To the editor:
We are honored that our work merits the attention of Dr Bunn and colleagues1 but disagree with some of their conclusions. Traditional risk factors for vaso-occlusive pain crisis, such as leukocytosis and high hemoglobin levels, incompletely predict vasculopathic events and mortality in adult patients with sickle cell disease (SCD). We maintain that the evidence supports the following propositions:
(1) Pulmonary hypertension (PH) is common in adults with SCD and is associated with a high risk of death.
The tricuspid regurgitation jet velocity (TRV) is a useful noninvasive screening tool for suspected PH; however, the diagnosis requires right heart catheterization. In epidemiologic studies, Doppler-estimated right ventricular systolic pressure more than 2 standard deviations (SD) above the mean (TRV > 2.5 m/s) is common in adults with SCD (approximate 30% prevalence) and is associated with a 9.2 to 15.9 risk ratio (RR) for early death.2,3–4 Similarly, NT-proBNP, a biomarker released from cardiomyocytes under pressure stress, is elevated in 30% of SCD patients and a value of 160 pg/mL and above identifies a subgroup with a 19.5% absolute increase in risk of death in the National Institutes of Health (NIH) and Multicenter Study of Hydroxyurea (MSH) cohorts.5
Based on these studies, a TRV less than 2.5 m/s or NT-proBNP less than 160 pg/mL are considered normal screening values associated with a low risk of death in SCD.2 Furthermore, we should agree that a TRV of 3 m/s or higher is more than 3 SD above the mean, present in 10% of SCD adults, and is associated with a high RR of death (> 10).6 We recommend that standard of care for these patients include clinical evaluation for PH risk factors and right heart catheterization.
The intermediate group (TRV 2.5-2.9 m/s; > 2 SD above mean) remains a source of controversy. In adults with SCD, this group overall appears to have decreased exercise capacity and increased mortality (RR of 4.4; P < .001). Refinement of risk definition in this group has great potential to advance the field.
A recent French study of PH in SCD appropriately highlights the limitations of the echocardiogram; the diagnosis of PH requires direct measurement of mean pulmonary artery pressure (mPAP) with PH defined as mPAP 25 mmHg or higher at rest. This study also validates our observations: even after excluding adult SCD patients with greater SCD disease severity, 6% of adults had mPAP 25 mmHg or higher, which meets the consensus definition of PH and is 3 SD above the mean. Furthermore, their patients with confirmed PH had more severe hemolytic anemia and, in a 2-year follow-up, deaths occurred only in the PH group.
(2) Hemolytic anemia impairs NO signaling and is a major and attributable risk factor for the development of clinical subphenotypes of SCD.
In humans, impairment in NO signaling, directly measured via venous occlusion strain-gauge plethysmography, correlates with increased levels of plasma hemoglobin or its surrogate LDH.7 This relationship between increasing plasma hemoglobin levels and direct measures of decreasing NO-dependent blood flow or low NO bioavailability has been confirmed in many studies in disparate diseases, such as malaria8 and paroxysmal nocturnal hemoglobulinuria.9 In patients with SCD, the development of vasculopathic complications—such as PH, priapism, and leg ulceration—are associated with markers of hemolysis.10,11 More than 18 cohort studies have consistently associated the severity of hemolytic anemia with these complications and risk of death, including the NIH-PH Screening,2 Duke University,4 University of North Carolina,3 Multi-center Study of Hydroxyurea,5 Pulmonary Hypertension and the Hypoxic Response in Sickle Cell Disease,12 and Cooperative Study of Sickle Cell Disease cohorts,13 a recently published Greek study,14 and our unpublished analysis of the Walk-PHASST screening cohort (July 7, 2010).
All mouse models of hemolysis studied to date develop spontaneous PH and right heart failure, including the Berkeley sickle cell mouse,15,16 the spherocytosis mouse,17 and the allo-immune hemolysis mouse.15 Pathologic evaluations in these models find no chronic thrombosis in the pulmonary vasculature; rather, a functional impairment in NO signaling driven by NO scavenging by plasma hemoglobin. NO depletion, PH, and systemic hypertension have been reported in animal studies by inducing intravascular hemolysis or by infusions of hemoglobin or hemolysate.18
Hemolytic anemia does not occur in isolation. Priapism is more frequent in SCD than in other hemolytic diseases, demonstrating the independent contributions of sickle vaso-occlusion and hemolysis-related pathology, analogous to the independent contributions of smoking and high cholesterol to the development of atherosclerosis.
(3) Failure of NO-based therapies in early clinical trials does not negate this mechanism of disease.
Less than 10% of all clinical trials are successful. The SCD field has seen failures of drugs targeting accepted mechanisms, such as poloxamer 188 to improve red cell rheology, senicapoc to inhibit the Gardos channel and to reduce HbS polymerization, and steroids to reduce inflammation. We cannot give up on clinical trials in SCD PH; rather, we need to improve trial designs and explore the insights into biology we have learned from all investigations.
What is crucial for the vitality of future hemoglobinopathy research is that we unite to critically analyze and learn from each trial, maintain the funding and infrastructure of the SCD research groups, and encourage the next generation of physician-scientists interested in SCD research.
Conflict-of-interest disclosure: The authors declare no competing financial interests.
Correspondence: Mark T. Gladwin, MD, Division of Pulmonary, Allergy and Critical Care Medicine, Vascular Medicine Institute, University of Pittsburgh, NW 628 Montefiore Hospital, 3459 Fifth Ave, Pittsburgh, PA 15213; e-mail: firstname.lastname@example.org.