The prognostic utility of heart rate and blood pressure response to regadenoson stress

Abstract

Background

Although heart rate response (HRR) to regadenoson stress has been shown to be a strong predictor of outcome, it has not been investigated in a large all-comers cohort. The prognostic utility of systolic blood pressure response (SBPR) has not been investigated in comparison with HRR.

Methods and results

In a retrospective cohort of 10,227 patients undergoing regadenoson stress single-photon emission computed tomography myocardial perfusion imaging (MPI), HRR, and SBPR were calculated as 100×(peak hyperemia value–baseline value)/baseline value. During 35 ± 21 months follow-up, 921 (8.8%) deaths were observed. The median HRR was 35% (interquartile range [IQR], 21% to 51%). The median SBPR was −9% (IQR, −17% to −2%). HRR and SBPR were independently associated with all-cause mortality with adjusted hazard ratio [HR] of .980 per 1% increment in HRR (CI, .977-.984) and .994 per 1% increment in SBPR (CI, .988-.999). Mortality rates increased with decreasing HRR quartile and SBPR tertile. HRR provided incremental prognostic value for all-cause mortality beyond clinical and imaging parameters (area under the curve [AUC] increase, .03; P < .001) and SBPR data (AUC increase, .11; P < 0001). SBPR did not provide significant incremental prognostic value beyond clinical and imaging parameters or HRR data. We derived and validated HRR of <20% as a cut-off that can improve risk stratification beyond clinical and MPI findings.

Conclusion

Among patients undergoing stress MPI, impaired HRR to regadenoson provided independent and incremental prognostic value for all-cause mortality beyond clinical, imaging, and SBPR data. SBPR positively correlates with HRR, but it does not provide incremental prognostic utility. HRR, but not SBPR, should be routinely reported and considered in assessing patients' overall risk. An abnormal HRR threshold of <20% can improve risk stratification.