In clinical routine atrial cardiomyopathy is typically categorized according to persistence of atrial fibrillation (AF) and plays a crucial role in treatment decisions. Supplemental information by biomarker would be beneficial for clinical ascertainment.
Purpose
We investigated the association of type of AF to known and novel cardiovascular biomarkers.
Methods
In 263 participants of the AFAB registry (Maastricht, the Netherlands) scheduled for AF ablation known and novel cardiovascular biomarkers (FGF23, BMP10, Ang2, IGFBP7, CA125, NT-proBNP, TNT_hs, sFlt_1, ESM1_7F89A5, DKK3) were analyzed prior ablation. With respect of the influence of ongoing rhythm on biomarker levels, we compared paroxysmal and persistent AF in a one-way ANOVA. In a logistic regression model association of biomarker level was calculated and adjusted for ongoing rhythm and known factors of developing AF (Sex, age, body mass index, heart failure, and hypertension). For prediction of type of AF, a receiver-operating analysis was performed.
Results
We found significant differences in biomarker level of paroxysmal compared to persistent AF depending on ongoing rhythm (table 1) for BMP10, Ang2, IGFBP7, FABP3, NT-proBNP, TNT-hs, and ESM1_7F8_9A5. In a logistic regression models adjusted for ongoing rhythm and factors of AF development Ang2 (OR 1.270, 95%CI 1.022-1.578, P=0.031) and NT-proBNP (OR 1.001, 95%CI 1.001-1.001, P=0.005) indicate a high probability of persistent AF (figure 1). The model to predict type of AF (including ongoing rhythm and known factors of AF development) had an area under the curve (AUC) of 0.69 (sensitivity 67%, specificity 59%), by considerung Ang2 and NT-proBNP in the model, the prediction improved to AUC 0.74 (sensitivity 74%, specificity 64%).
Conclusions
Based on our data we found significant different biomarker levels for paroxysmal and persistent AF. Biomarker levels vary depending on ongoing rhythm, this should be acknowledged. Ang2 and NT-proBNP showed significant association of type of AF in a model adjusted for factors of AF development. Adding Ang2 and NT-proBNP significantly improved prediction of persistent AF in our cohort.
Table 1 Biomarker level according to ongoing rhythm and type of AF. SR – Sinus rhythm
|
Paroxysmal |
Persistent |
One-way ANOVA |
|||
|
|
SR |
AF |
SR |
AF |
P |
|
FGF23 |
127±105 |
165±153 |
145±254 |
153±72 |
0.423 |
|
BMP10 |
1.8±0.4 |
2.1±0.4 |
1.8±0.3 |
2.0±0.4 |
<0.001 |
|
Ang2 |
2.0±0.9 |
2.7±1.5 |
2.2±0.8 |
3.5±2.1 |
<0.001 |
|
IGFBP7 |
73±14 |
79±16 |
74±10 |
79±15 |
0.030 |
|
FABP3 |
31±10 |
41±25 |
35±14 |
40±15 |
<0.001 |
|
CA125 |
12.7±7.6 |
12.8±7.3 |
13.3±8.8 |
13.1±6.0 |
0.969 |
|
NT_proBNP |
208±294 |
737±693 |
370±437 |
938±815 |
<0.001 |
|
TNT_hs |
8.5±4.2 |
16.8±25.3 |
18.6±28.0 |
16.0±19.8 |
0.004 |
|
sFlt_1 |
629±1064 |
723±1345 |
575±1240 |
922±1479 |
0.444 |
|
ESM1_7F8_9A5 |
2.2±0.9 |
2.6±1.3 |
2.5±1.5 |
2.6±1.0 |
0.028 |
|
DKK3 |
52±13 |
53±12 |
53±13 |
56±13 |
0.301 |
Figure 1 Odd's ratio for persistent AF, logistic regression model adjusted for sex, age, body mass index, heart failure, and hypertension
Figure 1 Odd's ratio for persistent AF
https://dgk.org/kongress_programme/jt2022/aP480.html