A novel biomarker, mid-regional pro-adrenomedullin (MR-proADM), has been shown to predict mortality among patients with heart failure better than the established biomarkers for prognosis that currently are available. The MR-proADM test is an indirect measure of adrenomedullin, a hormone that increases with blood vessel constriction and endothelial dysfunction, both of which frequently are present in patients with heart failure and are indicators of poor prognosis.

MR-proADM was compared with brain natriuretic peptide (BNP) and NT-proBNP (a biologic fragment associated with BNP) in the BACH trial (NCT00537628). BNP and NT-proBNP have been shown to be better predictors of mortality than clinical factors, such as age, gender, and renal function. The BACH trial is the largest biomarker study to be done in the setting of suspected heart failure, enrolling 1641 patients at 15 centers around the world. The patients all were evaluated for heart failure in the emergency department due to dyspnea that was unrelated to trauma or obvious myocardial infarction. Patients on dialysis were excluded from this study. Acute heart failure subsequently was diagnosed in 568 of these patients, and 1073 patients had non-acute heart failure.

The primary endpoint of the trial was the comparison of the ability of MR-proADM and BNP to predict 90-day mortality. The researchers also evaluated the predictive value for MR-proADM and NT-proBNP, additive predictive value, and prognostic ability among all patients who were evaluated for dyspnea.

Stefan D. Anker, MD, PhD, Campus Virchow-Klinkum, Charite Medical Center, Berlin, Germany, co-principal investigator of the study, reported that the MR-proADM test had better prognostic accuracy for 90-day mortality than BNP (73.5% vs 60.8%; p<0.001). The novel biomarker also was significantly superior to NT-proBNP in predicting death within 90 days (73.5% vs 63.6%; p<0.001). The prognostic power of MR-proADM was even stronger for 30-day than for 90-day mortality (area under the curve [AUC], 0.739 vs 0.674). The corresponding AUC values were 0.555 versus 0.606 for BNP and 0.641 versus 0.664 for NT-proBNP.

Dr. Anker noted that the prognostic accuracy for 90-day mortality was improved by adding MR-proADM to BNP (chi² statistic 23.9; p<0.0001) or to NT-proBNP (chi² statistic 15.3; p<0.0001). In contrast, adding either BNP or NT-proBNP to MR-proADM did not increase the prognostic value (p=0.906 and p=0.291, respectively).

The MR-proADM test was also evaluated in terms of prognosis among all patients in the study who had visited the emergency room due to dyspnea. The prognostic accuracy of MR-proADM was significantly better than BNP or NT-proBNP (chi² statistic 129.5 for log MR-proADM vs 60.1 for log BNP and 83.7 for log NT-proBNP; p<0.0001). Dr. Anker pointed out that the prognostic ability of MR-proADM was in fact better among patients who did not have acute heart failure than among those who did (interaction p=0.005).

The better prognostic ability of MR-proADM makes it a superior risk stratification tool, which can help lead to better patient management, said Dr. Anker. However, because the 90-day mortality is high for all patients with heart failure, the biomarker actually distinguishes patients who are at very high risk from those who are at high risk. This fact, coupled with the lack of different treatment options for patients who are at very high risk of death after heart failure, makes it unclear whether the biomarker has true clinical utility.