Induction and Long-term Follow-up With ABX464 for Moderate-to-severe Ulcerative Colitis: Results of Phase IIa Trial

ABX464 (Abivax, Paris, France) is a novel, first-in-class, orally administered small molecule for treatment of ulcerative colitis (UC). ABX464 interacts with the cap binding complex, allowing specific splicing of anti-inflammatory miR-124 (microRNA-124), leading to a cascade dextran sulfate sodium of modulating proinflammatory cytokines.1Tazi J. et al.Drug Discov Today. 2020; PubMed Google Scholar Additionally, ABX464 demonstrated a dual mechanism of action in HIV—generating miR-124 and splicing viral RNA— warranting further exploration in inflammatory bowel disease.2Vautrin A. et al.Sci Rep. 2019; 9: 792Crossref PubMed Scopus (42) Google Scholar Preclinically, ABX464 attenuated dextran sulfate sodium-induced colitis in mice and produced long-term protection and decreased miR-124 levels after stopping treatment. This phase IIa study evaluated safety and efficacy of ABX464 in in patients with moderate-to-severe UC. This randomized study included an 8-week, placebo-controlled, double-blind induction phase followed by an open-label long-term extension phase. The study enrolled adults aged ≤75 years with moderate-to-severe active UC. Patients who completed the induction phase—regardless of response—and were willing to continue into a long-term extension phase were eligible to continue. This trial was registered at clinicaltrials.com (registration numbers NCT03093259 and NCT03368118). In the induction phase, patients were randomized 2:1 to oral 50 mg ABX464 and placebo once daily for 8 weeks. In the long-term extension phase all patients received ABX464 50 mg once daily. The primary endpoint in the induction phase was safety, assessed as the rate of treatment-emergent adverse events (AEs). Efficacy endpoints were the proportion of patients achieving clinical remission at week 8 vs placebo, change from baseline (CFB) to week 8 in Mayo Clinic score (MCS) and partial MCS (pMCS), rates of endoscopic remission and improvement, CFB to week 8 in (histopathologic) Geboes score, change in fecal calprotectin, and miR-124 expression. For the long-term extension phase, the primary endpoint was long-term safety of ABX464. Additional efficacy endpoints included clinical and endoscopic rates of response and remission (Supplementary Methods). Overall, 32 patients were enrolled in the induction phase (ABX464, 23; placebo, 9). Of the 29 patients who completed the induction phase (ABX464, 20; placebo, 9), 22 continued into the long-term extension. The induction population was 62.5% men with a mean ± standard deviation age of 43.3 ± 16.2 years (Supplementary Table 1); 78.3% of patients receiving ABX464 experienced AEs vs 55.6% receiving placebo. The most common AEs in the ABX464 group were abdominal pain and headache (17.4% each). One patient in the induction phase receiving ABX464 prematurely withdrew because of an elevated transaminase 3 times the upper level of normal. The most common AEs through month 24 were nasopharyngitis and abdominal pain (31.3% each). There were no deaths. At week 8, 35.0% and 70.0% of patients in the ABX464 group achieved clinical remission and clinical response vs 11.1% and 33.3% in the placebo group, respectively. Endoscopic improvement and remission were observed in 50.0% and 10.0% of patients receiving ABX464, respectively, vs 11.1% each for placebo (Table 1). At month 12 of ABX464 treatment, 12 of 16 patients (75.0%) with an assessable endoscopy were in clinical remission: 4 of 12 (33.3%) sustained remission and 8 of 12 (66.7%) acquired clinical remission during the long-term extension phase. For the 22 patients who moved into the long-term extension phase, 55% were in clinical remission, with 18% and 36% sustaining remission and acquiring remission, respectively (Supplementary Figure 1A).Table 1Efficacy Outcomes at Week 8 of the Induction Study and Months 12 and 24 of the Long-term Extension StudyInduction PhaseABX464 (n = 20)Placebo (n = 9)Clinical remission PP7 (35.0)1 (11.1) FASan = 23 for ABX464 group.7 (30.4)1 (11.1)P-valuebFor the PP population..1588Clinical response PP14 (70.0)3 (33.3) FASan = 23 for ABX464 group.14 (60.9)3 (33.3)P-valuebFor the PP population..06Endoscopic remissioncP-value by analysis of covariance.2 (10.0)1 (11.1)P-valuebFor the PP population..9280Endoscopic improvement PP10 (50.0)1 (11.1) FASan = 23 for ABX464 group.10 (43.5)1 (11.1)P-valuebFor the PP population..0341Change from baseline inbFor the PP population. MCSMean (SD)−4.6 (2.8)−2.1 (2.5)P-valuecP-value by analysis of covariance..0742 Partial MCSMean (SD)−3.9 (2.2)−1.8 (2.0)P-valuecP-value by analysis of covariance..0462 Modified MCSMean (SD)−3.4 (2.0)−1.3 (1.9)P-valuecP-value by analysis of covariance..0508 Modified partial MCSMean (SD)−2.7 (1.4)−1.0 (1.4)P-valuecP-value by analysis of covariance..0212 Fecal calprotectin, μg/gMean (SD)−2698.5 (5955.2)−858.8 (1384.2)Median−898.8−557.1P-valuecP-value by analysis of covariance..4830 Fold induction of miR-124 in rectal biopsy7.71.5P-value.004Long-term Extension12 Months24 MonthsPrior ABX464 (n = 11)Prior Placebo (n = 5)All Patients (n = 16)All Patients (n = 16)Clinical remission PP9 (81.8)3 (60.0)12 (75.0)11 (68.8) FASan = 23 for ABX464 group.12 (75.0)3 (60.0)Clinical response PP10 (90.9)5 (100)15 (93.8)15 (93.8) FASan = 23 for ABX464 group.N/A5 (100)Endoscopic remission10 (90.9)3 (60.0)13 (81.3)7 (43.8)Endoscopic improvement PP11 (100)5 (100)16 (100)11 (68.8) FASan = 23 for ABX464 group.16 (100)5 (100)Change from baseline inbFor the PP population. MCSMean (SD)−1.8 (2.6)−4.4 (2.9)−2.6 (2.9)N/A Partial MCSMean (SD)−0.9 (2.0)−2.0 (2.8)−1.2 (2.3)N/A Modified MCSMean (SD)−1.4 (1.8)−3.2 (2.2)−1.9 (2.0)N/A Modified partial MCSMean (SD)−0.9 (2.0)−2.0 (2.8)−1.2 (2.3)N/A Fecal calprotectin, μg/gMean (SD)−223.3 (258.0)−497.5 (600.0)−309.9 (401.9)N/AMedian−104.7−270.9−111.9 Fold induction of miR-124 in rectal biopsyN/AN/A215.0N/AValues are n (%) unless otherwise defined. Clinical remission was defined as MCS ≤ 2, with no individual subscore > 1; endoscopic remission defined as Mayo endoscopic subscore = 0; endoscopic improvement defined as Mayo endoscopic subscore ≤ 1; and clinical response defined as reduction ≥ 3 points in MCS and ≥30% change from baseline score. FAS, full analysis set; MCS, Mayo Clinic score; N/A, not available; PP, per protocol; SD, standard deviation.a n = 23 for ABX464 group.b For the PP population.c P-value by analysis of covariance. Open table in a new tab Values are n (%) unless otherwise defined. Clinical remission was defined as MCS ≤ 2, with no individual subscore > 1; endoscopic remission defined as Mayo endoscopic subscore = 0; endoscopic improvement defined as Mayo endoscopic subscore ≤ 1; and clinical response defined as reduction ≥ 3 points in MCS and ≥30% change from baseline score. FAS, full analysis set; MCS, Mayo Clinic score; N/A, not available; PP, per protocol; SD, standard deviation. At month 24, of the 12 patients in clinical remission at month 12, 66.7% remained in clinical remission. Two patients lost clinical remission but remained in clinical response. Additionally, 1 patient withdrew at month 23 because of worsening UC; 1 withdrew at month 14 because of noncompliance. Per centrally read endoscopies at month 24, 68.8% showed endoscopic improvement, of whom 43.8% showed endoscopic remission. Of the 11 patients achieving endoscopic improvement, all showed histologic improvement. The numerically transformed Geboes score decreased by an average of 7 points, the Robarts Histopathology Index decreased by an average of 15 points, and the Nancy score decreased an average of 2 points. The average drop for all histologic indices was significantly larger in endoscopic responders vs nonresponders: Geboes score −6.7 vs −1.2, Robarts Histopathology Index −13.3 vs −3.8, and Nancy score −1.8 vs −0.3 (all P <.05). At week 8, the MCS and pMCS mean CFB ± standard deviation were −4.6 ± 2.8 vs −2.1 ± 2.5 for the ABX464 group and −3.9 ± 2.2 vs −1.8 ± 2.0 for the placebo group, respectively (P = .0292 and .0209, respectively) (Table 1, Supplementary Figure 1B–D). For the modified pMCS, there was a statistically significant difference between ABX464 vs placebo (P = .0212) (Table 1). At month 12, mean CFB ± standard deviation in the MCS was −2.6 ± 2.9 overall; −1.8 ± 2.6 vs −4.4 ± 2.9 for prior ABX464 vs prior placebo (Supplementary Figure 1E). Median fecal calprotectin decreased from 153.1 μg/g (baseline) to 27.9 μg/g and 31.6 μg/g at week 52 and month 24, respectively. At week 8, miR-124 expression increased 7.7-fold for the ABX464 group and 1.5-fold for the placebo group (P = .004) (Supplementary Figure 1F and G); at month 12, miR-124 expression increased a mean of 215.0-fold (Table 1). The safety findings here are corroborated by previous studies of ABX464 in healthy volunteers and HIV patients.3Scherrer D. et al.J Antimicrob Chemother. 2017; 72: 820-828PubMed Google Scholar Although 1 patient withdrew because of elevated liver enzymes, increased liver enzymes are not uncommon in patients with UC; ∼50% of patients with inflammatory bowel conditions may have elevated liver enzymes at any point.4Cappello M. et al.Clin Med Insights Gastroenterol. 2014; 7: 25-31Crossref PubMed Scopus (28) Google Scholar, 5Mendes F.D. et al.Am J Gastroenterol. 2007; 102: 344-350Crossref PubMed Scopus (109) Google Scholar, 6Riegler G. et al.Scand J Gastroenterol. 1998; 33: 93-98Crossref PubMed Scopus (60) Google Scholar The difference in endoscopic improvement on centrally read endoscopies after 8 weeks of treatment was statistically significant. Additionally, mean CFB in the MCS and pMCS at week 8 was substantially greater in ABX464 vs placebo. We also observed a high maintenance of remission rates. Up to 71% of patients in remission on ABX464 at week 8 sustained clinical remission until week 52. Of the patients who entered the long-term extension phase without clinical remission, 58.3% achieved this endpoint after 1 year. Additionally, patients receiving a total of 60 weeks ABX464 had median fecal calprotectin levels of 20.8 μg/g. This proof-of-concept study showed increased miR-124 expression in patients receiving ABX464 but no correlation with clinical response. This is being further evaluated in the phase IIb study. One limitation was 1-sided statistical testing. As a proof-of-concept study, it did not include a large enough population to be powered for significance. The primary efficacy outcome was not significant, although it trended in a positive direction. A phase IIb study powered for significance is underway. Additionally, we did not evaluate magnitude of effects and emphasize the difference in treatment groups; this will also be examined in the phase IIb studies. In conclusion, induction therapy with ABX464 50 mg once daily appeared to be safe and well tolerated. After 8 weeks of treatment, ABX464 appeared to be more effective than placebo in achieving endoscopic improvement and reduction of the MCS and pMCS. Maintenance therapy with ABX464 sustained remission and brought additional patients into remission. These clinical findings support the further development of ABX464 as a novel oral UC treatment. Severine Vermeire, MD, PhD (Investigation: Equal; Writing – original draft: Equal; Writing – review & editing: Equal). Xavier Hébuterne, MD, PhD (Investigation: Equal; Writing – original draft: Equal; Writing – review & editing: Equal). Herbert Tilg, MD (Investigation: Equal; Writing – original draft: Equal; Writing – review & editing: Equal). Gert De Hertogh, MD, PhD (Formal analysis: Equal; Writing – original draft: Equal; Writing – review & editing: Equal). Paul Gineste, PharmD (Data curation: Equal; Formal analysis: Equal; Validation: Equal; Writing – original draft: Equal; Writing – review & editing: Equal). Jean-Marc Steens, MD (Methodology: Equal; Project administration: Equal; Supervision: Equal; Writing – original draft: Equal; Writing – review & editing: Equal). The long-term extension study included a planned interim analyses in 8 centers at 12 and 24 months from the induction phase. The centers in France and Germany did not receive regulatory or ethical committee approval for the long-term extension phase. Both studies complied with the International Council for Harmonisation guidelines. The study protocol, amendments, and informed consent form were approved by each center's Ethics Committee, and all patients provided informed consent before screening. All authors had access to the study data, reviewed and provided input on analyses, provided input on drafting and revising the manuscript, and approved the final manuscript. Additional inclusion criteria were willingness to comply with concomitant medication requirements and hematologic and biochemical laboratory results within the defined parameters. In France, additional inclusion criteria were failure of or intolerance to ≥1 biologic drug to treat UC. Permitted concomitant medications included a stable dose of oral corticosteroids of prednisone or equivalent (≤20 mg/day), beclomethasone dipropionate (≤5 mg/day), or budesonide Multi Matrix (≤9 mg/day) for ≥2 weeks before screening. Immunosuppressants including azathioprine, 6-mercaptopurine, or methotrexate were required to be stable for 4 weeks before the screening visit. Previous anti–tumor necrosis factor-α therapy or vedolizumab had to be discontinued ≥8 weeks before the screening visit; cyclosporine or tacrolimus had to be discontinued ≥4 weeks before the screening visit. Other concomitant medications were kept at a constant dose during the study. Exclusion criteria were Crohn's disease, indeterminate colitis, abdominal abscess, history or imminent risk of colectomy, severe gastrointestinal complications, and active gastrointestinal infections. In France, patients were excluded if they received antibiotic treatment within 2 weeks before the screening visit. The 50-mg dose of ABX464 was based on previous clinical trials. Randomization was stratified by concomitant treatment with corticosteroids and previous anti–tumor necrosis factor-α antagonist exposure. Patients were evaluated weekly for the first month of treatment and then every 2 weeks during the second month. Flexible sigmoidoscopy with biopsy via central reading was performed at screening and at week 8. MCS and AEs were assessed and recorded at each visit. At the end of the induction phase (day 56), patients were able to enroll into the long-term extension phase; patients who did not continue completed an end-of-study visit at day 63. In the long-term extension patients were followed weekly for the first month, every 2 weeks for the second month, and then monthly through month 24. MCS and AEs were assessed and recorded at each visit. Biopsies, obtained in the sigmoid at 20 cm, were taken in 28 patients—20 receiving ABX464 and 8 receiving placebo—at baseline and day 56 (week 8). The biopsies were analyzed in a central laboratory and graded according to 3 indices: the Geboes score, Robarts Histopathology Index, and Nancy score. The Geboes score was transformed numerically by assigning 1 point for each increase in subgrade (range, 0–22). The Robarts Histopathology Index score ranges from 0 to 33, with higher scores indicating more severe inflammation. The Nancy core ranges from 0 to 4, with 4 representing the most severe inflammation. In previous comparative studies, both the Robarts Histopathology Index and Nancy score correlated well with the Geboes scores. The MCS consists of 4 items (stool frequency, rectal bleeding, flexible sigmoidoscopy, and physician assessment), with scores ranging from 0 to 12 points. The pMCS is a noninvasive assessment incorporating stool frequency, rectal bleeding, and physician assessment with a maximum score of 9 points. The modified MCS and modified pMCS use their respective MCS items but do not include the physician assessment score. Clinical remission was defined as MCS ≤ 2, with no individual subscore > 1; endoscopic remission was defined as a Mayo endoscopic subscore = 0; endoscopic improvement was defined as a Mayo endoscopic subscore ≤ 1; and clinical response was defined as reduction ≥ 3 points in the MCS and ≥30% change from baseline score. Efficacy analysis was conducted on the per-protocol set. Analysis of the primary efficacy endpoint on the full analysis set included all patients who received ≥1 dose of the study drug and had ≥1 baseline value. The safety analysis set included all patients. The 3 patients who had missing endoscopies were considered failures in the intention-to-treat population. The primary safety endpoint was the rate of treatment-emergent AEs and was compared between treatment groups by a likelihood ratio χ2Vautrin A. et al.Sci Rep. 2019; 9: 792Crossref PubMed Scopus (42) Google Scholar test at the 10% 1-sided significance level. AEs were tabulated (counts and percentage) by treatment group. All AEs were listed, and data were tabulated by body system/organ class. AE tabulations include all treatment-emergent AEs, which are further classified by severity, and relationship to treatment. Clinical laboratory results, vital signs, and electrocardiographic parameters were summarized using descriptive statistics (n, mean, standard deviation [SD], median, minimum, and maximum). Descriptive statistics (n, mean, SD, median, minimum, and maximum) were presented by treatment group for all efficacy endpoints for each time point by treatment group. The response rate was compared in patients who received ABX464 or placebo by a likelihood ratio χ2 test. The above sample size was sufficient to detect an increase in treatment-emergent AE rate from 10% to 50% with 86% power by likelihood ratio χ2 test at a 1-sided 10% significance level. For the long-term extension, interim analyses were summarized by using descriptive statistics (n, mean, SD, median, minimum, and maximum). For efficacy data, comparisons of 3 groups (total population, patients who received ABX464 in the induction phase, and patients who received placebo in the induction phase) were made using point estimates (mean, median, and ratio) and the spread of the data (minimum, maximum, and SD). Clinical remission and response data obtained at the end of induction at week 8 were used as baseline data for the maintenance phase of the study. Supplementary Figure 1Efficacy endpoints for the induction phase and open-label extension phase. (A) Percent of patients with relevant clinical outcomes at 12 months in the long-term extension phase. (B) Mean change from baseline in the MCS during the induction phase. (C) The MCS at baseline and day 56. (D) Mean change from baseline in the MCS during the long-term extension phase. (E) The pMCS and baseline and day 56. (F) Median fecal calprotectin level in the induction phase. (G) Median fecal calprotectin level during the long-term extension phase. Circles indicate pMCS; squares indicate MCS.View Large Image Figure ViewerDownload Hi-res image Download (PPT)Supplementary Table 1Patient Demographics and Baseline Clinical CharacteristicsABX464 (n = 23)Placebo (n = 9)Age, y43.0 (16.6)44.1 (16.1)Sex, male, n (%)12 (52.2)8 (88.9)Body mass index, kg/m225.6 (5.5)26.0 (3.8)Disease duration, y7.7 (6.7)6.5 (3.1)Concomitant corticosteroids,aPrednisone equivalent (≤20 mg/day), beclomethasone dipropionate (≤5 mg/day), or budesonide Multi Matrix (≤9 mg/day). n (%)8 (34.8)3 (33.3)MCS8.7 (1.5)7.9 (2.2)Modified MCS6.4 (1.2)5.8 (1.7)pMCS6.2 (1.4)5.6 (1.8)Modified pMCS4.9 (1.1)3.4 (1.4)Values are mean (SD) unless otherwise defined.a Prednisone equivalent (≤20 mg/day), beclomethasone dipropionate (≤5 mg/day), or budesonide Multi Matrix (≤9 mg/day). Open table in a new tab Values are mean (SD) unless otherwise defined.