A double-blind, placebo-controlled phase II study of the efficacy and safety of 2,2-dimethylbutyrate (HQK-1001), an oral fetal globin inducer, in sickle cell disease
Conflict of interest: R. G. Ghalie is an employee of HemaQuest Pharmaceuticals.
Abstract
This placebo-controlled phase II study evaluated the pharmacodynamics, efficacy and safety of 2,2-dimethylbutyrate (HQK-1001), a fetal globin gene-inducing short-chain fatty acid derivative, administered orally at 15 mg/kg twice daily for 48 weeks in 76 subjects with sickle cell disease (SCD). The median age was 26 years (range: 12–55 years) and 37 subjects (49%) were treated previously with hydroxycarbamide. Sixty subjects (79%) had Hb SS and 16 (21%) had S/β0 thalassemia. The study was terminated after a planned interim analysis showed no significant increase in fetal hemoglobin (Hb F) and a trend for more pain crises in the HQK-1001 group. For 54 subjects with Week 24 data, the mean absolute increase in Hb F was 0.9% (95% confidence interval (CI): 0.1–1.6%) with HQK-1001 and 0.2% (95% CI: −0.7–1.1%) with placebo. Absolute increases in Hb F greater than 3% were noted in 9 of 38 subjects (24%) administered HQK-1001 and 1 of 38 subjects (3%) administered placebo. The mean changes in hemoglobin at Week 24 were comparable between the two groups. The mean annualized rate of pain crises was 3.5 with HQK-1001 and 1.7 with placebo. The most common adverse events in the HQK-1001 group, usually graded as mild or moderate, consisted of nausea, headache, vomiting, abdominal pain, and fatigue. Additional studies of HQK-1001 at this dose and schedule are not recommended in SCD. Intermittent HQK-1001 administration, rather than a daily regimen, may be better tolerated and more effective, as shown previously with arginine butyrate, and warrants further evaluation. Am. J. Hematol. 89:709–713, 2014. © 2014 Wiley Periodicals, Inc.
Introduction
Compounds that reactivate the expression of Hb F can be therapeutically beneficial in sickle cell disease (SCD) because Hb F is a potent inhibitor of the polymerization of deoxyhemoglobin S 1. 2,2-Dimethylbutyrate sodium salt (HQK-1001), an orally bioavailable short-chain fatty acid butyrate derivative, was shown to stimulate γ-globin production and erythropoiesis in vitro and in animal models, including transgenic mice and anemic baboons 2. Although the molecular mechanism of action of HQK-1001 is not fully elucidated, similar short-chain fatty acid derivatives were found to reactivate the γ-globin gene by inducing a dissociation of histone deacetylase class 3 (HDAC-3) and its adaptor protein nuclear receptor corepressor (NCoR) from the γ-globin gene promoter, with coincident recruitment of RNA polymerase II to the γ-globin gene promoter 3. HQK-1001 was also found to displace HDAC-2 in chromatin immunoprecipitation assays and to suppress BCL-11A protein levels in treated erythroid progenitors 4. In addition, HQK-1001 enhances erythropoiesis and decreases erythroid apoptosis by phosphorylation and activation of STAT-5, c-cis, and Bcl-xL 5-7. On the basis of these preclinical observations, HQK-1001 was evaluated as potential therapy for patients with SCD and β-thalassemia.
In the first study in 24 subjects with SCD age 14 years and older, HQK-1001 at 10, 20, and 30 mg/kg daily for 12 weeks was well tolerated and resulted in modest increases in Hb F and total hemoglobin at the 20 and 30 mg/kg dose levels 8. A subsequent open-label study in 52 subjects with SCD age 12 years and older evaluated HQK-1001 administered at higher doses (30, 40, and 50 mg/kg/day) and for longer duration (up to 26 weeks). This study demonstrated that gastritis was the dose-limiting toxicity at 40 mg/kg/day and was exacerbated by oral iron use, and concluded that 30 mg/kg/day was the recommended dose for evaluation in future studies 9. The mean absolute increase in absolute Hb F was 2.7% in 31 subjects who had HQK-1001 added to a stable dose of hydroxycarbamide (hydroxyurea) and 2.0% in 21 subjects not treated with hydroxycarbamide and who received HQK-1001 alone, with maximum increases of 10% in both groups. Eight subjects with Hb F increases greater than 6% achieved their highest Hb F values after 12 weeks of therapy suggesting that longer treatment periods were needed for best response. The effect of HQK-1001 on pain crises could not be evaluated because there was no control group. The mean steady state maximum plasma concentrations (Cmax) achieved at 30 mg/kg administered as a single daily dose in three HQK-1001 studies (two in SCD and one in β-thalassemia) were in the range of 114–140 µg/mL (corresponding to 826–1045 µM), higher than the concentrations of 200–600 µM shown to induce Hb F and erythropoiesis in erythrocytes cultured from patients with β-hemoglobinopathies 4.
We conducted the present study, registered under NCT01601340, to evaluate in a placebo-controlled design the efficacy and safety in SCD of HQK-1001 administered at a total daily dose of 30 mg/kg, the maximum tolerated dose identified in prior studies. However, the daily dose was divided in two and administered as 15 mg/kg every 12 hr to reduce peak plasma concentrations and perhaps reduce the frequency of nausea and vomiting, the most common adverse events in prior studies. In addition, the duration of therapy was extended to 48 weeks to evaluate if a longer treatment period further improves Hb F response and to provide sufficient observation time to assess clinical endpoints. We limited the study to patients not on hydroxycarbamide to assess the efficacy of HQK-1001 without the confounding effects of concurrent treatment with another Hb F inducer.
Methods
Eligibility criteria
Patients with SCD between 12 and 60 years of age were eligible if they had hemoglobin SS or S/β0 thalassemia, symptomatic disease defined as at least one acute SCD-related complication or leg ulcers in 12 months prior to enrollment, and no current (i.e., within 3 months prior to enrollment) treatment with hydroxycarbamide. Main exclusion criteria included participation in a regular transfusion program, transfusion in the preceding 3 months unless Hb A had decreased to less than 20%, an acute vaso-occlusive event within 3 weeks prior to randomization, pulmonary hypertension requiring oxygen therapy, symptomatic untreated peptic ulcer or gastroesophageal reflux disease, history of pancreatitis, alanine aminotransferase, or aspartate aminotransferase (AST) greater than three times the upper limit of normal (ULN), serum creatinine, serum amylase, or serum lipase greater than 1.5 times the ULN, or known HIV infection.
Study design
This was a multinational, randomized, double-blind, placebo-controlled study. The protocol, consent forms, and subject diary were approved by central or local Ethics Committees for each participating institution. All subjects provided written consent (and/or assent if less than 18-years-old) before undergoing any study-specific procedure. Central randomization was stratified by the level of Hb F at screening (<5% vs. ≥5%). HQK-1001 (HemaQuest Pharmaceuticals, San Diego, CA) or placebo tablets were administered at the dose of 15 mg/kg twice daily for 48 weeks followed by a 4-week observation period. Study drug was temporary withheld in a subject if the hemoglobin increased to more than 13 g/dL at any time or increased by more than 3 g/dL in any 4-week period, or if the serum amylase or lipase increased to more than three times the ULN. Study drug could be withheld or the dose reduced to manage an adverse event or a clinically significant laboratory abnormality. All subjects received daily oral folic acid.
Study visits occurred every 4 weeks while receiving study drug and 4 weeks after completion of therapy. Efficacy assessments consisted of the evaluation of: (1) pain crises, defined as acute SCD-related pain events of at least 4-hour duration requiring visit to a medical facility and treatment with opioids or parenteral nonsteroidal anti-inflammatory drugs; (2) acute chest syndrome and other SCD-related pain events that did not qualify as pain crises; (3) daily pain and distress scores using a 10-point scale averaged for Days 1–7 after each study visit; (4) daily analgesic use averaged for Days 1–7 after each study visit; and (5) the fatigue component of the Functional Assessment of Chronic Illness Therapy (FACIT) quality of life scale assessed for the 7 days preceding each study visit 10. The English and Arabic versions of the FACIT questionnaire were licensed from FACIT.org.
Clinical safety assessments and safety laboratory tests, which included complete blood count with differential, reticulocytes count, and nucleated red blood cells, full chemistry panel, including amylase and lipase, prothrombin time and partial thromboplastin time, and urinalysis were performed at each study visit. Adverse events and laboratory abnormalities were graded using the NCI Common Terminology Criteria for Adverse Events Version 3.0. Globin fractionation by high-performance liquid chromatography was performed centrally (Sickle Cell Center Laboratory at Georgia Regents University, Augusta, GA) on samples obtained once during the 14-day screening period, on Day 1 prior to starting study drug, and then every 4 weeks. Because red cell transfusions increase the proportion of Hb A in blood, Hb F values at all study visits were corrected using the formula Hb F = 100 × Hb F/(Hb S + Hb F), where Hb S and Hb F are expressed in percent. Baseline values for Hb F and total hemoglobin for pharmacodynamic assessments were the average of all samples obtained prior to starting study drug. Predose and 1–2 hr postdose study drug plasma concentrations were measured every 3 months in all subjects to provide an estimate of compliance with therapy. In addition, and under a separate consent/assent, blood samples were collected at Week 4 from four subjects randomized to HQK-1001 for detailed pharmacokinetic analysis. Plasma concentrations were assayed at the end of the study, after subject treatment was unblinded.
An independent Data Monitoring Committee (DMC) was responsible for study oversight. A planned interim analysis was performed when all ongoing subjects had completed 24 weeks of dosing. The unblinded results were reviewed by the DMC, which could halt or modify the study if there was no evidence of efficacy or increased toxicity in the HQK-1001 group.
Study objectives, sample size calculation, and statistical analyses
The primary objective was to assess the effect of HQK-1001 on Hb F. The secondary objectives were to assess the effect of HQK-1001 on SCD pain crises and complications, daily pain, analgesic use, and quality of life, and the safety and tolerability of HQK-1001 administered at 15 mg/kg twice daily for 48 weeks. A sample size of 56 subjects was needed to show an absolute difference of 2% in the mean change in Hb F between the two groups, with a standard deviation (SD) of 3.0%, a power of 70%, and an α-error of 0.05 in a two-sided comparison. Assuming an early termination rate of 30%, 74 subjects had to be randomized to achieve the required power at Week 48. Standard descriptive statistics were used to summarize pharmacodynamic and efficacy parameters. Frequency distributions were compared by the Kruskal–Wallis test, mean values were compared by the two-sample t-test for differences between groups and the one-sample t-test for differences within-groups, and time-to-events were evaluated by the method of Kaplan and Meier.
Results
Baseline characteristics
Between August 2012 and May 2013, 77 subjects were randomized at 18 study sites; 11 in the United States, three in Lebanon, two in Egypt, and one each in Jamaica and Canada. One subject found to be ineligible after randomization did not receive study drug, was withdrawn from the study, and his data were not included in the analyses. Demographics and baseline characteristics were comparable between the two groups (Table 1). The study population was comprised mostly of young adults, with a preponderance of female (64%) and Hb SS genotype (79%). All subjects classified as White were enrolled at sites in Lebanon and Egypt and these sites enrolled proportionally more subjects with Hb S/β0 thalassemia (25%) compared to the other sites (16%). In the 39 subjects who had never received hydroxycarbamide, the reasons were no availability of compound in 16 (41%), treatment not indicated in 11 (28%), subject refusal in 6 (15%), and other reasons in 6 (15%). In the 37 subjects who had been treated previously with hydroxycarbamide, the reasons for stopping this therapy were subject decision in 13 (35%), poor tolerability in 9 (24%), financial reasons in 8 (22%), lack of efficacy in 6 (16%), and noncompliance in 1 (3%).
| Subject characteristics | HQK-1001 (N = 38) | Placebo (N = 38) | |
|---|---|---|---|
| Age (years) | Mean (range) | 27.8 (12–55) | 25.9 (12–46) |
| Age < 18 years | 5 (13%) | 8 (21%) | |
| Sex | Female | 25 (66%) | 24 (63%) |
| Male | 13 (34%) | 14 (37%) | |
| Race | Black or African-American | 24 (63%) | 22 (58%) |
| White | 14 (37%) | 15 (39%) | |
| American-Indian | 0 (0%) | 1 (3%) | |
| Country | United States | 13 (34%) | 11 (29%) |
| Lebanon | 8 (21%) | 10 (26%) | |
| Jamaica | 9 (24%) | 8 (21%) | |
| Egypt | 6 (16%) | 7 (18%) | |
| Canada | 2 (5%) | 2 (5%) | |
| Genotype | Hb SS | 30 (79%) | 30 (79%) |
| Hb S/β0 thalassemia | 8 (21%) | 8 (21%) | |
| Prior hydroxycarbamide | Yes | 20 (53%) | 17 (45%) |
| No | 18 (47%) | 21 (55%) | |
| No. of pain crises in the 12 months before enrollment | 0–1 | 13 (34%) | 18 (47%) |
| 2 or more | 25 (66%) | 20 (53%) | |
| Randomization stratum | Hb F < 5% | 18 (47%) | 18 (47%) |
| Hb F ≥ 5% | 20 (53%) | 20 (53%) |
| Hematologic parameters | HQK-1001 | Placebo | |
|---|---|---|---|
| Hb F (%) | Mean (range) | 8.3 (1.1–26.7) | 7.5 (0.5–23.4) |
| Hb F (g/dL) | Mean (range) | 0.7 (0.1–2.6) | 0.7 (0.0–2.1) |
| Hemoglobin (g/dL) | Mean (range) | 8.4 (5.6–10.4) | 8.6 (6.6–12.1) |
| Reticulocytes (%) | Mean (range) | 10.2 (1.0–20.3) | 9.80 (2.9–34.3) |
| Platelets (×109/L) | Mean (range) | 392.5 (138–590) | 426.9 (198–815) |
| White blood cells (109/L) | Mean (range) | 10.3 (5.3–21.1) | 12.1 (5.0–19.3) |
In this population of subjects not currently treated with hydroxycarbamide, the mean baseline Hb F was 7.9% (range: 0.5–26.7%), with 19 subjects (25%) having a baseline greater than 10% and three subjects (4%) having a baseline greater than 20%. Baseline Hb F values were comparable in 60 subjects with Hb SS (mean = 7.8%, SD = 5.7%) and 16 subjects with Hb S/β0 thalassemia (mean = 8.2%, SD = 5.9%). The variability of Hb F results for samples obtained at the screening and Day 1 visits was low, with a coefficient of correlation of 0.98.
Subject disposition and compliance with study drug
The study was terminated in November 2013 after the Week 24 interim analysis determined that HQK-1001 did not significantly increase mean Hb F and was associated with a higher incidence of pain crisis and acute chest syndrome. At the time of study closure, 7 subjects (9%) had completed the 52-week study period, 18 (23%) had withdrawn early, and 52 (68%) were still ongoing, with a median follow-up of 36 weeks for all subjects (range: 0–52 weeks). There were more discontinuations due to adverse events and withdrawal of consent in the HQK-1001 group (Table 2). Dose reduction or temporary dosing interruption for drug-related adverse events occurred in 7 of 38 subjects (18%) randomized to HQK-1001, including nausea and/or vomiting in six subjects, and in 2 of 38 subjects (5%) randomized to placebo. Drug compliance, using data from pill count and subject diary, was calculated as the ratio of the number of days study drug was taken to the number of days between the first and last dose. The mean compliance was 86.5% (range: 14–100%) in the HQK-1001 group and 91.9% (range: 41–100%) in the placebo group. A compliance of less than 90% was noted in 14 subjects (37%) in the HQK-1001 group and 10 subjects (26%) in the placebo group. In the HQK-1001 group, mean HQK-1001 plasma concentration was 51 µg/mL (SD: 22.9 µg/mL) in 48 predose samples and 77 µg/mL (SD: 30.2 µg/mL) in 57 postdose samples.
| No. (%) of subjects | HQK-1001 (N = 38) | Placebo (N = 39) |
|---|---|---|
| Completed the study | 3 (8%) | 4 (10%) |
| Ongoing at study closure | 22 (58%) | 30 (77%) |
| Early discontinuations | 13 (34%) | 5 (13%) |
| Reasons for early discontinuations | N = 13 | N = 5 |
| Withdrawal of consent/assent | 7 (54%) | 1 (20%) |
| Adverse eventa | 4 (30%) | 1 (20%) |
| Lost to follow-up | 1 (8%) | 1 (20%) |
| Investigator's decisionb | 1 (8%) | 1 (20%) |
| Subject ineligible | 0 (0%) | 1 (20%) |
- a HQK-1001: One case each of acute chest syndrome, nausea, acute pancreatitis, and cutaneous septic emboli. Placebo: Hodgkin's lymphoma.
- b HQK-1001: Progression of lupus requiring treatment with azathioprine. Placebo: Subject refused contraception.
Efficacy
In 54 subjects with data available at the interim analysis, the mean change in Hb F from baseline to baseline to Week 24 was higher in the HQK-1001 group but the difference between the two groups was not significant (Table 3). The mean absolute increase in Hb F was 0.9% in 24 subjects administered HQK-1001 compared to 0.2% in 30 subjects administered placebo. When all study visits were evaluated, increases in Hb F greater than 3% were observed in nine subjects (24%) in the HQK-1001 group compared to one subject (3%) in the placebo group (P = 0.007). The largest increase in Hb F was 6.4% in the HQK-1001 group and 3.1% in the placebo group. In the HQK-1001 group, the peak Hb F value occurred after Week 24 in 11 of 19 subjects (58%) who had data past Week 24, and 7 of these 11 subjects had Hb F increases greater than 3%. The mean changes from baseline to Week 24 in hemoglobin and reticulocytes were comparable between the HQK-1001 and placebo group, and the mean changes in total and direct bilirubin were significantly higher in the HQK-1001 group (Table 3).
| Parameter | Absolute change from baseline to week 24 | HQK-1001 (N = 24) | Placebo (N = 30) |
|---|---|---|---|
| Hb F (%) | Mean (95% CI) | 0.9 (0.1–1.6)a | 0.2 (−0.7–1.1) |
| Median (range) | 0.8 (−3.1, 6.4) | 0.1 (−6.9, 3.1) | |
| Hb F (g/dL) | Mean (95% CI) | 0.07 (0.01–0.15)a | 0.02 (−0.07–0.10) |
| Median (range) | 0.06 (−0.33, 0.51) | 0.0 (−0.49, 0.17) | |
| Hemoglobin (g/dL) | Mean (95% CI) | 0.01 (−0.27–0.29) | −0.12 (−0.46–0.23) |
| Median (range) | 0.00 (−0.22, 1.10) | −0.05 (−2.65, 1.45) | |
| Reticulocytes (%) | Mean (95% CI) | −0.4 (−1.7–0.9) | 0.1 (−1.7–1.9) |
| Median (range) | −1.3 (−3.9, 9.5) | −0.2 (−7.7, 11.7) | |
| Reticulocytes (×109/L) | Mean (95% CI) | −3.91 (−43.64–35.81) | 1.23 (−45.47–47.94) |
| Median (range) | −30.0 (−100.0, 315.0) | −10.0 (−245.0, 325.0) | |
| Total bilirubin (µMol/L)b | Mean (95% CI) | 11.41 (1.20–21.62)a | 0.75 (−2.85–4.36) |
| Median (range) | 9.5 (−48.3, 64.4) | −1.0 (−14.6, 33.3) | |
| Bilirubin direct (µMol/L)c | Mean (95% CI) | 7.02 (2.29–11.75)a | −0.54 (−1.85–0.77) |
| Median (range) | 3.45 (−10.09–41.05) | −0.86 (−5.98–11.29) |
- CI = confidence interval.
- a P-value < 0.05 for change from baseline within HQK-1001 group.
- b P-value < 0.05 for comparison between HQK-1001 and placebo.
- c P-value < 0.01 for comparison between HQK-1001 and placebo.
There was a trend for more frequent pain crises and acute chest syndromes in subjects randomized to HQK-1001, and the median time to first pain crisis was shorter in the HQK-1001 arm (Table 4). The mean scores for daily pain, daily distress, and interference of pain on daily activities, and the mean fraction of days when analgesics were used were comparable between the two groups. The results of the FACIT quality of life assessments were not computed for the Week 24 interim analysis.
| Parameter | HQK-1001 (N = 38) | Placebo (N = 38) | |
|---|---|---|---|
| Subjects with at least one pain crisis | 22 (58%) | 19 (50%) | |
| Annualized rate of pain crises | Mean (SD) | 3.5 (5.2) | 1.7 (2.2) |
| Median (range) | 1 (0, 20) | 1 (0, 8) | |
| KM time (weeks) to first pain crisis | Median (95% CI) | 15.1 (9.3–50.3) | 30.9 (10.1–NR) |
| No. subjects with acute chest syndrome | 6 (16%) | 2 (5%) |
- KM = Kaplan Meir; SD = standard deviation; CI = confidence interval, NR = not reached.
Safety
Adverse events reported in at least 10% of subjects in the HQK-1001 group are listed in Table 5. Nausea, vomiting, abdominal pain, headache, fatigue, and increased AST were more frequent with HQK-1001. Most drug-related adverse events were graded as mild or moderate. Grade 3 adverse events possibly related to HQK-1001 consisted of headache in two subjects, acute pancreatitis in one subject who also had cholelithiasis, and cutaneous septic emboli in one subject where the causality was unclear. These Grade 3 adverse events resolved after temporary or permanent HQK-1001 discontinuation. One subject randomized to HQK-1001 had worsening systemic lupus erythematosus, was treated with corticosteroids, and then, discontinued from the study prior to starting azathioprine for managing her lupus. Because two cases of acute pancreatitis were reported in the prior HQK-1001 dose escalation study, serum amylase and lipase were measured at each scheduled visit in this study. Grade 2 increase in amylase (between 1.5× and 3× UNL) was reported at a single visit in two subjects randomized to HQK-1001 and one subject randomized to placebo. Grade 3 increase in lipase (greater than 3× UNL) was reported in three subjects randomized to HQK-1001 and none in the placebo group.
| Adverse event | HQK-1001 (N = 38) | Placebo (N = 38) |
|---|---|---|
| Nausea | 14 (37%) | 4 (11%) |
| Headache | 14 (37%) | 11 (29%) |
| Vomiting | 13 (34%) | 5 (14%) |
| Abdominal pain | 8 (21%) | 6 (16%) |
| Fatigue | 8 (21%) | 6 (16%) |
| Pain in extremity | 6 (16%) | 7 (18%) |
| Dizziness | 6 (16%) | 6 (16%) |
| Fever | 6 (16%) | 6 (16%) |
| Aspartate aminotransferase increased | 5 (13%) | 0 (0%) |
Discussion
This placebo-controlled study showed that HQK-1001 administered at 15 mg/kg twice daily did not significantly increase Hb F or improve anemia in SCD. HQK-1001 increased Hb F in most subjects, but the increases exceeded 3% in just 24% of subjects, were below desirable targets, and were not sustained. Our assumption that better compliance with HQK-1001 resulting from improved tolerability due to twice daily dosing would improve Hb F response compared to the prior dose escalation study was not confirmed. In the subset of subjects with Week 24 data and receiving HQK-1001 alone, the mean absolute increase in Hb F was 0.9% in this study comparable to a mean increase of 1.4% in the prior study 9. The mean peak increase in Hb F was 1.9% in 38 subjects administered HQK-1001 in this study, a result comparable to a mean peak increase of 2.0% in 21 subjects receiving HQK-1001 alone in the prior study. Maximum Hb F increases where 6.5% in this study and 10% in the prior study.
The modest Hb F response observed in this trial is unlikely to be due to low HQK-1001 plasma concentration because the mean postdose plasma concentrations of 77 µg/mL (corresponding to 500 µM) was in the range shown to induce Hb F and erythropoiesis in preclinical models. As expected from pharmacokinetic modeling, administration of HQK-1001 at 15 mg/kg every 12 hr instead of 30 mg/kg once daily as used in previous studies has reduced peak plasma concentrations (Cmax) by approximately 50% but maintained similar area-under-the-curve concentrations.
Arginine butyrate, a prototype short-chain fatty acid, is a potent Hb F inducer due to its pan-HDAC inhibitory activity but suppresses erythropoiesis in culture due to cell cycle arrest, a characteristic of pan-HDAC inhibitors 2, 11. Arginine butyrate, administered by a pulsed regimen of 4 days of 12-hour intravenous infusion followed by 10–24 days without therapy resulted in a mean increase in Hb F of 13.8% in 11 patients with SCD, a threefold increase above baseline 12. In contrast, arginine butyrate administered by continuous infusion over 5–6 days each week resulted in only small and transient increases in Hb F 12, 13. A similar rise followed by a decline in Hb F response was seen in baboons after high-dose and prolonged administration of acetate, a product of butyrate catabolism, suggesting a possible association with erythroid suppression from continuous exposure to butyrates 14. Based on these observations, it is possible that a pulsed or intermittent, rather than daily, HQK-1001 regimen may achieve higher and more sustained Hb F responses than that were observed in this study. Iron was found to be important for Hb F and erythroid responses in other trials of butyrates 12. However, oral iron supplementation was not used in this study due to concerns about gastritis, and this may have contributed to the lower than expected erythroid response.
The most common adverse events of HQK-1001 affected the upper gastrointestinal tract, consisted of nausea, vomiting, and abdominal pain, and required dose reduction in several subjects and treatment discontinuation in one. The frequency of these adverse events was comparable to that observed in the prior dose escalation when HQK-1001 was administered at higher dose, but the severity was less. Although no subject developed acute gastritis, the dose-limiting toxicity in the prior study, the frequency of nausea and vomiting indicates that HQK-1001 at 15 mg/kg twice a day is not suitable for long-term continuous use in SCD. It is possible that an intermittent regimen may be better tolerated.
Asymptomatic rise in AST in five subjects in the HQK-1001 group may have been due to direct hepatotoxicity. Another possible explanation is a choleretic effect of HQK-1001, has been reported with other short-chain fatty acids 15, and as suggested by the significant increase in total bilirubin and direct bilirubin in the HQK-1001 group. One subject in this study and two subjects in the prior dose escalation study developed acute pancreatitis, two at 30 mg/kg/day and one at 50 mg/kg/day. Furthermore, three subjects in this study had reversible Grade 3 increases in lipase. The mechanism by which HQK-1001 causes pancreatitis in unclear, but two subjects in the 30 mg/kg dose level also had cholelithiasis, which may have contributed to this complication. The frequency of pain crisis and acute chest syndrome was increased in subjects randomized to HQK-1001 despite the small increase of Hb F in this group. A possible explanation is dehydration due to nausea and vomiting caused by HQK-1001. Other possible explanations are unknown off-target effects of HQK-1001.
In conclusion, HQK-1001 administered orally at 15 mg/kg twice daily is not recommended as a regimen for further investigations in SCD. An intermittent, pulsed dosing, and use in addition to hydroxyurea treatment, may be more effective and better tolerated, and would require further evaluation.
Acknowledgments
The following investigators enrolled subjects in the study: Miguel Abboud, MD (American University of Beirut, Beirut, Lebanon); Inas Asfour, MD (Ain Shams University Hospital, Cairo, Egypt); Kenneth Ataga, MD (University of North Carolina, Chapel Hill, NC); Rita Bellevue, MD (New York Methodist Hospital, New York City, NY); Amal El Beshlawy, MD (Pediatric Hospital of Cairo University, Cairo, Egypt); Adlette Inati, MD (Rafik Hariri University Hospital, Beirut, Lebanon); Johnson Haynes Jr., MD (University of South Alabama, Mobile, AL); Abdullah Kutlar, MD (Georgia Regents University, Augusta, GA); Deepa Manwani, MD (The Children's Hospital at Montefiore Medical Center, New York City, NY); Emily Meier, MD (Children's National Hospital, Washington, DC); Lynne Neumayr, MD (Children's Hospital and Research Center, Oakland, CA); Onyinye Onyekwere, MD (Howard University, Washington, DC); Marvin Reid, PhD, MBBS (University of the West Indies, Kingston, Jamaica); Bruce Sharon, MD (University of Illinois at Chicago Medical Center, Chicago, IL); Hedy Smith, MD, PhD (Tufts University Medical Center, Boston, MA); Wally Smith, MD (Virginia Commonwealth University Medical Center, Richmond, VA); Ali Taher, MD, PhD (Chronic Care Center, Beirut, Lebanon); and Richard Ward, MBBS (University Health Network Toronto General Hospital, Toronto, Canada).
The Data Monitoring Committee consisted of Theodore Colton, ScD (Chair), George Atweh, MD, Adam Lerner, MD, and Paul Swerdlow, MD.
