Inhibiting the Polymerization of Sickle Hemoglobin: It's Complicated

Although hundreds of different hemoglobin variants have now been identified, polymerization seems to be a unique property of HbS.¹  Polymerization has been characterized in precise detail over many years, with much of the seminal work in this area being performed by Dr. H Franklin Bunn and Dr. William A. Eaton — both co-authors of the Blood publication referenced in this article. Investigators have found a long cascade of pathologic processes resulting from polymerization in sickle cell disease (SCD), including red cell dehydration, vaso-occlusion and tissue infarction, hemolysis, inflammation, vascular endothelial dysfunction, platelet activation, hypercoagulability, oxidative stress, and reperfusion injury, with each of these processes contributing to the many different clinical manifestations of SCD. However, all these processes stem from the polymerization of deoxygenated HbS, and effective inhibition of polymerization would negate the need to manage any downstream consequences. Treatment of the manifestations of polymerization is typically only partially effective, as evidenced by many trials in SCD that have failed to show convincing clinical benefit. These include trials of prasugrel, sevuparin, nitric oxide, Poloxomer 188, and rivipansel, among others.^2-5  To date, the only pharmacologic treatment to show significant clinical benefits in almost every setting is hydroxyurea,⁶  whose main mechanism of action is to promote HbF synthesis, which slows the rate of polymerization, as discussed by Dr. Eric R. Henry and colleagues in their article.

Despite the attractive target, developing drugs to directly inhibit HbS polymerization has proven difficult, partly because of the very large quantities of hemoglobin present in the body. Interest in this therapeutic approach had waned as other “downstream” drugs and stem cell therapies were explored, though the development of voxelotor has now revived interest. Voxelotor was developed to bind covalently to the α-globin chain, to stabilize hemoglobin in the oxygenated form — a form that cannot polymerize, but equally does not release oxygen. Clinical studies have shown that voxelotor increases hemoglobin by 1 g/dL or more in 75 percent of patients, with an approximate 20 percent fall in reticulocyte percentage. Perhaps surprisingly, these changes were not accompanied by any detectable clinical benefits after 72 weeks, and in particular, no change in the frequency of episodes of acute pain or evidence of an improved quality of life.⁷  Dr. Henry and colleagues explored this paradox with a series of in vitro experiments and calculations and largely confirmed that, as expected, voxelotor increases the fractional saturation of sickle hemoglobin with oxygen, decreases the tendency of red cells to sickle with hypoxia, but also decreases the fractional oxygen delivered by the same hemoglobin. They go on to compare the theoretical effects of three different approaches to slow down polymerization: increasing oxygen affinity (voxelotor, pyruvate kinase activators), increasing HbF levels (hydroxyurea, decitabine), and decreasing the concentration of erythrocyte hemoglobin (senicapoc, ferroportin inhibitors). These calculations suggest that increasing oxygen affinity of hemoglobin is the least effective way to improve oxygen delivery, which to some extent, fits with the clinical observations that hydroxyurea is a far more effective drug than voxelotor.