Advances in the Management of Sickle Cell Disease: New Concepts and Future Horizons

Abstract

Sickle cell disease is a chronic and life-limiting disorder. Approximately 100,000 Americans are affected with sickle cell disease with most being African Americans. Newborn screening for sickle cell is available in the United States, leading to early detection and management of the disease beginning in infancy. According to the 2014 National Heart, Lung, and Blood Institute sickle cell disease guidelines, supportive care has been primary management of sickle cell disease, with hydroxyurea being the only FDA-approved, disease-modifying pharmacotherapy available and allogeneic hematopoietic stem cell transplant the only cure. Since 2017, three new disease-modifying therapies have been approved by the FDA: L-glutamine, crizanlizumab, and voxelotor. This review will discuss pertinent trials, dosing, interactions, side effects, access, cost, and their role in sickle cell management.

Introduction

Sickle cell disease (SCD) is an inherited, multisystem disorder characterized by distortion, stiffness, and adhesion of red blood cells (RBCs). In 1958, SCD was further characterized as arising from the substitution of valine for glutamic acid at the sixth amino acid position of the hemoglobin beta-globin chain.¹ Newborn screening for SCD became universally available in the United States, Puerto Rico, and the US Virgin Islands in 2006.² One in 13 African American newborns will have sickle cell trait. Approximately 100,000 Americans are affected with SCD, with 1 in 365 African Americans and 1 in 16,300 Latin Americans having the disease.³ The American states with largest population of sickle cell patients include New York, Florida, Texas, and Georgia with over 5000 patients living within each of these states.⁴

The most prevalent sickle cell genotypes include homozygous hemoglobin SS (HbSS) and heterozygous forms composed of hemoglobin Sβ⁰-thalassemia (HbSβ⁰-thalassemia), hemoglobin Sβ⁺-thalassemia (HbSβ⁺-thalassemia), and hemoglobin SC (HbSC) disease. Sickle cell anemia (SCA) refers to the most severe form of the disease, which includes HbSS and HbSβ⁰-thalassemia.⁵ Hemoglobin concentrations for HbSS and HbSβ⁰-thalassemia are typically 6 to 9 g/dL, while those for HbSC and HbSβ⁺-thalassemia are 9 to 14 g/dL. Complications can occur throughout a sickle cell patient's lifespan from infancy through adulthood. The more common acute complications include vaso-occlusive pain crisis, infection, and acute chest syndrome.⁶ Infection preventive measures for HbSS patients include antibacterial prophylaxis with penicillin. The medication starts as soon as possible after the infant is identified by the newborn screen and continues until the age of 5 years and completion of pediatric pneumococcal vaccine series. Patients may continue receiving penicillin prophylaxis for life if they have had a splenectomy or invasive pneumococcal infection. Eye examinations, urinalysis for proteinuria, and transcranial Doppler for stroke risk are recommended throughout these patients' lives to screen for other acute complications, allowing providers to act in the early stages of these complications.⁵

Sickle cell disease has complex physiologic changes including vaso-occlusion, anemia, hemolysis, inflammation, hypercoagulability, increased oxidative stress, and defective arginine metabolism. When deoxygenation occurs, sickle cells undergo polymerization and aggregation, producing sickling of the RBCs. This rigid shape causes the cells to be trapped in microcirculation, causing downstream tissue ischemic damage or death.¹ Because sickle cells have abnormalities of the Gardos channel, they are prone to dehydration. Hemolysis causes free hemoglobin to be released into the plasma, which acts as a scavenger of nitric oxide (NO) and causes formation of reactive oxygen species. Arginase-1 activity is lower in sickle cells, thus NO cannot be readily made.¹ Finally, abnormal adhesive properties can lead to activation of adhesion receptors, which results in abnormal interactions between RBCs, platelets, endothelium, and extracellular matrix proteins. This drives endothelial cell expression of procoagulant proteins. The selectins (E-selectin and P-selectin), which mediate adhesion, are upregulated in SCD.¹

The only cure for SCD is a hematopoietic cell transplant. This treatment is underused owing to lack of ideal donors (i.e., matched related or matched unrelated), as ethnicities are under-represented in donor registries. Using less desirable donors (i.e., mismatched related or mismatched unrelated donors) increases risk for graft versus host disease, leading to downstream effects of continued immunosuppression and risk for infections. Another possible cure is gene therapy, which is currently being investigated. Disease-modifying treatments recommended in the most recent guidelines from 2014 include hydroxyurea and chronic transfusion therapy. Both are widely available but remain underused. Hydroxyurea has been used to decrease vaso-occlusive pain crises, acute chest syndrome, and need for blood transfusions for HbSS and HbSβ⁰-thalassemia patients. Treatment with hydroxyurea should be offered to infants 9 months and older, children, and adolescents with SCA. Adults with history of acute chest syndrome and pain or anemia interfering with daily activities or quality of life should be treated with hydroxyurea.⁵ Hydroxyurea was previously difficult to administer to young children because it needs to be extemporaneously compounded, but with the advent of dissolvable formulation of hydroxyurea (Siklos, Medunik USA Inc, Rosemont, PA) this has become less of an obstacle. Chronic simple or exchange transfusion is used for children with elevated transcranial Doppler reading and adults and children with previous clinically overt stroke.⁵ Until 2017 these were the only disease-modifying therapies available for sickle cell patients (see Table 1 for FDA-approved disease-modifying medication for SCD comparison). This review will further explore the new disease-modifying therapies—L-glutamine, crizanlizumab-tmca, and voxelotor—which target oxidative stress, P-selectin inhibition, and polymerization pathologies, respectively, of SCD.

Table 1.

Summary of Disease-Modifying Therapies for Sickle Cell Disease

Drug	Dosing	Dose Modifications	Interactions	Side Effects	AWP: 12-mo Supply for 45-kg Patient	Patient Assistance
Hydroxyurea*	Initial: 20 mg/kg once daily by mouth Maximum: 35 mg/kg/day	Siklos: CrCL < 60 mL/min: Initial dose 10 mg/kg/day; titrate to response	May interfere with lactic acid, urea, or uric acid assays resulting in falsely elevated results	Myelosuppression, infection, macrocytosis	Hydrea (500 mg): $1180.80 Hydroxyurea (500 mg): $1058.40 Siklos (1000 mg): $22,680	Siklos: Yes
L-Glutamine†	<30 kg: 5 g (1 packet) twice daily by mouth 30–65 kg: 10 g (2 packets) twice daily by mouth >65 kg: 15 g (3 packets) twice daily by mouth	None known	No known significant drug or test interactions	Nausea, constipation, cough, headache, abdominal pain, extremity pain, back pain	$33,235.20	Yes
Crizanlizumab‡	Initial: 5 mg/kg IV every 2 wk for 2 doses then 5 mg/kg IV every 4 wk	None known	Interferes with automated platelet count, causing platelet clumping	Headache, pain, nausea, arthralgia, pyrexia, pruritus, vomiting, diarrhea, urinary tract infection, upper respiratory tract infection, chest pain	$110,315.40	Yes

AWP, average wholesale package price; CrCL, creatinine clearance

* Droxia (Bristol-Myers Squibb Company, Princeton, NJ); Hydrea (Bristol-Myers Squibb Company, Princeton, NJ); Siklos (Medunik USA, Inc, Bryn Mawr, PA).

^† Endari (Emmaus Medical, Inc, Torrance, CA).

^‡ Adakveo (Novartis Pharmaceutical Corp, East Hanover, NJ).

L-Glutamine

L-Glutamine (Endari, Emmaus Medical Inc, Torrance, CA), an amino-acid oral powder, gained FDA approval in 2017 for reducing acute complications in patients aged 5 years and older with SCD (Table 1).⁷ It was the first drug approved for SCD since hydroxyurea following several preliminary studies on the role of L-glutamine in reducing oxidative stress in pediatric and adult sickle cell patients.^8–10

L-Glutamine's exact mechanism in reducing pain crises is unclear but is related to its role in oxidative stress reduction. Oxidative stress potentiates the formation of reactive oxygen species (ROS) and hemolysis.^11–13 To balance the ROS in the RBCs, antioxidant pathways use reducing agents such as NO and nicotinamide adenine dinucleotide (NAD). Hemolysis impairs NO bio-availability, further leading to altered redox balance and increased inflammation.¹¹ The redox ratio of reduced NAD to total NAD is lower in sickled RBCs than healthy RBCs.^8,14 Erythrocyte production is also accelerated to compensate for the damage and impairment of sickled RBCs observed in this abnormal oxidative process. In patients with SCD, the escalated RBC production and greater need for NAD cause a higher demand for glutamine, a conditionally essential amino acid precursor of NAD.¹⁴ Oral supplementation of glutamine has been shown to increase NAD synthesis, which promotes the redox reaction that combats oxidative stress.⁸

Preliminary studies explored the use of glutamine in SCD. Niihara et al⁸ initially observed 7 patients older than 18 years with SCA and concluded 30 g per day of oral L-glutamine supplementation significantly increased NAD redox potential after 4 weeks. A subsequent study⁹ reviewed the mean adhesion of sickled RBCs to endothelial cells in 9 adult patients with SCA. Endothelial adhesion was significantly improved in the treatment group (n = 5) compared with the control group (n = 4) following at least 8 weeks of oral L-glutamine therapy. A phase II, randomized, double-blind, placebo-controlled, parallel, multicenter study was conducted in 2014.¹⁰ Randomly assigned in a 1:1 ratio, 62 patients (ages 9–58 years, 6 pediatric patients), with most diagnosed with SCA, received 0.3 g/kg of oral L-glutamine powder or placebo twice a day for 48 weeks. Primary and secondary efficacy results observed a significant decrease in frequency of pain crises, hospitalizations, and emergency visits at week 24 when compared with controls. At 48 weeks, results still showed improvement but were not statistically significant. Safety differences were not observed between groups, though there was a large withdrawal rate from both groups.

A 12-month, randomized, placebo-controlled, double-blind, parallel-group phase III trial was conducted in patients with a diagnosis of SCA to examine the primary outcome of L-glutamine in decreasing the number of pain crises, secondary to HbSS or HbSβ⁰-thalassemia¹⁵ (inclusion criteria per Table 2). A pain crisis was defined as “pain leading to treatment with a parenterally administered narcotic or ketorolac in an emergency department (ED) or during hospitalization.” Patients were randomly assigned in a 2:1 ratio to receive 0.3 g/kg/dose of L-glutamine or a placebo oral powder. Of the 230 patients enrolled, 156 participants completed the 53-week trial including an L-glutamine taper and observational period after the first 48 weeks. Mean age of participants was 22 years (range, 5–58 years) with 51.3% being children or adolescents. Approximately two-thirds of patients in each group were stable while receiving concomitant hydroxyurea at baseline, in addition to receiving either L-glutamine (n = 97) or placebo (n = 59). At baseline 15.1% and 20.5% of patients in the L-glutamine group and control group, respectively, experienced over 5 sickle cell pain crises in the year before the trial. Primary results showed a statistically significant 25% difference (p = 0.005) in the number of pain crises in the L-glutamine group (3.2 ± 2.24) vs placebo group (3.9 ± 2.54) at 48 weeks. One of the secondary endpoints resulted in a decreased average number of sickle cell pain-related hospitalizations (p = 0.005). Additional analysis found fewer occurrences of acute chest syndrome (p = 0.003) in the L-glutamine group. A reduced average number of cumulative days of hospitalization (p = 0.02), median time to first pain crisis (p = 0.02), and reduced median time to second pain crisis (p = 0.03) were also noted in the L-glutamine group compared with the placebo group. No significant differences in hemoglobin levels, hematocrit levels, or reticulocyte count were observed. Occurrence of adverse events between the 2 groups was similar; however, compared with the placebo group, there was greater than 5% incidence of nausea, extremity pain, and back pain in the treatment group. A larger percentage of patients in the treatment group withdrew from the trial (36.2%) than in the control group (24.4%), but the reasons for withdrawal were similar between both groups. Overall, L-glutamine was efficacious for patients more vulnerable to SCD complications.

Table 2.

Summary of New Disease-Modifying Therapies Trial Outcomes^15,19,23

Drug	Inclusion Criteria	Primary Outcome	Significant Secondary Outcomes
L-Glutamine	≥5 yr HbSS or HbSβ0-thalassemia ≥2 pain crises in past yr	Number of pain crises* 3.2 ± 2.24 (L-glutamine) vs 3.9 ± 2.54 (placebo) p value = 0.005	Hospitalizations for sickle cell pain* L-Glutamine 2.3 ± 1.99 vs placebo 3 ± 2.33; p = 0.005 Cumulative days in the hospital* L-Glutamine 12.1 ± 16.6 vs placebo 18.1 ± 27.4; p = 0.02 Days to first pain crisis† L-Glutamine 84 (62–109) vs placebo 54 (31–73); p = 0.02 Days to second pain crisis† L-Glutamine 212 (153–250) vs placebo 133 (115–179); p = 0.03 Acute chest syndrome - n (%) L-Glutamine 13 (8.6) vs placebo 18 (23.1); p = 0.003
Crizanlizumab	16–65 yr HbSS, HbSC, HbSβ0-thalassemia, HbSβ+-thalassemia, or other genotypes 2–10 sickle-related pain crises in past 12 mo	ITT: Median crisis rate per year (IQR) ‡ High dose 1.63 (0–3.97) vs low dose 2.01 (1–3.98) vs placebo 2.98 (1.25–5.87) High dose vs placebo p = 0.01 Low dose vs placebo p = 0.18 PP: Median crisis rate per yr (IQR) ‡ High dose 1.04 (0–3.42) vs low dose 2 (1–3.02) vs placebo 2.18 (1.96–4.96) High dose vs placebo p = 0.02 Low dose vs placebo p = 0.13	Time to first crisis, mo‡High dose 4.07 (1.31–NR) vs placebo 1.38 (0.39–4.9); p = 0.001Time to second crisis, mo‡High dose 10.32 (4.47–NR) vs placebo 5.09 (2.96–11.01); p = 0.02
Voxelotor	12–65 yr HbSS, HbSC, HbSβ0-thalassemia, HbSβ+-thalassemia, or other genotypes Hb level between 5.5–10.5 g/dL during screening 1–10 vaso-occlusive crises in past 12 mo	ITT: Hb response at wk 24 - n (%) Voxelotor 1500 mg 46 (51) vs Voxelotor 900 mg 30 (33) vs placebo 6 (7) Voxelotor 1500 mg vs placebo p < 0.001 PP: Hb response at week 24 - n (%) Voxelotor 1500 mg 44 (59) vs 30 (38) vs placebo 7 (9)	Absolute change in Hb, g/dL† Voxelotor 1500 mg 1.1 (0.9–1.4) vs placebo −0.1 (−0.3 to 0.2); p < 0.001 Mean % change indirect bilirubin† Voxelotor 1500 mg −29.1 (−35.9 to −22.2) vs placebo −3.2 (−10.1 to 3.8); p < 0.001 Mean change in percentage of reticulocytes† Voxelotor 1500 mg −19.9 (−29 to −10.9) vs placebo 4.5 (−4.5 to 13.6); p < 0.001 Annualized incidence rate of vaso-occlusive crisis – No. of crises per person-yr (95% CI)† Voxelotor 1500 mg 2.77 (2.15–3.57) vs voxelotor 900 mg 2.76 (2.15–3.53) vs placebo 3.19 (2.5–4.07)

Hb, hemoglobin; HbSβ⁰-thalassemia, hemoglobin Sβ⁰-thalassemia; HbSβ⁺-thalassemia, hemoglobin Sβ⁺-thalassemia; HbSC, hemoglobin SC; HbSS, hemoglobin SS; ITT, intention to treat; n, number; NR, not reportable; PP, per protocol

* Mean ± SD.

^† xMedian(95%).

^‡ Median(IQR).

L-Glutamine is dispensed in 5-g packets of oral powder. Dosing is 5 g, 10 g, or 15 g by mouth twice daily for patients <30 kg, 30 to 65 kg, or over 65 kg, respectively.⁷ The powder can be mixed with 8 oz of a cold or room temperature beverage or mixed with 4 to 6 oz of food that has a consistency similar to apple-sauce or yogurt. Full dissolution of the medication may not be achieved but can it still be administered. After mixing, medication should be promptly consumed. L-Glutamine will reach its peak effect in 30 minutes post-administration.⁷

L-Glutamine is predominantly excreted through metabolism, reused in processes like protein synthesis. It is also reabsorbed in the renal tubules; therefore, renal and hepatic dosing adjustments are not necessary.⁷ No efficacy data are available for patients with renal or liver disease. Patients who were pregnant were excluded from the phase III study, and women of child-bearing age were instructed to use contraception.¹⁵ Data are also not available on possible teratogenicity from L-glutamine.

Most common side effects reported in the clinical trials were nausea, constipation, headache, cough, abdominal pain, extremity pain, and back pain.^10,15 Several adverse effects leading to patient discontinuation of the study were hypersplenism, abdominal pain, dyspepsia, burning sensation in the feet, and hot flashes.¹⁵ Long-term effects of L-glutamine are presently unknown.

The administration of L-glutamine and the mild safety profile favor the use in pediatric populations. L-Glutamine currently does not have a guideline-recommended place in therapy; however, this therapy could be considered for those who cannot tolerate hydroxyurea adverse effects, or for those who seek additional clinical benefit in wanting to reduce the occurrence of pain crises. The average wholesale package price (AWP) for 60, 5-g packets of Endari is $1384.80, though a co-pay assistance program does exist allowing patient to pay the first $10 of their co-pay.^16,17 This medication can only be accessed by patients through a specialty pharmacy.¹⁷ Additional high-quality evidence may be necessary to confirm long-term efficacy of L-glutamine in reducing SCD pain crises, and post-marketing data are needed to further assess its safety profile.

Crizanlizumab-tmca

Crizanlizumab-tmca (Adakveo, Novartis Pharmaceuticals Corporation, East Hanover, NJ) is a humanized IgG2 kappa monoclonal antibody that binds to P-selectin and inhibits interactions with its ligands. It binds to P-selectin on the surface of activated endothelium and platelets and inhibits interactions between endothelial cells, platelets, RBCs, and leukocytes. It is FDA approved for patients with sickle cell who are 16 years and older to reduce the frequency of vaso-occlusive crisis (Table 1).¹⁸

The approval of crizanlizumab-tmca was based on phase II, double-blind, randomized, placebo-controlled trial (i.e., SUSTAIN trial).¹⁹ Patients were enrolled between August 2013 and January 2015 (see Table 2 for inclusion criteria). Hydroxyurea and/or erythropoietin could be used, but participants had to receive it for at least 6 months prior to study inclusion with a stable dose for at least 3 months. Dose adjustments, except for safety reasons, could not be made during the study period and participants could not start hydroxyurea therapy during the trial. Exclusion criteria included participants receiving chronic transfusions or preplanned exchange transfusion during the study period, chronic anticoagulation therapy other than aspirin, and stroke within the past 2 years.¹⁹

A total of 198 eligible participants were divided into 3 intention-to-treat (ITT) arms including high-dose crizanlizumab 5 mg/kg (n = 67), low-dose crizanlizumab 2.5 mg/kg (n = 66), and placebo (n = 65). Per protocol arms were reduced to n = 40 for high-dose crizanlizumab 5 mg/kg; n = 44 for low-dose crizanlizumab 2.5 mg/kg; and n = 41 for placebo. Baseline characteristics of age, sickle cell genotype, concomitant hydroxyurea, 2 to 4 pain crises, and 5 to 10 pain crises were evenly distributed amongst groups. As the primary outcome, the ITT high-dose crizanlizumab was associated with a significant reduction in median crisis rate per year of (IQR) of 1.63 (0–3.97), with placebo at 2.98 (1.25–5.87) (p = 0.01), resulting in a 45.3% difference from placebo. The low dose also was associated with a 32.6% difference from placebo, though not statistically significant (p = 0.18).¹⁹ The number of patients with crisis rate of zero at the end of the trial was approximately double for the high-dose crizanlizumab group vs placebo (24 vs 11). For the per protocol groups, high-dose crizanlizumab was associated with a median crisis rate per year (IQR) of 1.04 (0–3.42) with placebo at 2.18 (1.96–4.96) (p = 0.02). This was a difference of 52.3%, while low-dose crizanlizumab in comparison to placebo had 8.3% difference in median crisis rate per year. The number of patients in the high-dose crizanlizumab with a crisis rate of zero at the end of the trial for the per protocol group was triple that of placebo (15 vs 5).¹⁹

Evaluation of secondary outcomes showed crizanlizumab did not significantly reduce median rate of days hospitalized per year but there was a 41.8% difference for high dose vs placebo. Additionally, high-dose crizanlizumab did show a statistically significant reduction in time to first and second crisis (Table 2).¹⁹ Low-dose crizanlizumab was not statistically significant for any of the secondary outcomes. The number of patients with at least 1 serious or any adverse event was even amongst each arm. Pyrexia and influenza were the most frequent serious adverse events and occurred more in the treatment groups. Adverse events that occurred in more than 10% of patients receiving crizanlizumab were headache, pain, nausea, arthralgia, pyrexia, pruritus, vomiting, diarrhea, urinary tract infection, upper respiratory tract infection, and chest pain.¹⁹

A subgroup analysis stratified the primary outcome by use of hydroxyurea or not in participants. The patients receiving hydroxyurea and high-dose crizanlizumab had 32.1% decrease in median rate of crises per year than with placebo, while those not receiving hydroxyurea had a 50% decrease from placebo.¹⁹ A letter to the editor was published that explored whether the expected synergistic effect was reduced at higher concentrations of crizanlizumab or whether patients in the hydroxyurea group had more severe disease.²⁰ In a subsequent response, the subgroup analysis was further divided into 2 to 4 crises and 5 to 10 crises with and without hydroxyurea for high-dose crizanlizumab or placebo. High-dose crizanlizumab showed a 45.7% decrease in median crises rate per year in comparison to placebo for participants using hydroxyurea with 2 to 4 crises prior to enrollment, compared with 40.5% decrease in participants with the same number of pre-enrollment crises but without hydroxyurea.²⁰ The bigger difference was seen in those with 5 to 10 crises, as hydroxyurea users had 47.4% decrease in comparison to 85.7% for those not receiving hydroxyurea. The author explained the trial was not designed to detect a difference in this subgroup analysis and the number of participants in the 5- to 10-crises group without hydroxyurea was small, which may have led to the large difference rate.²⁰

Crizanlizumab was FDA approved on November 15, 2019. Dosing is two 5 mg/kg doses intravenously (IV) at week zero, followed by 5 mg/kg IV every 4 weeks thereafter. If a dose is missed but is within 2 weeks of the original dose, patients may continue on original schedule. If the dose is more than 2 weeks delayed, then the next dose is administered 4 weeks from when the most recent dose was given.¹⁸ Crizanlizumab is administered over a 30-minute period with a 0.2-μm inline filter with at least a 25-mL flush of dextrose 5% water or normal saline. A test interference with automated platelet counts has been observed with crizanlizumab that causes platelet clumping, particularly occurring if EDTA tubes are used. It is recommended to collect samples in a citrate-containing tube or to run blood sample within 4 hours of collection.¹⁸ The AWP is $282.86 per mL of crizanlizumab, which is supplied as a 100 mg/10 mL vial. Owing to the expense of this medication, there are a few access programs available through Novartis. Adakveo Access program assists in providing up to 4 treatments in the event of prior authorization denial and if the patient meets the FDA-approved indication regardless of insurance type.²¹ Other patient assistance includes universal co-pay card, which is for commercial insurances only, and will pay the remaining co-pay up to $15,000 per calendar year. Finally, Adakveo Connect provides insurance coverage guidance, drug and disease state education, and case management assistance.²¹

Voxelotor

Voxelotor (Oxbryta, Global Blood Therapeutics Inc [GBT], San Francisco, CA) is a hemoglobin S (HbS) polymerization inhibitor that is FDA approved for sickle cell patients 12 years and older (Table 1).²² It preferentially binds to RBCs and HbS, which increases the affinity of hemoglobin for oxygen. Voxelotor demonstrates a dose-dependent inhibition of HbS polymerization. Non-clinical studies propose that voxelotor may inhibit RBCs from sickling and deforming, and reduces whole blood viscosity.²²

The approval of voxelotor was based on the phase III HOPE trial, an international, multicenter, randomized, placebo-controlled, double-blind, parallel-group trial.²³ This trial accrued patients between January 2017 to May 2018 (see Table 2 for inclusion criteria). If participants were taking hydroxyurea, the dose must have been stable for the past 3 months. Patients who were receiving regular RBC transfusions, had had a transfusion in the past 60 days, or had been hospitalized in the previous 14 days before informed consent was obtained were excluded.²³

Eligible participants (n = 274) were identified and divided into 3 ITT arms: voxelotor 1500 mg (n = 90), voxelotor 900 mg (n = 92), and placebo (n = 92). Per protocol arms were further reduced to n = 74 for voxelotor 1500 mg; n = 79 for voxelotor 900 mg; and n = 76 for placebo. Baseline characteristics of age, sickle cell genotype, baseline hemoglobin, concomitant hydroxyurea, 1 pain crisis, and 2 to 10 pain crises were evenly distributed amongst groups. Most patients had HbSS or HbSβ⁰-thalassemia and two-thirds were receiving concomitant hydroxyurea. Median duration of follow-up for each group was 37 to 42 weeks. The primary endpoint was the percentage of participants who had a hemoglobin response, defined as 1-g/dL increase from baseline at 24 weeks. In the ITT groups, voxelotor 1500 mg had a significant hemoglobin response in comparison to placebo (51% vs 7%; p < 0.001). The difference was observed regardless of age, number of vaso-occlusive crises at baseline, or concurrent hydroxyurea or anemia severity at baseline. Voxelotor 900 mg was associated with a hemoglobin response in comparison to placebo though not statistically significant (33% vs 7%). When the primary endpoint was reviewed for per protocol arms, the hemoglobin response was 59%, 38%, and 9% for voxelotor 1500 mg, voxelotor 900 mg, and placebo, respectively.²³

Secondary outcomes included mean change in hemoglobin level from baseline to week 24, annualized incidence rate of vaso-occlusive crisis, and change in indirect bilirubin, absolute reticulocyte count, percentage of reticulocytes, and lactate dehydrogenase (LDH) levels to explore laboratory markers associated with hemolysis. Intention-to-treat analysis demonstrated a statistically significant difference in adjusted mean change in hemoglobin level for voxelotor 1500 mg vs placebo (p < 0.001).²³ Statistical significance was not observed for the per protocol groups. Voxelotor 1500-mg group had statistically significant change in hemolysis laboratory markers from placebo, including indirect bilirubin and percentage of reticulocytes (Table 2). The incidence of vaso-occlusive crisis was reviewed as annualized incidence rate of vaso-occlusive crisis, which was not statistically different for each voxelotor group. The indicator of vaso-occlusive crisis was also reviewed as participants having 1 or more vaso-occlusive crises during the study period and total number of crises, and neither of these were significantly reduced in comparison to placebo.²³ Most adverse effects recorded were grade 1 and 2, and the most frequently reported were headache and diarrhea at greater than 20%. There was no substantial difference between groups for grade 3, serious adverse events or treatment discontinuation.²³

A phase 2a open-label study reviewing safety and efficacy of voxelotor among participants aged 4 to 17 years recently finished and was presented at the European Hematology Association Congress 2021. Forty-five subjects aged 4 to 11 years were included in the trial. A dispersible tablet was used for weight-based dosing of 600 mg for 10 to 19.9 kg, 900 mg for 20 to 39.9 kg, and 1500 mg for 40 kg and above. The median age of subjects was 7 years, 84.4% used hydroxyurea concurrently, and 95.6% had HbSS genotype. Endpoints analyzed included change in hemoglobin level from baseline to week 24, percentage change in indirect bilirubin from baseline to week 24, percentage of reticulocytes, LDH level, percentage hemoglobin occupancy, and plasma and blood Cmax, Cmin, AUC, and t½. Forty-seven percent of subjects achieved a hemoglobin response defined as >1-g/dL increase in hemoglobin. Increases were detected as early as week 2. Mean percentage change from baseline to week 24 of hemolysis markers was 38.6% indirect hemoglobin, 2.6% LDH, and 3.3% percentage of reticulocytes. Pharmacokinetic parameter analysis was similar between patients aged 4 to 11 years and 12 to 17 years. Treatment-related adverse events occurred in 48.9% subjects. Most were grade 1 and 2, and only 2 subjects discontinued therapy owing to treatment-related adverse events. The most commonly reported were diarrhea, vomiting, and rash.²⁴

Voxelotor received accelerated FDA approval on November 25, 2019, based on increase in hemoglobin for patients 12 years and older. Continued approval for the indication of treatment of SCD may be contingent on verification and description of clinical benefit in future confirmatory trials.²² Voxelotor is supplied as 500-mg tablets with a recommended dose of 1500 mg by mouth daily given with or without food. Voxelotor undergoes hepatic metabolism via phase I and phase II metabolism, although it uses primarily CYP3A4 for metabolism with minor usage of CYP2C19, CYP2B6, and CYP2C9. Dose reduction is recommended for severe hepatic impairment (Child-Pugh C) and avoid use of strong or moderate CYP3A4 inducers, strong CYP3A4 inhibitors, or fluconazole. If avoidance is not possible, then a dose adjustment is necessary. The tablet must be swallowed whole, thus it should not be cut, crushed, or chewed. Of note, voxelotor may interfere with high-performance liquid chromatography measurement of Hb subtypes (HbA, HbS, and HbF).²² The AWP for each 500-mg tablet is $138.89, and voxelotor is distributed through specialty pharmacies. Global Blood Therapeutics, Inc offers GBT Source Solutions, which includes case coordinators to help with access, nurse support, and onsite patient navigators to assist with education and adherence. Patients with commercial insurance may qualify for a co-pay program to cover out-of-pocket expenses related to treatment up to a maximum of $15,000 per calendar year. Uninsured patients may qualify for patient assistance program through GBT.²⁵

Conclusion

Sickle cell disease is a chronic life-limiting disease. Disease-modifying therapies had been limited to hydroxyurea until the past 3 years when L-glutamine, crizanlizumab, and voxelotor were approved for varying pediatric age groups through adulthood. L-Glutamine and crizanlizumab demonstrated statistically significant reduction in pain crises in comparison to placebo, while voxelotor showed a significant increase in hemoglobin but did not display a difference in reduction of pain crises. Each has its own compliance issues. L-Glutamine and voxelotor require oral compliance with administration of an oral powder mixed with food or water twice daily or 3 tablets daily, respectively. Crizanlizumab is only available IV, requiring an infusion center for medication administration. Pharmacists can play a role in assuring access through obtaining insurance authorization and cost reduction strategies through patient and co-pay assistance as well as providing education to increase adherence to these medications. Each medication can be used concurrently with hydroxyurea, but there are no studies using any of the new disease-modifying therapies in combination. National Heart, Lung, and Blood Institute guidelines have not been updated since these new disease-modifying therapies have been approved, thus providers must determine the medication for their patients and when to initiate treatment. Future studies with the new agents include crizanlizumab use in younger pediatric patients, its use in priapism, and the efficacy of a larger dose (7.5 mg/kg). Additional studies include use of voxelotor for stroke risk reduction in children 2 to 15 years of age and safety of using higher voxelotor dosing (i.e., 3000 mg) in SCD.

ABBREVIATIONS

AWP

average wholesale package price

EDTA

ethylenediaminetetraacetic acid

FDA

US Food and Drug Administration

GBT

Global Blood Therapeutics, Inc

HbS

hemoglobin S

HbSβ⁰-thalassemia

hemoglobin Sβ⁰-thalassemia

HbSβ⁺-thalassemia

hemoglobin Sβ⁺-thalassemia

HbSC

hemoglobin SC

HbSS

hemoglobin SS

ITT

intention to treat

IV

intravenously

LDH

lactate dehydrogenase

NAD

nicotinamide adenine dinucleotide

NO

nitric oxide

NS

normal saline

RBCs

red blood cells

ROS

reactive oxygen species

SCA

sickle cell anemia

SCD

sickle cell disease