Page 140 - Biomarkers for risk stratification and guidance in heart failure

Prognostic impact of change in NT-proBNP versus change in eGFR.
outcome.17,18 In contrast in chronic HF, improvement in renal function defined as a decrease in creatinine of >0.3 mg/dl predicted lower mortality (HR 0.8, 95% CI 0.6-1.0).19 As can be seen in figures 2a and b, we found no significant impact on outcome of increase in eGFR of >20%.
Worsening renal function in HF patients can be caused by ominous processes
like forward failure, venous congestion, neurohumoral activation, and release of
vasoactive substances resulting in low renal perfusion.20 On the contrary, WRF
can also be caused by factors that are associated with favorable outcome like
titration of evidence based HF medication like ACE-inhibitors, AT-2 antagonist
and aldosterone receptor blockers.21-25 WRF can also reflect intravascular volume
depletion caused by diuretic treatment of HF.26 As a consequence, it is very likely
that the prognostic implication of WRF may depend on the underlying cause, and
can be quite different. The assumption that there are multiple triggers for WRF
with different pathophysiologic and prognostic backgrounds is strengthened by
the inability to predict patients at risk for WRF. Although WRF is related to various
factors like baseline renal function, hypertension, diabetes, diuretic use, age,
anemia, vascular disease, signs of congestion and many more,7 attempts to create
a predictive model for WRF have failed.27 In our study, patients with a decrease or
increase in eGFR more than 20% showed a wide variety of change in NT-proBNP ((IQR-40.0%-69.1%) and (-20.0%-94.4%)) respectively. This finding strengthens
the assumption that both worsening and improvement in eGFR can be caused by 6 a wide variety of pathophysiological processes, with different prognostic impact.
The ambivalent prognostic power of WRF is further illustrated by Testani et al.26 WRF, defined as a decrease in eGFR more than 20%, was related to worse prognosis, but was also associated with haemoconcentration, a factor clearly associated with lower 180-day mortality (HR 0.31, p=0.01). Metra et al showed that WRF in acute HF patients was not related to one year mortality or urgent heart transplant in patients that appeared euvolemic at discharge by physical examination.14 In patients with persistent signs of congestion, WRF was well related to worse outcome, but the increased risk appeared to be primarily driven by the presence of congestion. These findings are in line with our results, where changes in the probably most important biomarker of cardiac function, i.e. (NT- pro)BNP, had clearly superior prognostic impact as compared to WRF.
Changes in NT-proBNP have been reported to correlate with changes in clinical status, possibly giving insight in the success of HF treatment.28 Moreover, both in
139