Drug Therapies for the Management of Sickle Cell Disease

Inhibition of sickle hemoglobin polymerization

Therapeutic approaches targeting HbS polymerization, including 1) blocking intermolecular contacts in the sickle fiber, 2) induction of HbF synthesis, 3) reduction of intracellular HbS concentration, 4) increase in oxygen affinity, and 5) reduction of the concentration of 2,3-diphosphoglycerate, have been described²⁷. Table 1 lists ongoing studies of anti-sickling agents in SCD.

Induction of fetal hemoglobin. High levels of HbF ameliorate the severity of SCD^8,28,29. Hydroxyurea strongly induces HbF³⁰, although the mechanisms by which it does so are still unclear^30–33. Hydroxyurea improves erythrocyte deformability, lowers hemolysis, decreases circulating leukocytes and reticulocytes³⁴, reduces adhesion receptor expression^35–38, and is an NO donor^39,40. In two placebo-controlled, phase III trials, hydroxyurea significantly reduced the frequency of vaso-occlusive crises, acute chest syndrome, blood transfusion requirement, and hospitalizations in adults and children with SCD^23,41. Treatment with low, fixed weight-based dosing of hydroxyurea (10 mg/kg daily) also decreased the frequency of SCD-related complications with low toxicity^42,43. Hydroxyurea further ameliorates SCD morbidity by decreasing the risk of stroke and chronic kidney disease^44,45 and may reduce mortality in SCD^46–49. Recent studies to optimize hydroxyurea therapy are exploring dose maximization via pharmacokinetics-based dosing^50–54. Despite its multiple benefits, a substantial number of patients on hydroxyurea may not obtain an adequate clinical response.

Inhibitors of epigenetic enzymes which repress γ-globin genes, including DNA methyltransferase (DNMT1), histone deacetylases, and lysine-specific demethylase (LSD-1), and γ-globin gene co-activator agonists are being evaluated in SCD. Decitabine (5-aza-2’-deoxycytidine), a DNA hypomethylating agent which depletes DNMT1, increased HbF and total hemoglobin levels at non-cytotoxic doses in patients who had no responses to hydroxyurea⁵⁵. However, it has a very short plasma half-life and negligible oral bioavailability due to rapid inactivation by cytidine deaminase (CDA)⁵⁵. The combination of escalating doses of oral decitabine and an oral CDA inhibitor, tetrahydrouridine (THU), was well tolerated, and the highest tested dose increased HbF, doubled F-cells by approximately 80% of total RBCs, increased total hemoglobin, and decreased biomarkers of hemolysis, coagulation activation, and inflammation⁵⁶.

Histone deacetylase (HDAC) inhibitors increase HbF levels^57–63. Although treatment with intermittent doses of sodium butyrate produced sustained increases in HbF, F-cells, and total hemoglobin⁵⁷, the orally available sodium dimethyl butyrate (HQK-1001) did not significantly increase HbF and appeared to result in more pain crisis compared with placebo⁶³. A phase II study of the global HDAC inhibitor vorinostat was terminated early owing to poor accrual. Selective HDAC1/2 inhibition with ACY-957 increased HbF in vitro and represents a promising therapeutic approach with a better safety profile⁶⁴. Based on the success of the LSD-1 inhibitors RN-1 and ORY-3001 in increasing HbF and F-cells in preclinical studies^65–67, INCB059872 was investigated (NCT03132324). However, this phase I study was terminated early for business reasons. Pomalidomide, a third-generation immunomodulatory drug, produced modest increases of HbF, augmented erythropoiesis, and preserved bone marrow function following 8 weeks of treatment in transgenic sickle mice⁶⁸. Pomalidomide appeared to increase HbF and total hemoglobin only at the highest dose or with greater than 6 months of exposure⁶⁹. The safety and efficacy of increasing the expression of γ-globin gene co-activators—FOXO3 agonist⁷⁰ (metformin [NCT02983129]), nuclear factor-like 2 (Nrf2) agonist⁷¹ (dimethyl fumarate), and SIRT1⁷²—to augment HbF levels are being explored.

Allosteric modification of sickle hemoglobin to its high-oxygen affinity R-state. Voxelotor, recently approved under the FDA’s accelerated approval program, is an orally bioavailable small molecule which binds to α-globin chains of hemoglobin, increases hemoglobin oxygen affinity, and stabilizes the oxyhemoglobin state⁷³. Early studies showed a reduction in markers of hemolysis, with a rapid rise in hemoglobin by day 15⁷⁴. This response was not accompanied by worsening tissue hypoxia, based on serum erythropoietin levels and change in oxygen consumption (VO₂ max), even in participants receiving daily doses of 1,000 mg for 28 days and 900 mg for at least 90 days. Furthermore, voxelotor has been shown to reduce blood viscosity in vitro⁷⁵. In a multicenter, phase III study of patients 12 to 65 years of age, randomly assigned to once-daily oral voxelotor doses of 1,500 mg, 900 mg, or placebo, a significantly higher percentage of subjects on 1,500 mg had a hemoglobin response of 1.0 g/dL at 24 weeks compared with placebo⁷⁶. In addition, subjects on the 1,500 mg dose saw significant reductions in indirect bilirubin and reticulocyte counts from baseline. A dose-dependent increase in hemoglobin following treatment with voxelotor was not accompanied by an increase in pain crisis rate. As this trial was not enriched for subjects with frequent pain crises, appropriately designed studies of patients with frequent pain episodes are needed to determine the effect of voxelotor on reducing such crises. Open-label extension studies to assess the long-term effects of voxelotor are ongoing.

5-hydroxymethyl-2-furfural (5HMF, Aes-103) interacts allosterically with HbS, increasing oxygen affinity and decreasing HbS polymerization and RBC sickling during hypoxia. A phase I trial, with oral doses of up to 4,000 mg, showed no significant side effects⁷⁷, but a phase II trial was terminated early because of unblinding of drug groups (NCT01987908). FT-4202, a novel selective activator of RBC pyruvate kinase, decreases intracellular 2,3-diphosphoglycerate levels, with a resulting increase in hemoglobin–oxygen affinity. It demonstrated a favorable safety profile in healthy individuals⁷⁸ and is being investigated in a phase I study in SCD (NCT03815695).

The oxygen delivery therapeutic agents sanguinate (pegylated bovine carboxyhemoglobin) and MP4CO (pegylated human carboxyhemoglobin) are dual transfer agents which release carbon monoxide when delivering oxygen to hypoxic tissues. They stabilize HbS in its R-state and appear to exhibit anti-inflammatory and anti-apoptotic activity by induction of Nrf2 and heme oxygenase-1^79,80. Sanguinate had an acceptable safety profile^81,82. A phase II study evaluating its efficacy during acute pain crisis was recently completed (NCT02411708).

Increase red blood cell hydration. The inverse relationship between HbS concentration and delay time suggests that even small decreases in intracellular HbS might be beneficial⁸³. Senicapoc selectively blocks the calcium-activated, potassium efflux (Gardos) channel and improves anemia and hemolysis in SCD^84,85. Despite improvements in anemia and hemolysis, a phase III trial showed no significant decrease in the rate of pain crises compared to placebo⁸⁴. Similarly, studies of agents that block the potassium-chloride co-transport channel showed no clinical benefits^86–88. SCD101, a botanical drug with an unclear anti-sickling mechanism which might involve dilution of HbS by affecting the RBC membrane, was well tolerated in a phase Ib study, with a decrease in chronic pain and fatigue and improvement in leg ulcers⁸⁹.

Targeting downstream sequelae of sickle hemoglobin polymerization

Antioxidant therapy. Agents that upregulate antioxidant and/or reactive oxygen species scavenging processes have been evaluated in SCD. Glutamine, a conditionally essential amino acid, is a precursor for nicotinamide adenine dinucleotide (NAD) and improves NAD redox potential. In a randomized, multicenter trial of 230 patients with HbSS or HbSβ⁰-thalassemia, L-glutamine significantly reduced the number of pain crises, hospitalizations, cumulative hospital days, and frequency of acute chest syndrome compared with placebo⁹⁰. However, the effect size of L-glutamine in reducing pain crises was relatively small, with only a 25% reduction in the median number of pain crises. L-glutamine was well tolerated, although low-grade nausea, non-cardiac chest pain, fatigue, and musculoskeletal pain occurred more frequently than with placebo. Despite some concerns related to the high dropout rate in the trial (36% L-glutamine arm; 24% placebo arm), L-glutamine was approved by the FDA to reduce the acute complications of SCD in patients 5 years and older²⁴.

In an open-label pilot trial of oral N-acetylcysteine (NAC), treatment with either 1,200 mg or 2,400 mg daily for 6 weeks increased whole blood glutathione levels and decreased erythrocyte outer membrane phosphatidylserine exposure, plasma levels of advanced glycation products, and cell-free hemoglobin in both groups⁹¹. A randomized, placebo-controlled, double-blind trial of NAC at 600 mg twice daily for 6 months did not decrease the rate of SCD-related pain days per patient year, vaso-occlusive crises, hospital admission days, number of admissions, or days with home analgesic use compared with placebo⁹². The safety and efficacy of NAC, administered at a higher dose during pain crisis, is being explored in a phase I/II study (NCT01800526) (Table 2).

Anti-adhesive therapy. Agents targeting adhesion of blood cells to the endothelium have been investigated in SCD. Crizanlizumab, a humanized monoclonal anti-P-selectin antibody, was recently approved for use in patients 16 years and older for the prevention of vaso-occlusive crises²⁵ based on results of a randomized, double-blind, phase II study which evaluated the benefit of 2.5 mg/kg or 5 mg/kg doses versus placebo. Significantly lower median crisis rate, longer median times to first and second crises, and lower median rate of uncomplicated crises per year were observed with high-dose crizanlizumab compared with placebo following a 52-week treatment period⁹³. Multiple other studies of crizanlizumab are ongoing (Table 2).

Purified poloxamer 188, a nonionic block copolymer surfactant with hemorheologic and antithrombotic properties, was previously shown to significantly decrease the duration of pain episodes, especially in children and patients on hydroxyurea⁹⁴. However, a more recent phase III study in children and adults reported no significant effect of purified poloxamer 188 (vepoloxamer or MST-188) on duration of vaso-occlusive crises compared with placebo⁹⁵. Rivipansel sodium (formerly GMI-1070) is a small-molecule pan-selectin inhibitor that binds to E-, P-, and L-selectin⁹⁶. In a randomized, double-blind, adaptive, phase II trial, treatment with rivipansel during pain crisis produced a significant reduction in the mean cumulative intravenous opioid analgesic use compared to placebo⁹⁷. However, despite the promising phase II trial results, the recently completed multicenter phase III RESET trial failed to meet its primary (time to readiness for discharge) and key secondary (time to discharge, cumulative intravenous opioid utilization, and time to discontinuation of intravenous opioids) efficacy endpoints⁹⁸.

Heparins inhibit adhesive interactions via P-selectin^99,100. Tinzaparin, a low-molecular-weight (LMW) heparin, at therapeutic dose reportedly decreased the number of hospital days, the number of days with pain crisis, and the number of days with the most severe pain scores compared with placebo¹⁰¹. Sevuparin, a derivative of LMW heparin, which retains the P-selectin-binding domain of heparin but largely lacks anticoagulant properties, binds to P- and L-selectins, thrombospondin, fibronectin, and von Willebrand factor, inhibits the adhesion of sickle RBCs to stimulated cultured endothelial cells in vitro, and prevents vaso-occlusion with normalization of blood flow in a mouse model of vaso-occlusion¹⁰². A recently completed randomized, double-blind, placebo-controlled trial found that, although safe, sevuparin did not significantly reduce the time to resolution of vaso-occlusive crisis¹⁰³. A phase II feasibility study of therapeutic-dose unfractionated heparin (NCT02098993) was recently terminated for poor enrollment, but a phase III, randomized, placebo-controlled study of tinzaparin in acute chest syndrome is ongoing (NCT02580773).

Intravenous immunoglobulin (IVIG) binds to FcγRIII receptors, inhibits neutrophil adhesion to the endothelium, reduces RBC capture by leukocytes, and reduces Mac-1 activity due to recruitment of Src homology 2-containing tyrosine phosphatase-1¹⁰⁴. IVIG was tolerated at doses of up to 800 mg/kg during acute vaso-occlusive crises and decreased Mac-1 function from baseline¹⁰⁵. A phase II clinical trial of IVIG during acute pain episodes is ongoing (NCT01757418).

Anti-inflammatory agents. Multiple approaches to downregulate inflammatory pathways have been evaluated. Initial clinical responses with the use of steroids for pain crisis and acute chest syndrome were followed by “rebound” pain episodes and re-admissions after treatment discontinuation^106,107. Inhaled corticosteroids did not significantly reduce the morbidity of acute chest syndrome in a retrospective cohort study¹⁰⁸, but inhaled mometasone significantly reduced daily pain diary scores and levels of soluble vascular adhesion molecule-1 over the 16-week treatment period in another single-center, placebo-controlled trial¹⁰⁹.

The activation of iNKT cells is downregulated by the activation of adenosine A2A receptors (A_2AR)¹¹⁰. Regadenoson, a partially selective A_2AR agonist, was reportedly safe but did not decrease length of hospital stay, total opioid use, or pain scores compared to placebo during pain crises^111,112. In addition, there were no significant differences between regadenoson and placebo regarding the number of patients exhibiting a greater than 30% reduction in activated iNKT cells. The anti-iNKT cell monoclonal antibody NKTT120 resulted in rapid, specific, and continuous iNKT cell depletion in a single-ascending-dose study¹¹³, but further studies are required to determine its long-term safety and efficacy.

Statins have benefits independent of their cholesterol-lowering effect. In a pilot study, simvastatin was well tolerated, increased levels of NO metabolites, and decreased C-reactive protein (CRP) and interleukin-6 levels¹¹⁴. Daily treatment with simvastatin for 3 months significantly reduced pain crises, oral analgesic use, and levels of high-sensitivity CRP, soluble E-selectin, soluble ICAM-1, soluble ICAM-3, and VEGF¹¹⁵, providing supporting data for conduct of a placebo-controlled trial. Atorvastatin was well tolerated in a pilot study but did not improve endothelial function or decrease albuminuria¹¹⁶.

Long-chain omega-3 polyunsaturated fatty acids such as eicosapentaenoic acid (EPA [n-3]) and docosahexaenoic acid (DHA [n-6]) provide benefit to individuals with chronic inflammatory disorders. In a randomized, single-center study, daily omega-3 administered in capsule form significantly lowered the occurrence of vaso-occlusive events, severe anemia, blood transfusion requirements, and school absence compared with placebo¹¹⁷. More recently, a randomized, placebo-controlled study of different doses of SC411, a novel DHA ethyl ester formulation with high DHA bioavailability, found it was well tolerated, with significant reductions in d-dimer and E-selectin and an increase in hemoglobin¹¹⁸. SC411 also significantly reduced electronic diary-recorded pain episodes, analgesic use at home, and days absent from school but did not significantly lower pain crises in the pooled active groups compared to placebo.

Preliminary results suggest that the monoclonal anti-IL1β antibody canakinumab (ACZ885) is well tolerated and not associated with major side effects in SCD¹¹⁹.

Anticoagulant and antiplatelet therapies. With abundant evidence that SCD is a hypercoagulable state²¹, combined with data that coagulation and platelet activation may play roles in disease pathophysiology^120–124, multiple studies have evaluated the effects of anticoagulants and antiplatelet agents in SCD. Although treatment with the LMW heparin tinzaparin produced more rapid resolution of pain crisis and shorter duration of hospitalization than placebo¹⁰¹, it is uncertain whether the reported beneficial effects were a result of its anticoagulant or anti-adhesive effects. The results of a pilot study of the direct oral anticoagulant rivaroxaban in SCD are awaited (NCT02072668).

Multiple relatively small studies of aspirin have reported modest benefit at best^125–127, although one study reported an increase in oxygen affinity, hemoglobin level, and RBC lifespan¹²⁶. A pilot study of the glycoprotein IIb/IIIa inhibitor eptifibatide showed that it was safe during acute pain episodes but did not reduce the time to resolution of such episodes¹²⁸. Ticlopidine, a P2Y₁₂ ADP-receptor antagonist, reportedly decreased the number of pain episodes, mean duration of pain episodes, and severity of such episodes¹²⁹. However, a more recent phase III trial of prasugrel, a newer generation P2Y₁₂ receptor blocker, showed no significant reduction of the frequency of pain episodes in children with SCD¹³⁰. Ticagrelor, a reversible P2Y₁₂ receptor blocker, was well tolerated in children¹³¹ and young adults¹³² but had no beneficial effects on pain. A phase III study to determine the efficacy of ticagrelor in reducing vaso-occlusive crises in children is ongoing¹³³ (Table 3).

Nitric oxide and related agents. With the role of hemolysis in NO scavenging and subsequent endothelial dysfunction, NO and related agents may be beneficial in SCD¹³⁴. Inhaled NO did not improve time to resolution of pain crisis, length of hospitalization, opioid usage, or rate of acute chest syndrome compared with placebo¹³⁵. In another study, inhaled NO did not reduce the rate of treatment failure in adult patients with mild to moderate acute chest syndrome¹³⁶. However, L-arginine, an NO precursor, significantly decreased total parenteral opioid use and pain scores at discharge compared to placebo in children, although there was no difference in length of hospital stay¹³⁷. Furthermore, supplementation of adult patients on hydroxyurea with oral arginine significantly increased the level of NO metabolites and reduced the frequency of pain crises versus placebo following 4 months of treatment¹³⁸. Sildenafil, a phosphodiesterase-5 inhibitor which increases NO-mediated effects by inhibiting the degradation of cyclic guanosine monophosphate (cGMP), was associated with more frequent serious adverse events, predominantly hospitalization for pain, compared with placebo¹³⁹. No evidence of a treatment effect of sildenafil was seen on evaluated study outcomes. Studies of soluble guanylate cycle stimulators, riociguat (NCT02633397) and olinciguat/IW-1701 (NCT03285178), and the phosphodiesterase-9 inhibitor IMR-687 (NCT03401112; NCT04053803) are ongoing (Table 3).