB 24, statin 19%
• eGFR: 90 ml/min/1.73 m²

Interventions

• Eligible patients were binned into a group based upon their first HS troponin-T reading (in ng/L):
  • HS troponin-T <5
    • Those with EKG changes were excluded from this group
  • HS troponin-T 5-14
  • HS troponin-T >14
• Patients were identified by hospital registration
• Related laboratory and EKG data were extracted from hospital records
• Background characteristics (eg, PMH) were extracted from the Swedish National Patient Register
• Medications extracted from the Swedish Prescribed Drug Register

Outcomes

Presented as HS troponin-T <5 ng/L vs. 5-14 ng/L vs. >14 ng/L.

Primary Outcome

MI by 30 days

Includes fatal and nonfatal type 1 MIs, defined as MI from atherosclerotic plaque rupture, fissuring, or erosion resulting in intraluminal thrombus formation and subsequent myocardial ischemia.

0.17% vs. 3.08% vs. 26.2%

Secondary Outcomes

MI

By 180 days: 0.37% vs. 3.65% vs. 28.1%

By 1 year: 0.61% vs. 4.25% vs. 29.2%

All-cause mortality

By 30 days: 0.023% vs. 0.41% vs. 2.56%

By 180 days: 0.17% vs. 2.09% vs. 8.38%

By 1 year: 0.43% vs. 3.43% vs. 13.3%

Additional Analyses

Admission

21% vs. 44% vs. 82%

Most common discharge diagnoses

HS troponin-T <5 ng/L: Chest pain 50%, AF/SVT 6%, angina 5%, abdominal pain 3%

HS troponin-T 5-14 ng/L: Chest pain 34%, angina 8%, AF/SVT 8%, MI 3%

HS troponin-T >14 ng/L: MI 26%, chest pain 13%, HF 9%, AF/SVT 7%

Criticisms

• No report of symptom duration; those with false negatives tended to have an elevated HS troponin T in >3 hours, suggesting that the NPV is lower if performed <3 hours since onset of chest pain
• HS troponin-T may be associated with diurnal variation, which could affect the accuracy of the test
• The observational design means that individuals were discharged without defined criteria; the authors do not have the necessary evidence to conclude that patients with a low HS troponin-T can be safely discharged from the hospital
• Troponin tests were not measured serially in all patients, suggesting that they may have been inappropriately ordered in some
• As there was no active follow-up, the low incidence may have been related to incomplete data capture
• CKD was defined by a single creatinine measurement
• Diabetes was defined by treatment with hypoglycemic agent, missing those with untreated, undetected, or diet-controlled disease; those on hypoglycemic agents for other reasons (eg, PCOS) would have been inappropriately categorized
• Not a direct comparison between groups as those with EKG changes were excluded from the lowest troponin group but included in the other groups

Funding

• Not stated