N ‐terminal pro‐ B ‐type natriuretic peptide, tricuspid jet flow velocity, and death in adults with sickle cell disease

To the Editor: Both elevated N-terminal prohormone brain natriuretic peptide (NT-proBNP) levels (>160 pg/mL) and elevated tricuspid regurgitant jet flow velocity (TRV 2.5 m/sec) have been related to the presence of (echocardiography defined) pulmonary hypertension and the risk of early death in patients with sickle cell disease (SCD) [1–3]. Recent guidelines advise to use NT-proBNP levels to identify patients at risk for mortality when transthoracic Doppler to measure TRV is not available [4]. We studied the relation between elevated NT-proBNP levels (>160 pg/mL), elevated TRV ( 2.5 m/sec), and the risk of mortality in a cohort of 85 adult outpatients with sickle cell anemia (HbSS) or compound heterozygous states such as HbSb-thalassemia (HbSb-thal), HbSb-thalassemia (HbSb-thal), and sickle-hemoglobin C (HbSC) [5]. Patients were excluded if they had a history of chronic obstructive pulmonary disease or poorly controlled asthma, congestive heart failure, painful crisis, or acute chest syndrome in the preceding 4 weeks and/or blood transfusion within 3 months prior to performing study related tests. The mean age at baseline analysis was 34 years (median 30; interquartile range (IQR): 23–47). We followed patients for a median of 82 months (IQR 75–85) during which three patients were lost to follow-up and 12 patients (11 HbSS and 1 HbSb-thal) died. Median age at death was 53 years (IQR 37–60). Baseline trans-thoracic echocardiographic results were available from 81 patients and echocardiography was repeated every 2 years in steady-state conditions. TRV 2.5–2.9 m/sec was considered elevated, a TRV>2.9 m/sec severely elevated, and TRV was considered normal in patients with trace or no tricuspid regurgitation (undetectable values were assigned a value of 1.3 m/sec) [1]. In 25 patients (31%) a TRV 2.5–2.9 m/sec was measured at baseline (22 (39%) HbSS/HbSb-thal and 3 (12%) HbSC/HbSb-thal patients) and a TRV >2.9 m/sec in two HbSS patients. Twenty of 56 patients with a TRV <2.5 m/sec at baseline developed an increased TRV ( 2.5 m/sec) during follow-up (5.3% per year). Having a TRV 2.5 m/sec measured at baseline did not result in a significant increased hazard rate (HR) for mortality for these patients as calculated by Cox regression analysis (HR 1.6 [confidence interval (CI) 0.5–5.2], P 5 0.4). Figure 1A. Introducing age in a multivariate Cox-regression analysis did not change the HR for mortality in patients with TRV 2.5 m/sec (HR 1.1 [CI 0.3–3.7], P 5 0.9). The HR for mortality in a subgroup of HbSS/ HbSb-thal patients with a TRV 2.5 m/sec was 1.1 [CI 0.4–3.4], P 5 0.9. Plasma NT-proBNP levels were available from 77 patients at baseline after quantitation in EDTA anticoagulated plasma employing an ELISA (Roche). In 14 patients a NTproBNP value 160 pg/mL was measured (all HbSS/HbSb-thal patients), of whom 50% also had a TRV level of 2.5 m/sec at baseline. Figure 1C. Using the previously defined cut-off value for NT-proBNP of 160 pg/mL as a risk factor for early death [6], patients with elevated NT-proBNP levels had a HR of death of 10.0 [CI 2.9–34.4], P< 0.001, compared to patients with normal NT-proBNP levels (<160 pg/mL). Figure 1B. In the HbSS/ HbSb-thal group alone this HR was 6.3 [CI 1.8–21.6], P 5 0.003. Given the low number of deaths during follow-up, a multivariate Cox-regression survival analysis was performed for only two variables at the time, always including NT-proBNP levels. None of analyzed variables that may affect NT-proBNP plasma levels (such as age, renal function, TRV, ferritin, or hemoglobin levels) could solely explain the increased HR for death for patients with NT-proBNP plasma levels of 160 pg/mL. Therefore, an elevated NT-proBNP level identifies especially HbSS/HbSb-thal patients at high risk of death independent of an elevated TRV. We hypothesize that NT-proBNP levels are likely to be determined by several prognostic factors in SCD such as diastolic dysfunction, hypoxia, age, rate of hemolysis, inflammation, renal function and iron overload. NT-proBNP might therefore be useful as a simple prognostic biomarker in the clinically asymptomatic state reflecting the severity of the generalized vasculopathy and organ dysfunction in SCD. Importantly, both the lack of an association of TRV to mortality and the relatively low mortality in our cohort, as well as in a recent European SCD cohort, are in contrast with earlier studies in the United States (US) [1–3]. As TRV was significantly related to mortality in SCD in a study of comparable size [2], our relatively small sample size is not likely to be an explanation for this conflicting observation. Also the relative low number Figure 1. Bone marrow biopsies show progression of neoplastic spindle-shaped mast cells and osteosclerosis between 2009 and 2013 with no therapy (A– B), stable disease before and after bosutinib (B–C), and response to cladribine regarding mast cells (C–D). (hematoxylin and eosin (H&E), 103 objective). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.] CORRESPONDENCE