Clinical characteristics, resource use, and diagnostic accuracy of natriuretic peptides in cancer patients presenting with acute dyspnea to the Emergency Department

Abstract Background Previous studies suggest that dyspnea is a common presenting symptom in cancer patients and 6% of patients hospitalized for acute heart failure (AHF) have a history of cancer [1]. However, the clinical characteristics and diagnostic accuracy of natriuretic peptides in patients with active or past cancer presenting to the emergency department (ED) with acute dyspnea is largely unknown. Purpose We aimed to evaluate (1) the characteristics of patients presenting with acute dyspnea, (2) the prevalence of AHF, (3) the time to discharge and hospitalization rate, and (4) the diagnostic accuracy of natriuretic peptides (BNP and NT-proBNP) and the recommended algorithm to diagnose AHF in patients with and without cancer [2]. Methods Patients presenting with acute dyspnea to the ED were prospectively enrolled in a multicenter diagnostic study. Patient characteristics, including cancer status was prospectively assessed. Final diagnosis was centrally adjudicated by two independent cardiologists using all individual patient information including cardiac imaging, natriuretic peptides and follow-up data. Discrimination was assessed using the area under the curve (AUC). Results Among 2153 patients, 474 (22.0%) had cancer. Cancer patients were older, more frequently smokers, presented more often with productive cough and less often with paroxysmal nocturnal dyspnea, and had a higher prevalence of previous pulmonary embolism. Prevalence of AHF was 44.9% vs 51.0% in patients with and without cancer, respectively (P = 0.02), although it was the most frequent final adjudicated diagnosis for both groups. Cancer patients had a higher prevalence of pneumonia (18.1% vs 14.4%, P = 0.046) and lower prevalence of psychogenic hyperventilation (1.9% vs 5.2%, P = 0.002). Cancer-related dyspnea accounted for up to 30.4% of final diagnoses. Length of stay in the ED was similar in the two groups, although cancer patients had a higher hospitalization rate (85.0% vs 79.0%, P = 0.004) and the hospitalization was longer (median 9 vs 7 days, P<0.001). The diagnostic accuracy of NT-proBNP was high in cancer patients, but significantly lower compared to non-cancer patients (AUC 0.90 vs 0.93; P=0.02). The diagnostic accuracy of BNP was very high and comparable between groups (0.94 vs 0.95. P=0.40, Figure 1). In cancer patients, the AHF NT-proBNP algorithm maintained a very high safety for ruling out AHF but had lower efficacy with more patients remaining in the gray zone compared with patients without cancer (28.9% vs 18.9%, respectively; P<0.01, Figure 2). Conclusions Patients with cancer presenting with dyspnea have lower prevalence of AHF. Hospitalization rates are increased. The diagnostic performance of NT-proBNP, but not of BNP, is reduced. The efficacy of the NT-proBNP algorithm is reduced with more patients triaged to the grey zone and so is rule-in performance, possibly leading to overdiagnosis and mistreatment.Diagnostic accuracy of BNP and NT-proBNPNT-proBNP ESC algorithm