Persistent Biomarker Elevations in A-fib Patients Foretell Poor Outcomes

 Download this article's Factoid (PDF & PPT for Gold Subscribers)

In patients with atrial fibrillation (A-fib), persistent biomarker elevations indicate worse prognosis than transient or no elevations. The findings, from a prespecified substudy of the RE-LY trial published online May 2, 2014, ahead of print in Heart, suggest that repeat biomarker testing may help guide patient management, researchers say.

However, in a telephone interview with TCTMD, Sorin J. Brener, MD, of Weill Cornell Medical College (New York, NY), expressed skepticism about the utility of biomarker testing in A-fib patients.

Higher Rates of Vascular Death, Stroke at 2 Years

The noninferiority RE-LY trial, published in the New England Journal of Medicine in 2009, compared 2 doses of dabigatran (110 or 150 mg twice daily) with warfarin in 18,113 patients at risk of stroke from nonvalvular A-fib. After a median follow-up of 2 years, both dabigatran doses proved noninferior to warfarin with regard to the primary endpoint of stroke or systemic embolism, with the higher dose showing superiority over warfarin.

For the current substudy, Ziad Hijazi, MD, of Uppsala Clinical Research Center (Uppsala, Sweden), and colleagues looked at 2,514 patients who underwent serial measurement of cardiac troponin I and NT-proBNP at randomization and 3 months. Median follow-up was 2 years.

Using a cutoff of 0.1 µg/L for troponin I, elevations were seen in:

  • 48.5% of patients at neither randomization nor 3 months
  • 28.5% at either time point
  • 23.0% at both time points

With a cutoff of 778 ng/L for NT-proBNP, elevations were seen in:

  • 42.0% of patients at neither randomization nor 3 months
  • 18.4% at either time point
  • 39.7% at both time points

Rates of stroke or systemic embolism were highest over 2 years in patients with persistent elevations in troponin I or NT-proBNP (table 1).

Table 1. Cox Proportional Hazards Model: Stroke or Systemic Embolism

Level of Elevation

Annual Event Rate

HR (95% CI)

P Value

Troponin I



1.08 (0.43-2.70)
2.64 (1.23-5.69)





0.51 (0.19-1.35)
1.53 (0.87-2.69)


Vascular death was also more common with persistent elevations of each biomarker (table 2).

Table 2. Cox Proportional Hazards Model: Vascular Death

Level of Elevation

Annual Event Rate

HR (95% CI)

P Value

Troponin I



2.39 (0.94-6.08)
6.10 (2.64-14.09)

< .00001




2.68 (1.35-5.29)
4.15 (2.33-7.38)


 < .00001

Persistent elevations in both troponin I and NT-proBNP imparted even greater risk with annual event rates of 2.45% for stroke or systemic embolism and 5.41% for vascular death (P = .0046 and < .0001, respectively).

Serial measurement of the combined biomarkers provided better prognostic discrimination (C-statistics of 0.644 for stroke or systemic embolism and 0.765 for vascular death) than did the CHADS2 (C-statistics of 0.611 and 0.581) or CHA2DS2-VASc (C-statistics of 0.622 and 0.610) risk scores.

On multivariable logistic regression analysis, demographic variables associated with elevated troponin I were:

  • Male sex
  • Renal impairment (gfr < 50 mL/min)
  • CAD
  • Hypertension

The most influential variables associated with NT-proBNP elevations were:

  • Renal impairment (GFR < 50 mL/min)
  • Permanent A-fib (vs paroxysmal)
  • A-fib rhythm at screening

There was no significant interaction between biomarker level and study drug.

Usefulness of Serial Testing to Guide Treatment Unclear

“The present results are in accordance with previous studies in which serial measurement of cardiac biomarkers has [been] demonstrated to be powerful prognostic markers of mortality and morbidity in several patient populations such as [those with] acute coronary syndromes, stable coronary artery disease, acute and chronic heart failure, and in apparently healthy community-dwelling adults,” Dr. Hijazi and colleagues write, noting that the RE-LY findings extend that concept to a novel population.

“In a clinical setting, some of the factors [associated with biomarker elevation] might be possible to modify by pharmacological and/or interventional treatments,” they say, adding, “The use of serial measurements of cardiac biomarkers may provide novel strategies for risk identification and potential treatment tailoring in the [A-fib] population.”

Dr. Brener, however, countered that elevations are simply “a marker that bad things will happen.” Though there is some evidence backing the use of BNP to guide heart failure therapy, it is controversial, he reported. “Most studies demonstrated that there is no benefit…. It’s like measuring enzymes after PCI. [The elevations themselves are] not something you can affect.”

Importantly, he added, Dr. Hijazi and colleagues did not exclude events occurring in the first 3 months. “You don’t know here how the 3-month measurements are related to an event. If we measure them right after an event, they would be high. That doesn’t mean they predicted the event. Maybe they were because of it,” Dr. Brener proposed.

He also commented that the added prognostic value is “microscopic.”

Note: Coauthor Michael D. Ezekowitz, MD, is a faculty member of the Cardiovascular Research Foundation, which owns and operates TCTMD.


Hijazi Z, Oldgren J, Andersson U, et al. Importance of persistent elevation of cardiac biomarkers in atrial fibrillation: a RE-LY substudy. Heart. 2014;Epub ahead of print.



Related Stories:

  • The RE-LY trial was funded by Boehringer Ingelheim.
  • Dr. Hijazi reports receiving lecture fees and an institutional research grant from Boehringer Ingelheim.
  • Dr. Brener reports no relevant conflicts of interest.