Elevation of plasma cell-free hemoglobin in pulmonary arterial hypertension

Abstract

Background: Cell-free hemoglobin (CFH) is a potent nitric oxide scavenger associated with poor outcomes in several diseases. Pulmonary arterial hypertension (PAH) is characterized by reduced nitric oxide availability. We hypothesized that CFH would be elevated in PAH and would associate with hemodynamics and clinical outcomes.

Methods: We measured CFH in 200 consecutively evaluated patients with PAH, 16 unaffected bone morphogenetic receptor protein type 2 (BMPR2) mutation carriers, 19 healthy subjects, and 29 patients with pulmonary venous hypertension (PVH). CFH values were tested for association with hemodynamics, time to hospitalization, and death.

Results: CFH was elevated in patients with PAH and BMPR2 carriers compared with healthy subjects and patients with PVH (P ≤ .01 all comparisons). There were no differences in CFH across PAH subtypes. CFH modestly correlated with mean pulmonary artery pressure (ρ = 0.16, P = .03) and pulmonary vascular resistance (ρ = 0.21, P = .01) and inversely with cardiac index (ρ = -0.18, P = .02) in patients with PAH. CFH was not associated with hemodynamic response to nitric oxide or death. Patients with the highest CFH levels had increased risk of PAH-related hospitalization when adjusted for age, sex, and PAH cause (hazard ratio, 1.69; 95% CI ,1.08-2.66; P = .02).

Conclusions: CFH is elevated in patients with PAH and BMPR2 carriers compared with healthy subjects and patients with PVH. Elevated CFH levels are independently associated with an increased risk of hospitalization. Further study is required to understand the mechanism of CFH elevation and the potential pathologic contribution of CFH in PAH.

Figure 1 –

Cell-free hemoglobin (CFH) by group. CFH is significantly elevated in patients with PAH compared with healthy subjects and patients with PVH. No difference was found between PAH and UMC. UMCs had higher CFH than healthy control subjects and patients with PVH. Across-group comparison was initially performed using the Kruskal-Wallis test (P < .001). Between-group comparisons were performed using the Mann-Whitney U test. Bars denote median (interquartile range). PAH = pulmonary arterial hypertension; PVH = pulmonary venous hypertension; UMC = unaffected mutation carrier.

Figure 2 –

CFH by PAH subtype. No significant difference in CFH was found across different subtypes of PAH. Across-group comparison performed using Kruskal-Wallis test (P = .09). Bars denote median (interquartile range). CTD-PAH = connective tissue disease-associated PAH; HPAH = heritable PAH; IPAH = idiopathic PAH; POPH = portopulmonary hypertension. See Figure 1 legend for expansion of other abbreviations.

Figure 3 –

Correlation of baseline and repeat CFH values in patients with PAH. CFH was measured at the time of first clinical evaluation and repeated at the next clinical visit in patients with the highest and lowest 10 values among the IPAH, HPAH, and CTD-PAH populations. We found a strong correlation between initial and repeat measurements indicting that CFH values are relatively stable over time and do not simply reflect transient physiologic conditions. Line indicates best-fit Spearman correlation. See Figure 1 and 2 legends for expansion of abbreviations.

Figure 4 –

Impact of CFH level on time to first PAH-related hospitalization. A, Kaplan-Meier curves of PAH-related hospitalization for patients above and below the 50th percentile of CFH level. There was no significant difference between patients with CFH above or below the 50th percentile in time to hospitalization. B, Cox proportional hazards analysis. After controlling for age, sex, and cause of PAH, CFH above the 50th percentile was associated with a 69% increase in the risk of hospitalization. HR = hazard ratio. See Figure 1 legend for expansion of other abbreviation.