Risk Stratification of Patients With Severe Heart Failure Awaiting Heart Transplantation—Prospective National Registry POLKARD HF

Most methods used in the risk assessment of heart transplant candidates do not include new biomarkers.

Aims

The aim of the study was to examine the value of NTproBNP and hsCRP and their combined use together with HFSS score in the risk assessment of patients with heart failure enlisted for heart transplantation.

Methods

Data of 658 patients enlisted for heart transplantation in all active transplantation centers were stored in a prospective registry. The composite end point-death or urgent transplantation was recorded during the follow-up.

Results

Death or urgent transplantation was recorded in 161 (24%) of pts. 102 (15%) patients died and 59 (9%) were transplanted urgently. Kaplan-Meier curves for risk of death or urgent transplantation where highly significant when pts were stratified by the quartiles of NT proBNP (P < .000001) or quartiles according to the hsCRP level on admission (P < .002). In the multivariate Cox proportional hazard model, the significance was observed for NTproBNP (P < .01) and HFSS (P < .02), and borderline significance for hsCRP (P = .057). When ROC analyses of the area under the curve (AUC) values were considered, AUC area was for HFSS – 0.645, for NTproBNP – 0.653 and for hsCRP – 0.566. When all those variables were included together in the model, the AUC value rose to 0.6943. Based on those results a weighted risk model with all three parameters was proposed.

Conclusion

HFSS, NTproBNP and hsCRP levels are independent stratification variables of survival or need for urgent heart transplantation. Their predictive value is moderately increased when they are analyzed together.